Fine-Grained Sediments: Deep-Water Processes and Facies
Fine-Grained Sediments: Deep-Water Processes and Facies edited by D. A. V. Stow Grant Institute of Geology University of Edinburgh Scotland
and D. J. W. Piper Atlantic Geoscience Centre Geological Survey of Canada Dartmouth, Nova Scotia Canada
1984 Published for The Geological Society by Blackwell Scientific Publications Oxford London Edinburgh Boston Palo Alto Melbourne
Published by Blackwell Scientific Publications Osney Mead, Oxford OX2 0EL 8 John Street, London WC1N 2ES 9 Forrest Road, Edinburgh EH1 2QH 52 Beacon Street, Boston, Massachusetts 02108, USA 706 Cowper Street, Palo Alto, California 94301, USA 99 Barry Street, Carlton, Victoria 3053, Australia
DISTRIBUTORS USA and Canada Blackwell Scientific Publications Inc. PO Box 50009, Palo Alto California 94303 Australia Blackwell Scientific Book Distributors 31 Advantage Road, Highett Victoria 3190
First published 1984 9 1984 The Geological Society. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by The Geological Society for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that a base fee of $02.00 per copy is paid directly to CCC, 21 Congress Street, Salem, MA 01970, USA 0305-8719/84 $02.00 Printed in Great Britain at The Alden Press, Oxford and bound by Butler & Tanner Frome and London
British Library Cataloguing in Publication Data Fine-grained sediments.---(Geological Society special publications, ISSN 0305-8719; v.4) 1. Sediment transport 2. Sedimentation and deposition I. Stow, D.A.V. II. Piper, D.J.W. III. Series 551.3'5 GB850 ISBN 0-632-01075-4
Contents Preface: STOW, D.A.V. & PIPER, D.J.W .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
vii
Introduction STOW, D.A.V. & PIPER, D.J.W. Deep-water fine-grained sediments: history, methodology and terminology . . . . . . . . . . . . . . . . . . . . . . . . . .
Processes GORSLINE, D.S. A review of fine-grained sediment origins, characteristics, transport and deposition . . . . . . . . . . . . . . . . . . . . . . . MCCAVE, I.N. Erosion, transport and deposition of fine-grained marine sediments . EITTREIM, S.L. Methods and observations in the study of deep-sea suspended particulate matter KRANCK, KATE Grain-size characteristics of turbidites . . . . . . . . . . .
17 35 71 83
Terrigenous turbidites and associated facies VAN WEERING,T.C.E. & VAN IPEREN, J. Fine-grained sediments of the Zaire deep-sea fan, southern Atlantic Ocean . . . . . . . . . . . . . . . . . . . . MONACO, A. & MEAR, Y. Sedimentary sequences on the north-west Mediterranean margin during the Late Quaternary: a dynamic interpretation . . . . . . . . . . . . STOW, D.A.V., ALAM, M. & PIPER, D.J.W. Sedimentology of the Halifax Formation, Nova Scotia: Lower Palaeozoic fine-grained turbidites . . . . . . . . . . . . . . KIDD, R.B. & SEARLE, R.C. Sedimentation in the southern Cape Verde Basin: regional observations by long-range sidescan sonar . . . . . . . . . . . . AUFERET, G.A., LE SUAVE, R., KERBRAT, R., SICHLER, B., ROY, S., LAJ, C. & MULLER, CI Sedimentation in the southern Cape Verde Basin: seismic and sediment facies . . . . GOT, H. Sedimentary processes on the west Hellenic Arc margin . . . . . . CHOUGH, S.K. Fine-grained turbidites and associated mass-flow deposits in the Ulleung (Tsushima) Back-arc Basin, East Sea (Sea of Japan) . . . . . . . . . . . .
95 115 127
145 153 169 185
Carbonate turbidites and associated facies HEATH, K.C. & MULLINS, H.T. Open-ocean, off-bank transport of fine-grained carbonate sediment in the Northern Bahamas . . . . . . . . . . . . . . . . . . . FAUGERES, J-C., CREMER, M., GONTHIER, E., NOEL, M. & POUTIERS, J. Late Quaternary calcareous clayey-silty muds in the Obock trough (Gulf of Aden): hemipelagites or fine-grained turbidites? . . . . . . . . . . . . . . . . . . . . . . STOW, D.A.V., WEZEL, F.C., SAVELLI,D., RAINEY, S.C.R. & ANGELL, G. Depositional model for calcilutites: Scaglia Rossa limestones, Umbro-Marchean Apennines . . . . . . .
199
209 223
Contourites STOW, D.A.V. & HOLBROOK,J.A. North Atlantic contourites: an overview . . . . . . . 245 SHOR, A.N., KENT, D.V. & FLOOD, R.D. Contourite or turbidite?: magnetic fabric of fine-grained Quaternary sediments, Nova Scotia continental rise
.
. . . . . . .
257
GONTHIER,E.G., FAUGERES,J-C. & STOW,D.A.V. Contourite facies of the Faro Drift, Gulf of Cadiz . . . . . . . . . . . . . . . . . . . . . . . . . . . HALEMAN, J.D. & JOHNSON, T.C. The sediment texture of contourites in Lake Superior
275 293 V
vi
Contents
Hemipelagites and associated facies of slopes and slope basins HILL, P.R. Facies and sequence analysis of Nova Scotian Slope muds: turbidite vs 'hemipelagic' deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . MCGREGOR,B.A., NELSEN,T.A., STUBBLEEIELD,W.L. & MERRILL,G.F. The role of canyons in late Quaternary deposition of the United States mid-Atlantic continental rise . . . . BALLANCE,P.F., GREGORY,M.R., Gmson, G.W., CHAPRONIERE, G.C.H., KADAR,A.P. & SAMESHIMA,Z. A late Miocene and early Pliocene upper shelf-to-slope sequence of calcareous fine sediment from the Pacific margin of New Zealand . . . . . . . . . . . . PICKERING, K. T. Facies, facies-association and sediment transport/deposition processes in a late Precambrian upper basin-slope/pro-delta, Finnmark, N. Norway . . . . . . . . KRISSEK, L.A. Continental source area contributions to fine-grained sediments on the Oregon and Washington continental slope . . . . . . . . . . . . . . . . . . . THORNTON, S.E. Basin model for hemipelagic sedimentation in a tectonically active continental margin: Santa Barbara Basin, California Continental Borderland . . . . . . . . GORSLINE, D.S., KOLPACK, R.L., KARL, H.A., DRAKE, D.E., FLEISCHER, P., THORNTON, S.E., SCHWALBACH,J.R. & SAVRDA,C.E. Studies of fine-grained sediment transport processes and products in the Californian Continental Borderland . . . . . . . . . . . . . BOURROUILH, R. & GORSLINE, D.S. Fine-grained sediments associated with fan lobes: Santa Paula Creek, California . . . . . . . . . . . . . . . . . . .
311 319
331 343 363 377
395 417
Pelagites and organic-rich sediments ROBERTSON, A.H.F. Origin of varve-type lamination, graded claystones and limestone-shale 'couplets' on the lower Cretaceous of the western North Atlantic . . . . . . . . HAYWARD, A.B. Hemipelagic chalks in a clastic submarine fan sequence: Miocene SW Turkey CREVELLO, P.D., PATTON, J.W., OESLEBY, T.W., SCHLAGER, W. & DROXLER, A. Source rock potential of Bahamian Trough carbonates . . . . . . . . . . . . . . . . ISSACS, C.M. Hemipelagic deposits in a Miocene basin, California: toward a model of lithologic variation and sequence . . . . . . . . . . . . . . . . . . . . . ANASTASAKIS,GEORGEC. & STANLEY, DANIELJEAN. Sapropels and organic-rich variants in the Mediterranean: sequence development and classification . . . . . . . . . . THICKPENNY, A. The sedimentology of the Swedish Alum Shales . . . . . . . . . . ARTHUR, M.A., DEAN, W.E. & STOW, D.A.V. Models for the deposition of Mesozoic-Cenozoic fine-grained organic-carbon-rich sediment in the deep sea . . . . . . . . . .
437 453 469 481 497 511 527
Internal characteristics FAAS,R.W. Plasticity and compaction characteristics of the Quaternary sediments penetrated on the Guatemalan Transect--DSDP Leg 67 . . . . . . . . . . . . . . . . . MOON, C.F. & HURST, C.W. Fabrics of muds and shales: an overview . . . . . . . . WETZEL, A. Bioturbation in deep-sea fine-grained sediments: influence of sediment texture, turbidite frequency and rates of environmental change . . . . . . . . . . . .
563 579 595
Facies models: synthesis STOW, D.A.V. & PIPER, D.J.W. Deep-water fine-grained sediments: facies models
611
Preface This volume has been edited from the proceedings and discussion at an International Workshop on Fine-Grained Sediments held at Dalhousie University, Halifax, Canada, in August 1982, and co-sponsored by the Geological Society of London and the Geological Association of Canada. Additional contributions have been solicited to provide a more comprehensive publication, although we have retained the original narrow remit of the research meeting. This focused on modern and ancient, siliciclastic and biogenic, fine-grained sediments from a range of deep-water environments. We aimed to limit the scope of our discussions so that advances in understanding resulting in a useful synthesis might be achieved, but to leave the scope sufficiently broad that valuable cross-fertilization of ideas could take place. We hope that the result will be helpful to many research sedimentologists as a state-of-theart summary of a large and important class of rocks. The volume contains thirty-eight separate papers, including short regional contributions and longer review papers, organized into nine sections. There is a short introduction to the history, methodology and terminology of finegrained sediment studies. This is followed by a more substantive section dealing with processes of erosion, transport and deposition in deepwater. There are then sections on each of the main facies types including siliciclastic turbidites, carbonate turbidites, contourites, mixed (mainly hemipelagic) facies of slopes and slope basins,
pelagites and organic-carbon-rich sediments. A section on internal characteristics covers plasticity and compaction, mud and shale fabric, and bioturbation. Our final synthesis paper on facies models attempts to relate fine-grained sediment facies to depositional processes and deep-water environments. Each paper has been reviewed by at least two external referees as well as by the editors. In editing, we have not aimed for rigorous uniformity of style or terminology, as this was clearly not possible given the different approaches and persuasions of the seventy-five authors and coauthors who have contributed. Finally, our thanks to the many people who helped to make possible this publication: to all the participants at the meeting and to the contributors who were unable to attend; to Phil Hill, Tony Bowen and Martin Gibling for their help with the organization and running of the Halifax workshop; to those who refereed papers; to our respective institutions for secretarial, drafting and technical support; to the Natural Environment Research Council, UK, the Canadian Natural Sciences and Engineering Research Council, Dalhousie University, the Geological Survey of Canada, and to our sponsoring organizations for support; and to Nigel Palmer and Jane Grisdale of Blackwell Scientific Publications for their patient and diligent work. Dorrik A.V. Stow, David J.W. Piper
August 1983
Deep-water fine-grained sediments; history, methodology and terminology D.A.V. Stow and D.J.W. Piper SUMMARY: To introduce this collection of papers presented at an international research workshop in Halifax, Canada (1982), we highlight briefly three aspects of deep-water fine-grained sediments that are alluded to throughout the volume but never discussed specifically. These are: (a) an historical outline of the research that has made both possible and necessary the workshop and the volume; (b) a review of the methodology currently used in the study of fine-grained sediments; and (c) an assessment of the state of terminology as applied to this class of rocks.
History The development of knowledge concerning finegrained sediments in deep water has been closely related to the history of both sedimentology and oceanography, and their interpretation has evolved along with the science of geology as a whole. Major advances within each of these disciplines have, on the one hand, improved our understanding of fine-grained sediments, while on the other hand have acted as significant obstacles to progress in the years immediately following each breakthrough. It is as Kuhn (1970) suggested in his theory of scientific revolutions: the establishment of a paradigm both advances an area of study and leads to a period of blinkered normal science that serves to elaborate the paradigm but to digress little from it. We can illustrate this with several examples (Fig. 1).
Pelagic sediments The systematic study of deep-sea sediments and the birth of modern oceanography began with the voyage of HMS Challenger (1872-1876) which established the general morphology of the oceans and the types of sediments they contained. The report on Deep-Sea Deposits by Murray & Renard (1891), which documented the calcareous and siliceous pelagic oozes, their microflora and microfauna, the pelagic red clays, and deep marine ferromanganese deposits became the cornerstone of deep-sea sedimentology for over half a century. Although a major advance over previous knowledge, this new paradigm held that only pelagic clays and biogenic oozes were found in the deep sea and that all coarser-grained clastics were restricted to shallow water or continental environments. It also dismissed any suggestion that rocks now exposed on land were comparable with deep-ocean pelagic sediments. This, therefore,
had two main restricting effects on the study of ancient rocks: (a) clastic sediments, other than red clays, were not considered as deep-water; and (b) the interpretation of chalks, cherts and limestones was thrown into confusion. The growing body of evidence that invoked a deep water origin for thick flysch successions (e.g. Lesley 1892; Natland 1933; Bailey 1936; Pettijohn 1943) was held at bay for many years. Similarly, the Alpine geologists who claimed deep-water pelagic successions (e.g. Suess 1875; Fuchs 1877; Steinmann 1905; Heim 1924) and equivalent interpretation in other parts of the world (e.g. Hinde 1890; Molengraaf 1922), had to bide their time before being more generally accepted.
Systematic sedimentology If Murray & Renard were pioneers of modern oceanography, so Sorby was the father of sedimentology as we know it today. As early as 1861, even prior to the Challenger expedition, he had equated the English Chalk with deep-sea coccolith ooze based on the first reports of coccoliths from Atlantic sediments by Huxley (1858). His paper early in the 20th century (Sorby 1908), On
the application of quantitative methods to the study of the structure and history of rocks, was equally far ahead of his time and extremely perceptive about fine-grained sediments. 'Possibly many may think that the deposition and consolidation of fine-grained mud must be a very simple matter, and the results of little interest. However, when carefully s t u d i e d . . , it is soon found to be so complex a question, and the results dependent on so many variable conditions, that one might feel inclined to abandon the inquiry, were it not that so much of the history of our rocks appears to be written in this language.' H.C. Sorby (1908). It is not entirely clear why his plea for systematic sedimentology was not taken up by students of
4
D . A . V . S t o w and D . J . W . Piper /!
80-'
/
Information explosion
t/ / I
70"
968 Glomar Challenger sails (DSDP) #
966 Contourites (Heezen + others) I
6 0 - 1960s Plate tectonics revolution I I I ! I
shales for so long. Perhaps the rigorous scientific approach was difficult to apply to the finergrained rocks, or perhaps the methodology was lacking. There were, however, some notable exceptions: Helm's (1924) classic work on Alpine sediments; Archanquelsky's (1927) little known study of Black Sea sediments which he likened to the argillaceous fill of geosynclines; and Ruedemann's (1935) discussion of the ecology and sedimentology of Palaeozoic black shales in the Appalachians. Turbidite revolution
Walker (1973) has documented the four lines of research that effected the mid-20th century revolution in our thinking about deep-sea sediments. I These were: (a) the recovery of varied sediment i types other than pelagites on a number of Euroi 40I pean and American oceanographic expeditions in I I the first half of the century (e.g. Bdggild 1916; Long period of 'normal' science Andree 1920; Shepard 1932, 1948; Stocks 1933; data accumulation Bramlette & Bradley 1940; Arrhenius 1950); (b) 5 0 - no new paradigms the persuasive evidence presented by Daly (1936 ) and Johnson (1938) suggesting that sedimentladen density currents had excavated submarine canyons as they flowed downslope; (c) a series of flume experiments on both dilute and high-den20" sity flows carried out by Kuenen (1937, 1950); and (d) numerous observations of graded sand beds in probable deep-water successions on land (e.g. Natland 1933; Migliorini, 1946). The classic IO,~ paper Turbidity currents as a cause of graded 1908 Systematic sedimentology bedding (Kuenen & Migliorini 1950) became the I I (Sorby) manifesto of the revolution, and stimulated an intense period of systematic field, laboratory and II 1900oceanographic studies. I I However, attention became focused on the I sandstones and coarser-grained sediments to the I detriment of the finer-grained material. Much of 1891 Pelagic sediments the interbedded mudstone was dismissed as 90( Murray + Renard) 'background', 'ubiquitous' hemipelagic sediment. This has been more true of carbonate turbidites and the interbedded calcilutites which are still largely considered pelagic. The Bouma 80 (1962) standard structural sequence for sandy turbidites has stood for a long time as the sequence for turbidites, although it is clearly ~ 8 7 2 _ 7 6 HMS Challenger unsatisfactory for resedimented muds and silts or expedition for fine-grained biogenics. This neglect of fine70.grained turbidites was largely a consequence of the 1950s concepts of turbidity current dynamics. FIG. 1. Historical sketch showing the main concepts Although evidence for deposition of mud beds that have advanced our understanding of deep-sea from turbidity currents was recognized (Dzufine-grained sediments. The dashed line indicates the lynski &Kinle 1957) it was thought that little clay stepped increase in our level of understanding, as each new concept (or paradigm) leads both to a could settle out except at velocities that would significant advance and to period or relative give only very dilute turbidity currents, so that the standstill. deposit would be very thin (Dzulynski et al. 5 0 " 1950 Turbiditesl (Kuenen + Migliorini) I I I
Deep-water fine grained sediments," history, methodology and terminology 1959). Clay deposition was believed to require velocities of less than 1 cm/sec. The observational evidence for deposition of clay from typical turbidity currents could be explained by deposition as faecal pellets or shale chips, ponding, or repeated small turbidity currents (Dzulynski et al. 1959). In the 1960s, it became clear that clay could be deposited from suspension at velocities at which silt and fine sand could be transported (Einstein & Krone 1962), and increasing investigation of core samples clearly demonstrated that thick turbidite muds were a common deep-sea facies (van Straaten 1967; Hesse 1975).
Modern sedimentology The turbidite revolution also heralded the development of the modern science of sedimentology. Sedimentary structures were recognized as indicators not just of way-up in folded strata, but more importantly of processes (which could be quantified) and hence sedimentation facies and environments (Middleton 1965). Analysis of the sequence of sedimentation facies in a stratigraphic succession, enunciated in the last part of the nineteenth century by Walther (1893), was developed into a standard tool (de Raaf et al. 1965). These two techniques lead to the development of the facies model approach (Walker 1976). At the same time, the principles of carbonate rock classification developed by Folk (1959) provided the foundation for a major expansion of understanding of carbonate rock sedimentology accompanied by major investigations of modern shallow-water carbonate environments of Florida, the Bahamas and the Persian Gulf (Folk 1973).
Plate tectonics The major upheaval in geology that culminated in the early 1960s ideas of sea floor spreading and plate tectonics (Hess 1960; Dietz 1961; Runcorn 1962; Vine & Matthews 1963), had repercussions throughout the discipline. It finally sanctioned the idea that deep oceanic pelagic sediments may be found high up in mountain ranges, and threw new light on our interpretations of the very thick and largely fine-grained fill ofgeosynclines. It also spawned new major scientific investigation of the ocean basins and their sediments. However, even these undoubtedly profound advances had their drawbacks. In particular, scientific attention and effort was directed towards the large-scale picture, the placing of giant pieces in a new jigsaw puzzle, and this necessarily detracted from detailed sedimentology.
5
Contourites Only a decade after the explosion of research on turbidites many sedimentologists were beginning to see flaws in this undoubtedly elegant paradigm. Land geologists repeatedly described examples of 'turbidite' sequences having orthogonal palaeocurrent directions, even from bottom to top of the same bed (e.g. Kelling 1958; Craig & Walton 1962; Ballance 1964; Klein 1966). Marine geologists were documenting many other kinds of evidence that indicated an important role for bottom currents flowing alongslope in deep water (e.g. Heezen 1959; Heezen & Johnson 1963; Hubert 1964). Finally, it was the compilation of evidence by Heezen et al. (1966) on Shaping of the
continental rise by deep geostrophic contour currents that caught the imagination of many more geologists. Stow & Lovell (1979) suggested that this paper marked the beginning of a revolution in sedimentology comparable to the turbidite revolution. Important though this paradigm has been in adding a new dimension to our understanding of deep-sea processes and fine-grained facies, it has nevertheless proved an obstacle to progress in two respects: (a) research focused initially on sandy contourites and largely ignores muddy contourites, although the latter now appear volumetrically more significant; and (b) many authors apparently jumped onto the contourite bandwagon without a clear appreciation of just how to recognize a contourite, modern or ancient, and hence many erroneous interpretations have been published.
Information explosion In 1968, nearly 100 years after HMS Challenger set sail, the Glomar Challenger drillship put to sea at the beginning of what was to become the modern equivalent of that historic voyage. By the end of this international scientific mission (late 1983) nearly 100 separate legs of the expedition had been completed and over 600 holes drilled into the deep ocean floor. At the same time, many extensive national marine programmes have ventured into the deep sea; wide-ranging international, national and individual work has been carried out on land; and an enormous amount of geological data has been obtained in the search for hydrocarbons. Moreover, fine-grained sedimentology no longer takes the back seat. Research has included the muddy turbidites and contourites, hemipelagites, and fine biogenic oozes, red clays and black shales; the processes and controls on the dispersion and deposition of these facies; their micro-
6
D.A.V. Stow and D.J. IV. Piper
characteristics and physical properties; the compaction, disgenesis and association with both hydrocarbon generation and metallic ores. Clearly, it is difficult to assess the historical significance of such research and, similarly, difficult to see where we are being short-sighted or blinkered by new paradigms. One factor that mitigates against progress is the sheer volume of data now available, mostly undigested and not assimilated into the mainstream of knowledge. It is this information explosion that has made it necessary to focus our attention so narrowly for this workshop and volume in the hope that it may thus be able to contribute a little towards the synthesis and understanding of one small class of rocks.
Methodology Part of the reason that the study of fine-grained sediments has for so long lagged behind that of coarser-grained rocks is related to methodology. Although the approach and techniques are largely the same, there are certain problems unique to the fine end of the grain-size spectrum. These include: (a) the resolution of the individual particles, often less than 5 /~m in diameter, requires high-powered instrumentation and routine use of electron microscopy; (b) fine-grained rocks cannot be adequately characterized in the field or from cores: laboratory-based analyses are imperative; (c) because clays flocculate, textural analysis cannot be easily used to infer hydrodynamic processes (in contrast to sands); and (d) the dramatic post-depositional changes suffered by clays and fine biogenic material often make it extremely difficult to reconstruct the sediment characteristics at the time of deposition. However, these problems are by no means insurmountable. Many standard techniques are now successfully applied to fine-grained sediments, and new methods or refinements continually being developed. Coulter counters are routinely used for grain-size analysis, scanning electron microscopy for fabric, structure and composition, X-ray diffraction and X-ray fluorescence for mineralogy and geochemistry, CHN analysis for organic carbon determination, induction magnetometers for measuring the anisotropy of magnetic susceptibility, and X-radiography for study of sedimentary structures. In order to fully understand and accurately interpret a given sedimentary succession, it is important that the study be approached from many directions, at different scales of operation and employing a wide range of methods. For fine-grained sediments in particu-
lar, it is crucial that field examination is combined with laboratory analysis. Perhaps the best general approach to use is that described by Potter et al. (1980) as the 'Question Set Approach'. They developed a series of questions for the study of shale (mudrock) successions, that help to lead the investigator through the required series of techniques and analyses at the micro, meso and macro scales. We have reproduced their Question Set below (Table 1). Clearly, each particular study might require a modified question set, especially studies of finegrained biogenic sediments, for example. In addition, many investigators might choose to study in greater detail just one aspect of the whole (e.g. clay fabric, sediment grain-size, etc) but it is nevertheless useful to see where this micro-study fits into the overall picture. Clearly, we do not intend to describe all the variety of methods used in this short introduction. Instead we list below (Table 2) the main source books on methodology with a brief comment on their relevance to the study of finegrained sediments.
Terminology Fine-grained sediments are those rocks, both hard and soft, biogenic and clastic, that have a dominant grain size in the clay or silt grades (i.e. over 50% < 63 #m). Deep-water implies anything below wave base (say about 50m) and so includes the deeper shelf areas, shelf basins and deep lakes as well as the open ocean basins and marginal seas. Most of the papers in this volume, however, are concerned with sediments deposited at oce~/nic depths greater than about 200m. As a discipline grows, so the terminology within that discipline proliferates and confusion is inevitable. However, it is clearly important for improved communication within that discipline, as well as within the earth sciences as a whole (e.g. between sedimentologists and engineers), that a generally-accepted standard terminology is widely used. Fortunately, the state of confusion in the terminology applied to fine-grained sediments is not yet too severe, and so we are able here to propose a degree of standardization. We outline first a purely descriptive terminology for fine-grained sediment classification and description and, second, some of the terms used in the interpretation of deep-water facies.
Descriptive terminology Classifications of fine-grained sediments have been based most commonly on texture and
TABLE 1.
Question set for study of shale (mudrock) sequence (from Potter et al. 1980)
Describing outcrops and cores and using wire-line logs 1. Where is the section? 2, What are the major units? 3. What is to be done next? 4. What should be described? 5. What terms should be used in the field for the description of the major lithologic types of shales? 6. How should the observations be recorded? 7. What palaeontologic observations can and should be made in the field? a. Relative abundances of different macrofossils? b. Is the distribution ofmacrofossils patchy, uniform, or random? Are they concentrated within beds or on bedding planes? c. Are the macrofossils intact and well preserved or fragments and worn? Molds or casts? Recrystallized or replaced? d. What functional types of organisms are present? Encrusters, sediment trappers or binders, epifauna, or infauna? Mobile or sessile? Suspension feeders, deposit feeders, scavengers, or predators? 8. What is the gamma-ray profile--what is it good for and how is it obtained? 9. Wire-line logs--what are they and how can they be best used? 10. Why and how to describe cuttings? Laboratory studies 1. What samples should be selected? 2. What tools and techniques should be used? 3. What sequence of study should be followed? 4. How should the observations be recorded? 5. What components are present, what is their abundance, and what do they all mean (fundamental to the understanding of every rock, the key questions are always the same)? a. Large detrital grains, such as quartz, feldspar micas, heavy minerals, and carbonate grains? b. Detrital clays? c. Authigenic grains, including carbonates (calcite, dolomite, and siderite), quartz, feldspar, zeolites, and the authigenic clays, glauconite, sepiolite, etc.? i. How are authigenic minerals recognized in a mudstone or shale? Those formed after deposition either by precipitation in pores or by transformation of original detrital minerals (by solution and replacement). it. Significance? d. Mineral cements, such as silica, carbonate, or zeolites? e. Floccules? f. Pellets and pelaggregates? g. Organic particles? h. What can be learned from micropalaeontology and palynology? 6. What textural parameters should be measured and what do they mean? a. What proportions of clay, silt, and sand? b. Percentage of large micas? c. Size ranges and modes of diverse detrital grains? d. How well oriented are the different components of the shale sediment--the clay minerals, the silt and sand grains commonly found in t h e m - - a n d what significance, if any, does their orientation have for palaeocurrents? ~. Perfection of framework orientation? it. Silt and sand grains? e. Burrowed and/or mottled textures? f. Pore geometry--kinds and amounts? g. Is there any significance to the shape of shale cuttings? h. What can be seen by radiography? 7. What name is to be used and how should the shale be classified now that we know so much about it? 8. The petrographic report: What is the best way to organize the foregoing petrology and texture into a useful, concise, and coherent petrographic report? 9. What can be learned from the study of inorganic geochemistry? a. The major elements? b. The trace elements? c. Exchangeable cations? d. Pore-water chemistry? e. Stable isotope geochemistry? 10. Organic chemistry? a. What are the best indicators of thermal history? b. What does the study of palaeobiochemical indicators tell us? Making a synthesis and basin analysis 1. Where did the mud come from? 2. How was it transported to its final depositional site? 3. At what water depth, sedimentation rate, oxygenation, and toxicity to life was the mud deposited? 4. What has happened to the mud since deposition?
8
D.A.V. Stow and D.J.W. Piper
TABLE 2. Principal texts on methods in sedimentary geology BOUMA,A.H. 1969. Methods for the Study of Sedimentary Structures. Wiley-Interscience, New York. 446 PP. Comprehensive for sedimentary structures. Good on X-radiography, graphical presentation, impregnation. BRINDLEY, G.W. & BROWN, G. (eds) 1980. Crystal structures of clay minerals and their X-ray identification. Mineral. Soc. Monograph. No 5. The most thorough and up-to-date reference for clay mineralogy and methods. CARVER,R.E. 1971. Procedures in Sedimentary Petrology, Wiley-Interscience, New York. 653 pp. One of the most comprehensive books on laboratory techniques. Covers sedimentary structures, grain size, grain attributes, and textural, mineralogical and chemical analyses. Lacks many techniques developed over the past 10-15 years. COMPTON, R.R. 1962. Manual of Field Geology. John Wiley, New York. 378 pp. Early but thorough treatment of what to do in the field. CONYBEARE,C.E.B. & CROOK,K.A.W. 1968. Manual of Sedimentary Structures. Australian Dept. Natl. Development. Bull. Bur. Min. Res. Geol. & Geophys. 102, 327 pp. A useful guide to structures in all types of sediments. FLOGEL, E. 1982 (english edition), Microfacies Analysis of Limestones. Springer-Verlag, New York, Berlin, 633 pp. Does not detail methods used in the study of limestones at any great length, but has wide-ranging coverage and extensive references. FOLK, R.E. 1974. Petrology of Sedimentary Rocks.
Hemphill Pub. Co., Austin, Texas. 159 pp. General sedimentology text with useful sections and references on methodology. FREe, R.W., (ed.) 1975. The Study of Trace Fossils. Springer-Verlag, New York. Useful illustrations of trace fossils in sediments. GR1FFITHS,J.C. 1967. Scientific Method in the Analysis of Sediments. McGraw-Hill, New York. 508 pp. Good on texture, fabric, composition, data processing and statistics. KUMMEL, B. & RAUP, B. (eds) 1965. Handbook of Paleontological Techniques. Freeman & Co., San Francisco. 852 pp. Still one of the most thorough books on palaeontological techniques. MULLER, B. 1967. Methods in Sedimentary Petrology. Hafner. Fairly comprehensive, but slightly dated. PETTIJOHN, F.J. & POTTER,P.E. 1964. Atlas and Glossary of Primary Sedimentary Structures, SpringerVerlag, New York. 370 pp. Well-illustrated guide to structures in all rock types. POTTER, P.E., MAYNARD, J.B. & PRYOR, W.A. 1980. Sedimentology of Shale, Springer-Verlag, New York. 303 pp. Informative study guide and reference source for shales, including guide and references to methodology. TUCKER,M.E. (1982). The FieMDescription of Sedimentary Rocks. Open University Press, Milton Keynes, UK and Halsted Press, New York. 112 pp. A useful field manual covering all sedimentary rock types.
composition and secondarily on fissility, colour, degree of metamorphism and depositional environment. We propose here a twofold division of fine-grained sediments into mudrocks, with an implied siliciclastic composition, and, biogenic mudrocks, having either a calcareous or siliceous composition (Tables 3 and 4). The term mudrock seems more appropriate than shale although both are widely used as general class names (e.g. Potter et al. 1980). In subdividing the two groups we have tried to use generally-accepted simple terms that can be readily applied in the field and subsequently modified by the appropriate descriptors after laboratory analysis. The terminology presented in Table 2 is broadly in line with that proposed by many previous authors (e.g. Wentworth 1922; Ingrain 1953; Shepard 1954; Dunbar & Rogers 1957; Folk 1968; Picard 1971; Weser 1974; Pettijohn 1975; Blatt et al. 1980; Potter et al. 1980). It is also commonly used by sedimentologists, marine geologists, engineers and soil scientists. Minor differences that exist between classification systems are
mainly concerned with the exact percentage of sand that a mud must contain before it becomes a sandy mud, and so on. We suggest it is better to keep the terms more general in application, so that the textural or compositional contents are estimates rather than strict definitions. Mixtures of sand, silt and clay are commonly displayed graphically on triangular diagrams (Fig. 2). Colour, chemical composition or genetic terms can be used as additional descriptors to the basic terms as appropriate (e.g. black shale, uraniferous mudstone). These, and mineralogical terms such as chlorite illite mudstone, are mainly applicable only after detailed laboratory investigations. Spears (1980) notes that the percentage of quartz in mudrocks is generally proportional to the grain-size or siltiness. In ancient well-lithifled rocks it is often simpler to estimate quartzcontent than grain size directly, particularly by laboratory methods, but the same basic terminology should still be applied. Lewan's (1979) attempt to erect a laboratory classification of very fine-grained sediments is also based on textural
D e e p - w a t e r f i n e g r a i n e d sediments," history, m e t h o d o l o g y a n d t e r m i n o l o g y TABLE 3. Mudrock terminology (after Stow 1981) Mudrock (> 50% less than 63 llm)
Basic terms Unlithified
Lithified/non-fissile
Silt Mud Clay
Lith!fied/fissile
Siltstone Mudstone Claystone
Silt-shale Mud-shale Clay-shale
Approx. proportions/grain-size > ~ silt-sized (4-63/~m) silt and clay mixture (< 63/~m) > 2 clay-sized ( < 4/~m)
Metamorphic terms Argillite Slate
slightly metamorphosed/non-fissile metamorphosed/fissile
Textural descriptors
Approx. proportions
Silty Muddy Clayey Sandy, pebbly, etc
> > > >
Compositional descriptors
10% silt-size 10% silt- or clay-size (applied to non-mudrock sediments) 10% clay size 10% sand-size, pebble-size, etc.
Approx. proportions
Calcareous Siliceous Carbonaceous Pyritiferous t Ferruginous Micaceous and others
> 10~ CaCO3 (foraminiferal, nannofossil, etc) > 10~ SiO2 (diatomaceous, radiolarian, etc) > 1~ Organic carbon Commonly used for contents greater than about 1-5%
criteria and compositional modifiers. However, his redefinition ofmudstone, shale, claystone and marlstone are not very helpful. Pelite and lutite (pelitic and lutaceous) are synonymous with mudstone (muddy) and are not recommended terms, particularly since pelite is widely used to indicate a metamorphic rock. Siltite for siltstone is also a redundant term. Argillaceous sedimentary rock is more conveniently replaced by mudrock, but argillaceous as a strictly compositional term (meaning rich in clay minerals) is still useful in certain cases. C LAY
..........
~MUD
: s.~!t..... !i:-i\
SAND
80
silt and clay mixture silt and clay mixture
50%
80
SILT
FIG. 2. Terminology for sand-silt-clay sediment admixtures (after Shepard 1954). The shaded area denotes mud, a term much used for silt-clay mixtures following Folk (1968).
The terminology for biogenic mudrocks shown in Table 4 follows closely the various systems in common use (e.g. Gealy et al. 1971; Weser 1974; Berger 1974; Davies et al. 1977). Dean et al. (1984) have proposed that a more formal intermediate category is needed between ooze and mud for mixtures of biogenic material and mud (Fig. 3). They recognized marl as a very common and useful term for calcareous biogenic muds and introduced sarl for siliceous biogenic muds and smarl for mixed calcareous and siliceous biogenic muds. The general term for sediments in this category therefore becomes arl. Descriptors of all kinds (composition, texture, colour, etc.) can then be applied to the basic terms as for the siliciclastic sediments. A single descriptor indicates the dominant component, and a second descriptor (added at the beginning) indicates the second most important component when applied to oozes and muddy oozes (arls). Further classifications of lithified carbonate rocks, commonly after laboratory analyses, should follow the systems proposed by Folk (1959) or Dunham (1962). The former is based on the type of carbonate particle (ooliths, skeletal grains, pellets) and the type of cement (microcrystalline micrite or macrocrystalline sparry calcite). The latter is concerned more with the relative proportions of matrix and discrete particles. The terms calcilutite and calcisiltite, together with calcarenite and calcirudite for coarser-grained carbonates, can be useful when the grain-size of
Io
D.A. V. S t o w and D.J. W. Piper TABLE 4. Biogenic mudrock terminology Biogenic mudrock (50% biogenic, 50% less than 63 pm)
Basic terms Unlithified Ooze Muddy ooze* Mud
Descriptors Calcareous t Siliceous Foraminiferal Diatomaceous Radiolarian | etc. j Carbonaceous
Lithified
Approx. proportions
chalk, limestone, diatomite, chert, etc. Argillaceous chert, chalk etc. mudstone
> 2/3 biogenic biogenic/terrigenous mix > 2/3 terrigenous
Approx. proportions e.g. siliceous ooze, biogenic SiO2 > 50% foraminiferal ooze, forams > 50% foraminiferal-nannofossilooze, forams & nannos > 50% with nannos dominant diatomaceous muddy ooze, diatoms dominant biogenic component calcareous mud, 300,/0> CO3 > 10,~ organic carbon > 1%
* Dean et al (1984) suggest Arl=muddy ooze hence marlstone, sarlstone, smarlstone. the rock is considered its most important feature as, for example, with resedimented carbonates. Another area of descriptive terminology in which there is often confusion is that applied to sedimentary structures, and in particular to lamination and bedding. The most widely accepted system for stratification thickness is that of Ingram (1954) shown below (Table 5). Grading can be either positive (normal) in which the grain-size decreases upwards, or negative (reverse) in which the grain-size increases upwards. This is commonly applied to a single bed or discrete unit. Series of beds that show an overall grain-size change are termed finingupward or coarsening-upward sequences, although this usually refers only to the coarser (silt or sand) beds within the sequence. Similarly,
a graded laminated unit is an interval (commonly 2-20 cm thick) through which discrete silt laminae, separated by a finer-grained mud, show an overall upward decrease in grain-size. The terms massive, homogeneous and structureless are used synonymously to mean an ungraded and featureless sediment. In Fig. 4 we list the common sedimentary structures encountered in fine-grained sediments and show the symbols most often to represent them graphically. Standardization in this aspect of description and terminology would be very helpful. TABLE 5. Stratification types and thicknesses
(after Ingram 1954) Very thickly bedded Thickly bedded Medium bedded Thinly bedded Very thinly bedded Thickly laminated Thinly laminated Very thinly laminated
MUD
> 1m 30-100 cm 10-30 cm 3-10 cm 1-3 cm 3-10 mm 1-3 mm < 1 mm
Interpretative terminology
BIOGENIC CARBONATE
2/3
approx 2/3
BIOGENIC SILICA
FIG. 3. Terminology for biogenic carbonage, biogenic silica and terrigenous mud mixtures (as proposed by Dean et al. 1984).
The first phase of a study is, clearly, to describe the sediments and their characteristics, as far as possible objectively and using a standard format. The second phase involves interpretation of these primary data in terms of depositional processes, environmental setting, stratigraphic position and so on. International stratigraphic terminology is
Deep-water fine grained sediments," history, methodology and terminology SYMBOL
///
DESCRIPTION
SYMBOL
BEDDING PLANE STRUCTURES
massive, structureless
surface lineation
parallel bedding
flutes
parallel lamination
grooves
inclined bedding/lamination
load casts
cross-bedding/lamination
scour and fill
flaser bedding, fading ripples
flame (injection) structure
convolute bedding/lamination
mud cracks
slumped
INTERNAL STRUCTURES
oo@
lenticular bedding/lamination
~176
o~176
negative (reverse) grading
water escape pipes
II
dish structures
u~,s
positive (normal) grading
filled fracture
l
LAYER BOUNDARIES
microfault
sharp contact
concretion
sharp irregular (erosive) contact
OTHER
disturbed contact
disturbed section
gradational contact
fining-upward sequence
A V
BIOGENIC STRUCTURES
coarsening-upward sequence
bioturbation minor (0-50%)
tli 0
imbrication mud clast
wedge-shaped layer m
DESCRIPTION
STRATIFICATION
wavy bedding/lamination
II
interval over which structure occurs
bioturbation moderate (50-60~
(
bioturbation intense ( > 6 0 % )
I())
)
structure indistinct structure very indistinct
burrows
FIG. 4. Typical sedimentary structures and recommended graphic symbols (modified from Bouma 1962, and DSDP schemes). well established (Hedberg 1976; Holland et al. 1978). Environmental terms present more of a problem, particularly with regard to interpretation of palaeoenvironments. Deep-water morphological environments in which fine-grained sediments accumulate have been initially defined in large oceans (Heezen et al. 1959) as continental shelf (especially outer shelf), slope, rise and abyssal plain. The terms basin slope, basin rise and basin plain are more applicable to smaller basins (such as in the Mediterranean sea or California Borderland). Deep-sea fans are large morphogenetic environments that include both rise and plain settings. Many terms have been applied to particular
subenvironments or morphological elements within these larger-scale settings (e.g. Stow, in press). The most common include canyons, channels, gullies, levees, interchannel areas, lobes, slump masses, debris-flow masses, slump scars, sediment drifts, upper, middle and lower fans, and so on. Such present day morphological features are easily recognized by acoustic profiling in the oceans, but their recognition in ancient rock sequences requires a detailed knowledge of stratigraphic relationships and large-scale facies associations and sequences. This is particularly difficult in fine-grained rocks because of their paucity of diagnostic facies indicators and their common high degree of tectonic deformation. Extreme caution should therefore be used in
D.A.V. Stow and D.J.W. Piper
I2
TABLE 6.
Deep-waterfine-grained sediment facies (from Stow & Piper, this volume) Depositional process
Facies
Characteristics
Silt turbidite Mud turbidite Biogenic turbidite Disorganized turbidite
> 50,~ silt "~ > 50~ mud > 500/;,biogenics mud or biogenics or mixture
Silty (sandy) contourite
) > 50~ silt or fine "} commonly mixed sand size I> biogenic-terrigenous, k, bottom irregular "sequence', ( ( c o n t o u r ) > 50~o mud size bioturbated. ) current
Muddy contourite Pelagic ooze Hemipelagite Pelagic clay
> 2/3 biogenic "~ (planktonics) biogenick;. terrigenous m i x [ > 2/3 terrigenous clay (mud). ,,'
standard structural sequence no sequence
no structural sequence, rhythmic interbedding common, bioturbated
t
t
turbidity current
pelagic settling
Fine-grained sediment facies in deep water can making palaeoenvironmental interpretations for mostly be interpreted in terms of these depositancient fine-grained rock successions. There are three main groups of processes by ional processes, although it must be recognized which fine-grained sediments are deposited in the that the processes are in fact part of a continuum, deep-sea (Gorsline, this volume; McCave, this and that a similar continuum of facies therefore volume). (a) Resedimentation processes (synony- exists. Stow & Piper (this volume) have identified mous with mass gravity transport) are all those nine separate facies models (Table 6), each related processes that move sediment downslope over the to a specific depositional process and each characsea floor from shallower to deeper water and that terized by a standard sequence of structures or a are driven by gravitational forces. For fine- 9standard suite of sedimentary features. We suggest that these facies models, although grained sediments the most important of these are slides and slumps, sediment creep, debris-flows not exhaustive of the possible processes and facies and turbidity currents. (b) Normal bottom cur- in the deep sea, are all relatively well documented rents are all those deep currents that erode, and understood. Until similar evidence exists for transport and deposit sediment on the sea floor other facies (e.g. the 'nepheloidites' or 'suspenand that are driven by normal thermohaline or sion cascadites' invoked by some authors), it wind-driven circulation within the oceans. They seems better not to use the terms. Similarly, include, internal tides and waves, canyon cur- purely descriptive facies terms such as 'laminites' rents, bottom (contour) currents and deep surface and 'unifites' would be better replaced by the currents. (c) Pelagic settling through the water standard descriptive terminology outlined precolumn is an ubiquitous, slow and predominantly viously. vertical process under the influence of gravity. Much of the sediment settles as flocs and faecal pellets rather than individual particles.
References ANDREE, K. 1920. Geologic des Meeresbodens. Borntraeger (PUN.), Leipzig. 2, 689 pp. ARCHANQUELSKY, A.D. 1927. Ob Osadkak Chernov Morya i ik Znachenii v Pozanii Osadochnik Gornik Porod. Bull. Soc. Naturalistes Moscou Sec. Geol., N.S. 5, 261-89. ARRHENIUS, G. 1950. The Swedish Deep-Sea Expedition. The geological material and its treatment with special regard to the eastern Pacific. Geol. Fdr Stoekh. F6rh, 72, 185-91.
BAILEY, E.B. 1936. Sedimentation in relation to tectonics. Bull. geol. Soe. Am., 47, 1713-26. BALLANCE,P.F. 1964. Streaked out mud ripples below Miocene turbidites, Puriri Formation, New Zealand. J. sed. Petrol., 34, 91-101. BERGER, W.H. 1974. Deep-sea sedimentation. In: Burk, C.A. & Drake, C.L. (eds), The Geology of Continental Margins. Springer-Verlag, New York. 213-41. BLATT,H., MIDDLETON,G. & MURRAY,R. 1980. Origin
Deep-water fine grained sediments," history, methodology and terminology of Sedimentary Rocks. 2nd edn. Prentice-Hall, New Jersey. 782 pp. B6GGILD, O.B. 1916. Meeresgrundproben der Sigboga expedition. Sigboda-Expeditie Monographie, 65, 1-50.
BOUMA, A.H. 1962. Sedimentology of some Flysch Deposits. Elsevier, Amsterdam. 168 pp. BRAMLETTE, M.N. & BRADLEY, W.H. 1940. Lithology and geologic interpretations. Pt. 1. In: Geology & Biology of North Atlantic Deep-Sea Cores. USgeol. Surv. Prof. Pap., 196, 1-24. CRAIG, G.Y. & WALTON,K. 1962. Sedimentary structures and palaeocurrent directions from the Palaeozoic rocks of Kirkcudbrightshire. Trans. Edin. geol. Soc., 19, 100-19. DALV, R.A. 1936. Origin of submarine 'canyons'. Am. J. Sci., 31, 401-20. DAVIES, T.A., MUSICH, L.F. & WOODBURY, P.B. 1977. Automated classification of deep-sea sediments. J. sed. Petrol., 47, 650-6. DE RAAF, J.F.M. READING, H.G. & WALKER, R. G. 1965. Cyclic sedimentation in the lower Westphalian of north Devon, England. Sedimentology, 4, 1-52.
DEAN, W.E., STOW, D.A.V., BARROW, E. & SCHALLREUTER, R. 1984. A revised sediment classification for siliceous-biogenic calcareous-biogenic and nonbiogenic components. Init. Repts. DSDP, 75, US Govt. Print. Off., Washington, DC. DIETZ, R.S. 1961. Continent and ocean basin evolution by spreading of the sea-floor. Nature, 190, 854-7. DUNBAR, C.O. & ROGERS, J. 1957. Principles of Stratigraphy. John Wiley, New York. 356 pp. DUNHAM, R.J. 1962. Classification of carbonate rocks according to depositional texture. Am. Ass. Petrol. Geol. Mem., 1, 108-21. DZULYNSKI, S. 8t~ KINLE, S. 1957. Problematic hieroglyphs of possible organic origin from the Belovoza beds. Soc. geol. Pol. Ann., 26, 266-9. -KSIAZKIEWICZ,M. & KUENEN, P.H. 1959. Turbidites in flysch of the Polish Carpathian Mountains. Bull. geol. Soc. Am., 70, 1089-118. EINSTEIN, H.A. & KRONE, R.B. 1962. Experiments to determine modes of cohesive sediment transport in salt water. J. geophys. Res., 64, 1451-61. GEALY, E.L., WINTERER, E.L. & MOBERLY, R. 1971. Methods, conventions and general observation. Init. Repts. DSDP, 7, U.S. Govt. Print. Off., Washington,-DC 9-26. FOLK, R.L. 1954. The distinction between grain size and mineral composition in sedimentary rock nomenclature. J. Geol., 62, 344-59. - 1959. Practical petrographic classification of limestones. Bull. Am. Ass. Petrol. Geol., 43, 1-38. -1968. Petrology of Sedimentary Rocks. Hemphills' Pub. Co., Austin, Texas. 170 pp. - 1973. Carbonate Petrography in the Post-Sorbian Age. In: Ginsburg, R.N. (ed.), Evolving Concepts in Sedimentology. Johns Hopkins University Press, Baltimore. 118-58. FucrlS, T. 1877. Ober die Entstehung der Aptychenkalke. Sber. Akad. Wiss. l/Vien Math. Nat. Klasse, Abt. I, 76, 329-34.
13
HEDBERG, H.D., (ed) 1976. International Stratigraphic Guide. Wiley-Interscience, New York. 200 pp. HEEZEN, B.C. 1959. Dynamic processes of abyssal sedimentation: erosion, transportation and redeposition on the deep sea floor. Geophys. J. R. astr. Soc., 2, 142-63. - & JOHNSON, G.L. 1963. A moated knoll in the Canary Passage. Dt hydrogr. Z., 16, 269 pp. , HOLLISTER, C.D. & RUDDIMAN, W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science, 152, 502-8. - - - , THARP, M. & EWING, M. 1959. The floors of the oceans. 1. The North Atlantic. Geol. Soc. Am. Spec. Pap. 165, 122 p.p. HEIM, A. 1924. Uber submarine Denudation und chemische Sedimente. Geol. Rdsch., 15, 1-47. HESS, H.H. 1962. History of ocean basins. In: Engel, A.E.J. et al. (eds), Petrologic studies: A Volume in Honour ofA.F. Buddington. Mem. geol. Soc. Am., Boulder, Co., 569-620. HESSE, R. 1975. Turbiditic and non-turbiditic mudstone of Cretaceous flysch sections of the East Alps and other basins. Sedimentology, 22, 387~,16. HINDE, G.J. 1890. Notes on Radiolaria from the Lower Palaeozoic rocks (Llandeilo-Caradoc) of the south of Scotland. Ann. Mag. nat. Hist., ser. 5, 6, 40-59. HOLLAND, C.H. et al. 1978. A guide to Stratigraphical Procedure. Geol. Soc. Lond. Spec. Rep., 11, 18 pp. HUBERT, J.F. 1964. Textural evidence for the deposition of many western North Atlantic deep sea sands by ocean-bottom currents rather than turbidity currents. J. Geol., 72, 757-85. HUXLEY, T.H. 1858. On some organisms living at great depths in the North Atlantic Ocean. Q. Jl. microsc. Sci., 8, N.S., 203-12. INGRAM, R.L. 1953. Fissility of mudrocks. Bull. geol. Soc. Am., 64, 869-78. -1954. Terminology for the thickness of stratification and parting units in sedimentary rocks. Bull. geol. Soc. Am., 65, 937-8. JOHNSON, D. 1938. The origin of submarine canyons. J. Geomorph., 1, 230-43. KELLING, G. 1958. Ripple-mark in the Rhinns of Galloway. Trans. Edinb. geol. Soc., 17, 117-32. KLEIN, G. DE V. 1966. Dispersal and petrology of sandstones of Stanley-Jackfork boundary, Ouachita fold belt, Arkansas and Oklahoma. Bull. Am. Assoc. Petrol. Geol., 50, 308-26. KUENEN, PH, H. 1937. Experiments in connection with Daly's hypothesis on the formation of submarine canyons. Leid. geol. Meded., 8, 327-35. -1950. Marine Geology. John Wiley, New York. 568 pp. -• MIGLIORINI, C.I. 1950. Turbidity currents as a cause of graded bedding. J. Geol., 58, 91-127. KUHN, T.S. 1970. The Structure of Scientific Ret'olutions. Univ. Chicago Press. 210 pp. LESLEY, J.P. 1892. Summary description of the geology of Pennsylvania. Penns. Geol. Surf,. 2nd Rep., 1, 1-720. LEWAN, M.D. 1979. Laboratory classification of very fine-grained sedimentary rocks. Geology, 6, 645-8. MIDDLETON, G.V. 1965. Introduction, Primary Sedimentary Structures and their Hydrodynamic Inter-
I4
D.A.V. Stow and D.J.W. Piper
pretation. Soc. econ. Paleont. Min. Spec. Pub., 12, 1-4. MIGLIORINI, C.I. 1946. L'eta del macigno dell'Appennino sulla sinistra del Serchio e considerozione sul rimaneggiamento dei macroforaminiferi. Bull. Soc. geol. Ital., 63, (1944), 75-90. MOLENCRAAF, G.A.F. 1922. On manganese nodules in Mesozoic deep-sea deposits of Dutch Timor. Proc. Sect. Sci. K. ned. Akad. Wet., 18, 415-30. MtJRRAY, J. & RENARD, A.F. 1891. Report on deep-sea deposits based on specimens collected during the voyage of HMS Challenger in the years 1873-1876. In: Challenger Reports, HMSO, Edinburgh. 525 pp. NATLAND, M.L. 1933. Depth and temperature distribution of some Recent and fossil Foraminifera in the southern California region. Bull. Scripps Instn. Oceanogr., La Jolla, Tech. Ser. 3, 225-30. PETTIJOHN, F.J. 1943. Archaean sedimentation. Bull. geol. Soc. Am., 54, 925-72. -1975. Sedimentary Rocks. 3rd edn. Harper & Row, New York. 628 pp. PICARD, M.D. 1971. Classification of fine-grained sedimentary rocks. J. sed. Petrol., 41, 179-95. POTTER, P.E., MAYNARD, J.B. & PRYOR, W.A. 1980. Sedimentology of Shale. Springer-Verlag, New York. 303 p.p. RUEDEMANN, R. 1935. Ecology of black mud shales of eastern New York. J. Paleont., 9, 79-91. RUNCORN, S.K. (ed.) 1962. Continental Drift. Academic Press, New York. 338 pp. SHEPARD, F.P. 1932. Sediments of the continental shelves. Bull. geol. Soc. Am., 43, 1017-40. -1948. Submarine Geology. Harper & Row, New York. 348 pp. - 1954. Nomenclature based on sand-silt-clay ratios. J. sed. Petrol., 24, 151-8. SORBY, H.C. 1861. On the origin of the so-called 'Crystalloids' of the chalk. Ann. Mag. nat. Hist. ser., 3, 8, 193-200.
1908. On the application of quantitative methods to the study of the structure and history of rocks. Quart. geol. Soc. Lond., 64, 171-233. SPEARS,D.A. 1980. Towards a classification of shales. J. geol. Soc. Lond., 137, 125-9. STEINMANN, G. 1905. Geologische Beobachtungen in den Alpen, II. Ber. naturf. Ges. Freiburg, 16, 18-67. STOCKS, T. 1933. Die Echo profile Wissenschaftliche Ergebnisse der Deutschen Atlantischen Expedition auf den "Meteor' 1925-1927, vol. 2. STOW, D.A.V. & LOVELL, J.P.B. 1979. Contourites: their recognition in modern and ancient sediments. Earth-Sci. Reviews, 14, 251-91. - - i n press. Deep-sea clastics. In: Reading, H.G. (ed.), Sedimentary Environments and Facies. 2rid edn. Blackwell Scientific Publications, Oxford. SUESS, E. 1875. Entstehung der Alpen. W. BraumfiUes, Vienna. 168 pp. VAN STRAATEN,L.M.U.J. 1967. Turbidites ash layers and shell beds in the bathyal zone of the southeastern Adriatic Sea. Revue G(ogr. phys. Gdol. dyn., 9, 219-39. VINE, F.J. & MATTHEWS,D.H. 1963. Magnetic anomalies over oceanic ridges. Nature, 199, 947-9. WALKER,R.G. 1973. Mopping up the turbidite mess. In: Ginsburg, R.N. (ed.), Evolving concepts in Sedimentology. Johns Hopkins University Press, Baltimore. 1-37. 1976. Facies Models: I. General Introduction. Geoscience Canada, 3, 21-4. WALTHER,J. 1893. Einleitung in die Geologie als historische Wissenschaft. 3 vols. Fischer Verlag. Vena. WENTWORTH, C.K. 1922. A scale of grade and class terms for clastic sediments. J. Geol., 30, 377-92. WESER, O.E. 1974. Sediment classification based on Deep Sea Drilling Project drilling. Geol. Soc. Am., Program Abstr., 6, 1073-4. -
-
D.A.V. STOW, Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, Scotland. D.J.W. PIPER, Atlantic Geoscience Center, Geological Survey of Canada, Bedford Institute of Oceanography, P O Box 1006, Dartmouth, Nova Scotia, Canada, B27 4A2.
A review of fine-grained sediment origins, characteristics, transport and deposition D.S. Gorsline SUMMARY: Fine-grained sediments and sedimentary rocks make up as much as 75% of the present and past sedimentary records. River discharge is the largest single source of fine-grained material, followed by biological, volcanic and aeolian sources. The composition, texture and bulk properties that characterize different sediment facies are controlled primarily by climate, tectonics, sediment supply, oceanic dispersal systems and biological activity. Fine sediments are transported into deep water by low-concentration nepheloid plumes, turbidity currents of a range of concentrations, flows and mass-movements. Mass-movement may be the major process delivering fine-grained sediment to the deep-sea floor over long time periods. Abyssal plains, the end of the passive margin transport system, are more affected by turbidity currents. Much bottom current reworking of sediments occurs at all depths in the ocean with results that may be very subtle. In shallower waters, balances between supply and dispersal may allow for the accumulation of mud belts on shelves and much fine sediments passes through such systems. Organic matter concentrations in fine sediment may be more a function of accumulation rates of organic versus terrigenous matter than local biological productivity. Fine-grained sediments require the application of methodology and results of a broad range of scientific disciplines. Effective understanding of the fine record will require interaction between several sciences.
The theme of this symposium is fine-grained sedimentation in deep water. Viewed from global sedimentation perspectives, the oceans are the ultimate sink for world sedimentation. Although continental fine sedimentary environments are not discussed here, some papers on these topics have been included in a complementary symposium edited by Reinhard Hesse (Sedimentary Geology, in press). Effective understanding of the oceanic fine sedimentation system requires cooperative studies by geologists, biologists, chemists and oceanographers. Therefore, the introductory discussion will, in part, be aimed at non-geologists. On the other hand, since many geologists are not familiar with oceanography, I will include some brief discussion of ocean circulation features that affect the distribution and deposition of fine sediments. This paper complements the paper in this volume by Stow & Piper which discusses deepwater fine-grained sedimentary facies. I will briefly touch upon facies and some special sedimentary features, but primarily from an oceanographer's view of the environmental conditions and characteristics rather than specific discussion of sedimentary structures and textures. Readers interested in the oceanic eddies and rings noted later should see the symposium volume by Robinson (1983).
Fine sediment mass, source and characteristics Mass and Source
The stratigraphic importance of fine-grained sediments is evident from conservative estimates of the proportion of silts and clays and their lithified equivalents in the geologic record. Recent authors (Potter et al. 1980; Blatt 1982) estimate that 70-75% of sediments and sedimentary rocks are fine particulate aggregates, including fine biogenic sediments and rocks. Years ago, Clarke (1924) noted that average shales contain one-third clay minerals, one-third quartz and one-third organics, carbonates and other minor minerals. The primary source of fine deposits is river discharge; and, secondarily, volcanic explosive eruptions, aeolian transport and biological production. There are other interesting sources such as chemical precipitates and meteoritic debris which contribute very minor amounts of material to the overall budget. Recent estimates for the bulk contribution of the major sources in present times give river discharge as 1.5-2 • 101~tons yr - l (Garrels & Mackenzie 1971; Milliman & Meade 1983); aeolian transport, exclusive of volcanic ejecta, as approximately 107 tons yr-1 (based on data from Prospero & Bonatti 1969; Windom 1969; Pewe 1981; Prospero 1981); the contribution of explosive volcaniclastics as approximately
17
I8
D.S. Gorsline
107 tons yr -1 (estimated from eruption sizes, Huang et al. 1979; Smith 1979); and fine-grained biogenic sediment sources of the order of 109 tons yr-1, based on stratigraphic estimates and ocean carbonate budgets (Broecker 1974; Blatt 1982). River discharge
The bulk of the annual river discharge comes from the world's 20 largest rivers, with the three largest Asian rivers contributing almost onequarter (Milliman & Meade 1983). Because of the limited number of large rivers, areas of major accumulation in the oceans are relatively few in number. For example, the few giant submarine fans are mostly tributary to the largest Asian rivers (Curray & Moore 1971). River load is primarily a function of drainage area for given climatic regions and river drainage areas are, to a first order, a function of tectonics. Inman & Nordstrom (1971) have shown that the major rivers drain into trailing edge or marginal seas due to the displacement of continental divides towards the collision edge of continental plates. As Strakhov (1970) has discussed, the amounts of sediment discharged can be related on a global scale to latitude which, in turn, controls the factors of temperature, rainfall, vegetation and evapo-transpiration. Undoubtedly some of these relationships were different for pre-Middle Palaeozoic time, but since most of the sedimentary record is younger than this (Garrels & Mackenzie 1971) we need not consider a nonvegetated world. Mid-latitude streams, or streams draining high plateaus in the case of some of the large Asian rivers, deliver the largest amounts of sediment per discharge area (Milliman & Meade 1983). The major rivers deliver very large loads of relatively fine grain-size. As noted earlier, these tend to debouch from the trailing edge of the cratons. Gibbs (1967) has shown the importance of the highlands in the headwaters of big rivers in determining the characteristics of the river particulate load. Undoubtedly the variation in land-sea surface area with time has produced matching variations in river load delivered. It is also likely that there is a minimum land area necessary to provide significant sediment discharge. Volcaniclastic contribution
Volcanic explosive eruptions tend to be associated with the andesitic and dacitic volcanoes of active margins and plate convergences and with large silicic eruptions derived from major batho-
liths (Smith 1979). Rate of plate subduction may also be a factor. Variable proportions of the erupted material are ejected as local low altitude air-fall and/or as high altitude injection into the stratosphere or high troposphere (Sheridan 1979). The deposits resulting from these events are strongly influenced by the local winds and upper atmosphere circulation, respectively. Although initially from point sources, the high atmospheric injections are mixed and dispersed so that the resulting air-fall deposits are globally distributed. Thus the general distribution of this fraction is essentially the same as the global distribution of dust from other sources (Windom 1969). The ash-falls of this large scale and their depositional record form an intriguing source of data regarding the periodicity of major volcanic events. Many authors have considered the periodicity of volcanism in the oceans and globally (e.g. Stille 1924; Umbgrove 1947; Huang et al. 1975; Ninkovitch & Donn 1975; Vogt 1979; Kennett 1981 ). In deep ocean distal pelagic sediments, the lithic contribution is probably almost entirely from wind-borne sources and oceanic volcanic input. This was probably also true in the past (e.g. Lindstrom 1974, for Ordovician pelagites). Aeolian contribution
Non-volcanic aeolian transport to the oceans tends to be from arid lands or from large outwash systems. The geographic arrangement of these types of areas and the major wind-belts determines whether material is transported to the deep ocean. Windom (1969) examined the dust falls deposited in ice of the polar caps and large glaciers and compared these values with estimates of the wind-borne contribution in lower latitude deep-sea sediments as determined by applying the composition data derived from the ice cap deposits to the components observed in the oceanic accumulations. He noted that the influx is affected by latitude, distance from source and climatic factors. (See also Rex & Goldberg 1958; Rex et al. 1969.) Loess is a major continental fine deposit in periglacial areas and downwind from arid regions. These deposits are eroded by the major stream systems draining the continental interiors. In the present mass budget, a large portion of the silt fraction may come from erosion of these glacial deposits. Analysis of the influence of arid regions upwind of ocean areas has been summarized by Pewe ( 1981), Prospero (1981) and earlier workers (see Rapp 1974). Windom (1969) states that
A review o f f i n e - g r a i n e d sediment aeolian input contributes between 10 and 75~o of the nonbiogenic fraction of deep ocean sediments depending on distance from source, latitude, dilution and effectiveness of the tracers used. Heath et al. (1973) have shown the shift in the pattern of wind-borne quartz in deposits in subtropical latitudes in the eastern Pacific over late Pleistocene time. On a shorter time-scale, the dust plumes from the Sahara into the Atlantic shift in latitude with the seasonal shift in windbelts. Prospero (1981) has noted the generally low windborne concentrations in the distant ocean areas outside the arid belts.
Biogenic contributions Biogenic sediments are a world unto themselves. Many pelagic tests, capsules, plates and spines are silt size and smaller, and so primary biological productivity is an important fine particulate source. Bioerosion is an important source of fine lime particulates (see Futterer 1974; Hein & Risk 1975). Accumulation ofbiogenic sediments represents an intricate balance between production and solution (Berger 1974). Variations in biogenic content and in the stratigraphic appearance of biogenic-influenced sedimentation are a function of dilution of biogenic production by terrigenous influx. Ginsburg & James (1974) have discussed the importance of biogenic contribution to ocean margins other than the tropics, and note that although diluted by terrigenous influx, the middle and higher latitude contribution represents a major addition to the tropical biogenic source that is most generally used as the basis for biogenic budgets. Biological production of fine particulates is, of course, related to areas of high productivity. The older view was that these were essentially in areas of major upwelling along eastern ocean boundaries, the eastern equatorial system and major ocean water mass convergences as in the Circumpolar Antarctic Convergence. More recent work has shown that western boundary currents can generate zones of quite high productivity (Dunstan & Atkinson 1976; Janowitz & Pietrafesa 1980; Blanton et al. 1981;) and that the central gyres of the surface ocean circulation are also much more productive than was thought a few years ago (Shulenberg & Reid 1981; Jenkins 1982; Martinez et al. 1983). Parrish (1983) has introduced an interesting but preliminary multicomponent model in which the spectrum of organic-rich sediments and sedimentary rocks, cherts, chalks, phosphate rocks and glauconite are related to a range of oceanographic factors and mass sediment budgets from
19
land. Certainly, as noted earlier, the older simple view of geographically restricted ocean upwelling cannot be simply applied to palaeogeographic reconstructions. Organic content alone is not simply a function of original organic production but must include consideration of lithic dilution.
Composition and bulk properties Clays Clay is defined texturally as all material finer than 4 /~m (Udden 1914; Wentworth 1922). In this discussion, clay will be used in its mineralogical sense with a grain-size usually less than 10 #m and typically less than 2 #m (referring to discrete grains and not aggregates). Mineralogically, clay is composed of a family of silicate minerals with sheet structures similar to the phyllosilicate micas. The clays represent the stable product of the weathering of feldspars and micas at earth surface conditions. Clays possess the property of ion exchange to a greater or lesser extent depending on the clay species and this property reflects the unsatisfied bonds on and within the sheet structures. This property is an important one in that it provides a mechanism for transfer of organics and metal ions and also is the primary factor in the characteristic aggregation of clay grains in ocean environments (Gibbs 1981; Gibbs & Heitzel 1982). Due to the charges on their surfaces, clays have cohesion, which is a basic factor influencing the strength of fine sediments of high clay content. Clay mineralogy reflects climate and relief, and secondarily lithology. Several authors (e.g. Loughnan 1969) have shown that a specific feldspathic rock can deliver a wide variety of clay species depending on such factors as temperature, rate of soil drainage, and chemical environment. For large areas with heterogeneous lithologies, climate is the main control. Storage time increases with increasing drainage area and this also influences composition and texture. Changes of global climate therefore exert a strong influence on fine sediment characteristics; the proportion of silt to clay and of silt +clay to coarser grain sizes. As has been shown by clay mineralogists (Loughnan 1969) and by large scale oceanic studies (Biscaye 1965; Griffin & Goldberg 1968) there is a primary latitudinal variation in dominant clay species that reflects climate. In general, tropical clays tend to be laterites or kaolinites; mid-latitudes produce montmorillonites and illites; and polar latitudes produce illites and chlorites.
20
D.S. Gorsline
Silts
Bulk properties
Texturally, silts range between 4 and 63 /~m in grain-size, although both silt and sand are part of a continuous spectrum of particle sizes. Silt is a particularly diagnostic grain-size class for studies of deposition and reworking by bottom currents and of general directions of fine sediment transport (e.g. McCave 1982). Silts are typically heterogeneous mixtures of detrital primary minerals of which quartz is most common, and feldspars and ferromagnesian minerals are common accessories. Silt origin has been attributed to glacial grinding (see Smalley 1966; Kuenen 1969), volcaniclastic eruptions (Smith 1979) and weathering processes in soils in tropical and mid-latitude regions (Pye & Sperling 1983). Many pelagic biologic tests, capsules, spines and plates are silt size, and so primary biological production produces much silt as well as clay size particulates (Nahon & Trompette 1982). Silts, because of the small particle size, have large grain surface areas and so weathering will probably rapidly remove metastable or unstable minerals leaving quartz as the dominant mineral. If, as has been suggested by Pye & Sperling (1983), much silt is formed by chemical fracturing of strained quartz in the soil profile, then silt initially begins with a high quartz content. Hydrodynamically, silts can be defined as particles that are moved primarily in suspension, while sands are moved in traction or saltation. The boundary is a broad one but probably lies somewhere between 30 and 100/~m (Inman 1949). In most rivers, fine sand is moved as suspension load at times of flood (Brownlie & Taylor 1981). In the ocean, the levels of turbulence and shear stress are lower and particles larger than about 30 ~m generally move as bedload. Under strong wave surge in coastal and inner shelf regions, sands are rarely suspended more than a few tens of centimetres above the bottom (Cook & Gorsline 1972). Wildharber (1966) found that the typical suspension load a few kilometres off southern California was already limited to particles with a coarsest diameter of 30/~m, and most were below 20 Ftm. Since fine sediments of silt size move in suspension, they will tend to preserve their original grain shape and this may be a means of pinpointing source in a given sedimentary system. Ehrlich and his associates (e.g. Ehrlich & Weinberg 1970; Ehrlich et al. 1980) have pioneered use of Fourier statistical methods to analyse ~grain shapes as a means of making source determinations. Silt grain shape analysis is a promising area for research.
There are three important bulk properties of fine-grained sediments: (a) the sediment strength, which is a function of the cohesive properties of clay particles, internal grain friction and grain packing; (b) the bulk density, which is a function of water content, sediment composition and void ratio; and (c) sediment fabric, or the stacking arrangement of clay particles and aggregates. The behaviour of fine-grained sediments on slopes depends on these three bulk properties, in addition to the slope angle and sediment load. The water content (or bulk density) appears particularly important (Field 1981b; Almagor 1982; Bennett & Nelson 1983). Thus cohesive sediments with high water contents (low bulk densities) may fail even on low gradients (Lewis 1971; Coleman 1976; Coleman & Garrison 1977); and cohesive sediments with low water content (high bulk density) may be stable on relatively high gradient slopes. To a first approximation, for a given textural type, the water content is directly related to rate of accumulation, thus it is evident that rapidly deposited sediments on slopes may be metastable or unstable and subject to mass failure. On a smaller scale, several workers have examined the microstructure of clay fabrics, and related these both to the depositional mode and to bulk properties (Bennett et al. 1981; Hein, in press). Bennett and his co-workers have examined the arrangement of clay flakes, domains (stacks of clay flakes) and other aggregate forms (pellets) in sediments of various bulk densities. As a rule, the lower density sediments have more open packing with larger voids; flakes and domains are arranged in loose chains. In denser sediments, the structures are flatter and more collapsed and voids are flat and lenticular. Hein and her associates are presently examining the three dimensional packing of fine grains in sediments from a variety of environments in the California Borderland. Preliminary work has shown that the bulk properties of the various facies and depositional environments are different. It will be interesting to see if the flake fabrics show the influence of creep or mass failure of more advanced stages of movement. O'Brien et al. (1980) have examined microfabric in older sediments and sedimentary rocks and see systematic arrangements even in quite compact and lithified sedimentary deposits.
Cycles in fine sedimentation Hemipelagic deposits may preserve almost complete records of depositional history because they
A review of fine-grained sediment usually (not exclusively) accumulate in areas of low energy. They can therefore record the results of cyclic forces driving deposition. These cyclic periods range from seconds to millions of years (e.g. Fischer & Arthur 1977; Hesse & Chough 1980). Common cycles include seasonal variations, larger term climatic variations ranging from decades to tens of thousands of years, tectonic cycles which can overlap the longer period climatic cycles and pass on to hundreds of millions of years. On a regional basis, local uplifts or depressions, stream capture and channel avulsion are examples of other quasicyclic driving forces. Cycles with periods of 103 yrs and more are usually preserved in bioturbated sediments, but may require careful analysis for their recognition. Resolution of cyclic sedimentation is limited by accumulation rates and bioturbation; records of periods of up to a thousand years or so (depending on accumulation rates) are usually destroyed by bioturbation except where anoxic bottom waters exclude benthos (e.g. Emery & Hulsemann 1962; Soutar & Crill 1977; Malouta et al. 1982; Savrda et al. 1983). Bed thickness is another factor since bioturbation is often most intense only in the top 5-10 cm (Nittrouer & Sternberg 1981) of the accumulating fine sediment, and thick rapidly deposited units may never be entirely bioturbated even in well aerated benthic environments (Savrda et al. 1983). Variations in oxygen content of bottom waters can be rapid events geologically and where accumulation rates are slow, the anoxic accumulations may be bioturbated during the following period of oxygenation. Benthic burrowing activity can be surprisingly intense even at oxygen levels of less than 0.5 ml/L (Savrda et al. 1983). Dean & Arthur (in press) and Arthur et al. (this volume) have described interactions between cycles in the Cretaceous black shales of the deep Atlantic that influence the timing and magnitude of reduction of the sediments and the degree of bioturbation. The change from dysaerobic to anaerobic conditions is apparently triggered by the shorter frequency climatic cycles at times when the longer climatic or tectonic cycles bring the environments close to anoxia. These features represent cyclic periods of 10 4 yrs and more. Human influence on sediment discharges One major problem in the use of recent deposits and mass budgets of contemporary rivers to develop analogs for ancient deposits is the pervasive influence of man's activities on erosion (Meade 1969, 1982). Most workers agree that this effect has been to increase sediment discharges
21
from rivers as a result of overgrazing and other agricultural practices, and the much more recent discharge of wastes and dredge spoil on a large scale (National Research Council 1976). Conversely, as in southern California, many rivers have been truncated in drainage area by flood control dams and spreading basins which trap sediments (Brownlie & Taylor 1981). Estimates of effects of human activities vary by an order of magnitude, but work on the sedimentary accumulations in contemporary offshore southern California basins suggests a two- to three-fold increase in accumulation rates since late Pleistocene time (Nardin 1981; Schwalbach 1982) which may reflect the initiation of human influences superimposed upon the natural climatic and sea-level changes which have also occurred during that time span. Meade (1982) has noted that after such events as intensive agricultural alteration of a drainage area, the resulting sediment is stored within the drainage system and released over time periods of as much as centuries. Bourrouilh (pers. comm. 1983) has noted that the construction of the Aswan Dam on the Nile has generated major problems over the Nile delta. These effects include the influence on natural fertilization of the agricultural lands by cutting off the annual deposition of silts, and changes in habitats of economically important fish species in the areas off the mouth of the Nile.
Controls on fine sediment deposition Tectonics The tectonic setting of the source area or depositional site exerts a first order control on finegrained sedimentation. It affects the rates of uplift and denudation drainage patterns and volumes of river discharge, coastal plain and shelf widths, margin gradients, gross sediment budgets, the morphology of receiving basins and local sealevel changes. The relative activity, or style and frequency of seismicity and faulting, is also of primary importance to sedimentation. It may exceed and mask the effects of the other controls or, in areas of low tectonic activity, play a secondary role. Supply As is true for all sedimentary systems, the interaction of supply and dispersal energy can produce a variety of fine sedimentary deposits, facies and structures for given tectonic and sea-level condi-
22
D.S.
Gorsline
tions (Sloss 1962). As we have seen, the principle fine sediment supply is from river contributions. Initially this is deposited almost entirely in the continental margin as a result of capture, filtering and pelleting processes that have been reviewed. Where this supply is constrained by climatic, oceanographic or bathymetric barriers, the secondary sources become dominant as for biogenie contributions of either carbonate or siliceous composition (Ginsburg & James 1974).
Oceanic dispersal systems Dispersal is achieved by different dominant processes or agents as we progress from shelf to deep ocean floor. Shelves lie within the mixed surface layer of the oceans where strong wave, tide and wind stress currents are active. Although part of the ocean wave spectrum, internal waves should be specifically mentioned here. These have their largest development at density discontinuities in the ocean water column. Cacchione & Southard (1974) have discussed these wave forms and Swift (1973) has discussed their influence on shelves. Slopes receive energy from gravity, large-scale eddy interaction with the substrate and slope currents. Deep ocean floors are affected by tidal currents and large-scale deep slow circulations and possibly by transient stronger flows perhaps associated with large-scale rings and eddies. Where topography reduces the flow cross sections as over gaps in mid-ocean ridges or over and around seamounts, the tidally driven and large water mass circulations are accelerated and can produce subtle to strong substrate effects as noted earlier. In all environments transient effects can be important as in the passage of a turbidity current, or a mass failure or storm effects in shallower waters.
Sea-level As has been noted by Rona (1973), Pitman (1978) and Watts (1982), sea-level position relative to the shelf edge is a major factor in the delivery of continental sediments to the deep ocean. At high sea-levels, decreased stream relief, broad shelves and possibly slower atmospheric and ocean circulation rates (in the case of climatically driven sea-level changes) tend to produce conditions of trapping of sediments on the submerged craton edge. When sea-level falls to or below the shelf edge, delivery is almost entirely to the slopes and the deep-sea floor. Sea-level can be both tectonically and climatically driven (Morner 1974; Sclater et al. 1977) and the oscillations over time will show significant spikes in power spectral analysis that correspond to the two major driving
factors. The tectonic signal will be of the order of 106 yr and the climatic signal will be of the order of 104-105 yr (the Milankovitch cycles). The onset of late Tertiary continental glaciation and the resulting rapid sea-level changes, has been a major factor in world sediment budgets. Laine (1980) has shown that much of the sediment in the North Atlantic abyssal plains probably was delivered during the glacial epochs. This would have required very high sediment discharges from periglacial streams. Emery & Uchupi (1972) have noted the concentration of submarine canyons along the edge of the northeast Atlantic margin of the US and Canada matches the area and latitude of the Pleistocene ice extent. This indicates that at low glacial sea-levels, large quantities of glacial outwash were dumped at the shelf edge and redeposited downslope by turbidity currents and other massmovements.
Biological productivity Biological production occurs over large areas of ocean surface waters where nutrient recharging is active (Eppley & Peterson 1979), therefore the sedimentation patterns will reflect these areal inputs rather than point or linear sources. This differentiation or classification is of interest because it will affect the concentration of particles in the water column and thereby control the accumulation rates on the underlying substrates. Plumes and jets associated with areas of coastal upwellings approximate point sources, and regional coastal upwellings are typically linear sources (Gorsline 1978). Large-scale eddies and rings usually move over trajectories of considerable distance and thus their sedimentation effects are integrated over large areas (Swallow 1976). An exception may occur in the California Current System where the seasonally developed large eddies appear to remain essentially fixed in position (Koblinsky et al. 1983; Simpson et al. 1983). If this pattern has been characteristic of the past few thousand years of roughly stable sealevel position and ocean circulation, then there should be some record of their existence preserved in the sediments. Such eddies can be areas of relatively high productivity (Chelton et al. 1982; Haury 1983).
Biological pelletization Planktonic organisms use several means to filter particulates from the water column, aggregate them and then produce large pellets that sink rapidly to the sea floor (e.g. Osterberg et al. 1964; McCave 1975). The processes include filtering,
A review of fine-grained sediment digestion and excretion (e.g. Osterberg et al. 1964; Frankenberg & Smith 1967); capture on films or extended nets (Gilmer 1972; Bruland & Silver 1981; Silver & Bruland 1981); generation of particulate organic matter from dissolution of bubbles in sea-water (Johnson & Cooke 1980); and large-scale pellets from fish feeding (Robison & Bailey 1981). Some small fraction of fine, dispersed grains do get through to the deep open ocean, but it is probably less than 5% of the total sediment injected from the continents. This is supported by the slow accumulation rate of deep pelagic nonbiogenic sediments. These contain mainly oceanic volcanic material and aeolian material, so the transfer of fine particulates through the coastal and margin systems must be very small. Porter & Robbins (1978) have recognized preserved pellets in older shales and mudstones and point out their close similarity to contemporary copepod faecal pellets. Bioturbation In the deep oceans of the present world, bioturbation is almost ubiquitous except in very restricted basins where biological oxidation demands are high or where sills intercept low oxygen water from the Intermediate Water masses of the world oceans (e.g. California Borderland, Cariaco Trough). The degree of benthic reworking of deep floor substrates is related to the oxygen content of the bottom waters. Bioturbation is limited only by the lowest oxygen content (less than 0.5 ml/L) and some organisms (e.g. nematodes) can continue to live in near-anoxic conditions. These do not completely bioturbate the sediment but leave characteristic fine networks of tiny burrows in the otherwise unreworked primary sedimentary laminations. In the ancient oceans, times of deep ocean anoxic conditions are not uncommon and have been noted above and in other papers in this volume. Dilution In the southern California borderland basins, the trapping effect of the inner basins screens out much of the tcrrigenous supply and so the outer basins are zones of relatively slow sediment accumulation and biogcnic fractions approach half the volume of the fine hemipelagic sediments. Cores in the central and outer basins reveal that the carbonate content decreases during times of glacially lowered sea-level due to augmentcd influx of detritals from exposed banks or the direct delivery of river load to the exposed shelf edge on the adjacent continent. Use of dating
23
methods shows that in actuality, both biogenic accumulation rates and terrigenous rates increased at times of lower sea-level corresponding to times of increased ocean circulation rate. The biogenics increased by perhaps twofold while the terrigenous output sometimes increased almost an order of magnitude (Gorsline & Prensky 1975; Gorsline 1981). Similar patterns have been observed in the deep ocean (e.g. Broecker et al. 1958).
Processes of fine sediment transfer to deep water The discussion will consider the large-scale primary processes of transfer of particulates resulting from water motion and the secondary transfer processes that are related primarily to massmovement processes. Stow & Piper (this volume) discuss sedimentary structures characteristic of various transport processes which complements and extends this discussion. The transport processes include low concentration (less than 10 /agm L-1) nepheloid plumes, high concentration plumes at or near the mouths of large rivers (more than 10 mg L-1), resuspension of fine sediments by strong bottom currents, turbidity currents, flow as matrix of debris flows (high concentration flows; Lowe 1976, 1979, 1982), and by massmovement of various types (Nardin et al. 1979a). These processes can be grouped into those operating throughout the water column, which are generally continuous although with seasonal and longer term variations, and those that occur along the bottom, which are usually discontinuous and discrete events including turbidity currents and other mass-movements (Gorsline & Emery 1959). The continuous processes usually involve large volumes but contain low concentrations per unit volume and thus produce relatively slow accumulation rates at any single locality. The discontinuous processes involve smaller water volumes, but because of their high concentrations and high input rates, produce much larger accumulation rates at a given point. Viscous boundary layer effects are likely to be preserved as sedimentary structures in the latter (Hesse & Chough 1980; Stow & Bowen 1980); the influence of boundary layer processes on the low concentration continuous depositional processes may appear as regional textural gradients in the surficial sediments (McCave & Swift 1976). Plumes and nepheloid-layers The discharges from the world's large rivers typically exit from the seaward mouths of deltaic
24
D.S.
distributaries as turbid plumes (Meade et al. 1975; Coleman 1976; Gibbs 1976; Coleman et al. 1981). Seasonal variations may be very important and long term climatic cycles that affect run-off will exert a major effect. The turbid plume interacts with the shelf circulation. In some instances, the season of flooding may be contemporaneous with shelf current patterns that are different from those of other seasons. This is the case off the Oregon-Washington coast where the floods typically occur at times when the shelf circulation moves coastal waters to the north in contrast to the mean annual flow to the south. Discharge via deltaic and estuarine systems is typically as flocculates resulting from ion exchange between the clays and saline ocean waters. High particle concentrations and biological factors facilitate the flocculating process. Combinations of tidal exchange and salinity gradient processes tend to hold the suspended fine particulates in the head of the estuaries, producing turbidity maxima in or near the zone of maximum salinity gradient (Postma 1967). Tidal effects interacting with river flow can alter this position to some degree (Allen et al. 1980). When fine particulates enter the ocean via rivers, coastal erosion, or airborne infall, they slowly fall through the water column. Since aggregation occurs by various processes, the initial particle size distribution is rapidly altered (see Kranck 1975; Paffenhofer et al. 1979). Depending on the bulk density of the aggregates, they may fall much faster than the discrete fine particles. It should be remembered that biologically aggregated particles are attractive sites for bacteriological colonization. Bacterial degradation of biogenic pellets can be well advanced after a few days. As these particles fall, therefore, they are increasingly broken down and become less dense in bulk with concommitant decrease in settling velocity (Pomeroy & Diebel 1980; Pomeroy, in press). Secondary aggregation near-bottom can be effective when near-bottom pelagic populations of pelagic filter feeders are present as in the example of benthopelagic holothuroids in deep waters of California Borderland Basins (Carney, pers. comm. 1983). The fall velocity depends on water viscosity also (viscosity varies inversely with temperature). In deeper colder waters, the velocity will decrease. In addition, the ocean is a density stratified system with many density discontinuities. Each of these tends to slow particle fall because of turbulence at the boundary. As a result, particle concentration in the ocean is non-uniform with depth as was noted by Kalle (1939) and as further discussed by Jerlov (1953~ 1968). Their work,
Gorsline
based primarily on optical methods, has been verified by later workers using both optical and direct sampling methods (see Eittrem et al. 1969; Drake 1972). Rodolfo (1964, 1970) used water sampling and filtration in continental margin waters to define these zonations. He noted a surface turbid layer most strongly developed near the coast and river sources (Manheim et al. 1972; Meade et al. 1975), mid-water layers, and a bottom layer. These zones are typically layers with particle concentrations of/~gm/L and maximum concentrations of a few mg/L near the months of rivers, and are commonly known as nepheloid-layers (Eittrem et al. 1969; Ewing & Connary 1970). They are now recognized to be an important pathway for much of the fine open ocean particulate transport. The bottom nepheloid-layer may be up to 2000 m in thickness although the highest concentration is typically within a few tens of metres of the bottom. Larger aggregates probably pass rapidly through these nepheloid-layers to the bottom, although bacterial degradation of pellets falling through the water column can increase the layer concentration (Honjo et al. 1982; Pomeroy, in press). In the Pacific (Ewing & Connary 1970), the bottom nepheloid-layers appear to be moved in large gyres within the major deep basins of that ocean and draw much of their lithic fraction from the western continental margins of North America. Variations in thickness may in part result from bottom erosion. Recent work in the eastern tropical Pacific (Lonsdale 1976; Heath et al. 1976) suggests that winnowing occurs in narrow passages between basins, with the subsequent transfer of suspended particulate material to the bottom nepheloid-layers of the deeper basins. Thus, some of the fine material deposited in one depression may have its source from an adjacent basin or sill. In any event, a conservative estimate of the total suspended load in ocean water at any given moment is of the order of 10 9 tons (Lal 1977; calculations by Gorsline). McCave (1982) has noted that the fine particulates must pass into the benthic boundary layer at the base of flows. As the bottom shear increases these small scale processes become increasingly important at given concentrations. When this layer is developed the processes within the layer are both erosional and depositional on a micro scale. Thus effects of even weak bottom flow may be subtlely recorded in the sediments deposited.
Resuspension by bottom currents Bottom currents can resuspend and sort fine deposits and particles at all depths in the ocean (Heezen et al. 1966). High velocity flows are able
A review of.fine-grained sediment
to shape large-scale bedforms and act as an agent of major fine sediment winnowing (Lonsdale & Malfait 1974). More subtle textural and grain orientation properties of the sediments can reveal data about flow direction and winnowing in the silt and clay particle size ranges by lower velocity currents (Johnson et al. 1977; Ledbetter & Ellwood 1980). Drake & Gorsline (1973) have reported observations in California submarine canyons that suggest periodic reworking of canyon sediments. Their studies were made at the same time as near-bottom current-metre measurements by Shepard and his associates (Shepard et al. 1979), which show that near-bottom tidally generated water motions can resuspend fine sediment and move it progressively up or down canyon. Karl (1976, 1981) has given numerous examples of the reworking of shelf substrates by bottom currents in shallow shelf waters off southern California. His work also supports the laboratory studies by Cacchione & Southard (1974) in which the strong water motions are associated with shoaling internal waves. These typically concentrate at the base of the thermocline. Work by J. D. Smith and his colleagues at the University of Washington has modelled the influence of bottom current shear on fine sediments on shelves and predicted the depth of reworking and the resulting textural changes (e.g. Jumars et al. 1981; Nowell et al. 1981;). Nittrouer & Sternberg (1981) have also noted the reworking of fine mid-shelf mud belts associated with the Columbia River discharge and the progressive winnowing of fines over a period of decades (also see Baker 1976). Larsen (1982) has discussed the influence of seaward directed dispersive flow generated by wave groups and longer waves working on the coast as a mechanism for reworking and seaward transfer of fine sediments. McCave (1982) and McCave et al. (1982) have reviewed the results of the high energy benthic boundary layer experiment (HEBBLE) studies on the deep continental rise off the Atlantic margin of the US and showed the influence of transient strong bottom currents on the substrate. Yingst & Aller (1982) have described the effects of bioturbation on the same substrates. Thus both biologic and physical processes may overprint to produce a final sedimentary depositional structure assemblage. The large scale eddies and rings associated with Gulf Stream turbulence can reach to the deep-ocean floor, at least at inception of the eddy or ring, and may be the source of the deep water transient flow events noted in the HEBBLE area.
25
Turbidity currents
Turbidity currents are turbulent, relatively high concentration flows (Nardin et al. 1979b; Middleton & Bouma 1973) that require a high mud content to carry the sands and gravels characteristic of the early phases of this mechanism. As the coarser fractions settle out leaving distinctive structures and facies as a record of flow conditions (Bouma 1962), the remaining fine suspensates are laid down as distal turbidites on the lower slope or fan and abyssal floor, or in overbank deposits away from the conducting channels (Walker 1967; Haner 1971). These fine turbidites are typically finely laminated, graded, homogeneous or combinations of these features (Piper 1978; Stow 1979; Stow & Bowen 1980; Hesse & Chough 1980; Thornton 1981a). As contemporary turbidite basin deposits have been more closely studied, it is evident that thicker layers (up to several metres thick) are also common. Stanley (1981) and Stanley & Maldonado (1981) have called these massive layers unifites. They apparently represent rapid deposition from large muddy turbidity flows generated by slumping of unstable slope sediments. Thornton (1981 a) has described smaller examples from Santa Barbara Basin in the California Borderland, that contain transported microfauna from shallower slope depths. Since the recognition of the subtle grading and occasional faint lamination requires X-radiography and detailed textural analysis, these have probably been missed in the stratigraphic record. It is likely that turbidity current-deposited fine sediments are much more common in shales than previously supposed. Bioturbation may erase the thin-bedded forms, but the thicker units (more than 10 cm) will probably survive biologic stirring. Debris-flows and other high concentration flows
In a theoretical sequence, canyons store sediment from shelf and coastal sources and periodically release masses of the stored sediment in response to a variety of triggering forces (see Gorsline 1980). The sequence of processes that results is slides and slumps, progressing to high concentration flows and debris-flows and then turbidity currents. Changes in slope, concentration, channelization and textural composition produce a variety of facies and sedimentary structures (e.g. Mutti & Ricci Lucchi 1978; Nardin et al. 1979; Hein 1982, in press). Deposition of sands and gravels from high concentration flows and of pebbly muds from debris-flows is probably rapid. As the mass velocity slows to some critical level, the coarse mass 'freezes' as a slug.
26
D.S.
Gorsline
mud belts. The shift in position will be the response to the interaction of the two main Slope sediments are strongly affected by mass- variables (supply and dispersal) at given sea-level movements (Dott 1963; Emery & Uchupi 1972; positions. Inner shelf mud belts are exemplified by the Jacobi 1976; Embley 1976; Haner & Gorsline 1978; Nardin et al. 1979; Damuth 1980). This is Amazon muds that deposit along the Surinam recorded in the geologic record and in acoustic coast (Wells & Coleman 1981). These form seismic profiles in the contemporary oceans by aggrading belts that cause a general progradation contorted and sheared bedding, and by smaller of the coast. High mud supply is the basic factor scale evidence of displacement based on displaced (also see Augustinas 1978). The muds are of low faunal remains and textural anomalies (Douglas bulk density initially and actually damp out 1981). The magnitude of the material lost from a incoming wave energy. The cusps of mud accregiven slope or palaeoslope surface in a sedimen- tionary deposits have wave lengths of about 100 tary section can be determined by study of the km. Mid-shelf mud belts (of the order of 100 km bulk properties of the section. If a zone is overcompacted as compared to immediately long) have been studied offthe Washington coast, overlying beds, or if surface sediments are over- the New England shelf and off the Eel-Klamath compacted, estimates can be made of the unload- Rivers of northern California (Emery & Uchupi ing that has occurred (Booth 1979). Almagor 1972; Baker 1976; Field 1981a, respectively). All (1976) has shown that a regional slope profile can three areas represent dynamic accumulations be drastically modified by large-scale mass-move- where seasonal flood discharge of fine particuments (also see Emery & Uchupi 1972; Seibold & lates is greater than the capacity of shelf dispersal Hinz 1973). Mass-movement is a major factor processes. Offthe Columbia and the Eel-Klamath volumetrically in continental slope transport and Rivers, the flood stages come at times when the shelf current set is to the north and thus both mud is particularly characteristic of fine sediments. A variety of triggering forces can generate belts trend north from the source river mouths. mass-movement including earthquake shocks, The fines are constantly recycling through the sudden or rapid loading, undermining or under- deposits with periods of a few decades (Nittrouer cutting, and gas charging. Several writers have et al. 1979; Nittrouer & Sternberg 1981). The considered the influence of benthic sediment Washington deposit is graded with depth as a reworking on sediment strength. Richardson & result of the reworking of the accumulating silts Young (1980) have found that bioturbation can and clays over time and the basal, somewhat either increase or decrease strength, whereas coarser particles probably represent an equilibDrake (1976) suggested that bioturbation de- rium residuum of the process. The east coast belt creases cohesion and strength. Modelling studies probably has its source as the fine sediments of of the erodibility of substrates on shelves and the the Bay of Fundy and Gulf of Maine rather than a effects of burrowing organisms by Nowell et al. discrete river source. This deposit appears to be in (1981) and Jumars et al. (1981) showed the dynamic equilibrium. On many shelves, small mud patches are found theoretical effects on strength of a range of biological influences, and predict the resulting of scales an order of magnitude or more smaller substrate sensitivity, erosion and redeposition in than the mid-shelf mud belts discussed above. On response to strong near-bottom fluid stress as the west coast of the US the shelf systems are probably sufficiently close to dynamic equilibrium during storms. so that major changes in shelf sediment character can occur when the systems are perturbed. An example is the effect of the 1969-70 floods in the Depositional sites for fine southern California area which delivered the sediments in the ocean largest sediment discharges since 1938 (Curtis et al. 1973). These discharges overwhelmed the Shelves dispersal processes and the result was a mud Fine sediments can form appreciable deposits on deposit on the shelf that was over 30 cm thick shelves and shallow marine platforms when the inshore (Kolpack 1971; Drake et al. 1972) and supply exceeds the dispersal rate (Sloss 1962; required almost two years of normal wave reworkMcCave t972). Stanley et al. (1983) have dis- ing to remove. This shelf is normally sandy with an cussed the mudline as a textural marker that outer silty sand zone and occasional compact mud demarcates the positions of mud belts on shelves. patches a few hundred square metres in area. The Several scenarios have been examined: coastal reworking of the shelf deposits had a secondary mud belts, mid-shelf mud belts and shelf edge effect on the adjacent basin sedimentation which
Mass-movement: slumps and slides
A review o f f i n e - g r a i n e d s e d i m e n t has been discussed by Fleischer (1970b), Kolpack (1971) and Drake (1972). Outer shelf mud belts are also common. Off southern California, in the San Pedro area, Gorsline & Grant (1972) have described an outer shelf belt of very fine silty sand that is bioturbated and appears to be accumulating. Recent work by Drake et al. (in press) gives dynamic support to this since they show that the seasonal shift in the threshold depth for wave stirring of fine sediments essentially bounds the outer shelf fine silty belt. That some reworking by low velocity currents does occur is shown in profiles of water turbidity collected over a multi-year period by Karl (1976) in which increased turbidity at and above the bottom is seen at the depth of intersection of the local seasonal thermocline and the substrate. This may be the result of breaking internal waves striking the shelf at those depths. The dispersion of wave energy seaward and return flow generated by build-up against the coast by onshore wind stresses can be the sources of offshore net motion of fine particulates once they have been stirred into the water column by storm wave surge of tidal currents (Larsen 1982; Drake et al., in press). Slopes
Fine sediments moving off the shelf are typically deposited on the adjacent slopes. Both along slope currents (Drake, pers. comm.) and biological pelletization tend to favour their accumulation on the slope. Haner & Gorsline (1979) have reported that the slopes with most mass-movement activity off southern California are those that are in the path of turbid plumes passing seaward off promontories and from gyres centred over broad shelf segments (also see Davis 1980; Thornton 1981b). Those that are not so affected are typically the areas of highest stability, least mass-movement and slowest sediment accumulation rate. Similar convergences of nepheloid plumes and slopes along other coasts must determine the sites of highest accumulation rates and least stability. The extreme end member of the possible series of slope sedimentation types is the slope off the mouth of a major delta distributary. In these areas (e.g. Mississippi Delta, Coleman 1976) the slope sediments are deposited on gentle gradients at very high rates and creep and fluid failures are common. As discussed earlier, mass-movement and gravity influence dominates this depositional environment. Where sediment accumulations are large, the scale of the mass-movements also increases as witnessed by the large-scale failures on the Atlan-
27
tic margin of the US described by Emery & Uchupi (1972) in their comprehensive review of that margin. Such mass-movements must contribute a large amount of sediment to continental rises on passive margins and to trench floors on active margins. In the California Borderland, a transform margin, the basins of the northern inner borderland in the zone of highest fine sedimentation (Gorsline 1981; Schwalbach 1982) typically have slopes that are dominated by mass-movement features (Gorsline 1978; Nardin et al. 1979a). Deep sea and basin floors
The floors of marginal basins, trenches and abyssal plains are the end of the slope gradient from continental platform to deep ocean floor and, as such, represent the ultimate sediment sink in the oceans. Reworking and sediment transfer can still take place, particularly in gaps and sills in seamounts and oceanic ridges (Lonsdale & Malfait 1974; Chamley 1975; Diester-Haas 1975). In small marginal basins, sea-level position is an important control on sedimentation rate in general (Rona 1973; Pitman 1978; Watts 1982). This was also true of the Atlantic in Late Neogene times (Laine 1980) when glacial influx was a major source for the construction of the western abyssal plains and the North American slope and rise areas. In Pacific type oceans with active margins, the deep basins are starved and receive mainly hemipelagic sediments plus locally derived turbidity current and mass-movement deposits. In marginal basins, trenches and Atlantic type abyssal plains, turbidity current deposition is a major factor together with continuous nepheloid-layer transport. Bennetts & Pilkey (1976), Bornhold & Pilkey (1971) and Malouta et al. (1981) have described the area and volumes of a number of recent turbidites in Atlantic abyssal plains and marginal basin regimes. Elmore et al. (1979) described a turbidite from Hatteras Abyssal Plain that extended over a distance of at least 500 km. All of these turbidites involve volumes of sediment that range from 108 to 101~m 3, and that represent very large initial sediment accumulations in the source areas, although the resulting deposits are thin turbidites. Thus the important characteristic of this environment is an area which is large compared to bed thicknesses that are relatively small (centimetre range). As Ewing & Connary (1970) have stated, the nepheloid plumes in deep ocean basin plains are controlled by the deep-water circulation. As a result of earth rotation and tidal influences these deep circulations are gyres and thus suspension
28
D.S. Gorsline
transport is not directly to the basin centre but rather slope-parallel with much longer travel paths than a simple downslope flow over the gentle basin slopes. Such circulations could lead to radial transfer of suspension load to the low velocity centre of the basin-centred gyre with slightly higher sedimentation rates in mid-basin. Some limited evidence of this has been accumulated for small margin basins (Malouta et al. 1981). Isopach maps of Holocene basin sedimentation in the basins of the California Borderland show similar thicker accumulations in mid-basin rather than in the areas of associated fans (Schwalbach 1982). This reflects the low level of depositional activity in fans in the present time of high sea-level and also the probable influence of deep-basin water circulation on nepheloid transport. Data reported by Drake (1972) for Santa Barbara Basin in the borderland also suggests the presence of gyral flow in the deep-basin water as well as some possible long wave effects perhaps resulting from tidal resonance. Large-scale rings and eddies associated with large ocean boundary current turbulence and instability may affect the deep-sea floor. As shown in data from Koblinsky et al. (1983) the rings have diameters of 100-200 km and extend to depths of at least 1500 m (the maximum sampling depth) in the California Current System. The Gulf Stream eddies and rings may extend to the sea floor at times of formation and since many of these progress into the western north Atlantic, they may have an important effect on deep-sea floor substrates. The effect may be subtle and probably requires careful textural analysis together with grain orientation studies and represents an interesting problem for study. Fine-grained biogenic sediments are an important facies in pelagic sediments of the contemporary deep-sea floors. They occur in areas of high productivity, low terrigenous influx and relatively slow accumulation rate compared to ancient biogenous sediments (less than 5 cm/1000 yr versus 15 cm/1000 yr; Hakansson et al. 1974). The chalks and diatomites which are widely distributed in some parts of the geologic record appear to have been shelf or platform deposits (Hakansson et al. 1974; Surlyk & Birkelund 1979; Bottjer, pers. comm.). For example, the extensive chalks of the Cretaceous were deposited in water depths of less than 200-300 m. Diatomites represent times of high nutrient supply whereas the chalks formed at times of relatively reduced nutrient supply. Both are formed in starved environments where other factors have screened out the continental detrital contribution. Analysis of chalk stratigraphy has produced evidence for cycles with periods of the order of 104 yr that may match
the climatic cycles resulting from inequalities of solar insolation at high latitudes related to variation in the Earth's rotation, axial wobble and precession (Milankovitch cycles).
Conclusions It is obvious when looking at the broad spectrum of fine-grained sediment research, that it is an area of sedimentary petrology that requires interactions with several other disciplines. At least since late Palaeozoic time, most fine terrigenous particles have probably been worked upon by organisms either in transit or after deposition. Since fine sediments move in suspension for at least part of their transport history, oceanography is a necessary consideration in any palaeogeographic reconstruction of ancient depositional environments. Since fine sediments are associated with organic matter, partly due to similar transport pathways and also due to absorption of organics on clay flake surfaces, fine sediments are hosts to biotas that feed on this food source. Biological activity, including bacterial processes, alters the chemical environment after deposition and is a major factor in diagenetic change. Thus, microbiology and marine biology are important data sources for the stratigrapher and sedimentologist. Bioturbation can erase the primary physical structures that diagnose transport processes and so require very detailed analysis of mudstone and shale sections to see the evidence of textural or compositional differences that signal different processes such as pelagic infall versus fine turbidite deposition. It is evident that associations of organic matter and fine sediment are not restricted to a given ocean condition; e.g. upwelling in the classic sense. Rather, the organic content may be more a function of dilution by terrigenous input rather than high productivity (Gorsline 1981). The work to date and much of the work reported in this volume shows that the fine sedimentary record must be extensively restudied. Much of the shale record may actually be composed of small scale mass-movement (creep) and fine turbidites. Reworking may be much more extensive and it will require careful and very detailed analysis to define small hiatuses. Micropalaeontologists (e.g. Douglas 1981) are discovering that microfaunal remains are much more mobile than might be thought in the absence of evidence of large-scale turbidity current flow or large mass-movements. Thus many supposed hemipelagic records may contain mixed fossil assemblages in what may appear to be pelagic
A review of fine-grained sediment infall a c c u m u l a t i o n with no evidence of lateral transport. W h e n fine sediment depositional records are examined in detail, cyclicity is evident at a variety o f scales. This is an i m p o r t a n t research area in which m u c h w o r k needs to be done. W h e n fine sediments are deposited in anoxic environments, the fine detail of cycles as small as seasons can often be deciphered from the preserved primary structures. The bulk of the geologic record is fine-grained sediment. Therefore most of geologic history is in this part of the section. We have only begun to read this record.
29
ACKNOWLEDGEMENTS; M u c h of the research u p o n which this discussion is based has been sponsored by various grants over the past 10 years from the N a t i o n a l Science F o u n d a t i o n . Their continued support is most gratefully acknowledged. Drs. R. Bourrouilh, R. D u t t o n , R. Douglas, D. D r a k e and S. T h o r n t o n all read one or more versions of this paper and their c o m m e n t s were most helpful. Drs. D. Stanley, J. Syvitski and D. Stow all read earlier versions and also m a d e useful comments. The degree of incompleteness or u n d e r s t a n d i n g is solely the responsibility of the author.
References ALLEN, G.P., SALOMON, J.C., BASSCULLET, P., DU PENHOAT,Y. and DE GRANDPRE,C. 1980. Effects of the tide on mixing and suspended sediment transport in macrotidal estuaries. Sed. Geol., 25, 51-68. ALMAGOR, G. 1976. Physical properties, consolidation processes and slumping in recent marine sediments in the Mediterranean continental margin slope of southern Israel. Geol. Survey Israel, Rep., MG/76/4, 1 3 2 pp. - 1982. Marine geotechnical studies at continental margins: a review--Part II. Appl. Ocean Res., 4, 130-50. AUGUSTINAS,P.G.E.F. 1978. The Changing Shoreline of Surinam (South America). Unpub. Ph.D. Dissertation, Univ. of Utrecht, Netherlands. 232 pp. BAKER,E.T. 1976. Distribution, composition and transport of suspended particulate matter in the vicinity of Willipa Submarine Canyon, Washington. Bull. geol. Soc. Am. Bull., 87, 625-32. BENNETT, R.H., BRYANT,W.R. & KELLER,G.H. 1981. Clay fabric of selected submarine sediments: fundamental properties and models. J. seal. Petrol., 51, 217-32. - & NELSON,T.A. 1983. Sea floor characteristics and dynamics affecting geotechnical properties at shelf breaks. In: Stanley, D.J. & Moore, G.T. (eds) The Shelf Break: Critical Interface on Continental Margins. Soc econ paleo. Min Spec. Pub., 3 3 . BENNETTS, K.R.W. & PILKEY, O.H. 1976. Characteristics of three turbidites, Hispaniola-Caicos Basin. Bull. geol. Soc. Am., 87, 1291-300. BERGER,W.H. 1974. Deep-sea sedimentation. In." Burk, C.A. & Drake, C.L. (eds), The Geology of the Continental Margins. Springer-Verlag, Berlin. 213-41. BISCAYE, P.E. 1965. Mineralogy and sedimentation of recent deep-sea clay in the Atlantic Ocean and adjacent seas. Bull. geol. Soc. Am., 76, 803-32. BLANTON, J.O., ATKINSON, L.P., PIETRAFESA, L.J. & LEE, T.N. 1981. The intrusion of Gulf Stream Water across the continental shelf due to topographicallyinduced upwelling. Deep Sea Res., 28, 393-405. BLATT, H. 1982. Sedimentary Petrology'. W.H. Freeman & Co., San Francisco. 564 pp.
BOOTH,J.S. 1979. Recent history of mass wasting on the upper continental slopes, northern Gulf of Mexico as interpreted from the consolidation states of the sediment. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of the Continental Slope. Soc. econ. Paleo. Min. Spec. Pub. 27, 153-64. BORNHOLD, B.D. & PILKEY, O.H. 1971. Bioclastic turbidite sedimentation in Columbus Basin, Bahamas. Bull. geol. Soc. Am., 82, 1341-54. BOUMA, A.H. 1962. Sedimentology of some Flysche Deposits. Elsevier, Amsterdam. 168 pp. BROECKER, W.S. 1974. Chemical Oceanography. Harcourt Brace Jovanovich Inc., New York. 214 pp. BROECKER, W.S., TUREKIAN, K.K. & HEEZEN, B.C. 1958. The relation of deep-sea sedimentation rates to variations in climate. Am. J. Sci., 256, 503-1~/. BROWNLIE, W. R. t~ TAYLOR, B.D. 1981. Sediment management for southern California Mountains, coastal plains and shorelines; Part C--Coastal sediment delivery by major rivers in southern California. Calif. Inst. of Technology, Report EQL 17C, Pasadena. 314 pp. BRULANO, K.W. • SILVER,M.W. 1981. Sinking rates of fecal pellets from gelatinous zooplankton (salps. pteropods, doliolids). Marine Biol., 63, 295-300. CACCHIONE, D.A. & SOUTHARD,J.B. 1974. Incipient sediment movement by shoaling internal gravity waves. J. geophys. Res., 70, 2237-42. CHAMLEY,H. 1975. Influence des courants profonds au large du Bresil sur la sedimentation argileuse recente. IXth International Sedimentological Congress, Nice, Resumes des Publications, VIII, 1. CHELTON,D.B., BERNAL,P.A. & McGOWAN, J.A. 1982. Large-scale interannual physical and biological interaction in the California Current. J. mar. Res., 40, 1095-125. COLEMAN, J.M. 1976. Deltas - Processes of Deposition and Models for Exploration. Continuing Education Publications, Inc., P.O. Box 2590 Sta. A, Champagne, Illinois. 102 pp. -8~ GARRISON, L.E. 1977. Geological aspects of marine slope stability, northwestern Gulf of Mexico. Marine Geotech., 2, 9-44. --, ROBERTS, H.H. MURRAY, S.P. & SALAMA, M.
3o
D.S. Gorsline
characterization of grain shape. J. sed. Petrol., 40, 205-12. Err'rREM, S.L., EWING, M. & THORND1KE, E.M. 1969. Suspended matter along the continental margin of the North American basin. Deep Sea Res., 16, 613-24. ELMORE, R.D., PILKEY,O.H., CLEARY,W.J. & CURRAN, H.A. 1979. Black Shell Turbidite, Hatteras Abyssal Plain, western North Atlantic Ocean. Bullgeol. Soc. A m . , 90, 1165-76. EMBLEY, R.W. 1976. New evidence for the occurrence of debris flow deposits in the deep sea. Geology, 4, 371-74. EMERY, K.O. & HULSEMANN,J. 1962. The relationships of sediments, life and water in a marine basin. Deep Sea Res., 8, 165-88. 38, 51-76. & UCHUVl, E. 1972. Western North Atlantic DAVIS, C.C. 1980. LANDSA T Image Analysis of CircuOcean: topography, rocks, structure, water, life and sediments. Am. Ass. Petrol. Geol. Mere., 17, 532 pp. lation and Suspended Sediment Transport; California Continental Borderland. Unpub. M.S. Thesis, Univ. EPPLEY, R.W. & PETERSON, B.J. 1979. Particulate So. Calif., Los Angeles. 216 pp. organic matter flux and planktonic new production DEAN, W.E. & AgTI-IUR, M.A., in press. Origin and in the deep ocean. Nature, 202, 677-80. diagenesis of Cretaceous deep-sea black shales and EWlNG, M. & CONNARY, S.D. 1970. Nepheloid layer in multicoloured claystones in the Atlantic Ocean. US the North Pacific. Geol. Soc. Am. Mem., 126, 41-82. geol. Surv. Prof. Pap. FIELD, M.E. 1981a. Deposition on Pacific shelf DEMAlSON, G.J. & MOORE, G.T. 1980. Anoxic environe d g e - zone of constrasts. Bull. Am. Ass. Petrol. ments and oil source bed genesis. Bull. Am. Ass. Geol. Bull., 65, 924. Petrol. Geol., 64, 1179-209. FIELD, M.E. 1981b. Sediment mass-transport in basins: DIESTER-HAAS, L. 1975. Influence of deep oceanic controls and patterns. In: Depositional Systems of currents on calcareous sands off Brazil. IXth InterActive Continental Margin Basins. Pacific Section national Sedimentological Congress, Nice, Resumes Soc. econ. Paleo. Min Short Course, 61-84. des Publications, VIII, 2. FISCHER, A.G. & ARTHUR, M.A. 1977. Secular variaDox'r, R.H. 1963. Dynamics of subaqueous gravity tions in the pelagic realm. Deep-water carbonate depositional processes. Bull Am. Ass. Petrol. Geol., environments, Sue. econ. Paleo. Min. Spec. Pub., 25, 47, 104-29. 19-50. DOUGLAS, R.G. 1981. Palaeoecology of Continental FLEISCHER, P. 1970b. Mineralogy and sedimentation Margin Basins: a modem case history from the history, Santa Barbara Basin, California. J. sed. Borderland of Southern California. In: Depositional Petrol., 42, 49-58. Systems of A ctive Continental Margin Basins. Pacific FRANKENBERG,D. & SNXTH,K.L. 1967. Coprophagy in marine animals. Limnol. Oceanogr., 12, 443-50. Section, Soc. econ. Paleo. Min. Short Course, 121-56. FUrrERER, D.K. 1974. Significance of the boring sponge DRAKE, D.E. 1972. Distribution and Transport of SusCliona for the origin of fine-grained materials of carbonate sediments. J. sed. Petrol., 44, 79-84. pended Matter, Santa Barbara Channel, California. Unpubl. Ph.D. Dissertation, Univ. So. Calif. Los GARRELS, R.M. & MACKENZIE, F.T. 1971. Evolution of Sedimentary Rocks. W.W. Norton & Co., Inc., New Angeles. 358 pp. York. 397 pp. - 1976. Suspended sediment transport and mud deposition on continental shelves. In: Stanley, D.J. GIBBS, R.J. 1967. The geochemistry of the Amazon & Swift, D.J.P. (eds), Marine Sediment Transport River system- Part I, the factors that control the salinity and the composition and concentration of and Environmental Management. John Wiley, New York. 127-58. the suspended solids. Bull. geol. Sue. Am., 7 8 , 1203-32. - & GORSLINE, D.S. 1973. Distribution and transport of suspended particulate matter in Hueneme, - 1976. Distribution and transport of suspended particulate material of the Amazon River in the Redondo, Newport and La Jolla Submarine ocean. In: Wiley, M. (ed.), Estuarine Processes. Canyons, California. Bull. geol. Soc. Am., 84, Academic Press, New York. 35-47. 3949-68. DUNSTAN, W.M. & ATKINSON, L.P. 1976. Sources of - 1981. Sites of river-derived sedimentation in the ocean. Geology, 9, 77-80. new nitrogen for the South Atlantic Bight. In." & HEL'rZEL, S. 1982. Coagulation and the deposiWiley, M. (ed.) Estuarine Processes. Academic - Press, New York. 69-78. tion of mud. Abstracts Vol., XIth Intl. Sed. Congress, EHRLICH, R., BROWN, P.J., YARUS, J.M. & PRZYGOCKI, Hamilton, 4. GILMER, R.W. 1972. Free-floating mucus webs: a novel R.S. 1980. The origin of shape frequency distribufeeding adaptation for the open ocean. Science, 176, tion and the relationship between size and shape. J. sed. Petrol., 50, 475-84. 1239-40. - & WEINBERG, B. 1970. An exact method for GINSBURG, R.N. & JAMES, N.P. 1974. Holocene car1981. Morphology and dynamic sedimentology of the eastern Nile Delta shelf. In: Nittrouer, C.A. (ed.), Sedimentary Dynamics of Continental Shelves. Developments in Sedimentology. Elsevier. 301-26. COOK, D.O. & GORSLINE,D.S. 1972. Field observations of sand transport by shoaling waves. Marine Geol., 13, 31-55. CURRAY, J.R. & MOORE, D.G. 1971. Growth of the Bengal Deep-sea Fan and denudation in the Himalayas. Bull. geol. Soc. Am., 82, 563-72. CURTIS,W.F., CULBERTSON,J.K. & CHASE,E.B. 1973. Fluvial-sediment discharge to the conterminous United States. US geol. Surv. Circular, 670, 17 pp. DAMUTH, J.E. 1980. Use of high-frequency (3.5-12 kHz) echograms in the study of near-bottom sedimentation processes in the deep-sea. Marine Geol.,
A review of fine-grained sediment bonate sediments of continental shelves. In: Burk, C.A. & Drake, C.L. (eds), The Geology of Continental Margins, Springer Verlag, Berlin. 137-55. GORSUNE, D.S. 1978. Anatomy of margin basins. J. sed. Petrol., 48, 1055-68. - 1980. Deep-water sedimentologic conditions and models. Marine Geol., 38, 1-21. - 1981. Fine sediment transport and deposition in active margin basins. In: Depositional Systems of Active Margin Basins, Pacific Section Soc econ Paleo Min. Short Course, 39-60. - & EMERY,K.O. 1959. Turbidity current deposits in San Pedro and Santa Monica Basins off southern California. Bull. geol. Soc. Am., 70, 279-88. & GRANT, D.J. 1972. Sediment textural patterns on San Pedro Shelf, California (1951-1971): reworking and transport by waves and currents. In: Swift, D.J.P., Duane, D.B. and Pilkey, O.H. (eds), Shelf Sediment Transport. Dowden, Hutchinson & Ross, Stroudsburg, Pa., 575-600. - - & PRENSKY, S.E. 1975. Palaeoclimatic inferences for Late Pleistocene and Holocene from California Borderland basin sediments. In: Suggate, R.P. Cresswell, M.M. (eds), Quaternary Studies. Royal Soc. New Zealand. 147-53. GRIFFIN, J.J. & GOLDBERG, E.D. 1968. Clay mineral distribution in the Pacific Ocean. The Sea, 3, 728-4 1. HAKANSSON, E., BROMLEY, R . & PERcH-NIELSEN, K. 1974. Maastrichtian chalk of northwest E u r o p e - a pelagic shelf sediment. Spec. Pubs. lnternl. Assoc. Sed., 1,211-33. HA~R, B.E. 1971. Morphology and sediments of Redondo Submarine Fan, southern California. Bull. geol. Soc. Am., 82, 2413-32. -& GORSLINE, D.S. 1979. Processes and morphology of continental slope between Santa Monica and Dume Submarine Canyons, Southern California. Marine Geol., 28, 77-87. HAURV, L.R. 1983. An offshore eddy in the California Current System. Part IV: plankton distributions. Progress in Oceanography, in press. HEATH, G.R., DAUPHIN, J.P., OPDYKE, N.P. & MOORE, T.C. 1973. Distribution of quartz, opal, calcium carbonate and organic carbon in Holocene, 600,000 yr and Brunhes/Matuyama age sediments of the North Pacific. Geol. Soc. Am. Abstracts with Programs, 5, 662. --, MOOSE, T.C. & DAUPHIN, J.P. 1976. Late Quaternary accumulation rates of opal, quartz, organic carbon and calcium carbonate in the Cascadia Basin, northeast Pacific. Investigations of Late Quaternary oceanography and Paleoclimatology, Geol. Soc. Am. Mem., 145, 393-410. HEEZEN, B.C., HOLLISTER,C. & RUDDIMAN,W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science, 152, 502-8. HE1N, F.J. 1982. Depositional mechanisms of deep-sea coarse clastic sediments, Cap Enrage Formation, Quebec. Can. J. Earth Sci., 19, 267-87. --, in press. Conglomeratic deposits of braided channel systems: deep-sea and fluvial settings. In." Gravels and Conglomerates, Internl. Assoc. Sed. Mem. --, in press. Fine-grained deep-sea mass flow depo-
31
sits, sedimentary facies, mass physical properties and depositional mechanisms, offshore California Continental Borderland. J. sed. Petrol. -& RISK, M.J. 1975. Bioerosion of coral heads: inner patch reefs, Florida reef tract. Bull. Marine Sci., 25, 133-8. HESSE, R. & CHOUGH, S.K. 1980. The northwest Atlantic mid-ocean channel of the Labrador Sea. II. Deposition of parallel laminated levee muds from the viscous sublayer of the low density turbidity currents. Sedimentology, 27, 697-711. HONJO, S., MANGANIN1, S.J. & COLE, J.J. 1982. Sedimentation of biogenic matter in deep ocean. Deep Sea Res., 29, 609-25. HUANG, T.C., WATKINS, N.D. & SHAW, D.M. 1975. Atmospherically transported volcanic glass in deepsea sediments: volcanism in subAntarctic latitudes of the south Pacific during the late Pliocene and Pleistocene time. Bull. geol. Soc. Am., 86, 1305-15. --, WATKINS, N.D. & WILSON, L. 1979. Deep-sea tephra from the Azores during the past 300,000 yrs: eruptive cloud height and ash volume estimates. Summary, Part I. Bull. geol. Soc. Am., 90, 131-3. INMAN, D.L. 1949. Sorting of sediments in the light of fluid mechanics. J. sed. Petrol., 19, 51-70. - & Nordstrom, C.E. 1971. On the tectonic and morphologic classification of coasts. J. Geol., 79, 1-21. JACOBI, R.D. 1976. Sediment slides on the northwestern continental margin of Africa. Marine Geol., 22, 157-73. JANOWITZ, G.S. & PIETRAF~A, L.J. 1980. A model and observations of time-dependent upwelling over the mid-shelf and slope. J. geophys. Res., 10, 1574-83. JENKINS, W.J. 1982. Oxygen utilization rates in North Atlantic subtropical gyre and primary production in oligotrophic systems. Nature, 300, 246-48. JERLOV, N.G. 1953. Particle distribution in the ocean. Reports of the Swedish Deep Sea Expedition 1947-48, Ill, 71-98. 1968. Optical Oceanography. Elsevier, Amsterdam. 194 pp. JOHNSON, B.D. & COOKE, R.C. 1980. Organic particles and aggregate formation resulting from the dissolution of bubbles in sea water. Limnol. Oceanogr., 25, 653-61. JOHNSON, D.A., LEDBETTER,M. & BURCKLE, I.H. 1977. Vema Channel palaeooceanography-Pleistocene dissolution cycles and episodic bottom water flow. Marine Geol., 23, 1-33. JUMARS,P.A., NOWELL,A.R.M. & SELF, R.F.L. 1981. A simple model of flow-sediment-organism interaction. Marine Geol., 42, 155-72. KALLE, K. 1939. Bericht 2. Teilfahrt der deutsche nordatlantische Expedition "Meteor". V. Die chemischen Arbeiten auf der "Meteor"-fahrt, Januar-Mai, 1938. Ann. Hydrograph. Maritimen Meteorol., Beih. Z. 1939, 23-30. KARL, H.A. 1976. Processes Influencing Transportation and Deposition of Sediment on the Continental Shelf, Southern California. Unpubl. Ph.D. Dissertation, Univ. So. Calif., Los Angeles. 331 pp. -1981. Response of the suspended sediment trans-
32
D.S. Gorsline
port system to continental shelf dynamics. GeoMarine Letters, 1, 243-8. KENNETT, J.P. 1981. Marine Tephrochronology. In: Emilian, C. (ed.), The Sea. John Wiley, New York. 1373-437. -& THUNELL, R.C. 1975. Global increase in Quaternary explosive volcanism. Science, 187, 497-503. KOBLINSKY, C.J., SIMPSON, J.J. & DICKEY, T.D. 1983. An offshore eddy in the California Current System. Part II: Surface manifestations. Progress in Oceanography, in press. KOLPACK, R.L. 1971. Biological and oceanographical survey of the Santa Barbara Channel oil spill 1969-1970. II. Physical, chemical and geological studies. Allan Hancock Foundation, Univ. So. Calif. Los Angeles. 477 pp. KRANCK, K. 1975. Sediment deposition from flocculated suspensions. Sedimentology, 22, 111-23. KUENEN, PH.H. 1969. Origin of quartz silt. J. sed. Petrol., 39, 1631-33. LA1NE, E.P. 1980. New evidence from beneath the western North Atlantic for the depth of glacial erosion on North America and Greenland. Quat. Res., 144, 188-98. LAL, D. 1977. The oceanic microcosm of particles. Science, 198, 997-1009. LARSEN, L.H. 1982. A new mechanism for seaward dispersion of midshelf sediments. Sedimentology, 29, 279-83. LEDBETTER, M.T. & ELLWOOD, B.B. 1980. Spatial and temporal changes in bottom water velocity and direction from analysis of particle size and alignment in deep-sea sediment. Interpretative modelling of deep-ocean sediments and their physical properties. Marine Geol., 38, 245-62. LEWIS, K.B. 1971. Slumping on a continental slope inclined at 1-4 ~ Sedimentology, 16, 97-110. LINDSTROM,M. 1974. Volcanic contribution to Ordovician pelagic sediments. J. Ned. Petrol., 44, 287-91. LONSDALE,P. 1976. Abyssal circulation of the southeastern Pacific and some geological implications. J. geophys. Res., 81, 1163-76. & MALFAIT, B. 1974. Abyssal dunes of Foraminiferal sands on Carnegie Ridge. Bull. geol. Soc. Am., 85, 1697-1712. LOUGHNAN, F.C. 1969. Chemical Weathering of the Silicate Minerals, Elsevier, Amsterdam. 154 pp. LOWE, D.R. 1976. Grain flow and grain flow deposits. J. Ned. Petrol., 46, 188-99. 1979. Sediment gravity flows: their classification and some problems of its application to natural flows and deposits. In: Doyle, L.J. & Pilkey, O.H. (eds) Geology of the Continental Slopes, Soc. econ. PaleD. Min. Spec. Publ. 27, 75-84. 1982. Sediment gravity flows: II. Depositional models with special reference to deposits of highdensity turbidity currents. J. Ned. Petrol., 52, 279-97. MALOUTA, D.N., GORSLINE, D.S. & THORNTON, S.E. 1981. Processes and rates of basin filling in an active transform margin: Santa Monica Basin, California Continental Borderland. J. Ned. Petrol., 51, 1077-95. MANHEIM, F.T., HATHAWAY,J.C. & UCHUPI, E. 1972. Suspended matter in surface waters of the Northern Gulf of Mexico. Limmol. Oceanog., 17, 17-27.
-
-
-
-
MARTINEZ, L.-A., SILVER, M.W., KING, J.M. & ALLDREDGE, A.L. 1983. Nitrogen fixation by floating diatom mats: a source of new nitrogen to oligotrophic ocean waters. Science, 221, 152-4. MCCAVE, I.N. 1972. Transport and escape of finegrained sediment from shelf areas. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 225-48 pp. -1975. Vertical flux of particles in the ocean. Deep Sea Res., 22, 491-502. 1982. Processes of transport and deposition of fine-grained suspended sediment in the sea. -
-
Abstracts of Papers, Xlth International Congress of Sedimentology, Hamilton, Ontario, Canada. 3 pp. HOLLISTER,C.D., LAINE, E.P., LONSDALE,P.F. & RICHARDSON,M.J. 1982. Erosion and deposition on the eastern margin of the Bermuda Rise in the late Quaternary. Deep Sea Res., 29, 535-61. & SWIFT, S.A. 1976. A physical model for the rate of deposition of fine-grained sediments in the deep sea. Bull. geol. Soc. Am., 87, 541-6. MEADE, R.H. 1969. Errors in using modern stream load data to estimate natural rates of denudation. Bull. geol. Soc. Am., 80, 1265-74. -1982. Sources, sinks and storage of river sediment in the Atlantic drainages of the US. J. Geol., 90, 235-52. - - , SACHS,P.L., MANHEIM,F.T., HATHAWAY,J.C. & SPENCER,O.W. 1975. Sources of suspended matter in waters of the Middle Atlantic Bight. J. sed. Petrol., 45, 171-88. MIDDLETON, G.V. & BOUMA, A.H. (eds) 1973. Turbidites and Deep-water Sedimentation. Pacific Section Soc. econ. PaleD. Min. Short Course, 157 pp. MILLIMAN, J.D. & MEADE, R.H. 1983. World-wide delivery of river sediments to the oceans. J. Geol., 91, 34-42. MORNER, N.-A. 1974. Sea level variations and climatic fluctuations. Colloques Internationaux du CNRS, 219, 135-41. MUTTI, E. & RICcI-LUCCHI, F. 1978. Turbidites of the northern Apennines: Introduction to facies analysis. Internl. Geol. Rev, 20, 125-66 (Translation). NAHON, D. & TROMPETTE,R. 1982. Origin of siltstones glacial grinding versus weathering. Sedimentology, 29, 25-36. NARDIN, T.R. 1981. Seismic stratigraphy of Santa --,
-
-
Monica and San Pedro Basins, California Borderland. Late Neogene history of sedimentation and tectonics. Unpubl. Ph.D. Dissertation, Univ So. Calif., Los Angeles. 295 pp. , EDWARDS, B.D. & GORSLINE, D.S. 1979a. Santa Cruz Basin, California Borderland: dominance of slope processes in basin sedimentation. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. PaleD. Min. Spec. Pub., 27, 209-22. --, HEIN, F.J., GORSLINE, D.S. & EDWARDS, B.D. 1979b. A review of mass movement processes, sediment and acoustic characteristics and contrasts in slope and base-of-slope systems versus canyonfan-basin floor systems. In: Doyle, L.J. & Pilkey,
A review of fine-grained sediment
33
Effect on Man. Geol. Soc. Am. Spec. Pap., 186, C.H., (eds.), Geology of Continental Slopes, Soc. 71-86. econ. PaleD. Min. Spec. Pub. 27, 61-73. & BONATTI, E. 1969. Continental dust in the NATIONAL RESEARCHCOUNCIL 1976. Disposal in the . - atmosphere of the eastern equatorial Pacific. J. Marine Environment. National Academy of geophys. Res., 74, 3362-71. Sciences, NBC Report, Washington, DC. 76 pp. PYE, K. & SPERLING,C.H.B. 1983. Experimental invesN1NKOVITCH, D. & DONN, W.L. 1975. Explosive Cenotigation of silt formation by static breakage prozoic volcanism and climatic interpretations. Science, cesses: the effect of temperature, moisture and salt 194, 899-906. on quartz dune sand and granitic regolith. SedimenNITTROUER,C.A. & STERNBERG,R.W. 1981. The formatology, 30, 49-62. tion of sedimentary strata in an allochthonous shelf environment: the Washington Continental Shelf. RAPP, A. 1974. A review of desertization in Africa -Water, vegetation and man. Secr. Int. Ecology, Marine Geol., 42, 201-32. Rep., 1, Stockholm, 77 pp. , STERNBERG, R.W., CARPENTER, R. & BENNETT, J.T. 1979. The use of Pb-210 geochronology as a REX, R.W. & GOLDBERG,E.D. 1958. Quartz contents of pelagic sediments in the Pacific Ocean. Tellus, 10, sedimentological tool: application to the Washing153-9. ton Continental Shelf. Marine Geol., 31,297-316. SYERS, J.K., JACKSON, M.L. & CLAYTON, R.N. NOWELL, A.R.M., JUMARS, P.A. & EKMAN, J.E. 1981. - - , 1969. Aeolian origin of quartz in soils of Hawaiian Effects of biological activity on the entrainement of Islands and in Pacific pelagic sediments. Science, marine sediments. Marine Geol., 42, 133-54. 163, 277-9. O'BRIEN, N.R., NAKAZAWA,K. & TOLUHASHI,S. 1980. RICHARDSON,M.D. & YOUNG, D.K. 1980. Geoacoustic Use of clay fabric to distinguish turbiditic and models and bioturbation. Marine Geol., 38, 205-18. hemipelagic siltstones and silts. Sedimentology, 27, ROBINSON,A. 1983. Eddies in Marine Science. Springer47-62. Verlag, New York. 609 pp. OSTERBERG,C., eARLY, A.G. & CURL, H. 1964. AccelerROBISON, B.H. & BAILEY,T.G. 1981. Sinking rates and ation of sinking rates of radionuclides in the ocean. dissolution of midwater fish fecal matter. Marine Nature, 200, 1276-77. Biol., 65, 135-42. P A F F E N H O F E R , G.-A. & KNOWLES,S.C. 1979. Ecological RODOLFO, K.S. 1964. Suspended Sediment in Southern implications of fecal pellet size, production and California Waters. Unpubl. M.S. Thesis, Univ. So. consumption by copepods. J. mar. Res., 37, 35-49. Calif., Los Angeles. 91 pp. PARRISH, J.T. 1983. Upwelling deposits: nature of association of organic-rich rock, chert, chalk, phos- - - 1970. Annual suspended sediment supplied to the California Borderland by the southern California phorite and glauconite. Bull. Am. Ass. Petrol. Geol., watershed. J. Ned. Petrol., 40, 666-71. 67, 529. PEWE, T.L. 1981. Desert dust: an overview. In: Pewe, RONA, P.A. 1973. Worldwide unconformities in marine sediments related to eustatic changes in sea level. T.L. (ed.), Desert Dust: Origins, Characteristics, and Nature, 244, 25-6. Effect on Man. Geol. Soc. Am. Spec. Pap., 186, 1-10. PIPER, D.J.W. 1978. Turbidite muds and silts on SAVRDA, C.E., BOTTJER, D.J. & GORSLINE, D.S. 1983. Euxinic biofacies in anoxic basins: San Pedro and deep-sea fans and abyssal plains. In: Stanley, D.J. & Santa Barbara Basin, California Continental BorKelling, G. (eds), Sedimentation in Submarine derland. Bull. Am. Ass. Petrol. Geol., 67, 544. Canyons, Fans and Trenches, Dowden, Hutchinson SCHWALBACH, J.R. 1982. A Sediment Budget for the & Ross, Stroudsburg, Pa. 163-76. Northern Region of the California Continental BorPITMAN, W.C. III 1978. Relationships between eustacy derland. Unpubl. M.S. thesis, Univ. So. Calif., Los and stratigraphic sequences of passive margins. Angeles. 212 pp. Bull. geol. Soc. Am., 89, 1389-1403. POMEROY,L.R., in press. Microbial processes in the sea: SCLATER, J.G., HELLINGER, S. & TAPSCOTT, C. 1977. The palaeobathymetry of the Atlantic Ocean from diversity in nature and science. In: Hobbie, J.E. & the Jurassic to the present. J. Geol., 85, 509-52. Williams, P.J. Leb. (eds.), Heterotrophic Activity in SEIBOLD, E. • HINZ, K. 1973. Continental slope the Sea. Plenum Press, New York. construction and destruction. West Africa. In: Burk, - & DEIBEL,D. 1980. Aggregation of organic matter C.A. & Drake, C.L. (eds), The Geology of the by pelagic tunicates. Limnol. Oceanogr., 25, Continental Margins. Springer-Verlag, Berlin. 643-52. 179-96. PORTER, K.G. & ROaalnS, E.I. 1978. Zooplankton fecal pellets link fossil fuel and phosphate deposits. SHEPARD, F.P., MARSHALL,N.F., McLOUGHLIN, P.A. & SULLIVAN, G.G. 1979. Currents in Submarine Science, 212, 931-3. Canyons and other Sea Valleys. Am. Ass. Petrol. POSTMA, H. 1967. Sediment transport and sedimenGeol. Studies in Geology., 8, 173 pp. tation in the estuarine environment. In. Lauff, H.G. (ed.) Estuaries. Am. Assoc. Advancement Sci. Publ., SHERIDAN, M.F. 1979. Emplacement of pyroclastic flows: a review. In: Chapin, C.E. & Elston, W.E. 83, 158-79. (eds), Ash-flow Tuff`v, Geol. Soc. Am. Spec. Pap., POTTER, P.E., MAYNARD, J.B. & PRYOR, W.A. 1980. 180, 125-36. Sedimentology of Shales. Springer-Verlag, Berlin. SHULENBERG,E. & REID, J.L. 1981. The Pacific shallow 306 pp. oxygen maximum, deep chlorophyll maximum, and PROSPERO,J.M. 1981. Arid regions as sources of mineral primary productivity reconsidered. Deep Sea Res., aerosols in the marine atmosphere. In: Pewe, T.L. (ed.), Desert Dust: Origins, Characteristics, and 28, 901-20.
34
D.S. Gorsline
SILVER, M.W. & BRULAND, K.W. 1981. Differential stratigraphical study of fossil assemblages from the feeding and fecal pellet composition of salps and Maastrichtian white chalk of northwestern Europe. pteropods and the possible origin of the deep-water In." Kauffman, E.G. & Hazel, J.E. (eds), Concepts flora and olive green 'cells'. Marine Biol., 62, and Methods of Biostratigraphy. Dowden, Hutchin263-73. son & Ross, Stroudsburg, Pa. 257-76. SIMPSON, J.J., DICKEY,T.D. & KOBLINSKY,C.J., 1984. SWALLOW,J.C. 1976. Variable currents in mid-ocean. Mesoscale eddies in the California Current and their Oceanus, 19, 18-25. effects on climate. Am. Meteorol. Soc. International SWIFT, D.J.P. 1973. Continental shelf sedimentation. Conf. on Air-Sea Interactions of the Coastal Zone, In: Burk, C.A. & Drake, C.L. (eds), The Geology of The Hague. the Continental Margins. Springer-Verlag, Berlin. SLOSS, L.L. 1962. Stratigraphic models in exploration. 1 I%36. J. sed. Petrol., 32, 415-22. THORNTON, S.E. 1981a. Holocene Stratigraphy and SMALL, L.F., FOWLER, S.W. • UNLU, M.Y. 1979. Sedimentary Processes in Santa Barbara Basin - InSinking rates of natural copepod fecal pellets. fluence of Tectonics, Oceanography, Climate and Marine Biol., 51,233-41. Mass Movement. Unpubl. Ph.D. Dissertation, SMALLEr, I.J. 1966. The properties of glacial loess and Univ. So. Calif., Los Angeles. 351 pp. the formation of loess deposits. J. sed. Petrol., 36, 1981b. Suspended sediment transport in the 669-76. surface waters of the California Current off southSMITH, R.L. 1979. Ash flow magmatism. In: Chapin, ern California. Geo-Marine Letters, 1, 24-8. C.E. & Elston, W.E. (eds), Ash-flow Tufts. Geol. UDDEY, J.A. 1914. Mechanical composition of clastic Soc. Am. Spec. Pap., 180, 5-27. sediments. Bull. geol. Soc. Am., 25, 655-744. SOUTAR, A. & CRILL, P.A. 1977. Sedimentation and UMBGROVE, J.H.F. 1947. The Pulse of the Earth. climatic patterns in the Santa Barbara Basin during Martinus Nijhoff, The Hague. 358 pp. the 19th and 20th Centuries. Bull. geol. Soc. Am., 88, VOGT, P.R. 1979. Global magmatic episodes: new 1161-72. evidence and implications for steady-state midSTANLEY, D.J. 1981. Unifites- structureless muds of oceanic ridges. Geology, 7, 93-8. gravity-flow origin in Mediterranean basins. Geo- WALKER, R.G. 1967. Turbidite sedimentary structures Marine Letters, 1, 77-83. and their relationship to proximal and distal ..- , ADDY, S.K. & BEHRENS,E.W. 1983. The mudline: depositional environments. J. sed. Petrol., 37, variability of its position relative to shelf break. In: 25-43. Stanley, D.J. & Moore, G.T. (eds), The Shelf Break: WATTS, A.B. 1982. Tectonic subsidence, flexure and Critical Interface on Continental Margins. Soc. econ. global changes of sea level. Nature, 297, 469-74. Paleo. Min. Spec. Pub., 33, 279-98. WELLS, J.T. & COLEMAN,J.W. 1981. Physical processes - 81. MALDONADO,A. 1981. Depositional modes for and fine-grained sediment dynamics, coast of Surfine-grained sediment in the western Hellenic inam, South America. J. sed. Petrol., 51, 1053-68. Trench, Eastern Mediterranean. Sedimentology, 28, WENTWORTH, C.K. 1922. A scale of grade and class 273-290. terms for clastic sediments. J. Geol., 30, 377-92. STILLE, H. 1924. Grundfragen der t'ergleichenden Tek- - 1933. Fundamental limits to the sizes of clastic tonic. Verlag Gebruder Borntraeger, Berlin. 443 pp. grains. Science, 77, 633-34. STow, D.A.V. 1979. Distinguishing between fine WILDHARBER, J.L. 1966. Suspended Sediment over the grained turbidites and contourites on the Nova Continental Shelf oft Southern California. Unpubl. Scotian deep water margin. Sedimentology, 26, M.S. Thesis, Univ. So. Calif., Los Angeles. 98 pp. 371-88. WISDOM, H.L. 1969. Atmospheric dust in permanent --& BOWEN, A.J. 1980. A physical model for snowfields; implications to marine sedimentation. transport and sorting of fine-grained sediments by Bull. geol. Soc. Am., 80, 761-82. turbidity currents. Sedimentology, 27, 31-46. YINGST, J.Y. & ALLER, R.C. 1982. Biological activity STRAKHOV,N.M. 1970. Principles of Lithogenesis. 1. and associated sedimentary structures in HEBBLE Consultants Bureau. Plenum Press, New York. 535 area deposits, western North Atlantic. Marine Geol., PP. 448, M7-M 15. SURLYK, F. & BIRKELUND, T. 1979. An integrated D.S. GORSLINE,Department of Geological Sciences, University of Southern California, Los Angeles, California 90089-0741, USA.
Erosion, transport and deposition of fine-grained marine sediments I.N. McCave S U M M A R Y : Fine-grained marine sediments are cohesive but their degree of cohesion is not simply determined by grain size. Cohesion controls erodibility, and water content, mineralogy, cation exchange capacity, salinity of interstitial and eroding fluid, organic mucus content and Bingham yield strength are all parameters relating to cohesion that have been proposed. No unique relation to erodibility has emerged and for erosion of slowly deposited sediment, modified by biota, it seems that measures of surface properties such as aggregate strength should be more relevant than bulk properties such as yield strength. The latter may be more appropriate for rapidly deposited estuarine muds. Once eroded, suspended sediment is subject to flocculation and biological aggregation, which alters size and settling velocity distributions, thereby controlling distribution in the flow and rate of deposition. Disaggregation may also occur through turbulent straining of particles, but in most marine situations this will permit stable particles over 1 mm in diameter, though under higher shear close to the bed the upper limit may be 100-200/~m. For deposition where ~o < 0.1 Pa the maximum stable aggregate size is probably /> 100/~m. Deposition of fine sediment on smooth beds probably occurs by entrapment in, and settling through, the viscous sublayer. On rough beds particles are trapped in the interstices between roughness elements. We have only a rough idea of critical deposition conditions but can give a fairly good estimate of deposition rate in given conditions. The longer term net deposition rate as measured by radiometric methods over weeks to months probably owes as much to frequency of erosion as to frequency and rate of deposition and is not yet well predicted from fluid and sediment parameters.
Our ability to predict the behaviour of finegrained sediment is extremely poor because the particles stick together and the deposits are a wonderful medium for sustaining life. Most problems stem from one or both of these factors. Particles may adhere to one another because the van der Waals forces or electrostatic attractions are large relative to the weight of the particles. Alternatively they may be bonded by mucus secretions of bacteria, algae, diatoms or other members of the benthic in- and epiflora and fauna. Neither the physicochemical nor the biochemical bonding forces can be simply predicted or measured and thus the fluid stress required to remove particles from the bed cannot be predicted either. Moreover the same assemblage of particles may vary in water content or degree of compaction and biological components, thus the same type of mud may have several critical erosion stresses. Transport and deposition of fine-grained sediment depends to a large extent on settling velocity of the suspended particles and that depends on their state of aggregation which is part biological, part physico-chemical in origin. There are no good ways of predicting the state of aggregation of polydisperse, multimineralic suspensions without biota, thus in natural situations empiricism reigns, but based on very few measurements. This review considers mainly the behaviour of mud beds under Newtonian fluid flows and
deposition from dilute suspensions ( < 300 g m -3 in concentration) where particles do not interfere with flow structure. At high concentration ( > 5 - - 1 0 kg m -3) hindered settling begins, the suspension becomes fluid mud and behaves more like a consolidating soil than a turbulent suspension. The flow structure found in low-concentration cases over a smooth bed is the classic turbulent boundary layer comprising a viscous sublayer adjacent to the bed succeeded by a transition (buffer layer) and tubulent core whose velocity profile is fitted by the logarithmic von Karman-Prandtl relation (Rouse 1937). Both erosion and deposition involve interactions in the viscous sublayer whose thickness is ~ ' ~ 10 v/u. where v is kinematic viscosity (1 to 1.4 • 10 -6 m 2 s-1 in shallow to deep water) and u. is the shear velocity ( = ~ ) where ~o is the skin friction and p is the fluid density. Low values of u. in the deep ocean are ~ 1 mm s -1 while on the shelf they generally range up to 100 mm s-1. So ~' ranges from 14 mm down to 0.1 mm. Under depositional conditions the range is 6"= 1-14 mm, i.e. larger than almost all particle aggregates. Deposition then, and much erosion, involves transfer of particles across a viscous sublayer which covers a large part of most mud beds. Viewed at the sublayer scale the erosion and depositional processes should be similar to estuary, shelf, deep-sea and laboratory situations. The driving forces and concentrations will vary in
35
36
I.N. McCave
magnitude and frequency between these locations but at least for slowly varied flow, comparisons should be valid. Clearly we cannot be restricted to deep-sea information as there would be very little to report. This paper is thus general in character, avoiding only the very high concentrations typical of some estuaries, and high frequency oscillatory flow found under waves on the shelf. Although this paper concerns the behaviour of muds, the complexities which afflict them are not completely absent from sands. In particular, sands may suffer adhesion due to mucus and epipsammic flora. The latter are treated by Webb (1969) and Meadows & Anderson (1968), and their effect was strikingly demonstrated by de Boer (1981) who killed off the flora on a sandy tidal-flat (using CuSO4) and found the bedforms completely changed due to an altered erosion threshold. Nowell's (1982) experiments show that the erosion threshold of sands varies according to their muco-polysaccaride content and may be double the value for clean sand with no bacterial mucus. Grant et al. (1982) also present field and
laboratory data showing thresholds raised by biological activity on tidal-fiat sands. The curves of Miller et al. (1977) should be viewed as reference standards only rather than providing appropriate values for field situations.
Erosion Two facets of erosion are of particular interest: the critical conditions for initiation of erosion and the subsequent rate of erosion as a function of excess shear stress.
Critical conditions--physical parameters Erosion and the boundary shear stress at its onset can be measured with adequate precision in laboratory flumes. However, the property of the fine sediment to which this critical stress should be related is not generally agreed, and there have been many suggestions. In one of the most influential but misleading diagrams ever pub-
FIG. 1. Erosion, transport and deposition diagram according to Hjulstrom (1935, 1939) with mean velocity as the ordinate. Curve A is for critical erosion conditions while curve B is for critical deposition with the dashed part extrapolated.
Erosion, transport and deposition of.fine-grained marine sediments lished relating to fine sediment erosion (and deposition), Hjulstrom (1935, 1939) plotted critical erosion velocity against sediment size d (Fig. 1). This is quite acceptable for non-cohesive sediment but cohesion is by no means simply a function of size. Thus the left-hand limb of Hjulstroms curve, rising with decreasing size, expresses a truth that the sediments may be harder to erode than sand, (and in the case of boulder clay much harder); but smaller size is not the simple cause. We cannot predict critical erosion u,~ from d unless the sediment is noncohesive. The examination of fine non-cohesive sediments, pure glacial and crushed quartz silt, was made by Rees (1966), White (1970), Mantz (1977) and Uns61d (1982). This showed that critical values of u, continued to decrease with decreasing d down to sizes of 10/~m quartz in water. The range of u,c is from 8.8 mm s- 1 at 50 /~m to 6.1 mm s- l at 10/~m. For sizes smaller than 10 #m, quartz in water becomes cohesive (Uns61d 1982). The water content of the sediment has been related to U,c by several authors. Migniot (1968) plotted u,c versus sediment concentration C (in kg m -3) but found that different muds plotted on different curves and concluded the plot was not sufficiently general (Fig. 2). He also found that, for the same concentration, U,c was 1 to 2 times greater under salt water than under fresh water. The range of u,~ values, from 4 mm s- 1 to 60 mm
37
s - l for C = 80 to 700 kg m -3, is wide. The two regions shown in Fig. 2 have, for C > 300 kg m -3, critical shear stress r c ~ C 4 and for C < 3 0 0 kg m -3, rczcC 2 to C TM. Thorn & Parsons (1980) and Thorn (1981) summarizing work on four muds (Hydraulics Research Station 1977, 1979a, b) show a range ofz~ between 0.1 and 1 Pa (u,~= 10 to 31 mm s -l) for concentrations of 75 to 200 kg m -3 that are well fitted by a single line (Fig. 3). However, the lowest value ofz~ for each mud after two days consolidation was 0.05, 0.01 0.13 and 0.19 Pa (U,c=6, 10, 11.3, 13.8 mm s-l). Several American workers in the late 1950s and early 1960s sought to express critical erosion conditions for freshwater muds and agricultural soils in terms of the mechanical parameters plastic limit (Pt) and liquid limit (Lt) and plasticity index (Iw=Lt-Pi), (Dunn 1959; Smerdon & Beasley 1959; Gibbs 1962; Carlson & Enger 1963). Many samples of cohesive sediment, when plotted on a graph of Iw versus Lt, fall close to the 'A line', Iw=0.73 ( L / - 2 0 ) (Terzaghi & Peck 1968). These authors show that ~ increases with Iw, the data of Smerdon & Beasley giving -co= 1.31 Iw0"84 Pa for Iw between 10 and 50%. In general z~ increases along the A line, though at higher values of Iw Gibbs (1962) suggests some soils may be expandable and show lower thresholds. This apparent simplicity was confounded by Lyle & Smerdon (1965) who showed that the degree of compaction expressed as voids ratio e also exerts
60 50 40
30
'= 2 0 E E
100
| 1 200 300 C kg m -3
I
! 500
700
Fig. 2. Curves of critical erosion shear velocity versus bed concentration for four different muds from Migniot (1968).
38
I.N. McCave I
I
I
I
I
[
I
1.0
I
I
Y "
oo
I-*
/
0.1
FIG. 3. Curve of critical erosion shear stress versus bed concentration for three muds from Thorn & Parsons (1981).
0.01
I
]
I
i
it
BED
a strong influence (Fig. 4). No work of this type appears to have been published for marine sediments. There have been only two investigations using deep-sea sediments, calcareous ooze (Southard et al. 1971) and red clay (Lonsdale & Southard 1974). The range of concentrations examined is 150 to 800 kg m -3 (47-86~ water of wet weight) and the cohesion displayed by the ooze is slight (u,c = 4 to 8 mm s-1) while that of the red clay is very great (u,c = 8 to 36 mm s - 1(Fig. 5). Failure of the red clay when compacted to < 68% water content was by detachment of lumps, presumably as the bulk shear strength of the material was exceeded. This mode of erosion may result from form-drag on roughness elements created by an initial phase of particle-by-particle erosion. The lack of unique relationship between critical erosion u, and bed concentration for several muds caused Migniot (1968) to use the yield strength r.,. obtained from a rotating viscometer as FIG. 4. Critical erosion stress versus plasticity index with values of voids ratio as an additional parameter for Texas soils (Lyle & Smerdon 1965).
Ill
i
I
I
100
10
kg m -3
CONCENTRATION
2-0
i
T
!
z
I
1.5
1-0
/
.,.q,,
0"5
0
s
i
l
2's
plasticity index
,[w
35
Erosion, transport and deposition ojfine-grained marine sediments 40
!
\
l
!
(o) - " - - - - - M ass erosion (.)
30
Y ~
!
RED CLAY
20 ~4--Particle
ii
39
erosion
CALCAREOUS OOZE " ~ . , , ~ . ~ .
9
9
9
FIG. 5. Critical erosion shear velocity versus water content (percent water/total weight) for deep-sea sediments from McCave's (1978) recalculation of original data in Southard et al. (1971) and Lonsdale & Southard (1974).
--e
I
I
I
I
i
50
60
70
80
90
Water content {% by we/yht]
1962; 9 1975) (Fig. 6). These show for C = 2 0 - 3 0 0 k g m -3, ryocC 25, a dependence which Krone justifies theoretically. The Hydraulics Research Station (1979a) study of Brisbane mud also included some critical erosion shear stresses r,. and yield strengths r,, as a function of C, which were found to plot on the same line (Fig. 7). Nevertheless there is not one line here or on Fig. 2 so neither C nor water content relate uniquely to r,.. Comparisons between determinations of rB from viscometry and vane shear are made difficult by the variability associated with differing
the x-axis. This has the effect of compressing the data for 10 muds onto a single two-part curve with values ofu,c from 3 to 65 mm s -l (Fig. 6). On this curve at low ry, rc=0.32 ry89and for ~>.> 1.5 Pa, "cc=0.26 Vy. The yield strength measured in this way cannot simply be related to that obtained from shear vanes, and no measurements have been reported of in situ yield strength measured by viscometer. However, there have been measurements relating the Bingham yield strength, "rb extrapolated from viscometer measurements to the bed concentration C (Krone IO0
,
, w,,,
I
,
,
,
,
,,,l
I
i
!
i
,
,lit|
!
!
u
,
,,l,l
,
9
!
i
llllj
,
9
~ 9
. "~
/
9
e9
~ o~/" 9 ol~# ,~ilP
9
~ ''~
9
I0
9
9
r / , , , .
0 OiD 9
~
9
,
~:-~"
Ii 9
o.~
..~.~'~,,
i
i ,,,,I
9. 9
"~~
"
i
i
i
i
|llli
o.ol
i
|
i
i
i,l,I
i
I
I
oJ
i
iii
II
JD
i
I
1
i
zJltJ
j
J9
i
i
i
i1!
Joo
Y/e/d Strength Pa FIG. 6. Migniot's (1968) data for many muds collapsed onto a single curve using yield strength determined by vise 9
I.N. M c C a v e
4o
i
i
i
!
!
i
i
i
i
I
i/
I
/
1
// / /
0.8
/ / / ~.~/
0.6
ec
t.,i,J I-w
0.4
/u/
/
/
U
~7 ~)'/
0.2
I-Z 0.1
"" 0.08
,.-I,
0.06
~0.04
0.02
FIG. 7. Yield strength as a function of bed concentration showing many different lines with the same trend (slope ocC5/2 from Krone (1962) and Owen (1975). Also plotted is a line relating critical erosion stress to bed concentration on which 3 determinations of yield strength by rheogoniometer also plot, from Hydraulics Research Station (1979a).
0.at 0.00s o.oo6
0.004 10
equipment and operating conditions. The other experiments listed above are also hard to relate to natural situations because equality of bed concentration does not necessarily imply equality of all the factors contributing to sediment cohesion. In the late 1950s and early 1960s Krone (1959, 1962, 1963) conducted a classic series of experiments on erosion, transport and deposition of cohesive material. A more accessible summary of some of this is in Krone (1976, 1978). Repeated measurements of aggregate shear strength in a
Ill
2o
i
i
,o
I
r162
100
~o
~o
'
500
BED CONCENTRATION(Ckgm-3)
concentric cylinder viscometer yielded a few discrete values for shear strength and density for a given mud, rather than a continuum. Krone (1963) interpreted this in terms of distinct orders of aggregation--an assemblage of primary particles yields a zero-order aggregate (ps = 1269 kg m -3, ~y=2.2 Pa), several zero-order aggregates yield a first order aggregate (ps = 1179 kg m-3, Zy= 0.39 Pa), these combine to yield second order aggregates (ps = 1137 kg m -3, Zy= 0.14 Pa) and so on (values are for San Francisco Bay mud). A
Erosion, transport and deposition of fine-grained marine sediments series of erosion experiments conducted on the beds of Bay mud formed under varying deposition conditions showed that values of suspension concentration versus bed shear stress could be connected by a series of lines whose intercepts on the x-axis gave critical erosion stresses corresponding closely to the set of aggregate strengths deduced from the viscometer measurements (Krone 1962) (Fig. 8). These are interpreted by supposing that the bed has a discrete number of structural arrangements corresponding to the orders of aggregates 9Thus a zero-order aggregate at the surface of a bed of other zero-order aggregates will require a force equivalent to the strength of a first order aggregate to pull it off (Krone 1976). Krone's work leads us to think of erosion as the breaking of bonds between aggregates and the bed, requiring determination of bond strength, an idea that extends beyond the notion of pure electro-chemical flocs attached to a pure silt-clay bed. Compaction and erosion
With increasing overburden the aggregates appear to collapse and Krone (1963) shows a change in structure, probably to zero order at about 25 mm depth. The mass of sediment above the plane of collapse of first- to zero-order aggregates was 5.5 kg m -2. The shear strength of zero-order aggregates is in excess of 2.0 Pa, and in some cases up to 4.6 Pa for sediment with 75% < 2 /~m in diameter. This means that once the aggregates have been compacted to zero-order by overburden they become very difficult to erode, requiring values of u, > 50 mm s - 1. By contrast,
9
,,,-c~
A
4I
even first-order aggregates are much weaker, the greatest measured by Krone (1963) being 0.94 Pa. Aggregates of greater than zero-order are unlikely to be able to resist forces due to waves and tides combined, but could probably withstand those due to tides alone. The production of a bed of zero-order seems to be a prerequisite for the survival of mud deposits in high-stress zones, i.e. at least 5 kg m -2 should have been accumulated for the lowest layers to be ultimately preserved. A related feature of beds in high-stress areas is that they require higher stresses to erode them than materials of similar size and mineralogy in low-stress areas. A bed sample taken in an improperly closed box corer from the Nova Scotian Rise was found to have been eroded by water flowing through the corer leaving a rather irregular but firm mud surface. Samples of the same mud, taken undisturbed had a very soft surface and shear-vane tests showed the mud layer to have uniform and low shear strength down to 80 mm depth, below which it was hard and of different composition. It seems that the bed had 'hardened' under stress, the weak surface aggregates presumably being replaced by stronger ones of zero-order. The process is probably analogous to the stiffening of soils under sustained loading described by Mitchell (1976, p. 292). The sand content of mud increases its critical erosion shear stress and the presence of sand layers assists the drainage of mud layers resulting in their more rapid compaction (Terwindt & Breusers 1972). These authors found that a 20 mm thick mud layer with 37% sand, consolidated for 2 hours, could withstand stresses of u, up to
9
8
B
ev- c..~
~'-<2 ~.1 L.w
~,,'r o
0
0.2
0"4 BED SHEAR
ro
0"6 Pa
0"8
1.0
FIG. 8. Rate of erosion for San Francisco Bay mud beds prepared many times under different conditions from Krone (1976). The intercepts of the lines on the x-axis give the critical erosion stress which corresponds well with the aggregate strengths for this mud in Table 1.
42
I.N.
McCave
< 19 m m s -l, a value higher than that for the erosion of sand. After three hours' consolidation the mud was strong enough to bear a sand layer without deformation. This gives another way of compacting the mud down to zero-order: deposition of sand on top of the mud may provide the required overburden. However, before that can happen in most situations the mud must have consolidated to a surface shear strength greater than the critical erosion shear stress for the sand, i.e. > 0.15 Pa. Most estuarine and coastal muds have a significant admixture of several tens of percent of sand thereby giving the right conditions for preservation.
Erosion rate--physico-chemical parameters Several workers consider that the strength of the binding force between clay particles can be simply parameterized by the cation exchange capacity (m equiv/100 g) of the material. Sargunam et al. (1973) showed that the same soil but with different ionic concentration of interstitial fluid (0.004N-0.14N NaCI and 0.14N CaCI2) had different critical erosion stresses (z,= 1.1 to 8.5 Pa) under distilled water. However, when the eroding and interstitial fluids have the same composition critical erosion stress became very high. They concluded that migration of distilled
1
water into the soil and swelling due to the osmotic pressure gradient played an important part in the erosion process. This may be of some importance in estuaries where tidal and seasonal changes in the limit of saline intrusion occur. However, cation exchange capacity (CEC) alone does not parameterize erodability, as clay with adsorbed Ca + + is harder to erode than with adsorbed Na+: the sodium adsorption ratio S A R = N a + / (Ca + + + Mg + +) is important too. Arulanandan (1975) showed that Zc increased with increasing CEC for low SAR but decreased for high SAR. No single parameter indicative of critical shear stress emerges from this approach. Plots of erosion rate b (kg m -2 s- 1) versus shear stress show that for a given bed and fluid at a fixed temperature a straight line is obtained described by = M (vo - re)
where, if0 is in kg m -2 s -1 and -c is in Pa (Nm -2) M has units s m -1. Examples are given by Sargunam et al. (1973), Owen (1975), Ariathurai & Arulanandan (1978) and Thorn & Parsons (1980). Owen (1975) gives values of M of 1.07 to 2.04x 10 -3 s m -l for beds of density C = 199 to 246 kg m -3 at zero salinity, and 2.04 to 0.31 x 10 -3 s m -l for beds of 246 k g m -3 density at salinities from 0 to 32%. In his experiments,
1
1
13. Q. ,m
E
E O)
/~ //,//~
00
INDIVIDUALMUDS: o Brisbane 9 Grangemouth A Belawan 9Scheldt
I
10
(1)
I
20 /Xt (minutes)
3'0
FIG. 9. Rate of erosion per unit shear stress excess as a function of the measurement time-step drawn from Thorn (1980), Thorn & Parsons (1981).
E r o s i o n , t r a n s p o r t a n d deposition critical conditions were not affected by salinity but erosion rate was. Thorn & Parsons (1980) show that when a step change is made in shear stress the suspended sediment concentration initially rises rapidly but then levels off giving erosion rate constants M of 2.63 • 10 -3 s m -1 averaged over 10 min., 1.78 • 10 -3 S m -I over 20 min. and 1.39 • 10 -3 S m -1 over 30 min. for beds of C ,-~250 kg m-3 (Fig. 9). The experiments at the Hydraulics Research Station (Thorn & Parsons 1980; Thorn 1981) were conducted by depositing a mud bed from flowing water and allowing it to compact for two days. Such a bed shows a density profile increasing downwards. The step changes made in shear stress allow both the erosion rate and the critical erosion stress to be obtained because the concentration becomes steady after a while at which time the amount of sediment removed from the bed and thus, using the density profile, the density of the exposed surface, can be determined. The values of density C and zc of Fig. 3 were determined in this way. There may be some ambiguity here in that behaviour under clear water at the given stress would differ by lower "co (Allersma et al. 1967). Thus for clear water flows the curve of Fig. 3 would not be so steep. Ariathurai & Arulanandan (1978) show the erosion rate constant M to vary as a function of CEC, SAR and temperature. With increasing temperature the erosion rate rises more steeply and the critical erosion stress decreases (Fig. 10). These authors express their results using = MT(~0/1W--1)
(2)
thus their Mr = M~c. Values of M, are 0.5 to 5 • 10 - 3 kg m -2 s -1 corresponding, with ~c = 1 to 2.5 Pa, to M of 0.2 to 5 • -3 s m -~,
i
4
I
I
I
/
of fine-grained marine sediments
43
comparable with the results of Owen and Thorn & Parsons. Although in the deep sea, pore and eroding fluid composition, SAR and temperature may be regarded as constant, CEC is not. In estuaries the system may be very variable via several parameters. Clearly both the erosion rate and critical stress are functions of several physicochemical variables. Finally Ariathurai & Arulanandan (1978) note that presence of organic matter is also a factor in critical shear stress, and Young & Southard (1978) using natural marine muds, found that critical stress increased with organic carbon content, varying linearly between U,c = 8.3 mm s -1 at 0.8~ organic C and 12.9 mm s-1 at 2.0~o C. The observation of the temperature dependence of erosion rate has suggested to several authors, most recently Gularte et al. (1980), that rate-process theory should apply to erosion 0=X-~-exp-
~
exp
(3)
where k, h and R are the Boltzmann, Plank and general gas constants, T is absolute temperature, AF activation energy, Vz volume of 'flow units' (atoms or molecules in chemical reaction rates but probably clay particles here) and X a constant related to the frequency of activation. The process is thus considered in the framework of chemical theory developed for reaction rates but with application in soil mechanics (Mitchell 1964). If an experimental activation energy E is substituted for AF--V/rN/2 where N is Avogadro's number (Mitchell et al. 1968) then
lakT
= X-:- exp (--E)_~_
(4)
I
% x
~C
~,,,3 E
~2 nZ
o m
/,S
1
o r LU
0
"
i
0
1
2
3 SHEAR
I
1
I
4
5
6
STRESS
Pa
FIG. 10. T e m p e r a t u r e d e p e n d a n c e o f erosion rate and critical erosion stress 7
of compacted muds in a Moore & Masch apparatus from Ariathurai & Arulanandan (1978).
44
I.N. McCave
These two equations give a basis for determination of E and Vf from (4) ?In(b/T)
E
~(1/7)
R
(5)
and from (3) Oln b - -
~z
VU 2kT
-
(6)
From plotting log ~/T versus 1/T(Fig. 11) E may be determined, and from plotting log ,~ versus z, (Fig. 12) V; is found. The data of Gularte et al. (1980) determined from flow over an artificial illite-silt mixture in a water tunnel yield activation energies of 73 to 107 kJ mol-1 and flow volumes of 1.5 to 6.1 x 10 -19 m 3 (equivalent spherical diameter 0.66 and 1.05 pm, possibly illite particles) which decrease with increasing salinity. The activation energies are little less than those reported for soil creep but much greater than those for viscous flow of water. Thus Gularte et al. (1980) argues that breaking of strong interparticle forces is indicated. A flow volume decrease with increasing salinity (i.e. smaller 81n ~ / ~ at higher salinity) is much harder to explain convincingly in terms of flow (particle) volume but is
10 "3
E o~ 10 "4 v, I-
9
.,,.,o
10-5 3.2
i
.
I
I
I
I
3.3
3.4
3.5
3.6
lit
~
x 10 3
FIG. 11. Log-linear Arrhenius plot of erosion rate--temperature relationship shown by Gularte et al. (1980).
consistent with compression of the double layer resulting in stronger cohesion of aggregates. The approach gives some insights into the control of erodibility for pure materials but the actual values may not be realistic for the natural environment with biota, and in the constancy of the deep sea there is little relevance in notions of control of erosion by temperature. Experimental variation of temperature in situ would undoubtedly increase erosion rate (if only through the death-throes of the biota) but inference of rateprocess parameters would not be facilitated. Substantial data from coastal environments shows that suspended sediment concentration, probably due to sea-bed and tidal-flat erosion, becomes greater as the temperature falls (Buchan et al. 1967; Newton & Gray 1972). It seems unlikely that this is controlled by temperature, sodium absorption ratio, cation exchange capacity, water content, grain-size or any other of the mechanical properties investigated so far. Biological
influences
Few experiments have provided hard data on the subject but many field observations indicate that organisms may both stabilize and destabilize sediment. Stabilization of sediments by benthic diatoms and other micro-algae was investigated using an in situ bottom flume in the Bahamas by Scoff=in (1968, 1970) and Newmann et al. (1970). Mats of Enteromorpha and Lyngbya are extremely tough, the former withstanding flows up to 1 m s-J in a 0.1 • 0.1 • 3 m duct (Scoff=in 1970) probably equivalent to u,=50-100 mm s -l depending on roughness. These mats become so tough and thick (10 ram) that one cannot speak of critical erosion conditions for the sediment (which is not in contact with the flow), but rather 'mat breakup conditions' which is what happens in practice. The stabilizing influence of diatom-produced mucilage on sediment was shown by simple paddle-wheel stirring in culture flasks by Holland et al. (1974). Critical erosion stresses were not measured, but a significant difference in erodibility was demonstrated. Diatom populations on tidal-fiats produce mucus threads shown by Gouleau (1976) and Robert & Gouleau (1977) and presumed to be responsible for stabilizing the surface and promoting rapid mud deposition. Similar arguments were made by Coles (1979) and Frostick & McCave (1979) concerning mobile epipelic algae that leave a trail of mucus in sediment. Both authors found seasonal (summer) build up of mud on tidal-flats and winter erosion correlated with live/dead or absent algae. This suggested to Frostick & McCave (1979) an
Erosion, t r a n s p o r t a n d deposition o f f i n e - g r a i n e d m a r i n e s e d i m e n t s 10-4 -
45
/
/ /
r
/ /
J~
.
/ /
/
10--~
/
/
/Io
9 / /.
//. -
-
~
2', E
" "
~ /
.
/
9
/o
/ /.
J
9
/
/0
Io-eL /::)~ / i/. I- I /, I
10-)
I
0
I
I
0.1
0.2
I
0.3 l-, Pa
I
I
I
0"4
0"5
0"6
explanation for the winter increase in coastal suspended sediment concentrations mentioned above. Stabilization and destabilization may occur at different times of the year in subtidal sediments also. Rhoads et al. (1978) took undisturbed samples of the muddy bottom of Long Island Sound and inserted them into the bottom of a laboratory flume. Critical erosion experiments revealed a seasonal trend of low threshold in summer and high in winter during 1974-1975, but this trend was not continued into 1976 and the changes may be random. Nevertheless the values are significantly different, going from a minimum of Oc = 0.07 up to 0.11 m s- l in a flow 0.1 m deep, where fT<, is the critical mean velocity for movement of particles (generally aggregates). Rhoads et al. also conducted experiments on mobility of glass beads with and without coatings of microbial mucus. They found that for silt and fine sand the critical erosion velocity was increased by 57%, and the index of mucus concentration chosen, measurement of ATP (adenosine triphosphate), correlated well with increase in 0<, for the beads.
FIG. 12. Salinity dependence of erosion rate on shear stress from Gularte et al. (1980). Open circles are for salinity of 0 to 5%0, black dots for S = 7.5 and 10%0.
However, measurements of ATP on the cored sediments did not correlate particularly well with the change in bed stability because other organisms tend to destabilize the bed (Yingst & Rhoads 1978). The number of competing influences proved too great for their experimental design. In two series of carefully controlled experiments Nowell et al. (1981) and Ekman et al. (1981) examined particular aspects of stabilization/destabilization by organisms. Scour pits may be found around single animal tubes, but around clusters of tubes apparent stabilization is recorded. The experiments of Nowell & Church (1979) with different concentrations of roughness elements suggested stability fields for different tube densities, e.g. to stabilize the bed with tubes of 5 mm diameter more than 4 • 1 0 3 tubes m - 2 a r e needed. This result was confirmed by the results of Ekman et al. (1981), but most natural tube densities plot in the destabilizing field although the beds are observed to be stable. They suggest that mucus binding associated with organic activity around the tubes is responsible for bed
46
I.N. McCave
stability. The formation of tracks and trails roughens the bed causing it to be more easily eroded, with a 20% reduction in critical erosion velocity shown by the experiments ofNowell et al. (1981 ). These authors also show that faecal pellets may be rolled over a bed of cohesive mud quite easily. Their experiments together with those of Rhoads et al. give critical erosion speeds of biologically destabilized beds between 0~ = 0.50 and 0.25 m s -l, where O, is the flow speed one metre above the bed. Erosion of mucus-bound faecal coils starts at u,=12.0 mm s -l. These results are quite reasonable when compared with the only available in situ data from Young & Southard (1978). They found u,c of 3.2 to 8.4 mm s-1 using an inverted flume in Buzzards Bay but the same sediment in the laboratory after bioturbation for up to 60 days gave u,c = 9.2 mm s -1. These studies show that biological destabilization is most important, but the results of Young & Southard (1978) suggest that even carefully conducted laboratory experiments may over-estimate critical erosion speeds by a factor of two. Rather than the experimental range of u,~ = 8 to 18 mm s- Lan in situ range of 4 to 9 mm s- 1may be more correct. This would be equivalent to O~=0.10 to 0.25 m s -~ over a smooth bed. It suggests that the initial turbidity generated by deep-sea currents at speeds of 0~ ~0.2 m s -1 results from movement, breakup and entrainment of biologically destabilized parts of the bed. The rate of erosion of a bed could be entirely controlled by the rate at which such biological activity produced these less stable structures. In regions of higher stress such as tidal estuaries and shallow shelves, however, there may be more important physical control of erodibility.
ing a distinct layer beneath the normal dilute suspension of suspended sediment in sea water. Fluid mud occurs only in some estuaries and a few near-shore zones with very high concentrations of suspended sediment. Central to all these features of suspended sediment is the phenomenon of particle aggregation which may be brought about by biological means, electrochemical coagulation or polymer flocculation (Stumm & Morgan 1981).
Aggregation For aggregation to occur, two conditions must be fulfilled. Some process must bring particles close enough to collide and some of the collisions must result in coalescence, i.e. they must stick. Physical processes involved in bringing particles together are Brownian motion, laminar and turbulent shear and differential settling. The mechanisms are investigated in detail by many authors including Smoluchowski (1917) for the first formulation of the fundamentals, Friedlander (1977) and Pruppacher & Klett (1978) for mechanics of aerosols and Einstein & Krone (1962), Krone (1978), Hunt (1980, 1982) and McCave (in press) for applications to estuarine and marine particle coagulation. In Friedlander's (1977, chap. 7) general dynamical equation for the population balance of particles are two flocculation terms
On(v) ; c?t - 89 K(vi, v j - vi)n(vi)n(vj- vi)dvi0
f K(t)--t'i)n(v/)n(vi)dv/
(7)
0
Transport and aggregation in suspension During transportation, fine-grained sediment undergoes changes in its size and thus settling velocity distributions. This partially determines the vertical distribution in the flow, the rate of deposition, and relates to sorting of particles. It is in contrast to the inert behaviour of sand grains which are distributed in the flow under the controls of turbulence and their fixed settling velocity. At high concentrations (generally > 5-10 kg m -3, Krone 1962; Diephus 1966; Owen 1975) the mud may display hindered settling, i.e. a reduction in settling velocity, and start to develop cohesive strength (Krone 1963). This property is essential to the formation of 'fluid mud', high concentration suspensions (often 50-200 kg m-3) with a yield strength (i.e. non-Newtonian), form-
The first term on the right accounts for formation of particles of volume t~ by collision of particles of volumes Vi and (r i-vi) and the second term gives the loss of rJ sized particles by collision with all others. The size distributions are given by n(v) = dN/dv where dN is the number concentration of particles in volume v to v + dr. The terms K are the 'coagulation kernels' and are specific for different mechanisms (Friedlander 1977; Pruppacher & Klett 1978). Thus for Brownian motion with particles i and j, diameter d, and d o. = (di + dj) 2kT~j KBij "~" 3]2 di4
(8)
where T is absolute temperature and k is the Boltzmann constant. For turbulent shear the kernel is Ksij = m.d3ij (~/v) 89
(9)
Erosion, transport and deposition of fine-grained marine sediments where A is a constant equal to 0.163 based on Saffman & Turner (1956), and e is the turbulent dissipation rate. The kernel for differential settling is
time to halve the number of particles in suspension, can be calculated. For Brownian motion this is t,.B --
(10)
K~o = rc~J Aw~ij ( E c o + EDo)
47
4 where Aw~.o is the settling velocity difference between the particles, E c is the collision efficiency of the particles and ED is the collection efficiency by Brownian diffusion of particles onto a larger particle settling through smaller ones. Plotting these kernels for deep ocean conditions of temperature (viscosity) and turbulence shows (Fig. 13) that Brownian motion dominates for sizes below 1 pm, succeeded by capture through differential sedimentation and turbulent shear with increasing particle size. Using a somewhat different set of assumed values (high shear, constant particle density with size), Hunt (1982) suggests a region of shear dominance from 1 pm to 30/~m. However, oceanic suspensions usually contain a few large particles and these should be efficient scavengers even at low shear rates, because of the term d~ij in the turbulent shear kernel (McCave, in press). For more pure mud suspensions in regions of high shear (e.g. tidal estuaries) there may still be large particles present (e.g. Biddle & Miles (1972) show 800 pm flocs) whose bulk density is considerably less than that of smaller aggregates (Tambo & Watanabe 1979). If suspensions are initially monodisperse (all particles the same size) a 'coagulation time', the
3~t 4k TNo
(11)
and for turbulent shear t,.s -
1.04
.~ , Nod (e/v)~
(12)
where N,, is the initial particle concentration. For high-concentration deep-sea nepheloid-layers in the range d = 0.5-1 pm, No--~4.105 cm -3 and tcB "-, 889days. However, with a more realistic coalescence efficiency of 10%, t,8 becomes about 3 months. Coastal water concentrations are at least two orders of magnitude higher and tcB drops to 2 to 20 hours there. For shear with No = 6.103 cm -3 in the range d = 4 - 8 pm and deep-sea conditions, (e/v)l ,,, 0.084 s- I, &s = 4.4 months or, at 10% efficiency, 3.6 years. In coastal water with both a higher shear rate (-,- 1 s- 1) and concentration (N,, = 6.105 cm-3), t,s is in the range 2.7 to 27 hours. In both cases the figures suggest that significant coagulation can occur on the timescale of the tidal cycle in shallow water but that months to years are more appropriate to oceanic nepheloid-layers. In midwater ocean depths McCave (in press), shows that calculated coagulation rates are extremely slow with t,. in the region of tens to hundreds of years. Whether particles will be brought close enough together to stick depends on the surface chemistry
-6-7
-
di = 5 / z m .
.
.
.
-8
-9
kG kB
-10
k6
-11 -12
R = B~OWNIAN 5" = S'HEA~
ks
# = DIFFU~ENTIAL E~LDIMENTATION -13
--t
0.1
I
I
I
1
10
100
I
103
I
104
FIG. 13. Coagulation kernels plotted for collision of a 5 p m particle with particles of 1 to 104 pm diameter using particle and fluid parameters detailed by McCave (in press). Diagrams of this type were introduced by Friedlander (1965) and have been presented for aqueous suspensions by Lerman (1979) and Hunt (1980).
48
I.N. McCave
of the particles and the solution. In simple outline the balance between van der Waals attractive forces and electrical double-layer repulsion determines whether mineral particles in a pure solution will stick. If the solution has high ionic strength the particles' surface charge is balanced by a thin layer of counter-ions and the repulsive force is smaller a few nanometres away from the surface than it is in low ionic strength conditions. This permits particles to come close enough for van der Waals forces to dominate. The theory behind these models due to Derjagin & Landau and Verwey & Overbeek (DLVO) is given in Stumm & Morgan (1981) and van Olphen (1977). However, this is not the only mode of particle binding. Bacteria are observed to aggregate because they exude biopolymers. The exudates of both bacteria and algae give rise to 'bioflocculation' which may also involve mineral matter as the polymer can form a bridge between particles (Stumm & Morgan 1981; Busch & Stumm i968). This flocculation by polymers is an important process in waste water treatment dealt with by O'Melia (1972, 1978), Gregory (1978) and Lyklema (1978). Given the substantial bacterial populations of bottom sediments, binding by polymer bridging may be important in accounting for some of their cohesive properties. Binding of suspended clay and silt particles by mucus threads associated with diatoms is shown by Ernissee & Abbott (1975) (Thallasosira), and by Lewin et al. (1980) (Chaetoceras). Feeding by several groups of zooplankton (e.g. salps, doliolids, some pteropods, larvaceans) involves mucus nets used as particle traps which are then ingested or from which food is removed (Jorgensen 1966; Madin 1974; Gilmer 1974; Alldredge 1977; Muliin 1980; Bruland & Silver 1981; Alldredge & Madin 1982). These mucus structures are sometimes discarded and settle, acquiring more particles on the way down (Alldredge 1976). Thus
there are several ways in which organisms can promote aggregation in suspensions, apart from ingestion and pellet formation (Haven & Morales-Alamo 1972; Honjo & Roman 1979), and the phenomenon should certainly not be considered exclusively to be inorganically electrochemical. A further feature of natural suspended particles is that they tend to show the same (slightly negative) electrophoretic mobility in sea water, suggesting that some organic coating may determine important surface electrical properties (Niehoff& Loeb 1972, 1974; Hunter & Liss 1982). If all the particles are similar it may have the effect of reducing the proportion of collisions that result in coalescence. There are scarcely any determinations of the actual influence of all these effects on coalescence efficiency of natural particles. Swift & Friedlander (1964) obtained an efficiency of 37~ for polystyrene latex in artificial salt water for Brownian motion and laminar shear. Edzwald et at. (1974) obtain a value of 10% for natural estuarine sediment suspended in salt water, which is the best value available, although real world values are probably very variable.
Aggregate strengths and sizes Krone (1978) gives a pr6cis of his original (1963) elegant hypothesis of orders of aggregation. In essence he found that the same suspension could show different shear strengths and of volume concentrations of aggregates according to the shear rate. Furthermore the variation was discrete rather than continuous. Thus he supposed that at high shear only primary aggregates (zeroorder) could survive but at lower shear aggregates of these aggregates (first-order) would be stable and so on at even lower shears. He also calculated the densities of these aggregates as shown in Table 1. In most cases four orders are recognized
TABLE 1. Properties of sediment aggregates (from Krone 1976). Aggregate order 0 1 2 3 4 5 6
San Francisco Bay p kgm -3 r Pa 1269 2.2 1179 0.39 1137 0.14 1113 0.14 1098 0.082 1087 0.036 1079 0.020
Delaware Estuary p r 1250 2.1 1132 0.94 1093 0.26 1074 0.12
Brunswick Harbour, GA p r 1164 3.4 1090 0.41 1067 0.12 1056 0.062
Gulfport Channel MS p r 1205 4.6 1106 0.69 1078 0.47 1065 0.18
Bulk densities p are with interstitial fluid pw= 1025 kg m -3 Aggregate shear strength ~ was determined by viscometer
White River (saline) p 1212 4.9 1109 0.68 1079 0.47 1065 0.19
Erosion, transport and deposition o f f i n e - g r a i n e d marine sediments
49
FIG. 14. Particle volume size spectra determined by Coulter counter from Eisma & Kalf (1979), southern North Sea. Note higher concentration peaked and low concentration fiat winter distributions with no biological interference, and peaked late summer spectra dominated by biota. though seven are found for San Francisco Bay samples. Krone (1976) relates the strengths of these orders of aggregation to critical erosion stresses and the relation between aggregate strengths and stresses underlies the suggestion of 'stress-hardening' made in the previous section here. Size distributions from shallow and deep water have been principally determined using the Coulter Coulter (Brun-Cottan 1971; Sheldon et al. 1972; Eisma & Kalf 1979; Pak et al. 1980; McCave 1983). Recently some measurements
with the HIAC counter (Dickson & McCave 1982) and the laser forward-scattering particle counter (Bale et ai., in press) have appeared. Some authors have also used microscopic enlargement and particle counting (Schubel 1968, 1969, 1971; Wellershaus et al. 1973; Harris 1977; Baker et al. 1979; Lambert et al. 1981) to produce number spectra. Estuarine volume-size distributions at moderate concentrations ( ~ 100 g m -3) show a single well-developed peak (Kranck 1981) as do samples from the shelf when free of biological influences
5~
I.N.
McCave
100
00
5
m, E
T, ~ 1 7 6
10 ~ 9
~,g
"
o
'uWO
9 o l9l
...
.Ab
9
.o 0% o8 9
/
Iii
Cm -t" - - - - - - / " ~
9o
~.".~- b. "" "t
/ Ct
9"+" I' .% eee,.e-j
~
/\
\ \
- -
~ o
FIG. 15. Relationship between peakedness of distribution and concentration, from Kranck ( 1981).
I
(Kranck 1973, 1975; Eisma & Kalf 1979). The modal peak occurs mainly between 5 and 20/~m, but where plankton are important several modes may be encountered (Eisma & Gieskes 1977; Eisma & Kalf 1979) (Fig. 14). In the southern North Sea the peaked distributions are found nearshore while offshore at low concentrations ( < 10 g m -3) the size spectra are flat (Fig. 14, Eisma & Kalf 1979). This peakedness is clearly related to concentration in Kranck's (1981) estuarine data (Fig. 15). She maintains that this peak is produced by flocculation and that the kinetics of the process only result in a well developed peak when concentrations are high enough for it to develop rapidly. Kranck (1973, 1975) and McCave (1981) show aggregated and disaggregated distributions to demonstrate this essential feature of (all) suspended sediment size distributions (Fig. 16), a feature which in one data set from shelf and estuaries yields a regular relation between the flocculated (d/) and deflocculated (da) modal sizes, df=2.79 dd0'772 (Fig. 17). Some of this flocculation may involve organic binding of particles as samples may be readily deflocculated with hydrogen peroxide (Fig. 16). Data from Chesapeake Bay have figured in arguments that flocculation is not important in estuaries (Schubel 1968; Meade 1972) and that it is (Zabawa 1978). Although the microscopically counted size distributions from there appear not to relate closely to volume size distributions
I
10
I 100
C~/Ct (McCave 1979a, Fig. 6.11), nevertheless the data of Schubel (1968, 1969, 1971) do demonstrate the fact of fine background and intermittently resuspended populations of particles. The larger intermittently resuspended particles are shown by Zabawa (1978) to contain aggregates of both faecal pellet and electrochemical coagulation origins. Krone (1972, 1978) also argues persuasively that flocculation is a significant process in modifying size distributions and enhancing sedimentation through the consequent increase in settling velocity. Few determinations of oceanic particle size distributions have been made. Bader (1970) observed that particle-number size distributions in water tended to follow a power law relationship, and several workers (Brun-Cottan 1971, 1976; Sheldon et al. 1972: McCave 1975; Lerman et al. 1977; Baker et al. 1979; Pak et al. 1980; Richardson 1980; McCave 1983) showed that the exponent in the distribution N = a d -m was about m = 3 (i.e. a Junge distribution), where dis particle diameter and N is the number of particles bigger than d, for the region between the intense bottom and surface nepheloid-layers (Fig. 18). The presence of distributions comprising two segments of slopes m < 3 for sizes finer and m > 3 for sizes coarser than 5 pm is shown for shallow water by Brun-Cottan (1971), Bader (1970) and McCave (1975) and for nepheloid-layers both on the shelf (Pak et al. 1980) and in deeper water (Richardson
Erosion, transport and deposition o f fine-grained marine sediments 0.3
1.2
d. 0.2
0.8
g~
0.4
I-
,
~
i
1.4
t~ 0.6
1.2
0.4 O
0.8 0.4
o
9
0
'
1
~'~'
10
0
100
!
~
1
10
100
DIAMETER (pm)
Gently w e t
4 ~ (63~Jm)
sieved
==s
phi r ",2
4
8 /Jm
16
32
It]
Treated
with
H202
5
phi
Sample of suspended sediment taken at bottom +2m in 18m w a t e r depth off Walcott,Norfolk
FIG. 16. Raw and disaggregated size distributions from Nova Scotian coastal waters (Kranck 1973) (above) and southern North Sea coastal waters (McCave 1981) (below). 1980; McCave 1983) (Fig. 18). The m = 3 distribution is equivalent to a flat volume distribution, i.e. equal volumes of particles in logarithmic size grades (Sheldon et al. 1972) which appears to be the case in clear water, while the two-segment distribution is a peaked volume distribution with
5I
the peak in the region 4 to 8 pm found in more turbid water in direct comparison with the shelf and estuarine data given above, though at much lower concentrations. Lambert et al. (1981) measured and counted particles retained on 0.4/~m pore size Nuclepore filters. They find that the individual components (e.g. aluminosilicate grains, aggregates, o p a l + quartz, goethite particles) fit log-normal particle number distributions. In particular they find decreasing numbers of particles per logarithmic size interval for sizes < 0.8 pm, in contrast to the findings of Harris (1977). Similar results from the same method are reported by Baker et al. (1979) where decrease in dN/dd is demonstrated for d < 2.94 pm. Above that size the distributions are close to a power law with m > 3, thus there is a peak in the volume size distribution at about d-- 3 pro. Harris' (1977) data show a region where m = 1.65, close to the Brownian coagulation prediction of 1.5. The discrepancy between his data and those of other authors in the submicron range may either be undercount in the others as suggested by Wellershaus et al. (1973) or a true lack of such particles because they have already passed on up the size spectrum by aggregation. Aggregation in these deep-sea suspended sediment samples is also shown by the ultrasonic disaggregation experiments of McCave (1983) (Fig. 19). In these there were two modes of behaviour. High concentration, peaked distributions showed a substantial drop in the fine peak at 4 pm and an overall loss of material. This was thought to be due to disaggregation into submicron particles not now counted, being below the threshold. On the other hand, several of the very low concentration flat distributions actually gained material (Fig. 19). It was thought that in the process of becoming flat, much material in these distributions had been transferred into larger aggregates than the counter was registering (> 32 pm). Disaggregation of these particles was suggested to provide more material to the distribution than was lost by breakup of particles in the 1.26-32 #m counted window. Thus the flat distributions have gone beyond the stage of dominance by a single flocculated peak and have material more widely spread across the size spectrum (McCave, in press).
Breakup Analogous to the coagulation kernels of equation 7 are breakup kernels which account for the breakup ofvj to yield (L~-v;) and vi sized particles such that the rate of breakup of~!j per unit volume of suspension is
dN/dt
=
J(t~-; vi,
Vj--Vi) n (vj) dr~ dvi
(13)
I.N. McCave
52
64
16 E 9 0 /
m
9
8
0
0-.a U,.
4
de= 2"79d0772
FIG. 17. Relationship between modal diameters of flocculated and disaggregated suspensions from Nova Scotian estuaries (Kranck 1973).
I 2
according to Jeffrey (1982). A similar formulation giving the production rate of vi sizes from breakup ofvj sized particles was given by Valentas et al. (1966) and Valentas & Amundson (1966) (summarized by Spielman (1978)) as
dN/dt = J(vi, vj) n (vj) dvi dvj
(14)
but this presumes breakup into two equal halves. Jeffrey's expression yields these terms in the population balance equation for production and loss of vSs by breakup
~n(v) Ot
i
i
| J(vi+vfi vj, vi) n (?.)j--~ui) d v i Q/
o vj
f J(t); vi, v j - vi) n (vj) dvi o
(15)
However, it is by no means clear what is the functional form of J. This depends to some extent on the mode of particle breakup. Parker et al. (1972) proposed that particles either are fractured or are eroded particle-by-particle through surface shear. Their analysis showed that for particles smaller than the Kolmogorov microscale 2k = (I,'3//~) ~ the maximum size stable to surface erosion should be dm oc -~ and for d > 2k it should be dmocF -2, while in both regimes the
i
32
I
64
breakup size should be dm oc F- 89 (Note, as shear rate F oc e89where e is turbulent dissipation rate, these relations are equivalent to d,,,oce- 89dmoceand dmoc e-~ respectively). Parker et al. (1972) cite some support for these dependancies. Examination of floc size distributions shows many flocs larger than 2k which Parker et al. (1972) conclude must be undergoing surface erosion but are also maintained by high rates of aggregation. The rate of production of particles by surface erosion below the microscale was shown to be
dN/dt = KeF 2,
oo t'*
I
4 8 16 DEFLOCCULATED MODE ( d 4 p m )
(dN/dt oc e)
(16)
where Ke is a coefficient. Somewhat different relationships for floc fracture are proposed by Tambo & Hozumi (1979). They suggest that dmoC,g-3/2(3+kp) in the viscous subrange and dmocg-l/(l+kp ) in the inertial subrange, kp is the exponent in the relationship between floc size and density having a range kp = 1-1.5, so dm OCg -(0"38-0"33) and dmoCg -(0"5-0"4). If we say for the viscous subrange din oc e- 89then dmocU, -l or Zo- 89rather than z -1 suggested by Krone (1962). Their relationships for the viscous range are well supported by data from a stirred flocculator. They found dm~ 1.5 mm at e =2.10 -2 cm 2 s -3 and 200/~m at e = 10 cm 2 s -3 (F = 1.4 and 32 s -~) for clay-aluminium flocs. This suggests that in the generally low shear environment of the sea, flocs of a millimetre and more will be stable
Erosion, transport and deposition of fine-grained marine sediments
53
10 4
o~ 103"
|
E o
SAMPLE I-.z uJ O n"
15
KN83112110,
1000
|
mab
n..
("
)
2
u.= 10
,|
z
~ 101
~ lO.
N=7206d 3 2 3 2
..i
~10 ~
_o 0
'2
1
4.
8
1E;
1#
32
2 I
105.
4
8
16
32
@ (.) (.) (.)
104 .
(.) (.)
SAMPLE
2O
KN83/13/01
, 3 mab
(.)
~E 103,
9
(.) f,ILl
Iz
~
~ 15. LU Q.
Z LLI
"-- 10-
...I
w
rO
i1:
1 0 2,
_>
.~ 5"
101.
10o (.)
0
1(~ 1
1
2
4
8
16
32
dt~'n
,
~)
4.
8
16
',")
32
dlJm
FIG. 18. Examples of high and low concentration oceanic size distributions determined by Coulter counter from McCave (1983). (a) and (c) are particle-volume frequency curves while (b) and (d) are the corresponding cumulative particle-number distributions. while in higher shear zones near the bed 100-200 pm may be the limit. Under depositional conditions zo ~<0.1 Pa yielding, in the viscous sublayer, a shear rate dU/dz ~ 100 s -=. Equating this with the local turbulent shear rate (e/v) 89 in the vicinity of suspended particles suggests that an equivalent condition is e~< 100 cm 2 s -3. Extrapolating the results of Tambo & Hozumi (1979) indicates particles of diameter up to 100 /~m would be stable, and for deposition under Zo=0.031 Pa (u. = 0.55 cm s-=) particles up to 200 #m would be stable.
Numerical models such as that of Ives & Bhole (1973) show that breakup may exert a controlling influence on size distributions. Differing distributions are produced by breakup into 2, 3 or 4 fragments. None of the theoretical treatments yet go beyond total disintegration or breakage in two. Parker et al. (1973), Tambo & Hozumi (1979) and Jeffrey (1982) all consider some floc breakage mechanism involving strain exerted by a turbulent shear flow resisted by floc strength. However, no kernels for breakup (J terms in eqs. 13-15) have yet been proposed.
HIGH CONCENTRATION STA. 13
8
"g"
LOW CONCENTRATION STA. O1
/,,
,
o
~ 100 80
4
60
4
3 ",
-4
/
21.1 , . . . ,
4 0
2
/,, 9~
~..
2o
I 03
,~
0
~
100
/
21m
"~.~'-
~
/ "%,1
~ l
~"q2.6 \
~
"-.X~ -"~"I,,5
i I
I
8o
52. / "-. ._._I
,/
60
-'., ,,.
40
I
',
"""
5m
1 2
~, I
i
i"
2
01 0
u
I 4
I 8
~ 16
2
4
8
16
I 32
/.zm
FIG. 19. Disaggregated high- and low-concentration oceanic size distributions (dashed lines) compared with the original untreated spectra (solid lines) (McCave 1983).
1~176
'
I
.,
E
HINDERED SETTLING "
1.0 0-1
oE\
I
'~
~
9
1.0
I
10
CONCENTRATION
C kg m"3
FIG. 20. Owen's (1970) determinations of settling velocity of estuarine m u d as a function of salinity and concentration.
100
Erosion, transport and deposition of fine-grained marine sediments I
I
I
I
I
/
.
AJ
55
I
/
9
/ 9
9
/
0.1
/
9
9/
.
E E ;=... .m.4-, m
P= =,., 0"01
r .m .i.-,
/..
/.
/
.9
o12
o'.s
/
0.001
o11
I
1
I
2
I
6
C kg m -3
Settling velocity Settling velocity distributions are only reliable if determined in situ or, for shallow water, on deck in the sampler used to get the water (Owen 1971). Accordingly there is very little data available and none from beyond estuaries or the inshore coastal zone. Data from several sources shows that mean settling velocity increases with concentration and salinity. Owen's (1970) laboratory experiments with Avonmouth mud demonstrate the regions of free and hindered settling with a boundary at about 10 kg m-3 (Fig. 20). For waters of normal marine salinities this gives settling velocities of ~0.01 to 0.8 mm s -~ for concentrations of 0.1 to 5 kg m -3. These values are comparable to those determined by Owen (1971) in situ from the Thames estuary where ws = 0.22 mm s- l at 0.1 kg m -3. He also indicated WsOCC2"2 at neap tides but wsocC El at spring tides. Krone (1962), however, suggests wsoc C 4/3 and this agrees quite well with the trend of data from the Chao Phya in Thailand taken by Allersma et al. (1967) (Fig. 21). The
FIG. 21. In situ settling velocity of muds from the Chao Phya as a function of concentration (data points from Allersma et al. 1967) showing a fair agreement with the wsoc C4/3 prediction of Krone (1962). Curve for salinity= 30%0drawn from Owen's (1970) plot (Fig. 20).
latter data demonstrate settling velocities 30 to 70 times greater than those of the freshwater suspended sediment supplied to the estuary. Thus the flocculation effect yields a huge increase in ws. The low end of the Chao Phya data ( ~ 0.02 mm s-l) are comparable with the only available continental shelf values obtained by McCave (1970), from Postma's (1961) raw data, of 0.024-0.096 mm
S-I. A somewhat different presentation is used by the Hydraulics Research Station (1977, 1979a, 1979b) who determine settling velocity curves using the Owen tube (Owen 1971) and take the ratio of the settling velocity at each ten percentile to the mean concentration. For a given estuary these values of ws/C are averaged and a mean curve is produced (Fig. 22). It will be seen that there is a great range in the median (Ws50/C = 0.2 to 10 mm s-J/kg m -3) which partly reflects variation in grain-size and mineralogy. Apart from the high Brisbane value this implies median ws = 0.2-2 mm s-1 for high estuarine concentrations (1 kg m -3) and 0.02-0.2 mm s -1 for lower ones (0.1 kg m-3). The curves of Fig. 22 were
56
I.N. McCave
100 90 80
w
9
BEL
BELAWAN
SCH
SCHELDT
GRA
GRANGEMOUTH
BELBR2
SCH
GRA BR1
BR BRISBANE (1&2) 70 '- 6 0 I.LI
-
50
IZ
,,, 40 0
/ /
a.
/ /
2O
/
10
J
/
0
I
0.01
.....
0.1 SETTLING
VELOCITY
&
I
I
10
/ CONCENTRATION
,
ws/C
I00 -I
, rams
/ kgm- 3
FIG. 22. In situ settling velocity divided by concentration for percentiles at 10% intervals from Hydraulics Research Station (1977, 1979b) and Thorn (1980). (For the Scheldt ws/C 18 is plotted.) based on linear plots of Wsversus (7 for Belawan, Brisbane and Grangemouth mud but Wsoc C 1-8for the Scheldt (Thorn 1981). There are no data on distributions from the outer continental shelf to the deep-sea though descent rates of individual particles have been measured. It seems unlikely that the estuarine relations will extrapolate to low deep-sea concentrations as that would result in Ws--~10 - 4 t o l 0 - 3 mm s-1 or little more than the speed of single 2 ttm particles. The aggregation of deep-sea particles yields higher values. Settling velocities of individual large particles from the deep-sea have been measured by several authors. Values measured by divers or observed from submersibles for sinking speeds of marine snow, flocculent gelatinous particles of 0.5 mm in diameter, are in the region of 50 to 250 m day-l but are of doubtful accuracy (Kajihara 1971; Shanks & Trent 1980; Silver & Alldredge 1981). Far better known are the sinking speeds of crustacean faecal pellets (from copepod and euphausiids) measured by several workers and analysed by Komar et al. (1981). They obtained the expression ws=0.0694(ps-p)g E ~176 V2/3fl - l
(17)
for cylindrical faecal pellets with volume V, Janke (1966) shape factor E and density ps settling in fluid of density p and viscosity p. The range of settling velocities is 17-780 m d a y - l and of mean densities 1.15 to 1.29 with an average of 1.22 g cm -3. Settling velocities of salp faecal pellets is even greater, ranging from 320 to 2238 m d a y with a mean around 1000 m day -t in the
measurements of Bruland & Silver (1981) and Madin (1982). One other relationship worthy of note is Migniot's (1968) curve showing a flocculated to deflocculated settling velocity ratio as a function of primary particle (i.e. deflocculated) settling velocity (or diameter). The relationship shown for C = 10 kg m -3 is ( w f / w a ) = 0 . 4 4 6 wd -~ That is to say the greatest enhancement in settling velocity occurs for the finest particles. The generality of this expression has not been tested at other concentrations or by other authors. Vertical flux In the oceans much of the suspended sediment concentration lies in relatively small sized particles with low descent rates, but the vertical flux is dominated by large, relatively rare particles settling rapidly (McCave 1975). Certain small particles, for example calcareous nannoplankton, could not reach great depths at all (they would dissolve), unless they fell rapidly in aggregates (Honjo 1976). These hypotheses have been confirmed and extended in the large body of work emerging from the deployment of sediment traps from the mid 1970s onwards (for example, Wiebe et al. 1976; Rowe & Gardner 1979; Knauer et al. 1979; Honjo 1980; Dunbar & Berger 1981). The first candidates for the large, rapidly settling particles were faecal pellets. While early work concentrated on pellets from crustaceans (Smayda 1969; Schrader 1971; Fowler & Small 1972; Honjo & Roman 1978), more recent work
Erosion, transport and deposition of fine-grained marine sediments
57
has been on salp faeces (Bruland & Silver 1981: fine suspended sediment is almost uniformly Iseki 1981; Madin 1982) with very high descent distributed with depth. This is in accordance with rates. In addition several types of particle are the predictions of Rouse's suspended load equaalready large and do not require aggregation for tion (Thorn 1975). In depths of a few tens of rapid descent (foraminiferal tests, radiolarian metres when the tide runs up to 0.5 to 1 m s- ~this skeletons and pteropod shells). There has also distribution multiplied by the current velocity been a recent revival of interest in the marine yields fluxes in the region of 103--104 kg/m/tide, snow first described by Japanese workers in the though net fluxes will generally be an order of early 1950s (Suzuki & Kato 1953; Shanks & Trent magnitude less. Inglis & Allen (1957) applied this 1980). This comprises large (0.5-1000 mm) fragile method in the Thames estuary, but they also were aggregates and sheets of mucus. The source of the among the first to note that at high concentramucus is most likely one of the gelatinous zoo- tions the distribution in suspension is not uniform plankton (Gilmer 1972, 1974; Alldredge 1976, because of the formation of a dense near-bed 1977; Alldredge & Madin 1982), though mucus layer of fluid mud. This phenomenon is now produced in the palmelloid stage of coccolitho- known from several estuarine areas where it phorids has also been seen as a possible cause of forms the dominant component of the fine sediaccelerated particle transport (Smayda 1971: ment flux. It is also known from some coastal Honjo 1982). These large snow particles may areas where suspended sediment concentrations carry a significant load of particles acquired first are high, Surinam (Allersma 1966; Wells & in the surface while the mucus was part of a Coleman 1981) and the Bristol Channel (Parker feeding structure (Alldredge & Madin 1982) and & Kirby 1981). secondly acquired through sweeping out particles Farther away from the coast in shelf waters from the water column during descent (Honjo et some sort of stratification is the rule. This noral. 1982a, b; McCave, in press). mally results in surface and bottom layers of In areas where bottom currents are weak this suspended sediment with, in some areas, several vertical flux of large particles is the principal layers at intermediate depths (Drake et al. 1972; means of sedimentation. Little physical modifica- Harlett & Kulm 1972; Meade et al. 1975). Contion of the particles in the boundary layer is centrations are generally in the region 1-10 g m -3 expected and the deposition rate of stable com- but the mid-depth and surface layers may contain ponents should equal their long term vertical flux a large percentage of organic matter. These turbid (Lorenzen et al. 1981). Of course chemical and layers found on the shelf and the deep ocean are biochemical alteration, consumption and dissolu- termed nepheloid (cloudy)-layers. More detailed tion of particles will occur both in the water and examination of shelf layers shows they have a on the bed. Under stronger flows these large structure which at times is related to the bottom particles may be broken up, particularly if the mixed layer (Pak & Zaneveld 1977). However, peritrophic membrane of faecal pellets has been frequently they are only a few metres thick. Few damaged by bacterial activity (Honjo & Roman estimates of material flux exist for shelves, though Pierce & Siegel (1979) estimate an annual off1978). The rapid descent of large particles coupled shelf flux of 48 to 117 • 109 kg from the southern with their biological origin and seasonal surface Argentine Shelf and Drake (1976) gives an annual productivity means that there is now reason to down-canyon loss of 6.9 • 10v kg from the shelf of believe that there may be seasonality at abyssal southern California. depths. Seasonal flux of both biogenic and lithoNepheloid-layers are thought to detach from genic particles to ocean depths has been shown by the bottom at the shelf edge and at points on the Deuser et al. (1981) and Honjo (1982). The supply slope and in canyons where they are stirred up by of lithogenic materials to the surface and mid- internal waves. These internal nepheloid-layers water is not seasonal so its seasonal flux to the bed have been recorded off open shelves (Pak et al. suggests scavenging of biological origin (Honjo 1980), off the slope (Dickson & McCave 1982), 1982; Deuser et al. 1983). In most areas of the and in canyons (Drake 1974; Baker 1976). Materdeep-sea bioturbation removes this seasonal sig- ial settles out of them and is scavenged from them nal from the sediment record, but in a few by faster settling particles coming from surface near-shore restricted basins [e.g. Santa Barbara waters. The large literature on deep ocean nepheloidBasin (Emery 1960) varved sediments are encounlayers contains several examples of the relationtered. ship between nepheloid-layers and bottom mixed layer thermal structure (Biscaye & Eittreim 1974; Distribution in the flow and horizontal flux Armi 1978; McCave et al. 1980; Weatherly & In shallow tidal-shelf seas without stratification Kelley 1982; McCave 1983) (Fig. 23). The bottom
58
I.N. McCave 0 1.80 4700
1.85
i
0.8
0.7 0
i
i
15
06 9
05
04
i
_
300
4800
.J.
200
s
4900
I
100 ::
5000
.:
i -.
9 .,:" :!.'
.~:: .... o," 5100
2
4
8
16
32
,u.m
FIG. 23. Example of a nepheloid layer profiled by a transmissometer shown by attenuation coefficient e and bottom mixed layer demonstrated by potential temperature 0. Suspended sediment size distributions at various points in the flow shown at the right (McCave 1983).
mixed nepheloid-layer is typically 25-100 m thick and has a concentration of 50-500 mg m 3. Exceptional cases have been recorded on the Nova Scotian rise of concentrations up to 12 g m 3 (Biscaye et al. 1980). There is normally a sharp drop in concentration above the lowermost 25 to 100 m of the water column, followed by an exponential decrease to 'clear' water 500 to 1500 m above the bed. The global distribution of suspended sediment in the bottom nepheloidlayer corresponds very well to the path of intensified abyssal western boundary currents (Eittreim et al. 1976) but the load (as distinct from the flux) carried is slight, being generally no more than 30 g m -2. Estimates of sediment fluxes are more difficult and early values are given by Eittreim & Ewing (1972) for the North American Basin
(input 9 x 106 m 3 y - i , output 3 x 106 m 3 y - l ) and by Ewing et al. (1971) for the Argentine Basin. Biscaye & Eittreim (1977) calculate net northward particulate fluxes in the Argentine and Brazil Basins as 1 to 8 x 1 0 9 kg y - l . These transports are potentially subject to large errors, and studies in narrow deep ocean passages such as that of Hogg et al. (1982) yield more reliable data. They measured fluxes of water from the Argentine to the Brazil Basins of 4.05 x 106 m 3 simplying a sediment flux of about 3 x 109 kg y-~. By no means all the records of shelf and deep-sea suspended sediment and flux have been mentioned here, nevertheless the number of such studies is not large. In particular those detailing rates of fine-sediment transport are very few.
Erosion, transport and deposition of fine-grained marine sediments
Deposition Hjulstrom's (1935, 1939) diagram was earlier called misleading. This is most true for its line separating the transportation and deposition fields (Fig. 1). This was produced by extrapolating down to zero from Schaffernak's (1922) lowest point, at a particle diameter of 5 mm, taking account of WelikanotFs (1932) opinion that the curve should fall to very low velocities (Hjulstrom 1935). The starting point at 5 mm is based on the presumption that the lowest transportation velocity for fine gravel was 2/3 of critical erosion conditions. Even that would not now be thought true as Francis' (1973) data shows little difference between initiation and cessation of movement for angular and rounded sand and gravel in water. This deposition curve proved even more perplexing to a few authors who mistook it for a settling velocity curve and replotted the diagram with several versions of settling velocity curves displayed (e.g. Heezen & Hollister 1964). The misleading feature of the diagram was of course to suggest that finegrained sediment would only be deposited at very low mean flow velocities, <~0.15 cm s-~ for 20 ~m silt for example. One could reach the conclusion from this diagram that very fine sediment would never deposit at all. Thus we need to examine critical conditions for deposition and rates of deposition of suspended sediment from the perspectives both of theory and measurement.
Theory and laboratory measurements Einstein & Krone (1962) observed that at C < 300 g m -3 with flow speed below some critical value,
I
the suspended sediment concentration decreased logarithmically with time. A plot of the fractional depositional rate versus bed shear stress revealed no deposition for ro > 0.06 Pa. This is the critical or limiting shear stress for deposition ft. The concentration change is given by C, = Co exp ( - w ,
t p/D)
R = d m / d t = - Cws(1 - ro/rt)
l
I
I
-2
io-3
j~---exfrapolo fed for O = Consl = 0.4
I
-4
IO
I0
I
-3
I0
(19)
Other experimental determinations of ~t using natural muds gave values in the range 0.04 to 0.08 Pa. The difficulty with using natural muds is that they have a wide size range and that they tend to flocculate, so neither the size nor settling velocity of depositing material is known. McCave (1970) and McCave & Swift (1976) suggested that the physical reason for the apparently high threshold for mud deposition (as opposed to the very low one on Hjulstrom's curve) was entrapment of suspended sediment in the viscous sublayer. It was argued that 'bursting' from the sublayer would remove only particles of settling velocity below some critical value, i.e. slow settling particles would not have time to reach the bed before being caught up in a burst and ejected from the sublayer, thus z t = f ( w O . It was then argued that this functional relationship was given, as a minimum, by the critical erosion stress for non-cohesive fine sediment (McCave & Swift 1976). In a flow bearing quartz silt in suspension, reduction of flow speed below the
g
J
(18)
where t is time, D is depth and p is the overall probability of a particle sticking to the bed given by p = ( l - r o / ~ t ) (Einstein & Krone 1962). This expression was put in a form giving mass rate of deposition by Odd & Owen (1972)
jo-I
~fo
59
I
-2
I0 ws rams
I
-I
I
-I
FIG. 24. Critical deposition curve based on critical erosion of non-cohesive sediment from McCave & Swift (1976).
60
I.N. McCave
with D' = 2.7 • 10 - 13m 2 s - 1, R" = 1.88 • 10 - lOmg m -z s -I. Using equation (18) evaluated with C = I 0 mg m -3, Ws=3.6• -7 m s -l and -6 mg m -2 s -L (1-~o/t/)=0.75, R = 2 . 7 • Thus gravitational settling dominates over diffusive processes and must be the principal means of particle transport to the boundary in water, even of very fine (1/~m) particles. The viscous sublayer model has been invoked and elaborated by other workers seeking to explain sorting and formation of laminae in fine-grained sediments (Stow & Bowen 1980; Hesse & Chough 1980). However, Allen (in press) has pointed out that the length scales of lamination and sublayer bursting are so different that the larger scale organization of flow structures affecting the boundary layer detected by Falco (1977) and Head & Bandyopadhyay (1981) is more likely responsible for lamination. It is in this region that the maximum shear stress is encountered by particles, some of which may break up thereby preventing their deposition. However, Gust (1976) has found that suspended sediment concentrations in excess of a few hundred g m -3 may result in drag reduction and thickening of the viscous sublayer. There are clearly many aspects of the behaviour of fine aggregating suspensions in relation to flow structure that remain poorly understood. Extensive experiments by Partheniades and his associates (Partheniades et al. 1968; Mehta & Partheniades 1973) show that for high concentrations, C = 1 to 25 kg m -3, there is some deposition at to > zi and complete deposition for to < tl. The degree of deposition is expressed a s Ceq/Coand is a function of to (Fig. 25) where Co is the initial concentration and Ceq is the final concentration reached exponentially after a long time (2-20 hours). Ceq/Co is log-normally distributed against
critical erosion velocity would lead to cessation of saltation in the sublayer, reversal of the concentration gradient and thus deposition of the suspended material. In fact, if the material could stick to the bottom, the limiting shear stress might be even higher. This argument suggested the critical deposition curve given as Fig. 24 based mainly on the measurements of White (1970) but now augmented by those of Mantz (1977) and Unsold (1982). There are no published results relating limiting shear stress for deposition to the settling velocity of non-flocculating particles. This case should be examined before the more complicated one of natural muds. The viscous sublayer model presumes that particles reach the boundary by gravitational settling. In deposition of aerosols, transport to the boundary by Brownian diffusion and by eddy diffusion is important (Friedlander 1977). The rate of deposition is given by (20)
R' = - (O' + E') 6C/6z
where D' = k T / 3 n # d is the Brownian diffusivity and E'=v(z+/14.5) 3, in which z + =zu./v, is the eddy diffusivity. The region for which Brownian diffusion will be important is the diffusive sublayer 6b = (1 Or~u.) (D/v)~. For deep sea conditions and d = 1 ~m, (u. = 1 mm s-l, # = 1.5 10-3 Pa s, v = 1.43 x 10 -6 m 2 s-2), 6b=0.082 ram, a small fraction of the viscous sublayer thickness of 14.3 mm. The eddy diffusivity decreases rapidly into the viscous sublayer and D ' = E " at z + =0.083 or z = 120 #m. Thus Brownian diffusion is the most important diffusive process within 100/tm of the bed (for the deep-sea case with very fine particles). But how does this relate to settling flux to the boundary? Assuming a concentration gradient of 0 up to 10 mg m -3 of 1 /~m particles in the sublayer 14.3 mm thick 6C/6z is 700 mg m-4 and
1.0
I
I
I
I
I
i
I
ol
I
J 0.8
9 o o~162 q~ o ~ 9
0.6
9e
9
,:"
Ce._.]q Co
0.4
.2'_".-
.,/'.'e/ee~
0.2
O 0
...........
I 0.2
I O.4
I
I 0.6
I
1 0.8
I
I 1.0
FIG. 25. Change in concentration at equilibrium with shear stress demonstrating deposition for "co> "clat concentrations 1 kg m -3 (Mehta & Partheniades 1973).
Erosion, transport a n d deposition o f fine-grained marine sediments ('Co--Zt)/zt and against t/tso where t is time and ts0 is the time at which Ceq/Co= 50%. Thus there is a fraction of a given mud that will deposit at shear stresses well above the critical deposition stress (as defined by Einstein & Krone (1962)). This is of importance in estuarine deposition and the origins of fluid mud, but it also casts further doubt on critical erosion experiments where there is a large suspended load. There are no measurements of the properties of the material selectively deposited in this way. It is likely that only the strongest aggregates are deposited at the higher values of stress and that those particles capable of forming only weak bonds are continually disrupted near the bed and thus do not deposit. The process of deposition at the high values of stress (e.g. ~o= 0.6 Pa which could move 1 mm sand) must involve adhesion of sediment to the bed by electrostatic forces. Thus it seems likely that a bed deposited under such high shear conditions would have a much greater strength and critical erosion shear stress than one deposited under a slower flow. Field measurements
A large amount of the sediment in suspension in shelf and deep ocean waters is resuspended. Thus the short term deposition rate calculated from equation (18) rarely agrees with that obtained from radionuclides. In shallow marine environments where wave stresses are intermittently high, prediction of the net rate of deposition would require coupled models for fluid shear stress and flux and for rates of erosion and deposition. These exist for some estuarine areas, for example the Thames (Odd & Owen 1972), but not yet for shelves or the deep-sea, though progress is being made (Nowell et al. 1982). Thus we are forced back to measurement of accumulation rate. The radionuclides in sediments used for dating them are emplaced in two ways. Part of the signal at a given depth below the sediment surface arrives by being mixed down by biological activity and part arrives by the sediment surface moving up, by deposition. In conditions of no burrowing, for example in Santa Barbara Basin, the signal is entirely advective. But in most cases it is partly advective, partly bio-diffusive. In steady state this is expressed by
f v OpsU~ ~3psU -~xkr,~ff~-x j - ~ - ~ x + P s P - 2 p s N = 0
(21)
where KB is the biological diffusivity, S the sedimentation rate (LT-~), ps the sediment bulk density, N atoms g- 1 of nuclide and the third and fourth terms give nuclide production and decay
6I
rates (Goldberg & Koide 1962). The 21~ method has achieved great popularity recently because of its 22 year half-life. However, work with several tracers of differing half-lives shows that the biological diffusivity is not constant but decreases with depth. Sedimentation rates in Long Island Sound originally thought to be 1.1 mm y-~ on the basis of 21~ were halved to 0.5 mm y-~ when the diffusivity in the surface layer was found from 234Th (t~= 24d) profiles (Benninger et al. 1979). While 2t~ profiles may show a surface mixed layer underlain by a region of 21~ decrease, that decrease may not be due only to 2~~ decay. It may also contain some mixing effects on a rather longer time-scale. Thus the multi-layer model of Olsen et al. (1981) needs results from two or three nuclides to establish both depth-dependent diffusivities and deposition rates. Radionuclides with short half-lives are best in areas of rapid sediment accumulation. For example Kuehl et al. (1982) have determined rates from 6.5 to over 20 mm y-~ off the Amazon and Nittrouer et al. (1979) give 1-4 mm y-i (2-14 kg m-2 y-l) in the mid-shelf mud belt of the Washington Shelf using 21~ However, Bothner et al. (1981) found biological mixing too severe for that method and gave 0.2-0.5 mm y-1 using 14C for the mud belt off southern New England. Several measurements have been made on Long Island Sound sediments, giving values from 0.5 mm y-l to 3 mm y-l using zl~ 234Th and 7Be (Benninger et al. 1979; Aller et al. 1980; Krishnaswami et al. 1980). The advantage of using these short-lived isotopes is that short term rates of deposition can be determined. 2~~ (t 89 138d) and the bomb-produced ~37Csare also valuable in this respect. A comparison between them, the longer term rates given by ~4C and calculations based on equation (18) would give some insight into the amount of resuspension involved in the formation of shelf mud deposits. The same may also be possible in the deep-sea though few areas are likely to give fast enough rates of deposition. A notable exception is on the Nova Scotian rise where the creation of bedforms is believed to occur on a time-scale of a few weeks.
Conclusions Modern marine muds are complex mixtures of organic and inorganic components. Most experimental work has sought to characterize them in physical or chemical terms--Bingham yield strength, concentration, cation exchange capacit y - a n d relate these parameters to critical erosion stress and erosion rate constants. This approach
62
I.N. McCave median settling velocities generally fall in the range 0.05-0.5 mm s-~ an increase of up to two orders of magnitude over that for the disaggregated population. Shelf suspensions have velocities that fall at the lower end of this range, but there are few determinations and none at all for the deep sea. Aggregates are both produced and broken up by fluid shear. The breakup is rather poorly known at present but it may restrict aggregate sizes during deposition to be less than about 50/~m unless they have strong biological binding. The process will perform some sorting function so that only particles bound by strong bonds will be deposited in high shear zones. On this basis some clay mineral sorting may also be achieved by resuspension. Deposition is dominated by the concentration of the suspension and the settling velocity of the particles. The parameters all interact as settling velocity and critical deposition stress are functions of concentration. Nevertheless given these properties, it is reasonably predictable and can be incorporated in numerical models. Rates of shelf deposition would fall around 0.5 to 0.05 mg m -2 s-~ while deep-sea rates would be predicted to be
has had some success with some investigations in estuaries but has been less successful at shelf and oceanic depths because it is difficult to obtain in situ measurements comparable with those taken in the laboratory. Sediments in situ may have an important element of organic binding and their slow deposition rates render them most unlike the rapidly deposited slurries used in the laboratory. Thus we have a good idea of the general level of erosion stress (0.05-2 Pa) and erosion rate constant (10-4--2.10 -3 kg m -2 s -l Pa -l) but no good way of predicting it. It appears likely that deposits from fast flowing water will have a high erosion stress and that fast flowing water will 'harden' a previously deposited bed. Experiments in the field and laboratory currently being conducted will show the extent and importance of organic binding on clastic sediments. Such experiments have already shown the significance of sub-tidal algal mats on carbonates. Many changes occur to muds during transport but the most important is aggregation following particle collision by physical and biological means which increases particle settling velocity. For fully flocculated estuarine suspensions,
1000
I
% TRANSPORT SUSPENDED
lOO
AS LOAD
> W
~5
?0 ,0 k.
/ 10 F
I
tT I~
O
NO
MOVEMENT
I~n
t !
1.0 lffrn
tO0/~m
"lO/.~m
Imm
lOrnrn
GRA/N S I Z E d L
10-3
I
I
I
10 -2
10 -I
1.0
I
10
I
50
I
100
I
J
I
I
200 300 400 500
J
rnrns -7
SETTLING VELOCITY % F[G. 26. Proposed transport and deposition diagram for fine suspended sediment with erosion and transport fields ['or coarser materials.
Erosion, transport and deposition of fine-grained marine sediments one or two orders of magnitude less. Departure from this reflects variability in the environment; both of the depositing suspension's concentration and of the temporal distribution of resuspension. Earlier authors have sought to incorporate both erosion and deposition on one diagram (Hjulstrom 1935; Sundborg 1956; Postma 1967). However, as observed above, while deposition may be a function of settling velocity and thus size, erosion is not and so one cannot plot erosion and deposition on the same diagram, except for non-cohesive sediment. One can plot transport and deposition, however, and this is done in Fig. 26. The suspension transport curves are based on the exponent in the suspended load equation ( = w/tcu, describing the gradient of suspension concentration (where ~,-=0.4 is the von Karman constant), following the suggestion made first by Inman (1949) and subsequently taken up by others (Sundborg 1956). The critical deposition is the same as the critical erosion curve from White (1970) and Mantz (1977) following the arguments
63
of McCave & Swift (1976). The deposition field is contoured with lines of equal deposition rate divided by concentration giving a ~deposition velocity'. For erosion a separate diagram is required; however, it will appear from Figs 1-8 that no unique basis for such a diagram has yet been devised. ACKNOWLEDGEMENTS: In a synthesis of this type one draws heavily on the work of colleagues and other workers in the field. My sincere thanks go to them and my apologies in cases where credit is inadequately given. The manuscript benefited from reading by Drs. A.R.M. Nowell, R. Parker and J.P.M. Syvitski who have my thanks for the onerous chore. Thanks also to Barbara Slade who persuaded our word processor to accept an increasingly bulky manuscript. Finally I am grateful to the University of East Anglia and the US Office of Naval Research (Contract N00014-82-C-0019 NR083-004) for their support.
References ALLDREDGE, A.L. 1976. Field behaviour and adaptive strategies of appendicularians. Marine Biol., 38, 29 39. 1977. House morphology and mechanisms of feeding in the Oikopleuridae (Tunicata, Appendicularia). J. Zool., 181, 175-88. & MADIN, L.P. 1982. Pelagic tunicates: unique herbivores in the marine plankton. Bioscienee. 32, 655-63. ALLEN, J.R.L., in press. Fluid dynamics and sediment structures. In: Brenchley, P. & Williams, B.J. (eds) Sedimentology--Current Concepts and Applied Aspects. Geol. Soc. Lond. Spec. Pub. ALLER, R.C., BENNINGER,L.K. & COCHRAN,J.K. 1980. Tracking particle-associated processes in nearshore environments by use of 234Th/238U disequilibrium. Earth planet. Sci. Lett., 47, 161-75. ALLERSMA, E., HOEKStRA, A.J. & BIJKER, E.W. 1967. Transport patterns in the Chao Phya estuary. Delft Hydraulics Lab. Publ. no. 47, 24pp. AmAtHURAI, R. & ARULANANDAN,K. 1978. Erosion rates of cohesive soils. J. Hydraulics Die., Proc. Am. Soc. cir. Engrs., 104, 279-83. ARml, L. 1978. Some evidence for boundary mixing in the deep ocean. J. geophys. Res., 83, 1971-9. ARULANANDAN,K. 1975. Fundamental aspects of erosion of cohesive soils. J. Hydraulics Div.. Proc. Am. Soc. cir. Engrs., 101 (HY5), 635-9. BADER,H. 1970. The hyperbolic distribution of particle sizes. J. geophys. Res., 75, 2822-30. BAKER,E.T. 1976. Distribution, composition and transport of suspended particulate matter in the vicinity of Willapa submarine canyon, Washington. Bull. geol. Soe. Am., 87, 625-32. - - , FELLY, R.A. & TAKAHASHI,K. 1979. Chemical composition, size distribution and particle morpho-
-
-
logy of suspended particulate matter at DOMES sites A, B and C: relationships with local sediment composition. In: Bischoff, J.L. & Piper, D.Z. (eds), Marine Geology and Oceanography of the Pacific Manganese Nodule Province, Plenum Press, New York. 163-201. BALE, A.J., MORRIS, A.W. & HOWLAND, R.J.M., in press. Measuring the size characteristics of suspended particles in an estuary by laser Fraunhofer diffraction. In: Parker, W.R. & Kinsman, D.J.J. (eds), Transfer Processes in Cohesive Sediment Systems. Plenum Press, New York. BENNINGER, L.K., ALLER, R.C., COCHRAN, J.K. & TUREKIAN,K.K. 1979. Effects of biological sediment mixing on the 21~ chronology and trace metal distribution in a Long Island Sound sediment core. Earth planet. Sci. Lett., 43, 241-59. BIDDLE, P. 8r MILES, J.H. 1972. The nature of contemporary silts in British estuaries. Sed. Geol. 7, 23-33. BISCAYE, P.E. & EITTREIM, S.L. 1974. Variations in benthic boundary layer phenomena: nepheloid layer in the North American Basin. In: Gibbs, R.J. (ed.), Suspended Solids in Water. Plenum Press, New York. 227-60. -& -1977. Suspended particulate loads and transports in the nepheloid layer of the abyssal Atlantic, Marine Geol., 23, 155-72. - - , GARDNER,W.D., ZANEVELD,J.R.V., PAK, H. & TUCHOLKE, B.E. 1980. Nephels! have we got nephels! EOS, Trans. Am. Geophys. Un., 61, 1014 (abstract). DE BOER. P.L. 1981. Mechanical effects of microorganisms on intertidal bedform migration. Sedimentolog:v, 28, 129-32. BOTHNER,M.H., SPIKER,E.C., JOHNSON,P.P., REYDIGS, R.R. & ARUSCAVAGE,P.J. 1981. Geochemical evi-
64
I.N. McCave
dence of modern sediment accumulation on the continental shelf off southern New England. J. sed. Petrol., 51, 281-92. BRULAND,K.W. AND SILVER,M.W. 1981. Sinking rates of faecal pellets from gelatinous zooplankton (salps, pteropods, doliolids). Marine Biol., 63, 295-300. BRUN-COTTAN, J.C. 1971. Etude de la granulometrie des particules marines, measures effectuees avec un compteur Coulter. Cah. Oceanogr. 23, 193-205. - 1976. Stokes settling and dissolution-rate model for marine particles as a function of size distribution. J. geophys. Res., 81, 1601-6. BUCHAN, S., FLOODGATE, G.D. & CRISP, D.J. 1967. Studies on the seasonal variation of the suspended matter in the Menai Straits. I. The inorganic fraction. Limnol. Oceangr., 12, 419-31. BUSCH, P.L. & W. STUMM, 1968. Chemical interactions in the aggregation of bacteria: bioflocculation in waste treatment. Environ. Sci. Technol., 2, 49-53. CARLSON, E.J. & ENGER, P.F. 1963. Studies of tractive forces of cohesive soils in earth canals. U.S. Bureau of Reclamation, Denver, Hydraulic Branch Report, 504.
COLES, S.M. 1979. Benthic microalgal populations on inter-tidal sediments and their role as precursors to salt marsh development. In: Jeffries, R.L. & Davy, A.J. (eds), Ecological Processes in Coastal Environments. Blackwell, Oxford. 25~12pp. DEUSER, W.G., Ross, E.H. & ANDERSON, R.F. 1981. Seasonality in the supply of sediment to the deep Sargasso Sea and implication for the rapid transfer of matter to the deep ocean. Deep Sea Res., 28, 495-505. , BREWER,P.G., JICKELLS,T.D. & COMMEAU,R.F. 1983. Biological control of the removal of abiogenic particles from the surface ocean. Science, 219, 388-91. D1CKSON, R.R. & I.N. MCCAVE, 1982. Properties of nepheloid layers on the upper slope west of Porcupine Bank. Int. Council Explor. Sea, CM 1982/C:2, IIydrography Committee, 9pp. D1EPHUIS, J.G.H.R. 1966. The Guiana coast. T(jdschr. Kon. Ned. Ardrijksk. Gen., 83, 145-52. DRAKE, D.E. 1976. Suspended sediment transport and mud deposition on continetal shelves. In: Stanley, D.J. & Swift, D.J.P. (eds), Marine Sediment Transport and Environmental Management. Wiley, New York, 127-58. DRAKE, D.E., KOLPACK, R.L. & FISCHER, P.J. 1972. Sediment transport on Santa Barbara-Oxnard Shelf, Santa Barbara Channel, California. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment transport: Process and pattern. Dowden, Hutchinson & Ross, Stroudsburg, PA. 307-32. DUNBARR.B. & BERGER,W.H. 1981. Fecal pellet flux to modern bottom sediment of Santa Barbara Basin (California) based on sediment trapping. Bull. geol. Soc. Am., 92, 212-18. DUNN, I.S. 1959. Tractive resistance of cohesive channels. J. Soil Mech. Found. DIE., Proc. Am. Soc. cir. Engrs., 85 (SM3). EDZWALD, J.K., UPCHURCH, J.B. & O'MELIA, C.R. 1974. Coagulation in estuaries. Environ. Sci. Technol., 8, 58-63.
EINSTEIN, H.A. & KRONE, R.B. 1962. Experiments to determine modes of cohesive sediment transport in salt water. J. geophys. Res., 67, 1451-61. EISMA, D. & G1ESKES,W.W.C. 1977. Particle size spectra of non-living suspended matter in the southern North Sea. Netherlands Institute .for Sea Research, Internal Report 1977-7, 22pp. & KALE, J. 1979. Distribution and particle size of suspended matter in the Southern Bight of the North Sea and the eastern Channel. Neth. J. Sea. Res., 13, 298-324. EITTREIM, S.L. & M. EWING, 1972. Suspended particulate matter in the deep waters of the North American Basin. In: Gordon, A.L. (ed.), Studies in Physical Oceanography. Gordon & Breach, New York. 123-67. - - - , THORNDIKE, E.M. & SULLIVAN, L. 1976. Turbidity distribution in the Atlantic Ocean. Deep Sea Res., 23, 1115-27. EKMAN, J.E., NOWELL, A.R.M. & JUMARS, P.A. 1981. Sediment destabilisation by animal tubes. J. mar. Res., 39, 361-74. EMERY, K.O. 1960. The Sea off Southern California: A Modern Habitat of Petroleum. John Wiley, New York. 366pp. ERNISSEE,J.J. & ABBOTT,W.H. 1975. Binding of mineral grains by a species of Thalassiosira. In: Simonsen, R. (ed.), Third symposium on recent and fossil marine diatoms, Kiel 1974, Nova Hedwigia, Beiheft, 53, 241-8. EWING, M., EITTREIM, S., EWING J. & LE PICHON, X. 1971. Sediment transport and distribution in the Argentine Basin: part 3, Nepheloid layer and processes of sedimentation. In: Ahrens, L.H. (ed.), Physics and Chemistry of the Earth, 8, Pergamon Press. 49-77. FALCO, R.E. 1977. Coherent motions in the outer region of turbulent boundary layers. Phys. Fluids, Suppl., 20, 124-30. FOWLER, S.W. & SMALL, L.F. 1972. Sinking rates of euphausiid faecal pellets. Limnol. Oceanogr., 17, 293-96. FRANCIS, J.R.D. 1973. Experiments on the motion of solitary grains along the bed of a water-stream. Proc. Roy. Soc. Lond., A332, 443-71. FRIEDLANDER, S.M. 1965. The similarity theory of the particle size distribution of the atmospheric aerosol. In: Spurny, K. (ed.), Aerosols, Physical Chemistry and Applications, Czech. Acad. Sci., Prague. 115-30. 1977. Smoke, Dust and Haze: Fundamentals of Aerosol Behaviour. John Wiley, New York. 317pp. FROSTICK, L.E. & MCCAVE, I.N. 1979. Seasonal shifts of sediment within an estuary mediated by algal growth. Estuar. coast, mar. Sci., 9, 569-76. GIBBS, H.J. 1962. A study of erosion and tractive force characteristics in relation to soil mechanics properties. U.S. Bureau of Reclamation, Report Em-643. GILMER, R.W. 1972. Free-floating mucus webs: a novel feeding adaptation for the open ocean. Science, 176, 1239-40. - 1974. Some aspects of feeding in Thecosomatous Pteropods. J. Exp. Mar. Biol. Ecol., 15, 127-44. GOLDBERG, E.D. & KOIDE, K. 1962. Geochronological
Erosion, transport and deposition of fine-grained marine sediments studies of deep-sea sediments by the ionium-thorium method. Geochim. eosmochim. Acta, 26, 417-450. GOULEAU, D. 1976. Le role des diatomees benthiques dans l'engraissement rapide des vasieres Atlantiques decouvrantes. C. R. hebd. Seanc. Acad. Sci., Paris, D283, 21-3. GRANT, W.D., BOYER,L.F. & SANFORD, L.P. 1982. The effects of bioturbation on the initiation of motion of intertidal sands. J. mar. Res., 40, 659-77. GREGORY, J. 1978. Effects of polymers on colloid stability. In: Ives, K.J. (ed.), The Scientific Basis of Flocculation. Sijthoff & Noordhoff, Alphen a.d. Rijn. 101-39. GULARTE, R.C., KELLY, W.E. & NACC1, V.A. 1980. Erosion of cohesive sediments as a rate process. Ocean Engrg., 7, 539-51. GUST, G. 1976. Observations on turbulent-drag reduction in a dilute suspension of clay in sea water. J. Fluid Mech., 75, 29-47. HARTLETT, J.C. & KULM, L.D. 1973. Suspended sediment transport on the northern Oregon continental shelf. Bull. geol. Soc. Am., 84, 3815-26. HARRIS, J.E. 1977. Characterisation of suspended matter in the Gulf of Mexico--II. Particle size analysis of suspended matter from deep water. Deep Sea Res., 24, 1055-61. HAVEN, D.S. & MORALES-ALAMO,R. 1972. Biodeposition as a factor in sedimentation of fine suspended solids in estuaries. In: Nelson, B.W. (ed.) Environmental Framework of Coastal Plain Estuaries., Geol. Soc. Am. Mem., 133, 121-30. HEAD, M.R. & BANDYOPADHYAY,P. 1981. New aspects of turbulent boundary-layer structure. J. Fluid Mech., 107, 297-338. HEEZEN, B.C. & HOLLISTER, C.D. 1964. Deep-sea current evidence from abyssal sediments. Marine Geol., 1, 141-74. HESSE, R. & CHOUGH, S.K. 1980. The Northwest Atlantic Mid-Ocean Channel of the Labrador Sea: II. Deposition of parallel laminated leveemuds from the viscous sublayer of low density turbidity currents. Sedimentology, 27, 697-711. HJULSTROM, F. 1935. The morphological activity of rivers as illustrated by the river Fyris. Bull. geol. Inst. Upsala, 25, 221-527. 1939. Transportation of detritus by moving water. In: Trask, P.D. (ed.), Recent Marine Sediments. Am. Ass. Petrol. Geols, Tulsa, 5-31. HOGG, N., BISCAYE,P., GARDNER, W. & SCHM1TZ,W.J. 1982. On the transport and modification of Antarctic bottom water in the Vema Channel. J. mar. Res., 40 Supplement, 231-63. HOLLAND, A.F., ZINGMARK, R.B. & DEAN, J.M. 1974. Quantitative evidence concerning the stabilisation of sediment by benthic diatoms. Marine Biol., 27, 191-96. HONJO, S. 1976. Coccoliths: production, transportation and sedimentation. Mar. Micropaleont., l, 65-79. 1980. Material fluxes and modes of sedimentation in the mesopelagic and bathypelagic zones. J. mar. Res., 38, 53-97. 1982. Seasonality and interaction of biogenic and -
-
-
-
-
-
65
lithogenic particulate flux at the Panama Basin. Science, 218, 883-4. & ROMAN, M.R. 1978. Marine copepod fecal pellets; production, preservation and sedimentation. J. mar. Res., 36, 45-57. - - - , MANGANIN1, S.J. & POPPE, L.J. 1982a. Sedimentation of lithogenic particles in the deep ocean. Marine Geol., 50, 199-220. - - , SPENCER,D.W. & FARR1NGTON,J.W. 1982b. Deep advective transport of lithogenic particles in Panama Basin. Science, 216, 516-8. HUNT, J.R. 1980. Prediction of oceanic particle size distributions from coagulation and sedimentation mechanisms. In: Kavanaugh, M.C. & Leckie, J.O. (eds), Particulates in Water. American Chemical Society, Advances in chemistry series No. 189, 243-57. -1982. Self similar particle-size distributions during coagulation: theory and experimental verification. J. Fluid Mech. 122, 169-86. HUNTER, K.A. & L~SS, P.S. 1982. Organic matter and the surface charge of suspended particles in estuarine water. Limnol. Oceanogr., 27, 322-35. HYDRAULICS RESEARCH STATION, 1977. Properties of Grangemouth mud. HRS, Wallingford, Report EX781, 13pp. 1979a. Port of Brisbane siltation study. Fifth report: properties of Brisbane Mud. HRS, Wallingford, Report EX860, 18pp. -1979b. Properties of Belawan mud. HRS, Wallingford, Report EX880, 15pp. INGLIS, C.C. & ALLEN, F.H. 1975. The regimen of the Thames estuary as affected by currents, salinities and river flow. Proc. Inst. civil. Eng., 7, 827-78. INMAN, D.L. 1949. Sorting of sediments in the light of fluid mechanics. J. sed. Petrol., 19, 51-70. ISEKI, K. 1981. Particulate organic matter transport to the deep sea by salp faecal pellets. Mar. Ecol.-Progr. Ser., 5, 55-60. IvEs, K.J. & BI4OLE,A.G. 1973. Theory of flocculation for continuous flow system. J. Environ. Eng. Div., Proc. Am. Soc. cir. Engrs., 99, (EEl), 17pp. JANKE, N.C. 1966. Effects of shape upon the settling velocity of regular convex geometric particles. J. sed. Petrol., 36, 370-6. JEFFREY, D.J. 1982. Aggregation and breakup of clay floes in turbulent flow. Adv. Colloid. Interface Sei., 17, 213-8. JORGENSEN, C.B. 1966. Biology of Suspension Feeding. Pergamon Press, Oxford. KAJIHARA, M. 1971. Settling velocity and porosity of large suspended particles. J. Oceangr. Soc. Japan, 27, 158-62. KIRBY, R. & PARKER, W.R. 1983. Distribution and behaviour of fine sediment in the Severn Estuary and inner Bristol Channel, U.K. Can. J. Fish. Aquat. Sci., 40 (suppl. 1), 83-95. KNAUER, G.A., MARTIN, J.H. & BRULAND,K.W. 1979. Fluxes of particulate carbon, nitrogen and phosphorous in the upper water column of the northeast Pacific. Deep Sea Res., 28, 921-36. KOMAR, P.D., MORSE, A.P., SMALL, L.F. & FOWLER, S.W. 1981. An analysis of sinking rates of natural
66
I.N. McCave
copepod and euphausiid faecal pellets. Linmol. Oceanogr., 26, 712-80. KRANCK, K. 1973. Flocculation of suspended sediment in the sea. Nature, Lond., 246, 348-50. -1975. Sediment deposition from flocculated suspensions. Sedimentology, 22, 111-23. 1981. Particulate matter grain-size characteristics and flocculation in a partially mixed estuary, Sedimentology, 28, 107-14. KRISHNASWAMI, S., BENNINGER, L.K., ALLER, R.C. & VON DAMM, K.L. 1980. Atmospherically-derived radionuclides as tracers of sediment mixing and accumulation in near-shore marine and lake sediments: evidence from 7Be, 21~ and 239,24~ Earth planet. Sci. Lett., 47, 307-18. KRONE, R.V. 1959. Second annual report on silt transport studies utilizing radioisotopes. Hydraulic Engineering and Sanitary Engineering Research Laboratory, University of California, Berkeley. 123pp. -1962. Fhmw studies of the transport q[sediment in estuarial shoaling processes. Hydraulic Engineering and Sanitary Engineering Research Laboratory, University of California, Berkeley. 110pp. 1963. A study of rheologic properties of estuarial sediments. Sanitary Engineering Research Laboratory Report 63-8, University of California. Berkeley, 91 pp. 1972. A .field study of flocculation as a fiwtor in estuarial shoaling processes. Tech. Bull. 19, U.S. Army Corps of Engineers Committee on Tidal Hydraulics, Waterways Experiment Station, Vicksburg, Miss. 62pp. 1976. Engineering interest in the benthic boundary layer. In: I.N. McCave, (ed.), The Benthic Boundary Layer. Plenum Press, New York. 143-56. 1978. Aggregation of suspended particles in estuaries. In: Kjerfve, B. (ed.), Estuarine Transport Processes. University of South Carolina Press, Columbia, S.C. 177-90. KUEHL, S., NITTROUER,C.A. & DE MASTER, D.J. 1982. Modern sediment accumulation and strata formation on the Amazon continental shelf. Marine Geol., 49, 279-300. LAMBERT, C.E., JEHANNO, C., SILVERBERG,N., BRUNCOTTAN, J,C. 81, CHESSELET, R. 1981. Log-normal distributions of suspended particles in the open ocean. J. mar. Res., 39, 77-98. LERMAN, A. 1979. Geochemical Processes." Water and Sediment Environments. John Wiley, New York. 481pp. , CARDER, K.L. & BETZER, P.R. 1977. Elimination of fine suspensoids in the oceanic water column. Earth planet. Sci. Lett., 37, 61-70. LEWIN, J., COLVIN, J.R. & MCDONALI), K.L. 1980. Blooms of surf-zone diatoms along the coast of the Olympic Peninsula, Washington. XII. The clay coat of Chaetoceros armature T. West. Botanica Marina, 23, 333-41. LONSI)ALE, P. & SOUTHARD, J.B. 1974. Experimental erosion of North Pacific red clay. Marine Geol., 17, M25-M34. LORENZEN, C.J., SHUMAN, F.R. & BENNETT,J.T. 1981. In situ calibration of a sediment trap. Linmol. Oceanogr., 26, 580-4. -
-
-
-
-
-
-
-
LYKLEMA, J. 1978. Surface chemistry of colloids in connection with stability. In: Ives, K.J. (ed.), The Scientific Basis of Flocculation, Sijthoff & Noordhoff, Alphen a.d. Rijn. 3-36. LYLE, W.M. & SMERDON, E.T. 1965. Relation of compaction and other soil properties to erosion and resistance of soils. Trans. Am. Soc. Agric. Engrs., 8. MAD1N, L.P. 1974. Field observations on the feeding behaviour of Salps (Tunicata: Thaliacea). MarhTe Biol., 25, 143-7. -1982. Production, composition and sedimentation of salp fecal pellets in oceanic waters. Marine Biol., 67, 39-45. MANTZ, P.A. 1977. Incipient transport of fine grains and flakes by fluids--extended Shields diagram. J. Hydraulics Dir., Proc. Am. Soc. cir. Engrs., 103 (HY6), 601-15. -1978. Bedforms produced by fine cohesionless, granular and flakey sediments under subcritical water flows. Sedimentology, 25, 83-103. MCCAvE, I.N. 1970. Deposition of fine-grained suspended sediment from tidal currents. J. geophys. Res., 75, 4151-9. 1975. Vertical flux of particles in the ocean. Deep Sea Res., 22, 491-502. 1978. Sediments in the abyssal boundary layer. Oceanus, 21 (1), 27-33. 1979a. Suspended sediment, Chapter 6. In: Dyer, K.R. (ed.), Estuarine Hydrography and Sedimentation. Cambridge University Press. 13 l-185. 1979b. Suspended material over the central Oregon continental shelf in May 1974. I: Concentration of organic and inorganic components. J. sed. Petrol., 49, 1181-94. 1981. Location of coastal accumulation of fine sediments around the Southern North Sea. In: Postma, H. (ed.), Proceedings of the Workshop on Sediments and Pollutant Exchange. Rapp. ProcesVerbaux Reun. Cons. Int. Explor. Mer., 181, 15-27. 1983. Particulate size spectra, behaviour and origin of nepheloid layers over the Nova Scotian Continental Rise. J. Geophys. Res., 88, 7647-66. --, in press. Size spectra and aggregation of suspended particles in the deep ocean. Deep-Sea Res., 31. --, LONSDALE, P.F., HOLLISTER, C.D. & GARDNER, W.D. 1980. Sediment transport over the Hatton and Gardar contourite drifts. J. sed. Petrol., 50, -
-
-
-
-
-
1
0
4
9
-
6
2
.
& SWlFT, S.A. 1976. A physical model for the rate of deposition of fine-grained sediments in the deep sea. Bull. geol. Soc. Am., 87, 541-6. MEADE, R.H. 1972. Transport and deposition of sediments in estuaries. In: Nelson, B.W. (ed.), Environmental Framework of Coastal Plain Estuaries. Geol. Soc. Am. Mem., 133, 91-120. - - , SACHS, P.L., MANHEIM,F.T., HATHAWAY,J.C. t~ SPENCER, D.W. 1975. Sources of suspended matter in waters of the Middle Atlantic Bight. J. sed. Petrol., 45, 171-88. MEADOWS, P.S. & ANDERSON, J.G. 1968. Microorganisms attached to marine sand grains. J. Mar. Biol. Ass. UK, 48, 161-175. MEHTA, A.J. & PARTHENIADES, E. 1973. Depositional -
-
Erosion, transport and deposition o/ifine-grained marine sediments Behaviour of Cohesive Sediments. Coastal and oceanographic engineering laboratory, University of Florida, Gainesville, Tech. Rept., 16, 274pp. MIGNIOT, C. 1968. Etude des proprietes physiques de differents sediments tres fins et de leur comportement sous des actions hydrodynamiques. La Houille Blanche, 7, 591-620. MILLER, M.C., MCCAVE, I.N. & KOMAR, P.D. 1977. Threshold of sediment motion under unidirectional currents. Sedimentology, 24, 507-27. MITCHELL, J.K. 1964. Shearing resistance of soils as a rate process. J. Soil Mech. Found. Div., Proc. Am. Soc. cir. Engrs., 90 (SM1), 29-61. -1976. Fundamentals of Soil Behaviour. John Wiley, New York. 422 pp. , CAMPANELLA,R.G. & SINGH, A. 1968. Soil creep as a rate process. J Soil Mech. Found. Div., Proc. Am. Soc. cir. Engrs., 94, (SM1), 231-53. MULLIN, M.M. 1980. Interactions between marine zooplankton and suspended particles, a brief review. In: Kavanaugh, M.C. & Leckie, J.O. (eds), Particulates in Water. American Chemical Society, Advances in Chemistry Series, No. 189, 233-41. NEUMAN, A.C., GEBELE1N,C.D. & SCOFFIN,T.P. 1970. The composition structure and erodability of subtidal mats, Abaco, Bahamas. J. sed. Petrol., 40, 274-97. NEWTON, A.J. & GRAY, J.S. 1972. Seasonal variation of the suspended solid matter off the coast of North Yorkshire. J. Mar. Biol. Ass. UK., 52, 33-47. NIEHOFF, R.H. & LOEB, G.I. 1972. The surface charge of particulate matter in seawater. Limnol. Oceanogr., 17, 7-16. & LOEB,G.I. 1974. Dissolved organic matter in sea water and the electric charge of immersed surfaces. J. mar. Res., 32, 5-12. NITTROUER,C.A., STERNBERG,R.W., CARPENTER,R. & BENNETT, J.T. 1979. The use of Pb--210 geochronology as a sedimentological tool: application to the Washington continental shelf. Marine Geol., 31, 297-316. NOWELL, A.R.M. 1982. Measuring and modelling organism effects on sediment transport. Geol. Soc. Lond. Newsletter, II, (2), 38-9 (abstract). • CHURCH, M. 1979. Turbulent flow in a depthlimited boundary layer. J. geophys. Res., 84, 4816-24. , JUMARS, P.A. & ECKMAN, J.E. 1981. Effects of biological activity on the entrainment of marine sediment. Marine Geol., 42, 133-53. ODD, N.V.M. & OWEN, M.W. 1972. A two-layer model of mud transport in the Thames Estuary. Proc. Instn. cir. Engrs., Suppl, 9, 175-205. VAN OLPHEN, H. 1977. An Introduction to Clay Colloid Chemistry. 2nd edn. John Wiley, New York. OLSEN, C.R., SIMPSON,H.J., PENG, T.H., BoPP, R.F. & TRIER, R.M. 1981. Sediment mixing and accumulation rate effects on radionuclide depth profiles in Hudson Estuary sediments. J. geophys. Res., 86, 1020-8. O'MELIA, C.R. 1972. Coagulation and flocculation. In: Weber, W.J. (ed.) Physiochemical Processes for -
-
-
-
-
67
Water Quality Control. Wiley-Interscience, New York. 61-109. 1978. Coagulation in wastewater treatment. In: Ives, K.J. (ed.), The Scientific Basis of Flocculation, Sijthoff& Noordhoff, Alphen a.d. Rijn, 219-68. OWEN, M.W. 1970. A Detailed Study of the Settling Velocities of an Estuary Mud. Hydraul. Res. Station, Wallingford, Rep. INT 78, 25pp. 1971. The effect of turbulence on the settling velocities of silt flocs. Proc. 14th Congress int. Ass. Hydraulics Res., Paris, Paper D-4, 27-32. -1975. Erosion of Avonmouth mud. Hydraul. Res. Station, Wallingford, Rep. INT 150, 17pp. PAK, H. & ZANEVELO,J.R.V. 1977. Bottom nepheloid layers and bottom mixed layers observed on the continental shelf off Oregon. J. geophys. Res., 82, 3921-31. --, ZANEVELD,J.R.V. & KITCHEN, J. 1980. Intermediate nepheloid layers observed off Oregon and Washington. J. geophys. Res., 85, 6697-708. PARKER, D.S., KAUFMAN,W.J. &JENKINS, D. 1972. Floc breakup in turbulent flocculation processes. J. Sanit. Eng. Div., Proe. Am. Soc. cir. Engrs., 98 (SAI), 79 99. PARKER, W.R. & KIRBY, R. 1981. The Behaviour of Cohesive Sediment in the Inner Bristol Channel and Secern Estuary in Relation to Construction 0[" the Sez'ern Barrage. Inst. Oceanogr. Sci. (Wormley U.K.) Rept. 117, 54pp. Unpublished M.S. PARTHENIADES, E., CROSS, R.H. & AYORA, A. 1968. Further results on the deposition of cohesive sediments. Proc. l lth Conf Coastal Engineering, London. Am. Soe. civil Engrs., 723-42. PIERCE, J.W. & SIEGEL, F.F. 1979. Suspended particulate matter on the southern Argentine shelf. Marine Geol., 29, 73-91. POSTMA, H. 1961. Transport and accumulation of suspended material in the Dutch Wadden Sea. Neth. J. Sea Res., 1, 148-90. 1967. Sediment transport and sedimentation in the estuarine environment. In: Lauff, G.H. (ed.), Estuaries, Am. Ass. Advmt. Sci. Pub., 83, Washington, 158-79pp. PRUPPACHER, H.R. & KLETT, J.D. 1978. The Microphysics of Clouds and Precipitation. Riedel, Dordrecht. 714pp. REES, A.I. 1966. Some flume experiments with a fine silt. Sedimentology, 6, 209-40. RHOADS, D.C., YINGST, J.Y. & ULLMAN, W. 1978. Seafloor stability in central Long Island Sound: Pt. I. Temporal changes in erodibility of fine-grained sediment. In: Wiley, M.L. (ed.), Estuarine Interaction. Academic Press. 221-44. & BOYER, L.F., in press. The effects of marine benthos on physical properties of sediments; a successional perspective. In: Tevesz, M.J.S. & McCall, P.L. (eds.), Animal-Sediment Relations." The Biogenic Alteration of Sediments. Plenum Press, New York. RICHARDSON, M.J. 1980. Composition and Characteristics of Particles in the Ocean." Evidence for Present Day Resuspension. Ph.D. Thesis, Mass. Inst. Technol., Woods Hole Oceanogr. Inst., WHOI-80-52. ROBERT, J.-M. & GOULEAU, D. 1977. Confirmation -
-
-
-
-
68
I.N. McCave
exp6rimentale du r61e d'une diatom+e benthique, Navicula ramosissima (Agardh) Cleve, dans la production de formations mucilagineuses stabilisatrices des vasieres marines littorales. C.R. hebd. Seanc. A cad. Sci., Paris, D284, 1915-18. ROUSE, H. 1937. Modern conception of the mechanics of fluid turbulence. Trans. Am. Soc. cir. Engrs., 102, 463-543. ROWE, G.T. & GARDNER, W.D. 1979. Sedimentation rates in the slope water of the northeast Atlantic Ocean measured directly with sediment traps. J. mar. Res., 37, 581-600. SAFFMAN,P.G. & TURNER,J.S. 1956. On the collision of drops in turbulent clouds. J. Fluid Mech., 1, 16-30. SARGUNAM, A., RILEY, P., ARULANANDAN, K. & KRONE, R.B. 1973. Physico-chemical factors in erosion of cohesive soils. J. Hydraulics Div., Proc. Am. Soc. cir. Engrs., 99, (HY3), 555-8. SCHAFFERNAK, F. 1922. Neue Grundlagen fur Berechhung der Geschiebefuhrung in Flusslaufen. Franz Deutike, Leipzig & Vienna. SCHRADER, H.J. 1971. Faecal pellets: role in sedimentation of pelagic diatoms. Science, 174, 55-7. SCHUBEL, J.R. 1968. Suspended sediment of the northern Chesapeake Bay. Chesapeake Bay Institute, Johns Hopkins Univ. Technical Report, 35, 262pp. 1969. Size distributions of the suspended particles of the Chesapeake Bay turbidity maximum. Neth. J. Sea Res., 4, 283-309. 1971. Tidal variation of the size distribution of suspended sediment at a station in the Chesapeake Bay turbidity maximum. Neth J. Sea Res., 5, 252-66. SCOFFI~, T.P. 1968. An underwater fume. J. sed. Petrol., 38, 244-7. 1970. The trapping and binding of subtidal carbonate sediments by marine vegetation in Bimini Lagoon, Bahamas. J. sed. Petrol., 40, 249-73. SHANKS, A.L. & TRENT, J.B. 1980. Marine snow: sinking rates and potential role in vertical flux. Deep Sea Res., 27, 137-43. SHELDON,R.W., PRAKASH,A. & SUTCLIFFE,W.H. 1972. The size distribution of particles in the ocean. Limnol. Oceanogr., 17, 327-40. SILVER, M.W. & ALLDREDGE,A.L. 1981. Bathypelagic marine snow: deep-sea algal and detrital community. J. mar. Res., 39, 501-30. SMAYDA, T.J. 1969. Some measurements of the sinking rate of faecal pellets. Limnol. Oceanogr., 74, 621-5. -1971. Normal and accelerated sinking of phytoplankton in the sea. Marine Geol., 11, 105-22. SMERDON, E.T. & BEASLEY, R.P. 1959. Tractive force theory applied to stability of open channels in cohesive soils. Agricultural Expt. Sta., Unit'. Missouri, Columbia, Research Bull. 715. SMOLUCHOWSK1, M. 1917. Versuch einer mathematischen Theorie der Koagulation kinetik kolloider L6sungen. Ann. Physik. Chem., 92, 129-68. SOUTHARD, J.B., YOUNG, R.A. & HOLLtSTER, C.D. 1971. Experimental erosion of calcareous ooze. J. geophys. Res., 76, 5903-9. SPIELMAN,L.A. 1978. Hydrodynamic aspects of flocculation. In: Ives, K.J. (ed.). The Scientific Basis of -
-
-
-
Flocculation. Sijthoff and Noordhoff, Alphen a.d. Rijn. 63-88. STow, D.A.V. & BOWEN, A.J. 1980. A physical model for the transport and sorting of fine-grained sediment by turbidity currents. Sedimentology, 27, 31-46. STUMM, W. & MORGAN, J.J. 1981. Aquatic Chemistry (chapt. 10, the solid--solution interface). Wiley, New York, 2nd Edn. 780pp. SUND~ORG, A. 1956. The River Klaralven, a study of fluvial processes. Geogr. Annaler, 38, 125-316. SUZUKI, N. & KATO, K. 1953. Studies on suspended materials. Marine snow in the sea. I. Sources of marine snow. Bull. Fac. Fisheries, Hokkaido Unit,., 4, 132-5. SWIFT, D.L. & FRIEDLANDER,S.K. 1964. The coagulation of hydrosols by Brownian motion and laminar shear flow. J. Colloid Sei., 19, 621-47. TAMBO, N. & HOZUMI, H. 1979. Physical characteristics of flocs--II. Strength offloc. Water Res., 13, 421-7. -t~r WATANABE,Y. 1979. Physical characteristics of flocs--I. The ftoc density function and atuminium floc. Water Res. 13, 409-19. TERWINDT, J.H.J. & BREUSERS,H.N.C. 1972. Experiments on the origin of flaser, lenticular and sandclay alternating bedding. Sedimentology, 19, 85-98. TERZAGHI, K. & PECK, R.B. 1968. Soil Mechanics in Engineering Practice. John Wiley, New York. THORN, M.F.C. 1975. Monitoring silt movement in suspension in a tidal estuary. Int. Assoc. Hydraulic Res. 16th Cong., SaD Paulo, Pap C71. 596-603pp. 1981. Physical processes of siltation in tidal channels. In: Hydraulic Modelling Applied to Maritime Engineering Problems. Inst. Civil Engrs., London. 47-55pp. & PARSONS, J.G. 1980. Erosion of cohesive sediments in estuaries: an engineering guide. In: Third International Symposium on Dredging Technology, British Hydraulics Research Assocation, Cranfield. 349-58. UNSOLD,G. 1982. Der Transportbeginn rolligen Sohlmaterials in gleichformigen turbulenten Stromungen: eine kritische Uberprufung der Shields-Funktion und ihre experimentelle Erweiterung auf feinstkornige, nicht-bindige Sedimente. Doctoral Dissertation, University of Kiel, 145pp. VALENTAS, K.J. & AMUNDSON, N.R. 1966. Breakage and coalescence in dispersed-phase systems. Ind. Eng. Chem. Fund., 5, 533-52. - - , BILOUS,O. & AMUNDSON,N.R. 1966. Analysis of breakage in dispersed phase systems. Ind. Eng. Chem. Fund., 5, 271-9. WEATHERLY, G. & KELLEY, E.A. 1982. 'Too cold' bottom layers in the HEBBLE area. J. mar. Res., 40, 985-1012. WEBB, J.E. 1969. Biologically significant properties of submerged marine sands. Proc. Roy. Soc. Lond., B174, 355-402. WELIKANOFF, M. 1932. Eine Untersuchung uber erodierende Stromgeschwindigkeiten. Wasserkraft u. Wasserwirtschaft, 27 (17), 196-9. WELLERSHAUS, S., GOKE, L. & FRANK, P. 1973. Size distribution of suspended particles in sea water.
Erosion, transport and deposition of fine-grained marine sediments Meteor Forschungs-Ergbenisse, Series D, no. 16, 1-16. WELLS, J.T. & COLEMAN,J.M. 1981. Physical processes and fine-grained sediment dynamics, coast of Surinam, South America. J. sed. Petrol., 51, 1053-68. WIalTE, S.J. 1970. Plane bed thresholds of fine grained sediments. Nature, Lond., 228, 152-53. WIEBE, P.H., BOYD, S.H. & WINGET, C. 1976. Particulate matter sinking to the deep-sea floor at 2000 m in the Tongue of the Ocean, Bahamas, with a description of a new sedimentation trap. J. mar. Res., 34, 341-54.
69
Y1NGST,J.Y. & RHOADS,D.C. 1978. Seafloor stability in central Long Island Sound. Part II. Biological interactions and their potential importance for seafloor erodibility. In: Wiley, M.N. (ed.), Estuarine Interactions. Academic Press, New York. 245-60. YOUNG, R.A. & SOUTHARD, J.B. 1978. Erosion of fine-grained marine sediments: Sea floor and laboratory experiments. Bull. geol. Soc. Am., 89, 663-72. ZABAWA, C. 1978. Microstructure of agglomerated suspended sediments in Northern Chesapeake Bay Estuary. Science, N.Y., 202, 49-51.
I.N. MCCAVE, School of Environmental Sciences, University of East Anglia, Norwich, NR4 7T J, UK.
Methods and observations in the study of deep-sea suspended particulate matter S.L. Eittreim S U M M A R Y : Forward light scattering and light transmission are the two best techniques, coupled with direct sampling, to measure the spatial and temporal variations in suspended sediments of the ocean. It has been difficult to calibrate these methods with satisfactory precision in terms of suspended particulate matter (SPM) concentration; the best precision is approximately _+ 10 ~gL -I. The poor resolution is due to the many unknown factors in the relationship between optical effects and concentration of SPM, in addition to errors in direct sampling of SPM concentration. Because typical deep-sea concentrations are in the range of 5 to 50 FtgL-1, extreme care must be taken to avoid contamination by airborne dust and salt filter residues. As a short-lived tracer of water motions, SPM is useful because its settling rates range from about 10 I to 102 m day -I, rates giving half-life time-scales roughly comparable to time-scales of deep-water eddy motions and boundary layer events. More needs to be learned, however, to define these settling rates with some confidence. Much progress has recently been made in studies with sediment traps which directly measure the downward vertical flux of particles. The work with sediment traps is prompted by the realization that large fast-settling aggregates and faecal pellets are important in the delivery of material to the bottom, and due to their short residence time and hence low concentrations in the water column, they are not sampled adequately by water bottles. Introduction
The interest in fine-grained suspended particulate matter (SPM) studies in the deep sea stems from a n u m b e r of sources. First, SPM studies give information about the transport paths of deep-sea sediments. These transport paths span great distance and time-scales, and little is k n o w n of them aside from gross characterization of turbidity distributions and inferences from fine-grained sea floor sediments. Second, SPM can be of use in studies of deep-sea circulation as a tracer o f water motions. Because the particles settle out, SPM acts as a non-conservative property of the water whose 'half-life', if it is known, can be used to derive mixing rates. This is proving especially useful in benthic boundary-layer studies where brief boundary-layer mixing events can be detected and examined with this short-lived tracer. Third, it is now widely appreciated that element transports in settling particles are important links in the geochemical and biological cycles of the ocean. Therefore the ocean's dynamics of particle aggregation, dissolution, and settling, as well as particle chemistry and nutrient content, are important frontiers of study to understand these cycles. In this brief review paper I focus on deep-water methods, including optical methods used and the principles behind each. For a complete and rigorous treatment of optical methods in oceanography, see Jerlov (1976). Gibbs (1974) presents studies of applied optical methods in oceanography ranging from studies of ocean SPM to studies aimed specifically at the optical properties of sea
water. In addition, Sackett (1978) gives a comprehensive review of the state of the art of SPM work in sea water. Simpson (1982) presents a recent review of SPM work and some novel approaches toward sampling. The deep-sea environment is in general relatively time-invarient and the SPM concentrations are low, ranging from 1 to 103 /~gL-~. The coastal, shelf, and estuarine environments, on the other hand, are d o m i n a t e d by tidal and short-term storm effects, and in general higher concentrations prevail. Examples of data from different parts o f the world ocean will also be presented to give an idea of the range of values encountered in the deep sea. A strong impetus was provided towards the study of deep-sea SPM by the Swedish Deep Sea Expedition in the late 1940s, which systematically sampled the world ocean at all depths and found significant geographic variations as well as pronounced vertical stratification of SPM as determined by light-scattering properties of samples analysed o n b o a r d ship (Jerlov 1953). These studies contributed to the realization (e.g. Heezen & Hollister 1964) that the deep sea was not necessarily a 'stagnant' environment devoid of significant currents and that boundary-layer frictional effects are likely to be important in suspending sediments. Inspired by the Swedish Expedition data, two different approaches have evolved for studying the variations in ocean water turbidity ~. The two different approaches have 1 It should be noted here that 'turbidity', like cloudiness, is a term to describe the net optical effect of particles in water and is a non-quantitative term.
71
72
S.L.
Eittreim
been aimed at quite different objectives, and it is important to separate these different objectives in looking at instruments used today. One objective is to determine ocean water optical properties. These optical properties are strongly determined by, among other things, the SPM present. The other allied objective is to determine the amount of SPM in ocean water and its spatial and temporal variations. The traditional method of instrument deployment, that of vertical profiling from ships, augmented by discrete water samples, has provided the most geologically and oceanographically useful data. The recent evolution towards deployment of moored arrays reflects the fact that temporal information is needed on SPM concentrations as well as details of spatial variations near bottom in order to better understand boundary-layer processes which may be 'eventdominated'. Fine-scale spacing of sensors and long-term measurements can best be achieved by avoiding attachment to a surface vessel.
Optical methods Since direct sampling of SPM is for practical reasons spatially limited, optical methods offer the best means of obtaining information on SPM variations with high resolution. Another method for continuous data acquisition of SPM concentrations is acoustic scattering, but acoustic methods have proved useful only for relatively high concentrations of coarse material due to limitations imposed by sound wavelength and attenuation (Orr & Hess 1978). Optical methods can of course be applied in vitro, but their main advantage is in situ where continuous sampling in time and space can be achieved as opposed to discrete sampling, the latter of which is limited in interpretive value to an oceanographer. The two optical methods presently in widest use for determining SPM concentrations in the ocean are forward scattering and transmission. Forward scattering offers the highest sensitivity due to the fact that particles scatter light preferentially in a forward direction. However, absolute calibration is difficult, and most measurements based on scattering to date have been recorded in relative units. Instruments which measure relative scattering in a forward direction coupled with water samples taken for SPM concentration have been employed by Lamont-Doherty Geological Observatory (Lamont) and Woods Hole Oceanographic Institution (Woods Hole) through the 'GEOSECS' program, a reconnaissance geochemical survey of the world ocean. Reconnaissance surveys were carried out by Lamont in the
late 1960s and 1970s, occupying several thousand stations worldwide at which vertical nephelometer profiles were made (e.g. Sullivan et al. 1973, 1975; Jacobs et al. 1974, 1980; Eittreim et al. 1976). At a limited number of these stations, water samples were also collected for SPM extraction. GEOSECS included vertical nephelometer profiles and water samples for SPM extraction at most of their several hundred stations (Bainbridge et al. 1976, 1977). Transmission, or its reciprocal, attenuation, offers the attractiveness of simplicity of measurement and ease of optical calibration. The drawback of transmission measurements for deepocean work is their lower sensitivity to suspended particles and the relatively stronger effect of dissolved substances on the measurement. Thus, in very low concentration environments ( < 5 0 pgL-l), their utility may be less than forward scattering measurements. However, R. Zaneveld (pers. comm. 1981) of Oregon State University recently developed a red-light-emitting-diode transmissometer for which he claims _+0.1% absolute accuracy. This instrument may be able to resolve low concentration variations in the deep sea. To quantitatively define the optical properties of ocean water, well-calibrated instruments are required. Careful control on angles, irradiated volumes, wavelengths, source strength, and window transmission and reflectance must be maintained in order to define the two optical factors in an absolute sense: light-scattering as a function of angle and light absorption (Jerlov 1976). To determine scattering, the measurement must be made with high precision as a function of angle because of the steep increase in scattering with decreasing small angles. Quantitative in situ scattering measurements have been made to great depths in limited areas (e.g. Beardsley et al. 1970; Matlack 1972), but no broad reconnaissance surveys have yet been made. Advancement towards standardized measurements of factors that cause ocean turbidity variations, using a rugged package for survey work, would be a timely development (Austin 1974). Much of the complexity of the measurement is eliminated if one is simply interested in SPM concentration. Devices which measure a relative optical property that is strongly SPM-dependent and which are coupled with water samples to calibrate the measurement in terms of SPM concentration can be made simple and rugged, and such instruments have proved to be productive in the past. Scattering measurements About half of the scattered light produced by
The study of deep-sea suspended particulate matter particles is in a forward cone of 9 ~. Forward scattering observed against the black background of the deep sea below the photic zone is thus a highly sensitive method of measuring suspended particles. This pronounced forward lobe means that quantitative measurements of scattering must be made with respect to angle, 0./~(0) is the conventional symbol for scattering per unit volume as a function of angle. Most relative scattering instruments measure an integral of/~(0) over some range of angles. However, if the volume from which the scattering emanates is not limited to a small size and defined with precision, calibration in absolute units is not possible. By limiting the sample volume to a small size, the sampling statistics degrade (because of the relative rarity of the larger particles) so that the objective of obtaining useful SPM data which is representative of a spatial average is somewhat incompatible with obtaining quantitative scattering measurements- hence the advent of scattering instruments ('nephelometers' or 'turbidimeters') which measure in relative units over a range of forward angles (Thorndike 1975; Meade et al. 1975). The Lamont instrument uses light scattered from approximately 8 ~to 2 4 ~from an incandescent light and records photographically. The film for each station is calibrated prior to lowering to determine its optical density-exposure relationship (Thorndike 1975). The Woods Hole nephelometer uses a laser light source with similar small and adjustable angles of scatter, is self-contained, and has an acoustic telemetered link to a surface vessel (Meade et al. 1975). An instrument designed by R. Koehler of the Ocean Engineering Group at Woods Hole and used in abyssal benthic boundary layer studies by Armi & D'Asaro (1980) also uses forward light-scattering (20~'-40 '~) from a detector and light source of an industrial turbidimeter and also measures in relative units. Measurement of total scattering at all angles, an integral of/~(0) from 0 ~=to 180, is conventionally denoted by b in units of m - 1. Instruments that measure b, excluding for practical reasons the far forward and far aft angles, have been built and used at sea by Jerlov (1953), Beardsley et al. (1970), and Sternberg et al. (1974). These instruments also have generally recorded in relative units.
= 1/r In I / T , or = In l I T for a 1-m path-length instrument.
Workers have used both units, %T, or attenuation in m - l . Attenuation is a somewhat more convenient unit because its value varies directly with SPM concentration rather than inversely. Drake (1971, 1974), Pak et al. (1980), and McCave (1983) have reported on oceanic sampiing programs in which transmission was measured in conjunction with water sampling for SPM concentrations.
Water sampling objectives and methods Concentrations of SPM, particle size spectra, and particle composition are the three principal kinds of information sought through direct water sampiing. Concentrations of SPM in the deep sea are commonly measured gravimetrically and expressed in/~gL -1 (or #gkg-1), although parts per million (ppm) by volume is often used when the measurement is of particulate volume by Coulter Counter* or other electronic particle counting device. Particle number concentration NL -~, is also used when data are obtained with particle-counting instruments. Because of the steep increase in N with decreasing particle size, these units are highly sensitive to the limits of the size window being measured. Figure 1 gives the p.gL -1
measurements
Because light attenuates as a function of e ~", where ~ is the attenuation coefficient and r is a distance, the relation between light transmission, T, and ~ is
8
--
IM
~
-b.
A
I O
O
..4
-
\
r
~O
Attenuation
73
O
ppb by volume (assuming p= 2gcm -3) FIG. 1. Volume or mass concentration versus number concentration for various diameter spherical particles. * Any use of trade names is for descriptive purposes only and does not imply endorsement by the USGS.
74
S.L.
Eittreim
relationship between volume or weight concentration and number concentration for different assumed spherical particle sizes. Note that for a given constant mass or volume concentration, the number concentration will change by four orders of magnitude for a shift from 20 to 1 ~m in particle size. An adequate description of oceanic SPM usually requires information on its size spectra. The settling velocity and hence residence time of particles is determined largely by their size; thus to use SPM as a quantitative tracer of water motions requires knowledge of the size characteristics of the SPM. Studies of the size spectra naturally concern the state of aggregation of particles. Collection of particles into first, second, and higher order aggregates is common in the ocean, and unless the aggregation state is known, predictions regarding settling velocities as well as size definition of SPM with respect to bottom sediments may be incorrect (Krone 1962; Kranck 1973; McCave, 1983). The composition of SPM in ocean water can be examined by microscopic, chemical, X-ray or electron microprobe techniques (Sackett 1978). Automated counting methods coupled with a scanning electron microscope now allow extremely small samples to be dealt with quantitatively (Bishop & Biscaye 1982). Although centrifugation was a popular means of extracting particles for study in the past (Jacobs & Ewing 1969; Lisitzin 1972), since the advent of membrane filters with uniform and small pore size, most workers now use filtration as a means of collection. In situ filtration methods have had some success where large water volumes were necessary due to very low concentrations or when the desired analysis required a large sample (Bishop et al. 1977; Peterson 1977). The GEOSECS program produced a systematic sampling and analysis method for SPM which might be regarded as a model for future work (Brewer et al. 1976). Systematic procedures were designed to avoid contamination. Duplicate sampling and analyses were done by separate institutions (Lamont and Woods Hole) to assess the accuracy and completeness of the extraction techniques. The fact that these duplicate analyses showed significant differences indicates their sensitivity to sampling technique and the importance of procedure to good results (Brewer et al. 1976): Most sampling for ocean SPM is now done with open-ended polyvinylchloride bottles, which offer the least resistance to water passing through the bottles as the sampling apparatus is lowered. Where concentrations are less than 50 itgL -~, 20-30-1itre bottles are normally used to obtain
enough sample for handy analysis. 'Rosette' samplers, actuated by command from the surface, are the most convenient for accurately locating samples. Messenger-tripped bottles are commonly used and give good results as long as sufficient concern is given for possible contamination by the wire. The awareness of possible contamination, or the opposite, incomplete extraction, cannot be overstressed. It should be assumed that each sampling bottle has its own "signature' of contamination, and this should be measured by rotating bottles and taking replicate samples to determine what its 'blank' contribution is before accepting data from it. Incomplete extraction, due to particles settling out below the bottle spigot or sticking to bottle walls, has been documented (Gardner 1977) and can be minimized by agitation of the sample bottle and by filtering as soon as possible after bringing the sample aboard. When working with SPM concentrations in the tens of pgL-N which is the most common value for open-ocean SPM below the euphotic zone, serious errors can be introduced by laboratory contamination. Wherever filters are exposed, a clean air environment produced by an air filtration hood system or 'clean room' is desirable. Residual salts on filters are an equally serious contaminant which can increase filter weight concentrations by a factor of 10 or more. The GEOSECS program procedures call for washing filters with 7 to 10 aliquots of distilled, filtered water. Oceanic SPM studies prior to 1970 did not generally follow such rigorous procedures to avoid contamination and to completely rinse out salts, and generally reported higher concentration values (Harris 1972; Sackett 1978). Polycarbonate micro-filters, which are made by etching out neutron bombardment holes made in plastic sheet, became available about 10 years ago. These filters, as discussed by Biscaye & Eittreim (1974), make particularly good extraction devices for low concentrations of fine SPM because of (1) low or negligible weight loss of filter material during filtration and (2) a lower tare weight and less change with humidity than other filter materials. Brewer et al. (1976) estimated a sample weight precision of ___5 pg, based on repeated weighings, which translates to _+0.5 /tgk -~ for a 10-L sample. Standard deviation of SPM concentrations from replicate samples was 5.2 pgL -~ (Brewer et al. 1976) which gives a maximum error of the overall sampling and weighing techniques; a maximum because of the possibility that real variations or patchiness in SPM concentrations in the water contributed to the variations. Microscopic particle counting and identifica-
The stud), of deep-sea suspended particulate matter tion offers another means of analysis of filtered SPM. Before the advent of automated particle counting and scanning electron microscope methods of particle analysis (Lambert et al. 1981: Bishop & Biscaye 1982), such work was extremely slow and tedious but is now a useful technique, although in its infancy. Using an energy-dispersive X-ray spectrometer coupled to a scanning electron microscope, Lambert et al. (1981) and Bishop & Biscaye (1982) obtained information on A1, Si, P, S, K, Ca, Cr, and Fe which can be roughly translated to relative percentages of clay mineral species and biogenic versus mineralogenic percentages. Harris (1972) obtained size spectra data down to particle sizes of 0.02 /xm using scanning electron micrographs and manual measurements of particles. Number versus particle size can most efficiently be obtained with Coulter Counters or other electronic sizing devices, which measure volumes of particles as they pass through an orifice that is part of a monitored electrical path. These volumes are converted to diameters of equivalent spheres. In order to cover the size spectrum adequately, several different apertures must be used and the data overlapped, and in order to obtain data on particles smaller than 1 /ma, methods other than the Coulter Counter must be used. Bader (1970), Carder et al. (1971), and Sheldon et al. (1972) and others have successfully used Coulter Counters at sea. Measures must be taken to minimize mechanical and electrical noise which the instrument is quite sensitive to. The volume concentrations measured, when combined with concentrations by weight from filtration data, can be used to calculate apparent SPM bulk density (McCave 1983). As a general rule, oceanic as well as atmospheric particle sizes are distributed according to a power-law distribution, N = A D .... , where N is the number of particles greater than diameter D and A is a constant (Junge 1963; Bader 1970). A slope, m, of 3 implies equal volumes of particles in all size classes, a circumstance which Sheldon et al. (1972) found to hold generally for biogenic ocean particles ranging in size from bacteria to whales. For inorganic as well as organic SPM such a power law has been found to hold true in general, and attention is usually focused on cases of where and why the slope m deviates from a value of 3. Thus the emphasis in recent years has been on the description of SPM size spectra in terms of m and its variations up and down the water column and up and down the size spectrum. Figure 2 shows the correlations obtained between various kinds of optical measurements and SPM concentrations determined either by
75
filtered weight or by Coulter Counter volume. For low concentrations (e.g. < 50/tgL -~) it has been difficult to obtain a good empirical calibration with a low scatter of data points. There are many reasons for this. Primary among them is the fact that scattering as well as attenuation is sensitive to total particulate surface area rather than mass or volume. Thus, changes in the particle size spectra will change the optical effect per unit volume of particulates. Colour and index of refraction differences in the particle population will also have an effect, especially on attenuation, although this effect is less than that caused by size differences (Jerlov 1976). Finally, scatter introduced by sampling errors is probably roughly + 5 /~gL -t even for the most careful work. In any case, demonstrations of empirical calibrations such as those in Fig. 2 are necessary for deriving SPM concentrations from optical measures, and the scatter in these relationships should be kept in mind in the subsequent data analysis.
Sediment trap measurements The residence times of particles in the water must be considered when relating SPM trapped in sampling bottles to sedimentation on the bottom (McCave 1975). Bottles have a bias towards collection of particles with long residence time, since these particles occur in greater proportion in water than in the sediment. Sediment traps, which capture the downward flux of particles, offer another means of collecting oceanic SPM. Sediment traps collect particles which may, although contributing significantly to sediment, have a very short residence time in the water column. Hence, bottle samples are biased toward the small, slowly settling population of particles, whereas traps are biased toward the larger, high-flux population. Dymond et al. (1981) compared data from four different types of sediment traps deployed in Santa Barbara Basin. The maximum difference in flux measured among the four traps was a factor of 2, and the average flux measured over the 48-day period was in fair agreement with the longer term sedimentation rate in the basin determined by 21~ profiles in the sediment. Bruland et al. (1981) determined a ratio of SPM flux to bottom sediment flux of 0.93 for sediment traps set in Santa Barbara basin. It should be kept in mind that the fine SPM in the water column is scavenged to some unknown extent by the faster-falling large particles, the faecal pellets and particles which Honjo et al. (1982) called 'amorphous organic aggregates'. Honjo et al. (1982) demonstrated a consistent increase in the flux of lithogenic particles with
S.L. Eittreim
76 1000
(a) 9
600
O)
(c)
100
:k
60
400
d
: - .,,=|:..
t,-
~
80
(b)
-':~
9
40
".:i?i': " "
lO
200
Q.
20
1
I 10
0
I 100
0
I 600
Scattering ratio
I 400
I 200
I 0
-
I 2
0 0
Scattering pulse delay time
I 4
I 6
I 8
Relative scattering
(ms)
(d) 1200
L.j 8oo
(e)
9 " " : ...
1000 -
v
---.
"
9
6 c 0
. ; 9 ~149
(9
o
400
-
9. : :
9
: ""
9
9149149
..~.!~.~
""
o 9 ~~
9 ~ '-'.~ -
lOO
-.
9
9 o ,It,
0 0
Jil'l
I
0.5
1.0
I
I
1.5 2.0 Attenuation coeff., ( x ( m - 1 )
100
I
I
I
I
I
80
60
40
20
0
% Transmission
FIG. 2. In situ calibrations of various SPM-sensing instruments: (a) Lamont--Thorndike nephelometer (Biscaye & Eittreim 1974) and (b) Woods Hole---GEOSECS nephelometer (Meade et al. 1975) are both foward scattering meters. (c) University of Washington b-meter (Sternberg et al. 1974). (d) and (e) transmissometer calibrations of McCave (in press) and Drake (1974), respectively. Note the differences in scales of SPM concentrations. For (d), conversion of SPM concentration from ppb to/~g L- ] is based on a particle density of 2.0 g cm -3. depth that can best be explained by progressive scavenging during descent of large organic aggregates. The 'faecal pellet express' may be the primary means by which the fine SPM reaches the sea floor (Honjo & Roman 1978; Honjo 1982), and sediment traps offer the principal means of studying this process. Sediment trap studies are opening up a new field which directly addresses the question of residence time of particles in the water column. The VERTEX (Vertical transport and exchange of materials) program, for example (Martin & Knauer 1982), is an interdisciplinary study which attempts to define the downward flux
of particles and their chemical and biological changes during descent by sampling the upper and mid-water column with large sediment-trap arrays. Many questions still have to be answered regarding trapping efficiencies and the meaning of what is caught in the traps (Gardner 1980a,b). No trap deployments have yet attempted to catch non-vertical flux, for example. In a near-bottom nepheloid layer, where eddy diffusion in all directions may be important, the three-dimensional flux maybe important to understand. Although a trap may establish a downward net vertical flux, the true gross flux is unknown, since
The study of deep-sea suspended particulate matter it is possible, for example, that more (unsampled) particles moved up during the sampling period then moved down. To get at these kinds of questions, more work needs to be done in the future and the reader is referred to the above references as starting points.
Composition and particle size of ocean S P M Sackett (1978) reviewed the state of knowledge on the geochemistry, composition and concentrations of oceanic SPM, radionuclide estimates of residence times, and other aspects of SPM work which I will not repeat here. A few recent developments are worth mentioning, however. Automated particle counting and analysis techniques (Lambert et al. 1981, Bishop & Biscaye 1982) will soon produce more knowledge of SPM compositions. The two sources of ocean SPM are terrigenous detritus, transported via the atmosphere (Windora 1975), surface currents, mid-water turbid plumes emanating from the shelf break (Drake 1971) or bottom turbid flows, and biogenic detritus from the photic zone. In the mid-water regions, greater proportions of biogenic opal and Ca have been found (Bishop & Biscaye 1982), whereas bottom nepheloid-layers are known to be enriched in clay-mineral elements such as Fe (Betzer & Prison 1971) and Al (Feely et al. 1971) relative to the rest of the water column. The limited information that exists regarding particle size spectra of SPM suggests that aggregation state is important and that a bimodal distribution is often caused by the presence of a large aggregate mode in addition to the smaller primary particle mode. Second, third, and fourthorder aggregates are found, sometimes biologically bound, sometimes simply electrostatically bound (Kranck 1973; McCave, 1983). In general, the number vs. diameter distributions are loglinear, often with a break in slope at about 4/am, indicating a peaked volume spectrum in the vicinity of 4 ~tm (Harris 1972; McCave 1975).
Regional and temporal variations in the oceans For a compilation of worldwide SPM concentrations, the reader is referred to the GEOSECS volumes (Bainbridge et al. 1976, 1977) and to Brewer et al. (1976) for discussion of the Atlantic GEOSECS data. Water of lowest SPM concentration usually occurs in the middle of the water column (Fig. 3). The surface mixed layer and
77
euphotic zone supplies the upper water column with airborne and water-borne clay and silt and biogenic debris, all of which settle and diminish their concentration downward by aggregation (which increases their settling velocity), dissolution, and grazing by organisms (Lal & Lehrman 1973). The near-bottom part of the water column is supplied with resuspended fine sediment from the sea floor. The mid-water clearest zone is the region of overlap between the downward decrease from the surface and the upward decrease from the bottom and is usually a kilometre or more from these two sources. In areas where bottom resuspension is minimal the mid-water clear zone is very deep, or even extends to the bottom, such as in the central gyres of the north and south Pacific (Ewing & Connary 1970). The lowest concentrations in the mid-water zones occur in the low productivity areas of the temperate latitudes away from the continental margin sources of sediment or productive zones associated with divergences (Eittreim et al. 1976). The exact concentrations in these mid-water clear zones are difficult to establish because values are near the level of resolution allowed by sampling methods, about 5 #gL -1 (Biscaye & Eittreim 1974; Brewer et al. 1976). Figure 3 shows a range of turbidity profile types from low-energy, Iow-SPM concentration areas to high-energy, high-concentration areas. Note the broad clear zone in the mid-water region and the occurrence of bottom boundary layers capped by high gradients in the high-energy areas. Mid-water stratification is commonly associated either with advecting water masses of turbidity higher or lower than the surrounding water, such as the Mediterranean outflow, or due to a concentration of SPM at high density gradients caused by the reduction in mean settling speed of particles at the high-gradient regions (Jerlov 1959; Drake 1971; Pak et al. 1980). The most common high-turbidity feature of the water column is that associated with the seasonal thermocline. As Jerlov (1959) and Pak et al. (1980) note, and as is seen in many of the Lamont nephelometer profiles which were taken at night and thus avoid daylight interference (Eittreim 1970), the thermocline is commonly associated with a strong turbidity maximum which is presumed to be largely biogenic. Particle maxima can also be produced in theory by the effect of dissolution on settling speed of particles in the presence of slow upwelling (Lal & Lehrman 1973; Brun-Cottan 1976), but this effect has not been documented by measurements. The increase in turbidity commonly observed adjacent to bottom, termed a 'nepheloid-layer' by Ewing & Thorndike (1965), is observed more
78
S.L.
PACIFIC
~
J
Eittreim
7
o. S
ATLANTIC
J
U,o~ 4
I
I
I
o c .E s
15
112 38
/'xg L-1
? ....
183
~ 293
1
Log 6 -
( s c a t t e r i n g ratio)
}
138
FIG. 3. Selected nephelometer profiles (Eittreim & Ewing 1972; Sullivan et al. 1973, 1975) showing regional variations in profile shapes from a low-energy area (north and south-eastern Pacific) on the left to high-energy area (western Atlantic basins) on the right. SPM concentrations in pg L - l are indicated for bottom 100 m of water column, mostly from data of nearby GEOSECS stations (Bainbridge et al. 1976, 1977). Concentration values for the two profiles on the far right from Biscaye & Eittreim (1974) and McCave (in press), respectively. Numbers at the top of the profiles refer to station numbers of cruises Conrad 15 and 16 and Vema 23. Note that the units are on a log scale which suppresses the high- and enhances the low-amplitude features.
often than not in the ocean. In areas of extremely weak bottom water flow (e.g. < 5 cm s -1) this layer is sometimes not observed, the clear intermediate water extending to the bottom. But in most areas, some effects of bottom water interaction with sediment results in at least a minor increase in turbidity even if only a few p g L - 1 an amount difficult to document by sampling (Fig. 2). The exponential increases in SPM usually observed towards bottom often describe a series of straight-line segments on log-linear plots (Eittreim 1970; Sullivan et al. 1973, 1975), suggesting a hierarchy of layers on top of one another, with different exponents governing the diffusive particle transport within each layer. One explanation proposed by Armi & D'Asaro (1980) involves detachment from the bottom of turbulent and turbid benthic layers. Lateral migration of these detached layers along isopycnal surfaces, perhaps from relative topographic highs, into the interior of the water column may provide a mechanism for diffusing the SPM away from the bottom without invoking the unrealistically high coeffi-
cients of vertical eddy diffusion computed by Eittreim & Ewing (1972). Biscaye & Eittreim (1977) examined advected SPM fluxes for the bottom waters of the western Atlantic basins. The western Atlantic has a generally continuous bottom nepheloid-layer associated with the western boundary currents of polar bottom waters and a large data base of Lamont nephelometer stations exists in this area. Fluxes ranging from about 1 to 8 x 106 tons yr -1 were found at six section locations where bottom water transports have been measured. These fluxes are the same order of magnitude as the non-carbonate sediment deposition in the basins "downstream' of these sections (based on longterm sedimentation rates), suggesting that such fluxes are a significant factor in deposition of fines in these Western Atlantic basins. These fluxes are small, however, in comparison to loads carried by major rivers near their mouths such as the Amazon (900x 10 6 tons yr - l ) (Milliman & Meade 1983). However, most of the river particulate load, as is pointed out by Milliman & Meade, is deposited on subareal deltas, in estuaries, or on
The study of deep-sea suspended particulate matter the inner shelf. Data that will give average terrigenous sediment output beyond the shelf break to the deep sea is sorely needed. Until estimates, even good to an order of magnitude, can be made of offshelf terrigenous transport, attempts at making deep-sea sediment budgets in such areas as the western Atlantic basins, where turbid bottom water transport is significant, will be precluded. Another major unknown factor in constructing deep-sea sediment budgets is the residence times of nepheloid-layer particles. Estimates are needed of the number of deposition/resuspension cycles the average particle experiences before final deposition. If these estimates are not made, the link cannot be made between what is observed in the water column today and what is deposited on the sea floor. This question is presently being addressed by sediment trap work mentioned above. Extreme temporal variations such as those shown in Fig. 4 and those found in investigations in the HEBBLE area (40~ 62~ Shor et al., this volume) are probably confined to high energy regions ( > 10 cms -1 flow). Although seasonal fluctuations in biogenic production should be
""';
(a)
~5 far)
""
Po"~". \"...
:
m
I
_
. ....
I'
-(b) I _o q \
M
," 1
...
\:
.:
\
"
,
\
I oNo.,s, .on"]
400
A
200 ~
..
~oo~ ~ --9
(c)
/,-,
B60[:, ..~...o." ....~
~'~401," , .6
~
,'
\1
/
~ 20 0
5
I0
15
20
Days
FIG. 4. Examples of temporal variations in SPM observed in an active or high-energy area at two stations 110 km apart on the flank of the Blake-Bahama Outer Ridge (modified after Biscaye & Eittreim 1974). (a) and (b) are taken from bottle-sampled and filtered SPM. m.A.B, means metres above bottom. (c) Data from Lamont-Thorndike nephelometer bottom-most data point, approximately 10 metres above bottom.
79
expected as well as seasonal changes in river and airborne clay and silt debris, no significant seasonal changes have yet been detected in midocean areas. Perhaps this is because these changes are either too small to detect, or biogenic particles settle slowly enough so that the changes are smoothed out vertically in the water column. On the Lamont nephelometer stations repeatedly occupied, only the near-bottom part of the water column in high-velocity regions showed significant variations (Eittreim et al. 1976). Nephelometer profiles taken repeatedly over a 19-day period on the Blake-Bahama Outer Ridge flank showed a striking agreement in the region of the water column shallower than 3500 m, above the benthic boundary layer (Biscaye & Eittreim 1974). Studies in the HEBBLE area have shown that high concentrations ( > 1000 #g L -], Amos 1979; Biscaye et al. 1980) and short-term particle resuspension events ( < 1 day, Pak et al. 1983) do occur in the deep sea in restricted locations, such as along the narrow 'filament' of Western Boundary Undercurrent on the continental rise of the north-west Atlantic. To understand sedimentation of fines in the ocean, these local event-dominated processes must be more clearly understood. There is evidence that frequent high turbidity events can be caused by both downslope turbidity currents (Amos & Gerard 1979) and resuspension by contour-following bottom currents (Biscaye et al. 1980). The broad regional distributions shown in the GEOSECS and Lamont data reflect water mass interactions and mixing processes on a larger and slower scale than the processes observed in the HEBBLE area. An integration of these two different scale phenomena must ultimately be made.
Summary A few directions for future studies of fine SPM stand out as particularly fruitful at this time. Further work on aggregation and settling dynamics is important. Such studies may reveal new methods for characterizing resuspended versus 'fresh' SPM, i.e. that which has not yet experienced bottom contact. If the sea floor acts as an efficient aggregator as one might expect, then resuspended sediments might be predominantly in the form of aggregates, although the preliminary work of McCave (1983) does not suggest this. More studies of elemental characterization of SPM at all levels in the water column and comparison with adjacent bottom sediments may yield new insights into sources and pathways. Detailed high-resolution profiles in the bottom
8o
S.L. Eittreim
b o u n d a r y - l a y e r m a y yield i m p o r t a n t i n f o r m a t i o n on the nature of b o u n d a r y - l a y e r turbulence and h o w sediment is carried in suspension. Studies of particle flux m e a s u r e m e n t s with sediment traps will help our u n d e r s t a n d i n g of particle residence times and mechanisms of d o w n w a r d transport. In order to construct sediment budgets for deep-sea areas where resuspension of terrigenous sediment is c o m m o n , better data on offshelf terrigenous transport is needed to improve the u n d e r s t a n d i n g of the possible contribution from steady or aperiodic turbidity flows. Studies focused on particular pathways of S P M will be important. For example, what roles do canyons play, if any, in the offshelf transport of fine sediment? New types of deep-sea bedforms created by
u n k n o w n mechanisms related to steady and perhaps aperiodic b o t t o m flows have recently been found (Hollister et al. 1974; D a m u t h 1980; F l o o d 1981). These features are likely related to suspended loads carried in benthic b o u n d a r y layers, and their study by methods of SPM m e a s u r e m e n t with improved resolution and precision will be necessary. Needless to say, rigorous adherence to careful procedures in S P M w o r k to avoid c o n t a m i n a t i o n and to completely extract particles will help advance the state of the art. ACKNOWLEDGMENTS: D a v i d E. D r a k e and D a v i d Z. Piper of the US Geological Survey reviewed the manuscript and offered m a n y helpful suggestions.
References AMOS, A.F. & GERARD,R.D. 1979. Anomalous bottom water south of the Grand Banks suggests turbidity current activity. Science, 203, 894-97. ARMI, L. & D'ASARO, E. 1980. Flow structures of the benthic ocean. J. geophys. Res., 85, 469-84. AUSTIN, R. 1974. Instrumentation used in turbidity measurement. Proc., NOIC Turbidity Workshop. May 6-8. 1974, National Oceanographic Instrumentation Center, Washington, DC, 45-74. BADER, H. 1970. The hyperbolic distribution of particle sizes. J. geophys. Res., 75, 2822-30. BAINBRIDGE, A.E., BISCAYE, P.E., BROECKER, W.S., HOROWITZ,R.M., MATHIEU,G., SARMIENTO,J.L. & SPENCER, D. 1976. GEOSECS Atlantic bottom hydrography, Radon and suspended particulate atlas. GEOSECS Operations Group, Scripps hlstitution of Oceanography, Internal Report. , , , , , - -1977. GEOSECS Pacific bottom hydrography, Radon and suspended particulate atlas. GEOSECS Operations Group, Scripps Institution of Oceanography, Internal Report. BEARDSLEY, G.F., JR., PAR, H., CARDER, K. & LUNDGREN, B. 1970. Light-scattering and suspended particles in the eastern equatorial Pacific Ocean. J. geophys. Res., 75, 2837-48. BETZER, P.R. & PILSON, M.E.Q. 1971. Particulate iron and the nepheloid layer in the western North Atlantic, Caribbean and Gulf of Mexico. Deep Sea Res., 18, 753-61. BISCAYE, P.E. & EITTREIM, S. 1974. Variations in benthic boundary layer phenomena: nepheloid layer in the North American Basin. In: Gibbs, R. (ed.), Suspended Solids in Water. Plenum Press, New York. 227-60. -& -1977. Suspended particulate loads and transports in the nepheloid layer of the abyssal Atlantic Ocean. Marine. Geol., 23, 155-72. -& ZANEVELD, J.R.V., PAK, H. & TUCHOLKE, B. 1980. Nephels! Have we got Nephels! (abstract) LOS Trans. Am. geophys. Un., 61, 1014.
BISHOP,J.K.B. & BISCAYE,P.E. 1982. Chemical characterization of individual particles from the nepheloid layer in the Atlantic Ocean. Earth planet. Sci. Lett., 58, 265-75. --, EDMOND,J.M., KETTEN, D.R., BACON, M.P. & SILKER, W.B. 1977. The chemistry, geology and vertical flux of particulate matter from the upper 400 m of the equatorial Atlantic Ocean. Deep Sea Res., 24, 511-48. BREWER, P.G., SPENCER,D.W., BISCAYE,P.E., HANLEY, A., SACHS,P.L., SMITH, L.L., KADAR,S. & FREDERICKS, J. 1976. The distribution of particulate matter in the Atlantic Ocean. Earth planet. Sci. Lett., 32, 393-402. BRULAND, K.W., FRANKS, R.P., LANDING, W.M. 8~ SOUTAR, A. 1981. Southern California inner basin sediment trap calibration. Earth planet. Sci. Lett., 53, 400-8. BRUN-COTTAN, J. 1976. Stokes settling and dissolution rate model for marine particles as a function of size distribution. J. geophys. Res., 81, 1601-06. CARDER, K.L., BEARDSLEY,G.F., JR. & PAK, H. 1971. Particle size distribution in the eastern equatorial Pacific. J. geophys. Res., 76, 5070-7. DAMUTH, J.E. 1980. Use of high-frequency (3.5-12 kHz) echograms in the study of near-bottom sedimentation processes in the deep-sea: a review. Marine Geol., 38, 51 75. DRAKE, D.E. 1971. Suspended sediment and thermal stratification in Santa Barbara Channel, California. Deep Sea Res., 18, 763-9. 1974. Distribution and transport of suspended particulate matter in submarine canyons off southern California. In: Gibbs, R. (ed.), Suspended Solids in Water. Plenum Press, New York. 133-53. DYMOND, J., FISCHER, K., CLAUSEN, M., COBLER, R., GARDBER, W., RICHARDSON, M.J., BERGER, W., SOUTAR, A. & DUNBAR, R. 1981. Sediment trap intercornparison study in the Santa Barbara Basin. Earth planet. Sci. Lett., 53, 409-18. EITTREIM, S. 1970. Suspended Particulate Matter in the
The study of deep-sea suspended particulate matter Deep Water of the North American Basin. Ph.D. Thesis, Columbia Univ., New York (available in microfilm). 166 pp. -& EWING, M. 1972. Suspended particulate matter in deep waters of the North American Basin. In: Gordon, A. (ed.), Studies in Physical Oceanography, A Tribute to George Wust on his 80th Birthday, 2. Gordon & Breach, New York. 123-67. , THORNDIKE, E.M. & SULLIVAN, L. 1976. Turbidity distribution in the Atlantic Ocean. Deep Sea Res., 23, 1115-27. EWING, M. & CONNARY,S. 1970. Nepheloid layer in the North Pacific. In: Hays, J. (ed.)., Geol. Soc. Am. Mere., 126, 41-82. & THORNDIKE, E.M. 1965. Suspended matter in deep ocean water. Science, 147, 1291-94. FELLY, R.A., SACKETT, W.M. & HARRIS, J.E. 1971. Distribution of particulate aluminium in the Gulf of Mexico. J. geophys. Res., 76, 5893-902. FLOOD, R.D. 1981. Longitudinal triangular ripples in the Blake-Bahama Basin. Marine Geol., 39, M 13-M20. GARDNER, W.D. 1977. Incomplete extraction of rapidly settling particles from water samplers. Limnol. Oceanogr., 22, 764. 1980a. Sediment trap dynamics and calibration: a laboratory evaluation. J. mar. Res., 38, 17-39. 1980b. Field assessment of sediment traps. J. mar. Res., 38, 41-52. GIBBS, R.J., (ed.) 1974. Suspended Solids in Water. Plenum Press, New York. 320 pp. HARRIS, J.E. 1972. Characterization of suspended matter in the Gulf of Mexico - 1. Spatial distribution of suspended matter. Deep Sea Res., 19, 719-26. HEEZEN, B.C. & HOLL1STER,C. 1964. Deep-sea current evidence from abyssal sediments. Marine Geol., l, 141-74. HOLLISTER, C.D., FLOOD, R.D., JOHNSON~D.A., LONSDALE, P. & SOUTHARD, J.B. 1974. Abyssal furrows and hyperbolic echo traces on the Bahama Outer Ridge. Geology, 2, 395-400. HONJO, S. 1982. Seasonality and interaction of biogenic and lithogenic particulate flux at the Panama Basin. Science, 218, 883-4. -& ROMAN, M.R. 1978. Marine copepod faecal pellets; production, preservation and sedimentation. J. mar. Res., 36, 45-57. --, MANGANINI, S.J. & POPPE, L.J. 1982. Sedimentation of lithogenic particles in the deep ocean. Marine Geol., 50, 199-219. JACOBS, M.B. & EWING, M. 1969. Suspended particulate matter: concentrations in the major oceans. Science, 163, 380-3. JACOBS, S.S., BAUER, E.B., BRUCHHAUSEN,P.M., GORDON, A.L., ROOT, T.F. & ROSSELOT, F.L. 1974. Eltanin Reports; cruises 47-50, 1971; 52-55, 1972: Hydrographic stations, bottom photographs, current measurements, nephelometer profiles. Tech. Rept. Lamont-Doherty Geological ObservatoJ3, CU-2-74. , GEORGI, D.T. & PATLA, S.M. 1980. Conrad 17; Hydrographic stations, sea floor photographs, nephelometer profiles in the southwest Indian-
-
-
-
-
-
81
Antarctic Ocean. Tech. Rept. Lamont Doherty Geological Observatory CU-I-80- TR1. JERLOV, N.G. 1953. Particle distribution in the ocean. Rep. Swedish Deep-Sea Expedition, 3, 73-97. 1959. Maxima in the vertical distribution of particles in the sea. Deep Sea Res., 5, 178-84. -1976. Marine Optics. Elsevier, Amsterdam. 231 PP. JUNGE, C.E. 1963. Air Chemistry and Radioactivity. Academic Press, New York. 382 pp. KRANCK, K. 1973. Flocculation of suspended sediment in the sea. Nature, 246, 348-50. KRONE, R.B. 1962. Flume studies of the transport of sediment in esturarial shoaling processes. Hydraulic Engineering Laboratory and Sanitary Engineering Research Laboratory, Univ. of California, Berkeley. 110 pp. LAL, D. & LERMAN,A. 1973. Dissolution and behavior of particulate biogenic matter in the ocean; some theoretical considerations. J. geophys. Res., 78, 7100-11. LAMBERT, C.E., JEHANNO, C., SILVERBERG, N., BRUNCOTTON, J.C. & CHESSELET, R. 1981. Log-normal distribution of suspended particles in the open ocean. J. mar. Res., 39, 77-98. LISITZIN,A.P. 1972. Sedimentation in the World Ocean. Soc. Econ. PaleD. Min. Spec. Publ.., 17, 225 pp. MARTIN, J. • KNAUER, G.A. 1982. A comparison of particulate reactivities of Ag, Cd, Co, Cu, Fe, Mn, Mo, Ni, V and Sn observed during Vertex II. (abstract) LOS Trans. Am. geophys. Un., 63, 960. MATLACK,D.E. 1972. Deep ocean optical measurement (DOOM) report; Bahama Channels and northwestern Atlantic Ocean. Naval Ordnance Lab., Maryland, NOLTR-72-284, 1-33. MCCAVE, I.N. 1975. Vertical flux of particles in the ocean. Deep Sea Res., 22, 491-502. 1983. Particle size spectra, behavior and origin of nepheloid layers over the Nova Scotian Continental Rise. J. geophys. Res. 7647-66. MEADE, R.H., SACHS, P.L., MANHEIM, F.T., HATH-
-
-
8 8 ,
AWAY, J.C. & SPENCER, D.W. 1975. Sources of suspended matter in waters of the Middle Atlantic Bight. J. sed. Petrol., 45, (Appendix) 186-8. MILLIMAN, J.D. & MEADE, R.H. 1983. Worldwide delivery of river sediments to the oceans. J. Geol., 91, 1-21. ORR, M.H. & HESS, F.R. 1978. Remote acoustic monitoring of natural suspensate distributions, active suspensate resuspension, and slope shelf water intrusions. J. geophys. Res., 83, 4062-8. PAK, H., ZANEVELD, R.V. & KITCHEN, J. 1980. Intermediate nepheloid layers observed off Oregon and Washington. Jour. geophys. Res., 85, 6697-708. -& ZANEVELD,R.V. 1983. Temporal variations of beam coefficient on the continental rise off Nova Scotia. J. geophys. Res., 88, 4427-32. PETERSON, R.E. 1977. A Study of Suspended Particulate
Matter." Arctic Ocean and Northern Oregon Continental Shelf Ph.D. Thesis, Oregon State Univ., Corvallis. SACKETT, W.M. 1978. Suspended matter in sea water.
82
S.L. Eittreim
In: Riley, J.P. & Chester, R. (eds.), Chemical Oceanography. Academic Press, London. 127-72. SHELDON, R.W., PRAKASH, A. & SUTCL1FFE,W.H. JR. 1972. The size distribution of particles in the ocean. Limnol. Oceanogr., 17, 327-40. SIMPSON, W.R. 1982. Particulate matter in the oceans - sampling methods concentration, size distribution and particle dynamics. Oceanogr. Mar. Biol. Ann. Ree., 20, 119-72. SMITH, C.L., KADAR, S. • FREDERICKS, J. 1976. The distribution of particulate matter in the Atlantic Ocean. Earth planet. Sci. Lett., 32, 393-402. STERNBERG, R.W., BAKER, E.T., MCMANUS, D.A., SMITH, S. & MORRISON, D.R. 1974. An integrating nephelometer for measuring particle concentrations in the deep sea. Deep Sea Res., 21,887-92.
SULLIVAN, L., THORNDIKE, E., EWING, M. & EITTREIM, S. 1973. Nephelometer measurements, Hach turbidimeter measurements and bottom photographs from Conrad cruise 15. Tech. Rept. Lamont-Doherty Geological Observatory 8-CU-8- 73. --, THORNDIKE, E. & EITTREIM, S. 1975. Nephelometer measurements and bottom photographs from Conrad cruise 16. Tech. Rept. Lamont-Doherty Geological Ohsert'atory CU- 1 I- 75. THORNDIKE, E.M. 1975. A deep-sea photographic nephelometer. Ocean Engrg., 3, 1-15. WINDOM, H.L., 1975. Eolian contribution to marine sediment. J. sed. Petrol., 45, 2-8.
S.L. EITTREIM, US Geological Survey, Menlo Park, California 94025 USA.
Grain-size characteristics of turbidites Kate Kranck S U M M A R Y: Detailed sampling using a very small sample size and grain-size analysis with a Coulter Counter of three fine-grained turbidites enabled a distinction to be made between the well-sorted single-grain Stokes'-deposited and unsorted "whole suspension' floc-deposited grain-size populations. The results indicate that each turbidite is a continuous sequence deposited from the same source suspension. Particles settle to the bottom as flocculated masses. Initially flocs are broken up by near-bottom shear forces and only the coarsest silt and sand remains on the bed. The remaining mud forms a temporary mud suspension near the bottom which reflocculates and intermittently deposits at some critical concentration producing mud interlayers between silt laminae. Decrease in current velocity eventually allows simultaneous deposition of single grains and flocs with the latter becoming progressively more abundant resulting in formation of graded beds. Eventually, all deposition occurs in the form of mud flocs and a massive sediment results. The study essentially confirms the basic mechanism of the fine-grained turbidite deposition model proposed by Stow & Bowen (1978). The abundant presence of turbidites in the geological column is a result of the fact that transport of continental material to the ocean depressions is not a continuous one-step process. Terrestrial erosional products from river and aeolian transport initially sediment out relatively close to their source. Subsequent and intermittent resedimentation of this material downslope provides a large amount of clastic material to the deep sea. As a result approximately 50-70~o of oceanic sediment consists of continental clastics. Many of these are turbidites whose sole marks, graded bedding, mud-silt lamination and other characteristic features identify them as deposits from high turbidity subaqueous flows which travel distances on the order of tens to hundreds of kilometres. The 90% turbidite component of abyssal sediments attests to the importance of this process and justifies the attention given to turbidite sedimentation in the geological literature. Starting with the early work of Kuenen (1951, 1966a,b), studies of ancient, modern and experimental turbidite beds have provided a basic knowledge of the origin and physical characteristics of these deposits. The work of Bouma (1962) has proved especially useful by providing an idealized sequence which has formed a basis for most subsequent regional descriptions of turbidite deposits as well as for discussions of the hydrodynamic process creating turbidites (for reviews see, Middleton 1970; Allen 1982). It is noteworthy that almost all of this previous work has been based on descriptions of gross structural or stratigraphic features which can be identified directly in outcrops or sediment cores. These features include variability in apparent grain-size such as graded bedding and silt-mud laminae, but they can all be identified without the
aid of grain-size analysis. For example, the work of Harms & Fahnestock (1965), on the basis of similarities in the layering and surficial bedforms, compares Bouma's divisions with laboratory and river sediment deposited under different decreasing flow conditions. Some size distributions are presented in this and other work but more to give a general idea of grain-size range than as a diagnostic tool. Exceptions are Scheidegger & Potter (1965) and Potter & Scheidegger's (1966) studies of grain-size and vertical variability, and Middleton's (1967) distinction of distribution grading and coarse-tail grading. Middleton found that coarse-tail grading characterizes turbidites deposited from high concentration surges and distribution grading deposits from low concentration suspensions, interpretations which could not have been made from visual inspection alone. The value of grain-size analysis is excellently demonstrated by Stow & Bowen's recent use of detailed size analysis to postulate a model for the origin and grain-size distribution of the silt-mud laminae in turbidites from the Scotian margin (Stow & Bowen 1978, 1980). They postulate a mechanism of shear sorting whereby flocculated sediment arriving at the bottom-water interface is broken up by shear forces. Initially only the largest silt size of particles settle through the boundary layer and are deposited. The fine mud mud portion forms a layer of suspended mud of increasing concentration which eventually produces aggregates strong enough to withstand shear break up and rapidly deposits as a mud 'blanket'. The progressive fining of the silt laminae and the correspondence of the size at which particle percentages fall off at the coarse end of the mud grain-size distribution with the
83
84
K. Kranck
increase at the fine end of the silt peaks were seen as evidence that a series of mud silt laminae were deposited from one suspension in a waning flow. The purpose of this paper is to analyse further the grain-size characteristics of fine-grained turbidites and to examine the mechanism for laminae formation proposed by Stow & Bowen (1978, 1980).
overall stock distribution gives all grains the same relative settling rate and results in a bottom sediment which also has this same size distribution. The absolute settling rate will depend on the rate at which the flocs form, i.e. for any given suspension on the concentration. Stokes' settling, one size at a time, tends to produce well sorted sediment whereas the absence of sorting in floc settling produces a broad flat grain-size distribution similar to that of the parent suspension. Sediment formed through a combination of both types of settling will have a well sorted modal peak and a tail offloc deposited sediment (Fig. 1). Although these experimental results are derived from settling in still water there is a close parallel to turbidity current deposition. In both cases sediment settles from an initially high concentration suspension to form a deposit characterized generally by decreasing grain-size with time. In still water some Stokes' settling always occurred so that the modal and maximum sizes always decrease. In a turbulent flow, however, sediment may be kept in suspension by turbulence and floc settling will decrease the overall concentration, but the grain-size range does not change as in case B in Fig. 1. The spectral form of the Stokes'-settled portion of a sediment may be predicted from the size distribution of the source suspension. From a well mixed suspension the flux of given size particles to the bottom is given by
Settling model The grain-size analyses of sediments used in this study were examined in the light of a grain-size settling model (Fig. 1) built on earlier studies of depositional behaviour of sediments and described below. A series of settling experiments (Kranck 1980) performed with different concentrations of sediment suspended in both salt and fresh water showed that whereas in normal Stokes' settling, one size class at a time disappears from a given depth in the suspension, during settling of flocculated material all constituent grain-sizes settle at the same rate. In the former case, grain-size is the controlling factor of settling rate but in the latter case, the concentration of sediment in the suspension controls settling rate. This difference is due to the formation of flocs or settling entities each of which is made up of grains which have the same proportional size distribution as the whole suspension. This duplication within each floc of the STOKES'SETTLING
F=Cw
(1)
BOTH
FLOC S E T T L I N G
(.,9
Z
~
LOG d
LOG d
~
LOG d
---~ LOG d
LOG d
~
LOG d
FIG. 1. Sketch illustrating settling model used to interpret grain-size analysis. Numbers refer to portions of sediment populations removed from suspension and deposited progressively with time or distance away from source. Stokes' settling from an unsorted sediment in suspension leads to deposition of one grain-size at a time and produces a well sorted bottom sediment. Floc settling leads to deposition of flocs which contain a proportional amount of all grain-sizes and deposits a sediment with the same relative grain-size distribution as the parent source suspension. Sediment formed by a combination of both settling processes will be a mixture of both settling types depending on the relative importance of each for the particular locality and grain-size range.
Grain-size characteristics
where C = concentration and w = settling rate of a given grain-size in the source suspension. According to Stokes' Law the settling rate is proportional to the square of the diameter. If C were constant for all grain-sizes (i.e. independent of d) the size distribution of the resulting bottom sediment would be quadratic i.e. it would exhibit a slope of two on a logarithmic plot. The size distribution of the source suspension may be determined from the finest grained portion of the bottom sediment formed exclusively from grains which all deposit at the same rate while parts of flocs, and thus duplicate the size composition of the source suspension. If it is assumed that the size distribution of a source suspension is also described by a power law relationship, a bottom sediment may be seen as a two-component mixture of two sub-populations defined by equations (2) and (3) below (Fig. 2).
V f = aD'"
(2)
Vs = bD ''+z
(3)
where V(and V, are the volumes in logarithmic
of turbidites
85
intervals of particle diameter as percent of total volume, a and b are the intercepts of the floc and Stokes' settled sub-populations respectively, D is grain diameter and m, the slope of the source suspension (Fig. 2). The function of the whole population distribution becomes V = aD m + bD m+ 2
(4)
Equation (4) describes only the curve for particles smaller than the mode, i.e. positive slope segment of the distribution curve. The coarser negative slope to the right of the mode should be very steep reflecting truncation at the maximum grain-size carried in the suspension. This results in a strongly negatively skewed distribution.
Description of material Turbidites were sampled from three cores; two from the Laurentian Fan (Stow 1981) and one from the Sohm Abyssal Plain (Vilks et al., unpublished).
I0
$= f
0:2.4
3
=, o .I b :.062 rl
.01
I
1oo DIAMETER,(ffm)
FIG. 2. Example of grain-size distribution spectra illustrating method of deriving constants in equations 2, 3 and 4 (see text).
86
K. Kranck
VVvx I
I
0%
2
0%
6OO
UNIFORM MUD E3
2 2% 610 8%
3
20%
GRADED MUD E2
--620
40%
-4.
58%
-5
0%
63O
SILT-MUD LAMINAE El
-6
66%
I 640
,
#
,
0%
FIG. 3. Muddy turbidite subdivided according to sedimentary structures (Piper 1978) with grain-size spectra for selected samples. Units of axis same as in Fig. 2 with 10% the highest gradation shown on Y-axis of each graph. Percent values give floc settled proportion of sample. Three digit numbers give core depth in centimetres.
.....- _78 SILT-MUD D - - ' - ' - ' - - - - -9 ..... -IOll
65O
97%
/VVvv', i
I I I
#
i
Grain-size characteristics o[ turbidites Core 80-016-015 from the Sohm Abyssal Plain consisted of a 11.8 m sequence of distal silt and mud turbidites 0.5 to 0.2 m thick, bottoming out in sandy beds also of turbidite origin (Vilks et al. unpublished). In the turbidite chosen for sampling (Fig. 3) a thin (4 cm) sequence of silt and fine mud interlayered as 2-5 mm laminae forms the lowest layer (tentatively identified as a Bouma D division). The mud and silt are well separated and uncontaminated samples could be obtained of each. This is overlain by a 15 cm sequence of very finely laminated silt and mud. The laminations are diffuse and faint and no attempt was made to sample particular features of the layers. This appears to correspond to Piper's (1978) Et subdivision. Above this is another approximately 15 cm thick unit in which the laminations disappear and an even gradation of increasingly finer material is indicated by change in colour (Subdivision E2). A fairly distinct boundary separates this from a 15 cm uniform mud with no visable structures (Subdivision E3). In Core 73-011-9 from the Laurentian Fan one complete 25 cm long turbidite was sampled (Fig. 4). Over half the sequence consisted of sandy material. The lowermost clean-looking fairly uniform sand with only minor faint lamination is identified as Bouma B division. Next is a Bouma C division sand with several intervals of crosslamination and finally a sequence with distinct sand-mud laminae each several mm thick and easily sampled (Bouma D). Overlying this D division is a graded mud similar in appearance to the E2 sequence of core 80-016-015. These two turbidite beds from the Sohm Abyssal Plain and Laurentian Fan were chosen as representative of the cores as well as for their conformity to the ideal Bouma turbidite sequence. Between them they contain all the classical Bouma (1962) divisions except the A layer. In order to investigate further the origin of fine-grained bedding structures, a short less classical sequence was sampled in detail from Core 79-021-37, also from the Laurentian Fan. Much of this core was characterized by irregular intervals of short graded silt-mud sequences alternating with silt-mud laminations on different scales. A mud and a silt sample were collected from each of three silt-mud couplets and five samples from a graded bed near the interlayered sequence (Fig. 5).
Analytical methods All the cores were subsampled with a I mm wide spatula which was inserted parallel to the bedding to obtain approximately 1 g of sample. A minis-
87
cule amount of this was subsampled for size analysis. The grain-size distributions of the samples were analysed using a Model T A l l Coulter Counter interfaced with a HP-85 computer system. Calibration methods and sample processing followed methods as described in Kranck & Milligan (1979). The samples were disaggregated in a small amount of 10% (NaPOD6 solution, suspended in 3% NaCI solution and analysed using a 30 gin, and a 200 #m orifice tube for all samples as well as a 400/ml tube for the coarser sands. The samples were deflocculated by ultrasonification and the size analyses represent the single mineral grain distributions. The volume in each size channel (designated by size of channel midpoint) was calculated as a percentage of the total in all channels. Particle volume doubles in each consecutive channel (equivalent to l/3q~), resulting in 24 to 36 data points per analysis including some overlap between tubes. The results were plotted as log-log frequency distribution-spectra (Figs 3-5). This display method was chosen rather than the more commonly used arithmetic-log axis plots because portions of different samples with similar relative size composition plot as similar shaped curves irrespectively of the nature of the rest of the size distribution. For each turbidite the slope of the source suspension (equation (1)) was determined by measuring the slope angle of the size spectra between 1 /Lm and 2 #m. The slope values were averaged and a distribution for the floc settled portion of the sample was calculated from equation (1) using the 1 pm volume percent as the intercept. The calculated floc settled distribution was subtracted from the total size distribution to obtain the size distribution of the material that did not settle as flocs (i.e. the Stokes' settled material). In Figs. 3 to 5, each sub-population is marked on the spectra plot of the total distribution and the percent Stokes' settled material in each sample is listed beside the graph.
Results In all three turbidite sequences analysed, there is a general upward decrease in modal grain-size although in the D and El divisions the intermittent mud layers temporarily interrupt this trend. The size distributions are negatively skewed. The mud samples have a gentle positive slope which varies by less than five degrees from sample to sample, and a steeper negative slope. With increase in modal size, a modal peak becomes more prominent and in the coarser samples only the finest grain-sizes have the same slope as the fine
88
K. Kranck ~VVVX/v
210
••,
!
~
, 0%
O%
-2
GRADED MUD E2
~
-3
0%
I
SAND-MUD . . . . . . LAMINAED~.. / ~/"
I
I
-4 14%
-
/ t
CROSS- V . ' / i ' BEDDING
.Z 1-8
C I"/" I.'1." l
........... ~
/,.~
' 9
83%
!
230crnt
!
!
I
I
I
I
I
I
SAND B U
tt~ll L~
96%
/ ,
,
~
94% I I
I
I
I
, ~ s /I I
FIG.4. Sandyturbidite subdivided according to Bouma(1962)with grain-sizespectra for selectedsamples. Units on axis sameas in Fig. 2 with 10% The highestgradationshownon Y-axisof each graph. Percentvalues give flocsettledproportion of each sample. Three digit number givecore depth in centimetres.
I
92% .
I
95%
I0 /I
'
Grain-size characteristics of turbidites
89
FIG. 5. X-radiographs of turbidite portion sampled for graded layer (1-5) and sand mud cuplets (6-11) with grain-size spectra. Units on Y-axis same as in Fig. 2 with 10~ the highest values marked on Y-axis of each graph. Percent values give floc settled proportion of each sample. Three digit number give core depth in centimetres.
90
K. Kranck
muds. A few of the sand samples show a rise in the slope in the very finest sizes although this may be an experimental error related to electrical noise. Only in the E3 division of Core 80-016-015 was there no perceivable difference in grain-size between samples confirming that this material was indeed uniform. Subtraction of the postulated floc sub-population from the total population produced a nearly straight line for the Stokes' deposited sub-population. It is noteworthy that the slope of this exponential distribution is close to two in the non-laminated beds but becomes steeper and close to four in the lower coarser inter-layered portions of the turbidites. The percentage of Stokes' deposited material varies fi'om 97?,o to 0'~,o and reflects the general variation in modal size trend and increasing peakedness of the spectra.
Discussion The grain-size data from the three cores sampled appear to confirm the two-component model of settling (Kranck 1980): i.e. unsorted floc-deposited material deposited alternately or intermittently with single grain or Stokes'-deposited well-sorted coarser material. The presence of both of these types in one sample produces the negatively skewed modal peak and a fine, nearly straight tail characteristic of silty muds (Kranck 1975, 1980). These distributions plot as two nearly straight segments on cumulative probability plots and have been frequently described for various sediments (e.g. Inman & Chamberlain 1955) including turbidites (Piper 1973; Jipa 1974; Kepferle 1978; Stow 1981). The dissection of grain-size distribution into sub-population components is heavily dependent on high accuracy and high resolution analysis. Although the fine end portion of the size distribution could be satisfactorily defined mathematically, the function for the negative slope is not clear. It is difficult to determine empirically because in present methods of analysis the number of grains analysed in this region are very few, because an increase in sample concentration tends to cause coincidence from excessive numbers of small particles. Also it is not known if the rise in relative particle volumes at the fine end of the many sand samples (Fig. 4) is due to experimental errors associated with electrical noise or to some other effect such as trapping of very fine particles in pore spaces. The relationship of these results to the mechanics of turbidite deposition may be assessed by considering each division of the Bouma sequence in turn.
A Dit'ision: None of the core portions analysed contained examples of division A so its detailed granulometry could not be assessed. Kranck (1980) has likened the unsorted material which initially settles out of a flocculating experimental suspension to the basal portion, i.e. A division of natural deposits. Essentially this only adds flocculation to the earlier views (summarized by Middleton, 1970) of the basal or A division as a product of rapid deposition from a high concentration suspension where size sorting is inhibited. This division should have preserved within it the original size configuration of all but the coarsest size particles and a detailed study of its grain-size composition should prove profitable. B Dirision: This division, characterized by some plane parallel laminae, may be seen as a transition between division C, where traction transport of sand along the bottom is well established as indicated by ripple bedforms and the massive totally unsorted A layer, where transport presumably ceased when grains reached the bed. If there is sufficient floc disruption by bed shear to create a clean sand, this when superimposed on muddier A material may form the lower reverse graded portion of some turbidites. C Dirision: There is no perceptible difference between the size spectra of samples from B and C divisions which therefore is a classification based solely on structural evidence. Both B and C divisions have modal peaks almost entirely consisting of Stokes'-deposited material although a distinct tail of mud is also present. The slope close to 4 indicates that this material has been deposited and subsequently resuspended presumably to form a near-bottom traction-saltation load before again settling. Such near-bottom transport is in accord with the structures present and the need to expel the mud fraction. It is in agreement with reports of strong near-bottom turbulence reported from experimental turbidite flows. D Division: This is the division in which mud deposition first becomes a significant factor and the texture to which Stow & Bowen's (1978, 1980) lamination mechanism is most relevant. During its formation, turbulence has apparently decreased to a point where mud can deposit intermittently. A detailed examination of Stow & Bowen's model is outside the scope of this paper, but these results in general confirm their contention that the mud and silt laminae are deposited out of the same suspension. The rapid deposition of the mud suspension built up near the bottom is probably promoted by two factors: turbulence dampening and floc maturation. Heathershaw (1979) suggests that a concentration of suspended sediment (0.01 mm in diameter) as low as 15 rag/!
Grain-size characteristics of turbidites is able to modify turbulence and 130 rag/1 is the amount required to dampen it completely. The exact critical concentration threshold will be a function of many factors including grain-size and composition. Floc maturity is a term which may be used to describe the shear strength of the floc. Kranck & Milligan (1980) demonstrated experimentally that given sufficient time a suspension of fine-grained sediment will form flocs able to withstand very high shear rates. An important factor in this is the adhesive action of organic detritus and surface bacteria. The combination of turbulence dampening and floc maturity may cause deposition of mud laminae from even very dilute suspensions. E~ Division: The major difference between this subdivision and the D division may be one of scale. In the latter, mud uncontaminated by silt could be sampled and analysed; in the former the 1 mm thick randomly extracted samples all contained a portion of Stokes'-settled grains. Greater sampling resolution may or may not succeed in isolating the true sedimentation unit: the question has only limited relevance in explaining the depositional mechanism, which is assumed to be similar for both divisions. E: Division: As was the case in differentiating Divisions B and C, the El and E2 divisions can only be distinguished visually. The samples in both El and E2 have Stokes sub-populations with positive slopes of 2 indicating direct settling without resuspension. E.~ Division: The complete lack of Stokes'-settied fraction and of any changes in the grain-size indicates this sequence has settled entirely by flocculation. The suspension from which it has been deposited must be in equilibrium with the turbulence in the water so that only by forming flocs can the particles overcome upward advection. The only conditions under which a uniform sediment will form from a non-constant source is when the suspension is completely flocculated so that each particle which arrives at the bottom is a floc containing equal proportions of all grains from the suspension. In turbidite sedimentation
91
this may occur at the beginning of a flow when concentrations are very high, so that floc settling rates greatly exceed Stokes' settling rates, or alternatively at the end when there are no grains left which are able to settle singly from in the suspension. F Dirision: In these cores there were no signs of bioturbation or other features characteristic of normal pelagic sedimentation designated as F division by Piper (1978). If such material was laid down between the turbidite episodes it must have been eroded or cannibalized by the succeeding flow.
Conclusions The examination of the fine detail of the textural changes in individual turbidites shows the importance of grain-size properties in understanding depositional mechanisms. In this paper only deep-sea turbidites have been examined but similar studies are a prerequisite to the development of a truly genetic classification of fine-grained sediment. Modern electronic particle analysers and computer processing of results allow a large number of samples to be processed rapidly. This is especially valuable in the case of fine-grained sediments where individual grains cannot be seen directly. In the deep-sea sediments where depositional conditions are hard to observe directly, fine-grain granulometry should be a standard supplement to conventional studies. Similar investigations in the future of contourites, red and black pelagic muds and other facies would help settle many of the problems regarding their origin and stratigraphic significance. ACKNOWLEDGEMENTS: I am indebted to D. Piper and G. Vilks who provided the turbidite material used in this study. The size analyses were performed by T. Milligan and D. Nelson. Many colleagues have been helpful in the preparation of the final manuscript.
References ALLEN, J.R.L. 1982. Sedimentary Structures. Their Character and Physical Basis. Vol. II. Elsevier, Amsterdam, BOUMA, A.H. 1962. Sedimentology of some Flysch Deposits. Elsevier, Amsterdam, 162 pp. HARMS,J.C. & FAHNESTOCK,R.K. 1965. Stratification, bed forms and flow phenomena (with an example from the Rio Grande). Soc. econ. Palaeo. Min. Spec. Pub., 12, 84-115.
HEATHERSHAW,A.D. 1979. The turbulent structure of the bottom layer in a tidal current. Geophys. J. R. astr. Soc., 58: 395-430. INMAN, D.L. & CHAMBERLAIN,T.K. 1955. Particle-size distribution in near-shore sediments. Soc. econ. Palaeo. Min. Spec. Publ., 3, 106-29. JIPA, D.C. 1974. Graded bedding in recent Black Sea turbidites: a textural approach. In: Degens, E.T. & Ross D.A. (eds), The Black Sea-Geology, Chemistry
92
K. Kranck
and Biology. Am Ass. Petrol. Geol. Mem. 20. KRANCK,K. 1975) Sediment deposited from flocculated suspensions. Sedimentology, 22, 111-23. 1980. Experiments on the significance of flocculation in the settling of fine-grained sediment in still water. Can. J. Earth Sci., 17, 1517-26. -& MILLIGAN, T. 1979. The use of the Coulter counter in studies of particle size distributions in aquatic environments. Bedford Institute of Oceanography, Report Series BI-R-79-7. 1980. Macroflocs: Production of Marine Snow in the Laboratory. Marine Ecol, Progress Series 3: 19-24. KEPFERLE,R.C. 1978. Prodelta Turbidite Fan Apron in Borden Formation (Mississippi), Kentucky and Indiana. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 224-38. KUENEN,PH.H. 1951. Properties of turbidity currents of high density. Soc. econ. Paleo. Min. Spec. Pub., 2, 14-33. 1966a. Matrix of turbidites: experiment approach. Sedimentology, 7, 267-97. 1966b. Experimental turbidite lamination in a circular flume. J. Geol., 74, 523-45. MIDDLETON, G.V. 1967. Experiments on density and turbidity currents III. Deposition of sediment. Can. J. Earth Sci., 4, 475-505. 1970. Experimental studies related to problems of -
-
-
-
-
-
-
-
-
flysch sedimentation. In: Lajoie, J. (ed.), Flysch Sedimentology in North America. Geol. Ass. Can. Spec. Pap., 7. 253-72. PIPER, D.J.W. 1973. The sedimentology of silt turbidites from the Gulf of Alaska. In: Kulm, L.D., von Huene, R. et al., Initl. Repts. DSDP, 18, 847-67. 1978. Turbidite Mud and Silt on Deep-sea Fans and Abyssal Plains. In: Stanley, D.J. and Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. POTTER, P.E. & SCHEIDEGGER,A.E. 1966. Bed thickness and grain size: graded beds. Sedimentology, 7, 233-40. SCHEIDEGGER, A.E. & POTTER, P.E. 1964. Textural studies of graded bedding, observations and theory. Sedimentology, 5, 289-304. STOW, D.A.V. 1979. Distinguishing between finegrained turbidites and contourites on the Nova Scotian deep water margin. Sedimentology, 26, 371-87. & BOWEN, A.J. 1978. Origin of lamination in deep-sea, fine grained sediments. Nature, 274, 324-28. 1980. A physical model for the transport and sorting of fine-grained sediment by turbidity currents. Sedimentology, 27, 31-46. 1981. Laurentian Fan: morphology, sediments, processes and growth pattern. Am. Ass. Petrol. Geol. Bull., 65, 375-93. -
-
-
KATE KRANCK, Atlantic Oceanographic Laboratory, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, B2Y 4A2 Canada.
Fine-grained sediments of the Zaire deep-sea fan, southern Atlantic Ocean T.C.E. van Weering and J. van Iperen SUMMARY: The Zaire deep-sea fan is a large mud-dominated system fed mainly by the Zaire (Congo) River. A submarine canyon deeply incised into the outer estuary funnels part of the river's high suspended load directly out to the deep sea. Sediment is also contributed to the fan via slumping and debris-flows initiated on the upper slope. A number of 11-17 m long piston cores have been collected from this part of the SE Atlantic Ocean, and three of these are described in detail in terms of structure, texture and composition. Four sediment facies are recognized: (1) turbidite muds with mainly TEl, TE2 and TE3 structural divisions (after Piper 1978): (2) pelagic and hemipelagic biogenic muds: (3) homogeneous uniform muds that probably result from slow settling of dilute suspensions: and (4) debris-flow deposits. There are marked differences in facies distribution, turbidite thickness and organic carbon content between the upper and outer regions of the fan. The marine geology group of the Netherlands Institute for Sea Research has carried out a research programme in the Zaire (formerly Congo) River, its estuary, the adjoining part of the continental shelf and slope and in the Angola Basin during cruises held in 1976, 1978, and 1980 (Fig. 1). The aim of the investigations was the study of the physical and chemical processes involved in the outflow of the Zaire River to the Southern Atlantic Ocean and the study of the recent and sub-recent sediment dispersal pattern in the adjacent part of the continental slope and the Angola Basin. Results on the chemical and hydrographic processes in the Zaire outer estuary and plume have been reported by Eisma et al. (1978). This area was selected as the Zaire River has the world's second largest drainage basin (3.7x 106 km 2) and mean annual discharge (45 000 m3s -~) and is the main contributor of terrigenous sediments to the eastern South Atlantic Ocean, together with the River Niger. Moreover, although it is known that a part of the sediment load of the river is funnelled directly to the deep sea via a canyon which has incised itself already down to 100 m in the lower estuary, the Zaire deep-sea fan and its sediments have been studied only to a limited extent (Heezen et al. 1964; Bornhold 1973; Shepard & Emery 1973). In this article some selected cores will be described which are considered as representative of the depositional areas and processes on the Zaire fan.
Geological and oceanographical setting Geologicalsetting The Zaire fan lies in the Angola Basin, the deepest depression of the eastern South Atlantic. This
basin is a restricted basin bounded by the Walvis Ridge to the south, the Mid Atlantic Ridge to the west and the Guinea rise to the north (Fig. 1). The continental shelfofZaire and Angola is about 100 km wide and has its break at a depth of 100-120 m. Sediment studies of the Angolan shelf are not known to us; the shelf of Zaire, Cabinda and Congo has been extensively studied by Giresse and his co-workers (Giresse & Odin 1973; Giresse & Moguedet 1974: Giresse & Cornen 1976; Giresse et al. 1979) and is partly summarized in a recent article by Giresse et al. (1982). The typical shelf sediments are mainly finegrained, forming a blanket of muddy sands and sandy muds over the entire shelf. Clays are derived from the south and consist of 65% kaolinite, 251~0 illite-montmorillonite and some 10~ illite. Faecal pellets are found all over the shelf and are glauconitic at depths of less than 25 m and more than 110 m. Between 100 and 120 m relict shelly sands form a submarine terrace. The shelf is dissected by the Zaire canyon, which has eroded down to 100 m depth at 30 km from the river mouth in the outer estuary and down to 425 m at the river mouth. The canyon and fan valley have been studied by Heezen et al. (1964) and in greater detail by Shepard & Emery (1973), showing that the canyon and fan valley continue as far as 800 km to the west. The canyon has eroded deeply into the continental slope; in a zone landward of a belt of diapirs, the incision below sea bottom is from 400-1400 m and has a characteristic steep-sided V shape. Within the diapir belt, presumably salt diapirs of Aptian age (Baumgartner & van Andel 1971~ Emery 1972; Lehner & de Ruiter 1975), the canyon decreases in depth to about 300 m below sea bottom; its slopes are irregularly-shaped probably because of slumping and/or local uplift 95
96
T.C.E. van Weering and J. van Iperen
FIG. 1. SE Atlantic Ocean and study area. Depth contours are in metres. Main topographical features are the Mid-Atlantic Ridge, Walvis Ridge and the Guinea Rise.
through doming of the diapirs. Locally the canyon becomes less deep and more flat-floored seaward of the diapir belt; here it is flanked by levees up to 70 m high. The diapir zone has produced a sea bottom relief of several hundreds of metres through local updoming; it ends abruptly to the west with a steep escarpment, the Angola escarpment. The lower continental rise has some relict hills at depths between 4000 and 4500 m (Bornhold 1973). In the western part of the Angola Basin various seamounts interrupt the flat-lying sediments. Heezen et al. (1964) and Bornhold (1973) studied the sediments to some extent.
Hydrography The surface water circulation in the Angola Basin is primarily influenced by two main current systems, the Benguela Current and the South Equatorial Counter Current (SECC). The Benguela Current, dating from about 10 million years BP (Siesser 1980), transports relatively cold waters with velocities of 15-20 cms -1 northwards in a broad zone along the African coast up to about 15S., where the current curves to the west and meets the SECC (Fig. 2). The interaction between the currents leads to a complex pattern of gyres and local upwelling phenomena; the distribution
Fine-grained sediments of the Zaire deep-seafan
97
FIG. 2. Surface current pattern in the SE Atlantic Ocean. of silicoflagellates in a number of selected cores of the Angola Basin also points to local upwelling conditions (de Ruiter, pers. com.). A small branch of the Benguela Current continues towards the north and induces a northerlydirected longshore drift along the coast of Angola, probably resulting in the formation of beach barriers at the mouth of the Zaire River. There is evidence for a subsurface countercurrent beneath the Benguela Current, along the edge of the continental shelf with its core at 300-400 m water depth and with a southerly direction (Hart & Currie 1960). The SECC flows to the south-east at about 7-8~ and forms a gyre
where it meets the Benguela Current. A southerlydirected branch continues off the coast of Angola as the Angola Current, while another branch turns northwards off the coast of Zaire and Congo and probably influences the direction of the Zaire River plume. Recent research (van Bennekom & Berger 1983) shows that the freshwater outflow of the Zaire River is noticeable as far as 800 km offshore, the plume meanders and forms lobes in response to variations in windstress, river discharge and current strength. The direction of the plume is south-West or south-south-west in February and March, west or south-west from
98
T.C.E. van Weering and J. van Iperen
April to August and to the north-north-west in October and November and part of the southern summer. The deeper waters in the Angola Basin consist of North Atlantic deep water and Antarctic bottom-water which enter the basin via sills in the Romanche and Chain fracture zones (Wrist 1935). Current velocities at the bottom are very low (van Bennekom & Berger, 1983). Suspended matter Data on suspended load (Eisma & Kalf 1983) show that the Zaire transports annually an average of 43 x 106 tons of suspension load. About half of this load settles in the head of the canyon, and only 5% reaches the deep sea, the remainder being deposited on the shelf and slope or remaining near the bottom. The total content of suspended matter decreases from 10 m g L - 1 to 2 m g L - 1, going from the river mouth to the edge of the shelf. According to Eisma & Kalf (1983) flocculation affects only a few percent of the total suspension load. In the open ocean the concentrations of suspended matter are well below 1 mgL -l. The amount of organic matter in the suspension load varies from about 30% in the estuary to over 60% on the shelf. In the open ocean the suspended sediment comprises 70-90% biogenic and organic material (Emery & Honjo 1979), mainly organic aggregates (90~o), skeletal debris (8%) and mineral grains (2%).
Methods Piston cores of up to 18 m length were obtained by means of a modified Alpine piston corer. For the deep-sea coring a special Kevlar cable was used with the same breaking strength and stretch as a steel cable but considerably lighter. On board, the cores were cut in 120 or 150 cm long sections and stored at 4 C . The sections were split as soon as possible after retrieval, described and photographed. Colours were noted using the standard Rock Colour Charts of the Geological Society of America. On board, X-radiographs were made of the split sections in order to delineate internal sedimentary structures. The X-ray apparatus used was a Hewlett Packard Faxitron N-407 system, using Agfa Gevaert D-7 film. In the laboratory a number of fine-grained turbidites were selected for further detailed study and description, as well as for grain-size analysis and mineralogical determination. Grain-size analysis was performed with standard sieve and pipette methods, applying whole phi-intervals (Folk 1968). The carbonate content
of the samples was calculated through Ca determination of the samples by fluorometric titration with EGTA. Organic carbon was measured by means of a Perkin-Elmer CHN Analyser type 240-B. The sedimentary structures were described from the X-radiographs, following the modified turbidite terminology of Piper (1978). He subdivided Bouma's (1962) E division into turbidite mud (TEl laminated, TE2 graded, TE3 ungraded), and pelagic mud (F).
Results During the cruises a number of piston cores were taken in order to characterize sediments and depositional patterns of the shelf, the upper and lower slope, the deep-sea fan and the abyssal plain (Fig. 3). Only part of the cores revealed the presence of turbidite structures, in contrast to the rather structureless, homogeneous sediments forming the greater part of the shelf and slope (Fig. 4). The cores presented in some detail below were selected for further study as they show systematic changes in fine sediment dispersal across the fan. The main aim of the coring programme was to study the lateral extent and variation of turbidite sedimentation. (1) Core 38 was taken at a water depth of 5490 m in the outer deep-sea fan abyssal plain area, which is interrupted by some sea-mounts in the direct neighbourhood. Within the core, three turbidite intervals can be recognized, separated by rather homogeneous structureless, locally bioturbated and mottled pelagic sediments (Figs 5 and 6). (2) Core 42, obtained from a water depth of 4460 m, is considered representative of the outer part of a middle-fan area; probably this is an interchannel deposit cut off from the main turbidite supply. The lower 530 cm of the core is entirely turbiditic, whereas above this level, homogeneous partly bioturbated sediments are found (Figs 7 and 8). (3) Core 32 at a depth of 3600 m, was taken on a levee near the main channel of the upper fan. This core consists entirely of turbidites (Figs 9 and
10). Colour In core 38 the sediments range from dark greenish-grey (10 GY 4/4) via dark green (10 G 4/1) pelagic intervals, into olive-grey (5 GY 6/1), olive-brown (2.5 Y 4/3) and dark brown homogeneous sediments (10 YR 3/3). The turbidite intervals at the top are, in general, darker olive-
Fine-grained sediments of the Zaire deep-sea fan
99
"0
0 "0
O
0
"0
0 "0
"0
R r
.=. o~ fz~
R.6
0
.~ "0
9
0
I oo
T.C.E. van Weering a n d J. van Iperen
Fro. 4. Correlation of climatic zonation and occurrence of turbidite intervals (horizontal rule) in cores from the SE Atlantic Ocean.
grey and greenish-grey. The turbidites in core 42 have a grey (10 Y 4/1) to dark olive-grey colour (5 GY 3/1) and are only slightly darker than the greenish-grey (7.5 GY 4/1) to olive-grey (2.5 GY 5/1) non-turbidite sediments. The sediments of core 32 are generally dark and range in colour from olive-black (5 GY 2/1) in the lower part of the turbidites to slightly lighter, though still olive-black (7.5 Y 3/2) sediments, forming the top of a single turbidite. Within the lighter sediments, mottling occurs on a minor scale Structure
The turbidite intervals in core 38 show alternations of To (occasionally), TEl, TE2 and TE3
divisions. In the lower part of the turbidite interval TE1/TE2 alternations are most frequent, the TEl division having a sharp, clearly recognizable lower boundary. The TEl division is mostly thickly-laminated (3-6 mm thick laminae) or very thin-bedded (6-15 mm). The TE2 division is occasionally reversely-graded. Towards the top of the turbidite intervals, TEl divisions become scarce, TE2 and TE3 have a greater thickness and are slightly burrowed (Fig. 5). The TE3 interval grades gradually into the pelagic F division; upper boundaries of the turbidite intervals are therefore difficult to recognize. Syn-sedimentary deformation is seldom observed (Fig. 6). Towards the top of the core the turbidites are covered by rather homogeneous bioturbated
Fine-grained sediments o f the Zaire deep-sea f a n
I 01
FIG. 5. Columns showing the climatic zonation, the carbonate and organic carbon content and the turbidite intervals (T) of core 38. Climatic zonation and age after Jansen et al. 1983. Length of the core indicated in centimetres. mottled sediments, forming the greater part of the sediment column. The turbidite sections of core 42 consist almost entirely of alternations of TEz/TE3 divisions and of TE3/F divisions, in a setting which changes from turbidite-dominated into pelagic. TEl divisions occur only rarely, and are thickly-laminated (3-6 mm). The TE2 and TE3 divisions are thicker in the lower part (with a mean of 35 ram), changing gradually into thinner-bedded turbidites towards the top of the turbiditic interval (Fig. 8). Almost all the turbidites are covered by F divisions, separated by a clear boundary. The F divisions increase in thickness towards the top of the turbidite section; in the lower part they vary between 5 and 13 mm, with a mean of 7.5 mm, whereas upwards the F division is at maximum 32 mm and has a mean thickness of 16 mm. Over the complete turbidite interval, the ratio of pelagic sediments to turbidite sediments is 1: 3. Core 32 shows laminated TE~ divisions, TE: graded muds and TE3 muds. F intervals are scarce, but increase towards the top (Fig. 10). The
TE~ divisions are mainly thickly-laminated (3-6 mm), although many are thinner ( < 3 ram) and more rarely, thicker. Most Tvz divisions occur together with TE2 divisions, sometimes followed by TE3. Sequences without TEl divisions are also present. The mean thickness of a complete turbidite sequence is at maximum in the lower part of the core and decreases towards the top. Although the F intervals increase towards the top of the core, their thickness seldom exceeds 10 mm, and is mainly around 5 mm. The ratio of pelagic to terrigenous sediments is 1 : 20. Various levels with clay pebbles are found. The clasts are oval to circular, more rarely angular, and have a longest axis of 0.3-2.5 cm. Burrowing activity in the sediments is scarce in the lower part and somewhat more evident in the upper part. Most burrows are syn-sedimentary.
Grain-size and composition Grain-size determinations of some selected turbidite intervals were carried out. In core 38, the TD
T.C.E. van Weering and J. van Iperen
[o2
CORE
38
30-150
cm
IIo 3
7(
4
%
9
:'~ '~w~,~" ....
....-
"S
?
"S
i
"S
a. 0F-
T 50"
:
i i I
:
....
....
,S
']
~
70-
,,v-
,S .S
FIG. 6. Radiograph of sedimentary structures in lower turbidite interval of core 38. Scale in centimetres. S denotes samples used for grain-size determinations shown in Fig. 11. Note horizontal layering disturbed by bioturbation in lower part of core. Turbidites are TD/TEI and TEI/TE2 alternations changing into TE3 interval. Note also occurrence of burrows in TE3 intervals at 95 and 123 cm.
of the Zaire deep-sea fan
Fine-grained sediments
IO3
CORE 42 sediment column
•
g
o N
carbonate 0
2 ,
I
4 j
I
6 ,
I
8 ,
I
10 ~
I
12 ,
I
3m
,0-
00"
I
0 A
organic 1
J
carbon 2
I
5 % I
B, 29.7
18- 00_ 0
'3-
14 % ,
~9 - . ~ 2 0 9
'0)5-
O0" bi, 2 0 . 8
;20 O-
goo. I o~ to
_:--T
o. g. -9- ' ~
57.5
o
F~G. 7. Columns show climatic zonation, carbonate and organic carbon content, and turbidite interval (T) of core 42. Climatic zonation and age after Jansen et al. (1983). Length of the core indicated in centimetres. interval samples at 83 89 cm has a sand content (> 63/~m) of 20%, and a coarse silt (32-63 ~m) mode of about 55~Jo (Fig. 11). Several TEl intervals samples show lower sand content, still have coarse silts as the modal fi'action but contain a somewhat higher percentage of very fine silt and clay. There is apparently a change into a bimodal grain-size distribution from TD via TEt into TE2. The sand and coarse silt fraction contains abundant quartz, some mica, many diatoms, and sponge spicules, together with some radiolarians, and rare sponge spicules embedded in faecal pellets. Pyritization is relatively common. The finer fractions are mainly clays and siliceous ooze (fragments of diatoms, radiolarians and silicoflagellates). In core 42, the TEl division consists primarily of coarse and very coarse silts (63-16 pm) in addition to a considerable amount of fines ( < 4 #m), although the 4-8 ~m fraction is very small, and the distribution is therefore slightly bimodal. The TE2 and TE3 divisions are sometimes difficult to separate visually, and this is also reflected in their grain-size distributions, which are very similar,
both with around 80,~ of very fine silt and clay-sized particles (Fig. 12). In the TEl division, the sand fraction contains predominantly quartz and mica (rounded flakes), and considerable amounts of pyrite, diatoms (some partly pyritized) and sponge spicules. Some of the quartz grains are coated with haematite, as are some of the diatoms. Some of the diatoms in the turbidite intervals are clearly freshwater species. The silt and clay fractions consist of micas, clay minerals, coccoliths, fragments of foraminifera and phytoliths, and rare radiolarians. Plant remains occur throughout. The TEl divisions of core 32 characteristically have a coarse or very coarse silt modal grain-size, with a small amount of finer material. The TE2 divisions contain far more very fine silt and clay ( < 4 l~rn), and in the TE3 divisions this fraction accounts for over 60'~o of the sediment with only a minor coarser fraction (Fig. 13). The sand fraction consists of quartz at the base of some of the TEl divisions, with more mica in the upper part of the laminated units. The micas are often abraded and have rounded edges. Sponge spicules, dia-
IO4
T.C.E. van Weering and J. ~'an Iperen
FIG. 8. (a)-(b) Radiographs of core sections from core 42 showing increasing bioturbation in TE3 intervals towards the top of the core together with a decrease in thickness of TEl and TE2 intervals. Some of the light spots are pyritized burrows. Note local disturbance (at 25 cm) of TE2 interval.
Fine-grained sediments of the Zaire deep-seafan
Io5
Io6
T.C.E. t'an Weering and J. t'an Iperen CORE 5 2 sediment column
!I
0
carbonate 2 4%
organic carbon 5 4
5% I
~n
[,.,.
-=T~_
8-
g_ oO. FI~. 9. Columns of core 32 indicating climatic zonation, carbonate and organic carbon content. The entire core comprises turbidites. Core length in centimetres. Climatic zonation after Jansen et al. 1983.
S_0
toms and some foraminiferal fragments are also present. Plant fragments occur in all TEl divisions, orientated with their long axis parallel to the bedding plane. The silt and clay fraction consists of diatoms, coccoliths, mica, minor amounts of quartz, clay minerals, organic remains and phytoliths.
crease with increasing distance from the source (Fig. 14), being greatest in core 32 and least in core 38. The turbidites contain more organic carbon than non-turbidites in core 38, whereas there is almost no difference between the two in core 42.
Carbonate and organic carbon
Discussion
The carbonate content shows great variations due to a combination of dissolution and dilution with terrigenous material. In core 38 (Fig. 5), the peak value of 20.8% carbonate at the top of the core reflects recent depositional conditions, and indicates a slow dissolution of carbonates at the sea bottom. Other peaks in core 38 are the result of the occurrence of some carbonate rich clay pebbles of allocthonous origin. The carbonate in core 42 (Fig. 7) shows a number of consecutive peaks in the non-turbidite section, with peak values of 29.7~ at the top of the core. The turbidite section has a lower carbonate content with rather less clear peaks as a result of dilution by the increased terrigenous supply. This effect is still clearer in core 32 (Fig. 9) where the amount of carbonate is low compared to other cores, with values between 0.5 and ~o/ The content of J / o" organic carbon shows a clear tendency to de-
Piston cores from the continental slope, the upper part of the Zaire submarine canyon, and of the Zaire fan were studied by Heezen e t al. (1964). They recognized four sediment facies on the basis of sedimentary structures. Facies I is a homogeneous silty iutite, facies II a crumbly silty lutite, facies III contains graded silts and sands and facies IV is formed by laminated, alternately lutite-rich and lutite-free silt. These sediments are covered in the northern part by carbonate rich, grey and black lutites, and in the western and southern part of the study area by a reddishbrown lutite. The occurrence of facies IV and III was related to turbidity currents, triggered probably by high discharges from the Zaire River. A study of cable breaks between 1886 and 1937 showed that these invariably occurred in periods of high river discharge. Heezen et al. inferred that the Zaire
Fine-grained sediments of the Zaire deep-sea fan
IO7
FIG. 10(a)-(b). Radiographs of core sections from core 32. In the lower part of the core, thin To and TEl intervals alternate with TE2 graded muds. Distinction between TE2 and TE3 is difficult. F intervals increase in frequency and thickness towards the top. Light spots at 360 cm are angular and oval clay pebbles. Note slightly inclined lamination above 363 cm. Burrowing is scarce.
I08
T.C.E. van Weering and J. van Iperen
Fine-grained sediments of the Zaire deep-seafan
lO9
core58
% 60
83 ~,~- 8 4 cm
40 ~
8 6 - 8 6 ~2
Cm
93-93
~2
cm
~
200
63
4
63
4
,1 ~2-,2:m
0
63
4
6.3
4
63
4
_ _ 1131/2-,4 :r~
63
4
FIG. 11. Grain-size distribution of samples from core 38 showing decrease of very coarse silt component and increase in the < 4 #m fraction from TD--*TE1--*TE2.
submarine canyon today experiences 50 turbidity currents per century. This is a strikingly higher number than, for example, that suggested for basins off California, where a frequency of 1 per 400 years is estimated (Gorsline & Emery 1959), or for the western Mediterranean where a frequency was estimated of at least 3 per 2000 years (Rupke & Stanley 1974). In recent years the study of fine-grained deposits covering the deep-sea floor has increased considerably. A further subdivision of Bouma's (1962) turbidite division was made by Piper (1978). He subdivided fine turbidite muds into a laminated Ej division, a graded E2 division and an
ungraded E3 division; the overlying hemipelagic sediment he referred to as the F division, following Van der Lingen (1969) and Hesse (1975). Piper also stressed that, volumetrically, turbidite muds are two to ten times more important than turbidite sands. Previously, Rupke & Stanley (1974) subdivided fine-grained deposits from the Mediterranean into two genetically different mud types. Type A mud typically showed the same structures as Piper's El E3 and F divisions, and this type was recognized as turbidity current derived. Type B muds formed through pelagic settling. Stanley (1981) and Blanpied & Stanley (1981) described
core 42 % 80-
60 84
,~
,2-42~2cm
40
[~
,,_,,~0m
TE2
TE3
20,
o-
_
65
4
63
111-111 ~2cm
63
4
113~2-114cm
TE2
63
4
TE3
4
63
4
FIG. 12. Grain-size distribution of samples from core 42. Increase of the < 4 #m fraction from TEl ~TE2--*TE3 divisions reflects upward-fining trends.
T.C.E. ~,an Weering and J. t,an Iperen
I I0
core 3 2
i
6 0 84 2 1 - 211/2cm
231/2- 2 4 cm
27-271/2cm
40-
o
__.4-
63
4
63
4
63
4
63
4
63
4
63
4
o
60 :348 - 3 4 8 ~'2 cm
351 - 351 ~'2cm
40
TE 3
20
63
4
63
4
FIG. 13. Grain-size distribution of samples from core 32. Note difference in grain-size between TD and TEl intervals.
% orgonic ~ carbon ~
_
i
J
r r,-
~
I--
5
t--
i
613 b.-
I--
i
i
oo I,.-
co t.-
t--
I--
9 = meon volue
4
1
0
o
!
,oo
2~o
3oo'
4~o
-..
5oo'
6~o km
from
40 scource
FIG. 14. Relation between content of organic carbon and distance from source for a number of cores. T78-46 and T78-45 are indicated on Fig. 3 as 46 and 45 respectively.
Fine-grained sediments of the Zaire deep-sea fan "unifites' as 'lutite layers, often thick, comprising both structureless muds and faintly laminated muds revealing a fining-upward trend', from restricted basins in the Mediterranean. They ascribed the origin of unifites to a sediment gravity-flow continuum related to mud-charged turbidity currents and less dense turbid layer flows. They inferred extremely rapid sedimentation rates for the structureless muds of > 200 cm/1000 yrs. However, the unifite muds are not truly homogeneous and are partly comparable to Piper's laminated to graded and ungraded turbidite muds. An 'ideal' sequence for fine-grained turbidities comprising nine subdivisions (T0-T~) has been proposed by Stow & Shanmugan (1980), based on a study of both recent and ancient sediments. The complete sequence is rarely observed, and could not be established in the cores described in this paper either. Our own observations of the Zaire fan deposits show that four types of fine-grained deposits can be recognized in the area. Type 1 are turbidite muds, type 2 are hemipelagic sediments forming the F division on top of turbidites, type 3 are homogeneous, uniform sediments, and type 4 are the muddy debris-flow deposits found intercalated in core 32.
core 49
80-
II
I
Within the cores studied, only TD (occasionally), TEl, TE2 and TE3 divisions are present. TE~ divisions contain small amounts of sand and have their main mode in the coarse and very coarse silts. The TE~ division characteristically contains a small fraction of the 4-8 ~m fraction. The thickness of the TEj division varies but is < 20 mm in core 38 and < 10 mm in cores 42 and 32. The TE2 and TE3 divisions have a considerably greater thickness in cores 42 and 32. Most logically, the fine-grained turbidites from the Zaire fan form the distal end members of channelized, gravity-induced mass-flow in the canyon and upper fan. On the upper fan the channel levees are up to 70 m high so that only the finest sediments settle on the flanks, resulting in the deposition of thin TEl and mainly of TEe and TE3 turbidites. Any sand-sized material found outside the channels will occur as very thin sheet sands, interbedded with silts and silty clays. In the outer fan area, where the levees are less pronounced, the dilute turbid cloud or turbidity current tail will more easily overtop the levees, resulting in somewhat thicker TD and TEl intervals. However, the greater part of the turbid cloud will probably not reach that far into the deep sea, and so thinner TE2 and TE3 turbidites will be deposited. A contour current origin for the
53 cm
5 4 6 cm
6 8 5 cm
_
60-
I
I 40
20
I
65
80
63
4
core 44
6B
4
i8 cm
4
8 0 cm
4 2 5 cm
60 1
40
20
65
4
6
4
6B
4
FIG. 15. Characteristic grain-size distribution of samples from the continental slope, uninterrupted by turbidite sedimentation. Compare with Figs 11, 12 and 13.
I 12
T.C.E. van Weering and J. van Iperen
finely-laminated and graded silts can be excluded because the bottom current velocities in the Angola Basin are very low (van Bennekom, pers. comm.). For the deposition of the homogeneous finegrained deposits encountered in many cores from the Angola Basin and Zaire fan (Fig. 15) another transport mechanism is suggested. These deposits most likely include the crumbly silty lutites and the homogeneous lutites of Heezen et al. (1964); although the crumbly homogeneous muds that we cored can be ascribed to the formation of gas through rapid decomposition of organic-rich sediments. Typically, these crumbly sediments are found in the upper part of the fan, where the content of organic carbon is twice as high as in the outer fan area (Fig. 14). This type of homogeneous sediment could form through the settling out of low-density turbid-layer flows, perhaps caused by slumping on the continental slope. Slumping followed by debris-flows have resulted in the debris-flow deposits with oval, circular and angular mud clasts found in core 32. The slumping is probably caused by two principal mechanisms. Large amounts of the fine suspension load from the Zaire River are deposited on the continental shelf and slope on an uneven sub-bottom (Eisma & Kalf 1983). This uneven and irregular sea bottom has been induced partly by the intermittent updoming of salt diapirs. Sea floor instability is enhanced by both the high sedimentation rates and the irregularity of the sea floor, causing slumping followed by debris-flows across the slope and progressively more dilute suspension clouds further into the basin. Normal hemipelagic settling would occur at the same time as settling from these dilute suspensions, producing the homogeneous fine-grained sediments encountered between the turbidite intervals in core 38 and overlying the turbidites in core 42. Low-density suspension flows in the bottom boundary layer might also result in the deposition of the homogeneous muds. However, optical
measurements during one of our cruises (Zaneveld et al. 1979) gave no indication of nepheloidlayer sediment transport close to the bottom at that time.
Conclusions Four sediment types, reflecting different depositional mechanisms, can be recognized in cores from the Zaire deep-sea fan. Type 1: fine-grained turbidites, comprising thinly-laminated silts forming the TEl division, graded silts forming the TE2 and ungraded silts and clays forming the TE3 division (after Piper 1978). Type 2: pelagic sediments between turbidite beds particularly on the outer fan. Type 3: homogeneous, uniform sediments, which are thought to have accumulated slowly through settling out of dilute suspensions. These suspensions may have been caused by either slumping or by low density flows in the bottom boundary layer. Evidence of slumping is present in the debris-flow deposits found in a core at the base of the slope; bypassing of fines may have resulted in the formation of low-density turbid clouds responsible for the deposition of type 3 sediments. Type 4: debris-flow deposits found at the base of slope and upper fan area as mud clasts in a silty-clayey matrix, deposited from debris-flows derived from slumping on the slope. Difference in thickness between the various turbidite dl Asions over the fan is most likely the result of the interaction of sea floor morphology and the volume and height of the turbid cloud passing over the sea bottom. The ratio of pelagic to turbidite sediments is 1 : 3 on the outer fan and 1:20 on the upper fan. The content of organic carbon in turbidite sediments of the Zaire fan decreases considerably with increasing distance from the source; on the upper fan, organic carbon contents are slightly more than 4%.
References BAUMGARTNER, T.R. & VAN ANDEL, TJ. H. 1971. Diapirs of the continental margin of Angola, Africa. Bull. geol. Soc. Am., 82, 793-802. BENNEKOM,A.J. VAN& BERGER,G.W. 1983. Hydrography of the Angola Basin and modifications caused by the Zaire river. Neth. J. Sea Res., 17, (in press). BLANPIED, C. & STANLEY, D.J. 1981. Uniform mud (unifite) deposition in the Hellenic Trench, eastern Mediterranean. Smithson. Contrib. Marine Sci., 13, 40pp. BORNHOLD,B.D. 1973. Late Quaternary sedimentation
in the eastern Angola Basin. Techn. Rep. Woods Hole Oceanogr. Instit, 73--8. BOUMA, A.H. 1962. Sedimentology of Some Flysch Deposits. Elsevier, Amsterdam. 168 pp. EISMA, D. & KALr, J. 1983. Dispersal of Zaire river suspended matter in the estuary and the Angola Basin. Neth. J. Sea Res., 17, (in press). - - , DE BLOK, J.W., DORRESTEIN,R., POSTMA,H., NIENHUIS, P.H. & WEBER, R.E. (eds.) 1978. Geochemical Investigations in the Zaire River, Estuary and Plume. Neth. J. Sea Res., 12(3/4), 255-420.
Fine-grained sediments of the Zaire deep-sea fan
-
EMERY, K.O. 1972. Eastern Atlantic Continental Margin: Some results of the 1972 cruise of the R.V. Atlantis II. Science, 178, 298-301. -& HONJO, S. 1979. Surface suspended matter off Western Africa: relation of organic matter, skeletal debris and detrital minerals. Sedimentology, 26, 26-775. FOLK, R.L. 1968. Petrology of Sedimentary Rocks. Univ. of Texas. 170 pp. GIRESSE, P. & CORNEN, G. 1976. Distribution, nature et origine des phosphates miocen+s et +oc6nes sousmarins des plate-formes du Congo et du Gabon. Bull. BRGM. (2me-Serie), Section IV-no 1, 5-15. - - & MOGUEDET,G. 1974. La mati6re organique dans les s6diments du plateau congolais, facteurs de distribution, cons6quences sur les authigen6ses min6rales. Annls Univ. Brazzaville., T 10 (C), 15-29. & ODIN, G.S. 1973. Nature min6ralogique et origine des glauconies du plateau continental du Gabon et du Congo. Sedimentology, 20, 457-188. , KOUYOUMONTZAKIS,G. & MOGUEDET, G. 1979. Le quaternaire superieur du plateau continental congolais-exemple d'evolution paleooceanographique d'une plate-forme depuis environ 50 000 ans. In: van Zinderen Bakker, E.M. & Goetzee, J.A. (eds), Palaeoecology of Africa. Balkema Publishers, Rotterdam. 193-217. , JANSEN, F., KOUYOMONTZAK1S,G. & MOGUEDET, G. 1982. Les fonds du plateau continental congolais et le delta sous-marin du fleuve Congo. Traveaux et Documents de I'ORSTOM Nr., 138, 13-45. GORSLINE, D.S. & EMERY, K.O. 1959. Turbidity current deposits in San Pedro and Santa Monica Basins off Southern California. Bull geol. Soe. Am., 70, 279-90. HART, T.J. & CURRIE, R.I. 1960. The Benguela Current. Discovery reports., 31, 123-298. HEEZEN, B.C., MENZIES, R.J., SCHNEIDER,E.D., EW1NG, W.M. & GRANELL1,N.C.L. 1964. Congo Submarine Canyon: Bull. Am. Assoc. Petrol. Geol., 48, no. 7, 1126-49. HESSE, R. 1975. Turbidites and non-turbiditic mudstones and Cretaceous flysch sections of the Eastern Alps and other basins. Sedimentology, 22, 387-416. -
I 13
JANSEN, J.H.F., VAN WEERING, TJ.C.E., GIELES, R. & VAN IPEREN, J. 1983. Middle and Late Quaternary paleoceanography and climatology of the Zaire (Congo) fan and the adjacent eastern Angola Basin. Neth. J. Sea Res., 17, (in press). LEHNER, P. & DE RUITER, P.A.C. 1975. Structural history of Atlantic margin of Africa. Bull. Am. Ass. Petrol. Geol., 61(7), 961-81. VAN DER LINGEN, G.J. 1969. The turbidite problem. N.Z. Jl. Geol. Geophys., 12, 7-50. PIPER, D.J.W. 1978. Turbidite muds and silts on deep-sea fans and abyssal plains. In: Stanley, D.J. Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden Hutchinson & Ross, Stroudsburg, Pa. RUPKE, N.A. 1975. Depostion of fine-grained sediments in the abyssal environment of the Alg6ro-Balearic Basin, Western Mediterranean Sea. Sedimentology, 22, 95-109. -t~ STANLEY, D.J. 1974. Distinctive properties of turbiditic and hemipelagic mud layers in the Alg6roBalearic Basin, western Mediterranean Sea. Smithson. Contr. Earth Sci., Washington, 13, 40pp. SHEPARD, F.P. & EMERY, K.O. 1973. Congo Submarine Canyon and Fan Valley. Bull. Am. Ass. Petrol. Geol., 57 1679-91. SIESSER, W.G. 1980. Late Miocene Origin of the Benguela Upwelling System off Northern Namibia. Science, 208, 283-85. STANLEY, D.G. 1981. Unifites: structureless muds of gravity flow origin in Mediterranean Basins. GeoMarine Letters, (1), 77-83. STOW, D.A.V. & SHANMUGAM, G. 1980. Sequence of structures in fine-grained turbidites: comparison of recent deep-sea and ancient flysch sediments. Sed. Geol., 25, 23-42. WOST, G. 1935. Schichtung und Zirkulation des Atlantischen Ozeans II. Die Stratosph~ire. Wiss. Ergebn. D.A.E. Meteor VI, (1) 288 pp. ZANEVELD, J.R.V., SPINRAD, R.W. & MENZIES, D.W. 1979. Optical and hydrographical observations in the Congo river and the Angola basin during May 1978. Oregon School of Oceanography Data Report, Ref. 79-3. 202 pp.
TJ.C.E. VANWEERING• J. VANIPEREN,Netherlands Institute for Sea Research, PO Box 59, Texel, The Netherlands.
Sedimentary sequences on the north-west Mediterranean margin during the Late Quaternary: a dynamic interpretation A. Monaco and Y. Mear SUMMARY: The methods we have used in this study of pelitic sediments are based on the concept of fine-grained particle populations, and include: the counting of grains in the > 40 ~m fraction (lithoclastic, biogenic, authigenic and aggregates); grain-size analysis of the pelitic fraction expressed as a bi-logarithmic frequency curve giving the 'evolution index' of three granulometric classes--silt, clayey-silt and clay; mineralogical indices based on the ratios between particles with different hydrodynamic behaviour--quartz, mica-illite and smectite; and geochemistry of Mn, Cu and Fe to characterize the biogenic nature of the sediment. A succession of turbiditic (heterogeneous population), hemipelagic and pelagic (homogeneous population) sediments were examined from the NW Mediterranean margin. We identify three main sequences in the Pleistocene to Recent succession, each of which may be interrupted by episodic events: (1) a palaeoclimatic sequence ( ,-- 5 m thick) of Pleistocene to Holocene age, marking the end of deep channelized lobe construction on the Rhone deep-sea fan; (2) a dynamic sequence (repeated intervals of 20-50 cm thick) of Wfirmian and Holocene age respectively, in the lower fan and in the central basin: and (3) a uniform sequence (without granulometric, mineralogical or geochemical grading) corresponding to a ponded unit filled with reworked sediments (contourites) and slumps. Using several kinds of analysis, many authors have attempted to recognize the various divisions of sedimentary sequences and to establish a nomenclature for them (Rupke & Stanley 1974; Piper 1978; Stow & Shanmugan 1980; Stanley & Maldonado 1981). Their recognition in varied environments has often led to the association of these sequences with their fossil counterparts (e.g. Bouma 1972; Mutti & Ricci Lucchi 1978). The time-scale of such sedimentary phenomena remains poorly understood. In the present study, analysis of the coarsefraction composition, sediment grain-size, mineralogy of the < 2 #m fraction and the geochemistry (Fe, Mn, Cu) of the < 40 #m fraction of pelitic sediments has allowed the definition of various pelagic, hemipelagic and turbiditic facies. This method is applied to deposits from different physiographical domains (slope, deep-sea fan and basin) of the NW margin of the Mediterranean Sea. Based on these results and on morphological, structural, palaeographic and ecological considerations, we reconstruct the sedimentation mechanisms that have operated during the last Pleistocene-Holocene glacial-interglacial cycle. A standard sequence is recognized and its palaeoclimatic and dynamic significance are discussed.
Methods The methods used are specially adapted to the
fine-grained nature of the pelitic deposits studied. The principle of these methods, both granulometric and mineralogical, is based on the idea of populations of particles and their respective behaviour. Composition of the > 40/~m fraction
The overall proportion of this fraction in the sediments studied is generally low, but its composition is often revealing. This includes heavy minerals, quartz, feldspar, mica, calcite, microfauna and macrofauna (Fig. 1). Particular attention is given to two kinds of component that are indicative of different facies: clastic or biogenic aggregates (faecal pellets) and the products of diagenesis. They have been studied using a scanning electron microscope to reveal their fine structure and a microprobe for their chemical composition. Granulometry of the < 40 pm fraction
The grain-size distribution of a sediment plotted as a semilogarithmic graph or a 'canonic graph' (Rivi~re 1952) allows an estimation of the degree of 'evolution' or sorting of the sediment. The method using the slope of a bi-logarithmic frequency graph (Rivi~re 1960) has the advantage of permitting the calculation an 'evolution index' for a given granulometric class (Desprairies 1974). A facies can be defined from: the relative behaviour of three size fractions, silt (10-40/~m), clayey-silt II 5
I 16
A. Monaco
and
Y. Mear
FIG. 1. Vertical profiles of the pelitic facies for different grain-size fractions. Left, > 40/~m, relative proportions of elastic, biogenic and authigenic grains; Centre, < 2/~rn, variation in smectite/illite (rl) and illite/quartz (r2) ratios; Right, > 40 #m, variation in Mn, Cu and Fe.
(1-10 #m) and 'clay' ( < 1 pm); and the degree of sorting or 'evolution index' of each fraction in ascending order, unsorted facies (n=0), parabolic facies ( - l < n < 0 ) , logarithmic facies (n = - 1), and hyperbolic facies (n < - 1). In Fig. 2 we show an example of three typical facies. The fine-grained turbidite is a heterogeneous deposit composed of three differentlyevolved fractions. The silt fraction has a hyperbolic facies (n=2.8), the clayey-silt has a parabolic facies (n = - 0 . 5 ) characteristic of sediments deposited by excess loading, and the clay fraction is apparently unsorted, or has undergone subsequent disaggregation. The hemipelagite and pelagite both show hyperbolic facies for the silt fraction (n = - 2 . 4 or a little less), whereas the clayey-silt is logarithmic (n = - 1) for the hemipelagites, and hyperbolic (n = - 1.2 to - 1.4) for the pelagites. The clay fraction is parabolic ( n = - 0 . 7 ) and hyperbolic ( n = - 1.4) respectively, showing improved sorting for the hemipelagites, often coinciding with an enrichment in the finest clay (smectite). In fact, the graphs of bi-logarithmic frequency reflect the population dynamics of grains with different properties (shape, form, density) and the greater or lesser homogenization of terrigeneous
material. The degree of evolution or sorting of particles is a function of the energy and duration of transport. Horizontal or vertical variations in these indices allow us to follow the evolution of a sediment type of process in terms of its continuity or discontinuity.
Mineralogical criteria The method of mineralogical indices uses the ratios between characteristic peaks obtained by X-ray diffraction on oriented aggregates from the < 2/~m fraction. We have used minerals which have different hydrodynamic behaviours due to their different shape or size: smectite (Sin), illite (I), and quartz (Q) (Monaco 1981; Monaco et al. 1982). We can thus define rl = S m / I =
smectite 17 A (001) illite 10 A (001)
illite 5 A (002) r2 = I / Q = quartz 4.2 A (100) When these ratios change in parallel through a sedimentary section they indicate a homogeneity of the clay fraction, the absolute value being greater when the selection of clayey particles by
Sedimentary sequences on the north-west Mediterranean margin -I
Parabolic facies
Turbidite
n:--I
Logarithmic facies
Hemipelogite .......
n<-I
Hyperbolic facies
Pelagite
n=O
Unsorted facies
II7
n=Oegree of evolution
\
\ - I .4 ~176
~
~ ....
100%
, ,,- .).5 / \
",, ~
0
-~_x_~
""-
/
\-
~
9~
/
.'"
../
/
,,,
.~5,~
/-
,"
/ . . . . . ,...
~
-50%
.x.."" ""'.-I
/
/ ..
~
..~ ~
.
, , ~ 1 7 6 1~176176 7~6 1 7 6 o ,~176176176
.
.
-
.
.
.
.
.
"
'~176
2.8
I 0.I
I I
\\
I '-2"4 I0
/_Lm
FIG. 2. Cumulative frequency curves for three sediment types: turbidite, hemipelagite and pelagite. Superimposed on these are lines representing the 'gradients' of the frequency curves (after Rivibre 1952, 1960) which define different 'granulometric facies' (parabolic, logarithmic, hyperbolic, unsorted) on the basis of the value of the 'evolution index', n. The evolution index is calculated for different grain-size fractions ( < 1 pm, 1-10 #m and 10-40 #m). transport is greater. In Fig. 1, we pass upwards from a fine detrital facies (D, quartz-rich) to a hemipelagite facies (HP) and then a pelagite facies (P, smectite-rich). If the ratios are antagonistic, this indicates a heterogeneity of the deposit; for example, an increase in fine quartz (r2 decreases) may be concomitant with an increase in smectite (r~ increases). This heterogeneity in composition of the < 2 pm fraction also indicates a textural heterogeneity, as noted in the granulometric facies of turbiditic pelites (Fig. 2). If we take the relative abundance of the detrital (Q) and fine clayey (Sm) end members as reference points, we would have proximal turbidites (TP) when r2 is very low (Q abundant) and distal turbidites (TD) when rl is relatively high (Sm abundant). Geochemical criteria
We measure iron, manganese, and copper by atomic absorption spectrometry after treatment of the < 40 #m fraction with hydroxylamine-acetic acid (Chester & Hughes 1967). This selective method allows the extraction of elements fixed by the particles in the marine medium during transport, or incorporated into the precipitates follow-
ing diagenetic reactions. In all cases their relative abundance is related to the conditions of sedimentation. From many examples the following relation has most often been noted (Fig. 1): Mn and Cu are associated with deposits having a low sedimentation rate of pelagite type and concentrated at the water-sediment interface. They can therefore be considered as indicative of biogenic sedimentation; Fe is associated with conditions of high terrigenous input ( + o r g a n i c debris) and high sedimentation rates and is therefore indicative ofturbiditic or possibly hemipelagic sedimentation. General use of the methods
Using the granulometric, mineralogical and geochemical criteria outlined above, we have established two main categories of deposits (Monaco et al. 1982), based on the comparative behaviour of populations of fine particles, which occupy different granulometric classes due to their size, density and shape (Fig. 3). Quartz is dominant in the 40-10 pm class, mica-illite in the 10-1 pm class, and smectite in the < 1 pm class. Deposits composed of homogeneous popula-
I I8
A. Monaco and Y. Mear
%50 "
p
Homogeneous population
HP
I
I0
40
p.m %50 -
TP Heterogeneous population
0t
I I
I I0
40
/_zm FIG. 3. Smoothed grain-size frequency curves for homogeneous and heterogeneous facies; HP= hemipelagite, P = pelagite, TP = proximal turbidite, TD = distal turbidite. Smectite (Sm) dominates the < 1 #m fraction, illite (I) the 1-10/zm fraction, and quartz (Q) the 10-40 ~tm fraction.
tions are characterized by a parallel increase in the ratios r~ and r2 with increasing distance from the terrigeneous source. Long transport leads to considerable sorting of the clay particles and an enrichment in the biogenic-indicator elements, Mn and Cu, which also depend on the residence time of particles in the water. Sediments of this kind correspond to hemipelagie (n < - I ' , r1[• r2=4-6) and pelagic deposits (n< - 1; rL[• 10]=4-6; r2=6-10) which are differentiated on the basis of sorting and mineralogical indices. The Fe content increases with the sedimentation rate, and it is therefore generally higher in the fine detrital or hemipelagic sediments than in pelagites. Deposits consisting of heterogeneous populations are characterized by an antagonistic variation of the mineralogical indices, which is due to the presence of grain populations of different kinds and with different physical properties (cf. bi-logarithmic frequency graphs). The biogenicindicator elements, Mn and Cu, are relatively
low, whereas Fe tends to be more abundant. These deposits correspond to episodic sedimentation of a turbiditic kind. We have applied these methods to an analysis of Late Quaternary sediments collected in Kullenberg and box-corers from the N W Mediterranean margin, in an attempt to establish a standard sequence based on more than just structural criteria, to define the origin of this sequence, and to understand the discontinuity or continuity of the phenomena in their spatio-temporal order. Our interpretations are also based on certain other data including radiometric (14C), biostratigraphic (Groupe P R O F A N S ) and geotechnical (Chassefi6re et al. 1982) measurements.
Application to north-west Mediterranean margin Several earlier studies have been carried out in the study area (e.g. Maldonado & Stanley 1981),
Sedimentary sequences on the north-west Mediterranean margin some as part of the PROFANS (Project Fans) programme; in particular a bathymetric survey (sea-beam) and high resolution seismic reflection profiling (air gun, sparker~ 3,5 kHz) (Aloisi et al. 1981; Bellaiche et al. 1981; Bellaiche et al. 1982) have been useful for our own interpretations. We have also taken into account the structural setting and seismic facies. In the Rhone delta system we may consider the continental margin as comprising a series of depositional units with comparable lobe-like geometries that were constructed largely by glacial outwash during the last Pleistocene-Holocene glacial episode. These include deep-sea fan lobes. prograding lobes from the continental slope and pro-delta lobes (Fig. 4). According to Monaco et al. (1982) there is a clear relationship between the regressive phase and the construction of deep lobes on the Rhone fan, and a correspondence between the fan and the former Rhone channel, which occupied the axis of the so-called PetitRhone canyon up to the Wfirmian period. The
I I9
main locus of sedimentation shifted to the continental shelf during the Holocene and pro-delta lobes were constructed (Aloisi & Monaco 1980). There is a similarity in the volume of these lobes and the fan lobes, each comprising about 800 x 109T of sediment. In cores (1 6 m tong) from each of these depositional units or lobes, grey silt-laminated muds of Wiirmian age are overlain by more or less uniform beige-coloured Holocene muds. We describe the sediments in more detail below from the three main physiographic areas: the continental slope, the rise and the basin plain. The Rhone fan is a large elongate sediment accumulation that extends from the slope south of the Rhone delta across the rise and is prograding out onto the northern part of the basic plain.
Continental slope The upper part of the continental slope is and was a zone of sediment accumulation, particularly
FIG. 4. Map of the NW Mediterranean margin showing areas studied, and sedimentary lobe features.
I20
A. Monaco and Y. Mear
FIG. 5. Variations of the general sedimentary type-sequence in different physiographic settings (cf. Fig. 4). The two vertical profile curves, rl and r2, represent smectite/illite and illite/quartz ratios respectively in the < 2 #m fraction.
Sedimentary
s e q u e n c e s on the n o r t h - w e s t M e d i t e r r a n e a n
during periods of lowered sea-level when the margin was actively prograding. Below a depth of about 1000 m the various more recent depositional lobes pinch out and older strata outcrop. At about this point, hemipelagic Wiirmian grey muds, with a large, fine, detrital content and no clear structures, are overlain by a relatively thin layer (50-100 cm) of beige-coloured hemipelagic (fine homogeneous population) mud (Fig. 5). The transition from the grey to beige muds may be rather more abrupt on the lower slope. The superposition of water-saturated, plastic Holocene mud on cohesive Wiirmian sediments may give rise to a considerable bottom instability. We believe that this may in some way influence the bottom echo-character which is one of superficial hyperbolic diffractions interpreted by Damuth (1982) as indicative of the action of contour currents. It is only in the axes of the canyons feeding the fan that we observe sand and silt-graded turbidites.
Continental rise The main sediment accumulation that forms the Rhone fan (Menard et al. 1965) built up at the foot of the Rhone slope (1700-1800 m depth) at the outlet of the Petit-Rhone canyon is nearly 500 m in thickness. It is composed of six or seven superimposed lobes, each containing a main channel flanked by asymmetric sedimentary levees (Aloisi et al. 1981). The channel of the most recent lobe is further cut by a meandering thalweg 40-50 m deep and 500-600 m wide (Bellaiche et al. 1981). We have previously made a distinction, based on the litho-seismic facies, between these lobes and those o f a ponded unit (Monaco et al. 1982). The former are lenticular in shape, convex towards the upper surface filling the depressions between the lobate bodies, have an irregular seismic reflection at the surface and are more transparent at deeper levels. They are commonly crossed by a network of small channels. On the upper fan the pelitic facies linked to the late Wtirmian progradation are dominant. These are apparently homogeneous grey muds intercalated with centrimetre to millimetre thick silty to sandy laminae. They are turbiditic in the channels and hemipelagic in the more distal parts. On the levees of the upper fan the hemipelagites are dominant, these muds thus resembling deposits of the prograding slope. The levees of the middle fan corresponding to the oldest channels are of more complex structure with an alternation of turbiditic muddy and silty laminae. The frequency of the sands and silt laminae increases in the distal part
margin
~2I
of the fan in parallel with the reduction in thickness of the pelagic Holocene deposits (Nelson & Kulm 1973; Nelson & Wright 1980). The beige Holocene muds are relatively thin throughout the fan system, and accumulated at rates of between 3 and 30 mm/1000 yrs. In the channel without a thalweg, the facies are mostly pelagic but with some fine silty-sandy laminae showing evidence of non-deposition and erosion. The levees contain distal turbidites. The present blockage of the upper fan channel by pelagic deposits is shown by the presence of elements of anthropogenic origin (13VCs, Pb, Cd) (Cauwet & Monaco 1982). Recent research on the structure and chemical properties of suspended matter off the Rhone estuary (Aloisi et al. 1979) suggests that these sediments are deposited through a benthic nepheloid-layer formed at the river mouth. In the ponded units we distinguish two facies in the Holocene series: structureless sands, apparently associated with the most recently-eroded channel of the Rhone fan, and pelites with alternately turbiditic and pelagic laminae. No mineralogical or geochemical gradient is observed (Fig. 5). This heterogeneity is confirmed by the presence of mineral and organic products derived from shallow water (glauconite, epibathyal fauna). Sedimentation rates exceed 10 cm/1000 yrs.
Basin plain Where the Rhone fan grades into the basin plain, there is a zone of diapirs (Biju-Duval et al. 1974) overlain by an acoustically transparent sequence. These deposits often have distinct thin grey and ochre laminae, which give a varved appearance to the sediment and may represent periodic variation in the deep circulation. Towards the central part of the west Mediterranean basin beyond the deep-sea fans of the Rhone and the Gulf of Valencia, there is, paradoxically, a thickening of the Holocene series. This comprises five discrete sequences (Fig. 5), each 20-60 cm thick and made up of fine-grained, homogeneous pelites, that show colour-grading from grey at the base to maroon at the top. They also show gradation from a low fine-grained detrital content mixed with numerous clastic aggregates at the base, to pelagic division marked by an enrichment in the biogenic-indicator elements (Mn, Cu). Some of the sequences appear to be truncated. These five sequences are interpreted as finegrained turbidites that, according to C ~4 dating, were deposited during about the last 7000 years. The detrital, clayey and biogenic particles were
I22
A. Monaco and Y. Mear
picked-up from the slope and fan and redistributed by turbidity currents of various sizes and densities. The fineness of the original material and its transport to the basin has allowed hydrodynamic sorting within the current and the deposition of positive, finely-graded sequences. These deposits are similar to the faintly-laminated to homogeneous muds described by Stanley (1981) using the term 'unifites', from the western Mediterranean basin.
Sedimentary sequences Palaeoclimatic sequence Using these varied observations we have established a standard sequence of Late Quaternary sedimentation in the area, based on vertical changes in the various parameters and indices, i.e. the relative proportion of populations of grains which make up the <40 /~m fraction of the pelites; the evolution or sorting index of the different granulometric fractions; the comparative changes in mineral populations in the < 2/~m fraction, given by the ratios Sm/I and I/Q (rl and r2); and the content of biogenic-indicator elements Mn and Cu. From the base of most cores to the sediment surface we note (Fig. 1): (I) a decrease in the coarse lithoclastic populations (heavy minerals, quartz, feldspar, mica, silty or pelitic aggregates), more or less regularly replaced by biogenic populations and authigenic products (oxides of Fe and Mn); (2) an increase in the evolution (sorting) index of the silt and the clay fractions, which represents improved sorting of the different grain-size fractions; (3) a more or less continuous and parallel increase in the ratios Sm/I and I/Q in the < 2/~m fraction, i.e. the clayey populations sensustricto (smectite and illite) to the detriment of the fine quartz; (4) an increase in the hydroxylamine-acetic acid extractable Mn and Cu content, with a maximum of 2-3000 ppm and 30-40 ppm respectively at the water-sediment interface or immediately below: and (5) a systematic variation in geotechnical parameters, which show an increase in the water content and plasticity index, and a decrease in density and cohesion of the sediment. This sequence is related to the climatic changes that have taken place over the past 20 000 years (Wiirm-Holocene) and, on the deep-sea fan, is approximately 5 m thick. It represents the progressive reduction of terrigeneous input and its progressive sorting to finer and finer components. At the top of the sequence there is an increase in the rate of exchange processes at the sediment-
water interface, and the appearance ofauthigenic minerals (Fe and Mn oxides and hydroxides). It is assumed that this sequence is incomplete, in that it represents only the uppermost part of the fan lobe which is about 100 m thick on the upper and middle fan. We would anticipate coarse sands at the base of the lobe and consider these to form the lower part of our sequence. The sequence is also discontinuous in that it is interrupted by episodic turbiditic events, represented by centrimetre to millimetre thick laminae of silt and sand and distal mud turbidites. These turbidites are found throughout the Wiirmian, which testifies to the periodic functioning of the channels, the action of deep currents and, perhaps, of variable tectonic activity.
Dynamic sequences (turbidites) These consist of repeated turbiditic sequences similar to those described by Bouma (1972), Rupke & Stanley (1974), Mutti (1977), Piper (1978) and Stow & Shanmugan (1980). They are normally considered to result from a single sedimentary event and are based largely on structural and textural criteria. In the study area, such turbidites are common in the Wtirmian on the lower fan where the channels shallow and bifurcate. They are also found in the Holocene in the centre of the basin, where they are up to 60 cm thick. In detail, these sequences also appear to be discontinuous, interrupted by instantaneous sedimentary rhythms. They nevertheless have a discrete duration, and probably arise from a local geographical or hydrological change in the environment.
Ungraded sequences (contourites) On the rise and fan in the ponded units, there are ungraded beds with uniform mineralogical and geochemical composition throughout. They are, however, composed of alternating turbiditic and pelagic horizons. They also occur in certain other parts of the fan on probable abandoned sedimentary levees. In the most distal part of the convex prograding fan lobe, the pelitic and apparently homogeneous deposits, are colour-banded and appear to be varved. These ungraded deposits are mainly Holocene in age. We also note a reduced thickness of Holocene sediments on the lower slope, a hyperbolic bottom echo-character, and a whirlpool-like circulation pattern in the area with periodic current velocities up to 48 cm/s. We therefore interpret the ungraded sequences and varved structures as contourites. The good sorting of particle populations and
Sedimentary sequences on the north-west Mediterranean margin
I2 3
s
tl) r
J
._
.o
8 123
c___
0
I::I
0
l(", "!i .:~."
0
,.6
O
J
..Q 0
m
-~ I
o
o t"~
I
a_ w J " ~.~. II
II
tl
n ' w ~ m
0
II
'
12 4
A. Monaco and Y. Mear
the layering of the sediment are the result of synsedimentary reworking of deposits of varied nature by deep currents. These deposits accumulated as ponded units in the western part of the Pleistocene deep-sea fan.
Conclusions Our dynamic interpretation of these deep-water fine-grained facies is carried out on the basis of the relative degree of sorting of particle populations present in the pelites. Ultimately, it is the duration of transport which determines the mineralogical and geochemical characteristics of a sediment. Furthermore, our facies are not based on the relative content of biogenic and terrigeneous material, but are based solely on the composition of the terrigenous material and on authigenic mineral components. Thus, our so-called pelagic facies represent the most evolved part of a sedimentary sequence, that is a homogeneous deposit of fine clayey particles that has undergone a long transport in suspension. However, this does not imply vertical settling through the water column, as the presence of numerous planktonic organisms in the most pelagic sediments indicates. On the contrary, in most of the cases studied, there is an advective flow of particles within a benthic nepheloid-layer influenced by the general circulation pattern and by various morphological features. Our method also allows us to follow hydrodynamic variations both vertically (time) and horizontally (space). It reveals the importance of deep-water phenomena (currents, geochemical sediment-water exchanges) in sediments that had previously been considered homogeneous because of their fine-grained nature. Thus, the frequency of heterogeneous facies in the Holocene series, due to sedimentary resuspension through the action of contour or turbidity currents, is linked to the re-establishment of the general Mediterranean circulation during the last 10 000 yrs. Thus, the general form of mineralogical and geochemical gradients provides information on the condensed or expanded nature of
sedimentary sequences, on the presence of discontinuities and, finally, on the processes of sedimentation at any one site. A model has been established which takes into account the supply of material and the construction of depositional units or lobes in the context of deltaic systems (Fig. 6). The construction of pro-delta and fan lobes is linked to supply via channels. The general sedimentary sequence, which in all cases represents the search for equilibrium, here corresponds to the progradation of the middle and upper fan due to a massive sediment input. At present, much of the supply is trapped on the pro-delta lobes, and input to the fan system is limited to benthic nepheloid-flow over the edge of the shelf and to a certain amount of sand spillover. Sediment supply to the rest of the margin is via resuspension on the continental slope and rise and the transport of this material basinwards in turbidity currents forming dynamic sequences. The recent ponded unit corresponds to the reestablishment of the equilibrium profile, material being derived from the slope by slumping, contour currents and diffuse channelling and depositing ungraded sequences.
ACKNOWLEDGEMENTS: This work was supported by Centre National de la R6cherche Scientifique (ERA n ~ 962, RCP n ~ 512) and Comit6 d'Etudes Petroli6res Marines (CEPM) for the Project Fans (PROFANS). We thank A. Maldonado (Consejo Superior de Investigaciones Cientificas, Barcelona) and L. Mirabile (Istituto Universitario Navale, Napoli) for their contributions. We are deeply indebted to Y. Thommeret (radiometric data), Y. Le Calvez and L. Blanc-Vernet (biostratigraphy). J-C. Faug+res and T.C.E. Van Weering have improved the paper through their constructive review comments. We gratefully acknowledge D.A.V. Stow for improving the translation of the manuscript. Morphological and microchemical analysis were made with an SEM Montpellier (JEOL) and Orleans (Cambridge) and an Electron Microprobe (CAMEBA) in the Universit6 des Sciences et Techniques du Languedoc (Montpellier).
References ALOISI,J.C., MILLOT,C., MONACO,A. & PAUC,H. 1979. Dynamique des suspensions et mbcanismes sbdimentologiques sur le plateau continental du Golfe du Lion. C.R. Acad. Sei. Paris, 289, 879-82. , BELLAICHE,G., BOUYE,C., DROZ, L., GOT, H., MALDONADO,A., MIRABILE,L. dr,MONACO,A. 1981. L'Sventail sour-marin profond du Rh6ne et les d6p6ts de pente de l'Ebre: essai de comparaison
morphologique et structurale. In: Wezel, F.C. (ed.), Sedimentary basins of Mediterranean Margins.
--
Intern. Conf. Urbino, 1980, 227-38. & MONACO,A. 1980. Etude des structures s6dimentaires dans les milieux delta'/ques (Rh6ne). Apport /l la connaissance des conditions de s~dimentation et de diagen6se. C.R. Acad. Sci. Paris, 290, 159-62.
Sedimentary sequences on the north-west Mediterranean margin BELLAICHE, G., DROZ, L., ALO1Si, J.C., BOUYE, CH., GOT, H., MONACO, A., MALDONADO, A., SERRARAVENTOS,J. & MmABILEL. 1981. The Ebro and the Rhone deep-sea fan, a first comparative study. Marine Geol., 43, 75-85. - - , ALOISI,J.C., BERTHON,J.L., BOUYE, C., COUTELLIER, V., DROZ, L., GOT, H., MEAR, Y., MONACO, A., ORSOLINI,P. • RAVENNE,C. 1982. Morphologie au sea-beam de l'eventail skous-marin profond du Rh6ne (deep-sea fan). XXVIIIe Congr. Ass. Plen., Cannes, Dbc. 82. BIJU-DUVAL, B., LETOUZEY,J., MONTADERT,L., COURRIER, P., MUGNIOT,J.P. & SANCHO,J. 1974. Geology of the Mediterranean sea basins. In: Burke, C.A. & Drake C.L. (ed.), The Geology o f Continental Margins. Springer, New York. 685-721. BOUMA, A.H. 1972. Recent and ancient turbidites and contourites. Trans. Gulf Coast Ass. Geol. Soc., 22, 205-21. CAUWET, G. & MONACO, A. 1982. Pollution des zones profondes de M6diterran~e occidentale part les m&aux lourds. R61e de la dynamique s6dimentaire. XXVIIIe Congr. Ass. Plen., Cannes, Ddc. 82. CHASSEFIERE,B., ALOISI,J.C., MONACO, A., LE TIRANT, P. & MONTARGESA. 1982. Propri&6s m6caniques et stabilit6 de la couverture s6dimentaire du talus continental au large du delta du Rh6ne. XXVIIIe Congr. Ass. Plen., Cannes, Dbc. 82. CHESTER, R. & HUGHES, M.J. 1967. A chemical technique for the separation of ferro-manganese minerals, carbonate minerals and adsorbed trace elements from pelagic sediments. Chem. Geol., 3, 189-212. DAMUTH, J.E, 1982. Quaternary sediments and sedimentation processes on Atlantic continental margins as revealed by studies of sediment cores and high frequency echograms in the Western equatorial Atlantic. Bull. Inst. G~ol. Bassin d'Aquitaine, Bordeaux, 31, 127-130. DESPRAmmS, A. 1974. Degr6 de repr6sentativit6 de la distribution des groupes de grains, darts le rythme. Bull. Soc. gbol. Fr., 7, XVI, 624-33. GOT, H., ALOIS1, J.C., LEENHARDT, O., MONACO, A., SERRA-RAVENTOS,J. & THEILEN, F. 1979. Structures s~dimentaires sur les marges du Golfe du Lion et de la Catalogne. Rev. G~ol. Dynam. Gdogr. Phys., 21, 281-93. GROUPE PROFANS (ALOISI, J.C., BELLAICHE, G., BOUVE, C., DROZ, L., FERNANDEZ, J.M., GOT, H., MALDONADO,A., MEAR, Y., MIRABLE,L., MONACO, A., BLANC-VERNET, L., LE CLAVEZ, Y. & THOMMERETY.) 1982. S6dimentation r6cente et actuelle de l'~ventail sousmarin profond du Rh6ne. XXVIIIe Congr. Ass. Plen., Cannes, Dbc. 82. MALDONADO, A. & STANLEY, D.J. 1979. Depositional patterns and Late Quaternary evolution of two Mediterranean submarine fans: a comparison. Marine Geol., 31, 215-50. MENARD, H.W., SMITH,S.M. & PRATT, R.M. 1965. The Rh6ne deep sea fan. In: Whitard & Bradshaw, (eds,) Submarine Geology and Geophysics. Cols on Papers, 17, 271-85. MONACO, A. 1971. Contribution ~ l'6tude g6ologique et
I2 5
s6dimentologique du plateau continental du Roussillon (Golfe du Lion). ThOse Univ. Sc. et Tech. Languedoc, 295 pp. 1981. Argiles et m+canismes s6dimentog6n&iques en M6diterran6e. In: Wezel, F.C, Sedimentary Basins of Mediterranean Margins. Intern. Conf. Urbino, 1980. 299-312. 1982. Crit6res min&alogiques et g6ochimiques pour la reconnaissance des turbidit6s fines. C.R. Acad. Sc. Paris, 295, 43-6. 9ALOISI, J.C., BOUYE, CH., GOT, H., MEAR, Y., BELLA1CHE,G., DROZ, L., MIRABILE,L., MATTIELO, L., MALDONADO, A., LE CALVEZ, Y., CHASSEFIERE, B. & NELSON, H. 1982. Essai de reconstitution des m6canismes d'alimentation des Oventails sOdimentaires profonds de l'Ebre et du Rh6ne. Bull Inst. Gbol. Bassin d'Aquitaine, Bordeaux, 31, 99-110. MUTTJ, E. 1977. Distinctive thin-bedded turbidite facies and related depositional environments in the Eocene Hecho group. Sedimentology, 24, 107-31. -& RICO LUccH1, F. 1978. Turbidites of the Northern Apennines: Introduction to facies analysis. Int. Geol. Review, 20, 25-166. NELSON, C.H. & KULM, L.D. 1973. Submarine fans and deep-sea channels: In. Turbidites and Deep water Sedimentation. Pacific Sechan, Soc. econ. Paleo. Min. Short course, Anaheim. 39-78. 8/, WRIGHT, A.A. 1980. Variation in submarine fan sand facies and application to petroleum geology (abs.). Bull. Am. Assoc. Petr. Geol., 64, 756. NORMARK, W.R. 1970. Growth patterns of deep sea fans. Bull. Am. Assoc. Petrol. Geol., 54, 2170--95. 1978. Fan valleys, channels and depositionnal lobes on modern submarine fans: characters for recognition of sandy turbidites environments. Bull. Am. Ass. Petrol. Geol., 62, 912-31. P1PER, D.J.W. 1978. Turbidite muds and silts on deep fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds) Sedimentation in submarine canyons, fans and trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. RIVIERE, A. 1952. Sur la repr6sentation graphique de la granulom6trie des s~diments meubles. Interpr6tation des courbes et applications. Bull. Soc. gbol. Ft., 6, II, 145-67. 1960. G~n&alisation de la m~thode des 'facies granulom&riques' par extension de la notion d'indice d'+volution. C.R. Acad. Sci. Paris, 250, 2917. RUPKE, N.A. & STANLEY,D.J. 1974. Distinctive properties of turbiditic and hemipelagic mud layers in the Algero-Balearic Basin, Western Mediterranean sea. Smiths. Contr. Earth Sci., 13, 1-40. STANLEY, D.J. 1981. Unifites: Structureless mud of gravity-flow origin in Mediterranean basins. GeoMarine Letters, 1, 77-83. - - & MALOONAOO,A. 1981. Depositional models for fine-grained sediment in the western Hellenic Trench, Eastern Mediterranean. Sedimentology, 28, 273-90. STOW, D.A.V. & SRANMUGAM,G. 1980. Sequences of structures in fine-grained turbidites: comparison of recent deep-sea and ancient flysch sediments. Sed. Geol., 25, 23-42. -
-
A. MONACO & Y. MEAR, Laboratoire de S6dimentologie et G6ochimie Marines, Universit6 de Perpignan, France.
Sedimentology of the Halifax Formation, Nova Scotia: Lower Palaeozoic fine-grained turbidites D.A.V. Stow, M. Alam and D.J.W. Piper S U M M A R Y : The Halifax Formation is a several kilometre thick, shale-rich succession that
forms part of the (?) Cambro-Ordovician Meguma Group of Nova Scotia. The Meguma sediments were mostly deposited by sediment gravity flows on a prograding marine slope-rise complex of uncertain water depth. Eight separate facies of the Halifax Formation are recognized on a facies continuum from very shaley to very sandy. The two most sandy facies show partial Bouma sequences; the others are best interpreted in terms of a more detailed 'ideal sequence' for fine-grained turbidites. Vertical sequences show thinning-upward, thickening-upward and irregular patterns of sandstone and siltstone beds on a large-scale (100-200 m). There is also much wedging, subdivision and amalgamation of these coarse lithologies. Erosional features are locally common at the base of turbidite beds and still more widespread at the top. Slumps and bioturbation are less common. The depositional setting of most of the Halifax Formation is interpreted to be the mid- or upper-fan area of a muddy deep-sea fan, passing upwards into a prograding continental slope.
The Lower Palaeozoic Meguma Group of Nova Scotia (Fig. l) comprises a thick succession (at least 10 000 m) of sandstone, siltstone and shale that has been interpreted as a prograding deep water slope-rise complex (Schenk 1978; Schenk et al. 1980). These rocks were folded, intruded and regionally metamorphosed during the Devonian Acadian orogeny (Poole 1970; Williams 1978). The major folds are open, upright and low-plunging, and a marked cleavage parallels their arcuate trend (Fyson 1966). The metamorphism varies regionally from greenschist to amphibolite facies (Taylor 1967, 1969; Clarke& Muecke, 1980). The lower grade metamorphism results in slates becoming indurated so that alternating slate and siltstone lithologies are in places unusually well preserved and exposed, although where cleavage is intense outcrops are poor. Sparse faunal data are all of Early Ordovician age (Schenk et al. 1980), so that the lower part of the Meguma Group may be Cambrian in age. The occurrence of possible tillite suggests that the top of the Halifax Formation may extend into the Late Ordovician (Schenk 1972; Schenk et al. 1980). The Meguma Group comprises two partially coeval formations, the (mainly lower) Goldenville Formation of quartz metawacke and the (mainly upper) Halifax Formation, principally of slate. Halifax Formation lithologies overlie, intertongue with and may be intercalated with Goldenville Formation lithologies (Harris & Schenk 1975; Schenk et al. 1980). The Halifax Formation is overlain, probably conformably, by sandstones, shales and volcanic rocks of the White Rock Formation, which were deposited in a continental shelf environment (Lane 1976, 1981).
Harris & Schenk (1975) and Schenk et al. (1980) have reviewed the results of early mapping by the Geological Survey of Canada, and of previous sedimentological studies. In the last 20 years sedimentological work, chiefly on the Goldenville Formation, has established that the Meguma sediments were mostly deposited by sediment gravity flows (Phinney 1961; Campbell 1966; Taylor 1967; Schenk 1970; Harris 1971). The Goldenville Formation consists of sandy flysch, whereas the Halifax Formation is of shaly flysch. According to Schenk et al. (1980), on the facies classification of Walker & Mutti (1973), sand layers of the Halifax Formation correspond to Facies D ('classical' distal turbidites) and G (pelagic and hemipelagic muds and silts) with minor amounts of C (Bouma's Tae proximal turbidites). Sands of the Goldenville Formation are mainly of Facies B2 (massive sandstones without dish structures) and Facies C with minor A2 (organized conglomerates), A4 (organized pebbly sandstone), and B1 (massive sandstones with dish structures). Goldenville lithologies represent channel deposition in the mid-fan area of submarine fan systems (Schenk et al. 1980). The dispersal pattern appears almost constant throughout the time in which the Meguma Group was deposited (Schenk 1970). The palaeocurrents turn approximately 90 degrees from north in south-west Nova Scotia to east in the eastern part of the province (Fig. 2A) and local variations to this trend are consistently toward the north-west. The easterly palaeocurrent trend appears to parallel sedimentologic scalar properties, so that palaeocurrents trend along, not across, contours of such properties as layer thickness (Fig. 2B & C), sandstone-shale ratio (Fig. 2D) and percent I27
I28
D.A.V. Stow, M. Alam and D.J.W. Piper
Sedimentology of the Halifax Formation, Nova Scotia A
DIRECTIONAL STRUCTURES
B MAXIMUM TURBIDITE THICKNESS
2O
C MAXIMUM SANDSTONE THICKNESS
50
5 20 D SANDSTONE/SHALE RATIO
Fro. 2. Trend surface maps (from Schenk 1970) of Halifax Formation showing regional variation in (A) directional structures (sole marks, cross-stratification) (B) maximum turbidite thickness (cm) (C) maximum sandstone thickness (cm) (D) sandstone/shale ratio. Iithic clasts (Schenk 1970). Thus the general trend of the continental margin on which the Meguma Group accumulated is probably ENE-WSW, with downslope transport in proximal areas and along-slope transport in distal areas. Schenk (1970) originally suggested that the along-slope transport was due to contour currents, but later (Schenk 1971) pointed out the analogy with turbidites of the Hatteras Abyssal Plain, which have flow paths to the south parallel to the North
I29
American continental margin (Horn et al. 1971). The lack of stratigraphic markers seriously hinders regional sedimentologic interpretation of the Hali fax Formation. The upper contact with the White Rock Formation in the Annapolis Valley appears to be synchronous (Schenk et al. 1980; Lane 1981). The lower contact with the Goldenville Formation, which Schenk used as a datum in his early reconnaisance studies (Schenk 1970, 1971) could be highly diachronous (Schenk et al. 1980). In addition, the regional tectonic structure of the Halifax Formation is poorly known. The Halifax Formation of the Atlantic Coast of Nova Scotia is very variable on a small scale, from locally dominant sandstone to dominant shale facies. Packages of thin to thick sandstones (up to 1 m thick) occur in zones 1 to 20 m thick interbedded with siltstone-laminated shales. The sandstones are especially frequent near to the contact with the Goldenville Formation. The sequence in the Annapolis Valley region is generally finer-grained. Smitheringale (1973) identified three facies, from very silty, coarser and more quartz-rich to very shaley, finer and more chlorite-mica rich, and noted that these formed an overall coarsening-upward megasequence within a zone approximately 3000 m below the White Rock Formation. Lane (1976, 1981) described this silty-shale lithology in detail from the upper 1500 m of the Halifax Formation and noted a finer-grained slate with minor siltstone laminae for the 100 m or so immediately below the White Rock Formation. This present work is based on detailed measurement and study of sections at six locations (Fig. 1): Cape St. Mary and Cape View on the Bay of Fundy to the west of the Annapolis Valley, and sections at Rose Point, Ovens Park (Ovens Point, Thunder Cave and Felzen South), Blue Rocks and Hartlen Point on the Atlantic coast of Nova Scotia. In addition, we have re-examined some of the sections described by Lane (1981) and Schenk (Harris & Schenk 1975; Schenk et al. 1980).
Sedimentary facies Petrography The Halifax Formation comprises dominantly shale and slate with a variable proportion of sittstone laminae and, in lesser abundance, thin to thick sandstone beds. The shales are black to light grey or greenish-grey in colour and are commonly silty and laminated. They contain dominantly quartz and feldspar, with accessory minerals including epidote, apatite, tourmaline, zircon,
I30
D.A.V. Stow, M. Alam and D.J.W. Piper
magnetite, ilmenite, graphite and micas. The metamorphosed matrix in lower grade rocks is of chlorite, muscovite, biotite and sericite and where the metamorphic grade is higher garnet, andalusite and cordierite occur. The lighter-coloured shales are more silty and quartz-rich, the darker shales are more carbonaceous and pyritic. Most siltstones are light grey and rich in quartz and feldspar. The sandstones are mostly very finegrained, but rarely are medium-grained and contain shale rip-up clasts. They are mainly quartzo-feldspathic with a variable metamorphic matrix and a range of accessory minerals similar to those of the slates (Schenk et al. 1980). Both fauna and faunal traces are rare. Rare graptolites and trilobites have been found. Trace fossils have been recently described by Pickerill & Keppie (1981) and include invertebrate feeding tracks and vertical, sand-filled burrows.
(~3~ f0 t.O
Sedimentary structures Thicker-bedded sandstones show clear evidence of Bouma sequence turbidites. The thinner-bedded fine-grained facies of shale and siltstone include distinctive sedimentation units which correspond closely to fine-grained turbidite sequences defined by Piper (1978) and Stow & Shanmugam (1980). These units are variable and not always very distinct, but are typically 2- l 0 cm thick and comprise a thin basal siltstone or very fine sandstone layer, commonly cross-laminated or parallel-laminated, overlain by a thicker graded silstone-laminated shale. The cross-lamination in places occurs as 'fading' ripples in which ripple crests of siltstone pass gradually into ripple troughs of shale (Stow & Shanmugam 1980; see also Fig. 13a). The coarser siltstone and sandstone beds display a variety of erosional and synsedimentary deformational structures, including flute and irregular scour marks, shale intraclasts, convolute lamination, and load and injection structures.
Facies types In the measured sections presented in this paper, the slates have been sub-divided into six facies on the basis of sedimentary structures and the average siltstone/shale ratio (SIL/SH). The sandstones have been divided into two facies on the basis of bed thickness. These facies are described below and illustrated in Figs 3 and 4. We have to use such simple criteria for facies definition because weathering and cleavage frequently obsF1G. 3. Photographs illustrating siltstone-shale facies 1-6. For details, see text. 6
Sedimentology of the Halifax Formation, Nova Scotia
Facies
Facies
]
I
I
I
I
I
I
I
[
8 0 cm
FIG. 4. Photograph illustrating sandstone facies 7 (thin-bedded, 1-10 cm) and 8 (thick-bedded > 10 cm). For details see text. Facies cure primary structures, and there has been considerable plastic strain in many outcrops. Facies (1): Shale, SIL/SH < 5/95; structureless, rare ~ siltstone lenses. (Facies 7-9 of Lane 1980, are equivalent.) Facies (2): Shale dominant, SIL/SH < 25/75; infrequent, very thin (<1 mm), indistinct and/or discontinuous, parallel to wavy, siltstone laminae; gradational and sharp boundaries, rare normal grading. Rare thin sandstone beds. (Facies 4 and 6 of Lane 1980, are equivalent.) Facies (3): SIL/SH~40/60; frequent, thin (1-2 mm), regular, parallel siltstone laminae; rare low-amplitude longwavelength ripples; gradational and sharp boundaries, some normal grading. Rare thin sandstone beds. (Facies 5 of Lane 1980, is equivalent.) Facies (4): SIL/SH ~ 50/50; common, thin (2-4 i In the following facies descriptions the abundance of siltstone laminae per 10 cm of section is approximately quantified by rare ( < 1), infrequent (1-3), frequent (3-5), common (5-8) and very common ( > 8).
Facies
131
mm), regular, parallel siltstone laminae; low-amplitude long-wavelength ripples; rare lenticular laminae; gradational and sharp boundaries, some normal grading and loaded or scoured bases. Rare thin sandstone beds. (5): SIL/SH ~ 60/40; very common, thin to thick (3-6 mm), parallel siltstone laminae, lenticular laminae, convolute laminae; low-amplitude longwavelength ripples, fading ripples; gradational and sharp boundaries, normal grading, internal parallel and cross-lamination loading, scouring and injection structures. Thin sandstone beds rare to frequent. or: SIL/SH~60/40; very common, thin to thick (3-6 mm), regular, closelyspaced, parallel siltstone laminae. Thin sandstone beds rare to frequent. (6): Siltstone dominant, SIL/SH ~ 75/25; very common, thick (4-10 mm), parallel siltstone laminae, lenticular laminae, convolute laminae; fading ripples, gradational and sharp boundaries, normal grading, internal parallel and cross-lamination, loading, scouring, injection structures and mudstone rip-ups. Thin to mediumbedded sandstones rare to frequent. (7): Sandstone, thin-bedded (1-10 cm), commonly flat sharp base, also with flutes, scours and load casts, rarely gradational; sharp or gradational top; beds commonly lenticular, internal parallel and cross-lamination, with fading-ripples at top; normal grading common, rarely massive, partial Bouma sequences present (CDE). (8): Sandstone, medium to thick-bedded ( > 10 cm, up to 2 m); commonly flat sharp base, also scoured, loaded or channelled; sharp, gradational or eroded top; internal parallel and cross-lamination, the fading-ripples at top; shale partings, siltier and shalier zones indicate amalgamation common; shale rip-up clasts locally abundant, rare pebbly sandstone; normal grading, indistinct grading and massive, partial Bouma sequences present (AB, AE, BCD).
Facies associations and vertical sequences
The distribution of facies in vertical sections is
I3Z
D.A.V. Stow, M. Alam and D.J.W. Piper
shown in Figs 6 and 7. Facies (1) to (8) are convenient divisions on a facies continuum from very shaley to very sandy. The facies at the shaley end of this spectrum (1-4) tend to be closely associated with one another and, similarly, the more sandy facies (4-6) are closely associated. The thin-bedded (7) and thick-bedded (8) sandstones occur throughout the spectrum. These relationships are indicated on a facies transition probability matrix (Fig. 5), representing nearly 400 separate facies transitions (both sharp and gradational) in the measured sections at Ovens Point and Cape St. Mary. Siltstone-rich facies with frequent sandstone are commonest on the Atlantic coast of Nova Scotia, particularly close to the Goldenville Formation boundary. About 370 m of well-exposed section were measured at Ovens Point (Fig. 6(b)). The sequence is one of rapid alternations of sandy and shaley facies with no large-scale overall pattern evident except for an upward increase in the frequency of thick-bedded sandstones over
23 11 2
10 16
about the first 300 m. Medium-scale (10-20 m) coarsening-upwards and fining-upward sequences are common in the Thunder Cove section (also at Ovens Park), and the Rose Point and Hartlen Point sections. More random oscillations of sandstones and shales are also present. At the small-scale (1-2 m) too, coarsening-upwards, fining-upwards and irregular sequences are evident. These sequences are illustrated in Figs 7 and 8. The measured sections at Cape St. Mary (Fig. 6(a)) and Cape View (Fig. 7(c)) represent the shale-rich facies association of the Halifax Formation. At Cape View, shale facies (1) and (2) alternate randomly through 50 m of sequence (Fig. 7(c)). The 400 m thick section at Cape St. Mary is faulted, and may not be stratigraphically continuous. Its upper contact with the White Rock Formation is tectonic (Lane 1981). Thickbedded sandstones are restricted to the top 25 m. Both coarsening-upwards and finingupwards large-scale (100-200 m) sequences are present. Superimposed on these are medium scale (10-20 m) sequences showing fine-coarse-fine oscillations, and with more frequent, thicker (up to 10 cm) sandstone beds associated with the coarser facies. At the small-scale (1-2 m), no very clear sequences were observed. The shale rich facies association also occurs on the Atlantic coast of Nova Scotia, for example in parts of the Ovens Point (Fig. 6(b)), Rose Point (Fig. 8(b)) and Blue Rocks section (Fig. 8(a)).
13 5
21 18
Horizontal variability
UPPER FACIES 001
U.I
rr
3 4
4 5
0 -J7
6
19
2
8
10
3
4
9
1
1 10 4
7 19 6 5
2
7
11 24 18 2
\
FIG. 5. Transition matrix for facies 1-8 for Ovens Point and Cape St. Mary measured sections, showing (above) actual number of transitions and (below) those transitions that are more common than random (thin line) and more than twice as common than random (thick line).
The intersection of the shoreline with strike direction is such that individual beds of the Halifax Formation can only be traced laterally for 100-200 m at most. Lateral variation has been investigated in most detail in the sections on the Atlantic coast of Nova Scotia. Thin and mediumbedded sandstones and shales are commonly continuous over the length of the section (Fig. 9). The very thin siltstone layers are also laterally extensive and have been traced for over 50 m. The very thick (1-3 m) sandstones tend to wedge out over 50-100 m (Fig. 10) or, in some areas, occur as lenticular bodies over about 10 m of section (Figs 11 and 12). Beds of all thicknesses, however, can also vary considerably in thickness over relatively short distances (5-25 m ) - p i n c h i n g , swelling, subdividing and amalgamating (Figs 10, 11 and 12). Thin slump horizons have been traced over 5 m to 50 m of section.
Erosion and slumping On a local scale there is much evidence for erosion
Sedimentology of the Halifax Formation, Nova Scotia
(a) CAPE ST. MARY continued
(b) OVENS (SOUTH) continued .
4 0 0 - , ..........
1
I33
":'.'..1:.'.'.: '.::::::. ........ .............
. . . . . . . .
[7.'.'.'.'.
i/.-I [
,.%*,: :.':
~"(.y. " <.;.
section covered ......
t~
I I
_.
.......
j
/
- L '-1
.2.'2."
.....
---'I
covered
L
'I I
::".'.'"
..____J
'~"-.'::-
T
section covered ............
!,-;'?i:;;;;
covered,, 300-
100-
L
~'.i'::.:'::'....
J
]
~""" F":;'v""
-'1 ~-.
100
......
~t:~
]
J
-..,
' '! I
J
~--'
300-
:}y~:~!.:::..
9
...... ~I tt %,=*~.ltt 9
"":;.:-: ....
5!.
""'.'.'."
h
....... .......
E
!!!!i....t
.......
"-1 0
,
1 234
E
,
,
56
, 200
,
,
,
,
E ,
,
(1 to 6, m u d s t o n e )
0
FACIES
I
I
I
I
I
I
(7+8, sandstone)
7
[;;;;2 ..... a
FIG. 6. Measured sections of Halifax Formation showing large-scale sequences, (a) Cape St. Mary (b) Ovens Point. at the base of siltstone and sandstone beds, where there are scour and flute marks from a few mm to 20 cm deep (Fig. 13(e), (f)). These are locally very abundant, as in the Thunder Cove section at Ovens Park. In many cases, however, the base of a sand bed is relatively flat whereas the top has been eroded. Thin sandstone beds can be almost completely cut-out by small irregular depressions filled with shale. The tops of thicker sandstones may be irregularly eroded by up to 10 cm and, in one case, by a deep scour 50 cm deep (Fig. 13(c)). Thick, possibly channelized, lenticular sand-
stones are present in one section on Rose Point with abundant shale rip-up clasts on their lower margins (Fig. 13(d)). Commonly, rippled tops of siltstone or sandstone beds are eroded and overlain by shale. Many of these ripples are fading ripples (Stow & Shanmugan 1980), with crests of siltstones passing into troughs of shale. In some cases the eroded top of a ripple can be traced laterally into a perfectly gradational laminated transition from siltstone to shale (Figs. 13(a) and 13(b)). Individual contorted and convolute laminae
D.A.V. Stow, M. A/am and D.J.W. Piper
134 (a) HARTLEN
POINT
( b ) ROSE POINT (c) CAPE VIEW
,d
d
3O
approx lOOm shale
&
20
partly covered
<~
1
10
section covered
by tide
I FACIES I--678
SECTION 50
t
continued
111
t
continued
SECTION 17P
SECTION107
FIG. 7. Measured sections of Halifax Formation showing medium-scale sequences, (a) Hartlen Point (b) Rose Point (c) Cape View.
Sedimentology of the Halifax Formation, Nova Scotia (a) B L U E R O C K S
(b)
ROSE
[35 POINT
~ued
_
SECTION 53
_
SECTION 1 15
|i ji m i
_
SECTION 52
~ cont.
SECTION 114
SECTION 17
FIG. 8. Measured sections of Halifax Formation showing small-scale sequences, (a) Blue Rocks (b) Rose Point.
I36
D . A . V . Stow, M . A l a m and D . J . W . Piper
L
I
I
approx. 2 m
FIG. 9. Photograph illustrating horizontal continuity of bedded siltstone and shale, Halifax Formation, Rose Point. are relatively common within otherwise undisturbed sections. Small-scale slump-deformed laminae and probable synsedimentary microfaults are also evident in parts. Larger-scale slump horizons are much less frequent, but have been observed at Rose Point and Blue Rocks. Chaotically disturbed sandstone/shale slump horizons are between 10 cm and 100 cm thick (Fig. 14) and can be traced laterally for variable distances up to more than 100 m.
Discussion Depositional processes From earlier work as well as from our own observations the Halifax Formation is clearly a turbidite-dominated sequence. Facies (7) and (8) comprise of thin- to thick-bedded classical turbidites that are directly comparable with Facies D 1 and C respectively of Mutti & Ricci Lucchi (1972). However, the greater part of the Halifax Formation, Facies (1) to (6), is very thin-bedded shaly flysch equivalent to Facies D, E and G of Mutti & Ricci Lucchi (1972). Similar facies have been described recently by many authors from both modern (Piper 1978; Nelson et al. 1978; Stow 1979; Chough & Hesse 1980; Hesse & Chough 1980; Kelts & Arthur 1981) and ancient
sediments (Mutti 1977; Ricci Lucchi & Valmori 1980; Shanmugam 1980; Hicks 1981; Pickering 1981a,b; Stow et al. 1982). Much of the shaly flysch consists of finegrained turbidites, recognized by such distinctive sedimentary structures as graded laminated beds (Piper 1972, 1978) and climbing ripples in the silt beds. These fine-grained turbidites do not show the Bouma structural sequence but can be better documented with either Piper's (1978) or Stow's (1977) (Stow & Shanmugam 1980) structural schemes. In using this latter scheme we can interpret the Halifax Formation Facies (1) to (6) as comprising repeated partial to nearly complete ideal sequences (Fig. 15). Facies (1), (2) and (3) comprise base-cut-out sequences beginning with the structural divisions T6-7, T4-5 and T3 respectively. Facies (4), (5) and (6) are more commonly top-cut-out sequences beginning with division T01 but missing one or more of the upper divisions. Other variations occur as, for example, the alternative Facies (5) type (closely-spaced parallel laminae) which can be interpreted as closely-repeated T3~ sequences. Such partial sequences result from exactly the same process of downslope turbidity current evolution as has been invoked for partial Bouma sequences. On the whole, the top-cut-out sequences are more proximal and the base-cut-out ones more distal.
Sedimentology of the Halifax Formation, Nova Scotia
L I. N O I l O 3 S
. '..'-"
I37
t',
I
/
//
/
~'=
I
I
I -~~
I
I
/
~.~ ~8
/
,
I
~
01, N 0 1 1 0 3 8
/ / /
/I/ /
/ /
~
III
.~~ -i
/
\
0 N011038
~
I
=o .=-.=
\ \ \
\
=o
~:...'.?::::"-i.'~""' ,::,.
"" ~
g' N0110:18
! ~I
111
0
I
I
1
I38
D.A.V. Stow, M. Alam and D.J.W. Piper
6
4
D
E
2 9~
6m separation
o SECTION 2
SECTION 3
FIG. 11. Two measured sections from Thunder Cove (Ovens) showing lateral variability over 6 m of section. Note amalgamation, lenticularity, pinching and swelling. Current direction at high angle into plane of paper. Similar sequences have been described recently from the South Atlantic (Stow, in press). It should be noted that the base-cut-out sequences identified in facies (1) and (2) are not intrinsically very distinctive; and within facies (1) to (4) the proportion of relatively featureless shale is high. Although some of the silt beds in these
FIG. 12. Measured section from Kings Bay (Rose Point) showing thick lenticular sandstones with erosive bases separated by shale rip-up zones. Horizontal and vertical scales are equivalent.
facies are demonstrably turbidites, we cannot show that the entire sequence is turbidite. We see some similarities with high sedimentation rate continental slopes (i.e. Hill, this vol.; Pickering, this vol.). For example, the Pleistocene of the Scotian Slope (Hill, this vol.) shows abundant sediment similar to Facies (1)-(4), although generally lacking ripples, with rare interbedded turbidites. Schenk has suggested (e.g. Harris & Schenk 1975; Schenk et al. 1980) that parts of the Halifax Formation may be contourite deposits, based on criteria for contourites proposed by Bouma & Hollister (1973). Although this interpretation is consistent with the regional palaeocurrent pattern, we believe many of Bouma & Hollister's criteria to be incorrect (Piper & Brisco 1975). For example, starved ripples may result from deposition of all silt-sized material in a turbidity current, and are not necessarily evidence of contourites. Indee& the lateral passage of fading ripples into characteristically turbidite graded laminated silt to mud beds suggests that the ripples must also be turbidites. Structures now recognized as typical of modern contourites, such as wispy, irregular lenses of silt (Stow & Piper, this vol.) are absent. D'Orsay (1981, 1984) has demonstrated by sequence analysis in the Ovens Point and Thunder Cove sections that prominently rippled Facies (2)-(5), with regular lenticular or wavy lamination, are associated with Facies (6)-(8) containing Bouma AB sequences and common slump and scour features. Planar laminated developments of Facies (2)-(5) are spatially distinct. Such a distribution of facies is the inverse of that
1
0
metres
SECTION 21
Sedimentology of the Halifax Formation, Nova Scotia
139
"~
0
2~
o
o ~ 9
~ o
.,~
I40
D.A.V. Stow, M. A lam and D.J.W. Piper
F16. 14. Photographs of slump units in Halifax Formation, (a) Rose Point (b) Blue Rocks. expected for a contourite model, where silt starvation would be commonest furthest away from the source in the channel areas. In a turbidite model, the observation can be explained in terms of flow becoming both slower and steadier away from channel spill-over. D'Orsay (1981, 1984) suggests that facies (6)-(8) and the lenticular laminae development of facies (2)-(5) were deposited in a turbidity current channel-levee complex while the planar laminated development of facies (2)-(5) is inferred to represent overbank deposition further from channels. Erosional features are widespread in the Halifax Formation sections just above the Goldenville contact on the Atlantic Coast of Nova Scotia. Load, scour and flute marks at the base of beds are common. Very rapid variation in bed thickness, associated with shale rip-up clasts and slumped sediments indicates the existence of local
relief, associated with either channel-levee complexes or giant flute marks (Normark et al. 1979). The common erosion at the surface of sandstone layers is less commonly reported and less readily explained. In some rippled beds of fine sandstone or siltstone, erosion results from autoerosion by ripple migration during a single depositional event prior to mud deposition. In most examples the causal process is unknown. Schenk (pers. comm.) suggests such erosion may be evidence of a pro-delta environment.
Depositional environment Several authors have interpreted vertical sequences of facies in terms of progradation of depositional environments or cyclic depositional processes. Basin-fill or megasequences are commonly related to gradual evolution of the tectonic
Sedimentology of the Halfax Formation, Nova Scotia .."
'.'.'.'.: : ; : 2
I41
":2:;",
I7 j.,.~'2.:2%..~.~75 "..':.2~- ....... T3 ..,... ......
:=.-
.:L:/""./'-'Z:.
"i~i...7"~. i~
ii.:ii!!::::ii'i::j FACIES 3
FACIES 1
FAC,ES 2
- ~ k 3
~Cr
IDEAL SEQUENCE
T7 ~e~.
~ U / -Sx" CO?
~ .o,
15 ":-22.-'.2-'-22 ~
~
~.....
1
D]
~
~................ ,'~ -
-'~
FACIES6
I',"~.->";:'.'~ FIG. 15. Stow and Shanmugam's (1980) ideal scheme of sedimentary structures for silt-mud turbidites and its application to facies recognized in the Halifax Formation. Silt-stippled, shale-blank. regime, sediment supply or depocentre location. Ideal large-scale sequences include coarseningupwards, fining-upwards and uniform arrangements ofturbidites over tens to a few hundreds of metres. These have been interpreted as representing, respectively, mid-fan lobe, channel and basin-plain environments (Mutti & Ricci Lucchi 1972; Rupke 1977). More recently, however, it is being recognized that the 'ideal' sequence is rare and that partial, interrupted and mixed sequences are more common (Mutti 1977; Ricci Lucci 1977; Hicks 1981; Pickering 1981a,b). Shanmugam (1980) and Stow (1981) have used both the small-scale rhythms and individual fine-grained turbidite sequences, over a few millimetres to a metre of section, to characterize different subenvironments and controls of turbidite sedimentation. The following characteristics of the Halifax Formation are of importance in inferring its depositional environment: (1) the presence of large-scale (100-200 m) fining-upward, coarsening-upward and irregular sequences; (2) the oscillation between packets of shale-rich and sandstone/siltstone-rich facies over intervals of 10-20 m;
(3) the relative uniformity and flat bedding of the thin sandstones and finer-grained facies, together with a slight lenticularity of most thicker sandstones over 50-100 m of section; (4) local small-scale erosion, channelling and slumping. In addition, the stratigraphic position of the Halifax Formation between the proximal sandy flysch of the Goldenville Formation and the shelf sequence of the White Rock Formation places constraints on possible depositional settings. The large-scale cyclicity of the Halifax Formation may reflect progradation of muddy deep-sea fan lobes or levees (coarsening-upward), interrupted by muddy channel-filling sequences (fining-upward). Alternatively, the cyclicity may reflect progradation and retreat of a shelf source area, either deltaic or pro-glacial. The lack of coarse channel-filling sandstones within the Halifax Formation does not conform with a classical deep-sea fan interpretation. The lenticularity of thicker sandstones, the occurrence of packets of sandstone and siltstone-rich rock, and the evidence of erosion, channelling and slumping suggests that these sandier rocks may represent a 'channel' facies in a predominantly muddy sequence lacking coarse sediment supply.
142
D.A.V. Stow, M. Alam and D.J.W. Piper
There is no direct evidence for the water depth in which the Halifax Formation accumulated, so that the sediments could have accumulated in a relatively shallow pro-delta environment (Schenk, pers. comm.) similar to the turbidite facies described by de R a a f e t al. (1964), Kepferle (1978) and Pickering (1981b and this volume). The lower part of the Halifax Formation on the Atlantic coast of Nova Scotia shows many features characteristic of deposition on channellevee complexes of modern muddy deep-sea fans. The Holocene upper supra-fan of Navy Fan off California (Normark et al. 1979) can be used as an example. The levees have frequent erosional scours, often partially filled with fine-grained material. Back slopes of levees on modern deepsea fans often have gradients of several degrees and show slumping. Coarse spillover from channels is rare, and related to occasional levee breaching, thus producing a fining-up sequence of muddy flysch as the breach is healed. The abundant wavy bedding and autoerosion is probably related to unstable flow conditions associated with rapid flow expansion and entrainment as the current overtops the levee (Bowen et al. 1984). Whether the channel facies is fine-grained and represented by Facies (6)-(8) of the Halifax Formation, or whether it is represented by occasional intercalations of Goldenville Formation lithology is uncertain. We see no evidence for
more distal deep-sea fan deposition associated with lobes or basin plain: there is a lack of distal turbidite sands, and the muddy turbidites lack distal characteristics such as extremely fine laminated silts (Piper 1978). The finer upper parts of the Halifax Formation (as seen in the Annapolis Valley) show similarities with Quaternary continental slope sequences in areas where there is a large fluvial or glacial discharge of fine sediment, but coarse sediment is trapped in the continental shelf (Hill, this vol.).
Conclusions The Halifax Formation is a thick shaly flysch succession which shows various scales of cycles of resedimented shale and siltstone facies, with rare turbidite sandstone intercalations. The lower part of the Formation closely resembles modern deepsea channel-levee complexes; the upper may have accumulated principally from turbidity currents on a rapidly prograding continental slope. ACKNOWLEDGEMENTS: This work was initiated by P.E. Schenk, and was supported by N S E R C grants to Schenk & Piper, and by a Royal Society John Murray Travelling Scholarship to Stow. We thank our respective departments for technical and secretarial assistance.
References BOUMA, A.H. & HOLL1STER,C.D. 1973. Deep ocean basin sedimentation. In: Middleton, G.V. & Bouma, A.H. (eds), Turbidites and Deep Water Sediments. Soc. Econ. Paleon. Min. Short Course Notes, Anaheim. 79-118. BOWEN, A.J., NORMARK,W.R. & PIPER, D.J.W. 1984. Numerical modelling of turbidity currents on Navy Submarine fan, California. Sedimentology, in press. CAMPBELL,F.H.A. 1966. Palaeocurrents and Sedimentation in Part of the Meguma Group, Lower Palaeozoic of Nova Scotia, Canada. M.Sc. Thesis, Dallhousie University, Halifax, Nova Scotia. 91 pp. CHOUGH, S.K. & HESSE, R. 1980. The Northwest Atlantic Mid-Ocean Channel of the Labrador Sea: III Head Spill vs Body Spill deposits from turbidity currents on natural levees. J. sed. Petrol., 50, 227-34. CLARKE, D.B. & MUECKE, G.K. 1980. Igneous and metamorphic geology of southern Nova Scotia. Geological Association of Canada FieM Trip Guidebook, 21, 101 pp., Halifax. DE RAAF, J.F.M., READING, H.G. & WALKER, R.G. 1965. Cyclic sedimentation in the lower Westphalian of north Devon, England. Sed#nentology, 4, 1-52.
D'ORSAY, A.M. 1981. Depositional Analysis of the Upper Meguma (Halifax Formation) at Ovens Park, Lunenburg County, Nova Scotia. B.Sc. Hons. Thesis, Dalhousie University, Halifax, Nova Scotia 74 pp. - 1984. Depositional analysis of the Halifax Formation (Meguma Group) at Ovens Park, Nova Scotia. J. sed. Petrol., in press. FYSON, W.K. 1966. Structures in the Lower Paleozoic Meguma Group, Nova Scotia. Bull. geol. Soc. Am., 7 7 , 931-44. HARRIS, I.M. 1971. Geology of the Goldenville Formation, Taylor Head, Nova Scotia. Unpubl. Ph.D. Thesis, Univ. of Edinburgh. 226 pp. HARRIS, I.M. & SCHENK, P.E. 1975. The Meguma Group. Maritime Sediments, l l, 25-46. HESSE, R. & CHOUGH, S.K. 1980. The Northwest Atlantic Mid-Ocean Channel of the Labrador Sea: II. Deposition of parallel laminated levee-muds from the viscous sublayer of low density turbidity currents. Sedimentologv, 27, 697-711. HICKS, D.M. 1981. Deep-sea fan sediments in the Torlesse zone, Lake Ohau, South Canterbury, New Zealand. N.Z. Jl. Geol. Geophys., 24, 209-30. HORN, D.R., EWING,M., HORN,B.M. & DELACH,M.N. 1971. Turbidites of the Hatteras and Sohm Abyssal
Sedimentology of the Halifax Formation, Nova Scotia Plains, Western North Atlantic. Marine Geol., 11, 287-323. KELTS, K. & ARTHUg, M.A. 1981. Turbidites after ten years of deep-sea drilling- wringing out the mop? Soc. econ. Palaeo. Min. Spec. Pub., 32, 91-127. KEPVERLE, R.C. 1978. Prodelta turbidite fan apron in the Borden Formation (Mississippian), Kentucky & Indiana. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stoudsburg, Pa. 224-38. LANE, T.E. 1976. Stratigraphy of the White Rock Formation. Maritime Sediments, 12, 87-106. -1981. Stratigraphy and Sedimentology of the White Rock Formation (Silurian) Nova Scotia, Canada. M.Sc. Thesis, Dalhousie University, Halifax, Nova Scotia. 270 pp. MUTTI, E. 1977. Distinctive thin-bedded turbidite facies and related depositional environments in the Eocene Hecho Group (south-central Pyrenees, Spain). Sedimentology, 24, 107-31. & RICCI LUCCHI, F. 1972. Le torbiditi dell'Appenino settentrionale: introduzione all'analisi di facies. Soc. geol. Ital. Mere., 11, 161-99. NELSON, C.H., NORMARK, W.R., BOUMA, A.H. & CARLSON, P.R. 1978. Thin-bedded turbidites in modern submarine canyons and fans. In: Stanley, D.J. & Kelling, G., (eds), Sedimentation in Submarine Canyons, Fans, and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 177-89. NORMARK, W.R., PIPER, D.J.W. & HESS, G.R. 1979. Distributary channels, and mesotopography of Navy Submarine Fan, California Borderland, with applications to ancient submarine fan sediments. Sedimentology, 26, 749-74. PmNNEY, W.C. 1961. Possible turbidity current deposit in Nova Scotia. Bull. geol. Soc. Am., 72, 1453-54. PICKERILL, R.K. & KEPP1E, J.D. 1981. Observations on the ichnology of the Meguma Group (?CambroOrdovician) of Nova Scotia. Maritime Sediments and Atlantic Geology, 17, 130-38. PICKERING, K.T. 1981a. Two types of outer fan lobe sequence, from the late Precambrian Kongsfjord Formation submarine fan, Finnmark, North Norway. J. sed. Petrol, 51, 1277-86. 1981b. The Kongsfjord F o r m a t i o n - a Late Precambrian submarine fan in NE Finnmark, North Norway. Norges geol. Unders., 367, 77-104. PIPER, D.J.W. 1972. Turbidite origin of some laminated mudstones. Geol. Mag., 109, 115-26. PIPER, D.J.W. & BRISCO, C.D. 1975. Deep water continental margin sedimentation, Deep Sea Drilling Project, Leg 28, Antarctica. In: Hayes, D.E., Frankes, L.A. et al., (eds), Init. Rept., 28, 727-55. PIPER, D.J.W. 1978. Turbidite muds and silts on deep-sea fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans, and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. POOLE, W.H. 1970. Geology of south-eastern Canada. In: Geology and Economic Minerals of Canada. Econ. Geol. Report, 1, 227-304. RIccl LUCCHI,F. 1978. Turbidite dispersal in a Miocene deep-sea plain: the Marnoso-Arenacea of the
-
-
-
-
i43
Northern Apennines: Geologic en Mijnbouw, 57, 559-76. RIr162 Lucern, F. & VALMOR1, E. 1980. Basin-wide turbidites in a Pliocene over-supplied deep-sea plain: a geometrical analysis. Sedimentology, 27, 241-70. RUPKE, N.A. 1977. Growth of an ancient deep-sea fan. J. Geol., 85, 725--44. SCHENK, P.E. 1970. Regional variation of the flysch-like Meguma Group (Lower Paleozoic) of Nova Scotia compared to recent sedimentation off the Scotian Shelf. In: Lajoie, J. (ed.), Flysch Sedimentology in North America. Geol. Ass. of Can., Spec. Pap. 7, 127-53. 1971. Southeastern Atlantic Canada, northwestern Africa, and continental drift. Can. J. Earth Sci., 8, 1218-51. 1972. Possible late Ordovician glaciation of Nova Scotia. Can. J. Earth Sci., 9, 95-107. 1978. Synthesis of the Canadian Appalachians. In IGCP Project 27, Caledonian-Appalachian Orogen of the North Atlantic Region. Geol. Surv. Can., Pap. 78-13, 111-36. , LANE, T.E. & JENSEN, L.R. 1980. Palaeozoic history of Nova Scotia- a time trip to Africa (or South America?). Geol. Ass. Can. Field Trip Guidebook, 20, 82 pp. Halifax. SHANMUGAM, G. 1980. Rhythms in deep-sea, finegrained turbidite and debris-flow sequences, Middle Ordovician, eastern Tennessee. Sedimentology, 27, 419-32. SM1THERINGALE,W.G. 1973. Geology of parts of Digby, Bridgetown, and Gaspereau Lake map-areas, Nova Scotia. Geol. Surv. Can. Mere., 375, 78 pp. STOW, D.A.V. 1977. Late Quaternary Stratigraphy and Sedimentation on the Nova Scotian Outer Continental Margin. Unpubl. Ph.D. Thesis, Dalhousie University, Halifax Formation, Nova Scotia. 1979. Distinguishing between fine-grained turbidites and contourites on the Nova Scotian deep water margin. Sedimentology, 26, 371-88. 1981. Laurentian Fan: morphology, sediments, processes and growth patterns. Bull. Am. Assoc. Petrol. Geol., 65, 375-93. , in press. Turbidite facies, associations and sequences in the southeast Angola Basin. In: Hay, W.W. Sibuet, J.C. et al., Init. Rept. DSDP, 75. --, BISHOP, E.D. & MILLS, S.J. 1982. Sedimentology of the Brae Oilfield, North Sea: fan models and controls. J. Petrol. Geol., 5, 129-48. -• SHANMUGAM, G. 1980. Sequence of structures in fine-grained turbidites: comparison of Recent deepsea and ancient flysch sediments. Sed. Geol., 25, 23-42. TAYLOR, F.C. 1967. Reconnaissance geology of Shelburne and Yarmouth counties, Nova Scotia. Geol. Surv. Can Mere., 349, 83 p. 1969. Geology of the Annapolis - St. Mary's Bay map area, Nova Scotia (21A, 21B East Half). Geol. Surv. Can. Mere., 358, 65 pp. WALKER, R.G. & MUTTI, E. 1973. Turbidite facies and facies associations. In: Middleton, G.V. & Souma, A.H. (eds.), Turbidites and Deep-water Sedimen-
-
-
-
-
-
-
-
-
-
I44
D.A.V. Stow, M. Alam and D.J.W. Piper
tation. Pacific Section, Soc. econ. Min. Paleo., Short Course 119-57 pp. WILHAMS, H. 1978. Tectonic Lithofacies Map of the
Appalachian Orogen, Memorial Univ. of Nfld. Map No. I.
D.A.V. Sxow, Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, Scotland. M. ALAM,Department of Geology and Mining, University of Jos, Private Mail Bag 2084, Jos, Nigeria. D.J.W. PIPER,Atlantic Geoscience Centre, Geological Survey of Canada, Bedford Institute of Oceanography, PO Box 1006, Dartmouth, Nova Scotia, Canada B2Y 4A2.
Sedimentation in the southern Cape Verde Basin: regional observations by long-range sidescan sonar R.B. Kidd and R.C. Searle S U M MARY: A long-range sidescan sonar survey of the lower continental rise in the Cape Verde Basin has detected a dominant acoustic facies characterized by lineated backscattering patterns. These are interpreted as the sedimentary expression of turbidity current pathways across the area. Primary sediment transport directions can be interpreted from the backscattering patterns. Towards the north and west sediment distribution is increasingly affected by abyssal hill topography, either exposed oceanic basement features or hills draped with pelagic sediment. The largely fine-grained sediments of lower continental rise and abyssal plain environments have, until recently, been neglected by sedimentologists working in the marine field as locations for studies of sedimentary processes. Results of the HEBBLE experiment, mounted on the New England continental rise (Nowell et al. 1982), and the activities of the Seabed Working Group (SWG, an international consortium of scientists examining the feasibility of sub-seabed radioactive waste disposal in the deep basins of the North Atlantic and Pacific Oceans) have begun to question long-held concepts of sedimentation in such environments. This contribution, and the one which follows (Auffret et al., this volume), present some preliminary results of British and French studies under the auspices of the SWG in and around a study area (called CV 1) in the Cape Verde Basin of the Eastern Atlantic (Fig. 1). Here we concentrate upon the regional sedimentary environment as interpreted from a long-range sidescan sonar survey. The Institute of Oceanographic Science's longrange sidescan sonar system (GLORIA) is a dual sidescan that provides a regional sonograph of the ocean floor analagous to an aerial photograph taken at grazing incidence. The sonograph is in the form of a swath, up to 60 km wide, centred on the ship's track (Somerset al. 1978). The central portion of the swath, equivalent to one or two times the water depth, lies effectively in a transmission null of the sonar arrays. Thus, a zone about 5 km to each side of the ship's track is typically not insonified. The system uses a 6.5 kHz source transmitted as a 4-second pulse, in conjunction with a correlating receiver, to achieve an improved signal-to-noise ratio. The images interpreted for this study were analogue tape-recorded then replayed on to a line-scan photographic recorder and subsequently stretched photographically to remove variations in ship's speed and to set the cross-track and along-track scales equal. No slant-range correction was applied. In
practice this correction would only be useful in the one-third of the sonograph closest to the ship's track. The effect would be to move features detected by the sidescan at most two kilometres closer to the track. The nominal separation of targets downrange is about 45 m (for a 40-see transducer cycle setting) and the nominal resolution parallel to the ship is approximately 200 m at a 10 knots survey speed (40 sees • ship's speed in m/see). The surveys were run from the charter vessel, 'RV Farnella,' in 1981. Precision echo-sounder, magnetometer and gravimeter profiles were collected simultaneously along track, as well as seismic reflection profiles over selected areas. Ship's speed was generally around 10 knots for G L O R I A surveys but was limited to 7 to 8 knots for seismic-reflection profiling in and around the CV1 study area. G L O R I A has been used extensively over most environments of the deep ocean floor during its first decade of operation (Laughton 1981). Continental rise and abyssal plain areas typically show very bright echoes on the sonographs that are found to correspond to areas of outcropping rock (e.g. the sides of abyssal h i l l s ) j u t t i n g through a sediment cover of unlithified sediments that themselves backscatter sound much less strongly. In addition, however, the sonographs often show acoustic backscattering changes over these sedimented surfaces that do not correspond with identifiable changes in relief on echosounder profiles (Kidd 1982; Kidd & Roberts 1982; Roberts & Kidd 1984). Oceanographic echo-sounders typically have a resolution of _+ 1 metre. The backscattering changes have been interpreted as changes in bottom roughness that are comparable to the G L O R I A wavelength of around 25 cm. The roughness would result from sedimentary bedforms or sediment grain-size variations or a combination of both. In the extreme, sedimentary bedforms would be of mega-ripple or sand-ribbon dimensions while
I45
146
R.B. Kidd and R.C. Searle 30~
j~ 327-
oo Saharan G Seamounts
CVl
.
/~
.
.
CANARY ISLANDS
.
321
Kane
~,o00
/ ~
oO~
:
20~
% ~)
ISLANDS
0_
~
i
Z 3( Pw
20~
FIG. 1. Location of the R.V. Farnella GLORIA survey and the CV1 study area in the southern Cape Verde Basin. Bathymetric contours at 1000 m intervals are from GEBCO chart No. 5.08 (Searle et al. 1982). Along the ship's track are noted Julian day numbers for 1981 and the thickened line represents that part of the GLORIA survey that is illustrated in Fig. 2. grain sizes would require reaching cobble grade. Whatever interpretation is more likely for a particular area, the mapping of these changes in acoustic backscattering offers a clear potential for studies of distributions of sediment facies on a grand scale in low relief areas and for interpretation of regional sedimentary processes (Kidd & Roberts 1982). The survey over the lower Cape Verde continental rise provides just such an insight into regional sedimentary processes on the deep ocean floor.
Physiographic setting of the GLORIA survey The CV1 study reported in detail in Auffret e t al.
(this volume) is located at around 5200 m water depth in the southern Cape Verde Basin, almost 1000 km due west of the western Sahara (Fig. 1). The G L O R I A surveys conducted in 1981 extended from the continental rise across the Cape Verde 'Abyssal Plain' to the ridge flank topography at around 2 9 W . This contribution deals with the lower continental rise between 23 ~ and 2 7 W . , in water depths ranging from 4900 m to 5500 m. Regional slopes here range from about 1 in 400 in the south and east to 1 in 800 in the north and west and, with the exception of isolated abyssal hills, this is essentially the lower continental rise. There is no indication of a true abyssal plain in this area, with a gradient less than 1 in 1000 and flat-lying, ponded sediments. The sea floor east of the survey area rises
Sedimentation in the southern Cape Verde Basin gradually towards the Canary continental slope which begins about 500 km away. The Cape Verde Islands are 600 km away to the southsouth-east and the sea floor shows a gradual rise in this direction until the slopes of the Cape Verde Plateau are reached. To the south-west, the sea floor is relatively flat until it shoals southward across the Kane Fracture Zone near 21.5~ 28.5~ To the west and north-west, several isolated abyssal hills rise above the general level of the plain. The sea floor again rises immediately north of the area into a lower ridge flank province whose north-south and east-west tectonic structures have been characterized as a continuation of the Mid-Atlantic Ridge trend and its offset along the Atlantis Fracture Zone (Rona & Fleming 1973).
Description of the sonograph coverage A montage of the GLORIA sonographs for the area between 23 ~ and 25~N. and 2 3 to 27 ~W. is presented in Fig. 2. The lightest areas of the sonographs are those of maximum signal return while the darker areas are generally acoustic shadow zones or areas of high acoustic absorbency. Considerable signal loss occurs on these records in the far range because of sound propagation along thermo/haloclines in the water column. The maximum range of interpretable record in the sonograph swath varies from 15 to 30 km away from ship track. Figure 3 is a line diagram interpretation constructed to show the trends of the sediment lineaments and their relationship to the other acoustic facies. The CV 1 study area lies in the northern part of this portion of the lower continental rise. Features related to oceanic basement structures
Abyssal hills characterize the north-western quarter of the GLORIA survey area. They are identified on GLORIA sonographs by their high acoustic reflections and by their obvious relief, as indicated by shadows, and by echo-sounder and seismic profiles that cross some of them. They are mostly linear and are aligned north-north-eastsouth-south-west, parallel to the magnetic isochrons. Many appear fault-bounded on profile crossings and are interpreted as relict Mid-Atlantic Ridge topography. They stand up to 300 m above the surrounding sea-bed. Some abyssal hills, as at 25.0~ 24.5~ have 'aprons' of sediment around them of slightly different (usually darker) acoustic character. A small area in the south-western corner (26.2 W. to 27.OW.) of this GLORIA montage is
I47
characterized by alternating light and dark bands with individual north-north-east-south-southwest strikes. These features terminate at their northern ends along a west-north-west-eastsouth-east trending line (approx. 23.3N.). One or two very bright lineaments within this area look like typical abyssal hill basalt outcrops. Immediately west of longitude 27~ the ship's track ran over gently undulating topography with features tens of metres high that correlate with the same variations in G L O R I A backscattering intensity. This appears from seismic profiles to be an area where the topography of the oceanic basement is draped by pelagic sediment; it remains a hilly region around which any terrigenous or hemipelagic sediments derived by gravity-controlled transport from the continental margins might be expected to pond. The alternating light and dark bands must reflect differences between pelagic and hemipelagic or terrigenous sedimentation although, because of the lack of sufficiently precise sediment sampling at present, we cannot yet identify specific lithologies. The west-northwest-east-south-east line against which the acoustic facies variations terminate is, presumably, the trace of a fracture zone. Features related to surface sediments
By far the most common and consistent G L O R I A acoustic facies in our survey is found in the eastern part of the area. This is generally characterized by strongly-lineated bright images. The acoustic pattern comprises broad and narrow bands or streamers of high backscatter interspersed with occasional narrow, dark lineaments (e.g. at 23.7~ 25.8~ Some, but not all, of the lineaments in the south of the survey area (including the example quoted) correspond to local minima, of 10 m or less, on the 10 kHz echo-sounder profiles. The strike of the lineaments in the south and east is consistently about 305 ~. Further G L O R I A records east of those shown in Figs 2 and 3 indicate that the lineated acoustic facies variations, although less welldeveloped, extend as far as 24.7 N., 21.5~W. The area of brightest backscattering is between 25 and 26 W. at 23.5 N . Jacobi & Hayes (1982) mapped acoustic facies interpreted from 3.5 kHz high-resolution seismic reflection profiles immediately south of our survey area. They show linear depressions containing stratified sequences that they interpret as ~turbidity current pathways'; one extending with a 339 trend from the Cape Verde Plateau, the other with a 298 trend from the west African margin. These 'pathways' appear to coalesce at 23 N., 2 3 . 5 - 2 6 W . The lineated GLORIA
148
R.B. Kidd and R.C. Searle
o
E
3
~9
o ,< 9
Sedimentation in the southern Cape Verde Basin
I49
FIG. 3. Interpretation of the GLORIA sonograph coverage around the CVI study area.
acoustic facies with a 305 ~ trend that we observe could reflect the more distal parts of such a turbidity current flow pattern. Elsewhere on the montage (Fig. 2) the sediment lineament trend is less uniform. West of 27 :W. the lineaments appear to be diverted by the dominant north-north-east-south-south-west ridge-flank topography that just enters the south-western corner of the study area. In the far south-west, lineament directions change to westerly and even west-south-west. However, further west, outside the montaged area, it returns to about 284 ~. This is the regional fracture zone trend and it appears that the distribution of the surface sediment is being controlled by the underlying fracture zone morphology. Near the northern edge of the survey area, between 24~ and 25~ the strike of the sediment lineaments changes from west-northwest to west-south-west and between 25~ and 26~ it is consistently west-south-west. The CV 1 study area lies within this zone of changing directions and contains individual sediment lineaments striking west-north-west, west, and westsouth-west. Finally, between 24~ and 25~ and west of 25~ the lineament directions are even less consistent and a north-north-eastsouth-south-west trend becomes apparent, especially west of 26~
Interpretation of the acoustic facies Sediment lineaments similar to those described above were first detected on G L O R I A coverage of the southern Biscay Abyssal Plain (Kidd & Roberts 1982). Distinct variations in acoustic backscattering were recognized with diffuse ~ boundaries aligned oblique to the foot of the north Spanish continental slope. The lineaments there are clearly continuous with the bathymetrically-mapped interplain channel system (Laughton 1960) that begins north-west of Galicia Bank, with around 2 m of relief, and deepens southwestwards towards the gorge linking the Biscay and Iberia abyssal plains known as Theta Gap. Eastwards and upstream, the backscattering patterns occur where no relief is identifiable on echo-sounder records. They appear to originate at the mouths of a number of submarine canyons on the continental slope. The lineated G L O R I A acoustic facies on the Biscay Abyssal Plain was interpreted as marking the pathways of turbidity current flows that had imparted a bottom roughness effect to the sea-bed. The roughness was thought to result from either sedimentary bedforms or textural changes of the sediments themselves. During a G L O R I A traverse of the southern Hatteras Abyssal Plain, similar records show-
I50
R.B. Kidd and R.C. Searle
ing a lineated acoustic facies were obtained (Kidd 1982). Again, echo-sounder profiles showed little or no relief. To the east, a later, still unpublished, GLORIA survey detected further evidence of the lineated facies through Vema Gap and on to the Nares Abyssal plain. Again, an explanation was advanced that the G L O R I A acoustic facies depicts an effect of turbidity current flow across the abyssal plain sediments. The above interpretations have not been tested by near-bottom geophysical and photographic surveys nor by precisely navigated sediment sampling. However, one study of a different area has attempted to investigate the causes of backscattering on G L O R I A sonograph coverage. Prior to the 'Farnella' surveys reported here, the ship was used to obtain GLORIA and seismic records over another part of the continental rise of the Cape Verde Basin between 30~ and 3 2 N . referred to as the Saharan rise. A later sampling and survey cruise to that area by 'RRS Shackleton' provided bottom photographs and a small number of core samples confidently located with respect both to specific acoustic facies on the sonographs and to cross tracks taken with a 3.5 kHz high-resolution seismic profiler. Lithological interpretation remains tentative when these preliminary results are transposed to another part of this continental rise. Preliminary results, supported by analysis of the core samples from the northern area do suggest, nevertheless, the following general associations (Kidd et al., in preparation). First, debris-flow deposits give rise to generally lobate shapes in plan-view displaying strong backscattering within generally low backscattering sedimented areas. Presumably, the strong backscattering results from the hummocky surface microrelief of the debris-flows (Embley 1976) that has a roughness comparable with the G L O R I A acoustic wavelength (25 cm). Secondly, pelagic sediments appear to give very low levels of backscattering; isolated inliers of pelagic sediment within the Saharan sediment slide (Embley 1976) often appear completely black on the sonographs, presumably because they display a very smooth surface that produces specular reflection. Thirdly, proximal turbidites on the continental rise cause moderate backscattering with a lineated fabric, whereas fairly low backscattering areas lacking lineations characterize the abyssal plain to the west. The slope of the sea-bed also affects the degree of acoustic backscattering, although probably less strongly than does bottom roughness. Thus, moderate to steep slopes that face the sonar tend to return stronger echoes than do flat or gently-sloping areas. Areas sloping sufficiently steeply away from the sonar produce acoustic shadows. This effect is particu-
larly noticeable near maximum range on the sonographs where the sound rays are at almost grazing incidence. Finally, variations in acoustic impedance of the sediments should have some, but probably not a large, effect. The bright areas on the southern and eastern Cape Verde Basin sonographs we can identify, by analogy with these results further north, as turbidity current deposits possibly ranging to debris flows. Some of the dark lineaments within the region of generally bright acoustic facies may be true sediment channels. The group of eastnorth-east-west-south-west lineaments noted near 25.0~ 25.0~ are similar to those identified as channels by Kidd & Roberts (1982) north-west of Galicia Bank. The dark lineaments in the south of the survey area that correspond to local minima on the 10-kHz echo-sounder profiles resemble the 5 to 10 m deep channels they report north-east of Theta Gap. Near 23.5~ 26.7~ several such features run east-west near the northern limit of the G L O R I A swath. Here, channels may again occur and are clearly seen by G L O R I A because the sound rays strike the sea floor at near-glancing incidence, producing shadows behind the low-relief channels. Elsewhere, the dark lineaments may simply be regions of different sediment between streamers of brighter material. Even where true channels are not evident, it appears from the distribution of the bright and dark lineaments that they probably all reflect directions of bed-load transport. Sediment transport directions inferred from the lineaments are indicated with arrows on Fig. 3. Clearly, the interpretation that the sediment lineaments are produced by turbidity currents is consistent with Jacobi & Hayes's (1982) interpretations further south. If the lineated facies does represent a continuation of their coalesced 'pathways' it would seem that their eastern 'pathway', originating on the west African margin, has the greatest influence at the present day on the distal areas of the lower rise. The surface sediments of the CV 1 study area, as reported in the following contribution (Auffret et al., this volume), are predominantly marly nannofossil ooze. Coarse turbidite sands have been cored, but over a metre below the sediment surface. This would suggest that the cause of any bright backscattering there is unlikely to result from near-surface coarse sediment, although penetration of the GLORIA pulse to depths of several tens of centimetres is quite possible. Auffret et al. report three 3.5 kHz acoustic facies derived from their detailed high-resolution seismic profiling, and also east-west trending ridges and swells of 5 to 10 m relief detected from
Sedimentation in the southern Cape Verde Basin SEABEAM swath/bathymetric mapping. They correlated two of the acoustic facies with their core samples and identified a 'hemipelagic facies' that drapes slight topographic highs and a 'channel facies' that contains ponded turbidites. A third acoustic facies was not sampled but is suspected to represent true pelagic, possibly older, material over basement highs. Auffret et al. point out that the overall three-dimensional distribution of the sedimentary sequences in the area reflects acoustic basement trends at an orientation around 30 ~. However, the east-west trending swells and ridges in their bathymetry are more directly relatable to the G L O R I A lineaments. Bright backscattering correlates closely with the distribution of the 'channel facies'. Gardner & Kidd (1983) have discussed the fact that the horizontal dimensions of meso-scale sea floor features, their linearity, their facing slope angles, their orientation relative to the ship's track and their distance away from the track, all play a deciding role in whether or not a feature will backscatter enough energy to be resolved on analogue G L O R I A sonographs. In this particular case, we can be more confident of the dominant trends of the lineaments because the G L O R I A survey trackline in this part of the region provides overlapping sonograph coverage from a number of directions. The strong contrasts in backscattering may relate simply to changes in slope-angle but we suspect that a roughness effect imparted by sedimentary bedforms is also involved. We conclude that the sediment lineaments provide a record of the primary trends of the CVI channel facies at the present day. A general east-west (downslope) orientation would appear to dominate the region as a whole. The slight changes in direction of the sediment lineaments westwards across the CV1 area would be interpreted by us as a result of turbidity currents flows that encounter the edges of minor topographic perturbations caused by draped abyssal hills. The higher ridge-flank province north of 2 5 N . causes stronger deflections. Other major deflections can be interpreted elsewhere in the survey area, especially in the north-west. There remains the question of the sediment
15I
lineaments that occur where no identifiable relief changes occur. It seems likely that these lineaments could represent erosional, depositional or biologically-produced bedforms developed at the sediment surface. At CVI these effects could be present in addition to the recorded subtle changes in topography and slope angle. Future near-bottom surveys and more precise sampling, relative to the acoustic facies displayed on the sonographs and 3.5 kHz records, should solve this question.
Summary Acoustic backscattering changes on long-range sonographs of lower continental rise and abyssal plain environments appear to reflect the pathways of turbidity current flows across these areas. The exact cause of the bottom roughness that produces the variable acoustic backscattering is unclear at this time. It seems certain that, for the CV1 study area, whose surface sediments are mainly clay to silt size, the cause is not simply a grain-size effect. The sonograph coverage allows a dominantly east to west transport of sediment to be established.
ACKNOWLEDGMENTS: The crew and scientists aboard R.V. Farnella during the Cape Verde Basin leg of this charter cruise, in particular the G L O R I A technical support team, are gratefully acknowledged for their contribution to the data collection reported upon here. Amanda Bates and Peter Hunter prepared the illustrations at lOS. Internal reviews were provided by James Gardner, Robert Whitmarsh, Douglas Masson and Michael Somers. External reviewers were Nigel Fannin and Carol Pudsey. They are all thanked for their constructive and helpful comments. NOTE: These preliminary findings result from research carried out for the Department of the Environment as part of its radioactive waste management research programme. The final results will be used in the formulation of Government policy but, at this stage, they do not necessarily represent Government policy.
References EMBLEY,R.W. 1976. New evidence for the occurrence of debris flow deposits in the deep sea. Geology, 4, 371-4. GARDNER, J.V. & KIDD, R.B. 1983. Sedimentary processes on the Iberian continental margin viewed by long-range sidescan sonar. Part l: Gulf of Cadiz. Oceanologica Acta, 6, 245-54.
JACOB! R.D. & HAYES,D.E. 1982. Bathymetry, microphysiography and reflectivity characteristics of the West African Margin between Sierra Leone and Mauretania. In: Von Rad, U. et al. (eds), Geology of the Northwest African Continental Margin. Springer-Verlag, Berlin. 182-212. KIDD, R.B. 1982. Long-range sidescan sonar studies of
I52
R.B. Kidd and R.C. Searle
sediment slides and the effects of slope mass sediment movement. In: Saxov, S. & Nieuwenhuis, J.K. (eds), Marine Slides and Other Mass Movements. Plenum Press, New York and London. 289-303. & ROBERTS,D.G. 1982. Long-range sidescan sonar studies of large-scale sedimentary features in the North Atlantic. Bull. Inst. Geol. Bassin d'Aquitaine, Bordeaux, 31, 11-29. , SIMM, R.W. & SEARLE, R.C. 1982. A GLORIA survey of the Saharan Sediment Slide and its calibration with high-resolution seismic profiling, coring and near-bottom surveys, in preparation. LAUGHTON, A.S. 1960. An interplain deep-sea channel system. Deep-Sea Res., 7, 75-88. -1981. The first decade of GLORIA. J. geophys. Res., 86, 11511-34. NOWELL, A.R.M., HOLLISTER, C.D., JUMARS, P.A. 1982. High-energy benthic boundary layer experi-
ment: HEBBLE, EOS Transactions. Am. Geophys. Union, 63, 554-95. ROBERTS, D.G. & KIDO, R.B. 1984. Sedimentary and structural patterns on the Iberian continental margin: an alternative view of continental margin sedimentation. Mar. Petrol._ Geol., 1, 37-48. RONA, P.A. & FLEMING, H.S. 1973. Mesozoic Plate Motions in the Eastern Central North Atlantic. Marine Geol., 14, 239-52. SEARLE, R.C., MONAHAN, D. & JOHNSON, G.L. 1982. General bathymetric chart of the Oceans (GEBCO ) Sheet No. 5.08 (5th edition). Canadian Hydrographic Service, Ottawa. SOMERS, M.L., CARSON, R.M., REVIE,J.A., EDGE, R.H., BARROW, B.J. &ANDREWS, A.G. 1978. GLORIA II: an improved long-range sidescan sonar. Ocean. Int., 78, 16-24.
R.B. KIDD • R.C. SEARLE,Institute of Oceanographic Sciences, Wormley, Godalming, Surrey, GU8 5UB, UK.
S e d i m e n t a t i o n in the southern C a p e Verde Basin: seismic and sediment facies
G.A. Auffret, R. Le Suave, R. Kerbrat, B. Sichler, S. Roy, C. Laj and C. Muller SUMMARY: A detailed seismic and sedimentological study of a small area at the lower continental rise to abyssal plain transition off NW Africa has revealed a complex pattern of Plio-Quaternary sedimentation. The area is one of extremely low relief. However, the sediment facies distribution mirrors the slight morphological expression of basement structures and downslope shallow channels. Fine-grained hemipelagic sediments are dominant throughout, although in some cases the influence of low-density turbidity currents and of bottom currents is observed. Sandy turbidites, up to 2 metres thick in the channels, were emplaced during periods of lowered sea-level. The passive continental margin between the Madeira and Cape Verde archipelagoes off N W Africa is one of the best studied in the world (Fig. 1). Previous work has demonstrated the presence of large-scale morphological features including giant slumps, debris-flow masses and turbidity current channels, mostly within the Neogene and Quaternary sections (Belderson & Laughton 1966; Embley 1975; Arthur & Von Rad 1979). 3 0o 29
_
/--,,
':r
2 50 II
i
/,
v
Sarnthein & Diester-Haas (1977) reported the occurrence in this area of thick, massive, fine sand layers which they identified as turbidites derived from aeolian sediments. Sarnthein et al. (1982) have given a general review of the palaeo-environmental evolution since the Early Cenozoic, while Thiede (1977) and Crowley (1981) discussed, in particular, the evolution during the past 150 000 yrs. Work on the slope and upper ,
200 ",,:\
1 50 ~
~
;..,.
=
~, r \
V 1085
20~-
-//-
~VZ
17
\'
i s
29
L E
--_./// / : ~
-.
...
"-
gFZ
~ \\'
~
\ ~ W.-~ \ \
li7///////~. ,~.*"q 20
,[.
3 0o
2 50
1 5o
FIG. 1. Regional setting of the Cape Verde Basin showing inferred debris-flows (shaded), turbidity current pathways (dashed lines), and location of cores containing aeolian turbidites (solid and open circles). Study a r e a s CV1 and CV2 shown by solid black squares.
I53
I54
G . A . A u f f r e t e t al.
rise has concentrated on aeolian versus fluvial supply (Bein & Fiitterer 1977; Diester-Haas 1979; Lutze et al. 1979; Koopman in press), the influence of upwelling (Thiede 1977; Diester-Haas 1980; Labracherie 1980) and downslope transport (Bein & Ffitterer 1977; Sarnthein & DiesterHaas 1977; Lutze 1980). Relatively little work has been devoted to the present day sedimentary regime at the lower-rise to abyssal-plain transition (i.e. at about 5000 m water depth). Lonsdale (1978) has pointed out that tidal currents with velocities up to 17 cm/s occur at depths of 4000 m, and that these may be related to areas of hyperbolic echoes. A recent sidescan sonar study of the lower continental rise in this area (Kidd & Searle, this volume) has
identified an acoustic facies characterized by lineated backscattering patterns. This they interpret as the sedimentary expression of turbidity current pathways across the rise. In this paper we present results from an investigation conducted during the SEABED cruise ( C N E X O - - C E A of N/O J. Charcot in November 1980). During this cruise we carried out a detailed seismic and sediment coring survey of an area of the deep basin identified as Cape Verde 1 (CV1). The material collected has been studied using standard lithological procedures (visual descriptions, smear slides, carbonate content determinations, X-ray diffraction). The stratigraphy has been established using nannofossil zonation (Martini 1971; Mfiller & Rothe 1975;
Age
1
2
3
4
5
6
-0.5 "/~ -2
7 .
--I
8 .
.
Ft
I
82
--ll MII
105 Ty
125 170 2~
/' -
Mi
270 o
.
~" -- -/.
M
310 32O
z z __
/ 565 575 615 675 695
FIG. 2. Stratigraphic time-scale used in this study (from Pujol 1980). Column 1 = Magnetic polarity (black = normal) From La Becque et al. (1977). 2 = Nannofossil zonation from Thierstein et al. (1977). 3--Isotop!c stratigraphy from Shackleton (1977). 4 = Biozonation for tropical region. 5 = Termination from Van Donk. 6 = Age of high stands of sea-level. 7 = F1 -- Flandrian; Ty = Tyrrhenian; Mi = Milazian; Si = Sicilian; Em = Emilian; Ca -- Calabrian. 8 = Alpine glacial stages.
Si Em
Co
?
9
Sedimentation in the southern Cape Verde Basin Thierstein et al. 1977) as well as the magnetic stratigraphic scale of La Brecque et al. (1977) (Fig. 2). In addition to this we have studied the textural properties of the sediment and its magnetic fabric in order to characterize sedimentary processes prevailing during its deposition (Auffret et al. 1981; Auffret & Sichler 1982).
r55
have been well documented by Diester-Haas (1979) and Sarnthein et al. (1982). The present-day surface circulation is characterized by the southward-flowing Canary Current. At about 15 N., most of the surface water driven by the trade winds turns west, forming the westward flowing North Equatorial Current. Intermediate water mass circulation is complex with both northward and southward components (Sarnthein et al. 1982). Between 2000 and 4000 m, the North Atlantic Deep Water flows predominantly southward, and below this the Antarctic Bottom Water flows generally northward at very low velocities. The bottom current regime has been further investigated as part of the French contribution to the SEABED program (Vangriesheim & Madelain, CNEXO, unpublished report). Results from a six-month mooring with a current meter 10 m above the bottom indicate an average northward flow of 1 cm/s, with intermittent current speeds up to I0 cm/s. Embley (1975, 1982) and Jacobi & Hayes
Regional setting and seismic facies The north-west African margin is subject to much tectonic and volcanic activity. The tectonic activity is mainly related to the release of stress along the Atlantis and Kane fracture zones, whereas the Canary and Cape Verde Islands have been the site of important volcanic activity, up to 0.3 Ma in the Canary Islands (Schmincke & Von Rad 1979) and up to the present time in the Cape Verde Islands (Dillon & Sougy 1974). Past climatic changes 25~
25 ~
24~
25 ~
25 ~ ~176 l
/ J
J
, ~
C
o~149 ~
: ~ ~176
I
9 ~176
24 ~
24 ~ 50
50 /
/ i
J
24 ~ 40 25~
25 ~
24050
I~'IG. 3. Bathymetry of study area Cape Verde 1, from narrow beam echo-sounding. Contours in metres. Axes of valleys indicated by solid and dashed lines. Dotted lines show areas of basement high. Solid lines show areas of hemipelagic-type acoustic facies. Cores indicated by solid circles.
I56
G.A. Auffret et al.
(1982) have mapped the sea floor of the Cape Verde Basin using data from 3.5 kHz records as well as conventional echo-sounders. They interpret debris-flow deposits confined to the slope and upper rise (Fig. 1) and turbidity current pathways leading from the upper slope to the lower rise. A small abyssal plain, up to 5200 m deep, is flanked to the west by the abyssal hill province. A bathymetric map of the area (Fig. 3) has been constructed using data from the vertical beam of the J. Charcot Sea-Beam, and additional control from the 3.5 kHz echo-sounder. The average slope to the west is about 1.2 ~, with a probable system of east-west trending ridges and troughs of about 5-10 m relief and 10 km separation. A two-way travel time isochron map to the top of the acoustic basement is shown in Fig. 4. This shows the approximate NE-SW orientation of basement highs related to the palaeo-rift direction. The wave length is about 8 km and the maximum vertical amplitude between crests and troughs is about 1.5 km. As the surface of the sea floor is nearly flat, this map is also practically an isopach map of the sedimentary deposits, and
indicates preferential accumulation in the troughs. A map of acoustic facies (Fig. 5) derived from the 3.5 kHz records shows the distribution of the three main types of echo-characters identified. Facies 1 is characterized by maximum signal penetration, little diffusion, and good resolution of the reflectors. This facies is interpreted as hemipelagic drape over topographic highs. Facies 2 is less coherent than facies 1, with prolonged surficial or internal echoes that appear to be related to coarse layers. Figure 6 shows the facies transition between 1 and 2. Facies 3 is characterized by reduced penetration and maximum diffusion of the signal. Unfortunately, it has not been sampled and its significance is still not clear to us.
Lithostratigraphy The following stratigraphy has been established for the six cores recovered, on the basis of nannofossil and palaeomagnetic studies (Fig. 7): (1) Upper Plioeene: (older than 1 620 000 yrs Be) Two cores KS 11 and KS 14 from topographic
25~
24~
,x .
24~
24~
25~
24~
FIG. 4. Isochron map of depth to basement. Contours in seconds two-way travel time.
Sedimentation in the southern Cape Verde Basin
157
FIG. 5. Acoustic facies distribution. 1. Hemipelagic facies (shaded); 2. Channel facies (close hachure); and 3. Not identified (open hachure). highs reached Upper Pliocene marly ooze (nannofossil zone NN 16). A hiatus corresponding to nannofossil zone NN 18 and NN 17 (?), i.e. spanning about 800 000 yrs, is present in core KS 14. (2) Lower Pleistocene: (1 620000 to 900000 yrs
BP) The carbonate contents of the sediment sampled in cores 9, I !, 12, 13 and 14 are mostly high.
There are two coarse-grained layers of sandy, marly, calcareous ooze in core 14, each about 20 to 30 cm thick. Sedimentation rates range between 4 mm/1000 yrs (KS 1I) and 6 mm/1000 yrs (KS 14), the base of nannofossil zone NN 19 could be missing in this later core. The inferred age of the first coarse layer in core 14 is about 1000000 yrs, based on the estimated sedimentation rate.
FiG. 6. 3.5 kHz record of the transition between facies 1 and 2 (xxxxx).
FIG. 7. Lithostratigraphic correlation o f six cores from areas CV 1.
Sedimentation in the southern Cape Verde Basin (3) Middle Pleistocene: (900 000 to 458 000 yrs Bp) After the Jaramillo event ( ~ 900 000 yrs Bp) the carbonate content of the sediments decreased notably and there is a rhythmic alternation of marly ooze and clayey mud. A nannofossil ooze layer 20 to 50 cm thick is present in all cores and could correspond to an age of about 600 000 yrs Bp. Coarse-grained layers 20 to 30 cm thick are present in cores KS 7, 12, 13 and 14. Sedimentation rates range between 8 and 21 ram/1000 yrs. The top of zone N N 19 may be missing in core KS 09 where no sediments younger than 458 000 yrs Bp have been recovered. (4) Upper Pleistocene: (458 000 to 268 000 yrs Br,) Rhythmic alternations of marly nannofossil ooze and silty-clayey mud occur in core KS 11. In cores 7, 12 and 13 sandy mud or sandy marly foraminiferal ooze layers are interbedded, most commonly overlying calcareous mud. The thicknesses of these detrital layers which comprise mostly fine-grained quartz sand, increase from bottom to top of the interval, and are up to 200 cm in cores 7 and 13. Sedimentation rates are highly variable, up to 18 mm/1000 yrs in KS 07. (5) Uppermost Pleistocene--Holocene: (younger than 268 000 yrs BP) The surface sediments are predominantly marly nannofossil ooze, with an interbedded layer of calcareous mud in cores 7, 12 and 13. A hiatus is present in cores 9 and 14, both from topographic highs. The maximum sedimentation rate (3 ram/1000 yrs) is observed in core 12.
Facies and processes The lithological properties of the fine-grained sediments allow us to clearly distinguish two
I59
TABLE 1. Mean and standard deviation of
lithological properties from Formations 1 and 2. The water content is given relative to the weight of the wet sample CaCO3 Formation
Sand
~ H20
Moy.
s
Moy.
s
Moy.
s
36 59
25 25
6 10
12 9
82 66
4
CVI 1
2
5
s = standard deviation. Moy. = mean formations on the basis of the average carbonate contents (Table 1). Formation 1 is dominantly Upper Pleistocene (0 to 900000 yrs BP) marly ooze and calcareous mud, and Formation 2, dominantly Lower Pleistocene marly ooze and nannofossil foraminiferal ooze. Apart from well-defined sand layers of turbiditic origin, both formations are dominantly typical hemipelagic sediments characterized by: mixed biogenic and terrigenous composition, fine grain size, low sedimentation rates, the importance of bioturbation, the scarcity of laminated intervals and low sand contents. Only a few greyish muddy intervals that overlie the turbiditic sands may be interpreted as the fine-grained parts of turbidites. We have summarized in Tables 1 and 2 the lithological characteristics of hemipelagic sediments from Formations 1 and 2 in core 7. Figures 8 and 9 illustrate the sedimentary structures characteristic of these sediments. Formation I B is characterized by the occurrence of interbedded sandy layers up to 180 cm
TABLE 2. Lithology of sediments from core 7 Levels CaCO3 CaCO3 Plant (cm) * ** Nan. For. Ind. Quartz Clay debris 110 363-364 383-384 394-395 614 635 650 660 710 723
71 62 11 57 33 80 25 10 3 40
10 40 ~ 41 20 20 15 ~
50 1
30
3
3 5 ~
11 21 11 16 10 55 10 10
10 5
10 7
19 38 84 43 67 20 74 90 87 60
* Measured. ** Estimated. Trace.
~
Colour 7.5YR 7/2 7.5YR 7/4 7.5YR 5/6 7.5YR 8/2 7.5YR 8/2 7.5YR 8/2 7.5YR 6/2 7.5YR 6/2 7.5YR 5/6 7.5YR 8/4
% sand ** 8
0 I 1 0 0 1
Nature foraminiferal ooze nannofossil marly ooze calcareous clayey mud nannofossil marly ooze nannofossil marly ooze nannofossil ooze nannofossil clayey mud clayey mud clayey mud nannofossil marly ooze
G.A. Auffret et al.
I6O
(a)
(b)
(c)
(d)
FIG. 8. (a) (90 to 115 cm): Transition between the top of the 180 cm thick sandy turbidite and overlying marly ooze, the colour contrast emphasizes the bioturbation. (b) (340 to 345 cm): greenish sandy mud. (345 to 357 cm): clayey mud with Chondrites types burrows. (357 to 364 cm): nannofossil ooze. (c) (380 to 403 cm): Interbedded and strongly burrowed marly ooze (380-390 cm), nannofossil ooze (390-397 cm) and sandy turbidite (397-400 cm). (d) (550 to 575 cm): Sheared burrows in greenish mud. The dark specks are probably related to reduced iron sulfides. thick. Textural analyses of samples from the two thickest sandy layers show them to be positively graded (Fig. 10). The sand is composed mainly of quartz (60 to 80~o), foraminifers (up to 25%) and minor pyrite, glauconite and apatite. Figure 11 shows the stratigraphic distribution of the sand layers, expressed as thickness of sand deposited during interpolated 25 000 yr intervals. We have studied the magnetic fabric of finegrained sediments from Formations 1 and 2. The orientation of the axis of maximum susceptibility has been determined for a total of 300 samples. In Fig. 12 we have summarized the distribution of these directions in each core for all levels combined. Three situations can be distinguished: (a) one prevailing direction (downslope), cores KS 12, KS 13; (b) bimodal directions, cores KS 09, KS 14; (c) bimodal, with one dominant direction, cores KS 07-KS 11.
The distribution by age of the magnetic susceptibility orientations (all cores combined) shows that a 040 ~ direction was dominant during the 1.62 to 0.7 Ma period (i.e. Formation 2 and the base of Formation 1).
Discussion We shall discuss successively: (1) the correlation between past and present sedimentary patterns, the structural setting and the present morphological setting; (2) the characteristics of the hemipelagic facies; (3) the sandy turbidites; (4) the possible origin of the uppermost and Lower Quaternary hiatuses.
Correlation between sedimentary pattern and past and present morphology It is apparent from Figs 4 and 5 that the
Sedimentation in the southern Cape Verde Basin
I6I
One major channel crosses area CV 1 from east to west. Cores 7 and 13, recovered from this channel, both contain a 180 cm thick sand layer near the surface. Cores 12 and 14 sampled about 2 miles south and north respectively of the channel contain a less thick sand layer, which may correspond to the same turbidite. Both downslope resedimentation and local small-scale sediment instability have been important slope processes.
Hemipelagic sediments
FiG. 9. (633 to 637 cm): nannofossil ooze. (637 to 643 cm): marly ooze resulting from biological mixing between nannofossil ooze and clayey mud. (643 to 650 cm): nannofossil ooze infilled burrows in marly ooze. (650 to 658 cm): composite burrows including nannofossil ooze infilled Chondrites burrows (715 to 740 cm): strongly burrowed nannofossil marly ooze. distribution of acoustic facies mirrors the morphology of the acoustic basement. This observation implies that the basement structural framework has been a major control on sedimentation. A second major control has been the regional bathymetric trend which is approximately at right angles to the basement structural direction (Figs 3 and 4).
The fine-grained hemipelagic sediments recovered exhibit evidence of intense burrowing, as well as a magnetic fabric indicative of predominant downslope transport. The relatively low accumulation rate, as well as the intense burrowing, favour deposition from normal pelagic settling, whereas the magnetic fabric suggests some influence of downslope flow. The magnetic signature of most of the finegrained sediments studied indicates predominant NW-SE downslope transport. However, there is clear evidence of a different mode (45~ that could either be interpreted as a different turbidity current path or a contour-following bottom current direction. A very clear along-slope magnetic signature is characteristic of the 1 500 000 to 900 000 yrs BP stratigraphic interval (Fig. 13). This direction corresponds mainly to calcareous mud deposited in the period preceding the Jaramillo magnetic event. The same observation has been reported from AMS measurement on sediment collected to the east of the Great Meteor Seamount (Kuijpers, pers. comm.). This may correspond to higher bottom current activity, related to a high flux of Antarctic Bottom Water, toward the end of the Matuyama reversed polarity interval. The northsouth trending direction in core 7 also corresponds mainly to the deposition of calcareous clayey mud, that may have occurred around 500000 yrs BP, under an enhanced flow of Antarctic Bottom Water. The fluctuation in carbonate content appears to be related to carbonate dissolution, rather than dilution by higher terrigenous input or lowered carbonate productivity. According to DiesterHaas (1980), higher dissolution in this area prevailed during glacial periods characterized by a more humid climate and enhanced upwelling.
Aeolian turbidites The lithological characteristics of most of the sandy layers including the thickest one are very
G.A. Au.ffret et al.
I62
%
I
0,5
0,25
'
,
,
O,125mm /
"l,,
~
~
O"
(5"
O"
I
I
I
I ~
f//
75
(5~
(ram)
///
A//
KS 13 (180cm) 5cm
y
/
/1/
47 93
135 172
155 25
I
13
250 p
!
b
loop
210p
lOOp
FIG. 10. Grain-size distributions of the thickest sand layers from cores 7 and 13. similar to the so-called 'aeolian turbidites' described by Sarnthein & Diester-Haas (1977). These are thick (100 to more than 600 cm), massive, sandy units, ungraded and composed mostly of yellowish-red quartz grains of desert origin with frosted surfaces. The two massive quartz sand layers in core V 3253 (27-145 cm and 153-325 cm) are typical. They have a medium grain-size (median 180-220 mm), are well-sorted without significant silt fraction, and ungraded. On the basis of planktonic foraminiferal abundance, the Holocene/Pleistocene boundary is situated between 15 and 20 cm depth in this core. Sarnthein & Diester-Haas (1977) identified the
source of the aeolian turbidites as the shelf, south of Cap-Blanc (Fig. 14). According to these authors, during low stands of sea-level, desert sand dunes reached the shelf break, and provided a continuous but moderate flux of sand over the upper slope through avalanching. At the onset of an interglacial period, marine transgression favoured the triggering of massive slumps. The cores containing sand turbidites of upper Pleistocene age and the inferred path of the turbidity currents are plotted in Figs 1 and 12. Although the turbidites of the CV1 area seem to be older (470 000 to 270 000 yrs BP), the same pattern of resedimentation most probably
50ram
I_
m
14 I
TOTAL ~
SAND
THICKNESS
Olmm 3 0 0 x 103 yeors
400
CLIMATstode
12
500
600
16
700
18
FIG. 11. Stratigraphic distribution of the thickness of turbidite sand layers in cores 7, 12, 13 and 14 calculated in time intervals of 25 000 years.
Sedimentation in the southern Cape Verde Basin
~.~o.i/ i
i
I63 !
I
25~
24055
+KS 14
~
24~
o
I, I
I
i
24045
I
25~
~
I
i
i
[
I
i
I
25 ~
i
I
25 ~
i"
i
~
,
I
24~
i
i
,
i
24~
FIG. 12. Distribution of the azimuth of maximum magnetic susceptibility axis. All levels combined for each
core. applied. However, it is not known whether the source of the sand was the same. It is worth noting (Fig. 7) that most of the sand layers appear to have been emplaced following the deposition of calcareous mud, the nannofossils in which show evidence of moderate to strong dissolution. A turbidite layer of older age has also been recovered in core V 3252 from the top of an active diapir which is elevated about 4 m above the surrounding sea-floor (Lancelot & Embley 1975). This layer is richer in planktonic foraminifers which could result from hydraulic sorting within the turbidity current. The effect of diapiric piercement structures on the paths of turbidity currents across the lower rise has already been emphasized by Embley (1975). He also interprets the poor development of an abyssal plain in this area as the result of the ponding of turbidity currents in depressions caused by these piercement structures (Embley & Jacobi 1977). Hiatuses
There are two levels at which hiatuses occur in
area CV 1, a superficial one observed at the tops of cores 9 and 14 (470 000 to 0 yrs ~p and 350 000 to 0 yrs BP respectively), and an internal one in core t4 spanning about 700000 yrs between 2.3 and 1.6 my BP. The uppermost one is unusual in that it affects only two areas. However, in all cores (Fig. 15), one can observe a decrease of the sedimentation rate during the Upper Quaternary. Both cores 9 and 14 are located outside the "lows' (channels) on a slope of about 1- and there appears to be no obvious morphological feature that might induce higher current velocities at these points. One possible explanation for non-deposition at these sites is that the sediments were more easily eroded, perhaps as a result of pore-water circulation (Abbott et al., in press) and/or the type and intensity of bioturbation. Erosion by turbidity currents seems unlikely as this would have affected mainly topographic lows, whereas the opposite is observed. It is possible, but we believe unlikely, that the tops of the cores were missing due to coring disturbance. The older hiatus which occurs in core 14 coincides with the beginning of the Matuyama
O- 0,268
N= 19
040%
280
24,
;," U U
0 , 4 5 8 - 0,Tx 103y N= 119 0 20%
x 103y
80
28
80
20
24
:0
IbU
0,268
-
0,458xlO3y
02o%
280
80
24
'.0 200
0,7- 1,62x 103y
N=47
020%
280
N=87
80
24(
>-0
160
FIG. 13. Distribution of the azimuth of maximum magnetic susceptibility axis for successive stratigraphic intervals. All cores combined.
CANARIES
0
s
I STADE 2
CV1 011~ ~ J,,~U N E HOLOCE[
4,'~ .3AP
/ BLANC
FIG. 14. Probable pathways for turbidity currents across the N W African margin to the study area (CVI) and adjacent areas (from Sarnthein & Diester-Hass 1977).
I65
Sedimentation in the southern Cape Verde Basin 2
Age 0
2
0
4
6
8
~ _ _ L . - - _ _ I L
I0
I
9
12 m.
I
I
I
(103 71
GO -[-
ol c~I[
Z
zl
n,"
zl
nn
500
- -
KS
7
....
KS
9
---
KS
tl
-------. . . .
KS KS KS
12 13 14
%..
9..
\ .9
\
"..\
.
N ',
"~.
"~ \
\
\
\ \
\ \
Cr~l
\
- - i
\
'\
\
\
\ \ \
\
\
i
\ <~ >-
1500
\
\ \
I--
\
\
I | I I I I
Fic. 15. Sedimentation rates of cores from area CV1.
reversed magnetic polarity interval. A hiatus of the same age has been reported by Pujol et al. (1976) in a core from the Great Meteor Seamount. This may well correspond to higher bottom current activity, perhaps related to intensification of the circum Antarctic Bottom Water current, which is known to affect bottom circulation in many parts of the oceans (Huang & Watkins 1977).
Conclusion This detailed study of a small area on the lower rise off N W Africa has shown the complexity and variability of sedimentation in an area of lowrelief open ocean. Despite the flatness of the sea floor, areas of deposition and erosion co-exist within a few kilometres of each other. Low-relief channels are filled by both fine-grained hemipela-
166
G.A. A uffret et al.
gic sediment and thick sand layers. Accurate reconnaisance of this complex interfingering of sediment facies necessitates a detailed deep-tow investigation. This complexity is related to the interplay of sedimentary processes, which include gravity-controlled hemipelagic deposition, high-
density and high-velocity turbidity currents emplaced during periods of lowered sea-level, and m u d d y contourites c o m p o s e d mostly of calcareous clayey m u d emplaced between 1 500 000 and 900 000 yrs BP and, to a lesser extent at a b o u t 500 000 yrs BP.
References ABBOTT, D.H., MENKE, W., HOBART, M., ANDERSON, R.N. & EMBLEYR.W., in press. Processes influencing slope stability in marine sediments. J. geophys. Res. ARTHUR, M.A. & YON RAD, O. 1979. Early Neogene base-of slope sediment at site 397, DSDP Leg 47A: sequential evolution of gravitative mass transport processes and redeposition along the northwest African passive margin. In: von Rad, U. & Ryan. W.B.F. et al., Init. Rept. DSDP, 47, part 1. US Govt. Print. Off., Washington, DC. 603-618. AUFFRET, G.A., SICHLER, B. & COLENO, B. 1981. Deep-sea sediments texture and magnetic fabric. indicators of bottom currents regime. Oceanologica Acta, 4, No. 4, 475-88. - - & SICHLER,B. 1982. Holocene sedimentary regime in two sites of the north-eastern Atlantic continental slope. Bull. Inst. Gdol. Bassin d'Aquitaine, Bordeaux, 31, 181-194. BE1N,A & FUTreRER, D. 1977. Texture and composition of continental shelf to rise sediments off the northwestern coast of Africa: an indication for downslope transportation. "Meteor' Forch. Erg., C. 27, 46-74. BELDERSON,R.H. & LAUGHTON,A.S. 1966. Correlation of some Atlantic turbidites. Sedimentology, 7, 103-16. CROWLEY. T.J. 1981. Temperature and circulation changes in the eastern North Atlantic during the last 150000 years: evidence from the planktonic foraminifer record. Mar. Micropalaeont., 6, 97-129. DIESTER-HAAS,L. 1979. DSDP Site 397: Climatological, sedimentological and oceanographic changes in the Neogene autochthonous sequence. In: Von Rad, U. & Ryan, W.B.F. et al., Init. Rept. DSDP, 47, part 1. US Govt. Print. Off., Washington, DC. 647-70. - - 1 9 8 0 . Upwelling and climate off northwest Africa during the Late Quaternary. Palaeoecol. Afr., 12, 229-38. DILLON, W.P. & SOUGY, J.M.A. 1974. Geology of west Africa and Canary and Cape Verde Islands. In: Nairn A.E.M. & Stehli, F.G. (eds), The Ocean Basins and Margins, 2, The North Atlantic. Plenum Press, New York, London. 315-90. EMBLEY,R.W. 1975. Studies of Deep-Sea Sedimentation Processes Using High Resolution Seismic Data. Ph.D. Thesis, Columbia University. New York. 334 PP. & JACOBI, R.D. 1977. Exotic Middle Miocene sediment from Cape Verde Rise and its relation to piercement structure. Am. Ass. Petrol. Geol. Geol. Notes, 2004-9. 1982 Anatomy of some Atlantic margin sediment slides and some comments on ages and mechanisms. -
-
-
-
In: Saxov, S. & Niewenhuis, J.K. (eds), Marine Slides and other Mass-Movements. Plenum. Press, New York and London. 189-213. HUANG, T.C., & WATKINS, N.C. 1977. Contrasts between the Bruhnes and Matuyama sedimentary records of bottom water activity in the South Pacific. Marine Geol., 23, 113-32. JACOBI, R.P. & HAVES,D.E. (1982). Bathymetry, Microphysiography, And Reflectivity Characteristics of the West African Continental Margin. Sierra Leone to Mauritania. In: yon Rad, U. et al. (eds), Geology of the Northwest African Continental Margin. Springer-Verlag, Berlin. 182-212. KOOPMAN, B. (in press). Sedimentation von Saharastaub in subtropischen Atlantik w/ihrend der letzten 25 000 J~ihre. 'Meteor'Forsch. Erg., C. 35. LABRACHERIE~ M. 1980. Les radiolaires tbmoins de l'6volution hydrologique depuis le dernier maximum glaciaire au large de Cap Blanc (Afrique du Nord-Ouest). Palaeogeo. Palaeoclimat. Palaeoecol., 32, 163-84. LA BRECQUE,D., KENT, V. & CANDE,S.C. 1977. Revised magnetic polarity time scale for late Cretaceous and Cenozoic time. Geology, 5, 330-5. LANCELOT, Y. & EMBLEV, R.W. 1977. Piercement structures in deep oceans. Bull. Am. Ass. Petrol. Geol., 61, 1191-2000. LONSDALE,P. 1978. Bedforms and the benthic boundary layer in the North Atlantic: A cruise report of indomed Leg 11. SIO Reference 78-30. LUTZE, G.F. 1980. Depth distribution of benthonic foraminifera on the continental margin off NW Africa. 'Meteor' Forsch. Erg., C. 33, 31-80. - - , SARNTHEIN,M., KOOPMAN, B., PFLAUMANN,V., ERLENKENSER,H. & THIEDE, J. 1979. Meteor cores 12309: Late Pleistocene reference section for interpretation of the Neogene of site 397. In: Von Rad, U., Ryan, W.B.F. et al., Init. Rept. DSDP., 47, part l, US Govt Print. Off., Washington, DC. 727-39. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. 2nd Plankt. Conf. Proc. Rome 1970., 2, 739-85. MOLLER, C. & ROTHE, P. 1975. Nannoplankton contents in regard to petrological properties of deep-sea sediments in the Canary and Cape Verde areas. Marine Geol., 19, 259-73. PUJOL, C. 1980. Les foraminif~res planctoniques de l'Atlantique Nord au Quaternaire. Ecologie, Stratigraphie, Environnement. Mem. Inst. Gbol. Bassin d'Aquitaine, Bordeaux, 10, 254 p. --, DUPRAT, J. & PUJOS-LAMY, A. 1976. R~sultats pr61iminaires de l'&ude effectuOe par l'Institut de G~ologie du Bassin d'Aquitaine concernant la mis-
Sedimentation in the southern Cape Verde Basin sion MIDLANTE A dans l'Atlantique oriental entre 51 ~ N. et 35 ~ N. Bull. Inst. Gdol. Bassin Aquitaine, Bordeaux, 19, 3-32. SARNTHE1N,M. & DIESTER-HAAS,L. 1977. Aeolian sand turbidites. J. sed. Petrol., 47, 2, 868-90. --, THIEDE, J., PFLAUMANN,U., ERLENKEUSER,H., FUTTERER, D., KOOPMAN,B., LANGE, H. & SEIBOLD, E. 1982. Atmospheric and oceanic circulation patterns offnorthwest Africa during the past 25 million years. In: von Rad, U. et al (eds) Geology of the Northwest African Continental Margin. SpringerVerlag, Berlin, Heidelberg. 545-604. SHACKLETON,N.J. 1977. The oxygen isotope stratigraphy record of the late Pleistocene. Phil. Trans. Roy. Soc. Lond., 13, 280, 169--82. SCHMINCKE, H.V. & VON RAD, U. 1979. Neogene
I67
evolution of Canary Island volcanism inferred from ash layers and volcaniclastic sandstones of DSDP site 397 (Leg 47 A)./n: Von Rad, U., Ryan, W.B.F. et al., Init. Rept. DSDP, 47, part 1: US Govt. Print. Off. Washington, DC. 703-25. TmEDE, J. 1977. Aspects of the variability of the glacial and interglacial North Atlantic eastern boundary current (last 150 000 years). 'Meteor' Forsch Erg. C. 28, 1-36. THIERSTEIN, H.R., GEITZENHAUER,K.R., MOLFINO, B. SHACKLETON, N.J. 1977. Global synchroneity of late Quaternary coccolith datum levels: validation by oxygen isotopes. Geology, 5, V.S. 400-404. VANGRIESHEIM, A. & MADELAIN, F. 1982. Rapport contrat C.E.A. TMC.'14575 fitsc. 2, Oc&mographk' physique. (unpublished).
G.A. AUFFRET, R. LE SUAVE,R. KERBRAT& B. SICHLER,Centre Oc6anologique de Bretagne, B.P. 337, 29273 Brest, Cedex, France. S.RoY, C.LAJ, CNRS-CEA Centre des Faibles Radioactivit6s, B.P. l, 91190, Gif-sur-Yvette, France. C.MULLER, Consultante, 1 rue Martignon, 92500, Rueil Ma/maison, France. Contribution 824 du Centre Oc6anologique de Bretagne.
Sedimentary processes on the west Hellenic Arc margin
H. Got S U M M ARY: High-resolution seismic surveys across the western Hellenic Arc show that the
Plio-Quaternary cover is highly discontinuous along most of the margin. The slope displays a series of basins separated by ridges generally devoid of sediment. Sedimentological analyses of the recent sediment permit the origin and processes of sedimentation to be defined. The sediments provided by a specific area on land are successively trapped, released by the slope basins and transferred via channels and canyons towards the deeper trench basins. Several main provinces can be distinguished. Each of them includes one or more slope basins and trenches without either lateral connection or the mixing of sediment inputs. The diversity and complexity of sedimentary structures, the granulometric characters, and the variability in rates of sedimentation argue for the prevalence of mass gravity processes. The transfer of sediment from the upper slope basins to the outer trench is effected by an alternation of short periods of giant mudflows and regular but slight reworking between slope basins. This cascade-feeding process, characteristic of the Hellenic subduction margin, does not appear to prevent terrigenous material from reaching the deep trenches. We compare this briefly with sedimentation on passive margins in the western Mediterranean Basin.
This paper is a synthesis of earlier work that describes the distribution and nature of unconsolidated sediments covering the Ionian margin of Peloponnesus, between the islands of Zakinthos and Crete (Fig. 1), and attempts to identify sedimentation mechanisms. This margin represents the outer part of an active margin, which formed as a result of the convergence between the European and African plates in the region of the Ionian Sea and the Eastern Mediterranean basins. The data discussed here have been collected during four cruises (1974, 1976, 1977 and 1978) and include 4000 km of high-resolution seismic profiles, using a 3000-J sparker (IUN-Naples) and a 40 in 3 air-gun (Station de Geodynamique sous-marine, VillefranGhe), and 30 piston cores from the Peloponnesus margin and adjacent trenches. Previous studies of this area have primarily focused both on the. structure of the arc, including structural studies of the margin and Ionian trench (Le Quellec et al. 1978; Le Quellec & Mascle 1979; Le Quellec et al. 1980), and on its patchy sedimentary cover. Wong & Zarudzki (1969) and Ryan et al. (1971) have studied the general deposition and thickness of recent sediments, Ryan (1972) and Hinz (1974), their age and Pastouret (1970), Blanc & Blanc-Vernet (1971), Chamley (1971), Emelyanov (1972), Ryan (1972), Stanley & Knight (1978), Stanley et al. (1978), Vittori et al. (1981), Got et al. (1981), Blanpied & Stanley (1981), their nature.
Morphology and structural configuration The margin of the western Hellenic Arc is morphologically very complex, comprising a series of basins and terraces on the slope and isolated segments of a discontinuous trench at the base of the slope (Fig. l, Carter et al. 1972). Two major structural trends are evident, one at 140~ and the other at 70~176 These two trends divide the margin into numerous tectonically-controlled sedimentary units. From north to south the main basins and terraces are: (l) the Cephalonia-Zakinthos plateau (Hinz 1974) wedged between the Peloponnesus slope and Strophades Islands eastern slopes; (2) the Messenia and Laconia basins, separated by the Cape Matapan foreland; (3) and numerous smaller basins and sedimentary terraces primarily affecting the Kithira-Antikithira margin where a 10~ to 2 0 ~ trend, related to alpine tectonics is locally present (e.g. Cape Matapan). From north-west to south-east the trench segments are: The Zakinthos Trench (Got et al. 1977) and the North and the South Matapan Trenches. These trench basins, in particular the South Matapan Trench, are subdivided into a series of ponds ranging from 4200 to 4600 m in depth, and separated by ridges related to the base of the slope (Le Quellec et al. 1978; Vittori 1978; Le Pichon et al. 1979). A recent detailed seismic survey of the South Matapan Trench slope (Got 1981) together with a precision (Sea-Beam) bathymetric survey (Le
I69
I7o
H. Got
0
50km
PELOPONNESUS
AEGEAN
%
.,2o
SMT CRETE
Fig. 1. Chart of study area showing the core locations and DSDP sites 126, 127, 128. ZT = Zakinthos Trench; NMT =North Matapan Trench; SMT = South Matapan Trench; KB= Kiparissia Bay; MB = Messenia Bay; LB = Laconia Bay. Bathymetry in metres. Pichon et al. 1979) enable the relationship between slope basins and trenches to be more accurately defined, and it appears that the basins and trenches are connected by a network of tectonically-controlled channels and canyons which are generally devoid of sediments.
Stratigraphy Most of the seismic air-gun and sparker profiles show two major acoustic units (Fig. 2): a lower acoustic unit and a stratified sedimentary cover. These two units display different acoustic characteristics in different areas.
Deep stratigraphic unit South-west of the trench, along the Mediterranean Ridge, the lower acoustic unit is a severalhundred-milliseconds thick stratified sequence
(Fig. 2). Previous seismic (Wong & Zarudzki 1969: Hinz 1974) and sedimentological studies (Biscaye et al. 1972) relate this unit to the top of the Messinian (M Reflector). This series was cored on the Mediterranean Ridge during DSDP legs 13 and 42A at sites 125 and 374, respectively; the sediments recovered at site 126 contain Messinian evaporitic series (Ryan et al. 1973) and middle Miocene sediments. Furthermore, a submersible study of the Hellenic Trench (Le Pichon et al. 1981) has revealed outcrops of probably evaporitic nature attributed to the Messinian. The penetration obtained with our seismic systems suggests such an evaporite sequence. Locally, particularly in the zone of cobblestone morphology, this lower unit loses its stratified character. The transition between these two characters is gradational and can be best explained by active deformations affecting the whole sedimentary sequence.
Sedimentary processes on the west Hellenic' Arc margin
~7I
0
0
~9
o o
O
8
g
8
0
.~
9
,.~
t",lr,~
I72
H. Got
The lower unit along the inner wall of the trench is generally considered as an acoustic basement composed of alpine consolidated rocks. In slope basins, between the acoustic basement and the unconsolidated sedimentary cover, there are a series of parallel high-amplitude reflections, called unit C. This unit can be tentatively attributed to Miocene sediments trapped in subsiding basins. Upper stratigraphic units
The overlying sedimentary section includes two acoustically distinct units with different geometrical distributions. The lower acoustic unit (termed B) is characterized by weak internal reflections with poor lateral continuity. For this reason, the unit is referred to as a 'transparent layer'. However, near coastal areas, and particularly near areas of sedimentary input, there are more internal low and medium amplitude reflectors, although in areas of thin sediment cover, the series conserves its weakly reflective character. This unit generally overlies the deep units (S, M or C); the unconformity between the basement and unit B is related to a Miocene-Pliocene erosional surface, although on the Mediterranean Ridge, the two units are concordant. Structures which are interpreted as normal faults, commonly linked to faulted underlying basement, are observed within this weaklyreflective unit. In contrask the upper sedimentary unit (termed A) is characterized by parallel, highamplitude reflectors. Typically, these lie horizontally, and offsets of the order of 100 ms occur on the Zakinthos terrace, across vertical faults. Based on the seismic reflection profiles of the Mediterranean Sea, the results of the DSDP drilling at sites 125, 126, 127 and 128 (Ryan et al. 1973) and the structural configuration of the unconsolidated cover on land (Sorel 1976; Angelier 1976; Kelletat & Schroeder 1976: Dufaure 1977; Mercier et al. 1978) the acoustic stratigraphy can be dated as follows: Unit
Seismic character
Lithology
Age
Unit A
internally reflective unit weakly reflective unit unit
mud, sand, ash mud, marls
Quaternary
Unit B
Pliocene
Sediments Distribution
Mapping of the sediment thickness reveals continuity of fill from basins to trenches. The margin comprises a series of slope basins connected by narrow channels generally devoid of sediment.
These connected basins determine sedimentary units extending from a specific source terrain on land to a trench basin. Three main units can be distinguished (Fig. 3): Kiparissia basin to North Matapan Trench; Messinia and Laconia basins to northern basin of the South Matapan Trench; Kithira Strait to southern basin of the South Matapan Trench. In each of these connected basin systems the sediments contain lenticular layers, slump scars, gravitational slides, tensional scars, perturbed morphology and compressionally contorted beds (Fig. 4), all of which indicate that a large proportion of sediments have been reworked by mass gravity processes. Based on sedimentological and 3.5 kHz data, a ~giant' mud flow emplacement mechanism has been proposed by Stanley & Knight (1978) in the Hellenic Trench system. The overall nature of this sedimentation is referred to as cascade feeding (Got et al. 1981; Vittori et al. 1981) in which the continental slope both traps and/or serves as a conduct for sediments. The tectonic influence is clearly marked, not only in the erosion of the sedimentary units but also in the development of the valley and canyon network that funnels the sediments through the slope basins to the trenches. In the basin-valley system across the Kithira margin (Fig. 5), the deep-stratigraphic unit or acoustic basement (S) is overlain by a thick and relatively transparent upper unit. This upper unit contains a thin stratified series sandwiched between the more acoustically-transparent layers. This series is present in the successive slope basins from 1000 to 3000 m depth and can be identified as a key horizon of probably sandy composition. However, we note a change in the acoustic character of the stratified series. In the shallower basin, it is well marked by strong reflectors, whereas downslope its character tends to attenuate and its thickness increases. This evolution can be interpreted as a dilution of the reflective unit with the progressive reworking of sediments from basin to basin. General lithology
The study area is mainly located in a pelagic domain, and biogenic muds are predominant. However, within this superficial cover, two more distinctive lithotypes have been recognized: sapropels and ash layers. The sapropels were formed during 3 major episodes, at 7900 yrs, 23-25000 yrs and 30-38 000 yrs m~ (Nesteroff 1973; Stanley et al. 1978). The carbon content increases from the most recent layer (1.2-1.8%) to the most ancient (2.3-3.8%). These layers are present in about 50%
S e d i m e n t a r y p r o c e s s e s on the w e s t H e l l e n i c A r c m a r g i n
I73
FIG. 3. Reconstruction of the provenance and processes of sedimentation on the south-western Peloponnesus margin. This map shows: (a) the major sedimentary basins on the slope as delineated by the 200 ms isopach (after Vittori et al. 1981); (b) the main mineralogical provinces based on heavy and clay minerals; (c) the origin of inputs and the main axes of sediment dispersal (arrows). UCZB = Upper Cephalonia-Zakinthos Basin; LLB = Lower Laconia Basin; LCZB = Lower Cephalonia-Zakinthos Basin; KB = Kithira Basins; MB = Messenia Basin; WCB = Western Cretan Basin; of the cores, mainly on the continental margin. Their thickness varies from several centimetres in core 3 to 3 m in core 6 (Fig. 6). In addition to their high organic-carbon content, these sapropelic layers are characterized by mainly pyritic mineralization. The problems of their origin is not yet fully solved (Sigl et al. 1978). Preliminary results, based on element analysis, carbon content, and infrared spectrometry suggest that the organic matter is essentially of marine origin (Vittori et al. 1978). Two main ash layers have been recognized in
the Eastern Mediterranean: the Lower Tephra and the Upper Tephra, respectively dated at 25 000 and 3400 yrs BP (Ninkovitch & Heezen 1965). In cores from the south Peloponnesus margin, only the Lower Tephra occurs. The geographic distribution of the Lower Tephra is highly variable (Fig. 7); it is generally present on the slope, where its thickness is usually about 1 cm, but it can also occur as a succession of several centimetric layers. In channels (core 14, Fig. 6), ash is dispersed over more than 30 cm. This lithotype is rarely found in the trenches.
FIG. 4. Slumps and slump scars affecting the upper sedimentary unit in the Messenia Basin.
Sedimentary processes on the west Hellenic Arc margin
I7 5
FIG. 5. Line drawing of a seismic reflection profile across the Kithira margin showing the widespread correlation of the acoustic refectors. The arrows indicate the approximate core locations. In addition to these distinctive horizons, the biogenic muds show several variations (Fig. 6). Chromatic changes from yellow to grey are common and well marked. However, except for the black layers that result from the organic content, colour changes do not seem to be correlated with mineralogic or textural variations. The ash layers and biogenic accumulations form the main coarse layers. However, some cores retrieved in channels (core 14) or at channelmouths (core 11 l) show accumulation of detrital elements sometimes with graded bedding. We have also noted cross-bedding (core 3a), slumping affecting pieces of sapropels covered by horizontal deposits (cores 17 and 18), sharp contacts between muddy layers (core 6), mud and sapropel clasts, and changes of bed inclination in the same core (core 6).
Texture and composition The cores have been analysed for grain-size ( < 40 /~m), carbonate content, organic carbon, clay
minerals, and light and heavy minerals. Grainsize and mineralogy provide the most significant data. The percentage of the >40 /tm fraction is generally low. There is a slight decrease in the whole coarse fraction (biogenic and terrigenous) from the coastal gulfs to the trench basins. However, if the terrigenous content is considered alone, there appears to be no significant change in the percentage with depth (Fig. 8). One of the trench cores (111) contains more than 60~ terrigenous material. Similarly, the median grain-size does not change consistently with depth but is distributed according to the physiographic domain; the highest values (6-8/am) are encountered in transport zones (channels and canyons) and the lowest (1-2 /~n) are found in trenches. However, if we consider the evolution of the median grain-size in a sedimentary system (Fig. 9) such as the Laconia system (cores 14-12-11) or the Kithyra margin (cores 18-19), the values decrease slightly with depth. But it is noteworthy that the main pheno-
~76
H. Got
FIG. 6. Sections of cores showing the main lithological and structural features. The length is expressed in centimetres; f= foraminifera concentrations; s = sapropel layer; arrows indicate features discussed in text. menon is improved sorting with depth rather than a reduction of size. The cumulative grain-size curves (Fig. 10) are presented in a semi-logarithmic diagram. This permits the mode of deposition to be defined on the basis of the form of the curve (Riviere 1952). No sediment of the Peloponnesus margin, including trenches, shows a hyperbolic grain-size distribution characteristic of pelagic settling. On the
contrary, the form of this curve is generally parabolic, indicative of dominant gravity transport mechanisms. Clay minerals are dominated by illite (40-50%) followed by smectite and chlorite-kaolinite (30%); this association is characteristic of the Ionian Basin (Chamley 1971). From the quantitative standpoint, two main areas can be distinguished: (1) from Zakinthos to Cape Matapan,
Sedimentary processes on the west Hellenic Arc margin
I77
o
o
~9
0
0
0 0
2~
0
~2
o'~
o
o 2
I78
H. Got 9 Gulf and t r a n s i t zones
1oo2
9 slope basins
E9
9 Trenches
11'
q4
2O 9
12
50
36
4
6
9
5 9 18 9
nm
32
3
9
32
"19 9 9 35 34 9 9 9
37
9
9
22
2
3
4
5 x 1000m. Depth FIG. 8. Diagram showing the random variation of the detrital terrigenous content of the sediments versus depth. where the values of illite are around 50%; and (2) the Kithira area which is characterized by the highest values of smectite (30~I~) of probable Aegean origin (Venkatarathnam & Ryan 197 l). The clay-mineral assemblages cannot be correlated with either facies or colour; only the sapropelic layers show a decrease of smectite but this correlation has a diagenetic origin (Chamley 1971). Likewise, there are no systematic differences in clay mineralogy as a function of depth. ~5-
14
v
11
~4z _J oe"
o3 r
1'0
2'0
3b
20
AGE (X 1000 years)
FIG. 9. Evolution of the mean size of the pelitic fraction of sediments versus depth in two sedimentary units. Cores 14-12-11: Laconia Basin--Northern Basin of South Matapan Trench; Cores 18-19: Kithira Ridge--Southern Basin of the South Matapan Trench. Full lines: trenches. Dashed lines: Slopes.
On the basis of the heavy-mineral frequency distribution in the studied samples, four assemblages have been identified. These contain the following mineral species, listed in decreasing order of abundance: (1) garnet, spinel, and epidote (cores 8, 3, 1); (2) chloritoid, garnet, epidote, and blue amphiboles (glaucophane and crossite) (cores 4, 6); (3) epidote, chloritoid, and blue amphiboles (cores 17, 12, 14, 112); (4) magnesite, epidote, chloritoid, green hornblende (cores 19, 110. These associations, together with the clay mineral assemblages outlined above, allow us to define the three mineralogical provinces shown in Fig. 3. (a) Province 1 comprises Kiparissia Bay, CephaIonia-Zakinthos plateau, and North Matapan Trench. It is characterized by garnet, spinel and illite derived from the Alphee basin (Kiparissia Bay). (b) Province 2 includes Messenia and Laconia basins and the South Matapan Trench. It is defined by an association of chloritoid, epidote, glaucophane, illite, smectite and chlorite and the main source area is the Eurotas Basin (Messenia and Laconia Bays). (c) Province 3, coinciding with the Kithira area, is marked by an association ofchloritoid, epidote, glaucophane, hornblende, magnesite and smectite; this assemblage occurs in the slope basins of the Kithira Margin and in the eastern segment of the South Matapan Trench. The authigenic magnesite is probably linked, as well as the smectite, to the exchange of water masses across the Kithira Ridge (Lacombe & Tchernia 1972; Stanley & Perissoratis 1977).
I79
Sedimentary processes on the west Hellenic Arc margin 100~
/ < .<-"" ~< 1B/../..../.'/'y'~
, ~,o~
/" 9
/,/~~
~
..-"
o.~
..'"
...9 7
r
/
/
o.
.."~/l_q
/
/7 /
,
/
/ ......'"j " 9 I" ~ . / ..... ~ mO Q
.........""
I
I
I
!
I
I
I
J I I
I
s
J
~
,
~
J
Jl
I
1
I
io ]Jm
FIG. 10. Grain-size cumulative frequency graphs of selected sediments from the three physiographic provinces: gulfs (core 14), slope basins (cores 3, 12, 18) and trench (core 19) (see location Fig. 1). These mineralogical results support the model of sediment dispersal deduced from seismic data (Vittori et al. 1981). Each isolated segment of the trench is connected with a specific supply area: the Zakinthos and North Matapan Trenches are fed from the Kiparissia zone (Alphee basin), the South Matapan Trench from the Messenia and Laconia Gulfs, and the eastern segment of the southern trench basin from the Kithira Ridge. Furthermore, when we compare cores collected from the same sedimentary system (Fig. 11) we note that the ratio of biogenic to terrigeneous material remains approximately constant from one basin to the next. This implies a relatively continuous resedimenting process 9 Thickness and rates of sedimentation
The existence and thickness of a complete Late Quaternary sequence and the frequency and type of sedimentary structures are closely linked to the morphostructural setting (Fig. 7). The oldest horizons are exposed in the midslope area where pre-Pliocene sediments may outcrop as a result of the combined action of erosion and tectonics. On the other hand, the most complete sections occur
in the slope-basins. The trenches record the greatest thickness of Holocene sediments associated with increased homogeneity of the deposits. Considering the entire margin, the rates of sedimentation are highly variable. In the different sedimentary basins, however, the rates are comparable, averaging 10 cm/1000 yrs. This value, based on C ~4 analyses of cores, corresponds well to the results of seismic studies which show that the thickness of the unconsolidated cover is roughly the same in each of the slope and trench basins. However, there appears to have been a recent pause in sedimentation (Fig. 12) due to the migration of the maximum depocenter as a result of tectonic changes (Vittori et al. 1981). The basins probably show a similar history of long periods of active deposition alternating with periods of relative dormancy with little or no deposition 9
Processes and discussion Using a combination of seismic, lithoseismic, lithologic and sedimentologic data we have been
H. Got
I8o
MA I/+
MA
12
100%
IO0 %
0 I
I
I I
]
]"
I
I
I
I
\
i
.-"~'.~
.....
\
,
L
. __~~ - ~
\
I
I
7-9000
y. BP
---,-...~.
/\ /
)
' \
\
/
\
lm
/ \
I
,\
) \
/ \
2m
"
,
/
'
,
I \
,
I
\ 12 O O O
, ,
/
f
\
\ 25 0 0 0
x;. BP
!
\
6
'i
+
1
I
5b
/ \ ~
FI
,
'
TERR I G E N E O U S
I )
I
BIOCLASTIC
(0- 100%)
( 0 - 50%)
.........
,
\
I
,
/ ,
4m
V. BP
/
,
3m
/
\
r-N
/
/
/
)
\
/
/
/ ' \..
)5--~ ~ 1 0 N
\
I
,
/ ",,
~ i
i
,+
23600
X'. BP
9
50 FIG. 11. Correlation of two cores from two interconnected slope basins, based on the terrigenous versus biogenic content, showing good correlation of the main horizons.
able to infer the style and processes of sedimentation on this active margin (Vittori 1978; Got 1981; Got et al. 1981; Vittori et al. 1981). The west Hellenic Arc margin is highly discontinuous in its physiography as well as in the distribution of the unconsolidated Pliocene and Quaternary sedimentary cover. This morphology and sediment distribution pattern is controlled by two major structural trends related to the evolution of the arc.
Each separate segment of the Matapan Trench is connected to a different sediment supply area; this connection can be followed up-slope from basin to basin along tectonically-controlled channels. The slope-basins act as temporary traps and contain the thickest and most complete sedimentary sequences. The diversity and frequency of sedimentary structures, the variations in rates of sedimentation, the absence of marked sorting and the
S e d i m e n t a r y processes on the west Hellenic A r c margin
ISI
from reaching the trench as shown elsewhere (Horn et al. 1972; Scholl 1974; Scholl & Marlow v8 1974). This is certainly the result of the alternaW N tion of periods of active reworking and periods of relative dormancy. These processes of sedimentation result in poor sorting of the dominant W yfine-grained sediment and, when the environmental conditions are favourable, in the deposition of 'unifites' as described by Blanpied & Stanley (1981). The question we now pose is whether such a model of sedimentation and the resulting charac2 ter and distribution of deposits are characteristic of this type of compressional setting, or whether o they are more widespread? In fact, in the western \ Mediterranean Basin, an example of passive I OI I i 9 margin (Monaco & Mear, this volume), the 2 O0 4000 sedimentary cover is also discontinuous. Each DEPTH (in m.) separate physiographic area comprises isolated units of thick accumulation, continental-shelf FIG. 12. Sedimentation rate curves for different and pro-delta accumulations, slope and slide basins on the Kithira margin (see location Figs 1 masses, base of slope-basin and deep-sea fan and 5). lobes, interspersed with areas of relatively less accumulation. There are periods of higher and lower sedimentation rates and periods of nongranulometric facies types, all strongly support deposition which result in sequences of hemipelathe dominance of mass gravity flow processes of gites and reworked deposits (turbidites, unifites, sedimentation. This mechanism of sedimentation contourites). Moreover, linked sedimentary units can be termed a 'cascading system'. The sedi- transverse to the margin have been recognized for ments are initially deposited in the upper slope each deltaic source system of the western Medibasin; they are then transferred from the upper to terranean Basin. the lower slope by processes of spill-over and However, in spite of these similarities between channelization through a network of valleys and the two types of margins there are also differcanyons which are now filled to differing degrees. ences, particularly with respect to the distribution Thus, the slope environment plays a triple role, of the unconsolidated sedimentary cover. On the being a zone where: (1) sediments are trapped in east Mediterranean active margin the cover is tectonically-controlled slope basins; (2) sedi- highly discontinuous and, generally, the highs ments are reworked under the influence of seismic and slopes are devoid of sediments and the instability; (3) material is transferred towards the acoustic basement outcrops. The different sedibasins and trenches by a tectonically-controlled mentary basins are well isolated by tectonic physiographic network. Moreover, the NW-SE trends. On the passive margins of the western sense of sediment transport reflects the tectonic Mediterranean, outcrops of consolidated prestresses rather than the hydrodynamic regime. Pliocene series are scarce and the Quaternary The sedimentary system is therefore character- cover is more or less continuous, although variized by a spasmodic supply which results from an able in thickness and nature. alternation of trapping and transfer. Two The second difference regards the relative mechanisms act in the transfer processes: quantity of sediment types. On the passive mar(1) a mass reworking during short periods of gin, the ratio of hemipelagic to turbiditic seditime related to the seismic instability of the area. ments is between 3:10 and 4: 10, whereas on the This involves thick sequences of sediment, gener- eastern active margin it is between 1 : 10 and 2:10. ally located at the basin margins, and results in These differences are the result of different sedithe denudation of the slopes and highs. mentary mechanisms. On both margins, mass (2) a permanent but slight removal of the depo- gravity processes and channelling are important, sits during longer periods of time which enables but on the passive margin they are induced the key horizons to be conserved across the primarily by eustatic effects and secondarily by margin. tectonics, whereas on the active margin, seismic The cascade-feeding system of the Hellenic Arc instability related to the tectonic evolution of the does not seem to prevent the terrigeneous detritus arc is the main control on resedimentation. E
~
m
- :~<~.r9
1 82
H. Got
ACKNOWLEDGMENTS: The data presented have been collected during several cruises of the N/O Trident (University of R h o d e Island), Dectra (Instituto Universitario Naval of Naples), Marsili (Italian C N R ) and C o m m a n d a n t e Giobbe; thanks are expressed to the officers, crew and staff of these o c e a n o g r a p h i c vessels. We also thank D.J. Stanley (Smithsonian Institution of Wash-
ington), L. Mirabile (I.U.N. of Naples), A. Brambati (O.G.S. of Trieste) and G. Kelling (University of Keele) for their technical and scientific help. F u n d i n g for this ongoing investigation was provided by the Centre N a t i o n a l de la Recherche Scientifique, the Centre N a t i o n a l pour l'Exploitation des Oceans and the Italian Foreign Office.
References reflection measurements in the Ionian Sea. Geol. Jahh., 2, 33-65. HORN, D.R., EW1NG, J.I. & EWING, M. 1972. Graded bed sequences emplaced by turbidity currents north of 20 N in the Pacific, Atlantic and Mediterranean. Sedinwntolog.v, 18, 247-75. KELLETAT, D. & SCFIR6DER,B. 1976. Vertical displacement of Quaternary shorelines in the Peloponnesus. Bull. Soc. gbol. Fr. (7), XVIII, 2, 382. LE PICHON, X., ANGEL1ER,J. et al. 1979. From subduction to transform motion: a seabeam survey of the Hellenic Trench System. Earth planet. Sci. Lett., 44, 441-50. LE PICHON,X., HUCHON,P., ANGELIER,J., LYBERIS,N., BOULIN, J., BUREAU,D., CADET,J.P., DERCOURT,J., GLACON, G., GOT, H., KARIG, D., MASCLE, J., RICOU, L.E. & THIEBAULT,F. 1981. Active tectonics in the Hellenic Trench. Oceanologica Acta, no. SP, Geology of Continental Margins, 273-81. LE QUELLEC, P. & MASCLE, J. 1979. Hypoth~se sur l'origine des Monts Matapan (marge ionienne du P61oponn6se). C.R. Acad. Sci. Paris, 228, 31-4. 1-23. CHAMLEY, H. 1971. Recherches Sur La Sbdimentation - - , MASCLE,J., VITTORI,J., GOT, H. & MIRAB1LE,L. 1978. La fosse de Matapan (Mer Ionienne): nouArgileuse en Mbditerranbe. Thesis, Univ. Aix-Marvelles donn6es sur sa structure. C.R. Acad. Sci. seille. 401 pp. Paris, 287, 431-34. DUFAURE, J.J. 1977. N6otectonique et morphogen6se dans une p6ninsule m6diterran6enne: le P61o- - - - , MASCLE, J., V1TTORI, J. & GOT, H. 1980. Seismic structure of the Southern Peloponnesus conponn6se. Revue. Gdogr. phys. Gdol. dyn., XIX (1), tinental margin. Bull. Am. Assoc. Petrol. Geol., 64, 27-58. EMELYANOV, E.M. 1972. Principal types of recent 242-63. bottom sediments in the Mediterranean Sea: Their NESTEROEF,W.D. 1973. Petrography and mineralogy of mineralogy and geochemistry. In: Stanley, D.J. sapropels. In: Ryan, W.B.F., Hsu, K.J., et al., Init. (ed.), The Mediterranean Sea - A Natural SedimenRept. DSDP, 13 (1), US Govt. Print. Office, Washtation Laboratory. Dowden, Hutchinson & Ross, ington, DC. 713-20. Stroudsburg, Pa. 355-86. N1NKOVICH,D. & HEEZEN,B.C. 1965. Santorini Tephra. GOT, H., STANLEY, D.J. & SOREL,D. 1977. NorthwesIn: Whittard, X.F. & Bradshaw, R.D. (eds), Subtern Hellenic Arc: concurrent sedimentation and marine Geology and Geophysics. Colston Papers, 17, deformation in a compressive setting. Marine Geol., 413-52. Butterworth, London. 24, 21-36. PASTOURET, L. 1970. Etude s6dimentologique et 1981. M6canismes d'alimentation en cascade et pala6oclimatique de carottes pr61ev6es en M6diters6dimentation des fosses de l'Arc Hell+nique. In: ran6e Orientale. Tethys, 2, (1), 227-66. Wezel, F.C., (ed.), Sediment Basins of Mediter- RIVlERE, A. 1952. Expression analytique g6nSrale de la ranean Margins. Tecnoprint, Bologna. granulom6trie des s6diments meubles. Notion de facies granulom6trique. Bull. Soc. gbol. Fr., 6 (II), - - , MONACO, A., VITTORI, J., BRAMBATI,A., CATANI, 155-67. G., MASOLI,M., PUGLIESE,N., ZUCCHI-STOLFA,M., BELFIORE, A., GALLO, F., MEZZADRI, G., VERNIA, RYAN, W.B.F., STANLEY,D.J. et al. 1971. The tectonics and geology of the Mediterranean Sea. In: Maxwell, L., VINCI, A. & BONADUCE,G. 1981. Sedimentation A.E. (ed.), The Sea, 4, Wiley-Interscience, New in the Ionian active margin (Hellenic Arc); Provenance of inputs and mechanisms of deposition. York. 387-492. - 1972. Stratigraphy of late Quaternary sediments in Sed. Geol., 28, 243-72. the eastern Mediterranean. In: Stanley, D.J. (ed.), HINZ, K. 1974. Results of seismic refraction and seismic
ANGELIER, E. 1976. La n6otectonique cassante et sa place darts un arc insulaire: l'arc 6g6en m6ridional. Bull. Soc. gbol. Fr., (7), XVIII, 5, 1257-65. BISCAYE,P., RYAN,W.B.F. & WEZEL, F. 1972. Age and Nature of the Pan-Mediterranean subbottom reflector M. In: Stanley D.J. (ed.), The Mediterranean Sea - A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 83-90. BLANC, J . J . & BLANC-VERNET,L. 1971. Sur la pr6sence d'affleurements plioquaternaires immerg6s fi l'ouest de File de C6rigo (M~diterran6e nord-orientale). C.R. Acad. Sci. Paris, 267, (D), 271-73. BLANPIED, C. & STANLEY, D.J. 1961. Uniform mud (unifite) deposition in the Hellenic Trench, Eastern Mediterranean. Smith. Contrib. Earth Sci., 13, 1-38. CARTER, T.G., FLANAGAN, I.P. et al. 1972. A new bathymetric chart and physiography of the Mediterranean Sea. In: Stanley, D.J., (ed.) The Mediterranean S e a - A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa.
Sedimentary processes on the west Hellenic Arc margin The Mediterranean S e a - A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 149-69. --, Hsu, K.J. et al. 1973. Cleft in Mediterranean ridge, Ionian sea-site 126. In: Ryan, W.B.F., Hsu, K.J. et al. Init. Rept. DSDP, 13 (1), US Govt. Print. Office, Washington, DC. 219--41. SCHOLL, D.W. 1974. Sedimentary sequences in the North Pacific Trenches. In: Burk, C.A. & Drake, C.K. (eds), The Geology o f Continental Margins. Springer, New York. 493-504. SCHOLL, D.W. & MARLOW, M.S. 1974. Sedimentary sequence in modern Pacific trenches and the deformed Pacific eugeosyncline. In: Dott, R.H. & Shaver, R.H. (eds), Modern and Ancient Geosynclinal Sedimentation. Soc. econ. Palaeo. Min. Spec. Pub., 19. SIGL, W., CHAMLEY, H., FABR1CIUS,F., GIRAUD D'ARGOUD, G. & MULLER, J. 1978. Sedimentology and environmental conditions of sapropels. In: Hsu, K.J., Montadert, L., et al. [nit. Rept. DSDP, 42(1). US Govt. Print. Off. 13 (1), Washington, DC. 445-64. SOREL, D. 1976. Etude Nbotectonique dans l'Arc Egben Externe Occidental: les lles [on&nnes de Keffalinia et Zakinthos et l'Elide. Thesis, Univ, Orsay. 200 pp. STANLEY, D.J. & PERISSORATIS, C. 1977. Aegean sea ridge barrier and basin sedimentation patterns. Marine Geol., 24, 97-107.
---,
183
KNIGHT, R.S., STUCKENRATH,R. & CATANI,J.P. 1978. High sedimentation rates and variable dispersal patterns in the Western Hellenic Trench. Nature, 273, 110-12. & KNIGHT, J. 1978. Giant mudflow deposits in submarine trenches, Hellenic basins and slopes in eastern Mediterranean. Bull. Am. Ass. Petrol. Geol. Bull., 63, 532-3. -& MALDONADO, A. 1981. Depositional Models of fine-grained sediments in the Western Hellenic Trench, eastern Mediterranean. Sedimentology, 28, 273-90. VITTORI,J. 1978. Caract&'es Structuro-S~dimentaires de la Courerture Plio-Quaternaire au Niveau des Pentes el des Fosses Hell~,niques du Pdloponm;se (Grace). Th6se Sp6cialit6, Universit6 Paul Sabatier, Toulouse. 194 p. VITTORI, J., GOT, H., LE QUELLEC, P., MASCLE, J. & MIRABILE, k. 1981. Emplacement of the recent sedimentary cover and processes of deposition on the Matapan Trench Margin (Hellenic Arc). Marine Geol., 41, 113-35. VENKATARATHNAM,K. & RYAN, W.B.F. 1971. Dispersal patterns of clay minerals in the sediments of the Eastern Mediterranean Sea. Marine Geol., l l , 261-83. WONG, H.K. & ZARUDZKI, E.F.K. 1969. Thickness of unconsolidated sediments in the eastern Mediterranean Sea. Bull. geol. Soc. Am., 80, 2611-14. -
-
H. GOT, Laboratoire de Sedimentologie et G6ochimie Marines, Universit6 de Perpignan, Avenue de ViUeneuve, 66025, Perpignan, France.
Fine-grained turbidites and associated mass-flow deposits in the Ulleung (Tsushima) Back-arc Basin, East Sea (Sea of Japan) S.K. Chough S U M M A R Y : Acoustically laminated Quaternary deposits (up to 800 m thick) dominated
by turbidites overlie transparent hemipelagic sediments (Pliocene and older) in the Ulleung (Tsushima) Basin. The topmost layer in sediment cores is post-glacial hemipelagic sediment, about 2 m thick, underlain largely by a repetitive sequence of fine-grained, thinly- or un-laminated turbidites. Coarse-grained, graded and plane-laminated units rarely occur. Turbidites in the Ulleung Basin were accumulated at a rate of about 12 cm/103 years. The turbidites in the centre of the Ulleung Basin correspond to slump, slide and debris-flow deposits on the slope and base-of-slope environments. It suggests that the mass-flow processes occur successively from a broad, linear source on the prograding shelf break. Sediments originated from the adjoining margins of the Korean Peninsula and north western Honshu. Turbidites deposited during the major volcanic eruptions were derived from the Oki Spur and San-in coast, whereas those deposited between the volcanic events were mainly from the stable craton of the eastern margin of the Korean Peninsula. Submarine mass-flow processes such as slump, slide, grain-flow, debris-flow and turbidity currents are important in transporting large volumes of terrigenous sediments into the deep ocean basins. These processes have been active through geologic time both in active and passive margins (Embley 1976; Jacobi 1976; Nardin et al. 1979), where slopes are usually incised by submarine canyons which guide mass flows. Mass-flow processes also play an important role in seismically active back-arc basins. In the Ulleung (Tsushima) Back-arc Basin of the East Sea (Sea of Japan) (Fig. 1), the slope lacks submarine canyons and is characterized only by subdued hills on the base-of-slope. Yet the basin is filled by an accumulation of up to 800 m thick, Quaternary turbidites which predominate over hemipelagic sediment, which has probably accumulated since Pliocene time (Tamaki et al. 1978). These turbidites correspond to the existence of other mass-flow deposits on the slope. Although the basin became inactive a few million years ago (Uyeda 1979), the physiographic setting, i.e bounded by a seismically active margin on the east (Japan arc) and a rifted steep margin on the west (Korean margin), causes the occurrence of extensive mass-flow deposits. Submarine mass-flow processes in the Ulleung Basin have probably been more common during glacial periods when sea-level was lower, transporting terrigenous debris directly to the outer continental shelf via the adjoining rivers and streams. The present study reports on the occurrence and characteristics of turbidites and associated mass-flow deposits in the Ulleung Basin in an attempt to interpret the depositional processes
that operate in an inactive back-arc basin. The study is based primarily on the analysis of four piston cores (Fig. 2) and 3.5 kHz profiles (Fig. 3) obtained aboard R/V Hakurei-Maru of the Geological Survey of Japan (Honza 1978).
Geological setting According to Uyeda & Miyashiro (1974), the East Sea (Sea of Japan) was rifted as a result of collision and subduction of the hypothetical Kula-Pacific Ridge from late Cretaceous to Oligocene. Uyeda & Kanamori (1979) and Uyeda (1979) propose that the sea was opened by a tensional force, common in Mariana-type subduction, caused by a small-scale convection in the asthenosphere wedge located above the subducting lithosphere. Earlier, Bersenev (1971), Karig (1971), Hilde & Wageman (1973) and Ludwig et al. (1975) suggested that the sea opened from north-west to south-east, forming the Japan Basin first, followed by the formation of the Yamato and Ulleung Basins. Based on the prominent magnetic lineations found in the north-western coast of Japan, Honza (1978, 1979) assumed a clockwise radial spreading of the area with an axis of spreading near the southern coast of Korea. The age of oceanic basement, determined by an extrapolation of sediment accumulation rate in the DSDP site 299 in the Japan Basin, indicates late Oligocene to early Miocene (25 to 30 my, BP) for opening (Karig, Ing|e et al. 1975). This age agrees also with the initiation of recent subsidence of the Japanese Island Arc associated with the IN 5
S.K. Chough
I86
S I ~B E R I A . . .
.
9-
b,(~
./// JAPAN BASIN
~
2ooo" ..
E
A
S
T
::" . . ~,I
(JAPAN)
C
"
ULLEUNG ] BASIN /
KOREA
,J ;.Y
~
~ooO / ~ sPuR .~_.;y. ~ - 200m / ~ i " : hi
..,.~
.i
.
.
.
.
.
.
1~0 o
- ....... . . . . . .
1:350
.
"
:"":
~ ...
:.
. 9.
140 ~
FIG. l. Major physiographic features of the East Sea (Sea of Japan). I.G. = Inter-plain Gap.
subduction in the Japan Trench (Langseth et al. 1981). High heat-flow of up to 3 HFU (Watanabe et al. 1977) on the sea floor supports this timing of back-arc opening.
Physiography The Ulleung Basin is generally U-shaped in cross-section with a gentle slope to the north-east (Fig. 4). The base-of-slope also is transitional to the basin floor with a number of very subdued hills or fans. The basin is bordered by a steep continental slope of the Korean Peninsula and Korea Plateau on the west and north, respectively. The continental shelf on the east coast of Korea is narrow, not more than 25 km wide, and transient to the steep slope (slope gradient, cc. 6~ In the south and east, the basin is surrounded by the rather gentle slope of Oki Bank, the submarine extension of the Honshu shelf off Shimane
Peninsula and the slope of San-in district of south-western Honshu (Fig. 4). The basin extends through the gap between Ulleung and Dok Islands to the north-east, forming a narrow and long inter-basin plain between the Korea Plateau, Kita-Yamato and Yamato Ridges. It debouches onto the deeper Japan Basin near the southern end of KitaYamato Ridge. An inter-plain channel, Ulleung Inter-plain Channel, runs intermittently through the Ulleung Inter-plain Gap approximately along the long axis of the basin connecting the Ulleung Basin to the Japan Basin (Fig. 5). The channel is approximately 7 km wide and 35 m deep. It is erosional in origin and undercuts the turbidite sequence, and is most likely formed by strong bottom currents flowing through the gap. A small seamount (Ulleung Seamount) rises about 700 m above the basin floor in the basin approximately 105 km south-east of Ulleung Island. The continental slopes of the eastern Korean
Fine-grained turbidites and associated mass-flow deposits
187
FIG. 2. Location of core and grab samples taken from Ulleung Basin and the adjoining margins; P: piston core (R/V Hakurei-maru); RC: rock corer; others are grab samples from the shelves, rivers (R) and streams (S). Contours in metres. Peninsula and the Korea Plateau are steep, characterized by large-scale slump pits and scars. At the base-of-slope they are characterized by a morphology of slump and possible debris-flow deposits. Various scarp features were also found on the gentle slope at the foot of the margin offthe Oki Bank area (Ishibashi & Honza 1978).
Acoustic stratigraphy The thickness of the total sedimentary sequence (assumed velocity of 2 km/sec) in the East Sea (Sea of Japan) has been compiled recently by Gnibidenko (1979). The sequence is generally thick in the deep basins ranging up to 2.5 seconds and thin in the intervening ridges, rises and toward margins. It is also thick in some local basins along the margin. According to Tamaki et al. (1978) and Honza (1979) the sedimentary sequences in the Ulleung Basin may be divided acoustically into three units
(Fig. 6). The upper unit is an approximately 0.3 seconds thick, non-deformed, stratified Quaternary deposit with a high frequency reflection. Underlying slightly disconformably is the middle unit (Pliocene and early Pleistocene in age) which is also stratified and slightly deformed. Near the base of slope off San-in coast the middle unit attains a maximum thickness of 1.4 seconds. The thickness increases southward, continuing into the slope sequences. The lower unit of presumably Miocene age is acoustically transparent attaining thicknesses of more than 1.0 second.
Late Quaternary stratigraphy Abundant ash layers in the Ulleung Basin cores provide a tool for the correlation of individual turbidite and hemipelgic layers. The composition of each layer and the time of eruption have been detailed by Arai et al. (1981). Those identified in the Ulleung Basin cores include the Ulleung ash, composed primarily of colourless pumiceous
I88
S.K. Chough
FIG. 3. Ship tracks of the seismic survey in the Ulleung Basin; also shown is the facies distribution of mass-flow deposits. glass, that was erupted from Ulleung Island at about 9300 yrs BP and deposited in the top hemipelagic layer at a depth of about 180-190 cm (Fig. 7 and Arai et al. 1981). Another prominent ash layer is the Aira-Tn ash erupted from the Aira Caldera in southern Kyushu at about 21 000-22 000 yrs BP. This vitric ash consists mainly of bubble-walled and pumiceous glass, which is correlated in the basin occurring at depth of between 300 and 345 cm. A C 14 age determined at the depth of 360-364 cm (core P104; Fig. 7) gives an age of 18000+640 yrs, which is slightly younger than that postulated by the ash correlations. Other ash layers underlying the Aira-Tn ash include those of probable Yamato ash (25 000---35 000 yrs BP) from the Ulleung Island and Aso-4 ash from Aso Caldera (50 000 yrs Br,; Arai et al. 1981). In core P106, a number of ash-layers whose individual age is not certain were found below the latter ash layer. In addition to the ash particles pelagically settled, there are a
few ash layers composed mainly of ash debris that was resedimented by turbidity currents. The correlation of ash-layers and lithologic changes in the cores yield a sedimentation rate of 12 cm/103 yrs for the pre-Holocene turbidites. The average thickness of the individual turbidite layers is less than about 6 mm and this gives a minimum recurrence time of about 50 yrs, assuming that each individual layer was deposited in a relatively short period (days or months) of time.
Fine-grained turbidites Since about Pliocene time the accumulation of turbidites has been predominant over pelagic sedimentation in the deep basins of the sea (Fig. 6). Submarine mass-flow processes have probably been more common during glacial periods when sea-level was lower, transporting terrigenous debris directly to the outer continental shelf via the adjoining rivers and streams.
Fine-grained turbidites and associated mass-flow deposits
189
FIG. 4. Detailed bathymetric chart of the eastern margin of Korean Peninsula and Ulleung Basin; part of the entire map of the East Sea (Sea of Japan) by US Naval Oceanographic Office, Pacific Support Group (1969). Contours in fathoms.
Turbidite facies in the Ulleung Basin cores are characterized by finely-laminated and homogeneous mud with a minor occurrence of massive or graded gravel and sand layers, micro-crosslaminated sands and silts (Fig. 8). Hemipelagic mud occurs mainly in the uppermost 2 m of the cores, whereas turbidites are predominant below. Some fine-grained turbidite layers consist also of abundant biogenic sediments such as diatoms, silicoflagellates, spicules and radiolarians and others, similar to those of the hemipelagic facies.
Coarse-grained deposits, layers of sand and gravel, were found in the Ulleung Inter-plain Gap. Layers are up to 30 cm thick and show distinct graded or massive units bounded at the base by a scour surface and overlain by parallellaminated sand division. Gravels are up to 12 mm in long diameter and include large amounts of resedimented volcanic debris. The parallel or crudely parallel-laminated sand divisions (Tb) occur usually overlying the A-division. They are composed of terrigenous, volcanogenic and, in
I9o
S.K. Chough
FIG. 5. Seismic reflection profile (line 27, Fig 3) across the Ulleung Inter-plain Gap showing the erosional inter-plain channel (Ulleung Inter-plain Channel) incised by strong bottom currents: the distance between vertical lines is approximately 6 km; the thickness between horizontal lines is approximately 75 m. Profile courtesy of E. Honza, Geological Survey of Japan.
FIG. 6. Line drawings of seismic reflection profiles across the East Sea and Korea Strait. Modified after Honza (1979).
Fine-grained turbidites and associated mass-flow deposits m o
PI03
PI04
PI06
PI05 ....
1
I l . . . . . .
L
1
3 ,&&
!
[91
Another type of homogeneous unit consists of olive-grey (5Y5/2, 5Y5/1) mud alternating with the slightly olive-grey (5Y3/2) mud (Fig. 8(b), (c)). In the latter, the sediments consist mostly of biogenic material rich in diatoms and are largely bioturbated and contain abundant pyritized filaments. This olive-grey mud contains more than 15% terrigenous materials. A convolute mud unit also occurs and consists of slumped and sheared intervals of about 20 cm in thickness.
7 -4
&&&
5 6
8
LEGEND
~ (hemi) Pelagite ~"I ll Parall. lamin, Tephra turbidite
Sand Sand E~ Gravel
FIG. 7. Core descriptions (Ulleung Basin); asterisk for the depth of a C-14 age (18 000+640 yrs By). The top tephra (ash) layer is called the Ulleung ash (Oki ash) erupted from the Ulleung Island at about 9300 yrs BY; the second layer is the Aira-Tn ash erupted from the Aira Caldera in southern Kyushu at about 21 00022 000 yrs By; the third layer is probably either the Yamato ash (25 000-35 000 yrs ~v) or the Aso-4 ash from Aso Caldera (50 000 yrs BY) (Arai et al. 1981). Numbers indicate the depth points of subsampled turbidite and hemipelagic layers. some cases, foraminiferal sands. The micro-ripple cross-laminated and climbing micro-ripple laminated sand and silt (Fig. 8(a)) occur rarely, in layers exceeding 1 cm in thickness. Parallel-laminated mud (Td) is the dominant unit in the turbidites of the Ulleung Basin. It consists of olive-grey and grey-olive (5Y5/2-5Y 3/2) mud. Although it occasionally overlies the micro-ripple cross-laminated division, the unit commonly repeats itself. The individual depositional units are ill-defined, rarely exceeding 1 cm in thickness (Fig. 8(a)). The top of the parallellaminated unit is, in some cases, overlain by homogeneous mud (Te). In X-radiographs this latter division is devoid of parallel laminae. The thickness of each homogeneous mud unit ranges up to 4 cm and consists of terrigenous as well as biogenic ooze components.
Hemipelagic sediments The Ulleung Basin is presently covered with hemipelagic sediments rich in diatoms and finegrained terrigenous materials with minor amounts of silicoflagellates, spicules and others. Sediments on the Ulleung Basin slope are relatively rich in organic carbon of up to 1.5%. They are bioturbated largely by benthonic deposit feeders and contain abundant pyritized filaments of similar origin (Fig. 8(c)). In the sediment column, materials retain the evidence of near-bottom current activity. Other than the pyritized filaments, burrows rarely occur in the pelagic sediments and in the topmost sections of each turbidite layer. Chondrite and rind burrows were commonly found. In many sections the sediments are completely mottled.
Mass-flow deposits Closely-spaced high-resolution (3.5 kHz) seismic profiles obtained in the Ulleung Basin (Honza 1978) reveal various types of mass-flow deposits on the slope, base-of-slope and basin floor (Chough & Jeong, in prep.). Mass-flow deposits include slump, slide, grain- and debris-flow deposits and turbidites with corresponding slide scarps and scars (Fig. 9). These types of deposits have recently been recognized on many continental slopes and related environments both in active and inactive margins (Embley 1976; Jacobi 1976; Nardin et al. 1979). A variety of mobility is possible in mass-flows from short-distance rotational slumps to mobile debris-flows. They are recognized on the high-resolution profiles by the character of internal stratification and surface morphology (Fig. 10). Submarine slides are generally recognized on the profiles by minimal translation of massive slide blocks that are internally less deformed than the slumps and retain the original bedding. Slump deposits with minimal translation are seismically recognized by contorted internal beds or by discontinuous internal reflectors and/or blocky
I92
S.K. Chough
FIG. 8. X-radiographs of sediment cores from the Ulleung Basin (a), core P103, core depth, 279-299 cm; (b), core P106, core depth 655-675 cm; (c), core P103, core depth, 359-379 cm. Note in (a) micro-ripple cross-laminated (Tc) fine sand and silt (in the middle part) division is under- and over-lain by the repetitive sequence of parallel-laminated silt (Td). Burrow-rich hemipelagic mud layer in (b) and (c) alternates with homogeneous mud layer of turbidite origin. Large grains on the right edge in (a) are evaporite grains precipitated after the core was cut for X-radiography. For core locations see Fig. 2. Bar is 2 cm. surface m o r p h o l o g y (Fig. 10). W h e n the bedding is not always present the distinction between the two becomes difficult. The appearance of the displaced material on a seismic record also depends on the degree of lithification of the material. If the material is unconsolidated mud, even a short displacement can disrupt the continuity of internal reflectors sufficiently to make the material appear acoustically transparent.
Debris-flow deposits usually appear transparent due to poor sorting or lack of distinctive internal structure (Fig. 10), similar to transparent hemipelagic or pelagic sediments. However, the former are usually lens-shaped and internally discontinuous (Fig. 10). The surface m o r p h o l o g y and reflectivity of mass-flow deposits vary from h u m m o c k y to relatively smooth. In some cases, a relatively distinct echo is present whereas, in
FIG. 9. Air-gun profile across the eastern margin (western slope of Ulleung Basin) of the Korean Peninsula showing the slump scars on the slope. Figure courtesy of Korea Institute of Energy and Resources.
FIG. 10. 3.5 kHz profile of the south-eastern slope of the Ulleung Basin showing the slump-slide (s) and the debris flow deposits (d) (line 22 in Fig. 3). Also note the prograding sequence on the shelf. Vertical scale: two-way travel time in seconds; the thickness between horizontal lines is approximately 75 m; the distance between vertical lines is about 6 km. Note that the slump deposits are characterized by blocky surface, contorted internal beds or by discontinuous internal reflectors. Debris-flow deposits are recognized by the lack of reflectors, and by the acoustically transparent lens-shaped deposits. Profile courtesy of E. Honza, Geological Survey of Japan.
I94
S.K. Chough tation in the back-arc area in terms of origin, volume and rate of sediment supply. Heavy mineral assemblages, on the continental shelf of Korea, are dominated by green hornblende and stable minerals of metamorphic origin such as garnet, epidote, sillimanite, staurolite and actinolite-tremolite. On the Oki Spur and San-in coast of Honshu green hornblende is also predominant but is associated with unstable minerals such as ortho- and clino-pyroxene and basaltic hornblende with minor amounts of ferrohastingsite, epidote, brown hornblende and olivine. Basaltic hornblende (up to 47~o), ferrohastingsite (up to 26~/o) and olivine (up to 6~o) are the most diagnostic minerals in this region. The heavy mineral assemblage in the Uileung Basin is dominated by green hornblende with various mixtures of stable and unstable minerals in different turbidite layers. Some layers contain either the assemblages similar to those on the western margin or that on the other margin. Other layers contain a mineral assemblage common to both margins. This was further demonstrated by using Q-mode factor analysis (Bahk 1982). In Fig. 11 the samples from the east coast and continental shelf of Korea (represented by Factor I) are shown in good contrast to those from the San-in coast and Oki Spur (represented by Factor II). Turbidite layers (such as P103-1, 2, 3, P104-3, 4, 5, P105-6) were derived mainly from the stable craton of the east Korea continental margin and were deposited during the period between the volcanic activity. Turbidites such as P103-4, P105-3, 5, 7 and most of Pl06 fall near the Factor II axis, suggestive of a derivation from the
others, a diffuse or prolonged echo is observed. The irregularity produced during the flow or blocks of semilithified sediments have been observed on the surface of many sonar records (Embley 1976). Turbidites are acoustically distinctive due to the internal layers or bedding (Fig. 5). The beds are emphasized by an alteration of coarse and fine sediments. Mass-flow deposits in the Ulleung Basin and its slope are zoned along the contour in that the slump and slide deposits occur mainly on the slope which are, in turn, replaced in the deeper area below slope by debris-flow deposits (Fig. 3). The latter are transitional to the turbidites in the central basin. This zoned facies relationship is compelling evidence that mass-flows in the UIleung Basin were generated from a line source.
Provenance The provenance ofturbidites in the Ulleung Basin was deciphered in detail by Chough et al. (1981), Bahk (1982) and Bahk & Chough (1983). According to them, relative amounts of heavy minerals found on the margin of the basin compare with those in the turbidite layers in the cores from the basin. This was possible because the basin is bordered by the continental margin of eastern Korea, Korea Plateau and Oki Spur and north-western Honshu, in which rocks are of different assemblage and hence represent different detrital mineral provinces. This sheds light on the relative importance of tectonics and sedimenu
1.0
LEGEND
eo,e
9 Sonin Coast and Oki Spur 0 East Coast and Continental Shelf (Korea) 9 Ulleung Basin
\
3-4
l t,
9
I ./ /
/ / /
9
0.5
,,,~ 95-7
/ J
FIG. 11. Factor loadings for the heavy mineral data: Factor I, heavily weighted on samples from the east coast and continental shelf of Korea; Factor II is due to contributions from the Oki Spur and San-in coast. The first number of Ulleung Basin samples is the last digit of core, e.g. PI03 =3; the second number is for core depth shown in Figs 2 and 7. After Bahk (1982).
S-5 -A.
/
/
9 5-3
9
/ 9
t
I
9
4-1
.-
/
/
.~..+.- "~
/
0
/
C
A
i
. . . .
~-~4-3 ? 9 3-1 As-6 9 0
A3-3
o A
L
O~ ~ -,~ 0 oo
o n
0
0
0
/ /
.,4-5 ~-j_
I
0.5
.
,
I
o J
o
~176176 o i
i.o I
Fine-grained turbidites and associated mass-flow deposits San-in coast and Oki Spur. P103-4 and P105-7, deposited during the period close to the volcanic activity originated from the Oki Spur and San-in coast. Other turbidite layers (third group) such as P104-2, P105-4 belong to the region of either margin.
Sedimentary processes The presently inactive Ulleung back-arc basin is underlain by acoustically laminated, up to 800 m thick, Quaternary turbidites. The turbidites are dominated by a repetitive sequence of finegrained parallel-laminated and homogeneous mud units resulting mostly from low-density, dilute turbidity currents of low-flow regime conditions. At times, large-scale highly-concentrated turbidity currents were generated and resulted in the deposition of the graded, massive and crosslaminated units. Turbidity currents were generated from the adjoining steep slope of Korean
I95
margin and seismically active Honshu margin. Turbidity currents were not channelized, but most likely moved in the form of sheet flow with a line source in each event. Each turbidity current flow was generated most probably through a submarine slump, then debris-flow which then developed into turbidity current, forming a zonal facies distribution. Sediments were carried to the shelf edge by prograding deltaic processes (Fig. 12). Mass-flow processes in the basin have been operative since about late Pliocene through preHolocene time. As sea-level rose in the Holocene and submerged the continental shell turbidity current activity played a minor role in the basin except in local areas such as the Ulleung Interplain Gap. Here strong bottom currents also have incised the previously deposited turbidites. The turbidites in pre-Holocene times have accumulated at a rate of 12 cm/1000 yrs with a minimum recurrence time of about 50 yrs. Turbidites originated both from the stable craton of the
FIG. 12. Air-gun profile across the eastern shelf of Korean Peninsula off Buku, showing the prograding Neogene and Quaternary sequence. For location see Fig. 2. Figure courtesy of Korea Institute of Energy and Resources.
I96
S.K. Chough
eastern K o r e a n Peninsula and from the active arc of north-western H o n s h u . Hemipelagic sediments c o m p o s e d mainly of diatoms and silicoflagellates d o m i n a t e in the H o l o c e n e period. Calcareous sediments are deficient, due to the high oxygen content in b o t t o m water resulting in a very shallow calcite c o m p e n s a t i o n depth ( C C D ) of a b o u t 2000 m. ACKNOWLEDGEMENTS: The research was sup-
ported by the K o r e a Science and Engineering Foundation (KOSEF) and KOSEF-JSPS Exchange P r o g r a m m e . T h a n k s are due to E. H o n z a , E. Inoue, and A. N i s h i m u r a of Geological Survey of J a p a n for providing help and support. K.S. Bahk and G . H . Lee assisted in analysing the cores. K.S. Jeong analysed the 3.5 kHz seismic profiles. H.J. Kim, Instructional Media Center, did the drafting. All help mentioned is gratefully acknowledged.
References ARM, F., OBA, T., KITAZATO, H., HORIBE, Y. & MACHIDA,H. 1981. Late Quaternary tephrochronology and palaeo-oceanography of the sediments of the Japan Sea. Quat. Res., 20, 209-30. BANK, K.S. 1982. Provenance of Turbidites in the Ulleung Back-arc Basin, East Sea. Unpub. M.S. Thesis, Seoul National Univ., Seoul, Korea. 77 p. BAHK, K.S. & CHOUGH, S.K., (1983) Provenance of turbidites in the Ulleung (Tsushima) back-arc basin, East Sea (Sea of Japan). J. sed. Petrol., 53, 1331-6. BERSENEV,I.I. 1971. The origin of the Japan Sea basin. In: Island Arc and Marginal Sea, Tokai Univ. Press, Tokyo. 31-38. CHOUGH, S.K. & JEONG, K.S. (in prep.) Zoned facies of mass flow deposits in the Ulleung back-arc basin. CHOUGH, S.K., TAMAKI, K., BAHK, K.S., 1NOUE, E. & YUASA, M. 1981. Heavy minerals from the Oki Spur, Japan Sea. Bull. Geol. Surv. Japan, 32, 487-501. EMBLEY, R.W. 1976. New evidence for occurrence of debris flow deposits in the deep sea. Geology, 4, 371-4. GNIBIDENKO, H. 1979. The tectonics of the Japan Sea. Marine Geol., 32, 71-87. HILDE, T.W.C. & WAGEMAN,J.M. 1973. Structure and origin of the Japan Sea. In: Coleman, P.J. (ed.), The Western Pacific. Crane, Russak & Co. New York, and Univ. Western Australia Press. 675 pp. HONZA, E. 1978. Geological investigation of the Japan Sea. Geol. Surv. Japan, Cruise Rept., 8, 89-93. HONZA, E. 1979. Sediments, structure and spreading of Japan Sea. Nihonkai, 10, 23-45. ISHIBASHI, K. & HONZA, E. 1978. Bathymetric survey. In: Honza, E. (ed.), Geological investigations in the northern margin of the Okinawa Trough and the western margin of the Japan Sea. Geol. Sun'. Japan, Cruise Rept., 10, 12-14. JACOm, R.D. 1976. Sediment slides on the northwestern continental margin of Africa. Marine Geol., 22, 157-73. KARIG, D.E. 1971. Origin and development of marginal
basins in the western Pacific. J. geophys. Res., 76, 2542-61. KARIG, D.E., INGLE, J.C. Jr. et al. 1975. Init. Repts. DSDP., 31. US Govt. Print. Off. Washington, DC, 927 pp. LANGSETH,M.G., HUENE,R.V., NASU, N. & OKADA,H. 1981. Subsidence of the Japan trench forearc region of northern Honshu. Geology ()fi Continental Margins, Oceanologica A eta, 26th International Geological Congress, 173-9. LUDWIG, W.J., MURAUCHI, S. & HOUTZ, R.E. 1975. Sediments and structure of the Japan Sea. Bull. geol. Soc. Am., 86, 651-64. NARDIN, T.R., EDWARDS,B.O. & GORSLINE,O.S. 1979. Santa Cruz Basin, California borderland: dominance of slope processes in basin sedimentation. Soc. econ. Palaeo. Min. Spee. Pub., 27, 209-21. TAMAKI, K., MURAKAMI, F. & HONZA, E. 1978. Continuous seismic reflection profiling survey. In: Honza, E. (ed.), Geological investigations in the northern margin of the Okinawa Trough and the western margin of the Japan Sea. Geol. Sun,. Japan. Cruise Rept., 10, 39-42. U.S. NAVALOCEANOGRAPHICOFFICE, PACIFICSUPPORT GROUP, 1969. Bathymetric Chart of the Sea of Japan. Scale 1 : 2 000 000. UVEDA, S. 1979. Subduction zones: Facts, ideas and speculations. Oceanus, 22, 53-62. UYEDA,S. & KANAMORLH. 1979. Back-arc opening and the mode of subduction. J. geophys. Res., 84, 1049-61.
UVEDA, S. & MIYASHmO, A. 1974. Plate tectonics and the Japanese island. A synthesis. Bull. geol. Soc. Am., 85, 1159-70. WATANABE, T., LANGSETH, M.G. & ANDERSON, R.N. 1977. Heat flow in back-arc basins of the western Pacific: In: Talwani, M. & Pitman III, W.C. (eds:) Island Arcs, Deep Sea Trenches and Back-Arc Basins. Maurice Ewing Ser., l, 137-61.
S.K. CHOUGH,Department of Oceanography, Seoul National University, Seoul 151, Korea.
Open-ocean, off-bank transport of fine-grained carbonate sediment in the Northern Bahamas K.C. Heath and H.T. Mullins SUMMARY: The contribution of fine-grained, bank-derived versus pelagic-derived carbonate sediment in periplatform oozes has been evaluated by means of a detailed, quantitative carbonate mineralogy study of 35 surface sediment samples from north of Little Bahama Bank. Our data indicate that large volumes of aragonite and magnesian calcite are being exported along the open-ocean margin of Little Bahama Bank during the present high-stand of sea-level. Percentages of mud-size aragonite+magnesian calcite display a strong negative correlation (r = -0.912) with distance from the edge of the bank, whereas percentages of calcite show a strong positive correlation (+0.914). Immediately adjacent to the northern, windward margin of Little Bahama Bank, 86~ of the fine-grained carbonates are bank-derived. This value decreases in a linear fashion to 50~ bank-derived muds at a distance of 50 km from Little Bahama Bank. Direct seaward projection of our data suggests that fine-grained, metastable, shallow-water carbonate sediments are transported laterally at least 120 km from open-ocean bank margins. This has important implications for the global budget of CO2, the diagenetic potential of periplatform oozes, and the growth potential of carbonate slopes. Fine-grained carbonate sediments on modern, tropical, shallow-water platforms consist almost entirely of aragonite and magnesian calcite ( > 51'o MgCO3) with only minor amounts of calcite (Milliman 1974; Bathurst 1975). In such shallow, sunlit seas, fine-grained, metastable carbonate sediments are organically produced, primarily by calcareous algae (Neumann & Land 1975). In contrast, fine-grained pelagic carbonate sediments of the deep-sea consist almost entirely of calcite with only minor amounts of aragonite (Garrison 1981). These carbonate sediments are also produced organically; calcite by planktonic foraminifera and coccolithophorids, and aragonite by pteropods. The only deep-sea environments where aragonite and/or magnesian calcite are suspected of being precipitated inorganically as primary grains are the warm, hypersaline environments of the Mediterranean and Red Seas (Milliman et al. 1969; Milliman & Mi.iller 1973). Thus, in modern, open oceans there are two mineralogically distinct types of fine-grained carbonate sediments. However, around the margins of shallow-water carbonate platforms these two mineralogically unique sources of carbonate sediment mix to create yet a third type of fine-grained carbonate sediment, referred to as 'periplatform' ooze (Schlager & James 1978). Although we have been aware of periplatform ooze sedimentation in deep-water Bahamian troughs for some time (Pilkey & Rucker 1966; Rucker 1968; Kier & Pilkey 1971; Lynts et al. 1973), there have been very few studies that have attempted to quantify the contribution of bankderived versus pelagic-derived fine-grained sediment. Boardman (1978) first approached this
problem from a geochemical perspective for Holocene sediments in North-west Providence Channel, Bahamas. He concluded that more than 8000 of the total mud fraction is bank-derived, and that pteropods, the only known pelagic source of aragonite, do not contribute significantly to the fine-grained fraction. Thus, the two major sources of fine-grained periplatform oozes can readily be distinguished, simply on the basis of carbonate mineralogy. The very high percentages (80 + %) of bank-derived fines in North-west Providence Channel surface sediments may at first seem surprising, unless one considers the tremendous volumes of sediment being produced by calcareous algae on tropical shallow-water platforms. Neumann & Land's (1975) budget study demonstrated that calcareous algae alone are not only capable of producing all fine-grained carbonate sediments in the Bight of Abaco, a typical Bahamian lagoon, but that there is an overproduction of carbonate sediment by a factor of 1.5 to 3 times. Boardman's (1978) study convincingly showed that much of this excess shallow-water produced sediment is deposited along adjacent deep-water slopes and basins. However, Boardman's (1978) study was conducted on samples collected from North-west Providence Channel, an interplatform trough which receives fine-grained shallow-water carbonate sediment from both Little and Great Bahama Banks. Thus, it is still not known how far laterally fine-grained shallow-water carbonate sediments can be transported along open-ocean settings before being sedimented, and what compositional changes occur with increasing distances from their source. I99
K.C. Heath and H.T. Mullins
200
To help answer such questions, we have conducted a detailed study of fine-grained carbonate sediments along the open-ocean northern margin of Little Bahama Bank (Fig. 1). Our study area extends parallel to the northern margin of central Little Bahama Bank for a distance of approximately 100 km and extends 'offshore' for a distance of approximately 50 km. Our results confirm Neumann & Land's (1975) and Boardman's (1978) suggestions that large volumes of bankderived fines are presently being deposited along deep-water Bahamian slopes and basins. Our results also suggest that lateral transport in excess of 100 km occurs, and that the composition of periplatform muds varies in a linear fashion with distance from the shallow-water source.
Methods Thirty-five surface sediment samples were col-
lected from the slope north of Little Bahama Bank in July-August 1979 during cruise E-3A-79 of Duke University's R/V Eastward (Fig. 2). Twenty-four of the samples were collected by a Shipek grab sampler, and 11 samples were collected from the top 5 cm of pilot cores at piston core localities. Samples collected from the axes of submarine canyons, or from pilot cores that indicated evidence of sediment gravity-flow deposits at the sea floor, were excluded from our present analysis to minimize any possible effects of resedimentation by gravity-flow processes. Thus, we assume that our samples are mostly representative of only very recent hemipelagic sedimentation. However, at some grab sites it is possible that we have sampled parts of finegrained non-channelized turbidity-current deposits. Subsamples from each station were first soaked in chlorox to remove organics and various size fractions separated by wet sieving, isolating the 76 ~
78 ~
80 ~
i i
BLAKE
~
28 ~
LATEAU
NORTH
T
o o oJ
0
50 km
28 ~
INNNrn
LITTLE~ -..-(..:?
BAHAMA r ~BANK
9
n
200m
GREAT
BAHAMA~ BAN K
80 ~
25 ~
78 ~
76 ~
FIG. ]. Index map and generalized bathymetry of the northern Bahamas illustrating location of study area.
Open-ocean, off:bank transport o/i/ine-grained carbonate sediment
201
r~
<~ (.~
v
;L
Z <.
o
.=.
rn
F.
E
r~-
-o o
/
o
-~
r~
az
F-
W ..J
/
o
o ILl LLJ
F-
_1 L~
o
r~-
-o r~
O,J
e.i
202
K.C. Heath and H.T. Mullins
mud fraction (silt + clay) for mineralogical analysis. Quantitative carbonate mineralogy was determined by X-ray diffraction techniques (Neumann 1965; Milliman 1974) on a Phillips Norelco diffractometer using nickel-filtered, Cu-KT: radiation, with a fluorite internal standard added to the sample. Replicate diffractograms were obtained for each sample with peak areas calculated by planimeter and values averaged. Precision is estimated to be _+5% (Milliman 1974). The MgCO3 content of calcites was determined by measuring the shift of the main calcite peak towards that of dolomite (Miiliman 1974) and then comparison with the standard curve of Goldsmith et al. (1961). An example of a typical diffractogram containing aragonite, calcite, and magnesian calcite is given in Fig. 3. Our raw data (Table 1) were plotted on a base map and values contoured to depict regional trends, as well as versus distance from the 100 m contour along the northern margin of Little Bahama Bank. Pearson statistical correlation coefficients (r) for each set of data were calculated and regression lines fitted to the data (Till 1974). Samples from various off-bank distances were then chosen for SEM observations to determine lateral compositional changes.
Setting The slope north of Little Bahama Bank extends
down onto the surface of the Blake Plateau (Fig. 1) and consists of a relatively steep (~--4 ~) upper slope between depths of 200 and 900 m, and a more gentle ( 1 - 2 ) lower slope at depths of 900-1300+ m (Fig. 2). The upper slope is cut by twenty-two well-defined, but small (50-150 m in relief) submarine canyons (Fig. 2) which act as a qine source' for the transport of coarse debris to the lower slope (Mullins & Neumann 1979; Mullins et al. 1984). Piston core samples indicate that the upper slope is essentially all hemipelagic (i.e. periplatform) oozes, whereas the lower slope consists of approximately 60% sediment gravityflows interbedded with hemipelagic oozes and ahermatypic coral mounds (Mullins et al. 1981; Mullins et al. 1984). The depth range over which our samples were collected (237 m to 1280 m) is everywhere above both the aragonite (ACD) and calcite (CCD) compensation depths which occur at approximately 2500 m and 4500 m, respectively, in the North Atlantic (Broecker 1974).
Results Mineralogy Plots of our mineralogic data versus distance from the bank edge of northern Little Bahama Bank reveal that the percentages of all three carbonate phases identified (aragonite, magne-
hi
hi
I--
I--
--I
0 "J
-6O
--50
L)
hi
Z
I_
0
no 0
r e(
--40
j 1.1,.
).. F-
-(/) z bJ Iz
--30 b.l
FIG. 3 Typical X-ray diffraction pattern for fine-grained carbonate sediment containing aragonite, magnesian calcite, and calcite. Fluorite peak is an internal standard. Sample #3 in Fig. 2 and Table 1.
--20
--
mlO
.J ,,J
l-
n-
0 *2e
I
I
31
30
I
29
I ?_8
I 27
I 26
25
Open-ocean, off-bank transport of fine-grained carbonate sediment TABLE 1.
20 3
Mineralogical data jor surface sediment samples north of Little Bahama Bank Mineralogy (Mud) Distance Sample Water from bank % % % Mole %* # Type Depth (m) edge(km) Aragonite Calcite Mg-Calcite MgCo3 3 4 6 7 9 10 12 13 15 19 21 25 29 30 31 32 35 36 38 39 43 45 46 48 50 51 52 53 54 57 58 62 63 64 65
P P S P P S P S P P P S P S S S S S S S S S S S S S S S S S S P S S P
900 560 295 1140 645 330 1245 [165 1095 1280 850 285 498 237 1240 1080 439 295 880 1045 1174 t265 1190 1070 513 470 275 365 420 1055 1045 1023 880 970 975
24 12 3.5 48 14 3 42 33 31 41 16 2.5 4 4 31 23.5 3 1.5 17.5 26 33.5 42 43 25 8 6.5 1.5 4.5 7 35 43.5 22.5 18 21 23
37 36 38 11 44 37 23 27 29 22 37 36 43 30 25 31 43 28 32 31 28 25 24 30 41 39 37 42 39 25 26 31 35 32 30
32 22 11 49 16 21 40 46 36 41 24 17 13 21 44 31 16 12 26 33 32 52 43 32 13 19 22 14 23 34 41 22 29 34 39
31 42 51 40 40 42 37 27 35 37 39 47 44 49 31 38 41 60 42 36 40 23 33 38 46 42 41 44 38 41 33 47 36 34 31
15 13 12 14 13 15 13 15 15 14 13 9 14 15 17 14 14 15 13 13 15 11 N.D. 12 14 13 13 12 14 13 12 13 13 11 13
P = pilot core sample. S = Shipek grab sample. N.D. = not determined. * Average mole % MgCO3 = 13.5.
sian calcite, calcite) vary in a linear fashion with distance f r o m the b a n k edge; all at the 99% confidence level (Fig. 4). Percentages o f a r a g o n i t e a n d m a g n e s i a n calcite decrease with increasing distance f r o m the b a n k edge (Fig. 4)~ which reflects the effects o f increasing distance f r o m their source. In contrast, the percentages o f fine-grained calcite increase with increasing distance from the b a n k edge (Fig. 4), which reflects c o n t i n u o u s l y less dilution by b a n k - d e r i v e d aragonite a n d m a g n e s i a n calcite. Statistical analysis o f o u r d a t a a n d the construction o f regression lines allows us to predict the c o m p o s i t i o n o f p e r i p l a t f o r m m u d s simply as a
f u n c t i o n o f distance f r o m the b a n k edge (Table 2). F i g u r e 5 shows o u r results for the p e r c e n t a g e o f total b a n k - d e r i v e d m u d ( a r a g o n i t e + m a g n e s i a n calcite) versus distance f r o m the b a n k edge, which has a correlation coefficient (r) o f - 0 . 9 1 2 . At a distance o f 1 k m f r o m the b a n k edge, o u r d a t a indicate that 85% o f the m u d is b a n k - d e r i v e d , a n d that at a distance of 50 km, h a l f (50%) o f the fine-grained fraction is still b a n k - d e r i v e d . Direct seaward p r o j e c t i o n o f o u r d a t a b e y o n d 50 k m (the limit o f o u r study area) predicts t h a t b a n k - d e r ived m u d s will be present in surface s e d i m e n t s at distances o f up to 120 k m f r o m the n o r t h e r n m a r g i n o f Little B a h a m a Bank (Table 2).
204
K.C. Heath and H.T. Mullins MINERALOGY \
,,
OF
\
9
9
9
MUD
9
9
FRACTION
60-
!
% ARAGONITE
50-
O~ 9
50.00
1,14o
9
"
9
~ ~ m
~LdlTTLE~
.
B - ~ A K- ' ~
9
.
km
"
9
9
N : 35 9 r = -0,824
I0
y = 40,385-
-
0 60- 9 50- ~~
i /
9
20-
IO
9
[~-30 ~
2
~ 9
~
~
l
~
9
0
,
9
,
x
i
i
i
|
20
30
40
50
9
:o M g - C A L C I T E
20-
o 9
9
N : 35
9
r = -0.654
-
IO
Y = 45.694
o
s
'0
60q
o
50 t
~ C
-
0.313 x
30 '
4b
ALCITE
~o
9
~ _
3O
,
2O
9
9 ~,
o
9 0405
,
4 0 - ~8-- -g'~-~J-6"~---~ -
0
9 ~40
--
I0
0
~l~
O~-----_.._O .~O
-
Q
~,T~E'.~ ~.~A~. ~A.~.
...4r . ~ w 9
9
y = 13,922
io]~-I 9 ..
OI
,
0
N =35 r = ~0.914
I0
i
i
, 0.718
i
20 30 40 D I S T A N C E (km)
x
1
5q 50
FIG. 4. Distribution of carbonate minerals in mud fraction of surface sediment samples north of Little Bahama Bank. Surface sediments shown in map view (left) and versus distance from bank edge (right). Aragonite at top, magnesian calcite in middle and calcite at bottom. N = number of data points; r = correlation coefficient.
S E M observations
Samples of mud-size sediment collected at various distances from the northern bank margin of Little Bahama Bank, were examined by SEM to evaluate lateral variation in composition. Our observations reveal that samples in close proximity to the shallow-water bank are dominated by needles and blades of aragonite and magnesian calcite with only minor amounts ofcalcitic coccoliths (Fig. 6(a)). However, with increasing distance from the bank edge, the percentages of coccoliths gradually increase (Fig. 6) until they
constitute around half of the sediment at a distance of 46 km (Fig. 6(d)).
Discussion S e d i m e n t a t i o n processes
Our results are in agreement with Neumann & Land's (1975) and Boardman's (1978) suggestion that large volumes of fine-grained carbonate sediments are being exported from the Bahama Banks during the present high-stand of sea-level.
Open-ocean, off-bank transport of fine-grained carbonate sediment TOTAL
BANK- DERIVED
205
MUD
I0 km
09 ~
~
~
\
~
_
Q 7 0 0 . -9~ - ~ - - ~
\
\
..--
\
I00 % ARAGONITE
+ Mg-CALCITE
90 ~
QO ~
80 %
9
7O 6O 50
_
r
=
-0.912
y = 85.91-
0.7"12
x
40
30
,b
2'o
3'o
DISTANCE
4'o (krn)
Fine-grained carbonates on the bank-tops are stirred up during storms and swept off-bank in suspension by daily tidal action (Boardman 1978). Neumann & Land (1975) have observed large plumes of muddy sediment moving off the Bahama Banks after winter storms, and Proni et al. (1975) suggest that similar plumes sink because of their higher density and then spread out laterally along mid-water pycnoclines (Fig. 7). Vertical transport of such fine-grained bankderived carbonates probably occurs mainly as faecal pellets produced by zooplankton (Honjo
5;
FIG. 5. Areal distribution (top) and versus distance from bank edge (bottom) for total bank-derived carbonate mud (aragonite + magnesian calcite) north of Little Bahama Bank. N = number of samples; r = correlation coefficient.
1980). However, some material may also sink to the bottom and flow downslope as a low-density turbidity current.
Diagenetic potential Our data also indicate that the percentage of fine-grained metastable, bank-derived carbonates (aragonite and/or magnesian calcite) decreases in a linear fashion with distance from the bank edge. Thus, the composition of periplatform muds along open-ocean settings will
TABLE 2. Composition of periplatform muds versus distance from an open-ocean bank edge*
Composition(~o) Aragonite Magnesian Calcite Calcite Total Bank-derived Mud (Aragonite + Magnesian Calcite)
0
Distance from Bank Edge (km) 10 20 30 40 50 60 70 80 90 100 110 120
40 36 32 28 24 20 16 12 8 4 46 43 39 36 33 30 27 24 21 17 14 21 28 35 43 50 57 64 71 78
0 14 86
0 7 93
0 1 99
86 79 72 65 57 50 43 36 29 22
14
7
1
* Percentages beyond a distance of 50 km are predicted values based on direct projection of constructed regression lines.
K.C. Heath and H.T. Mullins
Z06
(a)
(b)
Co)
(d)
FIG. 6. SEM photomicrographs of mud fraction from surface samples north of Little Bahama Bank. (a) Sample #35, 439 m, 3 km from bank edge. Note abundance of aragonite needles, and micarb of magnesian calcite(?) and paucity of calcitic coccoliths; (b) Sample #38, 880 m, 17.5 km from bank edge. Note abundant needles and blades of aragonite and magnesian calcite plus common calcitic coccoliths; (c) Close-up of lower left-hand corner of (b); (d) Sample #46, 1190 m, 43 km from bank edge. Note abundance of calcitic coccoliths and common needles/blades of aragonite and magnesian calcite. Magnifications are all 10000 x, except for (c) which is 15000 x. Scales in the lower left hand corners are in microns.
Open-ocean, off-bank transport of fine-grained carbonate sediment MUD PUT IN S U S PE N S ION BY STORMS
~
OFF-BANK TRANS PORT BY T ID ES
~
' ~
P L A T F OR M
LATERAL
TRANSPORT
~
ALON 8
SURFACE
207
LAYERS
,..,~.!;'.:::.~ii::.:~:.~---,~:.~;.~:::.:~i?.: ., ~::::;:;:~.::.~i:~ "-2 S L VERTICAL~
L ~ . : ~ "' " ~ " " ~ : - : : .
TRANSPORTS .
....
VIA
-
....:.
" FAE A v
..
" v
I[,
~'~.I'~eS~:..~LATERAL .1 T~ANS~ORT~'l ALON~ ,I"~O-WATER],~'~r
9
SLOpE BASIN
Fro. 7. Schematic, generalized model for the lateral and vertical transport of bank-derived carbonate fines. vary with distance from the bank edge. This has important implications for the diagenesis of periplatform oozes, as Schlanger & Douglas (1974) have indicated that the original composition of a carbonate ooze predetermines its 'diagenetic potential'. Periplatform oozes deposited above the ACD in close proximity to carbonate platforms will have the greatest diagenetic potential because of their high content of fine-grained, metastable, bank-derived carbonates. However, the diagenetic potential ofperiplatform oozes will decrease in a linear fashion with distance from a carbonate platform until a stable, pure calcite ooze is reached at distances exceeding 100 km from the platform.
Global budget of COz The seaward projection of our observed trends implies that fine-grained bank-derived carbonates may be transported laterally as much as 120 km from carbonate platforms. This number is probably a minimum value, however, as we have analysed only carbonates contained in bottom sediment samples. Using open-ocean sediment traps, Berner & Honjo (1981) found magnesian calcite (13% MgCO3*)suspended in ocean waters more than 200 km from their nearest known source. Thus, lateral transport from shallowwater carbonate platforms of 200 km or more appears likely. Such large volume and long distance transport of fine-grained, metastable, shallow-water carbonates has important implications for the global budget of CO2. Bank-derived aragonite and magnesian calcite that settle below their respective compensation depths will dissolve and thus not be incorporated into the sediments. As Berner & Honjo (1981) point out, dissolution of such metastable carbonates ' . . . may constitute an * Note that the average mole % MgCO3 in magnesian calcites north of Little Bahama Bank is 13.5'!,,.
important mechanism for the neutralization of excess anthropogenic CO2 added to the oceans.' Any global budget for CO2 must consider the potential importance of calcium carbonate derived from over-production on carbonate platforms. We completely agree with Berner & Honjo (1981) that open-ocean sediment trap experiments are needed near carbonate platforms to quantify more precisely the flux of bank-derived aragonite and magnesian calcite to the deepocean.
Conclusions (l) Large volumes of fine-grained aragonite and magnesian calcite are being exported along open-ocean margins of the Bahama Banks during the present high-stand of sea-level. (2) Immediately adjacent to the northern margin of Little Bahama Bank 86% of the mud fraction is bank-derived. This value decreases in a linear fashion to 50% bank-derived mud at a distance of 50 km from the bank edge. (3) Direct seaward projection of our data suggests that metastable, bank-derived muds are likely to be transported more than 100 km from shallow-water carbonate platforms. (4) The diagenetic potential ofperiplatform oozes is likely to vary in a linear fashion with distance from a shallow-water platform. (5) The oil-bank transport of large volumes of metastable carbonate sediments during times of high sea-level is a potentially important factor in the global budget of CO2 and may be an important mechanism for the neutralization of excess anthropogenic CO2 in the deep-ocean. (6) Open-ocean sediment trap experiments are needed to quantify more precisely the lateral transport and vertical flux of bank-derived carbonate sediments.
208
K.C. Heath and H.T. Mullins
ACKNOWLEDGMENTS: U s e o f the R / V E a s t w a r d to collect samples used in this study was p r o v i d e d for by a U S N a t i o n a l Science F o u n d a t i o n G r a n t to Drs. A.C. N e u m a n n a n d A.C. Hine at the University o f N o r t h Carolina, C h a p e l Hill, a n d the D u k e University C o o p e r a t i v e O c e a n o g r a p h y P r o g r a m . T h e David a n d Lucille P a c k a r d F o u n -
d a t i o n p r o v i d e d travel s u p p o r t for Heath. S E M w o r k was c o n d u c t e d at H o p k i n s M a r i n e Station of S t a n f o r d University. This m a n u s c r i p t benefited from critical reviews o f R.E. G a r r i s o n a n d C.R. N e w t o n at the University o f California at Santa Cruz.
References BATHURST,R.G.C. 1975. Carbonate Sediments and their Diagnesis. Elsevier, Amsterdam. 658 pp. BERNER, R.A. & HONJO, S. 1981. Pelagic sedimentation of aragonite: Its geochemical significance. Science, 211,940-42. BOARDMAN,M.R. 1978. Holocene Deposition in Northwest Providence Channel, Bahamas: A Geochemical Approach. Ph.D. Thesis, Univ. North Carolina, Chapel Hill. 155 pp. BROECKER, W.S. 1974. Chemical Oceanography Harcourt-Brace-Jovanovich, Inc., New York. 214 pp. GARRISON, R.E. 1981. Diagenesis of oceanic carbonate sediments: A review of the DSDP perspective. Soc. econ. Palaeo. Min. Spec. Pub., 32, 181-207. GOLDSMITH, J.R., GRAF, D.L. & HEARD, H.C. 1961. Lattice constants of the calcium-magnesium carbonates. Am. Mineralogist, 46, 453-7. HONJO, S. 1980. Material fluxes and modes of sedimentation in the mesopelagic and bathypelagic zones. J. mar. Res., 38, 53-97. KIER, J.S. & PILKEY, O.H. 1971. The influence of sea-level changes on sediment carbonate mineralogy, Tongue of the Ocean, Bahamas. Marine Geol., 11, 189-200. LYNTS, G.W., JUDD, J.B. & STEHMAN,C.F. 1973. Late Pleistocene history of Tongue of the Ocean, Bahamas. Bull. geol. Sac. Am., 84, 2665-84. MILLIMAN, J.D. 1974. Marine Carbonates. SpringerVerlag, Berlin. 375 pp. & MOLLER, J. 1973. Precipitation and lithification of magnesian calcite in the deep-sea sediments of the eastern Mediterranean Sea. Sedinlentology, 20, 29-46. , Ross, D.A. & Ku, T.H. 1969. Precipitation and lithification of deep-sea carbonates in the Red Sea. J. sed. Petrol., 39, 724-36. MULLINS. H.T. & NEUMANN, A.C. 1979. Deep -
-
car-
bonate bank margin structure and sedimentation in the northern Bahamas. Soc. econ. Pahwo. Min. Spec. Pub., 27, 165-92. --, NEWTON, C.R., HEATH, K. & VAN BUREN, H.M. 1981. Modern deep-water coral mounds north of Little Bahama Bank: Criteria for recognition of deep-water coral bioherms in the rock record. J. sed. Petrol., 51,999-1013. , HEATH, K., VAN BUREN, H.M. & NEWTON, C.R. 1984. Anatomy of a modern open-ocean carbonate slope. Northern Little Bahama Bank. Sedimentology, (in press). NEUMANN, A.C. 1965. Processes of recent carbonate sedimentation in Harrington Sound, Bermuda. Bull. Mar. Sci., 15, 987-1035. -& LAND, L.S. 1975. Lime mud deposition and calcareous algae in the Bight of Abaco, Bahamas: A budget. J. sed. Petrol., 45, 763-86. PILKEY, O.H. & RUCKER, J.B. 1966. Mineralogy of Tongue of the Ocean sediments. J. mar. Res., 24, 276-85. PRONI, J.R.. RONA, D.C., LAUTER, C.A. & SELLERS, R.L. 1975. Acoustic observations of suspended particulate matter in the ocean. Nature, 254, 413-5. RUCKER, J.B. 1968. Carbonate mineralogy of sediments of Exuma Sound, Bahamas. J. sed. Petrol., 38, 68-72. SCHLAGER, W. • JAMES, N.P. 1978. Low-magnesian calcite limestones forming at the deep-sea floor, Tongue of the Ocean, Bahamas. Sedimentology, 25, 675-702. SCHLANGER, S.O. & DOUGLAS, R.G. 1974. The pelagic ooze-chalk-limestone transition and its implications for marine stratigraphy. Int. Ass. Sed. Spec. Pub. No. 1, 117-48. TILL, R. 1974. Statistical Methods for the Earth Scientist. Halstead Press, John Wiley, New York. 154 pp.
K.C. HEATH,Department of Geology and Moss Landing Marine Laboratories, San Jose State University, Moss Landing, California, 95039, USA. * H.T. Mullins, Department of Geology, Heroy Geology Laboratory, Syracuse University, Syracuse, New York 13210, USA. * Please direct all correspondence and reprint requests to Dr. Mullins.
Late Quaternary calcareous clayey-silty muds in the Obock Trough (Gulf of Aden): hemipelagites or fine-grained turbidites? J.-C. Faug~res, M. Cremer, E. Gonthier, M. Noel and J. Poutiers SUMMARY: Three cores have allowed us to survey Holocene deposits in a trough of the Gulf of Aden ridge. Within this particular environment (steep slopes, tectonic and volcanic activity) sediments are fine-grained, apparently homogeneous calcareous muds, rich in clay and planktonic microfossils: these features suggest that they are hemipelagic deposits. However, high sedimentation rates (20 to 50 cm/1000 yrs), lateral variation, and the occurrence of dynamic sedimentary structures (visible on X-radiographs and in thin sections), suggest that they have been emplaced mainly by gravity currents. These reworked hemipelagic materials are deposited on the shelf edge and on the trough slopes. This is confirmed by the sequential arrangement of sedimentary characteristics, as shown by analysis of microstruclures, microfacies, carbonates, clay minerals, grain-size and other components. Two types of sequences have been observed: the first we call the 'type sequence'. This commonly overlies an erosional surface and comprises three divisions from bottom to the top: Division I (equivalent to Piper's E1 division); mud with 8-10% sand fraction, rich in biogenic tests which occur as well-sorted horizontal laminae. Dicision H (equivalent to Piper's E2 and E3 divisions): mud with low sand fraction (1%), very poor in biogenic material, no lamination, but horizontal orientation of silt-sized bioclasts. Division I H (equivalent to Piper's F division); mud with low sandy fraction, poorly-sorted biogenic material, no dynamic sedimentary structures. The second type of sequence is less common and comprises two divisions: Division I'a; very thin layer (millimetric), rich in terrigenous silt, which scours underlying sediment and shows horizontal, oblique and cross lamination. Division I'b; mud with low biogenic sand content, showing horizontal more or less clear and irregularly-distributed lamination. These two sequences correspond to divisions E and F defined by Bouma (1962) and Piper (1978) in lerrigenous turbidite sequences. The 'type sequence" is comparable to turbiditic sequences described in pelagic or hemipelagic material, deposited in similar morpho-tectonic environments (e.g. the eastern Mediterranean Sea or Mid Atlantic Ridge fracture zone). Within the Orgon programme, sponsored by the French "Centre d'Etudes Petroli&es Marines' (CEPM) three cores have been sampled in the Obock Trough (Fig. 1), the deepest part of the Gulf of Tadjoura (Gulf of Aden). These piston and box cores allowed us to survey the Holocene sedimentary cover (Moyes et al. 1981). The Obock Trough (1070 m) is part of the Gulf of Aden rift system (Backer et al. 1973) bordered to the south and to the east by steep slopes. It is an area of much tectonic and volcanic activity (Fig. 2) (Clin & Pelissier-Hermitte 1981). Furthermore, hydrothermal activity is indicated by unusually high temperatures in bottom waters (18.6 ~ Pelet et al. 1980) and by clay minerals neogenesis (Monaco & Valette, pers. comm.). On first examination the Holocene silty-clayey muds, rich in pelagic microfossiis, appear to be homogeneous pelagic or hemipelagic sediments (the 'ubiquitous' sediments of Faug6res et al. 1979). However, detailed lithological survey has shown the role played by gravity currents in emplacing this fine-grained material. The follow-
ing methods were used: X-radiography of sediment slices (0.5-1.5 cm thick), thin-section analysis, grain-size analysis of the bulk sediment and of the calcareous and non-calcareous fractions (Gonthier et al. 1981 ), and compositional (mainly sand-sized bioclasts) analysis.
General sedimentary characteristics The 3.5 kHz seismic profiles (shot by Clin & Pelissier-Hermitte 1981) have allowed us to evaluate the thickness of the sedimentary cover (Fig. 3). This cover is very thin on the slopes bordering the Trough, relatively thick (20-40 m) on the Trough floor, and absent on newly-formed volcanic highs. We have studied samples from the thick section in the centre of the Trough. The sediments have been dated by planktonic foraminiferal associations and are entirely Holocene. This gives a sedimentation rate (25 to 45 cm/1000 yrs) significantly higher than elsewhere in the 209
2 IO
J.-C.
Faugbres et
al.
400
50 ~
% ~6o k ;::i~
/
I-
%.
",~
~-'":
\ - ~ --N',7... .~'N~(":
~BfE
-!:
. .:.:
"9 ::..:.:-..:!.'.:?" "" " : " ~ ~ e v ~- ~ , / /
--
43020 .
42040 ' !
9
I 9 .. 9 :. ,:.':'_.,~_:.~_ ~%01
I 1~
TADJOURA f 0 5 5 0
9
GHUBBET EL
eli Ill
f
KHARAB
0
0
DJIBOUT 9
.~
IOkm
11o30 '
""
"
"
d'Obock
..
,. - .,.
t
I
FIG. 1. Location and structural framework. Gulf of Aden (Moyes et al. 1981). The bacteria community in the sediments (Bensoussan et al. 198 I), shows an abundance of sulphate-reduction types consistent with rapid burial. The general lithological character of the sediment is exemplified by the longest core (KS 01, Fig. 4). The sediment is apparently a homogeneous and structureless olive-grey mud with some horizons richer in pteropod test debris. It is fine-grained (silty-clayey) with a carbonate content of 40 to 55%. The sand fraction represents, on average, 1.5 to 3% of the sediment and exceptionally up to lO ~ and comprises mainly calcareous tests (foraminifera and pteropods)
with some terrigenous material (feldspars, quartz and ferromagnesian minerals) (Figs. 5, 6(a). The silt fraction represents 50 to 70~ of the sediment. The silt-sized particles are generally smaller than 32 pm and mainly composed of calcareous biogenic particles (coccoliths) and detrital carbonates (Fig. 6(a)). The clay fraction (less than 2 /lm) represents around 30~ of the sediment and is dominated by smectites with some attapulgite and illite. Overall, pelagic biogenic particles are predominant. The less abundant terrigenous material results from the erosion of volcanic rock alteration products (smectite and ferrogmagne-
Late Quaternary calcareous clayey-silty muds in the Obock Trough
211
c q~
q~ c
o 0
0 0
"~
o~
o~
~0
U_
0
c-
O
o~
.>
%
,ez o~
.r
E
E
E ~5
",o~
(1)
o~
r-
~)
E
0
to
0
7-
~..
'%~
0
0
6 C
o
FIG. 3. 3.5 kHz profile record across the Obock Trough near to the KS 01 position (after Clin & Pelissier-Hermitte 1981).
FIG. 4. Lithologv of core KS 01 (after Moves et al. 1981).
Late Quaternary calcareous clayey-silty muds in the Obock Trough O
g._L 0ll,
--
001,
--
o~ ^ -Iv.
~
o
--
08
--
0Z
--
09
--
0g
--
0t;,
--
0s
--
-
-
e-
"--
J
~ e~
A
--
0 Ol
o00
E
Q;
DJaJ!U!LUDJOJ. 0!q~,uaq
--
00~
L
'~ :>Z
......'~
O00g
"O
A L
--
O
~
:n
~F::, .2.~ > ._,.2 E9 E
--
0~
::n
.s
"Z: "a
.~O
0~
2E~
G) I
"E 06
213
,&
O
o
r
(..)
i~.~
z
000
--
--
0o~ o
--
Oi
z
//j tA/L-
,
Oi
n
A
--
o00g
000
spodXoel~d
--
00~ 0
n"l
spodo~a~,d
N I ~'-
- -
OOOg
E~ .~
t-
w
--
ooog
.,..
0
r
I
o!UOPlUOld oo~
- -
,
,
o
g
"~
,
o
=_
\
--
~0
]
I
I
I
T
T
1
c:O0 ~ 9
r
~1) g~
'-g
~ 0 r
C: 0 ---
I!
IIIIitl
Jil - !11
o "'"
'"-
~ ~
~
~ "^
,'," w o
"~
z
r.:
,~'E
',-?;
N
,
..
z w ~o
.
'-~
g
E~
rS~
214
J.-C. Faugbres et al.
sian minerals), which outcrop around the Gulf of Tadjoura (Faug6res & Gonthier 1981: Faug+res et al. 1983). The Obock Trough sediments appear therefore to be a typically pelagic or hemipelagic facies. However, pelagic settling from the surface cannot explain the high sedimentation rates observed nor their important lateral variations, for there is a scarcity of continental input within the desert climatic zone and relatively low productivity in an area free of any upwelling effects. How, then, were these sediments emplaced?
Turbidite characteristics Detailed examination of one of the cores has revealed three sets of characteristics that suggest turbidity currents were responsible for the deposition of much of the sequence.
Sedimentary structures and microstructures of dynamic origin Three types of structures and microstructures have been recognized on X-radiographs and in
thin sections (Fig. 5; Fig. 6(a), 1, 2, 3; Fig. 6(b) and Fig. 6(c)). (1) Sharp often erosional contacts separate different-coloured muds. Overlying these contacts the sediment mostly shows horizontal or oblique plane lamination. It is richer in sand- or silt-sized particles and is sometimes graded. (2) Plane lamination is common at several levels in the core and mainly results from an alternation of laminae rich in sand-sized paralleloriented bioclasts with laminae rich in finegrained matrix and with smaller bioclasts. These laminae may be horizontal or oblique. (3) Parallel orientation of bioclasts or of siltsized terrigenous particles along horizontal or oblique laminae. All these structures are indicative of bottom dynamic processes. Moreover, the recurrence of laminated levels through the core suggests a periodic dynamic process.
Cyclicity in the abundance and composition of the sand fraction Analysis of the sand-fraction content through the core shows weak cyclic variation in abundance (Fig. 5). The 'peaks' correspond to an increase in
FIG. 6(a) Above: main components of sand (left) and silt (right). Below: X-radiographs of thin slabs of core KS01 (core width 5 cm); (1) erosive contact and lamination, (2) fine-scale lamination, (3) sharp contact and lamination.
FIG. 6(b) Thin-section photographs (maginification 14.8 x ), core KS01: division I, horizontal lamination alternately rich and poor in sand-sized biogenic material; (2) division I'b, horizontal lamination of silt-sized biogenic material, less clear than (1); (3) division I'a, oblique lamination and abundant terrigenous material. White arrows show way up.
A
v
FIG. 6(c) Thin-section photographs (magnification 14.8 x ), core KS01 (1) division II, slight orientation of silt-sized grains; (2) division III, no sedimentary structures, F = foraminifer; (3) sharp contact between divisions I and III: (4~ erosive contact between divisions I'a and III.
Late Quaternary calcareous clayey-silty muds in the Obock Trough
aoojaJ0lAs
1>2~-::.:':'.'::.:.:.:.:...,:.'-:.'.51
~
~laqs ad01s
~
:
+
,
--
--
--
--
--!':-.'.
--
......... ++. . . . . . . . .
:"..'...'.-.:.'..:"..'!
.::.: .:-.'. :: . .: 9. :.... :' :.. :::-.::.
I~.}:.:.::::
~
............ ,. . . . . .
......
:':.
.
.1
LL 0
(J
<~ n,. 1,1
Z
I--
Z
r~
m
o
,
2I 7
~ LL
tO
--
OJ
- - - - - -
E ::L
I
1
!
~S
O - -
Oh
--
GO
--
s
--
~)
--
QC
[i?/!!:iX::
9
~a
I::{{ S{.:::.
:
<% o
0 ea
O--
~.../.....-.v:..-....-...-.:..........:..
i
3•
7.:~.:
I
I 0 o o f/] W
8 0
w
-a W F-
p9u n q 0
4JaA--
+uopunq0--
Z +uasaJd-~JDJ - aJDJ ~ J a A - -
?:.:i..:....::-:..i:.:.:.:~::.i?::.{y::..::/:.:...~:.~{::. {:.?:?::::::
s n 0 u a 6 0 !q
i
~ I
I
o~
'
~ : ' " " " " " I
IIItrrr,J+ I
o~
.............
' .......
'
......
llll+ igl [11[:!1 o o~
":i: ~- snoua6u~a.
I
~
n-
0~
~W
I--
218
J.-C.
Faugbres et al.
both terrigenous and biogenic material (pelagic and benthic foraminifera, pteropods and pelecypods). However, the biogenic material shows marked variations of a distinctly rhythmic nature. There is a regular transition from azoic to very rich levels reaching a maximum just below the azoic level. Two hypotheses may account for this cyclic faunal distribution. It may have been the result of either successive primary productivity blooms (yielding biogenic-rich levels), or dilution of biogenic material by irregular input of terrigenous material (yielding the biogenic-poor intervals). Yet, the fact that the biogenic and terrigenous maxima are coincident cannot be explained by either hypothesis. A primary productivity increase would have resulted in a relatively lower percentage of terrigenous material, whereas an increase in terrigenous input would result in relatively less biogenic material. Neither of these in fact occurs. More reasonable explanations for the increase of both biogenic and terrigenous sand-fractions are, either winnowing of fine-grained material by bottom currents, or concentration of coarser material by gravityflows. In this very enclosed trough, bottom currents seem less likely than turbidity currents. A sequential analysis oflithological parameters in the cores provided further evidence of turbidity current emplacement.
Turbidite sequences
We have opted for a very detailed analysis of one core section, thought to be representative of the whole core KS 01 (Fig. 7), as this section shows each of the sedimentary structures mentioned above. It contains two horizons rich in fauna, which are separated by two nearly azoic horizons. The point of maximum fauna corresponds to the richest terrigenous sand and silt horizon. We have identified two types of sequences each overlying a sharp, sometimes erosional contact. (1) 1st sequence (between 1.95 and 1.42 m, Fig. 7) with three divisions. Division I (Fig. 6(b)-l), at the base, is calcareous silty-clayey mud which shows, in comparison with overlying divisions, the following characteristics: a relatively high sand-fraction (8 to 10% of bulk sediment); the abundance in this latter fraction of calcareous bioclasts; the highest carbonate content in the fine-grained fraction (45 to 55%); the lowest clay fraction (25%); the presence of sedimentary structures such as lamination, sand and silt-sized bioclast orientation; a good sorting of foraminifera and other sand-sized bioclasts, which are smaller than those in division |II, described later; an upward decrease in the abundance of biogenics, which could be considered as a grading indicator; the predominance of pteropods over pelagic foramini-
m
=3 E
~
E3
-2
"i
E2
: ____=~..~
I
I' b
i
EI
-----------------"=
D
~
C
BOUMA 1962
E
C
PIPER 1978
TYPE I
S E Q U E N C E S TYPE TT
PIPER 1978
BOUMA 1962
FIG. 8. Comparison of the two Obock Trough sequences with the sequences of Bouma (1962) and Piper (1978).
Late Quaternary calcareous clayey-silty muds in the Obock Trough fera and other biogenics; the constant occurrence of nearshore and continental-shelf species among the benthic foraminifera, which are also less numerous than the pelagic ones; a granutometric facies (Rivi6re 1977) that is logarithmic to parabolic (Gonthier et al. 1981), with slightly better sorting and a slightly coarser median diameter (5-8/~m) than those in the following divisions. At the top of this division, the biogenic fraction decreases, lamination disappears and there is a gradual transition to division II. Division H (Fig. 6(c)-1) is a silty-clayey mud, slightly less calcareous (40~o) than division I and characterized by: a very reduced sand fraction (1%) and therefore a low sand-sized bioclast content; among these bioclasts, pelagic foraminifera are predominant, but shallow-water benthic forms are always present; the highest clay content (30 to 35%); the absence of lamination, but some evidence of silt-sized bioclast orientation; a granulometric facies that is logarithmic, with a median diameter between 4 and 5/~m and very poor sorting. Division I I I (Fig. 6(c)-2), at the top, is always a calcareous silty-clayey mud characterized by: an intermediate sand fraction (2-3%); a greater biogenic particle content than in division II, though always lower than in division I; this biogenic fraction made up of pelagic foraminifera and pteropods of all sizes, and benthic foraminifera, some of which are shallow-water species: a clay content between 30 and 35o/ 9an absence of / o, lamination or bioclast orientation (except in transitional levels between divisions II and III); some evidence of bioturbation; a granulometric facies that is logarithmic, with very poor sorting and a median diameter of around 4/~m. (2) 2ndsequence (between 2.35 and 1.95 m, Fig. 7) with two divisions. Division I'a (Fig. 6(b)-3 and Fig. 6(c)~,)) at the base, has a sharp erosional lower contact, overlain by a thin millimetric lamina, rich in terrigenous medium-grained (30/~m) silts. Biogenic particles are very scarce. This division shows horizontal and oblique lamination due to variation in the abundance of terrigenous silt-sized material. Division I'b (Fig. 6(b)-2) directly overlies division I'a without any transitional zone. Its composition is very similar to that of division III in the first sequence. However, it shows many millimetric laminae (fairly well visible on X-radiographs), which seem to be caused by either, variation in the abundance of sand-sized biogenics or, by preferential orientation of this material. Moreover, within this division, homogeneous centimetric intervals without sedimentary structures alternate with laminated intervals: which shows a secondary rhythmic phenomenon.
219
Discussion Sequence interpretation For the first sequence, we have tried to understand both the meaning of each division and of the sequence itself. We have therefore compared the divisions of our sequence, mainly composed of calcareous fine-grained material, with those of turbiditic sequences observed by Bouma (1962) and Piper (1978) in terrigenous deposits (Fig. 8). Our division I seems to be equivalent to Piper's E1 division (graded laminated mud) and to Bouma's D and E divisions. Piper's E2 division (graded mud) has not been identified, whereas our division II and his division E3 (ungraded mud) appear to be equivalent. Our division III then corresponds to Piper's F division, although division III does not always show all the division F features. Division III appears rather to be a transition between the upper part of the turbiditic sequence and true pelagic sedimentation. This first sequence is therefore interpreted as a turbiditic sequence composed of pelagic and hemipelagic material with only the upper divisions of the classical turbiditic sequences represented. How then can the second, more terrigenous sequence, be compared to the Piper-Bouma sequences? We believe that our division I'a (silty lamination) is equivalent to Bouma's C division, and division I'b is equivalent to Bouma's D division. The whole sequence (I'a + b) may correspond to Piper's E1 division. The recurrent lamination in division I'b might have been the result of hydraulic jump effects (Komar 1971) or of other processes (see review in Stow & Bowen 1978). In comparison with the first sequence, the latter one would be truncated at the top and is composed of more terrigenous material,
Comparison with biogenic turbiditic sequences We have compared the first sequence, the most complete, with turbiditic sequences composed of pelagic or hemipelagic material and which have been described previously from comparable morphotectonic environments (Fig. 9): the eastern Mediterranean trough (Blanpied & Stanley 1981; Stanley 1981) and the northern Mid-Atlantic Ridge Charlie-Gibbs fracture zone (Faug6res et al. 1982, 1983; Gonthier et al. 1982). The following correspondence can be established between the divisions of our sequence and those of the sequence described by Blanpied & Stanley (1981) in fine-grained sediments without sedimentary structures or with fine lamination only (unifites). Our division I corresponds to their graded and faintly-laminated silty mud, with a
J.-C. Faugkres et al.
220
0
c~
~'~
r co (0 0 -~_
0 q~
~ 0
~_ 0
o~-. 0
W Of
~D t~
oo
~
.0 =
0
0
- -
0
O ~
o.
c
c
"~ ~-"
O~
~n
~E o9_
o ~-
~
~-~
0
0
tO
-0
0 bq
0
C
E
0
"--
0
LL
o
0
- -
CO
o
~
0 0
0
"~-
.-
nO nO 0
eO nO C 0 oo
0 c''~_
,= p. .o
if)
.=
rn ~o
U:
D
\
/
\
\
/
/
E \
co
\
.=
in~
W
I
C3
0
T
-
-
--
Z
W OE
b ,,,i
o~ , . ,
,..~\
o m 0
I
~
oE
0
0 ~0
I
o
\ ii~: ~ ,i~ii
I L.)
II II
n...
I
IIIIIi
I
Illl,
.0
9
0
Z
E
-~
-&
._J g ._j ~ w I
o_~
l-o-o
~E
c.E
~-a
o E
I
E
0
o4
-->~c M.--
- -
Late Quaternary calcareous clayey-silty muds in the Obock Trough largely biogenic sand-size fraction. Their graded and ungraded mud may correspond to our second division in which we don't observe any graded mud but consider it to be more terrigenous. The topmost fining-upward mud, with an increased proportion of sand-size biogenic components, corresponds to our division III, which present more textural fluctuations. In the Charlie-Gibbs fracture zone, the sequence is much more important and complete. We can still find in the upper divisions of that sequence the equivalent of the sequence found in the Obock Trough. The Charlie-Gibbs sequence comprises the following divisions, from the bottom to the top: (1) sedimentary breccia (1.5 m); (2) graded calcareous sands (3 m) mainly composed of planktonic foraminifera tests (equivalent to Bouma's A and B divisions); (3) silty-sandy calcareous mud with laminae (0.9 m) (corresponding to Bouma's C and D divisions, and to Piper's E1 division, and consequently to division I of our sequence); (4) clayey-silty homogeneous mud (3.5 m) with very scarce foraminifera (similar to our division II); (5) calcareous silty-clayey mud (similar to our division III), There are a number of other examples of biogenic turbidites, both calcareous and siliceous, that have been described from DSDP sites in deep oceanic settings (see Kelts & Arthur 1981, for summary). Similarly, numerous redeposited carbonates are recognized in ancient deposits on land, some of these being composed of essentially pelagic material (e.g. Stow et al., this volume). In all cases the sequences described show many similarities with those in the Obock Trough sediments, and so we are confident in our interpretation of the latter as turbidites. Steep slopes and high tectonic activity appear to be common controls for the redeposition of pelagic sediments.
Frequency and origin of the Obock Trough turbidites In the whole core the first sequence is the most common, and seems to be the most representative of sedimentation in the Obock Trough. We have chosen to call it the 'type sequence'. The second sequence (I'a and b) is the rarest and, although still dominantly biogenic and pelagic in composition, is marked by the influx of terrigenous material. The problem is, however, to establish the origin of these sporadic terrigenous inputs. They cannot be the result of aeolian inputs because the deposits have an erosional basal contacts. Furthermore, if they were aeolian (sensu stricto) they should be more frequent, because
22I
sand storms were a common occurrence during the Holocene period. We therefore propose two different origins for the turbidity currents: (1) most often these currents originated on the basin slopes from a dominantly hemipelagic source, giving rise to the ~type sequence'; and (2) sporadic arrival of material coming directly across the continental shelf from coastal, deltaic and barrier beach complexes, give rise to the second sequence; this, however, has not yet been confirmed by benthic micro-faunal studies. Nonetheless, this second type of turbidity current is relatively common, as at least 12 sequences are present in the Holocene deposits.
Conclusion The Obock Trough sediments are fine-grained, apparently homogeneous muds composed of pelagic or hemipelagic materials. Detailed analysis of structures and microstructures, microfacies, lithology and granulometry shows that turbidity currents played a major role in their emplacement. The predominance of these dynamic processes is typical of this type of morphotectonic framework: very uneven topography with steep slopes and high tectonic activity. It has lead to the accumulation of large quantities of ponded sediments in the Obock Trough which has acted as an efficient sediment trap. Turbidity currents reworked pelagic and hemipelagic fine-grained sediments previously deposited in the slope or on the shelf edge, and deposited beds with characteristic sequences. Compared to classical turbiditic sequences (Bouma 1962; Piper 1978, Stow & Shanmugan 1980) defined in terrigenous sediments, the sequences described here are more or less truncated, with (1) the absence of bottom divisions because of their originally fine-grained sediment nature, (2) the absence of the uppermost divisions due to the frequency and the erosional power of the turbidity currents, which have not allowed the deposition of a significant quantity of pelagic or hemipelagic sediments in the Trough. Moreover, sedimentary structures of dynamic origin are often less abundant and less clear than in the equivalent terrigenous turbidites. In the Obock Trough, as for the other examples described (Charlie-Gibbs Fracture Zone and Mediterranean trough), biogenic calcareous sand-sized material is a good indicator of the depositional process: it is concentrated and rather well-sorted in the bottom division of the "type sequence', where the sediments are laminated; it is almost absent in the middle division of the sequence and reappears, less abundant and poorly-sorted, at the top of the sequence where the sediments are free of dynamic structures.
222
J.-C. Faugkres et al. References
BACKER, H., CLIN, M. & LANGE, K. 1973. Tectonics in the Gulf of Tadjoura. Marine Geol., 15, 309-27. BLANPIED, C. & STANLEY, D.J. 1981. Uniform mud, (unifite) deposition in the Hellenic Trench, eastern Mediterranean. Smith. Contr. Marine Sci., 13, 40 PP. BENSOUSSAN, M., BIANCHI, A., BONNEEONT,J.L., BOUDABOUS, A. & SCODITTI, P.M. 1981. Les communaut~s bact~riennes des eaux et des s~diments profonds du Golfe d'Aden et de l a m e r d'Oman. II--Potentialit6s cataboliques des populations a6robies. In: CNRS (ed.), Gbochimie organique des sbdiments matins profonds, Orgon IV; Golfe d'AdenMer d'Oman. Paris. 23-42. BOUMA, A.H. 1962. Sedimentotogy of some Flysch Deposits. Elsevier, Amsterdam. 168 pp. CHN, M. & PELISSlER-HERMITTE,1981. Le cadre structural de la campagne Orgon IV. In: CNRS, (ed.), G~ochimie organique des s&timents marins profonds. Orgon IV: Golfe d'Aden-Mer d'Oman. Paris. 183-8. FAUGERES, J.C., GAYET, J., GONTHIER, E., GROUSSET, F., LATOUCHE, C., MMLLET, N., POUTIERS, J. & TASTET, J.P. 1979. Evolution de la s6dimentation profonde au Quaternaire r~cent dans le bassin nord-atlantique: corps s6dimentaires et s6dimentation ubiquiste. Bull. Soc. gbol. Fr. 7, t. XXI, no 5, 585-601. -& GONTHIER, E. 1981. Les argiles des sbdiments marins du Quaternaire rbcent dans le Golfe d'Aden et la mer d'Oman (mission Orgon IV). Oeeanologica Acta, 4, no. 4. , GAYET, J., GONTHIER, E., POUT1ERS,J., & NIANG, I. 1982. La dorsale m6dio-atlantique entre 43 ~ et 56~ N, faci6s et dynamique s6dimentaire dans plusieurs types d'environnement au Quaternaire r6cent. Actes colloque International CNRS, Bordeaux, Septembre 1981, Bull. Inst. Gbol. Bass& d'Aquitaine, Bordeaux, 1982, no. 31, 195-215. , GONTHIER, E. & POUTIERS, J. 1983. Facies and sedimentary dynamics in Charlie-Gibbs fracture zone (35 and 36 ~ W) during Late Quaternary. Marine Geol., 52, 101-19. GONTHIER, E., FAUGERES, J.C. & WEBER, O. 1981. Signification des m&hodes granulom&riques pour
la reconstitution des processus hydrodynamiques fi partir de s6diments terrig6nes carbonat6s de milieux marins profonds. Bull. Inst. G~ol. Bassin d'Aquitaine, Bordeaux, 1981, 30, 31-49. --, FAUGERES,J.C., & POUTIERS,J. 1982. Caract6res granulom&riques d'unc s~quence s~dimentaire de la fracture de Gibbs (dorsale m+dio-atlantique), courants de fonds et courants de turbidit6. 'Environnements sbdimentaires de l'Atlantique nord au Quaternaire', Bull. Inst. Gbol. Bassin d'Aquitaine, Bordeaux, 31,217-24. KELTS, K. & ARTHUR, M.A. 1981. Turbidites after ten years of deep-sea drilling-wringing out the mop? Soc. econ. Paleo. Min. Spec. Pub., 32, 91-127. KOMAR, P.D. 1971. Hydraulic jumps in turbidity currents. Bull. geol. Soc. Am., 82, 1477-88. MOYES, J., DUPRAT,J., FAUGERES,J.C., GONTHIER,E. & PUJOL, C. 1981. Etude stratigraphique et s6dimentologique. In: CNRS, (ed.), Gbochimie organique des skdiments marins profonds, Orgon IV, Golfe d'AdenMer d'Oman. Paris. 189-264. PELET, R. 1981. G6ochimie organique des s6diments marins profonds du Golfe d'Aden et de la met d'Oman: vue d'ensemble. In: CNRS, (ed.), Gbochimie organique des sbdiments marins profonds, Orgon IV, Golfe d'Aden-Mer d'Oman. Paris. 529--47. PIPER, D.J.W. 1978. Turbidite muds and silts on deep sea fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross: Stroudsburg, Pa. 163-176, RIVIERE, A. 1977. M6thodes granulom&riques, techniques et interpr6tation. Masson, Paris, New York. 170 pp. STANLEY, D.G. 1981. Unifites: structureless muds of gravity flow origin in Mediterranean basins. GeoMarine Letters, 1, 77-83. STOW, D.A.V. & BOWEN, A.J. 1978. Origin of lamination in deep-sea fine-grained sediments. Nature, 274, 324-8. --8s SHANMUGAM,G. 1980. Sequence of structures in fine-grained turbidites: comparison of recent deep sea and ancient flysch sediments. Sed. Geol., 25, 23-42.
J.C. FAUGERES,M. CREMER,E. GONTHIER,M. NOEL AND J. POUTIERS,D6partement G6ologie et Oc~anographie, Universit6 Bordeaux I, Laboratoire Associ~ CNRS no 197, Avenue des Facult6s, 33405 Talence Cedex France.
Depositional model for calcilutites: Scaglia Rossa limestones, Umbro-Marchean Apennines D.A.V. Stow, F.C. Wezel, D. Savelli, S.C.R. Rainey and G. Angell SUMMARY: The calcilutites of the mid-Cretaceous to uppermost Eocene Scaglia Rossa Formation in Italy have previously been interpreted as a continuous sequence of slowly deposited pelagic limestones. However, regional considerations of thickness, sedimentation rates and facies distribution, together with detailed analysis of composition, texture and sedimentary structures suggest that turbidity currents have played an important part in the deposition of, at least, the middle (Maastrichtian) member. Turbiditic calcilutites are very fine-grained and massive, whereas the pelagites are coarser-grained, mottled and bioturbated. There is, however, a continuous gradation of facies and processes between these two end-members, that we believe is applicable to many other modern and ancient deep-water carbonate systems. The common overlap of turbiditic and pelagic processes in fine-grained 'pure" carbonates arises partly from their relative lack of flocculation compared with muddier sediments. The Umbro-Marchean Apennines of central Italy comprise Mesozoic and Cenozoic sediments that have been crumpled into large open folds and broken by normal faults. The anticlines raise topographically more resistant Mesozoic limestones above easily eroded Tertiary clastics in the flanking synclines (Fig. 1). This has resulted in long NW-SE trending mountain ridges through which antecedent rivers flowing to the Adriatic have cut deep gorges and exposed good sections. In common with other Tethyan areas, the region evolved from a carbonate shelf in the early Mesozoic, to a series of carbonate basins and platforms active from the Jurassic to Palaeogene, that were finally levelled and filled by Miocene clastics. The whole area is probably parautochthonous (Bortolotti et al. 1970). We have concentrated in this paper on part of the basin-platform system of late CretaceousPalaeogene age, the Scaglia Rossa Formation. This is between 200 and 400 m thick and comprises dominantly limestones, with interbedded marlstones, shales and cherts. A small part of the Scaglia Rossa at Gubbio (Fig. I) has recently been the subject of detailed lithostratigraphic, biostratigraphic and magneto-stratigraphic studies and has been designated as the type section for the late Cretaceous-Palaeocene geomagnetic reversal time-scale (Alvarez et al. 1977). This area is also well known for the controversy surrounding a supposed high-iridium anomaly at the Cretaceous-Tertiary boundary and its implication for asteroid impact and Mesozoic extinctions (e.g. Alvarez et al. 1980; Wezel et al. 1981). There has been a further controversy in previous papers over the age and depositional processes of the Scaglia Rossa limestones (e.g.
Alvarez & Lowrie 198 !; Wezel 1981). In particular, studies of the section near Gubbio have emphasized the pelagic nature of the limestones (Arthur 1976; Arthur & Fischer 1977), whereas more extensive studies throughout the UmbriaMarchean Apennines have pointed to the turbiditic character of much of the Scaglia Rossa (Wezel 1979). We have therefore examined in greater detail the sedimentology of the fine-grained facies throughout the area of outcrop of the Scaglia Rossa, with a view to evaluating the conflicting evidence on the processes by which the sediments were deposited. Careful field observation, measurement and sampling has been followed-up by thin section petrography, X-ray diffraction and X-radiography of thin rock slices.
Stratigraphy The mainly rose-coloured limestones and associated sediments of the Scaglia Rossa formation straddle the Cretaceous-Tertiary boundary from the mid-Turonian to uppermost Eocene (Fig. 2: Renz 1936; Luterbacher & Premoli Silva 1964; Premoli Silva et al. 1976). In most parts of Umbria, the dark organic-rich fissile shales and lighter-coloured marls of the Aptian-Albian Marne a Fucoidi Formation pass conformably up into grey, greenish and whitish limestones and dark cherty black shales of the Scaglia Bianca Formation. These extend into the early Turonian and appear to be conformably overlain by the Scaglia Rossa. However, on the east coast at Mount Conero the upper part of the Marne a 223
D . A . V . S t o w et al.
22 4
I
13~
100 km
RIMINI 44 ~
%
URBINO 23 c FURLO
J/ I
ANCONA MONTE 7CONERO
i I
IlL
k
13
FIG. 1. Geological map of part of central Italy showing study area. Oblique hachure = upper Tertiary-Quaternary; blank = Mesozoic--lower Tertiary (including Scaglia Rossa Formation). Numbers refer to measured sections of Fig. 3 (cf. Wezel 1979). Regional names underlined are those areas referred to most frequently in the text. Fucoidi and the entire Scaglia Bianca are missing and there is a marked discontinuity near the base of the Scaglia Rossa (Wezel 1979). At the end of the Eocene, the Scaglia Rossa in most places passes conformably into Oligocene varicoloured marls and limestones of the Scaglia Variegata Formation, although locally there are disconformable contacts, and thence into Oligo-Miocene volcaniclastic and clastic turbidite basin-fill of the Scaglia Cinerea Bisciaro and Marnosa Arenacea Formations. The Scaglia Rossa itself has been subdivided into five members (Fig. 2;
Wezel 1979) that are broadly correlatable over the whole of the Umbria-Marche area (Fig. 3). From base to top these are as follows: The Lower Cher O, Member (LCM) comprises thin- to medium-bedded limestones interbedded with thin, irregular chert layers throughout and rare marlstone-shale partings in the lowermost 15-20 m. It ranges from 20 m to 160 m in thickness, being between 80 and 120 m in most sections and having an average sedimentation rate of 6-10 mm/1000 yrs. The Lower Marly Member (LMM) is a relatively
Depositional model f o r calcilutites Age
Thickness le
Lithology 9 "
,
,
.
.
Sedimentation r ,.rormat,on/,wemuer . /,, Rate mm 1000yr
,"
|
9 9'.'," ..'.-., Miocene
9
9
9
22 5
MARNoso ARENACEA
3'0
2'o
o
";, ",; ~,',; ;,"
~. . . . . . . . . .
26
BISCIARO/CINEREA
a.a.-a.-____.
Oligocene 37
SCAGLIA VARIEGATA
;!!!
U C H E R T Y MBR
:.
Eocene 53
Palaeocene --
a- ~
a-~ , "
.~
,.L--.a.~.a.
9
65
o 0
"
Maastrichtian r
9
U MARLY MBR
]i I
'~
~":':"'"'iiii
co
|"
~: .'..'..':.-'.. - ..
CALCARENITE MEMBER
....
co
' :-'-:-..."
0
.".".'.':
.':
:: . ' !:.:.::': :': :.'::'1
~'s163".":'.'-:.'.':'."."
--7O Campanian
o a.--a.-a.---~
L MARLY MBR
-
' "':'b' .
-J
9..:..:...
78
,~176
Santonian --
--
L CHERTY MEMBER
82 Coniacian 86 Turonian
0 co
"." .'. ['i" "
92
Cenomanian -- 100
SCAGLIA BIANCA
:':; "::
.1. a - . . ~ _1. .A_-I-_l._l.
.-.. ,.-
Albian E'
o o
, .
MARNE A FUCOIDI
30
20
i
l
lO i
Fl6.2. Stratigraphy, average thickness lithology and average compacted sedimentation rates of the mid-Cretaceous to Miocene succession in central Italy. Solid lines = shale; stipples = sandstone; v---volcanogenic sandstones; see Fig. 3 for other ornamentation.
thin horizon from 0-10 m thick. It comprises interbedded marlstones and shales and is commonly slumped and faulted. This deformation appears to be of both syn-sedimentary and probable tectonic origin related to the relative plasticity of the interval. The Calearenite Member (CAM) comprises thinto thick-bedded limestones with thin (average 2-3 mm, rarely > 1 cm) marlstone or shale partings;
in parts the limestones are calcarenites and calcirudites. It ranges from 40 m to 160 m in thickness, averaging about 80-100 m, and accumulated at a rate of 5-20 mm/1000 yrs, or locally over 30 mm/1000 yrs. The Upper Marly Member (UMM) is lithologically similar to the LMM, commonly slumped and faulted in a similar way especially at its relatively abrupt contact with the underlying
D.A.V. Stow et al.
226 SCAGLIA VARIEGATA
2
5 ikF
u..
k-1
- . sl:Z:~i -3-r-~
~
.1 ~!I l t"~! /'1I
bl /2-1 K:!--
~
.1,,. _..sl ~' e l Ii
SCAGLIA ROSSA
~I st]
CAM
11 lI ia I iI II II r~l
Sl, 'J,
!1 i If I II
I i s i II I
,_I.1
,
I1
'J
I
II
I II
I I ,i$ I -
li
i
LJ LCM
T''|~.
;iAi
II
-~
LJ
SCAGLIA BIANCA
U
LLJ ~ 9 ~ lJ i
ii
I
k-k-rk W-'~\
l il ~4--n
~'11~I ~
L_L]
rr-~
b- ".mes'~ limestone +chert maristone
~
~
I_.
.Rye.,?" Fkn-I
"S.L_~__I
N N i|1l l i
9
t~11
~
Ii
I
1
I1
N L! | ||| IIl
7
&
.
:_
":i II I I II CAM i l i
I I
N
9
",
!
I I I I All
13 UCMI~
I'--- I.I li I II'---I.I " "' -I I' I' I I"I" P"f" "'-I ""1 I-'I'---I . . . . I'I I ..... i-r-"-'l
I-+-.,!
~ ~
200m
UMM ~
-
+2.1
I I] I I
I I I I ! I ! It I I I
l-rJlJ-r IrE
!'"1"t , , ~, ,,, 9t l..~.,,.! , L" r'"h I I-', I
51~ /i
calcareniteLMM! +~- calcirudite calcirud te II I " " "I All disturbed bedding II I
i...-H t.t--.-i F--" ]-I ,,"i-4-~-i
I
II
i
il
I
--
/
FIG. 3. Measured sections of Scaglia Rossa Formation in various parts of Umbria-Marche regions of central Italy, based on Wezel (1979). 1. Fossombrone, 2. Pietralata (Furlo), 3. Acqualagna, 4. Bacciardi, 5. Moria, 6. Fosso del Mulino, 7. Monte Conero, 8. Apiro, 9. Frontale, 10. Monte Vicino, 11. Vittore di Genga, 12. Morello, 13. Gubbio, See Fig. 1 for locations. Scaglia Rossa Formation is divided into five members: LCM, Lower Cherty Member; LMM, Lower Marly Member; CAM, Calcarenite Member; UMM, Upper Marly Member; UCM, Upper Cherty Member. CAM, and is mainly thicker ranging from 10 m to 30 m (rarely more) with an average sedimentation rate of 1-5 ram/1000 yrs. The Upper Cherty Member (UCM) is lithologically similar to the LCM, but is thinner (up to 60 m) and has a somewhat lower sedimentation rate.
Chert is absent and marlstone-shale partings are rare in the lower part of interval, both becoming more frequent and thicker upwards. The shaley 'partings' are commonly over 2 cm in thickness and rarely over 10-15 cm, and the whole interval appears more varicoloured.
Depositional model for calcilutites Detailed biostratigraphic zonation of the Scaglia Rossa has been worked out by Renz (1936, 1951), Luterbacher & Premoli Silva (1962, 1964), Premoli Silva (1977) and Premoli Silva et al. (1976). Although Coccioni (1978) found some rather puzzling Miocene forams in certain shale interbeds, it now seems most likely that these were the result of some natural, perhaps tectonic, contamination process (Alvarez & Lowrie 1981; Wezel 1981, Coccioni 1983). Magnetostratigraphy of the Scaglia Rossa is summarized by Alvarez et al. (1977) and Lowrie & Alvarez (1977).
227
Sediment facies Five main facies can be recognized within the Scaglia Rossa formation: calcirudites, calcarenites, calcilutites/fine-grained limestones, marlstones/shales, and cherts. These are described briefly below (Fig. 4) before we examine in more detail the characteristics of the dominant facies, the calcilutites or fine-grained limestones. The calcirudites are mainly medium to very thick-bedded (up to 3 m) detrital limestones made up of bioclasts and rare mudstone lithoclasts up to several centimetres in diameter. They also
FIG. 4. Photographs of main Scaglia Rossa facies. (a) Calcirudite with large flute cast cutting into marlstone and calcilutite (b) Parts of two graded calcarenite--calcilutite turbidite beds (c) Calcirudite/coarse-grained calcarenite erosive into calcilutites; note Zoophycos trace fossil to left of hammer handle. (d) Interbedded calcilutite and calcarenite (whiter) beds with very thin shale partings.
228
D . A . V . S t o w et al.
occur as thin (0-10 cm), discontinuous, commonly negatively-graded layers. The bioclasts include molluscs, echinoids, bryozoans, corals and rudists, commonly fragmented and associated with sand-sized foraminiferal debris. The beds are commonly massive to graded with a scoured, erosive base (flutes up to 30 cm deep) and more gradational top. They are clearly the result of sediment gravity-flow (?turbidity currents) derived from a shallow-water carbonate platform to the east. The calcarenites are also often medium to very thick-bedded (up to 3 m) detrital limestones, occurring either separately or as part of a graded calcirudite-calcilutite bed. The most obvious sand-sized components are foraminiferal fragments with some probable molluscan debris, but recrystallization has obscured much of the original character. The beds are normally flat-bottomed and more or less distinctly graded with a range of internal structures including horizontal, cross and convolute lamination, water-escape structures and, near the top, bioturbation. Partial and complete Bouma sequences are common, and the beds clearly result from turbidity current or related sediment gravity transport. The calcilutites or, to use a less genetic term, fine-grained limestones, are the most abundant facies but more difficult to describe and least easy to interpret. It is these beds that are interpreted as pelagites, especially in the west near Gubbio, and as fine-grained calciturbidites near Monte Conero on the east coast. They are discussed in detail in the following section. The marlstones and shales form a continuum from beds with up to 50% CaCO3 especially in the LMM and U M M , to thin beds and partings of shale with less than 5% CaCO3. They occur throughout the Scaglia Rossa, interbedded with the harder prominent limestones with more or less regularity. Their crumbly and fissile nature obscure any primary internal sedimentary features. They appear to result partly from variation in the biogenic-terrigenous input and partly from diagenetic dissolution of carbonate. This facies is also discussed below. The cherts are thin, irregularly-bedded and nodular layers mainly confined to the LCM and UCM. They are mostly reddish in colour, but greyish cherts also occur towards the basal contact with the Scaglia Bianca, intermittently through the UCM, and at Monte Conero. The cherts are commonly very hard, intensely-fractured and recrystallized rocks, that probably derive originally from siliceous pelagic microfauna and flora. The regional and stratigraphic distribution of these five facies is notable (Fig. 3). Calcirudites
and calcarenites are dominant near Monte Conero in the east and also form a significant part of the succession in parts of the Furlo-Monte Vicino area. Further west, near Gubbio, they are completely absent and calcilutites are dominant. Stratigraphically, they mainly occur in the CAM although locally form part of the LCM, U C M and UMM. The other facies show less clear regional partition, apart from that associated with the thickness variation of the separate members. Stratigraphically, the marlstones and shales are more common and thicker-bedded in the early members and the cherts are confined to the cherty members. Each of the facies may be variously disturbed or contorted. In particular, the LMM and U M M and, less commonly, the UCM show highly chaotic bedding. To some extent this disturbance of the more shaley horizons may be a result of preferred tectonic slip along the more plastic intervals as the whole sequence was folded. However, some of the tightly-folded beds of the CAM (Figs. 3 and 4) enclosed in completely undisturbed flat-lying strata and showing a range of deformationed structures in a single slump, appear to be the result of slumping of soft and semiconsolidated sediments that had accumulated on a depositional slope. Preferential dissolution along folded bedding planes has in some cases produced a 'fish-bone' pattern that obscures the true nature of the folds. There is also evidence of relatively small-scale channelling, both in the coarser-grained facies at Monte Conero and in the mostly fine-grained calcilutites and calcarenites of the Furlo-Monte Vicino area. In the former area, the channels are best developed in the CAM, being several metres deep and several tens of metres wide, whereas in the latter they are mostly shallower and less broad, confined to the U M M and filled with lenticular calcarenite beds averaging 20-50 cm in thickness.
Calcilutite characteristics Bedding The most obvious characteristic of the Scaglia Rossa calcilutites is the rhythmic nature of the bedding (Fig. 5). Throughout the region the harder, more massive and thicker-bedded calcilutites are separated by thinner beds or very thin partings of marlstone and shale. The calcilutite bed thicknesses are very variable from less than 5 cm to as much as 120 cm, although within any one part of the section they tend to be more uniform, whereas the intercalated marlstone-shale layers are about an order of magnitude thinner, from
Depositional model for calcilutites
229
FIG. 5. Photographs of typical features in Scaglia Rossa calcilutites. (a) Thinning-upwards sequence of calcilutite and calcarenite (white) beds over about 15 m section (cf. Fig. 6(a)). (b) Detail of calcilutite and thin shale beds; lower calcilutite has 'over-run' a Zoophycosmound on underlying bedding surface and re-established a planar surface. (c) Lensing, channelling and slump features in calcilutites, near centre of photograph. (d) Irregular variation in calcilutite and calcarenite bed thickness over about a 16 m section (cf. Fig. 6(b)). (e) Slumped horizon overlain and underlain by horizontally-bedded calcilutites. 5 m m to 100 ram, or in some cases are completely absent. Mostly, the beds are very even and flat-bedded, locally with a b u n d a n t stylolites that m a y obscure the true nature and thickness of the bedding. M o r e rarely there is some m i n o r swelling and pinching. The thickest beds tend to occur in the C A M ,
particularly towards the top and in the middle parts. The upper part of the L C M m a y also be thicker-bedded than other parts of the succession. These thick beds are not strikingly different in appearance, although at M o n t e C o n e r o and in parts of the F u r l o - M o n t e Vicino area they are c o m m o n l y associated with thick-bedded
D . A . V . S t o w et al.
230
calcarenites and calcirudites. However, where calcarenites are not present, as at Gubbio, the two stratigraphically equivalent intervals in the CAM comprise thick-bedded calcilutites. The change in bed thickness is commonly (a)
BED o ,
i
lo i
1
|
THICKNESS 2O I
i
0
|
-III I I
lO t
i
20 i
i
30 i
I
40 I
' W
.[I
|
.
I '
i i It _1 I ,.i..I..
Ii :
I
(b)
(cm.)
30
I I
I I I I I
gradational over 5-40 m of section producing both thinning and thickening-upward sequences (Figs 5 and 6). In other cases, the thicker beds occur together in isolated packets, of 5-15 m in thickness, or more or less randomly through the succession.
m W
I I ,.1..I.. I
-J-q-
W
W W W W
w W
... "'I"1'I..."t
W
w
I:I: I..L..I..
W
W W W w w W
W W
w white bed . . . . . . calcarenite ---marl beds prominent (>3cm)
-I ':t'i' :::1:...F
w [80cm)
W
FIG. 6. Measured sections and bed thicknesses in the Furlo region: (a) Monte Cesana, Lower Cherty Member, thinning-upwards sequence over 15 m section (cf. Fig. 5(a)); (b) Monte Pietralata, Calcarenite Member, more irregular variation in bed thickness over 16 m section (cf. Fig. 5(b)). Only limestone beds measured (calcilutites and calcarenites): interbedded marls-shales mostly very thin ( < 5 ram) except as indicated towards top of Section (a).
Depositional model for calcilutites Composition The calcilutites are composed dominantly of CaCO3, with scattered pelagic foraminifera in a riannofossil or indeterminate micritic matrix. Re-crystallized silica occurs in minor amounts as sand-sized spheroidal blebs, probably after pelagic siliceous microfossils. Other minor non-carbonate components include feldspars, clay minerals, and rare magnetite, haematite and other heavy minerals, all of which are more abundant in the marlstones and shales than in the calcilutites. The organic-carbon content is very low, ranging from 0.07% in the calcilutites to 0.17% in some shales (Arthur & Fischer, 1977). X-ray diffraction analysis of some 60 samples of the fine ( < 4 #m) residues from crushed and acidified calcilutites shows the minor but significant presence of smectite (including expandable mixed-layer clays), illite, kaolinite, chlorite, quartz (probably biogenic silica), orthoclase and plagioclase. Non-acidified samples were also processed for the relative proportion of calcite to non-carbonates. The regional, stratigraphic and facies distribution of these components is illustrated in Figs 7 and 8 with respect to three geographical areas (Monte Conero, Furlo-Monte Vicino, Gubbio), four stratigraphic intervals (UCM, CAM and LCM of the Scaglia Rossa and the Scaglia Bianca), and red versus white calcilutites. In the east, Monte Conero is characterized by high or very high kaolinite/chlorite, orthoclase/ plagioclase and calcite/silica ratios, and by a relatively low silica/feldspar ratio. The central Furlo-Monte Vicino belt has a high kaolinite/ chlorite and more intermediate orthoclase/ plagioclase, calcite/silica and silica/feldspar ratios. And, furthest west near Gubbio, we find the lowest kaolinite/chlorite, orthoclase/plagioclase and calcite/silica ratios, but somewhat higher silica/feldspar ratio. The smectite/illite ratio also shows a slight decrease from east to west. A certain amount of local variability is noted within the Furlo-Monte Vicino belt but it is difficult to separate this clearly from the stratigraphic signature. The stratigraphic differences within the Scaglia Rossa formation are equally pronounced. The LCM and U C M (base and top) have relatively low kaolinite/chlorite, orthoclase/plagioclase and calcite/silica ratios but a higher silica/feldspar ratio. The intermediate CAM, by contrast, has higher kaolinite/chlorite, orthoclase/plagioclase and calcite/silica and lower silica/feldspar ratios. Smectite/illite appears to decrease slightly up-section. There were insufficient samples from the U M M and L M M to make an adequate compari-
2 31
son of mineral ratios, although they both appear to be somewhere in between the CAM and the cherty members. The Scaglia Bianca Formation is markedly different from the Scaglia Rossa, with a very much higher silica/feldspar ratio and low smectite/illite, kaolinite/chlorite and orthoclase/plagioclase ratios. The different facies show rather less clear mineralogical variation, though this may in part be due to an overprint of the regional and stratigraphic signatures. The few coarse-grained calcirudites and calcarenites analysed have very
SM/Z
50%
t 0
K/C
.....
,., (,.2)
.'"
".":':'.~'-"
5~176 t o.9 0
. . . .
-"
3.2 .
" '
" ," " , " "
so,~~
0
Q/F(O/P)
0.8
1.~ (2.5]
".=-"
""
1.8
I- , . , . . . ~ .
t.;:...L:
3.z (o.z)
. . . . . . .
0.6
0.6
:i9.1.'.:(0.3)
5O% SCAGLIA BIANCA (4)
SM
T
K
C
Q
F
FIG. 7. Calcilulite composition (X-ray diffraction results) and mineral ratios for four litho-stratigraphic intervals throughout the Umbria-Marche region. UCM (5 samples) Upper Cherty Member, CAM (28 samples) Calcarenite Member, LCM (13 samples) Lower Cherty Member, Scaglia Bianca (4 samples). SM = smectite + expanding mixed-layer clays, I = illite, K = kaolinite, C = chlorite, Q = quartz, F ---feldspar, O = orthoclase, P= plagioclase. Numbers indicate mineral ratios as shown. Percentages are based on weighted peak heights and are not considered absolute.
232
D.A.V.
s0%] 9
5o%
0
i
~
.
.
.
'
3.0
Q/F (o/1 o.7 (o.o)
56
,.2 (2.4)
.
,"
,'......,~
Stow
; . ; 9 .' .l
5O%
0
s0%t
1.o
1.3
(2.6)
e t al.
tish and reddish lithologies is that the former have more carbonate and the latter more terrigenous fine fraction. The mineral ratios show subtle and variable differences; for example, in the CAM the calcite/silica ratio is higher in the white than red facies, whereas the opposite is true in the LCM. Relatively few geochemical data are available for the Scaglia Rossa formation as a whole (Vannucci et al. 1979), although more attention has been paid to the section near Gubbio (Arthur 1976; Arthur & Fischer 1977) and to the Cretaceous-Tertiary boundary in particular (Alvarez et al. 1980; Wezel et al. 1981). Arthur (1976) has shown that geochemical profiles of metal contents (Fe, Ti, Ni, Mn), alumina and silica across limestone-shale couplets show maxima in the shales and minima in the middle of the limestone beds. Although the concentration in the shales is by a factor of 20 to 50, the total metal content in a given shale is about the same as that in the adjacent limestones, and concentration was probably the result of diagenetic dissolution of carbonate from the shales and its transfer to the adjacent limestones. There may have been some initial sea floor dissolution as well. Wezel et al. (1981) suggest that the enrichment of iridium in the shale interbed at the CretaceousTertiary boundary as well as at several other horizons within the Scaglia Rossa, is due to a combination of diagenetic concentration, very slow sedimentation rates and, possibly, volcanic contributions, rather than to extraterrestrial sources (Alvarez et al. 1980).
Colour 0~
. SM
.
. I
.
. K
. C
.
~ Q
F
FIG. 8. Calcilutite composition and mineral ratios for Calcarenite Member only of three separate regions, Monte Conero, Furlo-Monte Vicino and Gubbio, and for white and red lithologies of the Furlo-Monte Vicino area. See Fig. 7 caption for further explanation.
high calcite/silica ratios with little or no fine terrigenous fraction; the little there is shows high kaolinite/chlorite and orthoclase/plagioclase and low silica/feldspar ratios. The cherts, of course, are also distinctive; but the main difference between the calcilutites, marlstones and shales is simply one of CaCO3 content so that the calcite/ silica and CaCO3/terrigenous fraction ratios are highest in the calcilutites and lowest in the shales. The only consistent difference between the whi-
At the two ends of the colour spectrum of the Scaglia Rossa facies are the dark-brownish red shales and the white calcirudites and calcarenites. The calcilutites ranged through this spectum from darker red to pale red or rose and pinkish white to white. Most of the sequence at Monte Conero is white, including the calcilutites, whereas most of the Gubbio section is various shades of red and pink. In the central Furlo-Monte Vicino belt all the colours are represented and, most commonly, the white beds (coarse or fine-grained) are clearly distinct from the red. However, either white to red or red to white upward gradations can occur in the same bed and, in some cases, a mottled relationship is observed. As noted above, the compositional differences between red and white beds are gradational and slight. White contains more carbonate and red more terrigenous material as well as slightly more iron.
Depositional model for calcilutites
233
FIG. 9. Structural detail for calcilutites, shown by photograph of polished surface (left column), thin-section photograph of detail (right column) and sketch (centre column) for each of three Scaglia Rossa types. (a) massive, structureless, slightly mottled, very fine-grained; (b) burrowed, bioturbated, poorly-sorted, coarser-grained; and (c) indistinct or irregular lamination. Shown for comparison is the more clearly laminated calcilutite of the Scaglia Bianca (d).
234
D . A . V . S t o w et al.
Structure and texture After careful observations in the field, and of numerous polished slabs, X-radiographs and thin sections we can confidently say that the calcilutites are remarkably free from primary sedimentary structures. Apart from widespread bioturbation and burrowing (described below), there are three main structural types (Fig. 9). One is very fine-grained, massive and structureless, the second is more poorly-sorted, slightly coarsergrained, mottled and bioturbated, but otherwise structureless. The former rarely has indistinct banding or thin fine-grained foraminiferal lamination and more commonly some bioturbation, and the latter may show irregular patches and discontinuous lenses of coarser-grained foraminifera and wispy shale laminae. The third type shows very indistinct layering or lamination, picked out by foraminiferal concentrations, slight colour changes, shale wisps, grain alignment or an alternation of the other two structural types. All three types can occur in either reddish or whitish calcilutites throughout the Scaglia Rossa, and all gradations between the three are observed. A more systematic study than the present one might show a preferential regional or stratigraphic distribution of these three types. Our limited data suggest that the mottled, coarser-grained type is dominant overall, the massive, fine-grained type occurs preferentially in the CAM and in the Monte Conero area, and the indistinctly laminated type is most prevalent in the lower part of the LCM. At Monte Conero and in the Furlo-Monte Vicino area many of the clear turbiditic calcarenite and calcirudite beds pass upwards into the fine-grained massive calcilutite, and then this grades upwards into the coarser-grained mottled and bioturbated calcilutite (Fig. 10). These two structural types always occur in this same order and, in some cases, the lower one is white and the upper is red. The transition from calcarenite to calcilutite tends to be relatively abrupt, commonly disturbed by load-like and water-escape structures, and yet shows both compositional and positive textural grading. The transition from massive to mottled (white to red) calcilutites is completely gradational, with slight compositional and negative (reverse) textural grading (Fig. 10). Bioturbation and burrowing A wide range of burrows is found in Scaglia Rossa calcilutites, both on bedding surfaces and randomly throughout even the thickest beds (Figs 5 and 10). Most beds in fact are burrowed or, at
least, bioturbated to differing degrees. Where recognizable traces occur these include Zoophycos, Planolites, Chondrites, Helminthoida, Palaeodictyon, Granularia and Thalassinoides, several of which are systematic surface-combers that require relatively long intervals of time to thoroughly work over an area for food. No systematic variation of burrow types has been observed, but general bioturbation of the sediments appears to be slightly more pronounced in the Gubbio area and in the lower parts of the Scaglia Rossa (in the LCM, LMM and lowest part of the CAM).
Discussion Facies interpretation Most previous workers on the Scaglia Rossa have emphasized its pelagic character (e.g. Arthur 1976; Arthur & Fischer 1977) whereas Wezel (1979) argues that the formation is mostly turbiditic. Other studies have also pointed out the existence ofcalcarenite turbidites within a mainly pelagic sequence (Crescenti 1969; Crescenti et al. 1969). While there is little doubt that these coarser-grained lithologies were deposited by turbidity currents or related mass gravity-flow processes, our detailed investigations suggest that the more dominant fine-grained lithology (calcilutite or fine-grained limestone) cannot be explained simply as entirely turbiditic or entirely pelagic. We agree with previous arguments (Arthur & Fischer 1977) that the calcilutite-shale rhythmic bedding is due to a combination of episodic sedimentation and diagenetic dissolution of carbonate. There is certainly evidence of shale-filled burrows in the calcilutites indicating primary alternation of carbonate-rich and carbonatepoor episodes, whereas the compositional profile across a calcilutite-shale couplet, stylolites in the calcilutites and dissolution features in the shale (Wezel et al. 1981) all indicate that CaCO3 has been dissolved from the shales and partly transferred to the calcilutites. The rocks of the Scaglia Rossa formation and their mode of origin are undoubtedly very similar to many other limestone-shale sequences throughout the world (e.g. Hallam 1964; Fischer & Arthur 1977; Einsele 1982; Einsele & Seilacher 1982). However, our interpretation of the Scaglia Rossa departs from most earlier studies in suggesting, (1) that the input of both carbonate and of terrigenous material has fluctuated through different parts of the sequence; and (2) that the episodic increase in carbonate was influenced, not
Depositional model for calcilutites
235
FIG. 10. Structural, textual and compositional detail for two calcarenite-calcilutite turbidites, (a) from M o n t e Conero and (b) from Furlo. Composition of the five numbered levels shown in terms of mineral ratios: K = kaolinite, C = chlorite, Q = quartz (silica), F = feldspar, O = orthoclase, P = plagioclase, Ca = calcite. In each case, the shaded area refers to the first-mentioned mineral.
236
D . A . V . S t o w et al.
only by variation in biogenic productivity at the surface and by carbonate dissolution on the sea floor, but also by carbonate input into the system from turbidity currents. We summarize below the evidence for a turbidity current contribution to the Scaglia Rossa calcilutites. (1) Major variations in thickness (by a factor of 2 or 3) of different members and hence of sedimentation rates over a few kilometres in the Furlo-Monte Vicino area, particularly of the C A M and U M M . This seems difficult to explain by such extreme variation in pelagic sedimentation over such short distances. The succession at Monte Conero is thinner than those further west, which is in part due to internal erosion or non-deposition. (2) Coarse-grained calciturbidites (calcarenities and calcirudites) are common, particularly at Monte Conero and Furlo-Monte Vicino and in the CAM. It seems unlikely that these would be so prevalent and widespread and yet that fine-grained distal calciturbidites would be completely absent. (3) Locally there are clear slump and channel features, commonly involving calcilutites. This suggests local slopes, instability and downslope resedimentation. (4) The thickest calcilutite beds, presumably indicative of increased carbonate input, occur most commonly in the CAM member in association with coarsergrained calciturbidites. (5) The thinning-upward and thickeningupward sequences as well as the less regular variations in bedding thickness are strikingly analogous to those of clastic turbidite successions. (6) Many of the calcirudites contain shallowwater shelly debris, derived from the east, whereas the calcarenites contain much resedimented pelagic material in addition. A certain amount of terrigenous material was also being supplied to the basin(s), and we interpret the high kaolinite/chlorite and high orthoclase/plagioclase ratios as indicative of proximality to the land source. That these ratios are highest in the east, decreasing westwards, and higher in the CAM than the other members, suggest that pelagic-carbonate-charged turbidity currents incorporating the more proximal terrigenous material are most evident in the C A M and flowed from east to west. If the silica content of the calcilutites is interpreted as being derived from pelagic siliceous organisms, then the calcite/silica and silica/feldspar ratios support the concept of
more pelagic-dominated sedimentation in the west and also indicate greater primary productivity during deposition of the LCM and UCM. (7) A number of beds show a very clear sequence from the coarse-grained calcirudites/calcarenites through a very finegrained massive calcilutite to a coarsergrained, mottled and bioturbated calcilutite. The former we interpret as turbiditic and the latter as pelagic. The reverse-grading between the turbiditic and pelagic calcilutites is due to the presence of relatively large planktonic foraminifera (and recrystallized siliceous planktonics?) in the latter. These same two structural types, massive fine-grained and bioturbated coarse-grained, occur without the associated calcarenites throughout the Scaglia Rossa, and we presume are due to relatively more turbiditic and more pelagic inputs respectively. The indistinct lamination in some calcilutites is more difficult to interpret; in part it probably results from bioturbation, in part from diagenesis, and in part from current (turbidity current?) activity. Nevertheless, the overall aspect of the Scaglia Rossa calcilutites confirms the importance, and perhaps dominance, of pelagic deposition. (1) The very low rates of sedimentation are those most commonly associated with pelagites. (2) The limestone-marlstone (limestone-shale) rhythmic bedding, together with chert bands, is very typical of pelagic sequences. (3) The very high CaCO3 content, the silica of probable biogenic origin, and the abundance of planktonic fossils all indicate a pelagic composition. However, for the calcilutite turbidites, we suggest that it is largely pelagic material that has been redeposited. (4) The extensive bioturbation and burrowing throughout is reminiscent of many pelagites, but also occurs where there has been a very long time interval between turbiditycurrent events. The burrow assemblage appears rather mixed, but is mainly of the Nereites (bathyal) and Zoophycos (slope) zones (Seilacher 1967, 1978). However, both Crimes et al. (1981) and Wetzel (this volume) have shown that trace fossil assemblages in deep water are the result of a more complex set of variables than simple palaeobathymetry, including subenvironment, grain-size and sedimentation rate.
D e p o s i t i o n a l m o d e l f o r calcilutites Turbidite and pelagite characteristics We can identify, therefore, both turbiditic and pelagic calcilutites as distinct end members in the Scaglia Rossa Formation (Fig. 10). The turbidites are very fine-grained, micritic, very rich in CaCO3, probably originally mainly nannofossils and other fine carbonate detritus, and have a terrigenous signature in the small non-carbonate fraction. They are massive and structureless, with rare indistinct lamination and burrowing, and may be either white or pale red in colour. The pelagites are coarser-grained, similar in composition but with relatively less CaCO3, composed of nannofossils, large scattered planktonic foraminifera and recrystallized siliceous microplankton, and relatively more non-carbonate material with a non-terrigenous silicarich signature. They are mottled, extensively bioturbated and burrowed, and may have irregular foraminiferal concentrations and shale wisps. The characteristics we attribute to the turbidite end-member, although dissimilar from terrigenous mud turbidites (e.g. Piper 1978; Stow & Shanmugam 1980; Stow & Piper, this volume), have nevertheless been documented in a number of earlier studies of Palaeozoic to Recent limestones (e.g. Meischner 1965; Garrison 1967; Thomson & Thomasson 1969; Davies 1977; Anatra et al. 1980; Kelts & Arthur 1981, Faug6res et al. 1982 and this volume). All these authors describe fine-grained limestones or lime muds that are between 5 and 100 cm thick, even-bedded, commonly with thin argillaceous interbeds, massive, ungraded or with slight positive grading especially near the base, some with bioturbation near the top, and mainly of pelagic carbonate composition with or without a minor terrigenous component. They may occur individually or as part of clear calciturbidite beds (calcirudite-calcarenite-calcilutite), commonly in base-of-slope and basinal environments. Stanley's (1981) unifites and Wezel's (1973) omogeniti are both thick-bedded, featureless, homogeneous, calcareous muds which they interpret as a result of large fine-grained turbidity currents that became ponded in relatively small basins. Also recognized as fine-grained calciturbidites are beds that show distinctive current-induced structures, mainly parallel lamination, some of which are similar to our indistinctly-laminated structural type (van Andel & Komar 1969; Bissell & Barker 1977; Yurewicz 1977; Kennedy 1980; Faug6res et al. 1982 and this volume). However, many of the Scaglia Rossa calcilutites lie somewhere between the turbidite and pelagite end-member types, and thus appear to have had either a greater turbiditic or a greater
237
pelagic influence but not to be wholly one or the other. In fact, many previous authors have encountered similar problems in distinguishing between turbiditic and pelagic limestones and modern lime muds or oozes. Wilson (1969) drew attention to this in his description of deeper-water lime mudstones from basinal settings which he hesitated to interpret as true pelagic turbidites. Hesse (1975) attempted to give field and laboratory criteria for the recognition of fine-grained turbidites and non-turbidites in both carbonate and non-carbonate basins. His pelagic turbidites deposited in low-latitude basins above the carbonate compensation depth are very similar to our Scaglia Rossa ones, and show similar subtle differences from the interbedded pelagics. Carrasco (1977), Cook & Taylor (1977), Enos (1977) and Reinhardt (1977), in discussing thick limestone successions of various ages, all concluded that the dominant fine-grained facies was part pelagite and part turbidite (or contourite) but admitted difficulty in clearly distinguishing between the different processes. Homewood & Winkler (1977) recognized that the fine-grained parts of graded calcisiltite-calcilutite beds in Upper Jurassic rocks of the Pre-Alps were almost identical to the interbedded homogeneous pelagic calcilutites, and that there appears to be a complete gradation of facies and processes.
Depositional model for Scaglia Rossa It seems clear that deposition of the Scaglia Rossa formation took place in a basin, or more probably, a series of tectonically-controlled basins (Fig. 11) that became progressively deeper to the west, but that were never below the regional carbonate compensation depth. Sedimentation in the western (Gubbio) and central (Furlo-Monte Vicino) basins was dominantly pelagic during the LCM and again during the UCM periods (Fig. 11 (c)), with widespread accumulation of siliceous and calcareous oozes, now cherts and limestones. The rhythmic alternation of carbonate-rich and carbonate-poor facies was mainly due to the fluctuation in carbonate productivity, secondarily accentuated by diagenetic dissolution and transfer. There was also variation in turbidity current input of carbonate as, at Monte Conero in the east, there is a predominance of calciturbidites in the lower and middle parts of the LCM. The LMM and particularly the U M M mark a distinct change in the sedimentation regime, with the introduction of a large amount of terrigenous material (Fig. 1 l(a)). In the LMM this may in fact be a result of reduced carbonate input. The actual depositional processes of the marlstones and shales are unclear although the very slow sedi-
238
D . A . V . S t o w et al.
(a) narrow ~
pe~;g;c + h:m:pelagic s
" g d m'nant -~ L ~ _ - l a ' ~ , _~ ~
.qr_....~...--~ /
.....
9
9 ~'1 _o/
subdued tectonism
- ~ - ~i_-_~~ ~__.-;I*"/I_-_[_~_[- 1--]-*]*~ L - L-L.--'_I_ - A -__1_ 1-
_I - L-
high run-oTr
tcs minor~.~ . / . ~ ~,-~_-_~...-I_'IA-J-_N-I_-l-.]z
_~.m--~L-Z-_~
~
sheif
sea-level (?) low, productivity low
(b) s-I high, productivity medium
c) eL c t ~ ~(~ ~t "Jr ")c. L~ ~ ) ' ~ ) " pelagic
basinwide
(
settling "1" 8uspens'~
L
,
_ _ ~ L
?2
D
|
broad shelf
r ;I;t;orm ; i n ; ; .~ ~ t c fine tail
t c s , Lsl~
~.~
~
~
low
er~176 ~_~)~_..) "~ "~ / /
--
L'F 1-1_ ;,i-i.:l~
~.---
(trap;ed) run-off
active
i oo
,oc,on,oo
(c) s-I high, productivity high
pelagic settling dominant
broad shelf
.~___~.
~ tcs ~
~__tTr~~-~#-;
"
_
.-.~..
I-\7 7 L-": r: "-/
I T,-'N'~- 3 -: r --.I -.I "--It- I-- .-.
~--77/1~-I ~
EI-I~F~F-.I"
tc ( s)=turbidity current(s) calcilutite
I-~
"
"
/" . .
.
.
..--
run-off (trapped)
subdued
tectonism
s-I=sea-level
calcarenite ~
calcirudite I - ~
chert
I~
shale
FIG. l l. Depositional model and hypothetical cross-sections for the Scaglia Rossa Formation: (a) Upper Marly and Lower Marly members, (b) Calcarenite Member, (c) Upper Cherty and Lower Cherty Members. Model (b) is believed to be appropriate for many deep-water carbonate systems with evidence of turbidity current input. mentation rates suggest that they were, in large part, hemipelagic. It is interesting to note that both members, the L M M in mid-Cenomanian and the U M M at end-Maastrichtian, occur at times of global lowering of sea-level (Vail et al. 1977), though whether these eustatic changes would have had any influence on terrigenous supply in this region is uncertain. Certainly the occurrence of slumping in these members sug-
gests local tectonic activity was also important. The particularly abrupt and widespread lithological change at the C A M - U M M boundary may also be marked by a discontinuity. Deposition of the middle C A M (late CamCampanian-Maastrichtian) was more variable regionally and, on average, three to five times more rapid than the other members (Fig. 1 l(c)). A shallow-water carbonate platform in the east
Depositional model for calcilutites (present Adriatic Sea) shed material across its (faulted?) margin in mass gravity-flows. Thick calcarenitic turbidites were deposited up to 50 km to the west in the Furlo-Monte Vicino basin, although a more local platform origin for some of these turbidity currents was deposited directly over the coarser-grained calciturbidite, but much of it rapidly dispersed into the water column as a basin-wide suspension that settled more slowly and mixed with the background pelagic material. Thus the CAM, characterized by an increased rate of sedimentation, thicker calcilutite beds, higher CaCO3 content and a more proximal terrigenous signature, was greatly influenced by turbidity current input of fine-grained carbonates. The dispersion and relatively slow-settling of this fine material makes it difficult to distinguish from the background pelagic sediments and a true gradation of facies and processes appears likely. In the CAM, therefore, much of the increased carbonate of the limestone-shale couplets was due originally to turbidity current input.
Broader implications This depositional model (Fig. l l(b)) for finegrained and pure limestone (i.e. without significant terrigenous components), involving a mixed turbiditic-pelagic process, we believe is more widely applicable. There are many examples in the literature of slope and basinal carbonate systems where the facies recognized include resedimented calcirudites, calcarenites, calcidebrites, slump deposits and interbedded, distal pelagic limestones (e.g. Wilson 1969; Price 1977; Hubert et al 1977; Crevello & Schlager 1980). It seems clear to us that the calcilutite turbidites are conspicuously missing from or have been overlooked in these successions. Even McIlreath & James (1978) in their synthesis of facies models for carbonate slopes describe coarse peri-platform talus, lime breccias and graded calcarenites, but then simply pelagites and hemipelagites for the fines. Fine-grained pure carbonate, carried into a basin by a turbidity current or similar process, is
239
more likely to disperse into the water column than the equivalent fine-grained terrigenous material. This is because CaCO3 does not develop to the same degree the surface electrostatic charges that are so important in the flocculation of aluminosilicate clay minerals (van Olphen 1978; Flugel 1982). Hence, the fine-grained and low-concentration tails of carbonate-charged turbidity currents will mix with mid-basin pelagics and surface or mid-water suspensions from marginal carbonate platforms and settle slowly through the water column. This kind of process is very similar to that envisaged from the modern periplatform oozes of the Bahama basins (Crevello & Schlager 1980; Crevello et al. this volume; Heath & Mullins, this volume) and also postulated for some Mesozoic limestones in south central Turkey (Waldron 1981). Piper (1978) observes that carbonate commonly occurs towards the top of mixed clay-carbonate mud turbidites. Nisbet (1974) and Robertson (1976) also draw attention to the probable difference in hydrodynamic behaviour of carbonate and terrigenous grains, when discussing the origin of calciturbidites in Cyprus and Greece. It is not surprising, therefore, that calcilutite turbidites are difficult to recognize in both modern and ancient sediments. Careful study of regional, compositional, geochemical, textural and structural criteria, however, should allow the recognition of, at least, the end-member turbidites and pelagites, and some estimation of the relative importance of each process. ACKNOWLEDGMENTS:This is the first publication resulting from a cooperative project between the geology departments of Edinburgh and Urbino. D.A.V.S. acknowledges financial support from the Natural Environment Research Council while carrying out the field-work and research. S.C.R. received field expenses from British Petroleum, UK. F.C.W. and D.S. thank the CNR, Italy for research funds. We should also like to thank the secretarial and technical staff at the Grant Institute of Geology for typing, drafting and analytical work, in particular Patricia Stewart, Diane Baty and Colin Chaplin. Finally, our thanks to Peter Homewood and Michael Arthur for critically reviewing the manuscript.
References ALVAREZ,W. & LOWRIE,W. 1981. Upper Cretaceous to W.M. 1977. Upper Cretaceous-Palaeocene magEocene pelagic limestones of the Scaglia Rossa are netic stratigraphy at Gubbio, Italy. V. Type section not Miocene turbidites. Nature, 294, 246-8. for the Late Cretaceous-Palaeocene geomagnetic , ARTHUR, M.A., FISCHER, A.G., LOWRIE, W., reversal time scale. Bull. geol. Soc. Am., 88, 383-9. NAPOLEONE,G., PREMOL1SILVA,I. 8~ ROGGENTHEN ALVAREZ, L.W., ALVAREZ,W., ASARO,F. & MICHEL,
240
D . A . V . S t o w et al.
H.V. 1980. Extraterrestrial cause for the Cretaceous-Tertiary extinction: experimental results and theoretical interpretation. Science, 208, 1095-1108. ANATRA, S., ACKERMANN,T. & HOMEWOOD, P. 1980. Les faci& de l'Ultrahbv&ique du Monsalvens (Pr6alpes Externes) et de la r6gion d'Anzeinde (Pr6alpes Internes). Flog. geol. He&., 73/1, 283-92. ARTHUR M.A. 1976. Sedimentology of the Gubbio sequence and its bearing on palaeomagnetism. Mem. Soc. geol. Ital., 15, 9-20. --& FISCHER A.G. 1977. Upper Cretaceous-Paleocene magnetic stratigraphy at Gubbio, Italy. I. Lithostratigraphy and sedimentology. Bull. geol. Soc. Am., 88, 367-71. BISSELL, H.J. & BARKER,H.K. 1977. Deep-water limestones of the Great Blue Formation (Mississippian) in the eastern part of the Cordilleran miogeosyncline in Utah. Soc. econ. Palaeo. Min. Spec. Pub., 25, 171-86. BORTOLOTT1, V., PASSERIN1,P., SAGR1,M. & SESTINI,G. 1970. The miogeosynclinal sequences, In: Sestini, G. (ed.), Development of the northern Apennines geosyncline: Sed. Geol., 4, 341-4. CARRASCO, V, B. 1977. Albian sedimentation of submarine autochthonous carbonates, east edge of the Valles-San Luis Potosi platform, Mexico. Soc. econ. Palaeo. Min. Spec. Pub., 25, 263-72. COCClONI, R. 1978. Et~i miocenica della Scaglia Rossa del Bottaccione (Gubbio). Ateneo Parmense, Acta Nat., 14, 223-30. - - . 1983 Foraminiferi planctonici miocenici anomali nella Scaglia Rossa Umbro-Machigiana. Boll. Soc. Paleont. It., 22, l, in press. COOK, H.E. & TAYLOR, M.E. 1977. Comparison of continental slope and shelf environments in the Upper Cambrian and lowest Ordovician of Nevada. Soc. econ. Palaeo. Min. Spec. Pub., 25, 51-82. CRESCENTI,U. 1969. Stratigrafia della serie calcarea dal Lias al Miocene nella regione marchigiano-abruzzese (Parte I--Descrizione delle serie stratigrafiche). Mere. Soc. geol. Ital., 8, 155-204. , CROSTELLA,A., DONZELLI,G. & RAFFI, G. 1969. Stratigrafia della serie calcarea dal Lias al Miocene nella regione marchigiano-abruzzese (Parte II-Lithostratigrafia, biostratigrafia, paleogeografia). Mere. Soc. Geol. Ital., 8, 343-420. CREVELLO, P.D. & SCHLAGER, W. 1980. Carbonate debris sheets and turbidites, Exuma Sound, Bahamas. J. sed. Petrol., 50, 1121-48. CRIMES, T.P., GOLDRING, R., HOMEWOOD, P., VAN STUIJVENBERG,J. & WINKLERW. 1981 Trace fossil assemblages of deep-sea fan deposits. Gurnigel and Schlieren flysch (Cretaceous--Eocene) Switzerland. Ecol. geol. Helv., 74, 953-95. DAVIES, G.R. 1977. Turbidites, debris sheets and truncation structures in upper Palaeozoic deep-water carbonates of the Sverdrup Basin, Arctic Archipelago. Soc. econ. Palaeo. Min. Spec. Pub., 25, 221-48. EINSELE, G. 1982. Limestone-marl cycles (periodites): diagnosis, significance, causes--a review. In: Einsele G. & Seilacher A. (eds), Cyclic and Event Stratification. Springer-Verlag, Berlin. 8-53. - & SEILACHER A. (eds) 1982. Cyclic and Event Stratification. Springer-Verlag, Berlin. 536 pp.
ENOS, P. 1977, Tamabra limestone of the Pozo Rico trend, Cretaceous, Mexico. Soc. econ. Palaeo. Min. Spec. Pub., 25, 273-314. FAUGERES, J.C., GAYET, J., GONTHIER, E., POUTIERS,J. NYANG, I. 1982. La Dorsale M6dio-Atlantique entre 43 ~ et 56~ facies et dynamique s6dimentaire dans plusieurs types d'environnements au Quaternaire R&ent. Bull. Inst. Geol. Bassin d'Aquitaine, Bordeaux. 31, 195-215. FISCHER, A.G. & ARTHUR, M.A. 1977. Secular variations in the pelagic realm. Soc. econ. palaeo. Min. Spec. Pub., 25, 19-50. FLUGEL, E. 1982. Microfacies Analysis of Limestones. Springer-Verlag, Berlin. 633pp. GARRISON, R.E. 1967. Pelagic limestones of the Oberalm beds (Upper Jurassic-Lower Cretaceous), Austrian Alps. Bull. Can. Petrol. Geol., 15, 21-49. HALLAM, A. 1964. Origin of the limestone-shale rhythms in the Blue Lias of England: a composite theory. J. Geol., 72, 157-68. HESSE, R. 1975. Turbiditic and non-turbiditic mudstone of Cretaceous flysch sections of the East Alps and other basins. Sedimentology, 22, 387-416. HOMEWOOD, P. & WINKLER, W. 1977. Les calcaires detritiques et noduleux du Maim des M6dianes Plastiques dans les Pr6alpes fribourgeoises. Bull. Soc. Frib. Sc. Nat., 66, 116--40. HUBERT, J.K., SUCHECKI,R.K. & CALLAHAN,R.K.M. 1977. The Cowhead Breccia: sedimentology of the Cambro-Ordovician continental margin, Newfoundland. Soc. econ. palaeo. Min. Spec. Pub., 25, 125-54. KELTS, K. & ARTHUR, M.A. 1981. Turbidites after ten years of deep-sea drilling - wringing out the mop? Soc. econ. Pataeo. Min. Spec. Pub., 32, 91-127. KENNEDY,W.J. 1980. Aspects of chalk sedimentation in the southern Norwegian offshore. In: The Sedimentation of North Sea Reservoir Rocks. Norwegian Petroleum Society. LOWRIE, W. & ALVAREZ,W. 1977. Upper CretaceousPalaeocene magnetic stratigraphy at Gubbio, Italy. III. Upper Cretaceous magnetic stratigraphy. Bull. geol. Soc. Am. 88, 374-7. LUTERBACHER, H.P. & PREMOLI SILVA, I. 1962. Note preliminaire sur une revision du profil de Gubbio, Italic. Riv. Ital. Palaeont., 68, 253-88. & PREMOLISILVA,I. 1964. Biostratigrafia del limite Cretaceo-Terziario nel-l'Appennino Centrale. Riv. Ital. Palaeont., 70, 67-128. MCILREATH,I.A. & JAMES,N.P. 1979. Facies models 13. Carbonate slopes. Geoscience Canada, 5, 184-99. MEISCHNER,K-D. 1965. Allodapische Kalke, Turbidite in Riff-N/ihen Sedimentations-Becken. In: Bouma A.H. & Brouwer, A. (eds), Turbidites. Elsevier, Amsterdam. 156--91. NISBET, E.G. 1974. The Geology of the Neraida area, Othris Mountains, Greece. Unpub. Ph.D. Thesis, University of Cambridge. PIPER, D.J.W. 1978. Turbidite muds and silts on deep-sea fans and abyssal plains. In: Stanley D.J. & Kelling G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. PREMOLI SILVA, I. 1977. Upper Cretaceous-Palaeocene -
-
Depositional model for calcilutites magnetic stratigraphy at Gubbio, Italy, II. Biostratigraphy. Bull. geol. Soc. Am., 88, 371-4. --, PAGG1, L. & MONECHI, S. 1976. Cretaceous through Palaeocene biostratigraphy of the pelagic sequence at Gubbio, Italy. Mem. Soc. geol. ltal., 15, 21-32. PRICE, I. 1977. Deposition and derivation of clastic carbonates on a Mesozoic continental margin, Othris, Greece. Sedimentology, 24, 529-46. REINHARDT,J. 1977. Cambrian off-shelf sedimentation, central Appalachians. Soc. econ. Palaeo. Min. Spec. Pub., 25, 83-112. RENZ, O. 1936. Stratigraphische und mikropalaentologische Untersuchung der Scaglia (Obere KreideTertiar) im Zentralen Apennin. Ecol. geol. Hell'., 29, 1-149. 1951. Ricerche stratigrafiche e micropaleontologiche sulla Scaglia (Cretaceo superiore-Terziario dell'Appennino centrale: Mem. Descr. Carta Geologica d'Italia, 29, 1-173 (Italian trans, of Renz, 1936). ROBERTSON, A.H.F. 1976 Pelagic chalks and calciturbidites from the lower Tertiary of the Troodos Massif, Cyprus. J. sed. Petrol.. 46, 1007-16, SEILACHER,A. 1967. Bathymetry of trace fossils. Marine Geol., 5, 413-29. --. 1978. Use of trace fossil assemblages for recognizing depositional environments. In: Basan, P.B. (ed.) Trace Fossil Concepts. Soc. econ. Paleo. Min. Short Course No 5, Oklahoma. STANLEY, D.J. 1981 Unifites: structureless muds of gravity-flow origin in Mediterranean basins. Geomarine Letters, 1, 77-83. STOW, D.A.V. & SHANMUGAM,G. 1980. Sequence of structures in fine-grained turbidites: comparison of recent deep-sea and ancient flysch sediments. Sed. Geol., 25, 23-42. THOMSON, A.F. • THOMASSON,M.R. 1969. Shallow to deep water facies development in the Dimple Limestone (Lower Pennsylvanian), Marathon Region, Texas. Soc. econ. Palaeo. Min. Spec. Pub., 14, 57-78. VAIL, P.R., MITCHUM, R.M., THOMPSON, S., TODD, -
-
24I
R.G., SANGREE,J.B., WIDM1ER,J.M., BUBB, J.N. & HATFIELD, W.G. 1977. Seismic stratigraphy and global changes of sea-level. Am. Ass. Petrol. Geol. Mere., 26. VAN ANDEL T.H. & KOMAR, P.D. t969. Ponded sediments of the mid-Atlantic ridge between 22 ~ & 23 ~ North latitude. Bull. geol. Soc. Am., 80, 1163-90. VANNUCCI S., VANNUCCI R., MAZZUCOTELL1 A. & FRANCH1 R. 1979. Risultati preliminari dello studio petrografico e geochimico della Scaglia Bianca e della Scaglia Rossa della sezione del Bottaccione (Gubbio). Ateneo Parmense, Acta Nat., 15, 261-6. VAN OLPHEN, H. 1978. Clay Colloid Chemistry. 2nd edition. Interscience, New York. 300pp. WALDRON, J.W.F. 1981. Mesozoic Sedimentation and Tectonic Evolution o f the N E Anatalya Complex, Eyridir, S W Turkey. Unpub. Ph.D. Thesis, Edinburgh University. W1LSON, J.L. 1969. Microfacies and sedimentary structures in 'deeper-water' lime mudstones. Soc. econ. Palaeo. Min. Spec. Pub., 14, 4-19. WEZEL F.C. 1973 Diacronismo degli eventi geologici oligo-miocenici nelle Maghrebidi. Rivista Mineraria Siciliana. Anno. XXIV, N. 142-144, 219-232. 1979. The Scaglia Rossa Formation of Central Italy: results and problems emerging from a regional study. Ateneo Parmense, Acta Nat., 15, 243-59. 1981. Upper Cretaceous to Eocene pelagic limestones of the Scaglia Rossa are not Miocene turbidites (Alvarez & Lowrie), a reply. Nature, 294, 248pp. - - , VANNUCC1,S. & VANNUCCI,R. 1981. Decouverte de divers niveaux riches en iridium dans la 'Scaglia rossa' et la 'Scaglia bianca' de l'Apennh d'OmbrieMarches (Italie). C.R. Acad. Sci. Paris, 293, II, 837M4. YUREWICZ, D.A. 1977. Sedimentology of Mississippian Biasin carbonates, New Mexico and west Texas-the Rancheria Formation. Soc. econ. Palaeo. Min. Spec. Pub., 25, 203-20. -
-
D.A.V. STOW, S.C.R. RAINEY & G. ANGELL, Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW. Scotland. F.C. WEZEL & D. SAVELLI, Istituto di Geologia dell'Universitfi di Urbino, Urbino 61029, Italy.
N o r t h Atlantic contourites: an overview D.A.V. Stow and J.A. Holbrook SUMMARY: We present a brief overview of known bottom circulation and Tertiary to Recent sediment drifts in the North Atlantic Ocean, and outline the distinctive features of contourite sediments recovered from these drifts. There is a gradation of structural and textural characteristics between the finer-grained clayey and muddy contourites and the coarser-grained silty and fine-sandy contourites. There is also a gradation between the three compositional types identified: biogenic, mixed biogenic/terrigenous, and volcanogenic. Other features such as microfabric, or facies sequences are less well documented but potentially diagnostic.
Bottom circulation Our knowledge of deep-water circulation in the North Atlantic is perhaps better than for any other large ocean basin (e.g. Berggren & Hollister 1974, 1977; Shor & Poore 1978; Schnitker 1980). At the present day, several different water masses are involved in this circulation. The deepest, coldest water mass is the Antarctic Bottom Water (AABW), formed in the Weddell and Ross Seas bordering Antarctica and flowing north into both the western and eastern North Atlantic basins. Dense, cold water is similarly formed in the Greenland and Norwegian Seas and flows south through the Denmark Strait, the Faeroes Bank Channel and across several sills on the IcelandFaeroes Ridge. This Norwegian Sea Overflow Water (NSOW) mixes partially with Labrador Sea Water (LSW) formed south of Greenland and with AABW to form the North Atlantic Deep Water (NADW). Warmer saline Mediterranean Sea Water (MSW) flows out of the Straits of Gibraltar and spreads out at intermediate levels to both north and south. The possibility that MSW contributes to a relatively strong Eastern Boundary Undercurrent (EBUC) in the NE Atlantic as far north as the Rockall Trough has been noted by several authors (Swallow et al. 1977; Ellett et al. 1979). Bottom currents associated with the movement of these water masses have now been measured in many parts of the ocean (see Table 1 in Stow & Lovell 1979) mainly with short-duration current meters, and the complex circulation shown in Fig. 1 has been demonstrated. Velocities across the Greenland-Iceland-Faeroes Ridge are up to 30 cm/s. In the Hatton Rockall and Rockall Basins velocities average 5-15 cm/s, being consistently higher (2(~30 cm/s) in the core of the current and lower (0-10 cm/s) at the margins and over depositional ridges. Low velocity eddies spill over the ridge crests and flow perpendicular to the contours. A cyclonic loop of N A D W in the
Rockall Basin results in a narrow (10-20 km) current of up to 20 cm/s flowing northwards along the eastern margin of the Basin. Measurements over 9 months show that the NSOW flows predominantly westwards through the CharlieGibbs Fracture Zone with mean seasonal velocities of 0-8 cm/s and maximum of 20 cm/s. Similar velocities have been recorded from an anticyclonic loop of NSOW and LSW in the Labrador Sea. Numerous measurements in the western North Atlantic have shown persistent westerly and southwesterly currents of 5-20 cm/s associated with the Western Boundary Undercurrent (WBUC) (e.g. Shor et al. this volume). Over an 8 month period on the continental rise south of Cape Cod, Luyten (1977) demonstrated a mean westward flow of 5 cm/s and velocity pulses of up to 40 cm/s to both east and west. Both long-term and short-term measurements have confirmed bottom velocities beneath the Gulf Stream of 5-15 cm/s at depths of 3600-5000 m, with current eddies of up to 30 cm/s. Generally sluggish currents have been reported from the east central North Atlantic. However, bottom current velocities associated with the MSW are up to 300 cm/s in the Straits of Gibraltar and from 10-80 cm/s along various parts of the south Iberian margin in the Gulf of Cadiz (e.g. Faug+res et al. 1984; Gonthier et al. this volume). Although the movement of water masses can be generalized in terms of a bottom current system and short-duration current, measurements have yielded the sorts of values outlined above. It is important to note the great variability in time and space of such currents as well as the overall complexities of the circulation pattern. The currents vary from a few kilometres to tens of kilometres in width and flow at different levels within the water column. They flow downslope as well as along-slope, and large competent eddies peel off to move at right angles to the main flow. Insufficient measurements have been made to 245
246
D.A.V. Stow and J.A. Holbrook 8,0~
70
60
I
50
40
30
20 GREENLAND SEA
I /
ENLAND //
NORWEGIAN 9 . SEA
J
,/1# #
,
I
/1
LABRADOI:
/
v -~
v)oR ~ 2 t..J s I
I
o
/
,./
\... ~
//
,\ GADR
I
FA[
7
/
I
o
/t
t
%
I
/
) ,! I
I
I
i
FIG. 1. North Atlantic present-day deep-water circulation (arrows, see text) and the major sediment drifts (close stipple). FAD = Faro Drift, FD = Feni Drift, HD = Hatton Drift, G D - - G a d a r Drift, BD = Bjorn Drift, SD = Snorri Drift, ED = Eirik Drift, GRD = Gloria Drift, NOR = Newfoundland Outer Ridge, CR = Corner Rise, BR = Bermuda Rise, G S O R = Gulf Stream Outer Ridge, BBOR = Blake Bahama Outer Ridge, GAOR = Greater Antilles Outer Ridge, COR = Caicos Outer Ridge. Areas of bottom water formation shown by wide stipple.
d e t e r m i n e any characteristic periodicity in flow velocity, but b o t h seasonal (Shor et al. 1980) and tidal ( M c C a v e et al. 1980) fluctuations have been noted. Velocities decrease from the core to the margin o f the current, and reverse flows are c o m m o n l y measured. It appears that a variable period and variable d i a m e t e r eddy-like a d v a n c e is
m o r e typical o f deep b o t t o m currents than simple u n i f o r m unidirectional flow.
Sediment
drifts
We also show in Fig. 1 the sediment drifts in the N o r t h Atlantic that owe their origin largely to the
North
Atlantic
contourites:
system of bottom currents that has been developing since the early Tertiary. Most of these drifts have been at least sparsely sampled with piston and gravity cores or Deep Sea Drilling Project drill cores. We summarize below brief descriptions of the sediments recovered, together with the supporting evidence that allows their interpretation as contourites. In some of these examples there are intervals ofinterbedded slumps, debrites and turbidites. This list is based on that compiled by Stow (1982) with certain recent additions.
Feni Drift (Pujol et al. 1974; Faug6res et al. 1979, 1981; Dingle et al. 1981)
Sediments: calcareous silts, clays and oozes: gradational contacts between beds, homogeneous; fine-grained; mixed biogenic and terrigenous; sedimentation rate 6-12 cm/1000 yrs. Other evidence: inferred bottom currents, drift morphology, sediment waves, seismic characteristics.
Hatton Drift (Laughton et al. 1972; Shor & Poore 1978; Roberts & Montadert 1979; Montadert et al. 1979; McCave et al. 1980)
a n ot, e r v i e w
247
Bjorn Drift (Laughton et al. 1972; McCave et al. 1980)
Sediments: muds and calcareous muds; homogeneous, bioturbated, rare thin laminae; finegrained; biogenic and terrigenous, foram sand fraction; sedimentation rate 12 cm/1000 yrs. Other evidence: inferred bottom currents, drift morphology, seismic characteristics. West Reykjames Ridge (Luydendyk et al. 1978; Shor & Poore 1978)
Sediments: calcareous ooze and chalk, siliceous chalk; homogeneous, bioturbated, rare thin irregular coarse laminae; dominantly biogenic, some terrigenous. Other evidence: inferred bottom currents, drift morphology, seismic characteristics. Eirik Drift (Chough 1978; Latouche & Parra 1979; Chough & Hesse, in press)
Sediments: calcareous muds; intensely bioturbated and mottled, rare wavy parallel lamination, rare laminated coarse layers; fine-grained, poorly-sorted; biogenic and terrigenous. Other evidence: measured bottom currents, drift morphology, seismic characteristics.
Sediments: calcareous muds, oozes and chalk, diatomaceous chalk; mainly homogeneous and bioturbated, also wavy and contorted lamination (paper-thin lamination and other microstructures in diatomaceous chalk?); fine-grained with foram sand fraction; mixed biogenic and terrigenous: sedimentation rate 2 5 cm/1000 yrs. One sandy contourite identified, with 35-70'~o foram sand. Other evidence: measured bottom currents and nepheloid-layer, drift and small-scale morphology, seismic characteristics, hiatuses.
Sediments: calcareous muds; homogeneous, bioturbated; fine-grained; biogenic and terrigenous. Other evidence: measured bottom currents, drift morphology, seismic characteristics.
Gardar Drift (Shor 1980: Faug6res et al. 1979: McCave et al. 1980)
Sediments: calcareous mud; mainly homogeneous, some more sandy and silty intervals; fine-grained; biogenic and terrigenous. Other evidence: measured bottom currents and nepheloid-layer, drift morphology, seismic characteristics.
Sediments: calcareous muds; homogeneous; mixed biogenic and terrigenous: sedimentation rates 25--40 cm/1000 yrs. Sandy contourites identified, muddy sands 1- 10 cm thick; bioturbated and burrowed, no primary structures; fine sandsized, poorly sorted, positively skewed; mixed biogenic and terrigenous, broken and ironstained foram tests; sedimentation rate 2 cm/1000 yrs. Other evidence: measured bottom currents and nepheloid-layer, drift and small-scale morphology, seismic characteristics, hiatuses.
Gloria Drift (Laughton et al. 1972)
Newfoundland Ridge (Auzende et al. 1970; Pastouret et al. 1975; Latouche & Parra 1979)
Bermuda Rise and Corner Rise (Peterson & Edgar 1968; McGregor et al. 1973; Silva et al. 1976; Laine & Hollister 1981)
Sediments: calcareous muds; homogeneous and bioturbated; fine-grained with coarse biogenic debris; biogenic and terrigenous. Other evidence: measured bottom currents and nepheloid-layer, non-uniform drift morphology.
248
D.A.V. Slow and J.A. Holbrook
Blake-Bahama Outer Ridge (Heezen et al. 1966: Hollister et al. 1972; Klasik & Pilkey 1975; Flood 1981; Flood & Hollister 1980; Gradstein et al. 1981)
Sediments: calcareous clays and muds and marls: homogeneous, bioturbated, mottled, rare thin ungraded coarse layers, no other primary structures; mainly fine-grained, some with up to 20% foram sand, moderately poorly sorted: biogenic and terrigenous, carbonaceous in parts, cyclic vertical variations in composition in parts, alongslope compositional variation in parts. Other evidence: measured bottom currents and nepheloid-layer, large-scale drift and wide range small-scale morphological features, seismic characteristics, hiatuses.
Greater Antilles Outer Ridge (Bader et al. 1970: Tucholke 1973, 1975: Tucholke & Ewing 1974)
Sediments: calcareous muds and silty clays; homogeneous, mottled, rare concentrations of biogenic material in irregular layers; fine-grained, minor biogenic sand; biogenic and terrigenous, cyclic variations in carbonate content, carbonaceous in parts. Other evidence: measured bottom currents and nepheloid-layer, drift and small-scale morphology, seismic characteristics.
Gilliss Seamount (Taylor et al. 1975)
Sediments: calcareous muds and marls: homogeneous, structureless; fine-grained with foram sand fraction; biogenic and terrigenous, numerous broken foram tests: sedimentation rate 1 cm/1000 yrs. Other evidence: inferred bottom currents, drift morphology on flanks of seamount, seismic characteristics. Great Meteor Seamount (Von Stackelberg et al. 1979)
Sediments: calcareous oozes; bioturbated, rare lamination: fine-grained with nannofossil silt and foram sands; mainly biogenic, minor terrigenous. Other evidence: inferred bottom currents, drift morphology on flanks of seamount. Gibbs Fracture Zone (Faug~res et al. 1979; Shor et al. 1980)
Sediments: foram sands. Other evidence: measured bottom currents, drift and erosional morphology.
Faro Drift (Mougenot & Vanney 1982; Faug~res et al. 1984; Gonthier et al. this volume).
Sediments: calcareous muds, silty muds, silts and very fine sands; homogeneous, mottled, and remnants of primary structures: mostly intensely bioturbated: fine clay to fine sand-size, poor to moderately sorted: mixed planktonic/benthonic biogenic and terrigenous composition: regular horizontal variations and irregular vertical "sequence" of facies noted. Other evidence: measured bottom currents and nepheloid-layer, drift morphology, bottom photographs of current-induced structures, seismic characteristics.
Continental rise contourites Along much of the western North Atlantic continental margin, as well as along parts of the eastern margin, it appears that contourites are closely interbedded with resedimented and hemipelagic facies on the continental rise. The contourites described by Pastouret el al. (1978), Stow (1979) and Stanley el al. (1981) from these settings are closely comparable with those of the sediment drifts outlined above. However,, other authors have interpreted a facies comprising silt and fine-sand laminae and thin cross-laminated beds interlayered with mud as being of contourite origin (e.g. Heezen el al. 1966; Hollister & Heezen 1972: Bouma & Hollister 1973; Barrusseau & Vanney 1978). The distinction of facies types on the continental rise is still a matter of some controversy and is the subject of detailed study in an area of the Nova Scotian continental rise (HEBBLE study area, Shot et al., this volume). Until this problem is resolved we base our sedimentary characteristics of contourites on the less equivocal sediment drift examples.
Sedimentary characteristics The following is a synthesis of the sedimentary characteristics of contourites based on numerous descriptions of sediment drift samples (references as for individual drifts), and on previous syntheses by Stow & Lovell (1979), Lovell & Stow (1981) and Stow (1982). It does not apply to the gravel-lag contourites of narrow passages and straits which have only rarely been sampled. Contourite facies (Fig. 2)
Muddy contourites (silt and clay grade), sandy contourites (mainly fine sand grade) and gravel-
N o r t h Atlantic" contourites." an overview
I
i
249
,,_l
I0 cm
FIG. 2. Photographs of typical contourite facies: (a) muddy contourite, (b) mottled silty-muddy contourite, (c) sandy contourite (lower part). All from Faro Drift, Gulf of Cadiz. Core width approximately 10 cm. lag contourites (coarse sand and gravel grade) can be distinguished. These form a continuum from purely depositional fine-grained facies to purely erosional coarse lag deposits. Other facies divisions that have been used are mottled silts and muds and silty contourites.
Bedding style and thickness (Figs 2 & 3) Muddy contourites commonly occur in thick monotonous sections (from a few centimetres to tens of metres) with poor or absent bedding. Where silty or sandy contourites occur, they form thin irregular distinct to gradational layers (commonly < 3 cm thick), or thicker often distinct beds (commonly 10-30 cm).
Facies 'sequences' (Fig. 3). A characteristic succession of muddy, silty and fine sandy facies in positive and negative gradational sequences has been observed. These sequences are highly variable in thickness ( < 10 cm to > 100 cm) and often incomplete in the range of facies and structures present. A typical coupled negative-positive sequences is shown in Fig. 3 (from Gonthier et al., this volume).
Primary sedimentary structures (Figs 2 & 3) Irregular coarser (often shelly) concentrations, silty lenses and laminae occur in all facies types.
Wavy or wispy lamination is commonly noted in parts of the muddy and silty-mud facies; regular horizontal lamination is rare. Silt and fine-sand facies are commonly massive (bioturbated), or more rarely with internal horizontal and crosslamination. Layer contacts between facies can be entirely gradational, sharp and flat, irregular or erosive. Both tops and bottom of layers show these contact types with equal regularity; the same contact can vary in nature across the width of a core.
Bioturbation (Figs 2 & 3) One of the most characteristic features of contourites is their extensive bioturbation throughout. This bioturbation was clearly a continuous process, in many cases with several superimposed episodes, that has modified or destroyed much of the primary sedimentary structure, partially or completely altered the nature of the contacts, and is probably responsible in large part for the melange of silt, sand and mud that is commonly observed. Bioturbational mottling occurs at a pm, mm and cm scale together with iron sulphide mottles and filaments (mycelia). Distinct burrow types are also recognized, including Chondrites, Planolites, Scolicia, Teichichnus and Zoophycos. There appears to be a relationship between burrow and mottle size and type and the grainsize of the sediment (Faug6res et al. 1984).
D.A.V. Stow and J.A. Holbrook
250
Texture (Fig. 4) LITHOLOGY and STRUCTURE
MEAN GRAIN SIZE 4
8 16 32 64,urn
50
maximum velocity
In nearly all the samples studied to date there is a spectrum of grain-sizes from fine silt and clay grade (mean 8 /~m) to coarse silt/fine sand-size (mean 40-70/~m). They are mostly poorly sorted; some silts and fine sands are moderately well sorted. The shape of the cumulative frequency grain-size curves (following the terminology of Rivi+re 1977) are commonly parabolic-hyperbolic (coarse and fine tails) for the silts and sands, more or less hyperbolic (fine tail) for mixed silty-muddy facies, and close to logarithmic (uniform size distribution) for the finer muddy contourites. Progressive increase or decrease in grain-size (i.e. both positive and negative grading) is common in all facies on a small-scale, over a few centimetres, as well as over 50-100 cm of section.
Composition (Fig. 5)
~
bioturbation
Composition depends very much on the material available to the bottom current for transport and deposition. In the North Atlantic contourite drifts, there appear to be three end-member compositions: biogenic, terrigenous and volcanogenic. The pure end-member types are relatively rare, although pure foraminiferal sands and volcanogenic silty muds have both been described. Most commonly, there is a mixture of benthonic and planktonic biogenic tests with terrigenous and/or volcanogenic material. Part of the material may be far travelled (over 1000 km) and part may be derived locally, either in situ benthonic or overlying planktonic organisms. Current transport and long residence times on or near the bottom, causes fragmentation of biogenic tests and iron staining of grains. Organic carbon contents are commonly low (about 0.5%) Diagenetic minerals, especially iron sulphides and clays, are commonly present in small amounts.
~
discontinuous silt lenses
Fabric
....
gradational contact
~
sharp (irregular) contact
There have been few systematic fabric studies to date. Preliminary results suggest that elongate silt or fine sand grains and the anisotropy of magnetic susceptibility both probably parallel the bottom current direction. However, it is not clear what overall orientation pattern is attained by grains deposited from a bottom current that shows large eddies and reversals of direction. Limited scanning electron microscope studies indicate that the contourite mud fabric may be characterized by small particle clusters and individual clay plates relatively well aligned, parallel to bedding. How-
D p
silt and sand mud shell fragments
F1G. 3. Typical vertical succession or 'sequence' of contourite facies over 50 cm of section (from Faug6res et al., 1984, Faro Drift). Muddy, mottled silt and mud, and silty/free sandy contourite facies shown. Note variations in grain-size, sedimentary and bioturbational structures through the section.
North Atlantic contourites." an overview
25I
,~176 1 ~~1 60 % 40
20
. l ' I O0
I
I
I
i
I i i ;C,
60
,
!
i
i 4
!
~
l
ffm
I00-
80
60
4 0 84
i
(b) Blake Bahama O,Jter Ridge (mixed biogenic- terrigenous)
O,
/~d ' ' d o
.
.
.
.
~,'
. . . .
:..
'
'
~m
I00"
{'"~"~~.
8o
,., ":! % 40
(c) Feni, Hatton,Snorri Drifts (biogenic-rich, solid lines) Vogel
and Nashville
Seamount
Drifts
{volcanogenic-rich, dashed lines)
i i i i i i I00 60
9
i
i
i i I
I0
i
i
i
i
4
i
i
!
,Lzm
FIG. 4. Cumulative frequency grain-size distribution curves for North Atlantic contourites; semi-logarithmic plots. Examples from: (a) Faro Drift, (b) Blake-Bahama Outer Ridge and (c) NE Atlantic Drifts and Seamounts. (sources: Gonthier et al., this volume; Stow & Holbrook, unpublished data; Chough & Hesse, in press).
252
D.A.V. Stow and J.A. Holbrook 100% BIOGENICS b~~.....~
Feni, Snorri and Hatton Drifts
"~~....Gloria and Bjoro Drifts /
100%
, , ~".I~
V~rVVVt #'VVVV VvVvVVV vvv
Feni and Faro Drifts and Blake-Bahama O,,ter Ridge
VOLCANOGENIC ,,~,,,~iiii~i~~ (+SMECTITES)
X
100% TERRIGENOUS
Vogel and Nashville Seamount Drifts Flo. 5. Composition of North Atlantic contourites. Generalized triangular plot of three main components: biogenic debris, terrigenous material and volcanogenic material (including smectite clays). Based on smear slide estimates, and sand-fraction grain mounts and X-ray diffraction data (Stow & Holbrook, unpublished data). ever, this may depend to some extent on the pre-flocculated nature of suspended materials introduced to the bottom current from different sources. Both silt and mud fabrics will clearly be disturbed by subsequent bioturbation. Sedimentation rate
The rates of accumulation of many North Atlantic contourite drifts have been determined from either carbon 14 or micropalaeontological dating of cored sections (Davies & Laughton 1972: Hollister et al. 1972; Klasik & Pilkey 1975; Sheridan, Gradstein et al. 1981). These rates, mainly for Pliocene to Recent sections, mainly lie between about 5 cm and 10 cm/1000 yrs, ranging from as little as 1 cm/1000 yrs to around 15 cm/1000 yrs. Similar values are obtained from estimates of seismic thickness and stratigraphic age for certain drifts, and for contourites of the continental rise (e.g. Stow & Lovell 1979).
Discussion Processes
Bottom currents have an important effect on the nature and distribution of sediments in the North
Atlantic Ocean at the present day. Bottom circulation is complex, involving several different water masses, large areas of the ocean where flow is negligible, and other areas where flow velocity can reach over 100 cm/s for short periods of time. Long-term current measurements show that bottom currents vary greatly in time and space, having tidal, seasonal and irregular periodicities, current reversals and an eddy-like flow advance. Many bottom currents are capable of eroding, transporting and depositing fine-grained sediments (up to about fine or medium sand-size), of reworking medium and coarse-grained sands, and of winnowing finer material away from coarser-grained sandy and gravelly sediments. Bottom currents transport fine suspended material (average size about 12/,m) in thin to thick (110 m to over 1500 m?), very low concentration (0.01 to 0.3 rag/l), bottom nepheloid-layers, in some cases transporting the finest material over thousands of kilometres. Suspended sediment enters the nepheloid-layer from bottom current erosion, from sediment resuspension due to burrowing organisms, internal tides, etc. from the fine tails of turbidity currents, and from the vertical settling of hemipelagic and pelagic material through the water column. A relatively weak bottom current will only slightly effect this material during its
North Atlantic contourites. an overview
0 o
-
<
III
>-
< rr
~> -
~8
~
m
_0
r~
,,,
<
:~
_(2
-
Y'
~o
+
o ~
u_
<+s ~ ~ ~
+< ~ ~
+ ~zm o ,~ ~ <~ >__,
o
253
O_
~
<~
o~
~S
m
++
+o ~ z
-
@
z
0 o
+--
--
O
~
--
0
~-~
4
.~
m+s
~D --
r3';
z
~_
m
0
--
~1
u_ m
8
U_J
.B
o~ ~.,<~+~i.ii!1 I!:::~,.::.<~ i.'' ", ,. I~.I :I l~,/:,l,.l~i!/}il~.i'~ +0 o +,.:+.,,.,+5.i I}t(;i+.,. I-, ,',;~:/..1511:IaX.,/.IX~..., I +0 ,
II
i
.:'
'-
., l
:,
.,,0
.
;
,,
,:
;.
OO "ulD
8
Q
Z
<( rr
Z
co
~o
O
_0 m
Z
O
o
m
go~S9 0
z
B
+++.~~ ~ + o +. q ~
o++,+ ooo
u_
0, ++
n-
z _0
+ ? ~ y~ ,,< L.Uo_+ ~ LU
+~+~+:+
_
r rr
< o ,.u o. o ,,+._ +. <, w_S o2. =- >_ m ~< z .+ uu o
.+,,eo
<
~ ~
El
~
~+F:+-
~
(o
~-
<(
<m~>~
rr
o., .--
_+o m
+
+~
-~-
Oo Z
z
+ + + + M M
o
So
~ ~
,.?_.< 0m
-++? +-
<(
'
~176
+8++~
O O
M <
~ u_
<<:+~
+
+,--+
U.I
B
.+..+
0
lzI
,
I+., I'
I !. I{I I : :..-..". I -.:; I
i 9
+
I
,d
+L "U.ID
254
D.A.V. Stow and J.A. Holbrook
passage to the sea floor, and so in many cases the depositional process must be considered intermediate between bottom current and pelagic, hemipelagic or turbiditic, depending on the main mode of sediment supply. Stronger bottom currents will have a more marked effect on the sediment during its deposition. Facies
Given the complexity of bottom current flow and the common interaction of several different depositional processes in the deep sea, it is hardly surprising that there is a gradation of contourite facies as well as gradations between contourite and pelagic, hemipelagic or turbidite facies. Two main contourite facies can be identified as resulting from deposition by bottom currents: muddy contourites and sandy contourites (Figs 6 & 7). Muddy contourites are fine-grained, mainly homogeneous or structureless and thoroughly bioturbated, more rarely having irregular layering, lamination and lensing. They range from finer-grained homogeneous muds to siltier mottled silts and muds. Their composition varies with the primary source material, but is most commonly mixed biogenic and terrigenous. They most closely resemble hemipelagites. Sandy contourites occur as thin irregular layers ( < 1 to 5 cm) or thicker beds (5-25 cm), that are either structureless and thoroughly bioturbated or with some primary horizontal and cross-lamination preserved. They can show both negative and positive grading, or both, and have sharp or gradational bed contacts. Grain-size is commonly fine sand, more rarely medium sand, with poor to moderate sorting. In many cases the mean grain-size is the coarse silt grade and the facies may be more accurately termed 'silty to fine sandy' contourites. The composition is variable, commonly mixed terrigenous and biogenic. The facies may some-
times be confused with fine-grained turbidites. Muddy and sandy contourites commonly occur together in sediment drifts in irregular vertical ~sequences'. The effects of winnowing and reworking by bottom currents can result in contourite facies with rather different characteristics. Thin, irregular, poorly-sorted, structureless, mixed-composition, iron-manganese coated, coarse-sandy and gravel-lag contourites are formed by the winnowing and removal of all fines from a coarse-grained sediment by powerful bottom currents. The more or less in situ reworking of sandy turbidites can result in a bottom current modified turbidite sand. These are believed to be common on continental slopes and rises. Ancient contourites
We now have a relatively well-documented suite of sedimentary characteristics for contourites, based mainly on data from present-day sediment drifts but which we believe is applicable to contourites from other depositional settings. In theory, therefore, we should be able to recognize contourites with some degree of confidence from ancient rocks on land or in drill-cores from the oceans. In practice, however, these sedimentary criteria are not always definitive and may be partially lost during burial and diagenesis. The cautious triple-scale approach proposed by Lovell & Stow (1981) should still be followed in any identification of ancient contourites. ACKNOWLEDGEMENTS: We gratefully acknowledge secretarial, technical and drafting assistance at the Grant Institute of Geology. Financial support was provided by the National Environment Research Council and the Nuffield Foundation. Jean-Claude Faug6res and Alan E. Kemp reviewed an earlier version of the manuscript.
References (ed.), Studies in Paleoceanography. Soc. econ. Paleo. Min. Spec. Pub. 20, 126-86. marge du Grand Banc et la fracture de Terre Neuve. 1977. Plate tectonics and palaeocirculationC.R. Acad. Sci. Paris, 271, 1063-66. commotion in the ocean. Tectonophysics, 38, 11BADER, R.G. et al. 1970. Site 28. In: Bader, R.G. et al., lnit. Repts. DSDP, 4. US Govt. Print. Off., Wash48. BOUMA, A.H. & HOLLISTER,C.D. 1973. Deep ocean ington, DC. 125-43. basin sedimentation, ln: Middleton, G.V. & Bouma, BARUSSEAU,J.P. & VANNEY,J.R. 1978. Contribution ~i A.H. (eds.), Turbidites and Deep Water Sedimenl'6tude du mod61e des fonds abyssaux. Le r61e tation. 79-118. g6odynamique des courants profonds. Revue G~og. CHOUGH, S.K. 1978. Morphology, Sedimentary Facies phys. Gdol. dyn., 20, 59-94. and Processes of the North West Atlantic Mid-Ocean BERGGREN, W.A. & HOLL1STER,C.D. 1974. PalaeogeoChannel between 61 and52 N, Labrador Sea. Ph.D. graphy, palaeobiogeography and the history of circulation in the Atlantic Ocean. h~: Hay, W.W. Thesis, McGill University. 167 pp.
AUZENDE, J.M., OLIVET, J.L. & BONNIN, J. 1970. La
North Atlantic, contourites. an overview --
& HESSE, R., in press. Contourites from the Eirik Ridge, south of Greenland. Sed. Geol. DAVIES, T.A. & LAUGHTON, A.S. 1972. Sedimentary processes in the North Atlantic. In: Laughton, A.S. Berggren, W.G. et al. (eds), lnit. Repts. DSDP, 12, US Govt. Print. Off., Washington, DC. 905 34. DINGLE, R.V., MEGSON, J.B. & SCRUTTON, R.A. 1981. Acoustic stratigraphy of the sedimentary succession west of Porcupine Bank, NE Atlantic Ocean: a preliminary account. Marine Geol., 47, 17-35. ELLETT, D.J., DOOLEY, H.D. & HILL, H.W. 1979. Is there a northeast Atlantic shore current'? ICES, CM 1979, C:35, I l pp. FAUGERES, J.C., STOW, D.A.V. & GONTHIER, E., 1984. Contourite drift moulded by deep Mediterranean outflow. Geology. et al. 1979. Evolution de la s6dimentation profonde au Quaternaire r6cent dans le Bassin nordatlantique: corps s+dimentaires et s6dimentation ubiquiste. Bull. Soc. gOol. Fr., 21(5), 585-601. , GONTH1ER, E., GROUSSET,F. & POUTIERS,J. 1981. The Feni Drift: the importance and meaning of slump deposits on the eastern slope of the Rockall Bank. Marine Geol., 40, M49-M57. FLOOD, R.D. 1981. Longitudinal triangular ripples in the Blake-Bahama Basin. Marine Geol., 39, M 13-M20. & HOLLISTER, C.D. 1980. Submersible studies of deep-sea furrows and transverse ripples in cohesive sediments. Marine Geol., 36, M l-M9. GRADSTEIN, F.M., SHERIDAN, R.E. et al. 1981. Leg 76: Western North Atlantic Ocean. Joides Journal, VIII, 29-35. HEEZEN, B.C., HOLLISTER, C.D. & RUDDIMAN, W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science, 152, 502-8. HOLLISTER, C.D. & HEEZEN,B.C. 1972. Geologic effects of ocean bottom currents: western North Atlantic. In: Gordon, A.L. (ed.), Studies in Physical Oceanography, 2, 37-66. -EWING, J.I. et al. 1972. Init. Repts. DSDP, 11, US Govt. Print. Off., Washington, DC. HUGHES, T., DENTON, C.H. & GROSSWALD, H.G. 1977. Was there a Late-Wiirm Arctic Ice Sheet'? Nature, 266, 596-602. KLASIK, J.A. & PILKEY, O.H. 1975. Processes of sedimentation on the Atlantic continental rise off the southeast United States. Marine Geol., 19, 69-89. LAINE, E.P. & HOLLISTER,C.D. 1981. Geological effects of the Gulf Stream system on the northern Bermuda Rise. Marine Geol., 39, 277-310. LATOUCHE, C. & PARRA, M. 1979. La s6dimentation au Quaternaire R6cent dans le 'Northwest Atlantic Mid-Ocean Canyon' - apport des donn6es min6ralogiques et g6ochimiques. Marine Geol., 29, 137-64. LAUGHTON, A.S., BERGGREN, W.A. et al. 1972. Init. Repts. DSDP, 12, US Govt. Print. Off., Washington DC. LOVELL, J.B.P. & STOW, D.A.V. 1981. Identification of ancient sandy contourites. Geology, 9, 347-9. LUYTEN, J.R. 1977. Scales of motion in the deep Gulf Stream and across the Continental Rise. J. mar. Res., 35, 49-74. -
-
-
-
255
LUYENDYK, B.P., CANN, J.R. et al. 1978. Init. Repts. DSDP, 49, US Govt. Print. Off., Washington DC. MCCAVE, I.N., LONSDALE, P.F., HOLLISTER, C.D. & GARDNER, W.D. 1980. Sediment transport over the Hatton and Gardar contourite drifts. J. sed. Petrol., 50, 1049-62. MCGREGOR, B.A., BETZER,P.R. & KRAUSE,D.C. 1973. Sediments of the Atlantic Corner Seamounts: control by topography, palaeowinds and geochemically directed modern bottom currents. Marine Geol., 14, 179-90. MOUGENOT, D. & VANNEY, J.R. 1982. Les rides de contourites Plio- Quaternaires de la pente continentale sud-Portugaise. Bull. Inst. Geol. Bassin d'Aquitaine, 31, 131-9. MONTADERT, L., ROBERTS,D.G. et al. 1979. Init. Repts. DSDP, 48, US Govt. Print. Off., Washington DC. PASTOURET, L., AUEERET, G.A., HOEFERT, M., MELGUEN, M., NEEDHAM, H.D. & LATOUCHE, C. 1975. Sedimentation sur la Ride de Terre-Neuve. Can. J. Earth Sei., 12, 1019-35. - - , AUFFRET,G.A. & CHAMLEY,H. 1978. Microfacies of some sediments from the western North Atlantic: palaeoceanographic implications (Leg 44 DSDP). In: Benson, W.E., Sheridan, R.E. et al., Init. Repts. DSDP, 44, US Govt. Print. Off., Washington DC. 477-501. PETERSON, M.N.A. & EDGAR, T.N. (eds) 1968. Init. Repts. DSDP, 2. US Govt. Print. Off., Washington DC. PUJOL, C., DUPRAT,J., GONTHIER,E., PEYPOUQUET,J.C. & PUJOS-LAMY,A. 1974. R6sultats pr61iminaires de l'6tude effectube par l'Institut de g6ologie du Bassin d'Acquitaine concernant la mission Faegas dans l'Atlantique nord-est. Bull. Inst. Geol. Bassin Acquitaine, 16, 65-94. RIVIERE, A. 1977. M~thodes Granulom~triques: Techniques et Interpretation. Masson, Paris. 170 pp. ROBERTS, D.G. & MONTADERT, L. 1979. Margin palaeoenvironments of the northeast Atlantic. In: Montadert, L., Roberts, D.G. et al. (eds), Init. Repts. DSDP Leg 48, US Govt. Print. Off., Washington DC. 1099-118. SCHNITKER, D. 1980. Global palaeoceanography and its deep water linkage to the Antarctic glaciation. Earth Sci. Rev., 16, 1-20. SHERIDAN, R.E., GRADSTEIN, F.M. et al. 1982. Early history of the Atlantic Ocean and gas hydrates on the Blake Outer Ridge: results of the DSDP Leg 76. Bull. geol. Soc. Am., 93, 876-85. SHOR, A.N. 1980. Bottom Currents and Abyssal Sedimentation Processes South o f Iceland. Ph.D. Thesis, Woods Hole Oceanographic Institution. 246 pp. SHOR, A.N. & POORE, R.Z. 1978. Bottom currents and ice rafting in the North Atlantic: interpretation of Neogene depositional environments of Leg 49 cores. hi: Luyendyk, B.P. & Cann, J.R. et al., Init. Repts. DSDP, 49, US Govt. Print. Off., Washington DC. 859 pp. SHOR, A.N., LONSDALE, P., HOLLISTER, C.D. & SPENCER, D. 1980. Charlie-Gibbs fracture zone: bottom-water transport and its geologic effects. Deep-Sea Res., 27A, 325-45. SILVA, A.J., HOLLISTER,C.D., LAINE, E.P. & BEVERLY,
256
D.A.V. Stow and J.A. Holbrook
B. 1976. Biotechnical properties of deep-sea sediments: Bermuda Rise. Mar. Geotechnol., I, 195-232. STANLEY,D.J., SHENG, H., LAMBERT,D.N., RONA, P.A., MCGRAIL, D.W. & JENKYNS, J.S. 1981. Currentinfluenced depositional provinces, continental margin off Cape Hatteras, identified by petrologic method. Marine Geol., 40, 215-35. STOW, D.A.V. 1979. Distinguishing between finegrained turbidites and contourites on the deepwater margin off Nova Scotia. Sedimentology. 26, 371-87. 1982. Bottom currents and contourites in the North Atlantic. Bull. Inst. Geol. Bassin d'Aquitaire, 31, 151-66. LOVELL,J.P.B. 1979. Contourites: their recognition in modern and ancient sediments. Earth Sci. Rev., 14, 251-91. SWALLOW, J.C., FOULD, W.J, & SAUNDERS,P.M. 1977. Evidence for a poleward eastern boundary current in the North Atlantic Ocean. ICES, CM 1977, C:32, 21 pp. -
-
&
TAYLOR, P.T., STANLEY, D.J., SIMKIN, T, & JAHN, W, 1975. Gillis seamount: detailed bathymetry and modification by bottom currents. Marine Geol., 19, 139-57. TUCHOLKE, B.E. 1973. The Histo O' of Sedimentation and Ahvssal Circulation on the Greater Antilles Outer Ridge. Ph.D. Thesis, Cambridge, Woods Hole, MIT, WHOI. 1975. Sediment distribution and deposition by the Western Boundary Undercurrent: The Greater Antilles Outer Ridge. J. Geol., 83, 177-207. & EWlNG, J.I. 1974. Bathymetry and sediment geometry of the Greater Antilles Outer Ridge and vicinity. Bull. geol. Soc. Am., 85, 1789-1802. VON STACKELBERG,U., VON RAD, U. & ZOBEL,B. 1979. Asymmetric sedimentation around Great Meteor Seamount (North Atlantic). Marine Geol., 33, 117-32.
-
-
-
-
STOW, D.A.V. & HOLBROOK,J.A. Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, Scotland.
Contourite or turbidite?: magnetic fabric of fine-grained Quaternary sediments, Nova Scotia continental rise A.N. Shor, D.V. Kent and R.D. Flood SUMMARY: Samples of three piston cores and one gravity core from the Nova Scotia continental rise (depths 4210~1925 m) have been examined to differentiate parallel-to-slope and downslope depositional processes in Quaternary deposits from a region presently influenced by a strong contour current. Measurement of anisotropy of magnetic susceptibility of samples of a red-brown, silt-laminated lutite 'contourite' facies shows grain alignments which are consistent with both parallel-to-slope (contour current) flow and downslope (turbidity current) flow. We believe that these results provide support for the hypothesis that 'geologically significant" contour currents have influenced continental rise deposition during the Pleistocene. However, our observation that both alongslope and downslope alignments are present in lithologically similar units clearly demonstrates the need for studies on the relationship between lithofacies and process in this geological setting.
Contourites are sedimentary deposits in the deep ocean which are deposited by thermohalineinduced currents. The term 'contourite' was proposed by Hollister (1967) to describe a specific silt-laminated red-brown lutite facies observed in pre-Holocene samples from the continental rise offeastern North America. Subsequent investigations in regions of measured or inferred abyssal current flow have expanded the term to include a wide variety of sedimentary deposits ranging in texture from clay to sand-size, and with various structures and sorting characteristics (see Stow & Lovell 1979, for a review). Recent investigations of the facies first described as ~contourites' by Hollister (1967) have questioned the role of contour currents in formation of the silt laminae and reinterpreted these deposits as fine-grained turbidites (Stow 1979; Stow & Bowen 1980). The purpose of our present study is to identify sedimentary sequences which have been deposited beneath abyssal thermohaline flow so as to provide a baseline for subsequent descriptions of sedimentologic characteristics of the contourite facies. Few students of sedimentation processes on the eastern Canadian margin would doubt that both turbidity currents and contour-following thermohaline bottom currents have been active agents of transport and deposition at various times in the past. Recent studies by HEBBLE (High Energy Benthic Boundary Layer Experiment) investigators demonstrate the presence of "geologically significant' modern contour-following flow over the lower continental rise off Nova Scotia (Richardson et al. 1981). However, a fundamental question remains as to which process dominated on the continental rise during Pleisto-
cene low stands of sea-level when sediment influx rates were substantially higher than today (Stanley et al. 1972), and when different abyssal circulation patterns may have prevailed in the Atlantic. The method we use for identifcation of current-deposited sediments is the anisotropy of magnetic susceptibility (AMS), which is a measure of the preferential alignment or fabric of non-equidimensional silt- to sand-sized magnetic grains within a clay sediment matrix. The theoretical and experimental basis for using AMS as a determinant of flow during deposition of sediments is summarized by Hamilton & Rees (1970). The technique has been applied to abyssal marine sediments to confirm the sedimentologic prediction of turbidity current deposition in La Jolla Submarine Canyon (Rees et al. 1968), to define regional patterns of current flow (Ellwood 1980; Auffret et al. 1981; Bulfinch et al. 1982; Rees et al. 1982), and to indicate changes in current strength with time through restricted abyssal passages (Ellwood & Ledbetter 1979; Ledbetter & Ellwood 1980). These latter studies, however, have relied mainly on the magnitude of anisotropy rather than its direction. Ellwood & Ledbetter (1979) report ambiguous results in alignment directions from Vema Channel. We report here on our initial results from examination of AMS parameters in three piston cores and one gravity core from the Nova Scotia continental rise (water depths 4210-4925 m). Piston core samples are mainly from pre-Holocene deposits, and include both silt-laminated and visually structureless clays. Gravity core samples include both Holocene calcareous muds and pre-Holocene (Wisconsin?) gravel-rich clays. 257
258
A.N. Shor, D.V. Kent and R.D. Flood
Additional samples from a transect of piston cores from 4045 to 5025 m across the rise and from a set of box cores from 4800 to 5050 m are presently being examined.
Regional setting Bathymetric studies (using surface-ship echosounding) and side-scan sonar studies (using the SIO/MPL Deep Tow instrument; Spiess & Mudie 1970) show the presence of several lower rise channels and numerous shallow distributary channels or channel segments between 62 ~ and 63~ (Figs 1 & 2). Overall relief on the lower rise is extremely subdued, with regional gradients typically between 1:200 and 1:400. Channels rarely exceed 50 m in depth and have low, gently sloping walls (slopes typically < 1:). Where we have bottom photographs a dominantly clay
interface with contour-parallel bottom currentgenerated bedforms was observed even in channel thalwegs. The sediment cores come from relatively wellknown positions with respect to the local topography. The gravity core was located within one of the shallow distributary channels in the Deep Tow survey area (Fig. 2). The three piston cores were obtained from outside the Deep Tow survey area (Fig. 1). Although they are not precisely located relative to channel features, the shipboard 3.5 kHz records near the piston core sites do not show channel relief, and acoustic penetration is substantially greater than within the surveyed region to the west. Graded beds and thick sand layers are rare to absent in piston cores RC22-02, 14 and 15. These three cores were selected for study in part because they apparently do n o t represent channelized flow deposits.
FIG. 1. Bathymetry of the Novia Scotia continental margin (m). Rose diagrams at piston core sites summarize AMS alignment (kmax) in 10 intervals. Directions illustrated for gravity core EN-078-GGC 1 are mean values for three Holocene (Unit 1) and six pre-Holocene (Unit 3) samples.
Contourite or turbidite? 9 l- ....
..... !............
259 T
J
i
i
\
\
/'
1 , E
_EAVOR- 07 i
__l
:
I
:
+/
,
,
/i/ I
: /
A ~ O ~x
I S
',
'
40~
_
.
---t
~-~1~d~ <
UNIT
3
II
I
62
~
w
l
FIG. 2. Bathymetry near the site of the gravity core, illustrating distributary channel morphology. Contours from SIO/MPL Deep Tow survey. Ship was navigated relative to bottom transponders. Drift during coring is indicated as dotted tract. AMS alignments for the gravity core (kmax) as in Fig. 1.
Holocene sediments Holocene sediments were recovered in a series of four box cores in the vicinity of the Deep Tow surveys on the lower rise (depths 4843-5048 m) collected on Knorr cruise 83. These cores contain lithologies ranging from fine-grained muds to contorted silts. The description of these sediments is briefly summarized here to demonstrate the
complexity of depositional sequences and processes during the Holocene. The topmost surface layer in all box cores is a pale brown, faintly laminated slick clay from 2 to 10 cm thick, generally devoid of sand or silt, but with stringers of foraminiferal sand. This uppermost unit appears to be the one which is moulded into longitudinal bedforms by abyssal current flow (Tucholke 1982).
260
A.N. Shor, D.V. Kent and R.D. Flood
In two cores deeper than 5035 m it is underlain by a very stiff layer composed predominantly of silt-sized mineral grains, variable in thickness (5-15 cm thick) and strongly laminated with steeply dipping or strongly contorted beds. In the deepest core the lower contact is angular and irregular, and it contains clasts of the underlying clay. This unit is interpreted as a recent debrisflow, probably of limited extent, which may be related to the 1929 Grand Banks earthquake. The principal Holocene sediment underlies these units and is a moderate yellowish-brown, foraminifera-rich, extensively bioturbated clay. Few primary sedimentary structures are preserved, although some horizontal banding is present. The unit ranges from 15 to more than 30 cm thick, and contains Holocene foraminifera (G. menardii is present). It is the uppermost unit recovered in the gravity core and piston core 15 (piston cores 14 and 02 recovered no Holocene sediment). A pale brown clay with faint horizontal lamination, similar in general texture and appearance to surficial sediments, underlies the calcareous clay, and grades downward into a complex group of mud, silt, and fine sand layers from 5 to 10 cm thick which range in colour from pale brown to dark yellow-brown. This coarse unit is laminated and cross-bedded, with cross-beds highlighted by pale brown clay. Foraminifera recovered from the sand layers are Holocene, indicating deposition within the last ~ 11 000 yr. This unit appears to be the 'spillover sand' reported by Stanley et al. (1972), and it was recovered in the gravity core.
Channel deposits (core EN-078-GGC 1) A single large-diameter gravity core was recovered from the vicinity of 40~ 62~ I'W (depth 4830 m) during Endeavor cruise EN-078 (Fig. 2). This gravity core is about 20 km from the nearest box core discussed above. The ship position was determined relative to the Deep Tow survey through ranging on two longlife transponders remaining at the site, so that the core location is known relative to bathymetry defined by Deep Tow, except for problems introduced by drift during coring. The core lies near the base of the western edge of a 15 m deep, 1 km wide distributary channel. Three lithologies are recognized in the 83 cm long gravity core. The uppermost 17 cm is an olive-grey to yellowish-brown foraminiferal clay, of which the ~20% coarse fraction consists principally of foraminiferal fragments and terrigenous sand. Whole foraminifera are present, but not common. Globorotalia menardii are present in
this unit, indicating a Holocene Age (<11 000 yr). Neither biologic nor physical sedimentary structures are obvious visually or in the X-radiograph, although a grossly parallel-laminated fabric is indicated. Unit 2, from 17 to 22 cm, is a graded sand unit principally composed of quartz and feldspar with ancillary rock fragments and terrigenous mineral grains, foraminifera and minor authigenic grains. At least two cross-bedded units are present in the upper 2 cm of the unit. Structures cannot be defined in the basal section due to partial collapse (a result of freezing during transport from the ship to the laboratory in mid-January, and subsequent thawing in a horizontal position). The sand is very poorly consolidated, and cross-bedding structures show no evidence of biogenic disruption. G. menardii are present in this unit, indicating a Holocene age; Uvigerina spp. suggest a source on the continental slope ( < 2500 m). We interpret this layer as a Holocene turbidite. Unit 3, from 22 cm to the base of the core, is an olive-grey gravel-rich clay. Carbonate content is very low. No foraminifera are observed in coarsefraction separates. Angular clasts up to 4 cm in diameter are very common, and include basalt and one mollusc fragment. This gravel-rich layer is quite different from glacial-aged sediments recovered in piston cores from elsewhere on this rise, and was not recovered in the box cores.
Pre-Holocene sediments (piston cores) Over much of the Nova Scotia rise the pre-Holocene sediments consist of brown to reddishbrown slightly calcareous clays intercalated with discrete silt or sand layers. Although thick, coarse sand units are observed (and in some cases visibly graded (Hollister 1967; Horn et al. 1971 ; Hollister & Heezen 1972), the prevalent facies consists of thin, parallel, silt laminae (generally <1 cm thick), commonly in clusters, interbedded with a slightly calcareous silty clay. Silt layer frequency is highly variable, and in some instances may correlate from core to core (compare, for example, silt layer frequency in cores 14 and 15; Figs 4 & 6). Neither the age nor the origin of these deposits is well established, nor is it certain that the clay units and interbedded silt laminae are in fact derived from the same process. Both frequency of silt beds and individual bed thickness vary considerably. Average silt layer thicknesses appear to relate to proximity to major submarine channels (thicker beds near channels) based on examination of sand and silt layer frequency in the upper 5 m of 12 piston cores obtained during cruise RC22-09. In contrast, the frequency of occur-
Contourite or turbidite? fence of coarse layers shows a pattern related to depth of water rather than proximity to source. Caution must be exercised in interpreting this type of data, however, because neither stratigraphic continuity nor accumulation rates are established for these deposits, and individual cores show pronounced variations downcore in both coarse layer frequency and thickness. Previous interpretations of the silt-laminated pre-Holocene clays of the Nova Scotia rise have differed. Hollister (1967) suggests that they are most likely formed by bottom flows reworking the sediments, whereas Stow (1979) argues that they were deposited by turbidity currents. Therefore, we feel that it is important to determine whether the sediments retain a grain alignment, which might be preserved as an anisotropy in the magnetic susceptibility of the sediments, and which could then be used to define the direction of flow.
Methods, Anisotropy of magnetic susceptibility (AMS) measurements can be used to estimate the statistical grain orientiation of clastic sediments and therefore to allow inference about palaeocurrents and sedimentation conditions (Hamilton & Rees 1970). For an assemblage of highly magnetic silt-sized or coarser grains, such as magnetite (which is commonly present in deep-sea sediments), the direction of the maximum susceptibility axis in a sample will correspond to the direction of preferred orientation of the longest axis of the magnetic grains. This direction is hypothesized to correspond to the orientation of the non-magnetic grain fraction as well. The magnetic method of sediment fabric determination is thus indirect, but it provides a quantitative description of grain alignment in three dimensions more rapidly than other direct grain orientation measurements. We describe the measuring techniques and basis for sample rejection in detail here since understanding the magnetic technique and its limitation is essential to interpreting the AMS results. The susceptibility is estimated in the form of a second-rank symmetric tensor k which relates the magnetization Ji induced in a sample by a field Hi: Ji = kijHj
The susceptibility may be specified completely by six quantities, three relating to the magnitude of the principal axes (km~xkmt, and kmm) and three relating to their directions, which are orthogonal. We use a low-field torsion magnetometer (as described by King & Rees 1962) employing an
26I
alternating field of up to 5172 A/m (65 Oe) (rms) magnitude. Susceptibility differences are measured with this technique and a separate measurement of bulk susceptibility is made on an AC bridge to determine the magnitude of the principal susceptibility axes. A check on the reliability of the measured data is obtained by a redundant measurement scheme. The best data values are found by a least square fit, and the differences between these optimized values for the susceptibility tensor and the original readings are calculated. An estimate of data reliability is obtained by comparing the root mean square value of these residuals to that of the optimized values. This measure of data reliability rarely exceeds 10% in the specimens measured here, indicating that specimen susceptibilities are in general adequately represented by a second order symmetric tensor. The few specimens with a data reliability measure exceeding 151~,;may have more complex susceptibilities, and are rejected from further analysis (Tables 1-4). The results of the reliable AMS measurements are described in terms of the magnetic foliation plane, which is normal to the axis of minimum susceptibility, and the magnetic lineation within the foliation plane, which lies along the direction of the axis of maximum anisotropy. An indication of the degree of alignment is obtained from the parameter h, defined for this study as: kmax -- kmin
/1 = - x 100 (k .... + kmin+ ki.t)/3 and the parameter q (Hamilton & Rees 1970): kmax - kint
q - (k ..... + kind/2 - krnm The parameter h provides an estimate of the total anisotropy, the sum of both lineation and foliation. The parameter q is a measure of the relative importance of the two fabric elements: foliation produced by gravitational settling of grains on a flat surface, and lineation resulting from aligning forces tangential to the bed. Values ofq vary from 0 (pure foliation) to 2 (pure lineation) with the change from dominantly foliar to dominantly linear fabric at a value of 0.67. Both the orientations of the principal susceptibility axes and the parameter q can be calculated from susceptibility differences alone measured on the torsion magnetometer, whereas the parameter 17 is also dependent on the bulk susceptibility. Measurement of bulk susceptibility is made on the less sensitive instrument and includes contributions from paramagnetic and diagenetic sediment components in addition to the ferromagnetic fraction that is thought to be primarily
262
A.N. Shor, D.V. Kent and R.D. Flood
responsible for the observed susceptibility differences. Consequently, the magnitude of anisotropy (e.g. h) is much less precisely determined, and useful in a more qualitative sense than the orientation and shape of the susceptibility ellipsoid. The results of measurement on laboratorydeposited and natural sediments have established the characteristics of a primary magnetic fabric induced by flowing water (Granar 1958; Rees 1965; Hamilton & Rees 1970). The principal feature is a well-defined magnetic foliation in or near the bedding plane. The parameterf(Crimes & Oldershaw 1967) gives the angular deviation of the kmin axis from the pole to bedding; in undisturbed sediments values of f a r e usually less than 15~ (Hamilton & Rees 1970; Rees & Frederick 1974). Lineation is usually subordinate to foliation, and values of q typically range from 0.06 to 0.67. These characteristics reflect the expected dominant role of gravitation acting on grains during sedimentation. Several types of secondary fabrics have been recognized and often result from postdepositional alteration of a primary fabric. Prevalent in deep-sea sediments are deformed fabrics, either resulting from the activity of burrowing organisms in the sediment (Rees et al. 1968) or due to coring disturbance (Rees & Frederick 1974; Kent & Lowrie 1975). In order to discriminate and reject specimens which have suffered modification of primary fabric, we applied a strict rejection criterion based on the value of parameterf. Iffexceeds 15~ (i.e. if foliation dips more than 15~ the measured specimen fabric is considered anomalous and rejected from consideration as an indicator of current flow direction. We feel that the data reliability parameter andfconstitute minimum criteria to reject samples for which fabrics are poorly-defined and have dominantly secondary (post-depositional) characteristics. For perspective, we note that Rees & Frederick (1974) found in a study of DSDP cores that only 15 of 90 specimens with measurable anisotropy exhibited primary style fabrics. Cores in the present study are assumed to have been taken vertically. The cores can be oriented in the horizontal plane by reference to magnetic remanent declinations measured on the samples for which anisotropy is determined. The remanent magnetization of each sample was determined after partial alternating field demagnetization to 23874 A/m (to remove unstable or spurious components). Except for core intervals which show a pronounced trend (the top 3 m of core RC 22-14 and over the entire length of core RC 22-02), we believe that the mean declination provides an estimate of geographic north in these
normal polarity, late Pleistocene sediments. The rotation in the two cores is attributed to twist during either coring or extrusion (probably the latter, as values in core RC 22-14 change in a core-coupling break). Mean declinations were therefore calculated either as linear regressions on depth (for twisted sections) or as simple means (for sections exhibiting no twist). Azimuths of the kma• axes for each core were oriented relative to mean declinations to place the magnetic fabric results in geographic coordinates. Although an orienting technique using magnetic remanence of the sediment is the only one available for these cores, it presents a rather subtle problem that is difficult to resolve. It has been shown in laboratory redeposition experiments that certain processes of deposition can in some instances result in deflection of the remanence directions away from the geomagnetic field; alternatively, the geomagnetic field can exert an influence on anisotropy directions, especially in fine-grained sediment (Rees 1961; Rees & Woodall 1975). It is therefore possible that the horizontal component of remanence (declination), which we use to orient the cores, and the direction of maximum anisotropy, which we assume is related to the axis of current flow, may not be completely independent of each other. Lack of independence tends to bring the km~xand remanence directions into coincidence and to obscure any relationship ofkm~x axes with current flow in geographic coordinates. Figure 3 shows plots of the principal susceptibility directions (kmax and kmin) from cores RC22-14 and RC22-02. The reference azimith (N) in each plot is the mean remanent declination calculated for each sample level as outlined above. Also shown is the measured mean remanent inclination for each core and the inclination expected for a geocentric axial dipole field at the core latitude. The remanent inclination is on average shallower than the expected value in each core. The discrepancy can be interpreted as an 'inclination error ~, due to the influence of hydrodynamic and gravitational torques on the magnetic particles during acquisition of a detrital or depositional remanent magnetization. Although a bias of remanent inclinations into the magnetic foliation plane thus appears to exist, the extent of a corresponding bias of remanent declinations toward the magnetic lineations (kmax axes) is difficult to assess due to the absence of a suitable, independent frame of reference. However, considering the subordinate role of lineation compared to foliation in the magnetic fabric of these sediments, the remanent declinations would not be expected to be greatly deflected toward the
Contourite o1" turbidite?
263
R C 2 2 02 N
R C 2 2 02
f <- 15~
N
RC 22 14 N
f<
f > 15 ~
RC22 1,~
[50
N
f > 15 ~
KMAX O
KMIN
9
FIELD
[]
INCLINATION
REMANENT INCLINATION
FIG. 3. Stereographic plots of magnetic remanence and susceptibility anisotropy directions in piston cores RC22-02 (top) and RC22-14 (bottom). Panels on left show anisotropy data consistent with a primary fabric 0c~<15~ panels on right are anisotropy data rejected for paleocurrent analysis in each case due to high values o f f ( > 15~ The reference azimuth (N) is the mean remanent declination calculated for each sample level as discussed in text. Mean remanent inclinations (42.5 :: for RC22-02 and 49.8 for RC22-14) are compared to axial dipole values at core sites. lineation directions. Of more concern is the possibility of a systematic convergence of kmax and remanent declination directions due to field control on anistropy in such fine-grained sediments. However, the c o m b i n a t i o n of steep field
inclination (about 60 ) and selection of magnetic fabrics with near horizontal foliation planes 0Cless than 15 ) should help to minimize the effect of the geomagnetic field on the anisotropy data set used for palaeocurrent analysis in these cores (e.g.
264
A.N. Shor, D.V. Kent and R.D. Flood
Rees 1961). Detailed analyses of magnetic grain compositions and sizes are needed to further constrain the origin and significance of the magnetic susceptibility anisotropy in these sediments. With presently available information, we already see no obvious variation of down-core declination directions related to gross sediment lithoIogy. Finally, we point out that alignment azimuths (k ..... ) are not simple vector quantities, but rather are bi-directional; alignment directions are standardized here to refer to the downward dip of the k .... axis. However, the reader should be aware that all reported kma~ directions have conjugate values, 1807 different, and that the method described herein provides only an alignment and not an actual flow direction.
Results Piston core RC22-14 (4559 m) The magnetic fabric and remanence of 54 samples were measured from core RC22-14 between 14 cm and 965 cm (Table 1 & Fig. 4). Twenty-three samples were rejected for palaeocurrent analysis based on values of f exceeding 15 (foliation deviating from horizontal); two of these same samples also displayed a high rms error. Apparent declinations (corer frame of reference) show severe core twist over the upper 3 m (greater than 180 ~ twist between 80 cm and 295 cm). North directions at each sample level were calculated from a linear regression of declination on depth, and azimuths of magnetic anisotropy (kmax) were oriented to geographic coordinates using the regression line estimates of true north. The measured sample declination was used to orient the sample at 45 cm. Declinations from 300 cm to the bottom of the core showed no evidence of core twist, and the mean declination value of 65' was used to orient these samples to geographic north. Magnetic results show three discrete units in the core, each corresponding with a lithologic unit. The uppermost 225 cm (Unit I) shows a bimodal set of primary fabric kmax orientations, with four samples aligned along a trend of 115-145" and five samples along the nominally orthogonal trend of 210-245. Two samples in this unit were rejected. Unit II (225-475 cm), in which 11 samples were measured, contains no sample in which a primary fabric alignment determination is possible. Values of f exceed 30 ~ for all samples, and exceed 70 ~ in nine samples. The deepest unit (Unit III; 475-970 cm), displays a predominant alignment along 1 lff', with all but one of 22 samples between 90 and 130. Ten
additional samples were rejected from this unit. The dominant fabric orientation is near to the downslope direction at this site suggesting that these sediments may have been deposited by turbidity flows. The apparent disruption of primary magnetic fabric in Unit II is of some interest, since remanence of this unit shows normal results (positive, steep inclinations, with declinations consistent internally and with the unit below). The most reasonable processes for disturbing fabric but not N R M are bioturbation or possibly debris-flow emplacement involving severe liquifaction. Bioturbation might reasonably be expected to disrupt the original magnetic fabric while remanence is not locked in until sediments are buried below the burrowing zone (Kent 1973). If debris-flow transport is responsible, then it is necessary to infer that the entire deposited layer is remagnetized into the field at the time of deposition. Although there is a clear difference in carbonate content of Unit II relative to the remainder of the core, this unit is not necessarily allochthonous based on this criterion alone. At present we prefer the interpretation that bioturbation has altered the sediment fabric and possibly destroyed any laminations in this interval. Piston core RC22-02 (4925 m) Twenty-six samples were examined between 9 and 564 cm (Table 2 & Fig. 5). Six samples were rejected based on excessively large values o f / i Declinations showed no evidence of core twist, and the core mean of 100 ~was subtracted from all AMS azimuths to align samples in geographic coordinates. Sixteen of the 20 acceptable samples show km~x azimuths within 30 ~ of the regional contour trend of 235'. Unlike core RC 22-14, no correspondence was observed between interval s azimuth and lithology. Neither of two samples from the uppermost lithologic unit (Unit I; 0-22 cm) provided acceptable primary magnetic fabric, most likely because of disturbance during coring and subsequent handling (Kent & Lowrie 1975). This carbonate-enriched zone, which is commonly missed in the piston cores from the region, is inferred to be Holocene. A similar lithologic unit is discussed in the results for the gravity core. Lithologic Unit II (22-231 cm) is similar in most respectsto Unit I of core RC22-14. Three of eight samples were rejected. Of the remaining samples, four were aligned within 30 ~ of the contour trend. The lowermost sample in this unit shows alignment more closely approximating the downslope trend. There are two 2 cm sand layers in the 11 cm which are below this sample
265
C o n t o u r i t e or turbidite?
TABI~ 1. Piston core R C 2 2 - 1 4 41~
61~
4559 m
Sample Mean AzhTmth Geographic depth (cm) Inclination Declination declination+ (km,~,) alignment Unit I 14 45 80 91 110 123 164 176 194 201 212
37.5 70.9 59.8 62.0 63.9 57.5 40.0 52.6 37.6 55.4 44.4
063.4 186.4 261.7 292.7 301.2 294.1 334.5 345.9 357.8 009.9 015.8
Unit II 255 264 274 295++ 323 333 344 3655 372 385 465
30.4 43.8 44.9 51.0 66.6 66.6 60.6 56.7 57.2 55.4 76.7
073.1 088.8 085.3 085.2 050.9 050.0 050.6 067.0 057.5 046.8 048.0
Unit III 486 500 507 524 561 574 606 622 633 642 647 670 682 686 695 702 722 747 765 770 776 783 793 813 829 855 863 868 875 907 927 965
66.3 66.8 67.8 71.0 59.5 57.4 48.5 45.8 42.7 47.1 44.2 35.1 25.2 50.6 46.9 56.8 45.9 62.1 53.2 60.1 57.3 51.0 43.4 20.4 35.9 36.7 32.5 30.2 32.1 27.0 28.8 46.9
047.6 053.4 051.4 068.3 090.5 078.4 065.0 063.9 071.3 064.0 063.9 063.7 058.6 070.0 075.0 067.7 067.4 074.4 072.7 067.2 077.5 090.1 093.1 085.2 070.2 046.8 042.5 042.7 045.1 028.4 036.1 012.3
(186.4) 265.8 275.7 292.6 304.3 351.6 007.7 014.0 023.8
325.1 * 305.9 337.9 164.0 343.7 336.3 177.3" 204.2 125.6 324.9 345.5
119.5 072.1 248.3 051.1 032.0 212.6 117.9 310.9 321.7
186.6* 095.0* 279.3* 019.7" 335.0* 334.4* 166.1 * 33 ! .4* 330. l * 325.3* 014.3 *
64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9 64.9
344.9* 342.1 329.3 338.6* 347.8* 344.8 346.1 353.6 010.5 009.1 010.1 020.4 014.2 000.4* 011.7 160.2* 013.2 002.1 * 358.7 359.5* 192.3 016.5" 355.9 187.8 176.5* 004.9 006.4 356.4 004.5 003.5 019.2 193.7"
277.2 264.4 279.9 281.2 288.7 305.6 304.2 305.2 315.5 309.3 306.8 308.3 293.8 127.4 291.0 122.9 300.0 301.5 291.5 299.6 298.6 314.3
f 20.0 9.6 1.6 5.8 1.1 4.1 22.5 14.0 14.9 8.1 2.0
.91 1.01 .54 .56 .27 .25 .15 .21 .17 .20 .19
16.1 71.0 81.9 78.3 71.5 87.1 82.3 87.9 36.0 77.0 82.7
.37 8.3 .30 7.4 .84 5.3 .39 2.0 .31 6.4 1.05 6.9 .45 6.0 .35 39.4 1.85 9.1 1.24 4.6 1.46 8.4
17.9 10.1 11.2 34.3 28.3 6.0 8.9 3.3 9.6 8.2 5.1 7.4 8.9 89.2 13.8 69.3 5.9 31.8 6.7 89.4 8.4 29.8 13.5 9.5 78.4 7.5 6.6 7.2 4.2 10.2 10.0 24.8
.78 .18 .12 1.25 .70 .70 .88 .32 .21 .15 .19 .06 .15 .75 .49 .33 .29 1.33 .36 .51 .46 1.17 .26 .10 .56 .15 .04 .05 .18 .14 .05 .14
*f>15 t a = 1 0 . 5 for Unit I: a = 1 6 . 5 for U n i t I I I ~. rms error > 15% Inclination, declination following demagnetization to 23874 A/m.
7.7% 23.1 25.0 18.1 22.6 32.2 14.3 18.9 31.3 19.3 33.4
6.6% 11.6 24.7 4.0 I0.0 9.3 15.0 18.8 10.5 15.6 18.9 19.9 21.3 3.8 8.6 4.2 21.2 12.7 14.4 3.5 7.6 4.7 2.3 10.5 2.8 14.3 10.7 21.5 23.6 27.9 49.6 9.2
266
A.N. Shor, D.V. Kent and R.D. Flood
~ , ~, ~
~ I
,.:_~ ~1~
'
"~ e~ ~
=-- ~e
,#
.~. ~w.~ ~
II
.-.~
1 ~
[]
~
0
~"
~
:~ ~
~
~'~
9 ~'~
~
.~ ~ . ~
~
N [] .
~n
~i~T
"" " .
0
!
.
0
~
""
9 ~
0
~
~
~
~
0
~ 0 ~ ' ~ "~
0.--~
9
~
c~ [.~
i I
Q@
I
o.._~
~,
../.
/.
/
_.-'-.
Y
,D
. . .~ -. .
,
~_.-.._.__._
i=
~
~,~ ~-~
e
_ _ _
o~'~ 9~
o
~
. ~:
9- . ~ =
.
~ 0
~
./ ~,,. /,
!I
0~'~.~ 9 "%
,
' v
($32Y~3l~}
H.Zd30
""
,
9
14tO.Z.ZOBOD$
,,
"
~o
~
~=
0
~ ~ r ~:~B
,~.,~E E E
2
C o n t o u r i t e or turbidite? TABLE 2. Piston core 22-02 40~
61~
267
4925 m
Sample Mean Azimuth Geographic depth (cm) Inclination Declination declinationt (kmax) alignment Unit I 9 18
49.5 '~ 37.2
096.0 ~ 118.7
Unit II 87 107 132 145 173 190 211 220
28.1 18.5 30.8 33.5 42.4 52.1 39.1 47.8
102.1 098.4 076.5 072.4 051.2 098.3 106.8 123.2
Unit III 263 279 282 319 324 329 338 352 385 402
47.7 43.0 39.0 25.0 23.1 36.9 35.3 56.0 57.1 53.6
108.4 089.3 087.9 107.6 106.1 102.2 102.5 103.0 090.1 091.0
100.0 100.0 100.0 100.0 100.0
Unit IV 425 444 472 504 532 565
69.3 56.3 60.4 32.8 37.4 53.7
119.3 101.9 111.8 123.3 109.3 102.7
187.05. 101.2" 100.0~t 100.0
100.0 100.0 100.0
154.4 359.5 186.3* 352.1" 177.5* 153.4 309.6 094.7
100.0 100.0 100.0 100.0
324.7 333.6 327.8 338.4 329.3 000.7* 277.7 114.9 156.7 000.7
100.0 100.0 100.0 100.0 100.0 100.0
332.7 346.7 348.8 177.5 019.3 144.7
f 89.8 ~ 86.1
054.4 '~' 259.5
053.4 209.6 355.3 224.7 233.6 227.8 238.4 229.3
10.2 7.9 88.8 50.3 83.3 t3.1 2.2 10.0
.34 .32
4.6% 4.6
.85 2.4 .28 13.5 .37 4.0 1.54 5.5 .28 2.8 .53 12.9 .19 14.1 .33 3.6
177.7 014.9 056.7 260.7
7.1 8.3 4.9 5.8 13.2 16.8 6.7 2.4 2.3 1.3
.33 .26 .32 .18 .39 .33 .11 .22 .41 .21
7.6 8.3 10.6 26.9 5.5 11.1 7.9 7.7 11.4 11.0
232.7 246.7 248.8 077.5 279.3 044.7
3.5 7.3 7.9 13.6 3.0 7.1
.29 .18 .16 .06 .04 .25
7.9 20.5 17.2 24.8 26.8 15.0
* f > 15 ,-~
r a = 16.0:: Inclination, declination following demagnetization to 23874 A/m.
suggesting that we m a y have sampled a portion of a fine-grained turbidite. Unit III (231-402 cm) is visually similar to U n i t II of R C 2 2 - 1 4 with interbedded greyishb r o w n and reddish-brown clays. Silt laminations are more frequent in core R C 22-02, however, and Mn-oxide stains are not observed. Only one o f 10 samples were rejected for anisotropy, and seven of the remaining samples were closely aligned with the regional contours. Only one sample shows a p r e d o m i n a n t l y downslope alignment. Unit IV (402-587 cm) is similar to Unit II above, with somewhat more a b u n d a n t silt layering. All six measured values p r o d u c e d acceptable primary fabrics, and all lie closer to the c o n t o u r trend than the downslope direction.
Piston core RC22-15
( 4 2 1 0 m)
Thirty samples were analysed between 3 and 476 cm (Table 3; Fig. 6). Five additional samples between 483 and 585 cm were also examined (the deepest is from flow-in at the b o t t o m of the core). This latter group of samples is not discussed due to problems in interpretation of N R M below a tray break at 480 cm. Of the 30 samples, only one is rejected based on the criterion of large dip in anisotropy ( f > 15~ One sample with f = 15 ~ is retained in the present discussion. Lithologies in core R C 22-15 are not readily c o m p a r a b l e to those observed in cores R C 22-02 or R C 22-14. The core consists p r e d o m i n a n t l y of pale b r o w n clay, and contains a b u n d a n t silt laminae above 230 cm. F o u r units have been
268
A.N. Shor, D.V. Kent and R.D. Flood
d
[] o
E
N
CD
vn
[]
G
0
Z
.=.
_=
8 o
~'~ ~, 8
"]i ~
9
~
.
~
"
I f~
s
f~
"---- ~
r-
~
r~-,---a" r - - - ~ ' - - ~ - . - - r - . ' - - ' - -
"---.
.u 0 o
i
,
.u 9
9
9
v
~
9
v
(S3~.L31,fl) HJ.d30 WO.L..Logsns r
C o n t o u r i t e or t u r b i d i t e ? TABLE 3. Piston core 2 2 - 1 5 41~
62~
269
' W; 4210 m
Sample Mean A:imuth Geographic depth (cm) Inclination Declination declination+ (km,~x) alignment 3 33 44 53 74 93 106 113 116 134 147 167 193 221 234 253 284 305 345 352 375 384 393 404 415 437 441 446 464 476
25.6: 56.2 49.8 34.8 30.6 25.2 30.2 37.6 34.2 27.9 40.2 31.1 38.3 36.9 40.8 40.7 48.7 39.1 47.4 40.9 38.0 38.0 40.6 63.2 46.3 42.8 59.9 60.9 50.7 53.0
129.8 103.0 132.8 118.4 118.6 142.8 159.6 155.0 155.3 154.7 148.4 155.5 164.7 185.0 193.2 191.4 201.0 231.0 219.2 215.4 222.6 231.8 211.3 259.9 259.9 270.1 259.0 250.9 273.4 261.1
105.5 115.9 119.7 122.9 130.1 136.8 141.3 143.7 144.7 151.0 155.5 162.5 171.5 181.2 185.8 192.4 203.1 210.5 224.4 226.8 234.8 237.9 244.9 248.7 256.3 257.7 259.5 265.7 269.9
166.5 126.2 354.7 282.3 359.0 331.7 184.1 075.2 064.0 324.7 351.4 321.4 145.5 268.8 046.5 081.1 082.5 289.1 278.9 323.4 295.7 223.3 126.9" 241.5 248.8 263.3 285.2 353.0 274.1 284.5
061.0 010.3 235.0 159.4 228.9 194.9 042.8 111.3 279.3 173.7 195.9 158.9 334.0 087.6 220.7 248.7 239.4 078.6 054.5 096.6 060.9 345.3 356.6 000.1 007.0 027.5 093.5 008.4 014.6
f
q
h
7.2 4.7 3.7 1.8 0.9 4.6 4.3 2.7 7.9 8.3 4.5 7.1 9.8 5.3 15.0 5.7 3.1 5.0 6.7 2.5 3.0 2.7 59.1 12.7 9.8 1.0 14.7 5.1 6.7 2.9
.17 .13 .48 .03 .05 .05 .14 .07 .07 .02 .13 .06 .04 .12 .07 .08 .13 .10 .09 .10 .10 .02 .29 .26 .22 .51 .31 .15 .03 .04
51.9% 24.5 6.2 14.9 24.7 25.9 24.7 18.2 21.3 27.5 26.1 16.5 30.6 21.3 16.6 34.4 19.7 28.1 10.0 10.3 10.0 30.3 31.8 11.8 5.8 12.5 11.6 20.9 36.4 24.2
*f>15 t a=10.9 Inclination, declination following demagnetization to 23874 A/m. described as possible debris-flows, varying from 12 to 70 cm thick, based on their coarse texture, lack of parallel lamination, and the presence of larger clasts both of rock fragments and d e f o r m e d soft sediments. These units, however, show n o r m a l N R M directions, consistent with those of b o u n d i n g lithologic units. Azimuths ofkmax are more variable in this core than in either of the other two piston cores. N o notable correspondence is observed between iithologic properties and magnetic properties. Even excluding the samples from units described as debris-flow, no clear pattern emerges in kmax direction, and the sample values instead show a b r o a d distribution a r o u n d the compass. However, the region between 280 ~ and 330 ~ ( N W / S E ) contains only a single sample. We are not certain if further sample analysis will improve the separation o f azimuth groupings. The shallowest sample, from 3 cm, comes from the calcareous
unit c o m m o n l y observed near the top of gravity cores and box cores from the region, and shows a grain alignment parallel to contours, similar to that observed in the gravity core. A l t h o u g h the grain alignment deduced from A M S in core R C 22-15 does not show any clear pattern related to either contour-following or downslope currents, it is i m p o r t a n t to note that this core was recovered at a depth shallower than the region k n o w n to be influenced by the present deep b o u n d a r y current ( R i c h a r d s o n et al. 1981). Nevertheless, the palaeocurrent results remain enigmatic at this stage.
Gravity core EN-078-GGC 1 (4815 m) Nine samples were analysed from this 83 cm long large-diameter gravity core. Three samples came from the Holocene, carbonate-enriched Unit 1; six additional samples were obtained from a
27o
A.N. Shor, D.V. Kent and R.D. Flood
,,<
,.,..
I
~
o 0
:.}} 9
[]
.=-
tk 4..
oi"'t "" " 9
~1
9
"7
r (",1
o e~
.~
o
o
..~ ....~, "'~"
_
_z
z z
u~
~
.,.~ -'''''~
9
.
$, 9
9
v
{S3M.L3,~J
v
H.Ld3G I f / O . L l O 8 8 f l S
6~
Contourite or turbidite? gravel-rich clay (Unit 3) below 22 cm. Values o f f exceed 1 5 in one of three samples from Unit 1 and four of six samples from Unit 3. Because of the limited number of samples investigated in this core we have reported all values in Table 4. However, the samples with foliation planes exceeding 15~ should be interpreted cautiously. The gravity core was the only core in the present study which was sampled and analysed while still moist, which may cause some differences in sediment response to sampling and the AMS measurement procedure. Both units 1 and 3 are strongly anisotropic, and show internally consistent directions of alignment which are significantly different from each other. Unit 1 shows an alignment direction which is consistent with contour-current flow, whereas Unit 3 shows alignment of the long axis approximately parallel to the channel trend (downslope
flow).
Discussion The initial results of AMS studies of three piston cores and one gravity core from the Nova Scotia continental rise indicate that the silt-laminated reddish-brown clay lithofacies may not be strictly associated with a single flow process in this environment. Further examination of cores across the bathymetric interval between cores RC 22 14 and RC 22-02 are necessary to determine whether a progressive change from downslope to alongslope orientations is preserved or whether some more complex relationship exists between these two sampling areas. Cores in this bathymetric interval are presently being analysed. TABLE 4. Gravity core E N - O 7 8 - G G C 1 41~
27I
Based on our present results, the following inferences may be drawn about the processes of deposition over the depth interval of 4210-4925 m on the Nova Scotia rise: (1) Our only Holocene samples, from the calcareous facies in the gravity core and in piston core RC22-15 indicate that this unit retains a fabric related to alongslope flow in spite of some degree of bioturbation. Although further study is warranted, we believe that this unit may be considered a 'contourite' based on its age, setting and magnetic fabric. (2) The gravel-rich clay unit in the pre-Holocene section of the gravity core shows a strong downslope fabric alignment, and we consider this unit a type of fine-grained turbidite. Because it lacks primary sedimentary structures and contains significant gravel, we believe it deserves further sedimentological study to define the nature of the process(es) of deposition, and its regional distribution to determine whether deposition results from channelized flow. (3) Samples from the homogeneous clay intervals in piston core RC 22-14 clearly show a secondary disturbance of fabric which we believe results from bioturbation or possibly debris-flow deposition. If the latter is the transport agent, then we may infer that a change in magnetic remanence has taken place to statistically reorient magnetic minerals to the regional field following debris-flow deposition. (4) The silt-laminated sequences in piston core RC 22-14 show a magnetic fabric alignment which is most consistent with downslope flow
62~
4815 m
Sample Mean Azimuth Geographic depth (cm) Inclination Declination declinationt (kin,x) alignment
f
q
Unit 1 8 11.5 14
49.2 48.4 58.5
207.1 248.0 253.1
221.0 221.0 221.0
098.8 079.4 105.0
237.7 * 218.4 244.0
41.6 .63 2.5 .76 14.6 .99
Unit 3 23 28 38 48 58 68
45.6 44.7 49.2 45.1 37.3 48.0
231.8 212.0 222.1 200.0 196.3 219.0
221.0 221.0 221.0 221.0 221.0 221.0
019.8 201.4 177.6 219.1 181.5 059.9
158.8 340.4 316.6" 358.1" 340.5* 198.9"
12.3 10.8 21.5 22.9 18.7 34.4
*f>15 : t a = 20.0 Inclination, declination following demagnetization to 23874 A/m.
.18 .40 .ll .13 .23 .33
h 8.4% 8.3 6.6 34.4 57.5 48.2 50.7 50.8 41.0
272
A.N. Shor, D.V. Kent and R.D. Flood
during deposition. However, four samples within Unit I of core RC 22-14 show alignments consistent with contour-current flow rather than turbidity currents. (5) Silt-laminated sequences in core RC22-02, which comprise the entire 6 m long core, show a magnetic fabric alignment which is consistent with contour-current flow. Silt layers in core RC22-02 are generally thicker and less closely spaced than in core RC22-14, and the core is located beneath the present core of a strong thermohaline current, whereas RC22-14 is near the upslope margin of this flow. (6) Samples from core RC22-15 do not show a clearly defined preferential alignment related to either downslope or alongslope flow conditions. At the present time, the strongest alongslope currents are restricted to depths greater than this core. While it is possible that some complex interaction of downslope and alongslope processes results in a 'smearing' of the directional information preserved in AMS, we prefer to leave interpretation of these data until such time as unambiguous results relating lithofacies and depositional processes in other settings provide a less subjective basis for discussing complexities. (7) Highest percent anisotropy values are observed in Unit III of the gravity core (downslope alignment) and in piston core RC22-15 (no clear preferential alignment) rather than in core RC22-02 which shows the strongest alongslope alignment. Anisotropy values are low in Unit 2 of core RC22-14 which shows signs of post depositional deformation.
Conclusions We believe that the anisotropy of magnetic susceptibility technique (AMS) holds considerable promise in differentiating downslope and alongslope depositional processes of fine-grained sediments on the Nova Scotia continental rise. Our present results from four cores in general support the hypothesis of Hollister (1967) that contour-following bottom currents have played a role in sediment redistribution on the continental rise during the glacial stages of the Pleistocene. AIongslope currents are also important depositional agents in the Holocene. However, our preliminary results demonstrate that downslope magnetic fabric alignments occur as frequently as alongslope alignments. ACKNOWLEDGMENTS: Support for AMS studies was provided by the US National Science Foundation (OCE80-12897); cores were collected through support from US Office of Naval Research (T0210-82-11111). Archive facilities at L-DGO are supported by NSF through OCE81-22083. We are grateful to both agencies for their support of this work. N. Iturrino assisted in collection of gravity cores, B. Tucholke collected piston cores. J. Kostecki and F. Hall carried out AMS and remanence measurements. J. Broda loaned the gravity corer. V. Kolla was instrumental in obtaining support for this project. E. Free and C. Elevitch typed the manuscript. This paper was reviewed by C. Hollister, N. McCave, J. Damuth, W. Ruddiman, D.A.V. Stow, A. Bouma and A.I. Rees. L - D G O Contribution//3621.
References AUFFRET, G.-A., SICHLER, B. & COLENO, B. 1981.
Deep-sea texture and magnetic fabric indicators of bottom currents regime. Oceanologica Acta, 4, 475-88. BULWNCH,D.L., LEDBETTER,M.L., ELLWOOD,B.B. & BALSAM, W.L. 1982. The high-velocity core of the Western Boundary Undercurrent at the base of the U.S. continental rise. Science, 215, 970-3. CRIMES,T.P. & OLDERSrIAW,M.A. 1967. Palaeocurrent determinations by magnetic fabric measurements on the Cambrian rocks of St. Tudwal's Peninsula, North Wales. Geol. J., 5, 217-32. ELLWOOD, B.B. 1980, Application of the anisotropy of magnetic susceptibility method as an indicator of bottom-water flow direction. Marine Geol., 34, M83-90. & LEDBETTER, M.T. 1979. Palaeocurrent indicators in deep-sea sediment. Science, 203, 1335-7.
-
-
GRANAR, L. 1958. Magnetic measurements of Swedish varved sediments. Arkiv Geo~vsik, 3, 1-40. HAMILTON, N. & REES, A.I. 1970. The use of magnetic fabric in paleocurrent estimations. In: Runcorn, S.K. (ed.), Paleogeophysics. Academic Press, New
York. 445-64. HOLLISTER,C.D. 1967. Sediment Distribution and Deep Circulation in the Western North Atlantic. Unpubl. doctoral dissertation, Columbia University in the City of New York, USA. -& HEEZEN, B.C. 1972. Geologic effects of ocean bottom currents: Western North Atlantic. In: Gordon, A. (ed.), Studies of Physical Oceanography 2. Gordon & Breach, London. 37-66. HORN, D.R., EWING, M., HORN, B.M. & DELACH,M.N. 1971. Turbidites of the Hatteras and Sohm Abyssal Plains, Western North Atlantic. Marine Geol., 11, 287-323.
Contourite or turbidite? KENT, D.V. 1973. Post-depositional remanent magnetization in a deep-sea sediment. Nature, Lond., 246, 32~,. - & LOWRIE, W. 1975. On the magnetic susceptibility of deep-sea sediment. Bull. geol. Soc. Am., 87, 321-39. KING, R.F. & REES, A.I. 1962. The measurement of the anisotropy of magnetic susceptibility of rocks by the torque method. J. geophys. Res., 67, 1565-72. LEDBETTER, M.T. 1979. Fluctuations of Antarctic Bottom Water velocity in the Vema Channel during the last 160,000 years. Marine Geol., 33, 71 89. -& ELLWOOD, B.B. 1980. Spatial and temporal changes in bottom-water velocity and direction from analysis of particle size and alignment in deep-sea sediment. Marine Geol., 38, 245-61. RE~, A.I. 1961. The effect of water currents on the magnetic remanence and anisotropy of susceptibility of some sediments. Geophys. J.R. astron. Soc., 5, 235-51. REES, A.I. 1965. The use of anisotropy of magnetic susceptibility in the estimation of sedimentary fabric. Sedimentology, 4, 257 71. --, BROWN, C.M., HAILWOOD, E.A. & RIDDY, P.J., 1982. Magnetic fabric of bioturbated sediments from the northern Rockall Trough: Comparison with modern currents. Marine Geol., 46, 161 73. -& FREDERICK, D. 1974. The magnetic fabric of samples from the Deep Sea Drilling Project, Legs I-VI. J. sed. Petrol. 44, 655-62. - VON RAD, U. • SHEPARD, F.P. 1968. Magnetic fabric of sediments from the La Jolla Submarine
Canyon and fan, California. Marine Geol., 6, 145-79. REES, A.I. & WODALL,W.A. 1975. The magnetic fabric of some laboratory deposited sediments. Earth planet. Sci. Lett., 25, 121-30. RICHARDSON, M.J., W1MBUSH, M. & MAYER, L. 1981. Exceptionally strong near-bottom flows on the continental rise off Nova Scotia. Science, 213, 887-8. SPIESS, F.N. & MUDIE, J. 1970. Small-scale topographic and magnetic features. In: Maxwell, A.E. (ed.), The Sea 4. John Wiley, New York. 205-50. STANLEY, D.J., SWIFT, D.J., SILVERBERG, N., JAMES, N.P. & SUTTON, R.G. 1972. Late Quaternary progradation and sand spillover on the outer continental margin off Nova Scotia, southeast Canada. Smiths. Contr. Earth Sci. 8. Smithsonian Inst. Press, Washington, D.C. 88 pp. STOW, D.A.V. 1979. Distinguishing between finegrained turbidites and contourites on the Nova Scotia deep water margin. Sedimentology, 26, 371-87. -& BOWEN, A.J. 1980. A physical model for the transport and sorting of fine-grained sediment by turbidity currents. Sedimentology, 27, 31-46. --& LOVELL,J.P.B. 1979. Contourites: their recognition in modern and ancient sediments. Earth Sci. Rer. 14, 251 91. TUCHOLKE, B.E. 1982. Origin of longitudinal triangular ripples on the Nova Scotia continental rise. Nature, Lond., 296, 735-7.
SHOR, D . V . K E N T , R . D . FLOOD, Lamont-Doherty Geological Observatory of Columbia University in the City of New York, Palisades, New York 10964, USA.
A.N.
273
Contourite facies of the Faro Drift, Gulf of Cadiz E.G. Gonthier, J.-C. Faug~res and D.A.V. Stow SUMMARY: The Faro Drift is an elongate sediment body (50 km long, 300 m thick) that parallels the northern margin of the Gulf of Cadiz south of Portugal. On the basis of location, morphology and seismic character of the drift together with bottom photographs, sediment distribution and the known regional oceanography, we can be certain that the Faro Drift was constructed by bottom currents related to the deep outflow of Mediterranean water. Detailed study of a closely-spaced suite of cores has therefore allowed characterization of the sediments into three contourite facies: sands and silts, mottled silts and muds, and homogeneous muds. The first is equivalent to the sandy contourites and the two others to the muddy contourites of earlier studies. These three facies are arranged in irregular vertical 'sequences' that reflect long-term variations in bottom current velocity. They arc distinctly different from typical fine-grained turbidite and hemipelagite facies. There is currently much interest in the effect of bottom currents on sediments in the deep ocean, in the complex interaction of bottom current, turbidity current and pelagic depositional processes, and in distinguishing between their respective deposits (e.g. Richardson et al. 1981 ; Shor et al., this volume). These studies are of particular importance for the understanding of deep-sea sedimentation and for the reconstruction of palaeo-oceans. Many of the major contourite drifts in the North Atlantic have now been cored and their sediments examined in detail. We have thus been able to construct a general picture of contourite facies, refining the earlier work of Hollister & Heezen (1972). Two main contourite types have been described (Stow & Lovell 1979: Faug6res et al. 1979; Stow 1982): muddy contourites are relatively homogeneous and extensively bioturbated, whereas sandy contourites occur as thin, irregular, bioturbated lag deposits or more rarely as clean, cross-laminated beds. Both commonly comprise a mixture of biogenic and terrigenous material. Gradations between these two types as well as coarser-grained gravel lag deposits (gral,elly contourites) have also been observed. However, most of the cores studied to date represent either isolated or very widely-spaced sites on drifts that are many hundreds of kilometres long and tens of kilometres wide. We thus have very little idea of the horizontal distribution or vertical sequences of sediments within such drifts. We also lack detailed information on the small-scale variability of contourite facies in response to the known dynamic variability of bottom currents. We are not yet, therefore, sufficiently confident to unambiguously identify contourites in ancient rock sequences, nor to use facies variability as an accurate indicator of the development and fluctuation of palaeo-bottom currents, although several interesting attempts
have been made in these fields (e.g. Pastouret et al. 1978; Auffret et al. 1981; Lovell & Stow 1981). For these reasons, a much more detailed study of a contourite drift was clearly necessary, preferably one sufficiently small to allow a close-spacing of core stations and in an area where the physical oceanography was well-known. Several relatively small contourite drifts have been identified on the south Iberian margin (Vanney & Mougenot 1981), and were almost certainly constructed by the deep Mediterranean outflow through the Straits of Gibraltar. One of these, the Faro Drift, was selected for detailed study (Fig. 1). The French oceanographic vessel, RV Noroit, was taken to the area in November 1982 and completed some 300 km of 3.5 kHz seismic profiling, occupied 24 sites for piston coring (with associated gravity coring) and 5 sites for bottom photography (Fig. 1). Subsequent laboratory analyses included X-radiography of centimetrethick slabs of core, impregnation and thin-sectioning, grain-size determination by sedigraph and sieving, compositional determination of the sand and clay fractions, and geochemical, palaeomagnetic and biostratigraphical studies. The preliminary results of this work have been reported by Faug6res et al. (in press). In this paper we describe in detail the nature of contourite facies, typical contourite "sequences' and their hydrodynamic interpretation.
Oceanography and bathymetry At the present day there is a deep outflow ofwarm (> 13 C), saline ( > 36.4%0), low-oxygen (4.1 to 4.6 ml/1) Mediterranean Water through the Straits of Gibraltar (Madelain 1970; Zenk 1975; Reid 1978; Ambar & Howe 1979). This watermass flows north and west along the southern margin of Spain and Portugal at depths of 400-1400 m (Fig. 1). At Cape St Vincent, part 275
276
E . G . G o n t h i e r , J . - C . FaugOres a n d D . A . V .
Stow
FIG. 1. Study area with Faro Drift in large stipple, showing 3.5 kHZ seismic track (dashed lines), core sites (solid circles), camera stations (open squares), known flow (heavy arrows) and inferred flow (light dashed arrow) of Mediterranean water, and location of seismic profile shown in Fig. 2 (heavy zig-zag). Contours in metre. Inset map shows passage of Mediterranean outflow water over sea floor and location of study area. (From Faug6res et al., in press.) spreads westward into the mid-ocean and part turns north forming part of a larger scale eastern boundary current. In the Gulf of Cadiz there appear to be three main flows, between 1200-1300 m, 700-900 m and 500-700 m, controlled largely by the bottom topography and partly interconnected by down-canyon flow. There is an overall increase in flow depth from west to east. Flow velocities of up to 300 cm/s have been recorded in the Straits of Gibraltar, decreasing to 180 cm/sjust west of the Straits, 30-40 cm/s in the vicinity of the Faro Drift, and 10-20 cm/s at the western end of the Gulf (Meli6res 1974; Reid 1978). There are local variations due to the influence of bottom topography, some meandering of the flow and counterclockwise gyres, but an overall decrease in velocity concurrent with a westward broadening and deepening of the water mass is observed. The general effects of the Mediterranean Outflow have been documented by a number of workers (Giesel & Seibold 1968; Heezen & Johnson 1969; Meli~res et al. 1970: Meli~res 1974: Vanney & Mougenot 1981). Significant erosion takes place in the Straits of Gibraltar as we~l as in some of the deep channels along the margin, such
as the Fossa Alvares Cabral and Fossa Diego Cao to north and south of Faro Drift. Non-deposition is evident in some areas, whereas the accumulation of a series of sediment drifts up to 300 m thick is noted in others. Acoustic and photographic evidence reveals areas of rocky and current-swept bottom and areas of smooth or rolling sediment-covered morphology. Currentcontrolled bedforms, particularly evident in channelized areas include sediment waves and ripple marks. There has also been a long history of chafing and corrosion of submarine telegraph cables in the Gulf and the Straits of Gibraltar closely related to the Mediterranean Outflow. The overall constructional aspect of the Gulf of Cadiz with its relatively shallow slope gradient (1 : 300) is probably also, in part, a result of the Mediterranean Outflow, although detailed study reveals a much more complex interplay of controis. A series of papers by Mougenot and coworkers on the south Iberian margin (Baldy et al. 1975: Mougenot et al. 1979; Mougenot & Vanney 1980, 1982; Vanney & Mougenot 1981) have shown a shelf 15-40 km wide, a narrow and moderately steep slope (1:6.5 to 1 : 25 gradient), a series of channels or deeps parallel to the margin interrupted by downslope canyons and four
Contourite facies of the Faro Drift, Gul[o/Cadiz
277
3~ E~
.s E eq'~
o"~
~o
9~ az
~az
278
E.G. Gonthier, J.-C. FaugOres and D.A.V. Stow
constructional mounds or drifts along the length of the margin near the foot of the slope (Fig. 1). The Faro Drift itself is 40-50 km long, 10-20 km wide, and up to about 300 m in thickness with a maximum present day relief of 160 m (Fig. 2). It appears to have been constructed over a late Miocene-early Pliocene erosional surface and to have prograded steadily to the north or northwest over a distance of about 10 km during the past 5-6 million years (Mougenot & Vanney 1982). The mode of growth appears closely analogous with lateral accretion in a meandering fluvial system or with constructional levee development in a deep-sea fan.
Sediment facies There is little doubt that over 90~ o of the sediments cored are true contourites, including virtually all of the Faro Drift cores, as borne out by the microphysiographic and bottom photographic evidence (Faug6res et al., in press). Whereas some 5-70/0 of the material in the canyon to the west, or on the channel floors and steep eroded channel slopes, occurs in distinct turbidite beds and slumped units or is an older more
consolidated sediment of unknown origin. In the following sections we describe only the contourites, for which we recognize three main facies: sands and silts, mottled silts and muds, and homogeneous muds. Sands and silts
The sands and silts form about 5~o of sediment and occur in irregular beds from a few centimetres to 20 cm in thickness (Fig. 3). They are almost always surrounded by the mottled silt and mud facies. Primary sedimentary structures The top and bottom contacts of the sand and silt beds are mostly irregular and may be sharp but relatively flat, clearly erosive or completely gradational. Erosional contacts appear to be slightly more common at the base of beds. It is often difficult to distinguish contacts that are of primary dynamic nature from those caused by bioturbation, and many of the contacts that change from sharp to gradational across the width of the core have probably been affected by secondary bioturbation (Fig. 4). There is almost no lamination remaining in any
FIG. 3. Photographs of core sections with mottled silt and mud, and homogeneous mud contourites. Sharp and erosive bed contacts shown by solid lines, gradational contacts shown by dashed lines.
Contourite facies of the Faro Drift, Gulf of Cadiz
279
Fro. 4. Thin sections photographs of impregnated slabs of sections of two cores, showing detailed structures of different contourite facies. Note, in particular, nature of contacts and bioturbation. Teichichnus-likeburrow indicated with 'b'. Mottled aspect at levels 1 is an artefact. of the beds cored, although the presence of ripple marks and other current bedforms on the presentday sea floor would suggest that many of the sandy layers were originally laminated. Rarely, there are discontinuous and indistinct clayey laminae within the sands, although their origin is unclear. Size grading within the sands and silts is often present but very variable in nature (Figs 5 and 11). Positive and negative grading occur equally commonly and may be relatively continuous through a bed or occur irregularly in rapid succession. Several of the thicker beds increase in grain-size from the base towards the middle, where there are pockets of coarser (often shelly) material, and then decrease in size towards the top.
Biogenic structures Bioturbation is the most common structure encountered, although in many of the sands it is only visible as an indistinct mottling (cm-scale) on X-radiographs (Figs 3 and 10). Bioturbation at
the top and bottom contacts of beds introduces pockets of sand into the adjacent sediment and wisps or streaks of mud into the sand. Large (up to 10 cm long) mud-filled or sand-filled burrows occur as straight isolated protrusions some of which appear to be Lophoctenium, or as a tortuous network of smaller diameter tubes.
Texture Although some of the silts and sands contain rare coarse (up to 5 mm) shell fragments, they are mostly relatively fine-grained with a mean grainsize between 30 and 50/zm (Figs 4, 5 and 6). They are mostly, therefore, medium-coarse grained silts with 20-40~ sand and less than 10~0 clay. More rarely, the fine sand fi'action is dominant. They are mostly moderately well sorted. The shape of the cumulative frequency grain-size curves (Rividre 1977) tends towards parabolic or to a combination of hyperbolic (coarse tail) and parabolic (fine tail). This relatively large fine tail is found even in the middle of the thicker beds, which suggests that it is not solely due to
280
E . G . Gonthier, J . - C . Faug~;res a n d D . A . V . S t o w
/
pa~JOS
~lJOOd
~
E~
I
[-...
_~_
PalJos
IlaM
I "--
CO I,l
.
> rr
_. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
N
E C~
. . . . . . . . . . . . . . . . . . . . .
9
(~
....
(i,)
=~ CO _
_
.
.
.
Z
.
8"L
rr (-9
"I:~-
-
:-:--'-
- ~-'k-
~-
~-
-I~- - ~ - - ~ -
~I~ .
.
.
.
.
.
.
.
.
-~-
-
-~
--I~ ~ - I ~ - ~ -
-~-
(I) ci. [D
_v)
_
_
L~ I-<~ Z O m n~ <~ q.)__
:I
o~ 0
~
o
R~
J - -
j
~2
oo
S >(,.9 O ._J o "I-
U9
..,-~
r,_) p
pUDS .~l!S
,
,
r
,
,
,
,
,
r
,
,
r
-
SlAl
~
LU H
I
!
saldu~os
,
--
ual,~
.
r
" ~'"g g ' d g d o
o
-
o
u~
e~
~o ff~
'4 ~
"2 d" ~ ' ~ ~o
~
~
' ~2" ~-
oo
i
[
,
i
Contourite facies of the Faro Dr(p, Gulf o[Cadiz
28I
I00.
80
\ \
60 % 4O (a) HOMOGENEOUS MUDS 20.
O,
,
,
i
.
I00
.
.
.
.
.
60
i
9
.
i
.
.
I0
.
.
.
4
|
I/.zm
I00-
8O
60 % 40
MUDS 20
i
,
,
i
l
i
,
,
i
!
i
,
,
i
l
i
,
,
I
I00
60
I0
4
I /.zm
I00
60
I0
4
Iffm
I00-
8O
60 % 4-0
20
0
FIG. 6. Typical grain-size cumulative frequency curves plotted on semi-logarithmic paper for the three main contourite facies identified.
282
E.G. Gonthier, J.-C. FaugOres and D . A . V . Stow
bioturbation with the adjacent finer sediments but has a primary dynamic origin.
Composition The sands and silts clearly have the largest proportion of coarse fraction (> 63 #m) of the three main facies (Figs 5 and 7). The fraction > 150 ~m is dominantly biogenic, with roughly equal proportions of benthonic foraminifera, planktonic foraminifera and shell debris (mainly bivalves), and rather less ostracods and other species. The mean size of the planktonics is noticeably less than either the benthonics or shell fragments, but larger than the rare traces of quartz, mica, pyrite and amber. The 63-150/~m fraction contains notably more (60-70%) terrigenous material (quartz, mica, iron-coated grains, pyrite, rare glauconite), similar amounts of benthonic and planktonic foraminifera and rather less shell debris. The total carbonate content of the whole sediment varies between 30-40j% , whereas the clay fraction is mostly less than 10%. Many of the larger biogenic grains are fragmented and, more rarely, iron-stained. The quartz grains are subangular to subrounded.
X-radiographs, but it is always irregular and wavy. There is a lamination, continuous across the width of the core, that is a more or less distinct alternation of silty and muddy laminae; a more discontinuous lamination marked by the horizontal alignment of flattened lenses of silt or mud; and a gradation between these two types (Figs 9 and 10). Although the discontinuity of the laminae may be due to bioturbation, it seems clear that there remains at least some evidence of an original dynamic regime. Alternate positive and negative grading is evident on a small-scale within the mottled intervals, but there is also commonly a more gradual increase of grain-size upwards from an underlying homogeneous mud or decrease upwards from a sand through a thickness of 10-50 cm (Figs 5 and 11).
Biogenic structures
The mottled silts and muds form a visually distinctive facies that comprises either a rapid alternation of thin irregular mud, silt (and sand) layers or, more commonly, a completely irregular arrangement of these sediment types in pockets, lenses and streaks (Figs 8, 9 and 10). The facies can be further subdivided with respect to the relative proportions of sand, silt and clay: mottled silts containing lenses and mottles of mud and mottled muds containing lenses of silt. They form about 20% of the cored section in 'beds' that average 10-20 cm in thickness, but that vary from a few cm to 70 cm thick. They occur in association with both the sands-silts and the homogeneous muds.
The dominant feature of this facies is a thorough bioturbational mottling which occurs at a millimetre and centimetre scale, together with other more or less distinct and isolated pockets and streaks at a millimetre or centimetre scale (Fig. 8). Certain more regular or internally structured forms are also recognized and can tentatively be identified as specific trace fossils (Figs 9 and 10). These include, the small (#m-cm scale) elongate lenses and tubes, sometimes so densely-spaced as to form an interlocking network, that are probably a form of Chondrites; the more regular oval and elipsoid forms that are most likely Planolites; and some larger (centimetre diameter) chevronstructured burrows that appear similar to Scolicia and Teichichnus (Werner & Wetzel 1982). At certain horizons there are also very thin, long iron-sulphide filaments and larger pyrite-filled burrows. The bioturbation and burrowing has clearly been a continuous process that has served to modify or destroy much of the original sedimentary structure. Several superimposed episodes of bioturbation are commonly evident.
Primary sedimentary structures
Texture
The top and bottom contacts of the mottled intervals, as with the sand-silt facies, may be sharp and planar, erosive or gradational, or any combination of these types (Figs. 8, 9 and 10). Similarly, the pockets and lenses of different sediment types appear as more or less distinct bodies both to the naked eye and on X-radiographs. It is again difficult to distinguish primary dynamic from bioturbational contacts. Lamination is commonly evident, especially on
The mottled facies is clearly very heterogeneous and comprises pockets or layers of variable grain-size (Figs 4, 5 and 6). Our analyses have attempted to homogenize this variability by using relatively large samples, and have shown a gradation from a more clay-rich very poorly-sorted sediment (mean size, 8-16/~m) to a coarser silt, with poor but slightly better sorting (mean size, 16-30 /ml). The grain-size curves are nearly all parabolic, with a relatively small or absent coarse
Mottled silts and muds
Contourite.facies qf the Faro Dr([t, Gulf of Cadiz
Ca)
283
(b) SANDS
and
SILTS
(b)
Ca) MOTTLED
S I L T S and
(a)
MUDS
(b) HOMOGENOUS
MUDS
FIG. 7. Composition of the sand fractions of each of three contourite facies: (a) 150pm fraction, (b) 150-63pm fraction. FP = planktonic foraminifera; FB = benthonic foraminifera, B = other biogenic fragments, T = terrigenous grains, N = iron sulphide grains or tubes. Views through binocular microscope, scale at bottom left.
284
E.G. Gonthier, J.-C. Faugdres and D.A.V. Stow
FIG. 8. Photographs of core sections with mottled silt and mud, and homogenous mud contourite. Sharp and erosive bed contacts shown by solid lines, gradational contacts shown by dashed lines. tail and a large fine tail; although, in some of the finer samples the grain-size distribution is closer to logarithmic. The grading observed visually, and on X-radiographs is clearly confirmed by closely-spaced series of grain-size analyses.
Composition There is also a relative heterogeneity in the coarse fraction ( > 63/~m) composition of different samples of this facies (Figs 5 and 7). The three major components are benthonic foraminifera, planktonic foraminifera and terrigenous debris (especially quartz, silt and micas), as for the sands and silts, but their relative proportions are very
variable. The shelly debris is less important, and is mostly a similar size to the benthonic and planktonic foraminifera; many of the larger elements are fragmented. The total carbonate fraction oJ and the clays 10- 3U/o, '~~ comprisaverages 20-30/0 ing mainly illite with minor chlorite, kaolinite and smectite.
Homogeneous muds The homogeneous muds are the finest-grained, visually most monotonous and most abundant of the three contourite facies, forming nearly 70% of the cored section in intervals of a few centimetres
Contourite facies of the Faro Drflt, Gulf of Cadiz
285
FIG. 9. X-radiographs of core sections showing typical primary and biogenic structures in the different contourite facies. Hm = homogeneous mud, Mm =mottled mud, Ms = mottled silt, b = burrow (? Scolicia), 1 = discontinuous lamination, mp = mud pocket (? Planolites burrow), i = iron sulphide mottle or mycelia. Scale in centimetres.
286
E.G. Gonthier, J.-C. FaugOres and D.A.V. Stow
FIG. 10. X-radiographs of core sections showing typical structures of contourites (cf. Fig. 9).
Contourite facies of the Faro Drf/t, Gulf o/'Cadiz to a few metres in thickness. Although to the naked eye they appear more or less structureless or massive, X-radiography and detailed sampling has shown that they are not strictly ~homogeneous' (Figs 8, 9 and 10).
Primary sedimentary structures The homogeneous muds are most commonly in gradational contact with the mottled facies in sequences that are either positively or negatively graded. Less commonly, the contacts are sharp or erosive. Lamination occurs in some sections and may be very fine, distinct and planar, or less well-defined and wavy. It is sometimes accentuated by very thin parallel filaments of iron sulphides.
Biogenic structures Where not laminated, the homogeneous muds appear to have been thoroughly bioturbated, although the mottles, pockets and streaks visible on X-radiographs are mostly faint, indistinct and of a very small size (mm scale). Some small lenticular tube-like structures may be Chondrites, but none of the larger burrow types encountered in the other facies are visible. Iron-sulphide traces and pyrite-filled burrows are locally very common.
Texture Although fine-grained, these muds contain a large proportion of mostly fine and medium grade silt (40-60%), up to 50% clay and very little sand grade material (1-8%) (Figs 5 and 6). The mean size is less than about 10 or 12 ~m and the sorting is poor to moderate. The grain-size curves are close to logarithmic in shape with a small coarse tail giving a hyperbolic tendency.
Composition The small sand fraction contains mainly planktonic and benthonic foraminifera in approximately equal but variable proportions, and minor amounts of terrigenous and shell material (Figs 5 and 7). The total carbonate percentage is relatively low (mostly 20-30%) so that much of the silt fraction is terrigenous (especially quartz), with clay minerals dominating the finer fractions. Many of the planktonic foraminifera, bivalve and Dentalium shells and ostracods are whole rather than fragmented, and the sizes of the planktonics tend to be greater than those of the benthonics.
287
Vertical 'sequence' of facies Description of sequences Although we hesitate to use the term "sequence' for fear of confusion with the concept of a distinctive O'pe sequence, such as for turbidites, we do recognize that the vertical succession or arrangement of the three main facies may be of two characteristic types: a negatively-graded sequence in which the grain-size increases, and a positively-graded sequence in which the grainsize decreases upwards. These sequences are mostly between 10 and 100 cm thick (Figs 5 and 11). The complete negative sequence begins at the base with a fine-grained homogeneous mud, passes upwards through a clay-rich and then silt-rich mottled facies and is topped by a coarse silt or fine sand bed. Contacts between facies near the base are gradational but, towards the top, sharp and erosive contacts are more common. Each ofthe component facies may show the range of variability as described in the previous section, although the overall succession of facies is accompanied by progressive structural, textural and compositional changes. There is thus a tendency towards an upward increase in erosive contacts, increase in the size of burrows and bioturbational mottles, increase in mean grain-size and percent sand fraction, change in shape of the grain-size curves from logarithmic through parabolic to hyperbolic-parabolic, increase in the relative proportion of benthonic foraminifera, shell debris and total carbonate, and decrease in the relative size of the planktonic foraminifera compared with other biogenics. The positive sequence is the exact reverse, showing an upward change from the sand-silt through the mottled to the homogeneous mud facies, together with the corresponding structural, textural and compositional changes. There is considerable variability in these two sequences as they occur in different parts of the Faro Drift, particularly with respect to their thickness but also in terms of their completeness (Fig. 11). Commonly, the silt-sand facies is missing and, less commonly, part of the mottled facies or the homogeneous mud. The vertical succession of individual sequences through the cores is also variable. They are most often coupled negative-positive sequences, although these are not necessarily symmetrical, but either sequence can also occur independently. They may be very widely-spaced, separated by up to a metre or more of homogeneous mud, very closelyspaced without significant separation by a mud facies, or occur with a frequency somewhere between these two extremes.
E.G. Gonthier, J.-C. FaugOres and D.A.V. Stow
288 KC 8 2 2 1
KC
g E 5
8220
KC 8 2 1 7
E E5
EEg
k
o1_J
KC
8226
EE~
t i l l i
? /
/ I
I \
\ ! negative
sequence
positive
sequence
150-
net
gradual erosional .
.
.
.
si
\
/ ~contacls
bioturbatedJ : homogenous
cm
\
/,
muds
: m o t t l e d muds : mottled silts sJ Its s: sands
cm
/ I
FIG. 11. Sketches of core sections showing the different types of vertical sequence of contourite facies.
Hydrodynamic interpretation This vertical variation in the contourite facies is probably related to variation in velocity of the transporting current rather than to variation in supply to the system (see later discussion). Thus, a coupled negative-positive sequence in a given core represents a gradual increase, a maximum and then gradual decrease in current velocity at that particular site (Fig. 12). Such fluctuations can be slow and progressive, as shown by the thicker gradational sequences, or much more rapid and sudden, as shown by thinner sequences
and sequences with erosive surfaces and sharp contacts. Although our relative dating of the cores is not yet complete, there are clearly parts of the Faro Drift where current velocities capable of transporting and depositing coarser silts and sands have been very rarely attained, and fluctuations through the Late Quaternary have been infrequent. In other parts, however, the current velocities have been generally higher and fluctuations much more frequent. These fluctuations in time at any one site may be due either to a velocity change or pulsation e r a
C o n t o u r i t e f a c i e s o f the Faro Drift, G u l f LITHOLOGY and STRUCTURE
MEAN GRAIN SIZE 4
8 16 32 64,urn
50
\
/
maximum
velocity
t g
_
5;__ 0
y
*_.~-'-~-_
-silt and sand [-~
mud shell
fragments
bioturbation
discontinuous gradational
silt lenses
contact
sharp ( i r r e g u l a r )
contact
FIG. 12. Schematic vertical sequence of contourite facies from the Faro Drift showing the typical superposition of coarsening-upward (negativelygraded) and fining-upward (positively-graded) sequencies over 50 cm of section. The variation in mean grain-size can be interpreted in term of an increase and then decrease in bottom current velocity (inferred maximum of 10-15 cm/s) (from Faugdres et al., in press).
of C a d i z
289
general or local scale, or to a displacement in space of the whole or part of the current. We do not yet have sufficient data from the Faro Drift to comment on the latter possibility, but note that current instability and velocity pulsation has been shown to be a common characteristic wherever sufficiently long-term measurements have been made (Luyten 1977; Shor et al. 1980; McCave et al. 1980). However, the time-scales of these velocity variations are clearly very different from what we observe in the Faro Drift sediments, so that the facies changes in these cores would represent changes in the long-term average velocity of the Mediterranean outflow. We can make some estimate of this average velocity at the present day from current measurements, bedforms seen in bottom photographs and the grain-size of surface sediments (Faugdres et al., in press; McCave, this volume). The channelized flow to the south of the drift attains a velocity of between 50 and 80 cm/s, whereas that to the north may be similar or slightly lower. Surface sediments over most of the drift are muddy silts and our one camera station shows a partly buried and bioturbated ripple field. These features suggest a current velocity less than about 20 cm/s, although there may be a narrow vein of higher velocity flow over part of the drift marked by lack of deposition or slight erosion seen on the 3.5 kHz records (Faugdres et al., in press). For velocity fluctuations during the Late Quaternary, we must rely on the vertical sequences of grain-size variation observed. In a recent paper, McCave (this volume) has seriously criticized the use of Hjulstrom's (1939) well-known erosiontransport-deposition diagram, particularly with regard to fine-grained sediments. McCave has therefore constructed a new transport and deposition diagram for the finer sediments based on a large amount of observational and theoretical data. We have used this diagram (McCave, this volume, Fig. 26) to calibrate a typical coupled negative-positive sequence from the Faro Drift in terms of current velocity. McCave's shear velocity ( U , ) has been converted to flow velocity using an average value for the drag coefficient of 2.5 x 10 -3 (Daily & Harleman 1966). Using the median grain-size we see that the velocity varies from 8 to 18 cm/s; whereas, using the 5th percentile the velocity would be 15-25 cm/s. We prefer the lower estimate and suggest that the 5th percentile, although theoretically more reasonable, is unduly influenced by coarser more buoyant biogenic grains.
290
E.G. Gonthier, J.-C. FaugOres and D.A.V. Stow
Discussion Are they contourites? There is little doubt that the sediments recovered from the Faro Drift are true contourites. Even without reference to their sedimentary characteristics there is the combined evidence of an isolated drift morphology parallel to the continental margin, seismic records of its mode of construction since the early Pliocene, oceanographic evidence for an intermediate level water-mass derived from the outflow of Mediterranean water, and direct current measurements and bottom photographs that show strong bottom currents in the vicinity of the drift. In addition, we cored thick-bedded coarse sandy turbidites in the Faro Canyon area (Cores 30 and 35) and two very thin fine-grained turbidites interbedded with the normal (contourite) sediments at the western end of the drift (Cores 31 and 32). Structurally and texturally these are classical turbidites (Bouma 1962; Piper 1978), and comprise a mixed terrigenous-biogenic suite that is significantly different from the other drift sediments, but very similar to some equally distinctive sandy turbidites from a slope core to the north (Core 29). In particular, the turbidites are richer in glauconite, quartz, feldspar and a varied suite of accessory minerals.
Contourite characteristics and 'sequence' The drift sediments themselves are very similar to contourites that have been described from a variety of other deep-ocean drifts and from some continental rise sites (see summaries in Stow & Lovell 1979: Lovell & Stow 1981 : Stow 1982). Our first contourite facies~ the thin irregular and medium-bedded sand and silt, has sharp or gradational bed contacts~ is massive and bioturbated, fine-grained and moderately well-sorted, and has a thoroughly mixed terrigenous-biogenic (planktonics and benthonics) composition in which the biogenics are commonly fragmented and iron-stained (cf. sandy contourites of Lovell & Stow 1981). There are no internal primary structures remaining, such as cross-lamination or parallel-lamination, which were considered an important feature of the original contourites described by Hollister & Heezen (e.g. 1972). The second contourite facies identified from the Faro Drift is the mottled silt and mud facies. This is mostly well-bioturbated with large and small burrows and mottles, but shows some evidence of current deposition in the irregular wavy lamination and sharp or erosive bed contacts. It is fine-grained and poorly-sorted with
grain-sizes ranging up to coarse silt or fine sand. The mixed terrigenous-biogenic composition is similar to that of the other facies and shows markedly variable proportions of the different elements. The third and volumetrically most important contourite facies we have called homogeneous mud because of its massive structureless appearance to the naked eye. X-radiographic or microfacies examination, however, reveals sharp or erosive contacts with intervals of planar or wavy lamination within a thoroughly bioturbated, poorly-sorted, fine silty mud. The mixed terrigenous-biogenic composition has less of the coarser elements than the other facies and is less fragmented. Both the mottled and homogeneous facies have been described previously as muddy contourites (Stow & kovell 1979; Stow 1982). However, by recognizing an intermediate (mottled) facies it is possible to document characteristic vertical successions of the three facies. Ideally, we have a coupled negatively-graded to positively-graded sequence up to about 100 cm in thickness (Fig. 12), although in many cases only partial or reduced sequences occur. The full sequence represents more or less continuous deposition from persistent sediment-charged bottom currents. Sharp and scoured bedding contacts indicate short intermittent erosive episodes. The most likely explanation for the variation in grain-size and other features is a long-term variation in the average current velocity superimposed on the short-term variability that is known to occur in bottom currents. Such a long-term variation may be caused by an actual change in current strength as a result of climatic or oceanographic factors, or by lateral migration of the current axis. Variation in the grain-size of material supplied to the system is another possible explanation for the "sequences' developed in the Faro Drift cores. However, this is considered a less likely alternative for the following reasons: (a) other current-related features such as sharp or erosive contacts and the shape of the granulometric curve vary in addition to grain-size changes; (b) variation in the input of biogenic material might be expected to mirror climatic changes to some extent, but in fact the biogenic grain-size varies in parallel with that of the terrigenous material through the sequences; and (c) the silt and sand facies occur within both the Holocene and last glacial core sections. Bioturbation of the sediments has kept pace with deposition in most cases so that much of the primary depositional structure and fabric has been destroyed. The high degree of bioturbation
Contourite facies of the Faro Drift, Gulf of Cadiz is similar to that described from the Eirik Ridge contourites by Chough (1978) and Chough & Hesse (in press) and several 'storeys' or bioturbational depth zones are commonly present (Wetzel 1982; Werner & Wetzel I982). We suggest that the variation in size and type of burrows and mottles between the different facies is a direct result of variation in grain-size and/or current strength.
2 91
so far been described in the literature other than from bottom photographs (see Heezen & Hollister 1971). Neither did we encounter more purely biogenic contourites such as those described from the Snorri and Hatton Drifts in the North Atlantic (Laughton, Berggren et al., 1972; Montadert, Roberts et al. 1979).
Conclusions Comparison with turbidites and hemipelagites The contourites described here from the Faro Drift are unlikely to be confused with classical medium to thick-bedded sand-mud turbidites (Bouma 1962), or with other coarser-grained and thicker-bedded resedimented facies (Walker 1978). They are more similar in a number of respects to both fine-grained turbidites and hemipelagites and we concentrate therefore on their distinction from these facies (see also Stow & Piper, this volume). Fine-grained turbidites show a standard, commonly partial, sequence of primary currentinduced sedimentary structures. Close superposition of partial sequences may obscure the individual turbidite units but nevertheless preserve the characteristic structures such as fading ripples, low amplitude ripples, horizontal silt-mud lamination and so on. Bioturbation is restricted to or concentrated towards the tops of units, or may be completely absent. The grain-size is commonly from clay to fine sand size with positive grading through graded laminated units and moderately good sorting within single laminae. The composition is commonly exotic to the region and distinct from the interbedded sediments. Hemipelagites show no vertical sequence and are mostly monotonous, homogeneous, sandy silty biogenic muds. There is no evidence of any primary current structures and bioturbation is usually very thorough. The burrowing organisms are those that have adapted to fine-grained sediments, slow sedimentation rates and an absence of bottom currents. We suggest that the burrow assemblage is therefore different from that of contourites. The mixed terrigenous-biogenic composition is relatively constant throughout and many of the tests remain intact. We did not recover any coarse sand or gravellag contourites from the Faro Drift, although they may be present in the axes of the flanking troughs. This type of contourite facies would clearly have different characteristics but has not
(1) The Faro Drift in the Gulf of Cadiz is one of the clearest examples of a relatively small elongate sediment drift almost certainly constructed since the early Pliocene by bottom currents related to the deep Mediterranean Outflow. (2) A closely-spaced coring programme over the drift has therefore allowed the characterization of contourite facies in greater detail than has previously been possible. (3) On the continuous spectrum of deposits from muddy to sandy contourites described by Stow & Lovell (1979), we have chosen to document three facies types: sands and silts, mottled silts and muds, and homogeneous muds. (4) This threefold division has enabled us to recognize the vertical arrangement of facies in both positively and negatively-graded 'sequences'. The complete coupled negative to positive sequence can be interpreted in terms of a long-term regular increase then decrease in average current velocity of the order of 8 to 18 cm/s. (5) The detailed sedimentary characteristics of the Faro Drift contourites have been compared with those of typical fine-grained turbidite and hemipelagite facies. These should be helpful in the identification of both ancient contourites and modern contourites in areas where the interaction of different processes is more complex. (6) Further work is in progress on these sediments, particularly with regard to their horizontal and vertical distribution and its interpretation in terms of the late Quaternary pattern and history of outflow from the Mediterranean.
ACKNOWLEDGEMENTS: We thank the captain, officers, crew and scientists aboard the RV NOROIT for a pleasant and successful cruise. Technical and secretarial help from our respective departments at Bordeaux and Edinburgh was much appreciated. D.A.V.S. also acknowledges support from the Universit~ de Bordeaux I as a professeur associO. Bryan Lovell and Julian Holbrook reviewed an earlier version of the manuscript.
References AMBAR, I. & HOWE, M.R. 1979. Observations of the Mediterranean outflow--II The deep circulation in the vicinity of the Gulf of Cadiz. Deep Sea Res., 26A, 555-68. AUFFRET, G.A., SICHLER, B. & COLENO, B. 1981. Deep-sea sediments texture and magnetic fabric, indicators of bottom currents regime. Oceanologica Acta, 4, 475-88. BALI)Y, PH., BOILLOT,G., MO1TA, I. & MOUGENOT, D. 1975. Structure gOologique du plateau continental sud-portugais. C.R. Acad. Sci. ParLv, t271, serie D, 1063-6. BOUMA, A.H. 1962. Sedimentology 0[ sonw Fh'sch Deposits. Elsevier, Amsterdam. 168pp. CHOUGH, S.K. 1978. Morphology, Sedimentat 3" Facies and Processes of the North West Athmtic Mid-Ocean Channel between 61 and52 N, Labrador Sea. Ph.D. Thesis, McGill University. 167pp. -& HESSE, R., in press. Contourites from the Eirik Ridge, south of Greenland. Sed. Geol. DAILY, J.W. & HARLEMAN, D.R.F. 1966. Fluid Dynamics. Addison-Wesley Publishing Co. Inc., Massachussetts. 454pp. FAUGERES, J.-C. et al. 1979. Evolution de la sOdimentation profonde au Quaternaire rOcent dans le Bassin nord-atlantique: corps s~dimentaires et s6dimentation ubiquiste. Bull. Soc. geol. Fr. 21(5), 585-601. , STOW, D.A.V. & GONTHIER, E., in press. Contourite drift moulded by deep Mediterranean outflow. Geology. GIESEL, W. & SEmOLD, E. 1968. Sedimentechogramme von Ibero-Marollanischen Kontinentalrund. Meteor Forsch Erg., 1, 53-75. HEEZEN, B.C. & JOHNSON, G.L. 1969. Mediterranean undercurrent and microphysiography west of Gibraltar. Bull. Inst. Oceanogr. Monaco., 67, no. 1382, 95pp. -& HOLL1STER,C.D. 1971. The Face O[the Deep. Oxford Univ. Press, New York. 650pp. HJULSTROM, F. 1939. Transportation of detritus by moving water. In: Trask, P.D. (ed.), Recent Marine Sediments--a Symposium. Am. Ass. Petrol. Geol., Tulsa. 5-31. HOLLISTER,C.D. & HEEZEN, B.C. 1972, Geologic effects of ocean bottom currents: western North Atlantic. hT: Gordon, A.L. (ed.), Studies in Physical Ocetmography, 2, 367-6. Gordon & Breach Science Pubs. LAUGHTON, A.S., BERGGREN, W.A. et al, 1972. Init. Repts. DSDP. US Govt. Print. Off. Washington DC. LOVELL, J.B,P. & STOW, D.A.V. 1981. Identification of ancient sandy contourites. Geology, 9, 347-9. LUTYEN, J.R. 1977. Scales of motion in the deep Gulf Stream and across the Continental Rise. J. Mar. Res., 35, 49-74. MADELAIN, F. 1970. Influence de la topographic du fond sur l'6coulement m6diterranean entre le detroit de Gibraltar et le Cap Saint-Vincent. Cah. Oceans. XXII Ann. No, 1, 43-61. MCCAVE, I.N., LONSDALE, P.F., HOLLISTER, C.D. & GARDNER, W.D. 1980. Sediment transport over the Hutton and Gardar contourite drifts. J. sed. Petrol., 50, 1049-62. MELIERES, F. 1974. Recherches sur la Dynamique S~dimentaire du Golfe de Cadix ( Espagne ). Th~se d'Etat, Univerisit~ Paris VI.
--
NESTEROFF,W.D. & LANCELOT, Y. 1970. Etude photographique des fonds du Golfe de Cadix. Cah. Oceans. CCII Ann. No. 1, 63-72. MONTADERT, L., ROBERTS,D.G. et al. 1979. lnit. Repts. DSDP, 48, US Govt. Print. Off. Washington DC. MOUGENOT, D. & VANNEY,J.R. 1980. G6omorphologie et profils de reflexion sismique: interpr&ation des surfaces remarquables d'une platforme continentale. Ann. Inst. Oceanogr. Paris, 56 (s), 85-100. -& VANNEY, J.R. 1982. Les rides de contourites Plio-Quaternaires de la pentc continentale sud-Portugaise. Bull. Inst. Geol. Bassin D'Aquitaire, 31, 131-9. - - , MONTEIRO,J.H., DUPEUBLE,P.A. & MALOD, J.A. 1979. La marge continentale sud-portugaise: evolution structurale et sedimentaire. Ciecias da Terra (UNL), 5, 223-46. PASTOURET, L., AUFFRET, G.A. & CHAMLEY, H. 1978. Microfacies of some sediments from the western North Atlantic: palaeoceanographic implications (Leg 44 DSDP). In: Benson, W.E., Sheridan, R.E. et al., Init. Repts. DSDP, 44, US Govt. Print. Off. Washington DC. 477-501. PIPER, D.J.W. 1978. Turbidite muds and silts on deep-sea fans and abyssal plains. In: Stanley, D.J., Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. REID, J.L. 1978. On the mid-depth circulation and salinity field in the North Atlantic Ocean. J. geophys. Res., 83, 5063-7. RICHARDSON, M.J., WIMBUSH, W. t~ MAYER, L. 1981. Exceptionally strong new-bottom flows on the continental rise of Nova Scotia. Science, 213, 88?-8. RIVIERE, A. 1977. M~thodes Granulom~triques." Techniques et Interpretation. Masson, Paris. 170 pp. SHOR, A.N., LONSDALE, P., HOLLISTER, C.D. SPENCER, D. 1980. Charlie-Gibbs Fracture Zone: bottom water transport and its geologic effects. Deep-sea Res., 27A, 325-45. STOW, D.A.V. 1982. Bottom currents and contourites in the North Atlantic. Bull. Inst. Geol. Bassin d'Aquitaire, 31, 151-66, --& LOVELL,J.P.B. 1979. Contourites--their recognition in modern and ancient sediments. Earth Sci. Rev., 14, 251-91. VANNEY, J.R. & MOUGENOT, D. 1981. La platforme continentale du Portugal et les provinces adjacentes: analyse geomorphologique. Mere. Sure. Geol. Portugal. 28, 81. WALKER, R.G. 1978. Deep water sandstone facies and ancient submarine fans: Models for exploration for stratigraphic traps. Am. Ass Petrol. Geol. Bull., 62, 932 66. WERNER, F. & WETZEL A, 1982. Interpretation of biogenic structures in oceanic sediments. Bull. hTst. Geol. Bassin d'Aquitaire, 31,275-88. WETZEL, A. 1982. Biogenic sedimentary structures in a modern upwelling area: the NW African continental margin. In: Suess, E. & Thiede, J., (eds), Coastal Upwelling: Its Sedinwnt Record. Plenum Press, New York. ZEYK, W. 1975. On the Mediterranean Outflow of Gibraltar. Meteor Forsch. Erg., A 916, 23-43.
E. GONTHIER AND J.-C. FAUGERES, D6partement G6ologie et OcOanographie, Universit6 Bordeaux I, Avenues des FacultOs 33405 Talence Cedex, France. D.A.V. STOW, Grant Institute of Geology, Edinburgh University, West Mains Road, Edinburgh EH9 3JW, Scotland.
The sediment texture of contourites in Lake Superior J.D. Haifman and T.C. Johnson S U M M A RY: Previous sedimentological studies in Lake Superior indicated that deep water circulation is sufficiently intense in some regions to influence the deposition and erosion of postglacial clay. This is particularly evident off the Keweenaw Peninsula, where 3.5 kHz profiles, side-scan sonar, and cores reveal wavy bedforms and interbedded muds and sands interpreted to be contourites. This region is a good lacustrine analogue to the marine environment, yet is unique in its lack of interbedded turbidites and significant bioturbation. The sedimentary texture of the core top ( < 2 cm depth in core) samples in the region, determined by pipette and sieve analyses, changes from sand (md = 2.2 phi: 220 ~m) at the base of the slope, where the currents are strong, to clay (md = 10.8 phi: 0.6 ltm), farther offshore, where the currents are weak. Textural analyses reveal a distinct bi-modal distribution of sand and clay populations, with the sand dominating the sediment at the base of the slope, and clay dominating the sediment further offshore. The silt fraction of the sediment samples was analysed by an electronic particle analyser, which detected changes in the particle-size histograms from coarse-grained (md=4.79 phi, 36 l~m) and positively skewed (0.45) distributions to fine-grained (rod=6.06 phi: 15 /Lm) and negatively skewed (-0.09) distributions, that correlate to the previous sand and clay textures, respectively. There is a down-core increase in grain-size in two piston cores that suggests that bottom currents decreased in intensity from 9500 years ago to the present. Previous work indicates that contourites and turbidites can be distinguished both by regional variations and by criteria found in a single core. Regional variations useful for distinction (Stow 1979) included lateral fining and sorting, textural, mineralogic, and grain fabric trends that are either generally perpendicular to the bathymetric contours in the case of turbidities or are generally parallel to the contours in the case ofcontourites. In single cores, climbing ripples, dewatering structures, escape burrows, bioturbation restricted to the tops of beds, and unusually high organic content are more closely associated with turbidites than contourites (Stow 1979). The textural characteristics of single cores which were once designated as criteria to distinguish between turbidites and contourites (Hollister & Heezen 1972; Bouma & Hoilister 1973) are now known to be less useful discriminators (Piper & Brisco 1975: Stow & Lovell 1979.) The textural criteria for a single core are ambiguous for the two depositional regimes for at least three reasons: (1) They were established in sediments that probably have interbedded contourites and turbidites. Where both of these processes deposit sediment in the same area, the persistent contour currents may only partially rework and obliterate the deposits of infrequent turbidites. (2) There is a high degree of bioturbation in the marine environment. Intense bioturbation homogenizes and obscures the sedimentological record; therefore, identifying primary sedimentological parameters caused solely by
physical processes is difficult. (3) A wide range of flow regimes creates a wide range of deposits for both processes (e.g. distal vs. proximal turbidites and sandy vs. muddy contourites). The sedimentological features created by the different processes are more likely to be similar than if a smaller range of regimes existed. For this study, we selected a sedimentary environment that lacks both interbedded turbidites and significant bioturbation. Thus, we examined how the textural characteristics change solely under the influence of contour currents. Seismic reflection profiles (3.5 kHz), sedimentary structures, lake floor photographs (Johnson et al. 1980; Johnson et al. in press), and side-scan sonar records (Flood & Johnson, in press) provide evidence for contour currents influencing sedimentation at the base of the slope off the Keweenaw Peninsula in Lake Superior. Seismic reflection profiles revealed bedforms that are attributable to contour currents. The presence of a scoured trough at the base of the slope in this area is the strongest evidence for erosional processes at work. There is no evidence for turbidites interbedded with these contourites (Johnson et al. 1980). Lead-210 profiles in sediment cores throughout Lake Superior reveal a homogenized layer to a depth no greater than 1 to 2 cm that is usually attributed to biological mixing (Evans et al. 1981). X-radiographs of vertical sediment slabs indicate that fine laminations are well-preserved, with little or no disturbance by the benthic community. 293
J.D. Halfman and T.C. Johnson
294
\
b 0
E
0
tD
O.
.,,..a
O_ I
.ca
2~
,'~.k L~
~
I ~
~8
~.~
~
g~
'
0
~ I
oo
\
1 ,It
~
o
9
8~
,
~
~a
~e
The sediment texture o/contourites in Lake Superior The texture of the surface sediments can be a sensitive indicator of depositional mechanisms including the overlying bottom current velocity and flow direction (Visher 1969; Vernet et al. 1972; McCave & Swift 1976; Ellwood & Ledbetter 1977, 1979; Huang & Watkins t 977; Ledbetter 1979; Stow 1979; Allison & Ledbetter 1982). Both surface sediment samples and down-core samples were analysed in this study for trends in grain-size to determine the spatial characteristics of the bottom currents in the Keweenaw area and their variability during the past 9500 yrs.
Methods Twelve box cores (50 x 50 x 40 cm deep), ten gravity cores (2 m long) and two piston cores (6 m long) were obtained along two transects perpendicular to the slope off the Keweenaw Peninsula (Fig. 1). A gravity core was recovered adjacent to each box core to check the textural properties of adjacent sites. Approximately 10 g of wet sediment from the upper 2 cm of each core were sub-sampled for grain-size analysis. Both piston cores were subsampled every 4.8 cm from the surface to 10 cm below the transition from uniform postglacial muds to varved glacial-lacustrine clays that are about 9500 yrs old. The postglacial/glacial contact corresponds to the final retreat of the Laurentian Ice Sheet from the Lake Superior drainage basin (Farrand 1969). Core 14G was subsampled every 3 cm to the bottom of the core. The postglacial/glacial contact was not observed in this core. The silt fraction (5.85 to 71.44 microns) of all subsamples was sized by volume utilizing an electronic particle analyser (Elzone, Model 80XY, manufactured by Particle Data, Inc.). The silt fraction was isolated by wet sieving at 62.5 pm to remove the sand and by repeatedly centrifuging and decanting the clays ( < 4 ~m) from the silts. After the last centrifugation, the silt particles were dispersed in an electrolyte (Hematall by Fisher Scientific) by hand shaking. A 1 ml aliquot from the homogenized silt slurry was added to 250 ml of electrolyte for subsequent analysis of 20 000 particles through a 190/~m orifice by the particle analyser. The particle counts for each size range were converted to volume (or mass) assuming spherical particles with a diameter equivalent to the particle size measured. The diameter increment between adjacent data points was 0.03 phi for this study. The average deviation between duplicate analyses for the 178 subsamples analysed in this investigation was 0.04 phi, and ranged from 0.00 phi to 0.15 phi. The deviation exceeded 0.08 phi for only 6% of the analyses. The
295
detailed procedure of sample preparation, modified from Shideler (1976), is included in Halfman (1982). The surface samples from all of the gravity cores were analysed by pipetting (1 phi interval to 11 phi) and sieving (0.5 phi interval) before analysis of the silt fraction by the particle analyser. The remaining subsamples were analysed by a modified pipette and sieve procedure to determine the relative weight percent of sand, silt and clay before analysis of the silt fraction by the particle analyser. Statistical parameters (median, mean, mode, standard deviation, skewness, and kurtosis) were determined for all of the data sets utilizing the inclusive graphic formulas and nomenclature of Folk (1974), and is reported in Halfman (1982).
Results and discussion Core top analysis The trough at the base of the slope off the Keweenaw Peninsula is elongate (2 to 5 x 20 km), and the long axis is parallel to the regional bathymetry. Both 3.5 kHz profiles (Johnson et al., in press) and side scan sonar (Flood & Johnson, in press) suggest a decreasing influence of bottom currents on the sedimentation from the trough, where the currents are strong, to the open lake, where the currents are weak. Results from the pipette and sieve analyses of the total sediment shows that the trough sediments are coarser (mean = 162/~m, median = 219 /~m, Fig. 2), more skewed (0.67) and more leptokurtic (4.63 to 2.34) than the open lake sediments (mean =0.7/~m, median 0.6 gm, skewness = -0.45~ kurtosis = 1.22 to 0.57). The statistical differences correlate to the change in the sand content from 80 to 10% from the trough to the open lake (Fig. 2). The total sediment analyses indicate that the sediments in the trough are influenced by stronger bottom currents than further lakeward. Frequency histograms of the pipette and sieve data reveal a bi-modal sediment distribution, consisting of sand and clay populations (Fig. 3). The sand population has a characteristic mode of 250 microns (2 phi) and a Gaussian distribution for all of the samples except cores 12G, 13BX and 14G (Fig. 3). The uniform mode of the sand population suggests an origin that is either icerafted during the winter and spring months from the beaches on the south shore, or is eroded by bottom currents from the outcropping glacial till in the western part of the study area. Both mechanisms could account for the uniform sand
296
J.D. Haljman and T.C. Johnson
WESTERNCORETRANSECT -0.10
\ \
Z :
:
MEDIAN
\
\ %/
-
-
ILl
--
L ~
-
SKEWNESS
~
0.40
4O
,HI
Z 50 9-
E
-
:
-"
%
SILT
:
%
SAND
m
eL i00
0!;
Q
1
10
2
9
1
4
3
8
7
6
~100 z
~ 2O KM 0 1 2 3
J
~ l~176 f 250 ~
_
_
~
~_
-....
(a) FIG. 2. Grain-size fluctuations from the trough (right) to the open lake (left) as reflected in the silt fraction (determined by particle analyser) and in the total sediment (determined by pipette and sieve analysis). Stratigraphy from visual core descriptions is symbolized by sand (dots), sandy clay (dots and semi-circles), clays (blank), and glacial varves (parallel lines). (a) the western core transect; (b) the eastern core transect.
The sediment texture o[contourites in Lake Superior
297
EASTERNCORETRANSECT -0.10"
9 "1,1,1
-
""* S K E W N E S S
~
,40 \
/
t
4
9
0.40
I,#'t m
z LU
15 m
I00
+it 1
~11+++ 24 20 19 18 17
22 21
23 16 15
14 13
12 11
~oo,~~__ E+!1+1 loo]
i250
KM 0123
~u .
.
.
.
.~
FIG. 2 (b)
, _2
J.D. Halfman and T.C. Johnson
298
mode that does not decrease in size with distance from shore and the distinct normal distribution that indicates a sorting episode occurred sometime during or prior to sedimentation of the sand onto the present lake floor. Side-scan sonar records from the Keweenaw area reveal a more reflective background down current (north-eastward) than up current from the major outcrops of glacial till on the lake floor (Flood & Johnson, in press). More reflective backgrounds characteristically have more sand than less reflective backgrounds (R. Flood 1982, pers. comm.). Thus erosion of the brown outwash sand (Farrand 1969) may be a more likely source for the sand than ice-rafting.
I
I
The sand population of cores 12G, 13BX and 14G is less sorted and the mode is finer-grained than the other samples (Fig. 3). These cores were recovered from the north-east end of the trough. The textural variation observed in the trough from the south-west to the north-east are similar to those predicted theoretically by McCave & Swift (1976), and observed by Stow (1979) and Bulfinch & Ledbetter (1982) due to a decrease in current velocity. The above textural changes are in agreement with two geophysical observations: (1) 3.5 kHz seismic reflection profiles indicate increased sediment thickness to the north-east (Johnson et al., in press) and (2) side-scan sonar records reveal depositional furrows in shallower
I
I
I
I
0 O_ O~
LRTN 220 1 CN
C3
O_ OD
c3 CD_
I'~ Z LJ (_3 s LLI EL
0
0
>(_.D Z ELl
d_ o0
0
02~ ELi r-rEL 0 O_ CO
0 O_ CM
CD C)_
2
-"~:~"~ O0 -~- :" ' 0
2r
O0
~
4' O0
~/
6~
O0
8'
O0
I0' O0
12' O0
I 14 O0
GRAIN SIZE (PHI) (a) FIG. 3. Frequency plots of typical sediment distributions. (a) Core 22G, representing a typical bi-modal distribution. (b) Core 12G, representing trough sediment in the eastern transect. The sand population has more fine sand and coarse silt than the other samples. (c) Core 6G, representing trough sediment in the western transect.
The sediment texture of contourites in Lake Superior
299 CD 0
I
I
I
I
I
I
I
I
C, CD
0 0 -0
_o co
CD
-d
.,.--... i--.4
Z
LLI 1"-,4
"8" v
CO Z o
C]Z cr(_3
-d
o
-04
Q~ LO
Z
Z
Sr_J
o'ool
0"88
o'd6
o'dz.
0d9 ods od~, [ZN33Nqd) , , k g N 3 r q o 3 N . . 4
0"@
o'dz
0dt
o2
I
I
I
I
I
I
I
I
,d
1
-d
0__ c::, L.LJ C:Z, N v Z
r
6
-d
QD 04 Z I.--CC
5-(_)
._.j
0 "OOI
o'd6
0"d8
o'd/_.
0"d9 o'ds o'd~, o'dE (_LNqgN3a) ,kgN-qflO3U_.4
o'dz
o'd~
o
J.D. Hal~man and T.C. Johnson
300
the trough. The change in textural parameters determined for the silt fraction is parallel to the trend established by the pipette and sieve analyses. Simplification of the cumulative frequency curves of the pipette and sieve data was attempted by dividing the curves into straight line segments to represent different modes of transportation and deposition (Visher 1969). The cumulative frequency plots reveal some straight line segments (Fig. 5). Data points in the silt population, however, do not fall exactly on a straight line, and their deviation from the straight line is amplified
waters east of the trough (Flood & Johnson, in press). The results of the electronic particle analysis on the silt fraction reveal an increase in the median (14.99 to 36.15/~m) and mean (15.30 to 32.58/tin) grain-size, and the skewness ( - 0.09 to 0.45) from the open lake to the trough (Fig. 2). Mode, kurtosis and standard deviation, however, do not reveal any trends (Halfman 1982). The cumulative frequency curves generated by the particle analyser (Fig. 4) reveal a distinct change in the shape due to an increase in the relative proportions of coarse and fine silt from the open lake to
99.9
99
E '=) 0 >
t
90
9 Q
v
9
o
D fq.)
~
9
50
* D
O
t~
9
121
9
|
1
9
D" I:D
t
i
9 tt-
0
Y
O
9
,IF
z3
E
O
10
9
"IF
9
~3 'l-
99
9
1.0 9
,,
.5BX 9 7BX ~4G "2G
O "1~
7"
"
*9BX
0.1 Illl
70
t
50
t
I
I
I
I
I
I
30
I
I
10
Diameter (Fm) (a) FzG. 4. Cumulative frequency percent curves of the silt-fraction determined by the particle analyser (see text). Every fifth data point was utilized to construct the plots. (a) the western core transect; (b) the eastern core transect.
The sediment texture of contourites in Lake Superior
3os
m~ NG
99
,
#
..~ m , O
E 0
m
+, 0
m.
90
m.
O*
O. m, 9 . 9 9
0 9
(3) (D
50
.
-
"
O. 9
0
#
0 0
0 0 0 9
0
"
9
E
+
10
-)(-
0 -)(-
0
9
9 11 BX
o *
*
14G 016G 9
O.
1.0
-
9
@
9 21 B X
-)(..
o
0.1 Ill
70
I
I
50
i
I
I
I
1
I
3O
I
10
Diameter (p.m) FIG. 4 (b) by the continuous and smooth cumulative frequency plots from the electronic particle analyser (Fig. 4). Blaeser & Ledbetter (1982) and Bulfinch & Ledbetter (in press) published characteristic electronic particle size analyses of the carbonate-free silt fraction of the surface sediments underlying the Antarctic Bottom Water and the Western Boundary Undercurrent, respectively. Their results are analogous to the results of this study in that sediments underlying the core of the bottom currents are characterized by a relatively coarse grain-size. The marine studies, however, showed the skewness of the current-influenced sediments does not differ significantly from the uninfluenced sediments, but that the sorting does. This is contrary to our results. The differences between
the results of our study and the marine investigations can be explained by the magnitude of the change between the relative proportion of coarse and fine silt in the influenced versus the uninfluenced sediments. The marine investigations revealed a relatively smaller change than our study did in the relative proportions of coarse to fine silt. The size distribution, however, of the coarse and fine silt populations for both this study and the marine investigations are similar. The 3.5 kHz profiles illustrated by Johnson et al. (1980) clearly reveal that the coarse grain-size of the trough sediments results primarily from erosion of fne-grained material rather than from addition of coarse sediment. A typical histogram of the silt fraction of open lake sediment was changed theoretically by subtraction of fine silt
J.D. H a l / m a n and T.C. Johnson
302
i
LRTN 166 1 CN oDF-Z LJ (_0 Eli
>co Z
WCS~LLJ D_ LU > F-__J y-
18D
~2. O0
' 0
2'.O0
4~.00
B ~.00
6Rni N sIZE
8 ~.00
IO.O0 ~
12.00
1
. CIU
(PHI)
FIG. 5. Cumulative frequency percent curve determined by the pipette and sieve method for core 16G. This curve is typical of Lake Superior contourites. The curve can be broken down into three straight-line segments.
('erosion') and addition of coarse silt ('deposition') to see what proportion of the original material had to be removed or added to create a histogram that is characteristic of the trough sediment (Fig. 6). This exercise revealed that over 85 weight percent of the open lake sediment must be eroded to create the trough sediment. The results from the surface sediments of several cores recovered from the same core station often vary considerably because there is a patchy sand cover overlying the sandy mud. The lakeward margin of the eastern core transect is a good example (Fig. 2). The median grain-size of the silt fraction for 18BX is 30.61 #m (5.03 phi), whereas the median grain-size for the other cores are 15.52; 16.40 and 18.33 pm (6.01, 5.93, and 5.77 phi). The pipette and sieve analyses indicate similar inconsistencies. The X-radiographs of box cores that recovered anomalously coarser sediments than the adjacent gravity cores reveal an
atypical sandy layer overlying typical sediments for that core station. The textural analyses of 7BX (W.H. Busch 1982, unpub, data) at a depth below its sandy cap reveals a textural distribution similar to that in an adjacent gravity core (8G). Side-scan sonar records reveal crescentic depressions about 100 m in diameter that have hard reflectors on their floors (Flood & Johnson, in press). The seismic profiles sometimes indicate the presence of a hard reflector between overlapping acoustic echoes (Johnson et al. 1980; Johnson et al., in press). The evidence suggests that coarse lag deposits are preferentially winnowed in the localized depressions on the lake floor, and cause the patchy distribution in surface sediment grain-size. The median grain-size of the silt fraction in the trough sediments, revealed by the particle analyser, increases from the south-west to the northeast, contrary to the results of the pipette and
The sediment texture of contourites in Lake Superior
303
~-
NORMAL DISTRIBUTION
9
DISTRIBUTION 'DEPOSITED'
Q)
E
DISTRIBUTION
10-
'ERODED'
0
>
% %
>,, c
(1)
6-
IST (D LL E
2 -
E;
9
~_
9
I--1-11 60
[ 40
1
I
30
I
20
9
~
~
~
1 10
Diameter (/~m) FIG. 6. Hypothetical grain-size populations that were added to or subtracted from the average open lake grain-size distribution to determine how much deposition or erosion is required to create the trough sediments, Every fifth data point was utilized to construct the histograms. sieve data. The frequency histograms of the north-eastern trough reveal modes of sand to coarse silt in addition to the typical 2 phi mode for the sand population (Fig. 3(b)). The additional sand and silt modes, which are interpreted to represent material deposited due to the southwest to north-east decrease in current velocity, added coarse silt to the sediment distribution of the silt fraction to yield results inconsistent with the total sediment analysis. Down-Core
analysis
Cores 23P and 24P both reveal an increase in the median grain-size of the silt fraction with depth in the core to the postglacial/glacial contact (Fig. 7). The correlation coefficient (r) determined from a least-squares linear regression for core 23P is 0.42, which yields a confidence limit of over 99~. The down-core results can be interpreted in three ways: (1) The sediment source or distance to source changed due to a change in lake level. (2) The sediment source changed due to a change in the vegetation in the drainage basin. (3) The sediments in the Keweenaw area were influenced by a variable current velocity through time. Farrand (1969) and Drexler (1981) published results indicating that lake level did not fluctuate in the Keweenaw area after the last retreat of the
Laurentian Ice Sheet (c.9500 yrs BP). The area is close to the isobase of isostatic rebound passing through the controlling outlet for Lake Superior at Sault Ste. Marie (Saarnisto 1975). Palynological studies from the Lake Superior region indicate that the distribution of forest types since 9500 yrs BP has remained the same until the very recent (1880-1930 AO) deforestation by man (Wright 1969; Webb 1974; Maher 1977). The time required for first appearance of spruce trees for reforestation of a recently deglaciated area is about 40 yrs (Birks 1980). Therefore, the change in the median grain-size with depth is due to a change in the bottom current velocity. The results from the surficial sediment analyses Indicate that the fluctuation in the grain-size usually corresponds to the intensity of the regional bottom currents. Median grain-size is larger where the bottom currents are usually faster, although the patchy distribution of sandy caps offshore indicates that coarser size does not always imply stronger regional currents. It may reflect just a local perturbation caused by minor bathymetric features. The down-core trend from a smaller to larger grain-size is noisy, but the major deviations in the trend are consistent between 23P and 24P (assuming parallel sedimentation rates through the Holocene), perhaps due to a change in mean r
.
J.D. Halfman and T.C. Johnson
304
ELZONEMEDIAN GRAINSIZE (JlJM) 23P 22
24 P 20
+
=
18
20
+,
18
16
14 G 14
34 i
30 s
i
26 i
i
11
L
50
im
i' 50
100
Ir 100
i=
gE
1SG
F-
h
o
150 100
lb.
E 200
150 250
300
VARVES(-9500 YBP) 200
FIG. 7. The down-core fluctuations of the median grain-size of the silt fraction analyses (particle analyser) for cores 23P, 24P and 14G. The vertical line represents the average median grain-size for each core. bottom current intensity during the last 9500 years. The magnitude of the noise is less pronounced for 24P than 23P, which may correlate to the more distal location of 24P from the core of the bottom current. The noise is interpreted to reflect sporadic and perhaps more localized episodes of increased or decreased intensity of the bottom currents. The plots of the percent sand with depth in cores 23P and 24P reveal consistent results with the median grain-size plots derived from the particle analyser (Fig. 8), except at two depths in 23P where there are relatively large increases in the percent sand, yet the median grain size of the silt fraction decreases. The large quantity of sand indicates an episode of relatively strong bottom currents. The coarse silt was eroded away from this interval so the resulting grain-size distribution of the silt fraction appears finer than the sand fraction would predict. The difference between
the sand and silt analyses reveals the problem of relying on the silt fraction analysis alone to interpret palaeocurrent history. Core 14G reveals a decrease in the median grain-size of the silt fraction and the percent sand with depth in the core, contrary to what is observed in cores 23P and 24P. Unfortunately, a comparison between the grain-size stratigraphies of 14G and the parallel stratigraphies of 23P and 24P is difficult because chronostratigraphic markers for the three cores are lacking. Both 23P and 24P penetrated to glacial varves, so the complete postglacial record was recovered for these cores. The complete postglacial sequence was not recovered for 14G, however. The down-core trend for 14G may reflect a very localized development of a sandy cap as a crescentic depression formed near the core site in relatively recent times. Alternatively the two piston cores or all three cores may reflect very localized
The sediment texture o/contourites in Lake Superior
3o5
PIPETTE:SAND& SILT (WT. PERCENT) 23P 30 20 10
0
20
24P 10 0
40 r
14G 30 20 r
50 5~i
b.
100
p,
150
100l ul
0
U
Z
,5oi
F-
2oo~
~" 150
250 " ~ _ ~ i
!
I !
300
VARVE~ S>950Y 0BP/
-
200 FIG. 8. The down-core fluctuations of percent sand (dots) and silt (stars) by weight from the pipette and sieve analyses. The vertical line represents the average percent sand for each core.
accelerations of the bottom currents, and not reflect the history of regional bottom current strength in the study area. These results from the down-core analyses are not totally consistent with similar results of other studies in the Laurentian Great Lakes. Median grain-size of a core from Lake Michigan decreases with burial depth (Rea et al. 1980). The Lake Michigan core, however, like core 14G from this study, did not recover the complete postglacial record. The radiocarbon date from the base of the Michigan core was 3500-5000 yrs BP. A core recovered from Thunder Bay in northern Lake Superior reveals a noisy but recognizable trend of increasing grain-size with depth to the postglacial/glacial contact (Mothersill 1979), similar to cores 23P and 24P of this study. A Lake Huron core reveals a uniform grain-size with depth (Mothersill & Brown, in press). Slight increases of the grain-size are observable at the top of the core and between 8000 to 9500 yrs ~P.
The uniform grain-size is not surprising because the core was recovered from the centre of the Goderich Basin and probably is isolated from strong bottom currents. Other Lake Huron cores reveal a decrease in grain-size with burial depth (Graham & Rea 1980). Graham & Rea (1980) attribute the change in grain-size to a change in lake level. The grain-size stratigraphies from this investigation suggest that bottom current velocities were faster just after the retreat of the Laurentian Ice Sheet (about 9500 yrs BP) than in the recent past. The bottom currents in the study area are attributed to the downward extension of the northeastward-flowing, wind-driven Keweenaw Current that presumably is accelerated during storms (Johnson et al. 1980). Thus, a possible decrease in the storm intensity or frequency may have occurred between 9500 yrs BP and the recent past. This is the first evidence reported for the fluctuation of storm intensity or frequency with time in
3o6
J.D. Haljman and T.C. Johnson
the Lake Superior region. More studies such as this will have to be conducted before the 'palaeostorm' record is clearly defined, however, because some cores such as 14G may be indicating very localized rather than regional conditions.
Conclusions Size analysis of the silt fraction usually differentiated the sediments which are under the influence of bottom current activity at the base of the slope off the Keweenaw Peninsula in Lake Superior. The sediments from the trough at the base of the slope are characterized by a coarser median grain-size and an increased skewness, and are influenced by larger current velocities than the open lake sediments. Within the trough, the grain-size decreases and sediment thickness increases from the south-west to the north-east, which correlates to a parallel decrease in the bottom current velocity. The textural characteristics of the trough sediments probably result from the contour current winnowing as much as 85% by weight of the silt-sized sediment supply. Regional textural parameters are successful in delineating the lateral extent of the contour currents. The analysis of the silt fraction, however, occasionally yields results inconsistent with the total sediment analysis.
Uniform sandy layers are concentrated locally in crescentic depression along the lake floor. The textural parameters from the cores that presumably penetrate these sandy layers suggest that textural analysis of a single core to identify contourites might produce anomalous results. The down-core investigation revealed an increase in the grain-size with depth in the postglacial record in two of the three cores examined. The change in the grain-size may have resulted from a gradual decrease in the intensity of the bottom currents from 9500 years ago to the recent past. More studies are needed, however, to detect confidently the history of regional bottom current activity and distinguish it from local aberrations caused by shifting bedforms at the individual core sites. ACKNOWLEDGEMENTS: We thank E.B. Nuhfer, D.J.W. Piper, and C.F.M. Lewis for their suggestions and comments that greatly improved the manuscript. We thank the officers and crew of the R/V Laurentian. We thank R. Flood and W.H. Busch for their personal comments and use of their unpublished data. Contribution number 007 of the Limnology Program, University of Minnesota, Duluth, Minnesota. This work was funded by National Science Foundation grants OCE 8018339 and OCE 8109833.
References ALLISON, E. & LEDBETTER, M.T. 1982. Timing of bottom-water scour recorded by sedimentological parameters in the South Australian Basin. Marine Geol., 46, 131-47. BIRKS, H.J.B. 1980. The present flora and vegetation of the moraines of the Klutlan Glacier, Yukon Territory, Canada: a study in plant succession. Quat. Res., 14, 60-86. BLAESER, C.R. & LEDBETTER, M.T. 1982. Deep-sea bottom-currents differentiated from texture of underlying sediment. J. sed. Petrol., 52, 755-68. BOUMA, A.J. & HOLLISTER, C.D. 1973. Deep ocean basin sedimentation. In: Middleton, G. & Bouma, A.H. (eds), Turbidites and Deep Water Sedimentation. Soc. econ. PaleD. Min. Short Course, Anaheim, 79-118. BULFINCH, D.L. & LEDBETTER, M.T. 1982. Western Boundary Undercurrent delineated by sediment texture at the base of the North American Continental Rise. Deep Sea Res., in press. DREXLER, C.W. 1981. Outlet Channels Jor the PostDuluth Lakes in the Upper Peninsula of Mich(Tan. Ph.D. Thesis, University of Michigan, Ann Arbor. ELLWOOD, B.B. & LEDBETTER, M.T. 1977. Antarctic Bottom Water fluctuations in the Vema Channel:
-
effects of velocity changes on particle alignment and size. Earth planet. Sci. Lett., 35, 189-98. 1979. Palaeocurrent indicators in Deep-Sea sediment. Science, 203, 1335-37. EVANS,J.E., JOHNSON',T.C., ALEXANDER,E.C., LIVELY, R.S. & EISENREICH,S.J. 1981. Sedimentation rates and depositional processes in Lake Superior using 21~ geochronology. J. Great Lakes Res., 7, 299-310. FARRAND,W.R. 1969. The Quaternary history of Lake Superior. Proe. 12th Conf Great Lakes Res., International Assoc../'or Great Lakes Res., 181-97. FLOOD, R. & JOHNSON,T.C., in press. Side-scan sonar targets in Lake Superior---evidence for current transport of bottom sediments. Sedimentology. FOLK, R.L. 1974. Petrology of Sedimentao, Rocks. Hemphill Publishing Co., Austin. GRAHAM, E.J. & REA, D.K. 1980. Grain size and mineralogy of sediment cores from western Lake Huron. J. Great Lakes Res., 6, 129-40. HALFMAN, J.D. 1982. Textural Anah'sis o/" Lacustrine Contourites. M.S. Thesis, University of Minnesota, Minneapolis. HOLLISTER,C.D. & HEEZEN,B.C. 1972. Geologic effects of ocean bottom currents: Western North Atlantic. -
The sediment texture of contourites in Lake Superior In: Gordon, A.L. (ed.), Studies in Physical Oceanography, 2, 37-66. HUANG, T.C. & WATKINS,N.D. 1977. Contrast between the Brunhes and Matuyama sedimentary records of bottom water activity in the South Pacific. Marine Geol., 23, 113-32. JOHNSON, T.C., CARLSON, T.W. & EVANS, J.E. 1980. Contourites in Lake Superior. Geology, 8, 437-41. - - , HALFMAN,J.D., BUSCH,W. & FLOOD, R., in press. Effects of bottom currents and fish on sedimentation in a deep water, lacustrine environment. Bull. geol. Soc. Am. LEDBETTER, M.T. 1979. Fluctuations of Antarctic Bottom Water velocity in the Vema Channel during the last 160 000 years. Marine Geol., 33, 71-89. MAUER, L. 1977. Palynological studies in thc western arm of Lake Superior. Quat. Res., 7, 14-44. MCCAVE, I.M. & SWIFT, S.A. 1976. A physical model for the rate of deposition of fine-grained sediments in the deep sea. Bull. geol. Soc. Am., 87, 541-6. MOTHERSILL, J.S. 1979. The paleomagnetic record of the late Quaternary sediments of Thunder Bay. Can. J. Earth Sci., 16, 1016-23. & BROWN, H., in press. Late Quaternary stratigraphic sequence of the Goderich Basin: texture, mineralogy and palaeomagnetic record. J. Great Lakes Res. PIPER, D.J.W. & BRISCO, C.D. 1975. Deep water continental margin sedimentation. In: Hayes, D.E., Frakes, L.A. et al., Init. Repts. DSDP, Leg 28, US Govt. Print Off., Washington, DC. 727-55.
3o7
REA, D.K., BOURBONNIERE,R.A. & MEYERS,P.A. 1980. Southern Lake Michigan sediments: changes in accumulation rate, mineralogy, and organic content. J. Great Lake.; Res., 6, 321-30. SAARNISTO, M. 1975. Stratigraphical studies of the shoreline displacement of Lake Superior. Can. J. Earth Sci., 12, 300-19. SHII)ELER, A.L. 1976. A comparison of electronic particle counting and pipette techniques in routine analysis. J. Ned. Petrol., 46, 1017-25. STOW, D.A.V. 1979. Distinguishing between finegrained turbidites and contourites on the Nova Scotian deep-water margin. Sedimentology, 26, 371-87. & LOVELL, J.P. 1979. Contourites: their recognition in modern and ancient sediments. Earth Sci. Ret'., 14, 251-91. VAN ANDEL, J.H. 1973. Texture and dispersal of sediments in the Panama Basin. J. Geol., 81,434-57. VERNET, J.P., THOMAS, R.L., JAQUET,J.M. & FRIEDLI, R. 1972. Texture of the sediments of the Petit Lac. Eclogae Geol. Heh,., 65, 591-610. VISHER, G.S. 1969. Grain-size distribution and depositional processes. J. sed. Petrol., 39, 1074-106. WEBB, T. 1974. A vegetational history from Northern Wisconsin: evidence from modern and fossil pollen. Tile American Midland Naturalist, 92, 12-34. WRIGHT, H.E. JR. 1969. Glacial fluctuations and the forest succession in the Lake Superior area. Proc. 12th Conf Great Lakes Res., hTternational Assoc.jbr Great Lakes Res., 397-405. -
-
J.D. HALFMANl and T.C. JOHNSON, Limnology Program and Department of Geology, University of Minnesota, Duluth, Minnesota 55812, USA. I Present address: Duke University Marine Lab., Pivers Island. Beaufort, North Carolina 28516, USA.
Facies and sequence analysis of Nova Scotian Slope muds: turbidite vs 'hemipelagic' deposition P.R. Hill S U M M A R Y: Five fine-grained mud facies were identified in piston-cores obtained from the Nova Scotian Slope. The sedimentary structures of the muds indicated deposition under varying bottom-current conditions and rates of sediment supply. Two facies associations were recognized using Markov sequence analysis: (a) turbidite and (b) non-turbidite associations. The problems of recognizing turbidite, contourite and hemipelagic deposits in fine-grained sequences are discussed in the context of low-concentration, low-velocity flows. The dynamics of unconfined flows indicates that turbidity currents become unstable and are deflected by Coriolis to near contour-parallel flow. Dilution and deceleration of the flow may produce a transition to hemipelagic-type deposition. Transitions of this type within individual flows make recognition of discrete turbidite, contourite or hemipelagic processes very difficult in parts of the geological record. The shelfbreak marks an important transition between a shelf regime dominated by wave, tide and current processes (Johnson 1978) and the oceanic regime where waves and tides are much reduced in amplitude. Currents in the oceanic regime are controlled mainly by the large-scale circulation in the ocean, with the secondary effects of topography (e.g. internal wave refraction) often being important on the continental slope. Strong circulation-driven currents have been recognized on several continental margins, generally on the rise (Heezen et al. 1966; Stow & Lovell 1979). Sediment gravity-flows are also commonly associated with the margin seaward of the shelfbreak. Turbidity currents in particular may transport large volumes of sediment across the slope via canyons to the rise and abyssal plain. The continental slope is generally thought to be a zone of coarse sediment bypass and the dominant form of sedimentation is thought to be fine-grained hemipelagic settling (Doyle & Pilkey 1979). The term 'hemipelagic', although defined on the basis of composition (Rupke & Stanley 1974), is generally considered to mean deposition under near-zero current velocities. However, studies of bottom currents on slopes (Hill & Bowen 1983; McGregor, pets. comm.) indicate that velocities are rarely zero, even when there is very little net drift. The transport of fine sediment on the slope by bottom currents and turbidity currents is poorly understood, largely due to problems associated with distinguishing the products of different processes in the sedimentary record. This paper uses Markov sequence analysis to identify mud facies associations on the Nova Scotian Slope and discusses the intergradational nature of the process. The study area (Fig. 1) lies on the continental slope due south of Halifax, Nova Scotia, in a
region where the shelfbreak has a depth of approximately 250 m. Twenty-four piston-cores were collected and ten were used in the sequence analysis. The cores were split and X-radiographed, some sections in thin (0.5 cm) slabs for additional resolution.
Physiography A detailed acoustic survey of the study area revealed a complex slope morphology (Fig. 1, details in Hill 1981), related largely to massmovements on various scales. Two turbidity current channels have been identified in the area. The 'main channel' decreases in size downslope from > 1 km width and 70 m depth on the upper slope to < 200 m width and 15 m depth by 1000 m water depth. The lower reaches of the main channel cross a small depositionat lobe. The second small (20 m deep) channel (Fig. 1) in the west of the area appears to be discontinuous and traverses a small area of undisturbed slope. The remainder of the study area shows a very irregular morphology, with relief varying from over 100 metres to a few metres. This irregularity is related to late Pleistocene slumping on various scales (Hill 1981). The presence o f a 1 to 2 metre mainly Holocene mud drape over the whole area indicates that the features are largely pre-Holocene in age. The palaeo-oceanography of the Scotian Shelf during the late Wisconsin is not well defined, but there is evidence to suggest that a floating ice sheet may have been present over submerged parts of the shelf during this time (Mudie 1980). Sediment supply to the slope may have been influenced by the ice cover.
3II
312
P.R. Hill
FIG. 1. General location map showing
morphological features determined from GLORIA II sidescan and seismic profiles, from Hill (1981). Numbers refer to piston core locations.
Stratigraphy Core sequences show a very consistent division between (1) a lower red-brown to brown mud unit with associated silt, sand and gravel beds, and (2) an upper, olive-grey and dark yellow-brown mottled mud unit (Figs 2 and 3). These units are traceable for at least 300 km eastward on the slope (Stanley et al. 1972). The lower brown unit is highly variable in lithology, such that correlation of individual beds was not achieved between cores as closely spaced as 0.5 kin. In contrast, the upper unit shows a generally uniform sequence of interbedded olive-grey and dark yellow-brown beds which can be correlated between almost every core. Radiocarbon dating of dispersed organic material in the cores suggests that the age of the boundary between units 1 and 2 lies between 18 000 and 20 000 yrs BP (Fig. 3), although this may be an overestimation since the samples probably contain significant quantities of reworked 'dead' carbon (Hill 1981). The single date in core 58 and the more frequent preservation of
sedimentary structures within unit 1 suggests that the sedimentation rate decreased markedly above the unit l/unit 2 boundary.
Sediments Facies 1
Mottled mud
This facies includes a range of grain-sizes, from sandy-silt to silty-clay and is characterized by the presence of pervasive bioturbation, giving rise to a mottled or patchy appearance (Fig. 4(a)). The X-radiographs often reveal a variety of burrow structure, with Zoophycos being particularly common. Facies 2
Parallel-laminated mud
Generally finer-grained silty-clay, facies 2 muds show thin to very thin parallel laminations in X-radiographs (Fig. 4(b)). The laminae sometimes exhibit thinning-upward sequences on the
Facies and sequence analysis of Nova Scotian Slope muds
313
CORE II CM 0 T . '
2d
~live
Dark
2c
grey
kioturl~ated
sand)'
silt.
yello~,-I town bioturbated
mud.
I ' I . * . 0-- i
2b-
t)live
2o
Dark yellow-brown mud w i t h and bands of olive grey
grey
sandy" s i l t .
101.5
I00
cm,
red-brown
thin mud.
laminae
lamina.
Red b r o w n c l a y e y mud. N a i n l y s t r u c t u r e l e s s , with occasional fine laminae and red/grey brown colour banding near top.
Ib
I:rown,
mainly
structureless
mud.
200 Irm~n to yellm,-I mott 1 iny.
rown mud ~ i t h
sulphide
I
.-'ro~,n f:~ud ~ , i t h
fin{" p a r a l l e l
laminations.
FIG. 2. Type stratigraphic core from the Nova Scotian Slope defined by Hill (1981). This core was obtained outside the study area, but illustrates the main stratigraphic units which can be traced for 350 km along the Nova Scotian Slope.
scale of 2 to 3 cms, but more commonly show little organization and form units several tens of centimetres thick. Facies 3
Wispy-laminated mud
The laminae in Facies 3 are very thin and discontinuous, with an irregular wispy form (Fig.
4(c)). In places, climbing ripple-forms can be seen amongst the less regular wispy laminations. The facies includes a range of grain-sizes from clayeysilt to silty-clay. A similar lithology was recognized by Stow (1977) as thin intervals within turbidite units. Banerjee (1977) generated wispy laminae in a flume study under waning-current conditions.
SCOTIAN 54
55 d
d
GULF
55G d
c
C
2--
57
58
--23 el
9,920
_
2--
0
a I 8,860
tO0
(3
1117450
I
18,790 E
200 I t
"-rI(2_ I_lJ a
a
a I
al
1 28,000 300 i
400
I
500
FIG. 3. Carbon-14 dates (yrs BP) on cores from the Nova Scotian Slope. Date with asterisk is unreliable as it was sampled across a possible discontinuity between Units 1 and 2.
FIG. 4. Representative X-radiographs of the five fine-grained sediment facies, (a) Facies 1, mottled mud; (b) Facies 2, parallel laminated mud; (c) Facies 3, wispy-laminated mud; (d) Facies 4, homogeneous mud; (e) Facies 5, sandy-silt to mud graded couplets.
Facies and sequence analysis of Nova Scotian Slope muds Facies 4
Homogenous mud
t2)
Mud units in which no sedimentary structures are seen, even in X-radiographs, are included in this facies (Fig. 4(d)). Stanley (1981) has recognized a similar lithology in the Mediterranean, and has used the term 'unifite' to describe it. On the Scotian Slope, the facies generally occurs in thin ( < 10 cm) units.
Facies 5
315
\
4
Graded sandy silt to mud couplets
This facies consists of sharp-based thin silt or sandy-silt divisions which grade upwards into silty-clay divisions to form a couplet (Fig. 4(e)). These couplets may occur in bundles or be isolated within a muddy sequence. In places, the couplets may be separated by thin or very thin beds of silty-clay of a distinctly different colour, suggesting rapid deposition of the couplets between periods of more continuous mud deposition.
ORIGINAL
5
Facies associations
Apart from Facies 5, the facies generally have gradational upper and lower boundaries. To help define significant facies associations, a Markov sequence analysis was applied to ten core sequences. The method is identical to that used by Cant & Walker (1976). The final difference matrix (Table 1) highlights the statistically significant transitions between facies as positive values greater than 0.05 (Cant & Walker 1976). Fig. 5 shows the original, observed facies relationships in the form of a flow diagram and is compared to the preferred relationships derived from the Markov chain analysis. Although the total number of transitions is minimal for this method of analysis, two associations are usefully distinguished: (1) Turbidite association, Facies 2, 4, 5.
TABLE 1. Final difference matrix obtained from Markov analys&. Positive numbers greater than 0.05 indicate significant transition probability and are used in Fig. 5. Details of other steps in the analysis given in Hill (1981) 1
2
3
4
-0.34 0.32 -0.11 1 0-10 0.20 2 - 0.22 3 0.31 -0.30 0.43 4 - 0.42 0.31 -0.17 -0.13 5
5
PREFERRED
FIG. 5. Facies sequence analysis. Top: original observed transitions (numbers in brackets indicate number of observed transitions); bottom: preferred facies transitions from Markov analysis. Dashed line indicates insignificant transition probability. The analysis shows the close association between silt/mud couplets of Facies 5 and muds of Facies 2 and 4 (Fig. 5). The dominant sequence is a fining-upward sequence from the silt mud couplets (Facies 5), through parallel laminated mud (Facies 2) to homogeneous mud (Facies 4). The complete sequence is not always present but three typical sequences are recognized from visual examination of cores (Fig. 6): (a) Multiple silt/mud couplets. (b) Complete sequences of silt/mud couplets, parallel laminated mud and homogeneous mud. (c) Parallel laminated mud with slightly coarser basal silt laminae and homogeneous mud. Type (a) sequences are not recognized in the Markov analysis because only gradual transitions were included. Type (c) sequences often consist only of Facies 2 units, with the slightly coarser base allowing identification of individual units. Such distinctions can sometimes be made only in X-radiographs.
316
P.R. Hill
2
2
5
4
E3
!:i:'...
2
12 :.
EI
D
(a )
( b}
(c )
PIPER (1978)
FIG. 6. Typical turbidite mud sequences observed in Nova Scotian Slope cores. Numbers refer to facies defined in the study. Piper's (1978) model sequence for fine-grained turbidites shown in right hand column. The general sequence recognized here is very similar to the standard sequence proposed by Piper (1978) for turbidite muds and silts (Fig. 6). The basal part of Facies 5 is equivalent to a Bouma D division. Piper's model includes three subdivisions of the Bouma E division: E1 laminated mud, E2 graded mud and E3 ungraded mud. Subdivisions E1 and E2 are equivalent to Facies 2 of this study and E3 is equivalent to Facies 4. The sequence analysis suggests that the wispylaminated mud (Facies 3) might make up a fourth member of the association (Fig. 5). This facies shows a strong Markov dependence on the massive mud (Facies 4) and an insignificant dependence on the laminated mud (Facies 2). Stow (1977) in a detailed analysis of muddy turbidites, suggests that a wispy-laminated division occurs between divisions E 1 and E2 of Piper (1978), rather than between E2 and E3 as would be implied by this analysis. Experimental studies of fine sediment transport (Banerjee 1977) support Stow's results. Thus, it seems unlikely, in most cases here, that Facies 3 is part of the turbidite association. (2) Non-turbidite association, Facies 1 and 3. The transition analysis indicates a close twoway relationship between the mottled muds of Facies 1 and the wispy-laminated muds of Facies 3 (Fig. 5). There are generally no textural changes across these transitions; the main change is in the preservation of sedimentary structures. Facies 1 muds are clearly bioturbated to the extent that all primary structures have been destroyed, whereas intervals of undisturbed sediment usually reveal the characteristic wispy lamination of Facies 3. Bioturbated muds on continental slopes are generally interpreted to result from hemipelagic sedimentation (Doyle & Pilkey 1979) as they reflect relatively low rates of sedimentation and contain much biogenic sediment. The close two-
way association of these bioturbated muds with wispy-laminated muds suggest that their respective depositional processes are also related. The preservation of sedimentary structures is the only discernible difference between the two facies, suggesting that the rate of deposition may be the only difference in process. The facies in this association are not directly related to turbidity currents as indicated by the Markov analysis. Yet, the association is not typically hemipelagic, with the preservation of primary sedimentary structures indicating relatively rapid deposition under significant dynamic conditions. It is suggested that both facies were deposited under much the same kind of current regime, with only the rate of sediment supply causing differences in sedimentary structures. Extension of the term 'hemipelagic' to include the wispy-laminated facies is probably justified, rather than invent a new term.
Discussion
Mud turbidites have been successfully identified in this study by making use of the characteristic sequence of structures proposed by Piper (1978) and Stow & Shanmugam (1980). Furthermore, non-turbidite muds can be distinguished by the Markov sequence analysis despite the fact that they show sedimentary structures which are constituents of the Stow & Shanmugam (1980) sequence. It is not surprising, of course, that similar sedimentary structures are observed in muds deposited from different processes, because the range of current velocities from which mud deposits is relatively small. Hence, the near-bed mechanics of deposition may be very similar and concentration of sediment in the flow may be the only distinction between turbidite and non-turbidite deposition. Banerjee's (1977) experiments support this and also suggest that the rate of current deceleration is important to the type and sequence of sedimentary structures that develop. Where sand and silt interbeds are not present, it may be very difficult to identify discrete turbidite or non-turbidite intervals. Gradational bed boundaries may predominate and individual sedimentation units may vary greatly in thickness from a few millimetres up to tens of centimetres. This may be a source of annoyance to the geologist who wishes to neatly categorize the sediments into turbidite, contourite or 'hemipelagite' strait-jackets but it should be recognized that gradations between the three processes are probably common in the ocean. The following analysis attempts to illustrate this point. Consider a turbidity current flowing down a
Facies and sequence analysis of Nova Scotian Slope muds slope channel such as seen in the study area (Fig. 1). Komar (1969) has demonstrated that the Coriolis force is important in channelled flow and explains the difference in the height of channel levees by this effect. However, if the flow becomes unconfined at the end of the channel or by overflow of the channel walls, the flow will tend to be turned by the Coriolis force (to the right in the northern hemisphere and to the left in the southern). The alongslope acceleration is the sum of the Coriolis acceleration and the gravitational component as the flow spreads laterally. This can be expressed by: cqu
~h
where u and v are the alongslope (x) and downslope (y) velocity components, f is the Coriolis force, h the depth of the flow and g" = g ( p t - p ) / p where p, is the flow density, and p the density of sea water. For the latitude of Nova Scotia,f is in the order of 10 -4 s -~ so that for an initial flow velocity of 1 ms-~, the 'turning radius' would be of the order of 10 km. A faster current would have a correspondingly longer turning radius. The time-scale of this turning effect is given by T = 2g/f, i.e. in the order of 15 hours. The final direction attained by the flow (described by the angle 0) is given by: tan 0 -
fh Co V
Where Co is the drag coefficient and V the final velocity of the flow which is considered to be in a sheet-like steady-state. Assuming a reasonable value for Co (in the order of 10-3), the final flow direction is dependent on the ratio h/Vo. For a
317
flow of height 100 m and final velocity 1 ms-1, tan 0 = 10 or 0 ~ 90 ~ giving flow almost parallel to contours. Similarly a 10 m thick flow of the same velocity or a 100 m flow of 0.1 m s - 1 would result in tan 0 = 1 or 0 = 4 5 ~ giving flow oblique to contours. It is reasonable to assume that turbidity currents would vary considerably in flow thickness and velocity so that a complete spectrum of flow directions from downslope to contour-parallel is likely on the slope and rise. The flow velocity would decrease due to lateral spreading and loss of momentum until the flow is driven by the ambient currents only. Thus turbidity currents may both turn into 'contour-currents' if the ambient alongslope currents are strong, or supply sediment to the bottom boundary layer for deposition under 'hemipelagic' type of conditions. Under such a set of intergradational processes, it may be more useful to think in terms of sediment supply rates and flow concentrations as controls of sediment facies, rather than a discrete set of physical processes. Field identification of turbidite, contourite and hemipelagite sediments as such, is never likely to be possible due to the absence of unique distinguishing criteria and their gradational nature. More time-consuming methods such as the Markov analysis applied here, and current orientation by clay fabric analysis must be applied to allow such interpretations to be made.
ACKNOWLEDGEMENTS: My thanks go to Tony Bowen for discussions and calculations concerning the dynamics of unconfined turbidity currents. Jim Syvitsky, Dorrik Stow and Kevin Pickering read an earlier draft of the manuscript.
References BANERJEE,I. 1977. Experimental study on the effect of deceleration on the vertical sequence of sedimentary structures in silty sediments. J. sed. Petrol., 47, 771-83, CANT, D.J. & WALKER,R.G. 1976. 'Development of a braided-fluvial facies, model for the Devonian Battery Point-Sandstone, Qu6bec'. Can. J. Earth Sci., 13, 102-19. DOYLE,L.J. & PILKEY,O.H. 1979. Geology of continental slopes. Soc. econ. Palaeo. ?din. Spec. Pub. 27, 374 Pp. HEEZEN, B.C., HOLLISTER, C.D. & RUDDIMAN W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science, 152, 502-8. HILL, P.R. 1981. Detailed Morphology and Late Quaternary Sedimentation on the Nova Scotian Slope, South of Halifax. Unpubl. Ph.D. Thesis, Dalhousie Univ., Halifax, N.S., Canada. 331 pp.
--
& BOWEN,A.J. 1983. Modern sediment dynamics at the shelf-slope boundary offNova Scotia. In: The Shelf-Slope Boundary: a Critical Interface on Continental Margins. Soc. econ. Palaeo. Min. Spec. Pub. 33. JOHNSON, H.D. 1978. Shallow siliclastic seas. In: Reading, H.G. (ed.), Sedimentary Environments and Facies. Elsevier, New York. 207-58. KOMAR, P.O. 1969. The channelised flow of turbidity currents with application to Monterey Deep Sea Fan Channel. J. geophys. Res., 74, 4544-58. MUD1E, P.J. 1980. Palynology of Late Quaternary Marine Sediments, Eastern Canada. Unpubl. Ph.D. Thesis, Dalhousie Univ., Halifax, N.S., Canada. 464 pp. PIVER, D.J.W. 1978. Turbidite muds and silts on deepsea fans and abyssal plains. In: Stanley D.J. & Kelling G. (eds), Sedimentation in Submarine
318
P.R. Hill
Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-176. - & NORMARK W.R., in press. Turbidite depositional patterns and flow characteristics, Navy submarine fan, California. Sedimentology. RUPKE, N.A. & STANLEY,D.J. 1974. Distinctive properties of turbidite and hemipelagic mud layers in the Alg6ro-Balearic Basin, Western Mediterranean Sea. Smith. Contr. Earth Sci., 13, 40pp. STANLEY, D.J. 1981. Unifites: structureless muds of gravity-flow origin in Mediterranean basins. GeoMarine Letters, 1, 77-83. --, SWIFT, D.J.P., SILVERBERG,N., JAMES, N.P. & SUTTON, R.G. 1972. Late Quaternary Progradation
and Sand 'Spillover' on the Outer Continental Margin off Nova Scotia, Southeast Canada. Smith. Contr. Earth Sci., 8, 88pp. STOW, D.A.V. 1977. Late Quaternary Stratigraphy and Sedimentation on the Nova Scotian Outer Continental Margin, Unpubl. Ph.D. Thesis, Dalhousie Univ., Halifax, N.S. Canada 360pp. - - & LOVELL,J.P.B. 1979. Contourites: their recognition in modern and ancient sediments. Earth Sci. Rev., 14, 251-91. --• SHANMUGAM,G. 1980. Sequence of structures in fine grained turbidites: comparison of recent deepsea and ancient flysch sequences. Sed. Geol., 25, 23-42.
P.R. HILL, Department of Geology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 3J5. Present address: Atlantic Geoscience Centre, P.O. Box 1006, Dartmouth, Nova Scotia, Canada B2Y 4A2.
The role of canyons in late Quaternary deposition on the United States mid-Atlantic continental rise B.A. McGregor, T.A. Nelsen, W.L. Stubblefield and G.F. Merrill SUMMARY: The continental margin in the US mid-Atlantic region is dissected by many downslope-trending canyons, extending from near or at the shelf-edge out onto the rise. Sediments from the shelf and slope are transported seaward through the canyon system to the rise. Sand-size material from the shelf is introduced by means of spillover into the canyons and onto the upper slope. A dendritic gully system dissecting the entire slope also provides pathways for slope sediments, mainly silts and clays, to be introduced into the canyons. Distinct grain-size distributions of the shelf sands which are interbedded or mixed with the fine-grained sediments of the slope and rise can be used as tracers for transport pathways on the continental margin. Canyon erosion on the rise also may provide a local source of sediment which must be considered. Sand-size-sediment distribution and DSRV Alvin observations show that erosion and deposition take place periodically in the mid-Atlantic canyons and on the rise seaward of New Jersey and Delaware. The continental slope and rise seaward of New Jersey and Delaware are dissected by numerous submarine canyons. This study focuses on the slope and rise in the vicinity of Wilmington Canyon, a large canyon seaward of Delaware Bay (Fig. 1). The detailed morphology and canyon processes on the slope and rise can be evaluated by integrating narrow-beam bathymetric data, seismic reflection profiles, mid-range sidescan sonar data, piston cores, grab samples, and observations from DSRV Alvin. Initial study of the US mid-Atlantic continental slope morphology was by Veatch & Smith (1939), whose bathymetric survey indicated that the slope was dissected by numerous valleys (canyons). Energy and resource needs have renewed interest in the continental margin and resulted in new studies and research tools. Recent detailed bathymetric surveys (Kelling & Stanley 1979; Merrill & Bennett 1981; McGregor et al. 1982) and long-range sidescan sonar (GLORIA) images (Teleki et al. 1981; Twichell & Roberts 1982) have refined the interpretation of the slope morphology of Veatch & Smith in the vicinity of Wilmington Canyon. The slope is very dissected by canyons, and the walls of the canyons on the upper to midslope are cut by numerous gullies (Twichell & Roberts 1982). Seismic reflection profiles of the slope and upper rise show the depositional and erosional history of the margin during the Tertiary and Quaternary in the Wilmington Canyon area (Kelling & Stanley 1970; McGregor & Bennett 1981). Sediment samples collected to evaluate the activity of slope and canyon processes have resulted in differences in interpretation because of the varying objectives of the data collectors. After analysing a regionally distributed group of cores
collected as four core transects normal to the shelfbreak and spaced 35 km apart, Doyle et al. (1979) and Keller et al. (1979) concluded that the continental slope of the US mid-Atlantic is a region of deposition of fine-grained sediment and of little transport or deposition of sand-size material. Short-term, current-meter observations in the canyons also suggested that transport processes in the mid-Atlantic canyons are relatively inactive (Keller & Shepard 1978). In contrast, Kelling & Stanley (1970) and Forde et al. (1981), using seismic-reflection and sedimentologic surveys focused on specific canyons, suggested that the transport of sediment, including sand-size material, has taken place periodically in the mid-Atlantic canyons since the late Pleistocene. Data that are integrated in this study include mid-range sidescan images of the topography of the slope and rise (Fig. 2), which provide greater resolution of the gully system first observed on the GLORIA images, and details of canyon-floor morphology. The sidescan images were collected with a mid-range sidescan sonar system (Sea MARC I) as part of a cooperative study between the US Geological Survey and Lamont-Doherty Geological Observatory of Columbia University during August and September 1980. In September 1980, sidescan data were used in dives planned by the Marine Geology and Geophysics Laboratory of the National Oceanic and Atmospheric Administration (NOAA) in Miami. Features identified on the sidescan images were directly observed and sampled from DSRV Alvin. The canyon floors of Wilmington and South Wilmington Canyons were sampled in detail. A series of 6 m piston cores on the slope and rise and 25 grab samples on the shelf were 319
32o
B.A. McGregor et al.
OCEANOGRAPHER, ~ ~ 43 ~ 73 ~
42 ~ 74 ~
HUDSON
41 ~
i
75 ~
l
L
40 ~ 76 ~
~ [ LINS DP EE NN KC OEHRL _
A
,
RE
WlLM~NGTON~ !
BALTM I ORE
39 77
?,
38
36 ~ 79 ~
78 ~ 35 ~
77 ~
34 ~
76 ~
75 ~
33 ~
74 ~
FIG. 1. Index map of the US Atlantic coast showing location of study area outlined by heavy black lines.
Bathymetry is in metres. collected at sites selected by using detailed bathymetry and single channel seismic reflection profiles collected during 1975, 1977, and 1978 as part of a NOAA program to assess the sea floor stability of the continental margin between Wilmington and Lindenkohl Canyons (Fig. 2). The purpose of this discussion is to integrate these different types of data, each with different scales of resolution, in order to evaluate the
processes and transport pathways of the continental slope and rise.
Physiography The slope is dissected by canyons that can be grouped into two broad classes: those that indent the shelf-edge (e.g. Wilmington & Spencer) and
The role of canyons in late Quaternary deposition
32I
e~
r~ o
Q
o
3 o~ O.t~
.=..=
.~.=-
~<
322
B.A. McGregor et al. o
t~
g
<
o o
O
~
m
e~ d ~
~..~
~
f2~
.,.~
The role of canyons in late Quaternary deposition those whose heads terminate on the upper slope (Fig. 2). The general morphologies of the two types of canyons are different. The canyons whose heads terminate on the upper slope have relatively straight valleys trending downslope (Figs 3 & 4), whereas Wilmington Canyon, whose head indents the shelf-edge, has a tightly meandering channel on the slope and upper rise (Fig. 5). Sidescan-sonar images of the canyons whose heads terminate on the upper slope south of Wilmington Canyon indicate scarps around the heads that are suggestive of headward erosion (Fig. 3). The walls of both types of canyons on the slope are gullied, although those that head on the upper slope are more extensively gullied (Figs 3 and 4). In some places, the gullies from two adjacent canyons have completely dissected the intercanyon divide (Fig. 3). Both types of canyons have relatively flat floors and the details of the floors cannot be detected by standard surface ship
323
bathymetric profiling instruments. On the lower slope, the floors of North Heyes and South Heyes Canyons have lineations trending downslope (Fig. 4). Because of this down-slope trend, these lineations are believed to be related to downcanyon flow. Similarities between fluvial systems and the meander patterns and channel morphology seen in Wilmington Submarine Canyon from DSRV Alvin were discussed by McGregor et al. (1982) and Stubblefield et al. (1982). The general trend of the canyons is toward the south-east, with the exception of Wilmington Canyon, which abruptly changes trend to the east at the base of the slope (Fig. 2). The canyons that dissect the slope just north of Wilmington Canyon feed into Wilmington on the upper rise. The many canyons (approximately 22) on the slope merge; only five valleys are present on the rise at approximately 2600 m depth (Fig. 2).
FIG. 4. Sidescan images of gullies on walls of South Heyes Canyon. Dark areas are the canyon floor and gullies. A dendritic pattern in the gullies is also shown. See Fig. 2 for location of area shown in this sidescan image. Large arrow indicates downslope direction. Small arrow points to lineations on the canyon floor.
324
B.A. M c G r e g o r et al.
.=. o
0
~9
o o
0 0
E
N
The role o f c a n y o n s in late Q u a t e r n a r y deposition
Stratigraphy The surficial sediments that have been sampled in the study area are Quaternary in age (late Pleistocene to Holocene). Radiocarbon dating of a core on the slope 45 km south of this area indicates an accumulation rate of 22 cm/1000 yrs for late Pleistocene through Holocene (Doyle et al. 1979). Three cores have been dated (Nelsen 1981) in this study area (Fig. 2: slope samples 8046 and SLP- 1 and rise sample 8044). Although dating of core SLP-1 suggests a sedimentation rate in close agreement with the previously reported rate (Doyle et al. 1979), dates on cores 8046 (slope) and 8044 (rise) indicate rates that are, respectively, nearly one third and twice that value. Dating of the top of core 8046 (slope) suggests a decrease in the sedimentation rate during the Holocene to approximately one tenth the value given by Doyle et al. (1979). The sedimentation pattern on continental margins is greatly influenced by changes in sea-level (Vail et al. 1977). An increased sediment supply from rivers discharging on the east coast at or near the shelf-edge during Pleistocene low stands of sea-level must have resulted in a higher sedimentation rate on the outer shelf and upper slope. The shelf in this area is covered by relict Pleistocene sediments (Emery & Uchupi 1972). Rotary cores from the Baltimore Canyon trough area, show that Pleistocene terrigenous clastic deposits are thick on the slope and in places exceed 300 m (Poag 1979). Seismic-reflection profiles can be used to map the thickness of Quaternary sediments on the slope and rise. In the part of the mid-Atlantic continental rise in the study area, the thickest Quaternary sediments (300-400 m) appear to be in the area of Wilmington and Spencer Canyons (McGregor & Bennett 1981).
Sediments Sediments in the slope and rise (the slope/rise boundary is defined here as the 2100 m isobath) portion of the study area are generally dominated by silts and clays with an average sand content of < 10% while the adjacent shelf typically is > 90% sand. Characteristic grain-size distributions for these areas are shown in Fig. 6. There are frequent exceptions in the sand percentages for both intercanyon and canyon areas, and these will be discussed later. Texturally, the dominant fine-grained component of the slope and rise sediment is transitional from sandy-silts to silty-clays and has a ubiquitous, nearly uniform olive-grey to dark grey (5Y 4/2-4/1) colour. With respect to sedi-
(a) L) z 2o o h
~I _0z
(b)
4~
*~ ~AL ~v r
DOMINANT SHELF SAND MEAN = 1.42 STD. DEV. = 0.63
-I
0
4O
(c)
I
2
3
4
SUBORDINATE .SHELF SAND MEAN = 2.40~ I
,I
STD. DEV.= 0.73i
z~ o-?_.
325
-I
',
I/I,
i
,
~0
I
2
3
4
DOMINANT SLOPE SAND
3O z ZO I.L
o_
"I
0
I PHI
3
4
FIG. 6. Diagnostic sand populations from the study area (from Nelsen 1981). mentary structure, slope cores are generally different from rise cores. A typical slope core, as noted by Forde & Nelsen (1981), exhibits no primary or secondary sedimentary structures, either visually or in X-radiograph images (i.e. 8050, Figs 2 and 7(a)). They interpret this to be the result of active bioturbation on the slope. On the other hand, rise cores typically exhibit abundant primary sedimentary structures dominated by interlayered turbidite sands and silty-clays with secondary features resulting from bioturbation. Cores 8044, 8053, and 8045 (Fig. 2) provide good examples of typical rise core sedimentation patterns and structure. Fine-grained sedimentation on the rise is interrupted by sandy turbidite sedimentation of two general types. The first is exemplified by core 8044 which was taken on a relative topographic high. Fine-grained sedimentation (light-layers) is interrupted, on a millimetre to centimetre scale, by fine to very-fine sand turbidites (Fig. 7(b)). The second general type of turbidite sedimentation is associated with topographic lows in the canyon thalwegs. In an X-radiograph of such a rise thalweg core (8053, Fig. 7(c)), fine-grained sediment (light area) alternates with thicker and coarser sandy turbidites. These turbidites can be highly erosive of the indigenous fine-grained sediments, as seen in Fig. 7(c) for the base of the upper turbidite. With
326
B.A. McGregor et al.
FIG. 7. X-radiographs of typical cores from the study area shown in Fig. 2 featuring (a) a bioturbated, structureless slope core (8050), (b) a sequence of thin, non-erosive fine to very-fine sand turbidites diagnostic of rise sedimentation on relative topographic highs (core 8044), (c) fine-grained sediments (light area) interbedded with massive, normally erosive coarse sand to pebbly turbidites diagnostic of relative topographic lows (i.e. rise canyon thalwegs, core 8053), and (d) bioturbated rise canyon turbidite sequences (core 8045). particle sizes ranging upwards to pebbles, rise thalweg turbidites are normally significantly coarser than turbidites from relative topographic highs. It was shown by Nelsen (1981 ) that some of these pebble-size particles were really 'clayballs' which could be mineralogically traced from rise depositional sites (8053, Fig. 2) to probable source areas on the upper slope (8033, Fig. 2) and head of Spencer Canyon (8015, Fig. 2). Bioturbation in rise canyon sediments serves to increase the sand content in the fine-grained sediment layers and hence the mean grain-size. This process can be seen in core 8053 (Fig. 7(c)) for the lower turbidite, where mild bioturbation has disturbed the top portion of this layer with sandy sediment distributed upward into the overlying finegrained sediment. Taken to an extreme, bioturbation can destroy essentially all primary bedding leaving anomalously sandy intervals within otherwise much finer sediments. This is seen in
Fig. 7(d), for rise thalweg core 8045, where the resulting average sand content is elevated to approximately 27% in this otherwise silty-clay interval. In general, unless a diagnostic mineral tracer is available, as in the 'clayball' example above, fine-grained sediments are difficult to interpret in terms of source, transport pathways, or processes. To aid in the source, pathway and process question, coarse-grained (sand) sediments can be used as a tracer originating from the source area of the outer shelf. On the basis of groupings derived by factor analysis of grain-size distributions from 25 grab samples and 17 cores, Nelsen (1981) showed that two distinct populations of quartz sand can be mapped on the shelf and that a third population dominates the sand fraction in the predominantly mud facies of the slope and rise. Whereas the dominant shelf population (Fig. 6(a)) is repre-
The role of canyons in late Quaternary deposition sented by a moderately well-sorted medium sand, and the subordinate shelf population (Fig. 6(b)) is represented by a moderately sorted fine sand, the sand population for the midslope to rise is dominantly well-sorted and is very-fine sand (Fig.
6(c)). Using these three sand populations as tracers, the original source of the sand fraction cored on the slope and rise can be determined. These tracers mixed or interbedded with the finegrained sediment can also be used to recognize transport paths on the margin for shelf-edge, spillover and downcanyon transport. Sand spillover from the shelf was described by Stanley et al. (1973), in this portion (Fig. 1) of the mid-Atlantic slope. Spillover occurs in both canyon and intercanyon areas. The process was demonstrated for Wilmington Canyon by Stanley (1974) and for Spencer Canyon by Forde & Nelsen (1981). In the intercanyon areas, spillover can be traced by the grain-size distribution noted above. These spillovers can be locally extensive and have been shown (Nelsen & Bennett 1981) to result in mesoscale ( ~ 0.5 m) soft-sediment deformation structures on the upper slope or, more subtly, to result in dispersed sand in bioturbated intercanyon slope cores (Nelsen 1981: Neisen & Bennett 1981). On the slope, the quartz sand content in canyon cores averages 18.6% and that in intercanyon cores averages 6.2%. In Spencer Canyon, the average sand content is 29.5 ~ and some intervals contain as much as 58% sand, a value more than twice the maximum found on the intercanyon slope (Nelsen 1981). As noted above for Fig. 7(a), sand is rarely found in layers on the intercanyon slope. The exceptions to the generally structureless intercanyon slope cores are the mesoscale soft-sediment deformation structures noted above and cores 8033 and 8016 south of, and adjacent to, Spencer Canyon. In the latter two cores, sand content in the fine-grained sediment averaged 32% in the upper 2 m and had values as high as 75~ in the top 10 cm. No clear trends exist in the sand content of other intercanyon slope cores. In the slope portion of Spencer Canyon, however, the non-bioturbated sandy sequences in the top 5 m contain high proportions of sand similar to that currently found on the adjacent shelf, suggesting late Pleistocene and Holocene seaward transport of this material (Forde & Nelsen 1981). The sand in the rise turbidites has been shown, through the tracer method noted above, to be of the same type as that currently mantling the outer shelf in the study area (Nelsen 1981). As these shelf sediments underwent seaward dispersal as turbidity flows, their commonly erosional nature
327
(Fig. 7(c)) affected fine-grained slope sediments in two ways. First, turbidites which erode the indigenous fine-grained sediments resuspend and incorporate these sediments into the flow thus providing a mechanism for their net down-slope transport. Secondly, downslope transport of fine sediments can and does take place as 'clayballs' which are transported along with the turbidity flow and deposited in the coarse basal portion of the turbidite. In the study area Nelsen (1981) showed this process to be active to a water depth of at least 2700 m. These processes, however, have undergone an apparent decrease in frequency and magnitude on the rise. In contrast to the slope cores, the rise cores show a clear trend upsection with a decrease in total sand content as well as a decrease in number of sand layers and lenses from the late Pleistocene and Holocene (Forde & Nelsen 1981). Sediment samples were collected from the small topographic features on the floors of Wilmington and South Wilmington Canyons on the rise (2300-2400 m depth) from DSRV Alvin in order to show localized variability of the sediments. The sediment type is generally a light-grey lutite. Extensive bioturbation in the form of pits, mounds, scratches, burrows, tracks and trails was observed on the floors of both canyons, and burrows were observed in the 130 cm-long cores. Cores from the inside and outside of a meander of Wilmington Canyon showed differences in the percentage of dispersed sand. On the inside of the meander (area comparable to a point bar in a fluvial system) the core contained 6% sand; whereas a core from the deeper part of the floor adjacent to the outside wall of the meander contained as much as 40}/0 sand. Lead-210 measurements in a core from Wilmington Canyon indicate that the accumulation rate of fine sediments on the floor of the canyon is a few millimetres per year (C. A. Nittrouer, pers. comm. 1982). Besides sediments originating from the shelf and slope, a local source on the rise was also observed from DSRV Alvin. A loosely bound gravel conglomerate together with disaggregated gravels and a sandstone unit were observed and sampled in South Wilmington Canyon (Stubblefield et al. 1982). Through erosion these coarse materials are believed to have been exposed in the adjacent north wall of South Wilmington Canyon. We believe that this wall may be cut into a large slide block of displaced shelf-edge sediment (Stubblefield et al. 1982). Ridges having as much as 20 m of relief on the canyon floor were composed of upturned clay units having sediment cover ranging from zero on the ridges to more than 100 cm on the troughs. Grain-size analysis of
328
B.A. McGregor
the lutite in troughs between ridges showed a sand content of 20%. Piston cores from the continental rise show that the style of deposition varies with the morphology. Detailed sampling from DSRV Alvin indicates that grain-size variations are also related to the microtopography. These variations are important in understanding processes responsible for sediment transport and deposition.
Processes The extensive dissection of the slope results in a large number of canyon heads at and indenting the shelf-edge. Proximity of the canyon heads to the shelf allows the sand-size material mantling the shelf to be swept (spillover) into the canyons. The extensive gully system on the slope, observed feeding into the canyons, may affect the amount of fine-grained silts and clays available for transport. Because the depth of water in which the gullies are found (400-1500 m) is significantly greater than the lowest known stands of sea-level, they must have formed in the marine environment, probably as a result of mass wasting (McGregor et al. 1982). The threefold difference in the accumulation rate determined from two slope cores only 6 km apart may reflect the efficiency by which gully systems introduce silt and clay into the canyons. The decrease in sand-size material in some canyon axes on the mid-slope (Forde et al. 1981) may reflect either dilution by fine sediment from the gully system or sand bypassing in the midslope portions of the canyons. In the thalweg cores from the rise in the study area (Fig. 2), turbidite sequences range from a few centimetres to 42 cm in thickness, and sand contents range from 17% to 100%. In contrast, the turbidite sequences in the rise core 8044 taken on a relative topographic high are wispy, thin intervals of very fine sand showing only a trace of material identifiable as present-day outer shelf sand (Fig. 2). Sand-tracer data indicate that sources for this area are either on the slope itself or possibly on the shelf south of Wilmington Canyon from which samples were not available for this study. Even though the character of the sand and the sedimentation patterns differ between topographic highs and lows on the rise, they show a common trend in the general decrease in the number of layers and total sand content upsection. The presence of fewer valleys on the rise than on the slope implies that large areas of the slope drained by several canyons have a common channel on the rise where deposition and/or
et al.
erosion can occur. Flow events in the slope portion of the canyons may occur less frequently than they do in the rise portion of the canyons. Coincident flow events from different canyons may be reflected as a series of events on the rise. Doubling of the slope sediment accumulation rate in rise core 8044 may reflect this funnelling of sediments. On the rise, both deposition and erosion are taking place within the canyons. Graded sand layers in the cores indicate that the transport events are periodic. The 21~ data suggest that deposition is taking place at a rate of a few millimetres per year, and bioturbation of the surficial sediment (upper 130 cm) indicates lack of very recent transport. Observations of clean outcrops on the walls of the rise canyons from DSRV Alvin suggest that erosion is also taking place. Undercutting and slumping of the outside wall of the meanders in Wilmington Canyon serves to introduce sediment into the canyon. Therefore, not only are the shelf and upper slope sources of sediment, but a local source on the rise is also present. Erosion also takes place in canyons that do not meander. The ridges on the floor of South Wilmington Canyon suggest differential erosion. Sand and loosely bound gravel conglomerates in the canyon wall provide a local source for coarse-grained material deposited on the rise. Without & situ Alvin observations of the local source, down-canyon transport resulting in a turbidite with a coarse-grained basal unit might have been inferred. Depositional processes within the channel of Wilmington Canyon vary in a localized area on the rise. As noted above, more sand is differentially deposited on the outside than on the inside of the meanders, suggesting deposition from down-canyon flow. The location of the canyon thalweg along the outside of the meander, as well as steepness and outcropping horizons on the outside wall, suggests that erosion occurs in this portion of the channel. Deposition and erosion may be coincident with down-canyon density flows. The X-radiograph of the core seen in Fig. 7(d) shows that bioturbation was active in this area of the rise. Inspection of X-radiographs and sediments from the study area reveals that bioturbation is present in all rise cores but is greatest in sediments devoid of lamination or that contain thin, wispy, turbidite sequences. On the other hand, sedimentary sequences containing thicker, coarser, and more complete Bouma sequences had less apparent interturbidite bioturbation; perhaps because the sequences are thicker, the burrowers could only rework the upper finegrained parts. In the latter core types, however, as
The role of canyons in late Quaternary deposition the frequency and size of turbidites decreased upsection, a seeming concomitant increase in bioturbation was observed (Forde & Nelsen 1981).
Summary and discussion Physiography and microtopography are important in evaluating the sedimentary processes on the continental margin. Both spillover of shelf sands and headward erosion of canyons results in sand deposition in canyon heads. Grain-size distribution patterns allow several different sand populations to be identified. Although there is a slope sand population, fine-grained silts and clays are the dominant sediment type on the slope. An extensive gully system dissects the slope and provides pathways for the transport of large amounts of fine-grained slope sediments into the canyons. Mass wasting is the most likely process for transporting sediments within the gullies because they are located on the steep canyon wall. This large volume of fine-grained sediment dilutes the shelf and upper slope sands along the midslope, where a lack of graded sand layers is noted. Lineations on the floor of South and North Heyes Canyons parallel to the canyon axis suggest that down-canyon transport of sediment takes place. Differences in canyon morphology (meandering versus straight channel) imply that processes may vary between canyons. Radiometric dating indicates that deposition is taking place on the rise, whereas direct observations from DSRV Alvin show that erosion is also taking place. This erosion can provide local sources for sediments on the rise, in addition to sources on the shelf and slope from which sediment is transported seaward through the canyon systems. Piston cores on the rise contain graded beds having sand
329
populations similar to that of the adjacent shelf. On the rise north of Wilmington Canyon (Fig. 2), at approximately 2600 m depth, four valleys receive and channel sediment that has been eroded from a 55 km length of outer shelf and slope which is dissected by more than 22 canyons on the midslope. On the basis of sediment distribution, observations of the microtopography and surface sediment distribution, we believe that the canyons of the US mid-Atlantic continental slope are periodically the conduits for sediment transport to the rise. Because the large source region of the slope contains fine-grained materials, these are the dominant materials transported and deposited on the rise. Mixtures of shelf sands with fine-grained material serve as tracers that indicate downcanyon transport. ACKNOWLEDGMENTS: We wish to thank E.A. Shinn and J.M. Robb of the US Geological Survey for providing helpful comments and review of the manuscript. Special appreciation to C.A. Nittrouer (North Carolina State University) and D.J. Stanley (Smithsonian Institution) for radiometric dating analyses. Thanks to the officers and crews of the NOAA ship Researcher, R/V Gyre, and R/V Lulu and the pilots and support team of the DSRV Alvin. The submersible study was supported by NOAA's Undersea Research Program and Marine Geology and Geophysics Laboratory in Miami, Florida. The latter also supported the coring program. Funds for the sidescan survey of the slope south of Wilmington Canyon were provided by the US Bureau of Land Management (BLM) under Memorandum of Understanding AA 851-MUO-18 and Interagency Agreement AA 851 -IA 1-17 between BLM and USGS.
References DOYLE,L.J., PILKEY,O.H. & Woo, C.C. 1979. Sedimentation on the eastern United States continental slope. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleo. Min. Spec. Pub. 27, 119-29. EMERY, K.O. & UCHUPI, E. 1972. Western North Atlantic Ocean topography, rocks, structure, water, life, and sediments. Am. Ass. Petrol. Geol., 17, 532 PP. FORDE, E.B., STANLEY,D.J., SAWYER,W.B. & SLAGLE, K.J. 1981. Sediment transport in Washington and Norfolk submarine canyons. Applied Ocean Res., 3, 59-62. - & NELSEN, T. 1981. Variability of sedimentary textures and processes on continental margin north
of Wilmington Canyon. Am. Ass. Petrol. Geol. (abs.), 65/9, 1662. KELLER, G.H. & SHEPARD, F.P. 1978. Currents and sedimentary processes in submarine canyons off the northeast United States. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans, and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 15-31. , LAMBERT, D.N. & BENNETT, R.H. 1979. Geotechnical properties of continental slope deposits-Cape Hatteras to Hydrographer Canyon. In: Doyle, L.J. & Pilkey, O.H., (eds), Geology of Continental Slopes. Soc. econ. Paleo. Min. Spec. Pub. 27, 131-51. KELLING, G. & STANLEY,D.J. 1970. Morphology and
330
B.A. McGregor et al.
structure of Wilmington and Baltimore submarine canyons, eastern United States. J. Geol., 78, 637-60. MCGREGOR, B.A. & BENNETT, R.H. 1981. Sediment failure and sedimentary framework of the Wilmington geotechnical corridor, US Atlantic continental margin. Sed. Geol., 30, 213-34. - - - , STUBBLEFIELD,W.L., RYAN,W.B.F. & TWICHELL, D.C. 1982. Wilmington Submarine Canyon: A marine fluvial-like system. Geology, 10, 27-30. MERRILL, G.F. & BENNETT, R.H. 1981. Bathymetric map of geotechnical corridor on US Atlantic continental margin southeast of Cape May. Am. Ass. Petrol. Geol. Bull., 65/9, 1667. NELSEN, T.A. 1981. The Nature of General and Mass Mocement Sedimentao' Processes on the Outer She!/', Slope, and Upper Rise Northeast of Wilmington Canyon. Ph.D. Dissertation, University of Miami, Miami, Florida. 303 pp. -& BENNETT, R. 1981. The signatures of surficial sand and small scale mass wasting of sediments at the shelfbreak and upper slope seaward of the Baltimore Canyon Trough. Am. Ass. Petrol. Geol. (abs.), 65/9, 1667. POAG, C.W. 1979. Stratigraphy and depositional environments of Baltimore Canyon trough. Am. Ass. Petrol. Geol. Bull., 63, 1452-66. STANLEY, D.J. 1974. Pebbly mud transport in the head of Wilmington Canyon. Marine Geol., 16, M I-M8.
--,
SWIFT, D.J.P., SILVERBERG, N., JAMES, N.P. & SUTTON, R.G. 1973. Recent sand spill-over offSable Island Bank, Scotian Shelf. Geol. Surv. Can. Pap., 71-23, 167-94. STUBBLEFIELD, W.L., MCGREGOR, B.A., FORDE, E.B., LAMBERT,D.N. & MERRILL,G.F. 1982. Reconnaissance in DSRV ALVIN of a 'fluvial-like' meander system in Wilmington Canyon and slump features in South Wilmington Canyon. Geology, 10, 31-6. TELEK1, P.G., ROBERTS, D.G., CHAVEZ, P.S., SOMERS, M.L. & TWICHELL, D.C. 1981. Sonar survey of the US Atlantic continental slope: acoustic characteristics and image processing techniques. Offshore Technology Conference, Paper OTC 4017, 91-102. TWlCHELL & ROBERTS,D.G. 1982. Morphology distribution and development of submarine canyons on the United States Atlantic continental slope between Hudson and Baltimore Canyons. Geology, 10, 408-12. VAIL, P.R., MITCHUM, JR., R.M. & THOMPSON III, S. 1977. Seismic stratigraphy and global changes of sea level, Part 3: Relative changes of sea level from coastal onlap. Am. Ass. Petrol. Geol. Mere., 26, 63-81. VEATCH, A.C. & SMITH, P.A. 1939. Atlantic submarine valleys of the United States and the Congo Submarine Valley. Geol. Soe. Am. Spec. Pap., 7, 101 pp.
B.A. MCGREGOR, US Geological Survey, 915 National Center, Reston, Virginia 22092, USA. T.A. NELSEN, W.L. STUBBLEEIELDand G.F. MERRILL, National Oceanic and Atmospheric Administration, Atlantic Oceanographic and Meteorological Laboratories, 4301 Rickenbacker Causeway, Miami, Florida 33149, USA.
A late Miocene and early Pliocene upper slope-to-shelf sequence of calcareous fine sediment from the Pacific margin of New Zealand P.F. Ballance, M.R. Gregory, G.W. Gibson, G.C.H. Chaproniere, A.P. Kadar and T. Sameshima SUMMARY: Upper Miocene and lower Pliocene sediments at East Cape, North Island, New Zealand, comprise an upwards-coarsening and shallowing sequence. Calcareous sandy siltstones, with coccoliths and about 10% clay, were deposited in upper bathyal depths. Upwards they became sandier, with an increasing ratio of benthic to planktic foraminifera and increasing macrofauna, and grade into fossiliferous very fine silty sands, with many coccoliths and a few percent clay; the sands were deposited in shelf depths of 50 to 100 m. Comparison with modern equilibrium shelves suggests that the shelf sediments were deposited on a storm-dominated, wind- and wave-driven ocean margin characterized by long-period swells and normal wave heights of I to 3 m. Clay-size sediment seems to have been winnowed out of all the sediments and transferred to greater depths, apart from a few percent trapped between framework grains. The prolific trace fossil fauna contains several ichnotaxa common to all the sediments; however, Teichichnus, a Zoophycos-type and a compound Nereites type, along with minor forms and non-calcareous tube fossils (cf. Terebellina) are confined to the bathyal facies, while Asterosoma extends from the shelf facies part way into the bathyal facies. Coccoliths comprise a restricted flora of temperate-water aspect, lacking both deep water and tropical forms. Macrofauna of some 50 species in the bathyal sediments is dominated by thin-shelled Nuculid and Nuculanid bivalves, while some 80 or more species in the shelf sediments are dominated by the thick-shelled bivalve Cucullea. Debris-flow deposits in the bathyal facies include both bathyal and shelf sediments; they suggest a shelf-slope topography. Numerous interbeds of graded volcanic sand occur in all the facies studied. They were deposited by air-fall and turbidity currents, are mostly rhyolitic in composition, and were derived from the active New Zealand magmatic arc.
Marine sediments of latest Miocene to earliest Pliocene age outcrop over some 200 square kilometres between East Cape and the Awatere River, in the North Island of New Zealand (Fig. 1). They have been subjected only to slight folding and faulting, and shallow burial. The sequence can be regarded as consisting of calcareous background sediment whose accumulation was punctuated by the periodic arrival of volcanic tephras. The latter invariably show normal size grading, and some show elements of the Bouma sequence and contain fossils, indicating that some at least are turbidites. The majority are rhyolitic in composition (Chaproniere 1969). They are not considered further in this paper, except as a possible source for the sand fraction found in the background sediments. The sequence in general is thus a type of flysch (Fig. 2). This paper is concerned only with the sequence of background sediment, which coarsens upwards from sandy, calcareous muds of inferred upper slope origin to calcareous very fine sands of inferred shelf origin. Macro-, micro- and tracefossils are abundant throughout. Previous work on the rocks comprises early regional mapping by the New Zealand Geological
Survey (Ongley & MacPherson 1928), and recent 1:50 000 mapping (I. G. Speden & P. R. Moore, in prep.). Kennett (1966) described the foraminifera of one section, and Chaproniere (1969, 1973) made a regional study at the western end of the block. The rock specimens collected for this study are numbers 33401 to 33446 inclusive, in the University of Auckland rock collection.
Stratigraphy and facies Two formations are distinguished. The lower Pohutu Formation (Chaproniere 1969, unpublished) consist of strongly calcareous, massive, sandy mudstones, with tephra interbeds occurring every metre or two. Intraformational and extraformational debris-flow deposits occur commonly, and backwards-rotated, nested slump packets are exposed at one place south of East Cape. Thickness is unknown because of minor structural complexity, but is only a few hundred metres at most. The base is not exposed; it probably rests unconformably on early Miocene mudstones (Chaproniere 1969). This formation 331
332
P.F. Ballance et al. z
//~ /
Te Araroa
/ ~/I
. ~ . ROAD "~ J.:l." SECTION
//~O/ "'-.'~/ i 0 ~ -. 1 /
2 <
//
East Cape
j....
~-~// FIG. 1. Locality map showing the sections studied. Upper Miocene to Lower Pliocene sediments in the East Cape block are shaded.
COAST SECTION
....
0
corresponds to the upper slope facies of this paper. The upper Paeoneone Formation (Chaproniere 1969, unpublished) consists of calcareous, massive, muddy, very fine sand, with tephra interbeds every few metres. It is richly fossiliferous. Thickness is more than 300 m, and nowhere is the top preserved. The base is gradational from the underlying Pohutu Formation. This formation outcrops around Te Araroa. It comprises the shelf facies of this paper, while the
km
60
few metres of sediment gradational between the two formations is inferred to approximate to a shelf-slope break facies.
Lithological descriptions The lithological descriptions incorporate field, binocular microscope, and scanning electron microscope (SEM) observations and grain-size and carbonate percentage analyses; mineralogy from X-ray diffraction is discussed in a separate section. The descriptions are grouped into lower calcareous, sandy, clayey siltstone; intermediate calcareous, very sandy, clayey siltstone; and upper calcareous, muddy, very fine sandstone units, corresponding to the upper slope, shelfslope break, and shelf facies, respectively.
Lower calcareous, clayey, sandy siltstones
FIG. 2. Calcareous, slightly sandy and clayey siltstones of the upper slope facies at East Cape, with interbedded graded volcanic tephras. Hammer (circled) for scale.
In the field (Fig. 2) these sediments are very pale grey to white, massive, and moderately indurated. They do not fritter on exposed surfaces in the way that clay-rich mudstones normally do. Little bedding is apparent, and the prolific trace-fossil assemblage indicates that the sediment has been churned and homogenized. Large calcareous concretions occur. Macroscopic pieces of pumice occur in some beds. Binocular microscope examination reveals a large percentage of elastic grains, including microfossils, euhedral feldspar, quartz, glass, pumice, white and black mica, other mafic minerals, carbon and pyrites. SEM examination of broken surfaces (Fig. 3(a)-(f)) reveals sand and silt-size clasts coated with flakes of clay minerals and coccolith plates. Porosity appears to be high. Recognizable components include pumice grains and glass shards, foraminifera, rare radiolaria, spines and spicules, and pyrite framboids. Many cavities in pumice grains and fossils are unfilled. Rare euhedral
(a)
(b)
(c)
(d)
(e)
(f)
FIG. 3. SEM micrographs of the upper slope calcareous siltstone facies: (a) showing scattered framework grains with adhering coccolith plates and flakes of clay. Specimen 33401, Awatere Valley road section, near base; (b) with an unidentified spicule. 33422, East Cape; (c) with euhedral ?apatite crystals growing on the inner surface of a ?radiolarian test. 33435, NZ Grid Reference N63/813618, Wharariki Pt; (d) showing books of clay minerals, a spicule and a pyrite framboid. 33427, NZ Grid Reference N63/874605, Ahipaepae; (e) from a bed containing the 'heiroglyph' trace fossil Teichichnus. Note sand grains, a spicule, and grain of pumice upper left. 33432, NZ Grid. Reference N63/839609, Taunahahana; (f) from a mottled, mixed tephra and upper slope siltstone. Note the abundance of pumice, with unfilled vesciles. 33409, NZ Grid Reference N72/917495, Waikori Bluff;
334
P.F. Ballance et al.
FIG. 4. Triangular plot of CaCO3 against sand and mud for all facies. Sands are the shelf facies; muds are the upper slope facies. Note the uniform CaCO3 content. The overlap between sands and muds arises in part because of mixing of tephra into the muds by bioturbation, and in part from the gradational contact. crystals, probably apatite, of undoubted diagenetic origin occur (Fig. 3(c)). The mudstones typically contain between 10 and 25~ CaCO3 by weight (Fig. 4). The acid residues contain 13 to 25~ sand, 54 to 62% silt, and 10 to 14~ clay-sized material, by weight percentage of the entire sample (Fig. 5). They are thus texturally sand and clay-bearing siltstones.
Intermediate calcareous, very sandy, clayey siltstones In the field the boundaries of this intermediate lithology are entirely gradational. It is only a few metres thick. It is marked off from the underlying mudstones initially by increases in the proportion, but not size, of sand, and in the content of macrofossils and large benthic foraminifera. At
FIG. 5. Triangular plot of sand, silt and clay contents for all facies, with CaCO3 removed. Sands are the shelf facies; muds are the upper slope facies. Note the uniform low clay content, and the gradation from mud to sand.
SAND
MUD
higher levels the grain-size also increases to a maximum of medium sand, and large Cucullea come in to complete the transition to the upper sandy unit. In binocular microscope and SEM examination (Fig. 6(a)) there is little to distinguish the intermediate lithologies from the underlying unit. All the components are similar in kind. Coccoliths are abundant. CaCO3 content ranges from 10 to 16%, and clay-size content from 9 to 12~, while sand and silt-sized grains are subequal in amount. XRD analysis (Table 1) suggests clay mineral contents of the order of 30%.
Upper calcareous, muddy, very fine sandstones In the field these are thick-bedded to massive sands, calcareous and slightly muddy, but other-
SIX
CLAY
Slope-to-shelf sequence of calcareous fine sediment from New Zealand
(a)
(c)
335
(b)
(d)
Fro. 6. (A) SEM micrograph of the intermediate, shelf-slope transition, very sandy mudstone. Note the prominent glass shard and central pumice grain, and spicule. 33406, Awatere Valley road section. (B) SEM micrograph of muddy, very fine sandstone shelf facies with Asterosoma, showing very fine sand grains including cleavage-bounded? feldspar at lower left, central pyrite framboid and elongate shell fragment. (c) as (B), showing abundant coccoliths. (b) and (c), 33444, Te Hekawa, NZ Grid Reference N63/792618. (O) as (b), with Cucullea, showing a relatively well-sorted phase with a tri-radiate spicule and a large pumice grain. 33445, Te Araroa, NZ Grid Reference N63/783619. wise well-sorted and very fine grained. They are lightly indurated, extensively bioturbated, lacking in primary sedimentary structures, and richly fossiliferous. The most conspicuous and abundant fossil is Cucullea (Fig. 7). Most fossils are clearly not in life position. In binocular microscope examination, grainsize extends up to the lower part of medium sand. Sorting appears poor, and recognizable components include pumice, foraminifera, spines and spicules, and macrofossil fragments. SEM examination (Fig. 6(b)-(d)) shows fairly even-sized very fine sand grains, pumice grains, fossils, high porosity and interstitial clay flakes and numerous
coccoliths. From XRD analysis (Table 1) clay mineral content is perhaps as high as 30%. CaCO3 content ranges from 9 to 2 2 ~ clay-size material from 5 to 9%, silt from 21 to 25%, and sand from 46 to 65% (Figs 4 and 5). The sand mode is very fine sand.
Mineralogy by XRD Nine samples were analysed by conventional techniques. Quantities of the constituent minerals were roughly estimated by comparing the corresponding peak areas with those of a standard
P.F. Ballance et al.
336 TABLE 1.
XRD Estimated Mineral Composition of Selected Sediments.
Sample No.
01
02
06
07
27
32
35
44
45
Quartz Plagioclase Potash-feldspar Calcite Dolomite Chlorite Illite Kaolin Montmorillonite Interstratified Calcite (104) Dolomite(104)
D D
D D
D D
D D
D D
D D
D D
D D
G
G
G
F
G H F F
G F F F
G G G F
E E G F F F F G G
D= E= F = G: H=
3.035
26-35% 16-25% 11-15% 6-10% 2-5%
3.032
3.033
F F G G H 3.036
3.038 2.899
G F F G F 3.035
G H F G G G H 3.030 2.889
G H F G G G 3.037 2.899
G H G F H F H 3.033 2.898
Notes: Sample numbers are 33401 and upwards. Calcite (104) and dolomite (104) spacings in A. 'Interstratified' is Chlorite-smectite interstratified mineral. 'Illite' includes glauconite.
mineral mixture with known relative abundances similar to the rock samples. Thin sections were cut from two rock samples in order to clarify the mode of occurrence of dolomite which was detected by XRD.
Results of the X-ray Analyses The samples represent a stratigraphic and geographic coverage of the transition from slope sediments to shelf. Most contain about 30% each of quartz and plagioclase, and 6-15% each of
calcite, chlorite, illite (plus glauconite), kaolin and montmorillonite (Table 1). To clarify whether volcanic glass in the rock has been altered, five pumiceous rock samples from the tephra interbeds were examined by non-oriented mount X R D record, and the results are shown in Table 2. No zeolite was detected, but three samples were found to contain large amounts of the clay minerals halloysite and montmorillonite. The nine rock samples have a consistent mineral composition (Table 1), except for 33427,
FIG. 7. Bedding plane view of an upper calcareous, muddy, very fine sandstone bed. Note the large Cucullea shells, the smaller shells, the accumulation of shell fragments by the hammer head, and the large trace fossil burrows which is referred to the ichnogenus cf. Bergaueria, a possible fish-feeding excavation. Te Araroa.
Slope-to-shelf sequence o f calcareous fine sediment.from N e w Zealand
337
TABLE 2. XRD Estimated Mineral Composition of Pumiceous Sediments. Sample No.
08
14
16
33
40
Quartz Plagioclase Calcite Chlorite Illite Kaolin Halloysite Montmorillonite Halite (NaC1)
E D G F G G
G G H
G F H
F A
E D H
E
F
E
H
C G
D G
G G
A = 56-65% B =46-55%
E = 16-25% F =11-15%
c=36-45%
G= 6-10%
D = 26-35%
H= 2-5%
H H
H
Notes: 'Illite' includes glauconite. Sample numbers are 33408 and upwards.
which contains a considerable amount of dolomite. The dolomite is slightly calcic, approximately 53 Mol% CaCO3, which is close to the mean composition of the diagenetic dolomite of Miocene marine formations of California described by Murata et al. (1972).
Fauna and flora Microfauna
Binocular microscope and SEM examination of the sediments reveals abundant foraminifera, rods, spicules, radiolaria and diatoms. Only the foraminifera have been studied to date. Kennett (1966) reported foraminifera indicative of water depths between 60 and 300 m in the upper very fine sand facies at Te Araroa. In the road section (Chaproniere 1969, 1973; and this study) the percentage ofplanktic forms falls from 60 to 70 at the base of the sequence, to around 20 at the transition from sandy muds to sands, to about 14 in the middle of the upper very fine sand unit. Above that, however, it rises again to about 45% at the top. The percentage of arenaceous forms varies inversely, rising to 6 at the transition, and 7 in the middle of the sand unit, but falling to 2 at the top. A shallowing trend from outer neriticupper bathyal to neritic is indicated, using Boersma's (1978) criteria. The coastal section shows high planktic percentages of 40 and 60 at East Cape (Auckland University foram samples 1213, 1215), but all other samples from the calcareous, sandy, clayey siltstone (1214, 1216-18, 1224, 1225, 1227, 1229) show planktic ratios ranging from 15% to zero. This pattern of planktic abundance is paralleled by a limited number of benthic deep-water indi-
cators, Globocassidulina subglobosa, Hoeglundina
elegans, Pullenia bulloides, Karreriella cylindrica, Cibicides molestus, and Sigmoilopsis schlumbergeri (Vella 1962); several of these species occur in the samples with the highest planktic percentage, and only odd specimens occur in samples with low planktic values. Such low values of planktic abundance are normally interpreted to mean neritic to inner neritic environments (Boersma 1978). However, such an interpretation in these beds would be at variance with the evidence of trace fossils, macrofossils and lithology, and the low planktic abundances represent an anomaly. Macrofauna
Macrofossils are locally abundant throughout the sequence, but have not been studied in detail. In the calcareous, sandy, clayey siltstones, the commonest macrofossils are small thin-shelled Nuculid and Nuculanid bivalves, with valves conjoined in some cases. The fauna of 45 to 50 species comprises mostly mollusca, but includes large solitary corals (Flabellum) and burrowing heart urchins with spines still attached. Fewer than 10 species are common. Upwards in the sequence the fauna becomes more diverse, but this is in part a reflection of displaced forms brought in by volcanic gravity-flows. In the calcareous, muddy, very fine sand facies, a fauna of around 80 species includes many thick-shelled forms, the most conspicuous of which is Cucullea. The latter occurs mostly as single valves, and rarely with both valves together and closed. The valves are commonly bored, randomly orientated, and are rarely in life position.
338
P.F. Ballance et al.
Microflora
Trace fossils
The calcareous nannoplankton was studied on a reconnaissance basis by Kadar, using light microscopic techniques. Nine samples yielded 30 taxa, of which only two Coccolithus pelagicus and Reticulofenestra pseudoumbilica are common in all samples. Three factors in the assemblage indicate a temperate water mass, the low diversity, the low Discoaster to Coccolith ratio, and the absence of the warm-to-tropical living genus Ceratolithus (Norris 1965). Four species in the assemblage, Braarudosphaera bigelowi, Discolith-
The entire sequence is bioturbated. In places the bioturbation has been so thorough that bedding is obliterated, while mud from the background sediment and sand from the tephra interbeds are mingled in a mottled texture. Generally, however, the tephra interbeds retain their identity and essential characteristics, even though they are penetrated by individual burrows. Similarly, the sandy mud generally contains discrete, individually recognizable burrows. However, the upper sand facies is more thoroughly churned, and individual burrows are less easily recognizable. Even here, however, tephra interbeds retain their identity. In exposures away from the intertidal rock platform, bioturbation is evident, but details are not visible. On the intertidal platform, however, the trace fossils are superbly displayed. In addition to the ichnotaxa shown on Fig. 8, there are several less common forms that are not yet described. Figure 8 shows an obvious change in ichno-
ina japonica, D. multipora, Scyphosphaera sp. seldom occur in deep oceanic sediment (Bukry 1973), suggesting that the sediments are not of deep bathyal origin. The calcareous nannoplankton do not show a clear distinction corresponding to the various facies described in this paper. Coccoliths were observed in almost all the samples studied under the SEM, and are abundant in the sand facies (Fig. 6(c)). Thus all the sediments in the section were open to free access by oceanic waters.
LOCALITY
Te Hekawa Te Araroa
Wharariki Pt. Maruhau Pt.
Te Ahipari
Taunahahana
East Cape
Horoera Pt.
Sediment-type
ICHNOFOSSILS
SAND
MUDDY SAND
SANDY MUD
Asterosoma c f.Bergaueria Chondrites Scolicia Thalassinoides/ Ophiomorpha Planolites Teichichnus Zoophycus -type compound Nereites-type Rorschachichnus (nov.) compound bowtie structure non-calcareous tubes (cf. Terebellina)
FIG. 8. Distribution of the most abundant trace fossils in the coastal section. Localities shown all lie between Te Araroa and East Cape (Fig. 1). Sand=calcareous, muddy, very fine sand of the shelf facies; muddy sand = calcareous, very muddy, very fine sand of the shelf-slope break facies, sand mud = calcareous, sandy, clayey siltstone of the upper slope facies.
Slope-to-shelf sequence o f calcareous fine sediment f r o m New Zealand fauna from the sandy mud to the sand facies. The forms confined to the calcareous, sandy, clayey siltstones include two Zoophycos-type and a Nereites-type which are ubiquitous in slope and flysch facies (Crimes 1977). In addition, an undescribed ink-blot or web-like burrow system, which occurs rarely in the sandy mud, is widespread in northern New Zealand in flysch and flysch-like deposits of Miocene age and younger. Similarly, the small, white, non-calcareous tube fossil cf. Terebellina, which is common in places in the sandy mud, is widespread in flysch sequences of Cretaceous age and younger both in New Zealand and California (personal observations of the authors). Teichichnus may allow a more specific designation of environment, being listed by Crimes (1977) among a group of sediment eaters characteristic of offshore environments, but not canyon, fan or trough. Thus, the ichnofauna and tube fossils of the calcareous, sandy, clayey siltstones suggest an off-shelf, upper continental slope environment of deposition. No trace fossil is confined to the calcareous, muddy, very-fine sand facies. However, the most abundant form, the ostrich foot-like Asterosoma, a probable crustacean structure extends only part way down through the sequence into the upper levels of the sandy mud, and cf. Bergaueria, possible fish-feeding excavations, have been seen only at two localities, one in the sands and one in coarse interbeds in the sandy muds (Fig. 7). The trace fossils that occur throughout the sequence are forms which, as presently distinguished, have a very wide facies distribution. There is an indication at the locality Te Ahipari (Fig. 8) that substrate type may also exert a control on the distribution of trace fossils. Here the incoming of Asterosoma and cf. Bergaueria coincides with the incoming of thick, coarsegrained volcanic-rich sediment gravity-flow deposits.
Debris-flow deposits Several horizons of debris-flow deposit occur in the slope facies. They range in thickness from one to several metres, and extend laterally for some hundreds of metres. Some of them are intra-formational and consist of masses of sediment identical to the enclosing sediment. The masses reach a metre or more in length, and are commonly folded and enclosed in a matrix of identical sediment showing a streaky and swirled texture. At least two extra-formational debris-flows contain masses of sand which are identical both in lithology and in fossil and trace fossil content with the overlying shelf facies. Thus the debris-
339
flows establish that there was a sufficient palaeoslope during accumulation of the slope facies for a limited thickness of cohesive but plastic mud to slide. They also establish that a shallower facies of identical character to the overlying shelf facies existed contemporaneously with, and subsequently prograded over, the deeper facies. Associated in situ slump features are restricted to one occurrence of stacked or nested rotational slump packets, south of East Cape.
Palaeocurrents All palaeocurrent data is from the volcanic interbeds in the slope facies. Small-scale crosslamination, flame structures and aligned cf. Terebellina tubes, in several beds low in the sequence, indicate current flow to an easterly quadrant (Fig. 9). There is no palaeocurrent data from the background sediment, but alignment of long axes of clasts in debris-flow (Fig. 9) indicates a northstriking palaeoslope. A palaeoslope down to a trench lying to the east is what would be expected from regional considerations (Ballance et al. 1982). Walcott & Christoffel (1981) reported that Miocene sediments in the vicinity of East Cape show no apparent rotation of palaeomagnetic declination.
Interpretation of facies On the basis of the changes in grain-size, microfauna, macrofauna, trace fossils, and the asso-
(o)
(b)
t N
(d)~ FIG. 9. Rose diagrams of palaeocurrent data from the volcanic interbeds (a) to (c) and from an olistostrome (d). (a) small scale cross lamination. (b) strikes of flamestructures, which are mostly overturned towards the SE. (c) alignment of cf. Terebellinatubes. (d) alignment of long axes of clasts in an olistostrome at East Cape. All data are from the upper slope facies.
340
P.F. Ballance et al.
ciated debris-flow and rotational slumps, the coarsening-upwards background sediment in the flysch-like sequence described here is interpreted as encompassing three facies: upper slope, shelfslope break, shelf.
Discussion Notable features of all the sediments are the small amount of clay-size material (5-14%), and the constant amount of carbonate (10-25%). Much of the latter consists of pelagic organic tests introduced from Pacific oceanic waters, which had free access to both slope and shelf. The paucity of clay-size sediment implies that it was being winnowed through the system, and transferred to deeper water. Many authors have considered the mechanisms of transfer of fine sediment across the continental shelf (e.g. McCave 1972; Komar et al. 1974). Many factors are involved, including wave action, tidal currents, bottom-hugging turbid flows, storms etc. Table 3 represents an attempt to compare the shelf facies of this study with modern shelves. The only criteria available for direct comparison are grain-size and inferred water depth. Few modern shelves appear to have reached equilibrium with post-glacial processes (Howard 1978); those which have, and which are comparable with the East Cape sediments in showing a systematic offshore fining, occur on storm-dominated, windand wave-driven coasts (Howard 1978). Table 3 lists a number of such modern shelves; in the case of Roussillon, on the French Mediterranean coast (Jago & Barusseau 1981), the sediment is finer than the maximum that the wave regime could theoretically drive, presumably because coarser sediment is not available, thus if the sediment at Roussillon were used to calculate
wave regime and depth, the results would indicate a greater water depth than actually exists. The closest comparison from Table 3 is with shelves adjacent to major oceans, suggesting that the East Cape shelf was open to high, long-period waves, which regularly stirred the bottom sediment and moved the finer sediment progressively offshore. This is in agreement with palaeogeographic reconstructions which place the East Cape area in a similar position, with respect to the offshore trench and Pacific Ocean, to the present east coast of New Zealand (Ballance et al. 1982). A more detailed comparison between the East Cape shelf facies and modern shelves, in terms of sediment transport regimes, is not possible with present data. When we move from the shelf to the upper slope sediments, via the transitional sediment, the replacement of a very fine sand mode by a silt mode suggests a continuum in the sediment transport processes. It suggests that the latter were competent to transport silt and clay across the outer shelf, leaving a residue of very fine sand, but were only competent to transport clay-size material on the slope, leaving a residue of sandy silt. It is of interest to consider possible mechanisms of winnowing and transport on the continental slope. Huthnance (1981) reviews waves and currents near the shelf-edge, and indicates that thermohaline currents, high frequency internal waves, and internal tides are all possible mechanisms. Southard & Cacchione (1972) showed experimentally that breaking internal waves are capable of moving bottom sediment, and that they move it preferentially downslope. Dickson & McCave (1982) show that a bottom nepheloid-layer on Porcupine Bank, west of Eire, is much thicker at a depth of 450 m than it is above and below; they also demonstrate that the thickened bottom layer extends outwards
TABLE 3. Comparison o f water depth f o r veryfine sand with wave height andperiod for several
wind- and wave-driven shelves.
Roussillon, French Mediterranean coast (Jago & Barusseau 1981) Adriatic shelf, Italy (Passega et al. 1967) Bristol Bay, Bering Sea (Sharma et al. 1972) Washington shelf, NE Pacific (Larsen et al. 1981) South Australian shelf (Larsen et al. 1981) East Cape sediments
Water depth
Wave height
Wave period
10--40 m
0-1 m
4 sec.
10-30 m
--
--
60-100 m
--
--
< 90 m
1-3 m
13-17 sec.
> 75 m 50-100 m
1-2 m
12-17 sec.
Slope-to-shelf sequence of calcareousfine sediment from New Zealand into deeper water as an intermediate nepheloidlayer, and conclude that it originates by sediment erosion as a result of mixing at the boundary of differing water masses, possibly by internal tides enhanced by impingement on a characteristic bottom slope. Bottom slope is critical in determining whether internal waves and tides are amplified or attenuated (Huthnance 1981). Thus, there are several possible mechanisms which can winnow sediment delivered to the continental slope by spillover from the shelf. This may explain why some slopes are blanketed by silts, e.g. East Cape, eastern USA, north of Cape Hatteras (Doyle et al. 1979), eastern North Island of New Zealand (Lewis & Kohn 1973), while others, which are covered by more clay-rich mud, e.g. Plio-Pleistocene slopes in northern California (Piper et al. 1976), may lack either the mechanisms or the critical bottom slope. In the case of the East Cape slope, the frequent rhyolitic turbidity currents and ash showers which blanketed the slope do not seem to have been
341
reworked except by burrowing. This would suggest that any winnowing processes acted only on a thin surface layer of sediment. The few percent of clay and pelagic carbonate preserved in both the shelf and slope sediments were presumably trapped between framework grains or deposited in the form of faecal pellets. ACKNOWLEDGMENTS: Field support from the University of Auckland Research Committee is gratefully acknowledged. The United States Geological Survey kindly provided institutional support for Ballance during the preparation of the paper. SEM micrographs were taken on the USGS machine at Menlo Park by Robert Oscarson. David Piper and Tony Hayward reviewed the manuscript. Carbonate and grain-size analyses were done at Auckland University by Keith Johnstone. The figures were drafted by Queenie Ballance, and the manuscript was typed by Lillian Wood and William Ballance.
References BALLANCE,P.F., PETTINGA,J.R. & WEBB, C. 1982. A model of the Upper Cenozoic evolution of northern New Zealand and adjacent areas of the Southwest Pacific. Tectonophysics, 87, 37-48. BOERSMA, A. 1978. Foraminifera. In: Haq, B.U. & Boersma, A. (eds), Introduction to Marine Micropalaeontology. Elsevier, New York. 376 pp. BUKRY, D. 1973. Low latitude coccolith biostratigraphic zonation. In: Edgar, N.T., Saunders, J.B. et al. (eds), Init. Repts. DSDP, US Govt. Print. Off., XV. Washington, DC. 685-703. CHAPRONIERE, G.C.H. 1969. Geology of the Te Araroa Area, East Cape. Unpub. thesis, University of Auckland, New Zealand. 1973. On the origin of Globorotulia miotumida conomiozea Kennet, 1966. Micropalaeontology, 19, 461-8. CR1MES,T.P. 1977. Trace fossils of an Eocene deep-sea sand fan, northern Spain. In: Crimes, T.P. & Harper, J.C. (eds), Trace Fossils 2. Seel House Press, Liverpool. DICKSON, R.R. & MCCAVE, I.N. 1982. Properties of nepheloid layers on the upper slope west of Porcupine Bank. Report to the Hydrography Committee of the International Council for the Exploration of the Sea. 9 pp. & 10 figs. DOYLE,L.J., PILKEY,O.H. & WOO,C.C. 1979. Sedimentation on the eastern United States continental slope. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Palaeo. Min., Spec. Pub., 27, 119-29. HOWARD, H.D. 1978. Shallow siliciclastic seas. In: Reading H.G. (ed.), Sedimentary Environments and Facies. Blackwell Scientific Publications, Oxford; Elsevier, New York. 208-58.
HUTHNANCE, J.M. 1981. Waves and currents near the continental shelf edge. Prog. Oceanog., 10, 193-226. JAGO, C.F. • BARUSSEAU,J.P. 1981. Sediment entrainment on a wave-graded shelf, Roussillon, France. Marine Geol., 42, 279-99. KENNETT, J.P. 1966. Four Upper Miocene to Lower Pliocene sections, Hawke's Bay to East Cape, New Zealand. Trans. Roy. Soc. N.Z. Geology, 4(10), 189-209. KOMAR, P.D., KULM, L.D. & HARLETT, J.C. 1974. Observations and analysis of bottom turbid layers on the Oregon continental shelf. J. Geol., 82, 104-11. LARSEN, L.H., STERNBERG,R.W., SHl, N.C., MARSDEN, M.A.H. & THOMAS,L. 1981. Field investigations of the threshold of grain motion by ocean waves and currents. Marine Geol., 42, 105-32. LEWIS, K.B. & KOHN, B.P. 1973. Ashes, turbidites and rates of sedimentation on the continental slope off Hawkes Bay. N.Z.J. Geol. Geophys., 16, 439-54. LIPPMANN, F. 1973. Sedimentary Carbonate Minerals'. Springer-Verlag, New York. 228 pp. MCCAVE, I.N. 1972. Transport and escape of finegrained sediment from shelf areas. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport." Process and Pattern. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 225-48. MURATA, K.J., FRIEDMAN, I. & CREMER, M. 1972. Geochemistry of diagenetic dolomites in Miocene marine formations of California and Oregon. US geol. Surv. Prof. Pap. 724-C, 12 pp. NORRIS, R.E. 1965. Living cells of Ceratolithus cristatus (Coccolithophorinae). Arch. Protistenk., 108, 19-24. ONGLEY, M. & MACPHERSON,E.O. 1928. The geology
342
P.F. Ballance et al.
of the Waiapu Subdivision, Raukumara Division. N.Z. Geol. Surv. Bull., 30. PASSEGA, R., RIZZINI, A. 8L BORGHETTI, G. 1967. Transport of sediments by waves, Adriatic coastal shelf, Italy. Am. Ass. Petrol. Geol. Bull., 51, 1304-19. PIPER, D.J.W., NORr~ARK, W.R. & INGLE, J.C. 1976. The Rio Dell Formation: a Plio-Pleistocene basin slope deposit in Northern California. Sedimentology, 23, 309-28. SHARMA, G.D., NAIDU, A.S. & HOOD, D.W. 1972. Bristol Bay: a model contemporary graded shelf. Am. Ass. Petrol. Geol. Bull., 56, 2000-12. SOUTHARD,J.B. & CACCHIONE,D.A. 1972. Experiments
on bottom sediment movement by breaking internal waves. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds). Shelf Sediment Transport: Process and Pattern. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 83-97. VELLA, P. 1962. Biostratigraphy and palaeoecology of Mauriceville District, New Zealand. Trans. Roy. Soc. N.Z. Geology, 1(12), 183-99. WALCOTT, R.I. & CHRISTOFFEL,C. 1981. Palaeomagnetic measurements in Miocene sediments of eastern New Zealand. Abstract, Ann. Conf. Geol. Soc. o/ New Zealand, Hamilton, N.Z.
P.F. BALLANCE, M.R. GREGORY, G.W. GIBSON, G.C. CHAPRONIEREl, A.P. KADAR, T. SAMESHIMA,Geology Department, Auckland University, Auckland, New Zealand. 1 Present address: Bureau of Mineral Resources, P.O. Box 378, Canberra City, ACT 2601, Australia.
Facies, facies-associations and sediment transport/deposition processes in a late Precambrian upper basin-slope/pro-delta,, Finnmark, N. Norway K.T. Pickering SUMMARY: The lower Naeringselva Member of the 2.5 to 3.5 km thick Bfisnaering Formation varies from 60 to 540 m thick, and is part of the 9 km thick late Precambrian Barents Sea Group cropping out on Varanger Peninsula, north-east Finnmark, N. Norway. It is underlain by a minimum 3.2 km of submarine fan deposits and passes up into delta front and fluvio-deltaic deposits of the Bhsnaering Formation. Within the lower Naeringselva Member (LNM) four main facies are recognized: parallel-laminated mudstone; current-rippled siltstone; structureless mudstone, and wave-rippled fine-grained sandstone. Soft sediment deformation was abundant as slide, slump and in situ disturbed layers up to tens of metres thick. Erosional-depositional discordances are evident between packets of beds. Palaeocurrent data, together with a limited number of slide fold readings, suggest a west to south-westerly source area for the submarine fan, upper basin-slope/pro-delta and delta deposits. The stratigraphic position and sedimentology of the LNM suggests that it forms part of an ancient upper basin-slope/pro-delta, with fan deposits on the lower basin-slope. Apart from sporadic storm events visibly affecting the slope, it was below wave-base. Modern analogues for the LNM are the upper part of continental slopes separating fan from delta deposits, as in the Mississippi Delta-Slope-Fan system.
Ancient submarine basin slopes, separating shelf from deep water basins, have been described from many parts of the world and throughout the geological column (Doyle & Pilkey 1979). Possible ancient continental slopes mainly comprise relatively thinner bedded, finer grained sediments compared to underlying submarine fan and other continental rise deposits, and overlying shelf deposits (e.g. Piper et al. 1976; Tillman et al. 1981). Many modern continental slopes have been shown to consist of silts and sandy silts, with small-scale irregularities and gullies incised into the slope sediments (MacIlvaine & Ross 1979). Also, modern continental slopes are characterized by abundant sediment slides, ranging from centimetres to tens of metres in thickness (Moore 1961; Lewis 1971; Almagor & Garfunkel 1979; Knebel & Carson 1979), and such soft-sediment deformation appears to have contributed to substantial amounts of the postulated ancient continental slope successions (Doyle & Pilkey 1979). Recognizing ancient slide deposits and their overall dimensions may be difficult because of limited outcrop (see Woodcock 1979). Taking cores from modern continental slopes involves similar problems with interpretation, and Kelts & Arthur (1981) suggest the following features for the confirmation of slides in DSDP cores: sheared, stretched, or highly microfaulted beds, folded or overturned layers, and frequent rever-
sals in dip. These criteria are appropriate for interpreting sediment slides in ancient successions where outcrop is limited. Continental slopes and pro-delta slopes often occur as discrete, separated, depositional environments, especially where there is a wide continental shelf. However, there are cases where the delta-slope-submarine fan system forms a continuous basin slope, incised by deep marine channels, canyons and gullies (e.g. the Mississippi Delta-Slope-Fan, Moore et al. 1978; the Niger Delta-Slope-Fan, Allen 1964, 1965; Oomkens 1974). Coarse-grained sediments generally bypass the upper continental slopes and prodeltas and they are funnelled onto the continental rise and basin plains to form various slope aprons such as submarine fans. The distribution of continental slope facies associated with delta-fan systems is controlled by a variety of processes, including sea-level changes (Moore et al. 1978; Sidner et al. 1978), delta lobe switching (Scruton 1960; Dailly 1976), tectonism (Sangree et al. 1978), subsurface intrusions such as salt-dome development (Amery 1978) and the nature of the sediments that are fed into the deep water environment (Sheridan 1981). The purpose of this paper is to describe the salient features of an ancient upper continental slope and pro-delta, separating submarine fan from fluvio-deitaic deposits. 343
K.T. Pickering
344
Study area The late Precambrian Barents Sea Group (Siedlecka & Siedlecki 1967; Johnson et al. 1978), cropping out on Varanger Peninsula, NE Finnmark, N. Norway, is at least 9 km thick and comprises the following formations from the base upwards: (1) the Kongstjord Formation, submarine fan deposits at least 3.2 km thick (Siedlecka 1972; Pickering 1981a); (2) the B~snaering Formation, fluvially-dominated deltaics between 2.5-3.5 km thick (Siedlecka & Edwards 1980); (3) the B~ttstjord Formation, shallow marine intertidal and supratidal carbonates and siliciclastics about 1.5 km thick (Siedlecka, pars. comm.), and (4) the Tyvofjell Formation, shallow marine and fluviatile siliciclastics about 1.5 km thick (Johnson et al. 1978). Fossil acritarchs from the Bhsnaering and B~ttsfjord Formations give ages of Upper Riphean and Lower Vendian, respectively (Vidal 1981). The Barents Sea Group only occurs north of the Trollfjord-Komagelv Fault, interpreted as a palaeotransform fault with 500-1000 km of dextral displacement (Kjode et al. 1978). The group has undergone one major syn-metamorphic deformation (Roberts 1972), and Rb-Sr isotopic I
FORMATION
r
t
Environments
1.5
Shallow
o BATSFJORD
1.5
Shallow marine intertidal supratidal carbonates
[
m cn 3; < zv- Q. (..) II 0 "'1 ill It"t"
IIIlllllllllllllllllllllllllll
,,,t- m < O r,o_
II
o BASNAERING
~
Barley~
Fluvio-deltaic 3.5
B
marine/fluvial
sandstones and
2.5--
KONGSFJORD
i
Sedimentary
TYVOFJELL
z _
I
km
age determinations suggest that this event occurred about 520 Ma (Taylor & Pickering 1981). Metamorphism did not exceed the lowest greenschist facies (Roberts 1972) and, generally, primary sedimentary structures are well preserved in essentially continuous coastal exposures. The lower Naeringselva Member (LNM) forms the oldest part of the B~,snaering Formation (Fig. 1) and occurs immediately above the Kongsfjord Formation Submarine Fan. The LNM is mainly siltstones and mudstones, whereas the upper Naeringselva Member is predominantly medium (10--30 cm thick) to thick-bedded (30-100 cm), fine to very coarse-grained sandstones (Fig. 2). The upper Naeringselva Member is interpreted as delta front (Siedlecka & Edwards 1980), and the LNM as pro-delta deposits (Siedlecka & Edwards 1980; Pickering 1982a). The LNM varies between 60-540 m thick from west to east on Varanger Peninsula, respectively (Fig. 3). The transition from submarine fan deposits to the LNM is gradual and the boundary is arbitrarily fixed at the top of a transitional zone varying between 20 m, on the SE side of Kongsfjord, to 160 m thick at Blodskytodden, in which thick to medium-bedded turbidites are intercalated with
>3.2 A
R
E
Submarine N
sandstones
I
tJttitl]ltlJlTtlltll]lt]ltYltlllllltrrztlttt
T
S
fan
~
sandstones S
E
A
.o e ,C'.
dskytc, dde n
Okset Storflogdalen
Z. .,NGSELV.
MEMBE.
?
'k~
FIG. 1. Map showing the outcrop of the Naeringselva Member of the B~tsnaering Formation and its stratigraphic location within the Barents Sea Group (vertical line legend). The sections referred to in the text are shown by black arrows that also indicate the younging directions. TKF is the Trollfjord-Komagelv Fault, and 't' is the thickness of the formation in kilometres. Map based on Siedlecka & Edwards (1980).
Facies, facies-associations and sediment transport~deposition processes
FIG. 2. Typical outcrop of upper Naeringselva Member coarse-grained sandstones just above the top of the LNM., SE side of Kongsfjord. Hammer shaft is 35 cm long.
345
the LNM facies (Figs 3 & 4). Immediately below the transitional zone, there are sandstone-rich sections of very thick-bedded (100+ cm) to thin-bedded (3-10 cm) turbidites showing obvious channel morphology similar to middle fan deposits recorded from much lower down in the Kongsl]ord Formation (Pickering 1981a, 1982b). However, some of the 'channel deposits' show unusually steep erosional bases (Fig. 5) and most of the turbidites, regardless of bed thickness, tend to be lenticular, suggesting that bed shape in the transitional zone from fan to non-fan basin slope may have been controlled by the increased gradients associated with the upper basin-slope rather than the location of beds within the submarine fan (Pickering 1981b, 1983). The LNM thickens and coarsens upwards into medium and thick-bedded sandstones of the upper Naeringselva Member, in a transitional zone approximately 20 m thick, defining the upper portion of this lower member. The Naeringselva Member is abruptly overlain by medium to coarse-grained, heavy-mineral laminated, cross-bedded, reddish-coloured sandstones of the Seglodden Member (Fig. 6), interpreted as delta top deposits by Siedlecka & Edwards (1980). In contrast to the greyish colour of the underlying KongsOord Formation, and the reddish colour of the Seglodden Member of the Bfisnaering Formation above, the LNM typically is composed of greenish-grey (chlorite-rich), very thinbedded (1-3 cm) siltstones and mudstones. The
FIG. 3. Stratigraphic sections to show the thickness changes in the LNM, and the immediately overlying and underlying parts of the Barents Sea Group. Shading indicates different stratigraphic elements.
346
K. T. Picketing
FIG. 4. Turbidity current deposits in the transitional zone between submarine fan and upper basin-slope deposits from Storflogdalen (see Fig. 1). Scale is 15 cm long. Note the lenticularity of these fine-grained Tbcde beds (e.g. arrowed).
Naeringselva Member consists of immature to submature subarkoses, containing 5-15% feldspar, with clay matrices mainly ofchlorite-sericite mixtures (Siedlecka & Edwards 1980). These rocks are highly cleaved, and because of their incompetence compared to the surrounding litho-" logics, the exposures of the L N M are easily eroded and seldom show beds with lateral continuities greater than about 30 m.
LNM Facies
FIG. 5. Typical outcrop of the Kongsfjord Formation submarine fan deposits immediately below the LNM on the SE side of Kongsfjord. Height of cliff is approximately 25 m. Note the lenticularity of the beds and the dramatic erosional base to the sandstone that is arrowed. Quartz-veined rock in foreground is a dolerite dyke.
The LNM was measured as ~bundles' or packets of beds, either as discrete multi-facies (faciesassociation) packets, or as discrete single facies packets. From the vertical arrangement of these facies and facies-associations, a depositional model is developed that may be appropriate to other ancient and modern upper continental slope successions. Four facies are recognized as the primary constituents of the LNM: since most post-depositional sedimentary deformation of the deposits still allows for their classification in terms of Facies I-IV (below), soft-sediment deformation facies were not erected.
Facies I, 'parallel-laminated silty mudstone' Facies I is very thin, parallel-laminated, graded, very fine-grained sandstone, or siltstone, to mud-
Facies, facies-associations and sediment transport/deposition processes
347
FIG. 6. Cross-bedded, heavy-mineral laminated sandstones of the Seglodden Member (delta top that may have been reworked). SE side of the Kongsfjord. Hammer shaft (circled) is 35 cm long.
stone beds (Fig. 7). The grading is always normal and generally gradual, although there are some relatively abrupt transitions from siltstone to mudstone giving the impression of couplets. The beds are typically Bouma Tde turbidites with micro-loading as the only visible sole structure. Individual beds of Facies I are traceable laterally for a maximum of 50 m, and over such distances there are no changes in sedimentary structures. Facies I mainly occurs in association with Facies III in packets up to about 100 m thick. The lateral continuity of individual beds, the absence of any basal erosive features, the very thin beds, the grading, and small grain-sizes suggests sedimentation from low density, low velocity turbidity currents. The origin of the lamination in Facies I, forming an overall grading as successive silt laminae progressively become thinner in each bed (Fig. 8), are unlikely to be the result of a series of thin turbid flows, one for each lamina. Also, cohesive sediments (which Facies I in its pre-lithified state almost certainly was) are unlikely to have been deposited and resuspended by the same current (Pierce 1976), therefore, the laminations must be the result of deposition from a discrete (waning) flow event. The laminations may be explained by a depositional sorting process as proposed by Stow & Bowen (1978, 1980) or the multiple bursting-cycle mechanism of Hesse & Chough (1980). Similar fine-grained, thin-bedded deposits have been described by Piper (1978) as E1 laminated mud although, in
FIG. 7. Facies I, 'parallel-laminated silty mudstone'. Scale is 15 cm long, LNM, SE side of KongsI]ord.
348
K. T. Pickering
FIG. 8. Facies I with limited amounts of Facies III, 'structureless mudstone' shown by black arrow. White arrow shows an example of a Facies I bed in which the grading is picked out by progressively thinner laminae of silt. Pentax lens cap for scale. contradistinction to Facies I, the E1 m u d s are often l e n t i c u l a r - - a feature that is rare to Facies I.
Facies II, 'current-rippled siltstone' Facies II comprises thin-bedded and very thinbedded, graded, current-rippled, fine to very fine-grained sandstone to siltstone or silty mudstone (Fig. 9). The beds are B o u m a Tcde turbi-
dites. The ripples are asymmetric, round-crested single sets with internal laminations showing either stoss-side erosion or preservation. The ripples have wavelengths of approximately 7-8 cm, amplitudes of about 2 cm and m a x i m u m lee-side laminae inclinations of 20 ~. The internal structures are emphasized by subtle grain-size variations from fine to very fine-grained sandstone to siltstone/mudstone.
FIG. 9. Facies II, 'current-rippled siltstone'. Scale is 15 cm long. Palaeoflow from left to right, and cleavage intersects bedding at an acute angle. LNM, Blodskytodden Section.
Facies, facies-associations and sediment transport~deposition processes The top surfaces of Facies II beds mainly show straight-crested ripples although linguoid, intermediate and irregular ripple types occur. It was not possible to establish a relationship between ripple crest type and the internal lamination of ripples. The Bouma Tcd~ divisions, small grain-sizes, bed thinness, lack of directional sole structures and non-erosive bases to the beds, suggests sedimentation from low density, low velocity turbidity currents. In the cases where stoss-side erosion of the ripples was observed, it seems likely that the depositing currents were not always at maximum capacity and that they were sufficiently competent to move sediment by traction without significant sedimentation from suspension 'fallout' at that flow stage. Facies I and II may contain contourite deposits (Heezen et al. 1966) belonging to the group of muddy contourites as defined by Stow & Lovell (1979) but the distinction between turbidite and contourite deposits still remains equivocal enough to prevent such a distinction in the case of the LNM.
Facies III, 'structureless mudstone' Facies III occurs as structureless, dark-grey to black mudstone that may be silty (Fig. 8). Its thin-layered nature militates against identification in anything other than optimum weathering conditions--the best exposures being in wavewashed faces and well-weathered surfaces. Facies III was observed up to only a few mm thick, occurring above the turbidite Te division of the underlying beds of other facies. The distinction between Te and Facies III is made on the basis of colour changes between relatively lightcoloured structureless mudstone or silty mudstone with a gradational transition from the Bouma Td division, and darker coloured structureless mudstone or silty mudstone above--there being a sharp discordance between what is reasonably designated Te and the overlying mudstone. Facies III is the result of the relatively slow settling of sediments from a suspension, possibly from a nepheloid layer (see Pierce 1976 and the references therein). Such a nepheloid layer could have been fed by the outward diffusion of fluvialdeltaic discharge, low density flows down submarine canyons, or over the shelf edge, turbidity currents and resuspension by bottom currents (Pierce 1976). Further research on the LNM mudstones, such as X-radiography, may help to clarify the true relationship between Facies I, II and III. Piper (1978) recognizes a threefold division of
349
the Bouma Te interval which, from the base upwards, is defined as: El, laminated mud; E2, graded mud, and E3, ungraded mud, possibly being covered by hemipelagic sediment of the 'F' division (after Hesse 1975) or ep division of Kuenen (1966). Facies III may belong to the ungraded mud interval of turbiditic affinity or it may be derived from a nepheloid layer, but for the time being its origin remains uncertain.
Facies IV, 'wave-rippled fine/very fine-grained sandstone' Facies IV vary from beds with symmetrical cuspate to round-crested, graded or ungraded siltstones (Siedlecka & Edwards 1980, Fig. 4), to couplet-like beds without obvious grading between a lower fine-grained and an upper structureless siltstone part (Fig. 10). These beds were previously reported as very rare within the LNM (Pickering 1982a), but Siedlecka & Edwards (1980) report that the ripple cross-lamination (Facies II) and symmetrical ripples (Facies IV) increase in abundance upwards through the Naeringselva Member. Rarely, interference ripples occur. Within the couplet-like beds, the lower part is often graded medium to fine-grained sandstone showing symmetrical to asymmetrical ripple forms, with internally form-discordant, centimetre-sized, bundles of laminae (Fig. 10). These laminae are less than 1 mm thick with azimuthal differences of up to 70 ~ in the same ripple form. The internal laminae show more than one flow direction for a given ripple train, although there is always a dominant direction, similar to the current-ripple direction in the surrounding Facies II beds. The upper part of these beds is structureless siltstone or silty mudstone, draped over the underlying layer to produce a form surface to the bed. The grain-size of Facies IV beds is often coarser than the surrounding beds of Facies I and II, and the immediately underlying beds or layers are, in some cases, disturbed such that they appear to have been disrupted in situ. Some of these beds stand out because the lower parts appear to have been selectively weathered (Fig. 10). The couplet-like nature of some of the beds suggest two discrete related depositional processes. The lower part of the bed appears to be due to a current that carried the fine-grained sand, silt and mud into the depositional environment but was not a steady unidirectional flow. The grading at the base may indicate turbidity current flow as the mode of transport of the sediment, but the internal structure militates against this being the depositional process, since
35 o
K.T. Pickering
Fro. 10. Facies IV, 'wave-rippled fine/very fine-grained sandstone'. Note the superposition of two beds of this facies: the lower graded sandstone part is internally laminated, and the upper siltstone part is a structureless drape. The Facies I beds beneath show disruption of the originally horizontal laminae. LNM, SE side of Kongsfjord. the laminae imply oscillatory flow. The current seems to have been capable of eroding and depositing sediment in roughly the same place and with a variety of flow directions. This suggests that the current fluctuated in strength several times during the sedimentation process, but with an overall decrease in strength to produce the grading. The dominant flow direction suggests that there was a net sediment transport direction. The upper part of the beds implies that the final depositional stage was the slow accumulation of sediment from suspension. Invariably, there was insufficient silt and mud in suspension to fill and level the topography created by the ripple form. Thus, the complete depositional process appears to have been from a waning, initially oscillatory, flow or current. The sediment suspension mechanism may have been a wave-induced turbidity current (under storm conditions--see below) transporting sediment from further upslope, or from the shelf, where grain-sizes were slightly coarser than the L N M ambient sedimentation. Exceptional storm events are invoked because of the paucity of Facies IV, at least lower down in the succession. Towards the top of the LNM, and in the upper Naeringselva Member, where Facies IV become more common, the sedimentary environment may have been closer to, or above, wave-base for most storm events. Since the dominant transport direction, inferred from these beds, is similar to the current-ripple
directions that are interpreted to show the approximate downslope direction of flow, it is concluded that the net sediment transport direction during the storms was also downslope. The return to 'fair weather' conditions allowed the slow settling of the silt-mud drape, from suspension fallout. Perhaps, the build-up of the storm, prior to the deposition of the ripples, eroded and/or disturbed the beds below, producing the soft-sediment deformation observed in some of the underlying beds. Facies similar to this, with form-discordant internal structures, poor bi-directionality, and local late-stage draping, have also been interpreted as wave-generated (de R a a f et al. 1977). The symmetrical ripples are more typical of bedforms that have been shaped by oscillatory current motion. Like the unusual type described above in detail, they are also interpreted as wave-generated.
Soft-sediment deformation The LNM contains many different styles of soft-sediment deformation that tend to be most common in those parts of the succession where there are greater proportions of Facies II. There are essentially four main styles: (1) slides; (2) post-depositional sediment mass-flows such as intraformational breccias; (3) in situ liquefaction and fluidization structures, and (4) syn-sedimen-
Facies, facies-associations and sediment transport/deposition processes tary faults excluding those directly associated with sediment slides. Table 1 shows the salient features of these soft-sediment deformation deposits with possible mechanisms to explain their formation. A more comprehensive treatment of the various soft-sediment deformation styles may be found in Pickering (1982a). Sediment slide deposits, including slumps, are the most common post-depositional features of the LNM, and the other sedimentary deformation styles occur as parts of, or in isolation from, larger scale slides. The distinction between the various slide sheets (Table 1) is made on the degree of disruption of the beds within the slides, but undoubtedly there is a continuum between these styles that depends upon such factors as the rate of strain, the amount of pore fluid, distance of transport and the geotechnical properties of the sediments that were involved in sliding. Intra-slide beds may be essentially coherent, forming folds with a relatively consistent sense of over-turning (Fig. 11), semi-coherent such that beds are partially brecciated, disharmonically folded and locally destroyed through liquefaction and fluidization (Fig. 12), to slides in which most of the primary bedding has been destroyed to leave vestigial bedding as wisps and 'balls' of the primary sediments (Fig. 13). Cook (1979) describes many features, from a late Cambrianearly Ordovician continental slope succession in central Nevada, USA, that bear direct comparison with those found in the LNM; in the Nevada rocks, it appears that the intensively deformed
351
sediments tend to occur towards the base and margins of translational slides. The bounding surfaces of the slides are difficult to delineate in the LNM exposures because of limited outcrop widths that seldom exceed 30-40 in. However, a number of criteria were used to define the lower surfaces, for example, steeply inclined erosion surfaces above which there are disturbed sediments, and surfaces that juxtapose packets of beds with significant differences in the dips of the enclosed beds. Also, there are erosion surfaces that abruptly cut down into the underlying beds and show sediment draping above (Fig. 14): such surfaces may be due to sliding events leaving a slide scar, or they may represent current scours into which subsequent sediments were deposited to level any uneven topography (cf., Clari & Ghibaudo 1979). Most of the LNM soft-sediment deformation shows evidence of lateral movement and, or, rotation of the deformed sediments in the form of: ( I ) erosion into underlying beds by the deformed unit: (2) glide planes along which sediment failure occurred (see paragraph above), and (3) folded beds in sheets between undisturbed, normally bedded deposits. Generally, the deformation occurs within single sheet-like deposits, but in a few cases it occupies channel-like depressions in cross section (Figs 15 & 16). In these examples, the depression or scour is filled by a single slide event that presumably excavated the depression. Usually, it is possible to examine the pre-slide nature of the incorporated deposits, and they
FIG. 11. Coherent soft-sediment fold sheet (slide) showing a consistent sense of over-turning to the left (down-slope). Pentax lens cap for scale (arrowed). LNM, SE side of KongsOord.
K.T. Pickering
35 z
09
w z
z
_o~ ~
o
<
oc
I-J J,.
0 -u.
w "r co
r,
_
r,
z 8
0 z co w
m
,~
o o.
~:
~:
i
~
0")
U)
CO
O0
(")
Z
C]
2
a.
a.
0
F-
0
I
W I--
- -
u. o
0
Z
w
0
0
w I~
~
0
z
~_
0
0
i-
-.~
z 0
o-.~ w z w oc cc w (D i_ w Z t~ _
W I--
W
cc 0 ~ w r7
W I--
ILl I--
O
W W
n" W
nW
W
CC W
CO: W
nLJJ
OC ca
W co
0 :E
~
:E
w u)
0 :E
0 :E
0 I-
<
0
LJ-
I-z
0
0
LU n," "-~ I-'~ Z
0 D: UJ O_ O.
cc
w _l co ~c
n7I0 0 ~ (/)
0 w 0: r~_
~
7I0 0 ~ cO
0 w nn.-_
7I0 0 ~ of)
7t0 0 ~ cO
7I0 0 ~ oO
cc 0 u_ z
0 0
I--
u_ 0 I
0 r~
LU n-
w
I-Z 0 0 CC LU
z
._;
:E z~
co 00 w z
x
o
,< :E
"r F-
m w- < ~
~ ~
~
~o oo
--I
w
Z 0 k-
_j
0
~c
~
w- < ~
~
wz
w
~ ~
~o oo
~o oo
~ o
~
E
E
E
E
E
E
E
~
o§
c~i .
co.
~,_
.,-
,,-.
~
CO
0
0
I-w
~-~
~_
z o
0 o
< -
0 ~
~
I-
111
w
w w o T (~o~oo0
8133HS
o
w
-
0
w m
.~ co
~~
-7
z
._i u_
,,,
_
0 ,,
oo i
w
3(]17S
w 0 -r oo :z: ~o
0 o
Facies, facies-associations and sediment transport~deposition processes
353
FIG. 12. Semi-coherent deformed soft-sediment sheets (3 are arrowed). Hammer shaft is 35 cm long (circled). LNM, SE side of Kongsfjord.
FIG. 13. Close-up view of a semi-homogenized slide sheet, the base of which is arrowed. Note the almost complete assimilation of the incorporated Facies I beds into the 'massive' siltstone 'matrix'. LNM, SE side of Kongsfjord.
354
K.T. Pickering
FIG. 14. Small-scale slide surface with the overlying slope deposits draping the depression created by sediment sliding (arrowed). Pentax lens cap for scale (circled). LNM, SE side of Kongsfjord. appear to comprise similar deposits to the surrounding facies. Both tend to contain a high proportion of Facies II. Although it is impossible to appreciate the three-dimensional geometry of the slide-filled structures, they are tentatively interpreted as transverse sections through slope gullies. Since the width-to-depth ratios of these depressions ranges from about 5:1 to 8: 1, outcrop widths of at least a couple of hundred metres are needed to
see if some of the thickest slide deposits infill depressions, or if they rested above the surrounding sediments at the time of their emplacement as features with positive relief.
Other features A feature that commonly occurs in the LNM are packets of beds, 10-100 cm thick, separated by
FIG. 15. Western part of slide deposit that filled a channel-like depression in slope sediments. White arrows delineate the margin of the slide (with normal bedding on the right and at the top of the plate. 30 cm-high shoulder bag for scale (circled). LNM, Blodskytodden Section.
Facies, facies-associations and sediment transport~deposition processes
355
FIG. 16. Small-scale slide deposit that filled a gully-like depression in slope sediments. Arrows show margin of slide, and normal bedding is visible in the foreground and parallel with the 15 cm scale (circled). LNM, Storflogdalen Section. erosional-depositional discordant surfaces (Fig. 17). The bounding surfaces to the packets are inclined up to a few degrees from the bedding planes in the underlying beds. These surfaces are formed either by erosion upon which subsequent sediment draping occurred, or they are purely depositional discontinuities formed by sedimentation upon an uneven topography. Some of these
discordances may be attributable to syn-sedimentary erosion as described by Clari & Ghibaudo (1979) and by Dott & Bird (1979), or to slump scars subsequently infilled by slope sediments (Lewis 1971; Roberts et al. 1976). Dott & Bird (1979) have demonstrated that channels in the Elkton Siltstone Member of the Tyee Formation (Eocene, SW Oregon, USA)
FlG. 17. Packets of slope beds defined by: (1) Erosional surfaces upon which subsequent sediment draping occurred (immediately below 15 cm scale), or (2) purely depositional sloping surfaces as seen just above the continuous bed that is arrowed. LNM, Blodskytodden Section.
356
K.T. Pickering
were incised into, and then filled in by, finegrained, thin-bedded deposits. By analogy with the Elkton Siltstone Member, many of the LNM discontinuities may be a result of sedimentation on a pervasively gullied basin slope but without firm evidence to support this hypothesis, the exact nature of many of the discontinuities must remain speculative. Siedlecka & Edwards (1980) describe a unit of pink sandstone approximately 360 m from the base of the LNM near Storflogdalen (Fig. 1) that measures approximately 10 x 2 m. This sandstone unit is unique to the LNM, resembles the sandstones seen in the younger Seglodden Member of the Bfisnaering Formation and comprises softsediment deformed beds. Therefore, it appears that sediments from the shelf occasionally were caught up, e n m a s s e , in resedimentation events such as slides from near to the shelf-break.
Synthesis The LNM thickens eastwards from approximately 60 m in the Oksetoppen Section and the section south-east of Kongsl]ord, to 540 m in the Blodskytodden Section (Fig. 3). Current-ripples indicate palaeoflow towards the east-north-east (Fig. 18) and this is consistent with observations and measurements from the underlying Kongst]ord Formation (Pickering 1981a), and also the
palaeocurrent data of Siedlecka & Edwards (1980) for the four members of the Bgsnaering Formation (Figs. 19(a) 8,: (b)). A limited number of wave-ripple crests were observed, indicating flows approximately opposite to those inferred from the current-ripple directions. A small number of slide fold orientations were measured and they suggested a slope that was inclined towards the east-north-east. Palaeocurrent data from the Kongs0ord and B~snaering Formations (above) suggests that the basin slope was inclined towards the east to north-east. Therefore, the LNM current-ripples are most likely to have formed from downslope flowing currents. Also, the asymmetry of some of the wave-generated ripple forms (seeFacies IV) is most easily explained as a consequence of the net~ downslope, sediment transport under storm conditions. In the three coastal sections~ the relative proportions of Facies I and II were estimated, and packets of Facies-association I/II were defined (Fig. 20). The paucity of Facies IV within the LNM~ although its importance must not be under-rated, led to its omission from Fig. 20, and since Facies III most often is associated with Facies I, they were included together as 'Facies I'. Typical bed thicknesses and grain-sizes were also estimated, with the rarity or abundance of softsediment deformation in each defined packet of beds (Fig. 20). From Fig. 20~ two conclusions
Fro. 18. Palaeocurrent data and slide fold data from the LNM. The rose diagrams are for current-ripples and the arrows show the inferred down-slope direction deduced from measuring slide folds. 'N' is number of readings and 'm' is the mean direction of current flow. Compare with Figs 19(a) & (b).
Facies, facies-associations and sediment transport/deposition processes
NAERINGSELVA
& SEGLODDEN
357
MEMBERS
~f~,~,,..~..~.__..~ BARENTS SEA I'-~
'35
(a)
GODKEILA
& HESTMAN
MEMBERS
26,":' " ' ' ' ' "
9" . : - :
""'"~" " "
""'"""""
"
~
BARENTS
SEA
'(:'?
z
~ "'i15
(b)
FIG. 19. Outcrop of the B~,snaering Formation with rose diagrams of cross-bedding and current-ripple flow directions for the four members. T K F = Trollfjord Komagelv Fault. Numbers next to rose diagrams refer to the number of readings, and the average current flow direction at sites of less than 10 readings is indicated by an arrow. Diagrams are redrawn from Siedlecka & Edwards (1980). (a) Naeringselva Member readings as dashed lines and Seglodden Member as solid lines; (b) Godkeila Member readings as dashed lines and Hestman Member as solid lines. Note that the overall palaeocurrent directions throughout the B~}snaering Formation tend to show east to north-easterly flows.
K.T. Pickering
358
E 0 i Z
I
- -
_
- -
9
c..~
i
i
- -
LU
|
I
I
i
l
i
I
I
_
I
"~
~'~
~
.
~
~
>-,
9 _J
m
=
~
m
E
~
~'~ . ..,.,
"-d "
b,'h
i=
I
i
,0 b"q I
0 m
0 ~0
i
i
r',"hu'ht,,% i
i
"~
m
~'~
~
r"4
I
O _ i
i
i
~
~
i
i
0
Z UJ
"~
~
"~
ouil
AJa^
"-
tm
.~
~ 0
'=="I,,
~
,-,,,
,-,-
I
I
l
I
I
]
J
I
~=~
:
I
o
O +-,
I'--
o
i
_,_
~
~
i
I
/
I
--
O
,e: ,,,
O,-~
-
O
123 0
~
?
E
,.
~
I
i: ]
~ I
1
I'
I o
(2)
~ ,--
~N
~N ~N m~ ~
d
Facies, facies-associations and sediment transport~deposition processes may be drawn: (a) soft-sediment deformation is abundant where Facies II is dominant and is rare, or absent, wherever the section is mainly Facies I, and (b) the sections comprise alternations of relatively disturbed and undisturbed deposits, often up to tens of metres thick, associated with variations in the amount of Facies I and II (see first conclusion).
LNM depositional model The common occurrence of slide sheets and other types of soft-sediment deformation, the thin-bedded fine-grained deposits, and the stratigraphic framework of the LNM, suggests sedimentation on an upper basin-slope. The slope was constructed by east-north-easterly, downslope flowing, low density, low velocity turbidity currents, a minor amount of hemipelagic deposition, some storm event beds, and a substantial proportion of slide and slump sheets. Deposition was essentially below normal storm wave base. The upward-coarsening, and the transition into fluvio-deltaics, suggest that the LNM is, at least towards the top, an ancient pro-delta (Siedlecka & Edwards 1980; Pickering 1982a). However, the thickness of the LNM and the occurrence of deep water submarine fan deposits beneath the member suggest that much of the succession represents upper basin-slope deposits: fan sediments having accumulated on the lower slope. Into this slope, gullies of varying dimensions were excavated and filled by similar sediments. Sediment instability was a common feature of this slope, presumably producing many of the structures described by Coleman & Prior (1980, see Figs. 35, p. 90, 42, p. 103 & 48, p. 113) from the pro-delta and delta front of the Mississippi Delta. Sediment slides are widely reported from many ancient basin-slopes (e.g. Mutti & Ricci Lucchi 1972: Cook 1979; D o t t & Bird 1979) where they constitute substantial volumes of the slope successions, exhibiting many of the characteristics found in the LNM. The dramatic thickening of the slope succession by more than 400 m, from west to east on Varanger Peninsula, over an estimated downslope distance of 5-10 km (Pickering 1982a) suggests a NNW-SSE striking basin margin. The accumulation of 2.5-3.5 km of fluvio-deltaics above the fan-slope system indicates substantial subsidence of the depositional basin that may have occurred because of fault-controlled subsidence on the basinward side of a basin hinge line oriented NNW-SSE Since Facies I and II are the product of the same sediment transport process, with Facies II
359
as Tcde and Facies I as Tde turbidity current deposits, the former is probably the more 'proximal' equivalent of Facies I: therefore, the LNM succession may be showing alternations in the proximality of the sediment-feeding system, in terms of packets of the slope beds. A possible cause of the packeting of slope deposits could be delta lobe switching (Scruton 1960), combined with a pendulum effect on the location of the main delta-fed slope sediments that may have been induced by the weight of submarine fan sedimentation on the lower basin-slope (compare with the model proposed by Daiily 1976, for the Niger Delta). When the delta lobe feeding a given part of the upper basin-slope switched, or migrated, away from an area, thinner bedded and finer grained sedimentation would have ensued. The ~quieter' depositional environment would have favoured the accumulation of the possible hemipelagic deposits, thereby explaining their association with Facies I. Sea-level changes, either in conjunction with, or in isolation from, the processes already mentioned could also account for the facies-association packets in the LNM. Finer grained, thinner bedded sediments would accumulate during relatively high stands of sea-level. If sea-level changes were the primary factor controlling the nature of slope sedimentation, then in turn they may have been influenced by local tectonic effects triggered by the vertical and lateral accretion of the BS.snaering Formation Delta. Many of the possible ancient analogues for the LNM that have been documented (e.g. Walker 1966; 1970; Laird 1968; Galloway & Brown 1972; Keith & Friedman 1977; Kepferle 1978; Dott & Bird 1979; Siemers 1981) mention channels incised into the basin slope that are filled with coarser grained, thicker bedded sandstones than the ambient slope sediments (e.g. Walker 1970). Such channel fills appear to be absent from the LNM, and any channels that may be present are filled with identical sediments to those into which they were cut, analogous to the slope channels described by Dott & Bird (1979).
LNM modern analogues The thickness of the Barents Sea Group (at least 9 kin), with a minimum 3.2 km of submarine fan deposits and 2.5-3.5 km of fluvially-dominated deltaics covering an area greater than about 3000 km ~ suggests deposition on an ancient continental margin. The scale of the delta-slope-fan system, where the submarine fan is believed to have had a radius of at least 100 km at one stage (Pickering 1981a), favours an analogy between
360
K.T. Pickering
some of the major deltas of the world such as the Mississippi and Niger Deltas, and their associated deeper water sediments, and the late Precambrian Kongsl]ord-BS.snaering Formations. Studies of the Mississippi (Huang & Goodell 1970; Sangree et al. 1978; Woodbury et al. 1978; Moore et al. 1978; Booth 1979) and Niger Delta-Slope-Fan systems (Allen 1964, 1965; Oomkens 1974) have revealed upper continental slopes, and pro-deltas, of fine-grained thin-bedded muds/silts with gullies and channels incised into the slopes, and abundant soft-sediment deformation structures such as slides and slumps--also, characteristics of the LNM. Booth (1979) attributes the omnipresent condition of slope instability on the upper continental slope of the northern Gulf of Mexico to the rapid deposition of fine-grained sediment: he states the following series of events that lead to mass-wasting of the slope: rapid deposition; excess pore pressure above hydrostatic develops; abnormally low effective overburden pressure producing inadequate shear strength in the sediments and consequently failure. Although other factors may have contributed to sediment failure by sliding, in the L N M slope, fast rates of sedimentation probably were the single most important cause. Other causes of mass-wasting of upper continental slope sediments that are regarded as having been of particular importance in the construction of
modern slopes include earthquake activity (e.g. the Gulf of Alaska as described by Carlson 1978) and flowage of evaporites within fine-grained sediments, for example, in the continental margin of Israel and Northern Sinai (Almagor & Garfunkel 1979), but there is no evidence to indicate that such processes were important for the deposition of the LNM.
Discussion Channel fills of coarse-grained deposits were not observed in the L N M although it is possible that they exist in unexposed parts of the succession. However, their apparent absence suggests that the sections studied were lateral to the main area of sediment supply to a submarine fan, assuming that fan sedimentation was still occurring. Although comparisons are made between modern upper continental slope sedimentation and the LNM, the thickness of the ancient succession, up to a maximum 540 m, seems to be considerably less than might be expected. Postdepositional compaction and tectonic deformation of the LNM may have reduced considerably the original thickness such that the discrepancies in thickness between modern and postulated ancient upper continental slopes is more apparent than real.
References ALLEN,J.R.L. 1964. The Nigerian continental margin, bottom sediments, submarine morphology and geologic evolution. Marine Geol., 1, 289-332. -1965. Late Quaternary Niger Delta and adjacent areas: sedimentary environments and lithofacies. Bull. Am. Ass. Petrol. Geol., 49, 547 600. ALMAGOR, G. & GARFUNKEL, Z. 1979. Submarine slumping in continental margin of Israel and northern Sinai. Bull. Am. Ass. Petrol. Geol., 63, 324-40. AMERY, G.B. 1978. Structure of continental slope, northern Gulf of Mexico. In: Bouma, A.H., Moore G.T. & Coleman J.M. (eds), Framework, Facies, and Oil-Trapping Characteristics of the Upper Continental Margin. Am. Ass. Petrol. Geol. Studies in Geology, 7, 141-53. BOOTH,J.S. 1979. Recent history of mass-wasting on the upper continental slope, northern Gulf of Mexico, as interpreted from the consolidation states of the sediment. In: Doyle L.J. & Pilkey O.H. (eds), Geology o[ Continental Slopes. Soc. econ. Paleont. Min., Tulsa, 27, 153-64. CARLSON, P.R. 1978. Holocene Slump on Continental Shelf Off Malaspina Glacier, Gulf of Alaska. Bull. Am. Ass. Petrol. Geol., 62, 2421-6. CLARI, P. & GHmAUDO,G. 1979. Multiple slump scars
in the Tortonian type area (Piedmont Basin, northwestern Italy). Sedimentology, 26, 475-96. COLEMAN, J.M. & PRIOR, D.B. 1980. Deltaic Sand Bodies. Continuing Education Course Note Series No. 15, Department o['Education. Coastal Studies Institute, Louisiana State University. 171 pp. COOK, H.E. 1979. Ancient continental slope sequences and their value in understanding modern slope development. In: Doyle L.J. & Pilkey O.H. (eds), Geology 0[" Continental Slopes. Soc. econ. Paleont. Min., Tulsa., 27, 287-305. DAILLY, G.C. 1976. Pendulum effect and Niger Delta prolific belt. Bull. Am. Ass. Petrol. Geol., 60, 1543-75. DOTT, R.H. JR. & BIRD, K.J. 1979. Sand transport through channels across an Eocene shelf and slope in southwestern Oregon, USA In: Doyle L.J. & Pilkey O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleont. Min., Tulsa, 27, 327-42. DOYLE, L.J. & PILKEY, O.H. (eds) 1979. Geology of Continental Margins. Soc. econ. Paleont. Min., Tulsa, 27. GALLOWAY, W.E. & BROWN, L.F. JR. 1972. Depositional systems and shelf-slope relationships in Upper
Facies, facies-associations and sediment transport/deposition processes Pennsylvanian rocks, north-central Texas. Unit'. Texas Rep. Invest. Bureau econ. Geol., 75, 62 pp. HEEZEN, B.C., HOLLISTER, C.D. & RUDDIMAN, W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents Science, 152, 502-8. HESSE, R. 1975. Turbiditic and non-turbiditic mudstone of Cretaceous flysch sections of the eastern Alps and other basins. Sedimentology, 22, 387-416. & CHOUGH, S.K. 1980. The Northwest Atlantic Mid-Ocean Channel of the Labrador Sea: II. Deposition of parallel laminated levee-muds from the viscous sublayer of low density turbidity currents. Sedimentology, 27, 697-711. HUANG, T.C. & GOODELL, H.G. 1970. Sediments and sedimentary processes of eastern Mississippi Cone, Gulf of Mexico. Bull. Am. Ass. Petrol. Geol., 54, 2070-2100. JOHNSON, H.D., LEVELL, B.K. & SIEDLECKI, S. 1978. Late Precambrian sedimentary rocks in East Finnmark, north Norway and their relationship to the Trollfjord-Komagelv fault. J. geol. Soc. Lond., 135, 517-33. KEITH, B.D. & FRIEDMAN, G.M. 1977. A slope-fanbasin-plain model, Taconic Sequence, New York and Vermont. J. sed. Petrol., 47, 1220-41. KELTS, K. & ARTHUR, M.A. 1981. Turbidites after ten years of deep-sea drilling--wringing out the mop? In: Warme J.E., Douglas R.G. & Winterer E.L. (eds), The Deep Sea Drilling Project: A Decade of Progress. Soc. econ. Paleont. Min., Tulsa., 32, 91-127. KEPFERLE, R.C. 1978. Prodelta Turbidite Fan Apron in Borden Formation (Mississippian), Kentucky and Indiana. In: Stanley D.J. & Kelling G. (eds), Sedimentation in Submarine Canyons, Fans. and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 224-38. KJODE, J., STORETVEDT,K.H., ROBERTS, D. & GIDSKEHAUG, A. 1978. Palaeomagnetic evidence for largescale dextral displacement along the TrollfjordKomagelv Fault, Finnmark, North Norway. Phys. Earth & Planet. Interiors, 16, 132-44. KNEBEL, H.J. & CARSON, B. 1979. Small-scale slump deposits, middle Atlantic continental slope. Marine Geol., 29, 221-36. KUENEN, PH. H. 1966. Experimental turbidite lamination in a circular flume. J. Geol., 74, 523-45. LAIRD, M.G. 1968. Rotational slump scars in Silurian rocks, western Ireland. Sedimentology, 10, 111-20. LEWIS, K.B. 1971. Slumping on a continental slope inclined at 1 - 4 . Sedimentology, 16, 97-110. MACILVAINE, J.C. & Ross, D.A. 1979. Sedimentary Processes on the Continental Slope of New England. J. sed. Petrol., 49, 563-74. MOORE, D.G. 1961. Submarine slumps. J. sed. Petrol., 31, 343-57. MOORE, G.T., STARKE,G.W., BONHAM, L.C. & WOODBURY, H.C. 1978. Mississippi Fan, Gulf of Mexico-physiography, stratigraphy and sedimentation patterns. In: Bouma A.H., Moore G.T. & Coleman J.M. (eds), Framework, Facies, and Oil-Trapping Characteristics of the Upper Continental Margin. Am. Ass. Petrol. Geol. Studies in Geology 7, 155-91.
-
36I
MUTTI, E. & RICCI LUCCHI, F. 1972. Le torbiditi dell'Appennino settentrionale: introduzione all'analisi di facies. Mere. Soc. geol. ltal., 11,161-99. OOMKENS, E. 1974. Lithofacies relations in the Late Quaternary Niger Delta complex. Sedimentology, 21, 195-222. PICKERING, K.T. 1979. A Precambrian Submarine Fan and Upper Basin-Slope Suceession in the Barents Sea Group, Finnmark, North Norway. Unpubl. Ph.D. Thesis, Univ. of Oxford, England. 239 pp. 1981a. The Kongsfjord Formation--a Late Precambrian submarine fan in Northeast Finnmark, North Norway. Norg. geol. Unders., 367, 77-104. 1981b. Transitional submarine fan environments IAS 2nd EUR. MTG., Bologna, Italy, Abstract vol.,
-
-
-
-
-
1 4 2 - 4 .
-
-
1982a. A Precambrian upper basin-slope and prodelta in northeast Finnmark, North Norway--a possible ancient upper continental slope. J. sed. Petrol., 2, 171-86. 1982b. Middle-fan deposits from the late Precambrian Kongsfjord Formation Submarine Fan, Northeast Finnmark, northern Norway. Sed. Geol., 33, 79-110. 1983. Transitional submarine fan deposits from the late Precambrian Kongsfjord Formation Submarine Fan, N.E. Finnmark, N. Norway. Sedimentology, 30, 181-99. PIERCE, J.W. 1976. Suspended sediment transport at the shelf break and over the outer margin. In: Stanley D.J. & Swift D.J.P. (eds), Marine Sediment Transport and Environmental Management. John Wiley, New York. 437-58. PIPER, D.J.W. 1978. Turbidite muds and silts on deepsea fans and abyssal plains. In: Stanley D.J. & Kelling G. (eds), Sedimentation in Submarine Canyons, Fans, and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. - - , NORMARK,W.R. & INGLE, J.C. JR. 1976. The Rio Dell Formation: a Plio-Pleistocene basin slope deposit in Northern California. Sedimentology, 23, 309-28. RAAF, J.F.M. DE, BOERSMA, J.R. & GELDER, A. VAN 1977. Wave-generated structures and sequences from a shallow marine succession, Lower Cretaceous, County Cork, Ireland. Sedimentology, 24, 541-83. ROBERTS, D. 1972. Tectonic deformation in the Barents Sea Region of Varanger Peninsula, Finnmark. Norg. geol. Unders., 282, 1-39. ROBERTS, H.H., CRATSLEY, D.W. • WHELAN, T. 1976. Stability of Mississippi Delta sediments as evaluated by analysis of structural features in sediment borings. Off[~hore Tech. Conf, Texas, Paper No. OTC 2425. SANGREE,J.B., WAYLETT,D.C., FRAZIER,D.E., AMERY, G.B. & FENNESSY, W.J. 1978. Recognition of continental-slope seismic facies, offshore Texas-Louisiana. In: Bouma A.H., Moore G.T. & Coleman J.M. (eds), Framework, Facies, and Oil-Trapping Characteristics of the Upper Continental Margin. Am. Ass. Petrol. Geol. Studies in Geology 7, 87-116. SCRUTON, P.C. 1960. Delta building and the delta -
-
362
K.T. Pickering
sequence. In: Shepherd F.P. & Andel Tj. H. van (eds), Recent Sediments, Northwest Gulf of Mexico. Am. Ass. Petrol. Geol. 8-102. SnEmDAN, R.E. 1981. Recent research on passive continental margins. In: Warme J.E., Douglas R.G. & Winterer E.L. (eds), The Deep Sea Drilling Project: A Decade of Progress. Soc. econ. Paleont. Min., Tulsa., 32, 39-55. SIDNER, B.R., GARTNER,S. & BRYANT,W.R. 1978. Late Pleistocene geologic history of Texas outer continental shelf and upper continental slope, ln: Bouma A.H., Moore G.T. & Coleman J.M. (eds), Framework, Facies, and Oil-Trapping Characteristics of the Upper Continental Margin. Am. Ass. Petrol. Geol. Studies in Geology 7, 243-66. SIEDLECKA, A. 1972. Kongsfjord Formation--a late Precambrian flysch sequence from the Varanger Peninsula, Finnmark. Norg. geol. Unders., 2 7 8 , 41-80. - & EDWARDS, M.B. 1980. Lithostratigraphy and sedimentation of the Riphean Bfisnaering Formation, Varanger Peninsula, North Norway. Norg. geol. Unders., 355, 27-47. - - - & S~EDLECK~, S. 1967. Some new aspects of the geology of Varanger Peninsula (northern Norway). Norg. geol. Unders., 247, 288-306. SIEMERS, C.T. 1981. Sedimentological core analysis of deep-water clastic sediments in the down-dip Woodbine-Eagle Ford Interval (Upper Cretaceous), Typer County, Texas. In: Siemers, C.T., Tillman, R.W. & Williamson, C.R. (eds), Deep- Water Clastic Sediments. A Core Workshop. Soc. econ. Paleont. Min., Tulsa, Core Workshop No. 2. 249-371. STOW, D.A.V. 1980. Distinguishing between finegrained turbidites and contourites on the Nova Scotian deep water margin. Sedimentology, 26, 371-87. & BOWEN, A.J. 1978. Origin of lamination in deep-sea fine-grained sediments. Nature, 274, 324-28.
--
& BOWEN, A.J. 1980. A physical model for the transport and sorting of fine-grained sediments by turbidity currents. Sedimentology, 2 7 , 31-46. - & LOVELL,J.P.B. 1979. Contourites: their recognition in modern and ancient sediments. Earth Sci. Rev, 14, 251-91. TAYLOR, P.N. & PICKERING, K.T. 198l. Rb-Sr Isotopic Age Determination on the late Precambrian Kongsfjord Formation, and the Timing of Compressional Deformation in the Barents Sea Group, East Finnmark. Norg. geol. Unders., 367, 105-10. TILLMAN, R.W., HUGHES, S.M., SCOTT, R.M., & DAILEY, D.H. 1981. Upper Cretaceous deep water Winters Sandstone, Cities Service Nixon Community No. 1, Soland County, California. In: Siemers, C.T., Tillman, R.W. & Williamson, C.R. (eds), Deep-Water Clastic Sediments. A Core Workshop. Soc. econ. Paleont. Min., Tulsa, Core Workshop No. 2, 45-76. VIDAL, G. 1981. Aspects of problematic acid-resistant, organic-walled microfossils (acritarchs) in the Upper Proterozoic of the North Atlantic Region. Precamb. Res., 15, 9-23. WALKER, R.G. 1966. Shale Grit and Grindslow Shales: transition from turbidite to shallow water sediments in the Upper Carboniferous of northern England. J. sed. Petrol., 36, 90-114. - 1970. Deposition of turbidites and agitated-water siltstones: a study of the Upper Carboniferous Westward Ho! Formation, North Devon. Proc. Geol. Ass., 81, 43-67. WOODBURY, H.C., SPOTTS, J.H. & AKERS, W.H. 1978. Gulf of Mexico continental-slope sediments and sedimentation. In: Bouma, A.H., Moore, G.T. & Coleman, J.M. (eds), Framework, Facies, and OilTrapping Characteristicw of the Upper Continental Margin. Am. Ass. Petrol. Geol. Studies in Geology 7, 117-37. WOODCOCK, N.H. 1979. Sizes of submarine slides and their significance. J. struct. Geol., 1, 137-42.
K.T. PICKER1NG, Department of Geology, University of London, Goldsmith's College, London SE8 3BU, UK.
Continental source area contributions to fine-grained sediments on the Oregon and Washington continental slope L.A. Krissek SUMMARY: Chemical and quantitative mineral abundance data for the 2-20/~m fraction of 78 samples from the Oregon-Washington continental slope have been modelled using linear programming to estimate sediment contributions from continental source areas and evaluate the effects of mixing material derived from these sources on the formation of hemipelagic sediments. The mineral data distinguish three source areas, while the chemical data define only two. Results from the two data sets indicate similar patterns and magnitudes of source area influence on the continental margin. The contributions are used to construct a sediment budget for 2-20/~m-sized sediments on the continental slope. The slope sediments contain approximately 47% Columbia River, 32% Klamath Mountain, and 21% California Coast Range material. Approximately 25-500/o of the Columbia River, 29-46% of the Klamath Mountain, and 7-12~o of the California Coast Range annual suspended sediment input is retained on the slope, indicating that the slope acts as a more effective trap of 2-20/~m material from proximal than distal source areas. This study also demonstrates the importance of multiple sediment sources and sediment mixing in the formation of hemipelagic sediments on a continental margin. Past studies of fine-grained sediment dispersal in marine environments have generally inferred the relative importance of multiple source areas on the basis of the abundance of single minerals, especially the clay minerals (Biscaye 1965; Griffin et al. 1968; Kolla & Biscaye 1973; Kolla et al. 1976; Pinet & Morgan 1979; Griggs & Hein 1980; Karlin 1980). However, the proposed contributions cannot be quantified by such a univariate approach, particularly if a sedimentary deposit receives contributions from two or more sources. Several recent efforts have used multivariate chemical data sets and mixing models to quantify contributions from various sources. Heath & Dymond (1977) applied normative analysis to a data set of elemental abundances from deep-sea sediment samples in order to describe the sediments as mixtures of hydrothermal, detrital, hydrogenous, and biogenous material endmembers. Dymond & Eklund (1978) used a linear programming solution of elemental abundances to estimate constituent abundances in metalliferous sediments. Heath & Pisias (1979) also used linear programming to convert X-ray diffraction analysis results to mineral abundances using talc as an internal standard. Cobler & Dymond (1980), Dymond et al. (1980), Dymond (1981), and Heath & Dymond (1981) have continued to expand the application of linear programming to deep-sea chemical data, and Heath & Dymond (1981) and Leinen & Pisias (1982) have incorporated Q-mode factor analysis as an objective technique for the determination of end-member compositions from the data set.
The dispersal of hemipelagic sediments on the Oregon and Washington continental slope has been discussed previously in a qualitative manner by Russell (1967), Duncan et al. (1970), Spigai (1971), Olmstead (1972), Baker (1973), Karlin (1980), and Krissek & Scheidegger (1983). Only the work by Krissek & Scheidegger (1983), however, has involved both analyses of material coarser than 2 ~m and the use of quantitative X-ray diffraction techniques (Gibbs 1967a). The first point is important because particles in the 2-20 pm size range form approximately 40% of the hemipelagic surface sediments on the continental slope; the second is important because only quantitative mineral abundances can be compared with other absolute (e.g. geochemical) data from the same samples (Heath & Pisias 1979). On the basis of their recent work, Karlin (1980) and Krissek & Scheidegger (1983) concluded that the major continental sources of sediments found on the Oregon-Washington continental slope are the California Coast Range, Klamath Mountains, and the Columbia River (Fig. 1). Both studies indicated that material derived from the southerly sources is transported northward on the margin for distances of hundreds of kilometres. While Karlin (1980) emphasized the role of a poleward undercurrent, Krissek & Scheidegger (1983) concluded that the transport could be attributed to either the poleward undercurrent, wind-driven winter circulation on the continental shelf and subsequent downslope transport of turbid layers (Spigai 1971), or a combination of the two.
363
364
L.A. Krissek 127 ~
125 ~
~ , ."
123 ~
"',
COAST RANGE
tinuously alter the quantitative composition of material derived from a source area (Gibbs 1976b; Arcaro 1978).
47 ~
Sampling and analytical methods
9 . "~:;' :~U-:-~ ~:~%:., ~',u WASI?..% {~dR~"I
9 4i ././~ ~i / z~ LL...
o ~ /
t
OREGON COAST RANGE
. ~. 1
"
\ I
i ~-~,
~'/ ' ('
I ; L .......... ~ . _
RIVER
"'~'~'} "
~ ~ \. ORE. ~I: ~,CALIE
' ~~
::,:" , :2:~/ _
4~'? I
KLAMATH MOUNTAINS
, CALIFORNIA _J C O A S T RANGE
FIG. 1. Bathymetry of the Oregon-Washington continental slope and location of the samples used in this study. Contours are in metres. The locations of the major continental source areas of the Pacific Northwest are shown, and river samples used are listed in Table 1.
As demonstrated by Krissek (1982), fluvial sediments derived from each of these source areas in the 2-20 #m size fraction have recognizable mineral signatures. In addition, the 2-20 #m fraction forms a major portion (averaging approximately 40%) of the hemipelagic surface sediments on the continental slope. Therefore, the combined fluvial and marine data sets of Krissek (1982) and Krissek & Scheidegger (1983), respectively, provide an excellent opportunity to use linear programming modelling to quantitatively investigate the dispersal and deposition of continental margin sediments. By restricting this investigation to the 2-20/~m size fraction, a major portion of the slope sediments are considered, a n d size-selective transport and source area compositional heterogeneity do not pose major problems. A similar attempt at modelling source area contributions to the < 2 #m size fraction was unsuccessful because between-source mineral signatures in the finer fraction were less distinct than in the 2-20 #m fraction, within-source compositional heterogeneity was greater (Krissek 1982), and size-selective transport processes may con-
Surface sediment samples from 78 cores taken on the Oregon-Washington continental slope were used in this study (Fig. 1). Samples from 14 streams entering the Pacific Ocean were used to characterize the mineral composition of material derived from the three major continental source areas. These streams are listed in Table 1. Because of the greater time involved in chemical analyses, only six of these stream samples were used to characterize the chemical composition of the fluvial sediments. Because these six streams (Eel and Mad Rivers, Redwood Creek, Klamath, Rogue, and Columbia Rivers) carry approximately 88% of the total annual suspended sediment discharged from the three major source areas and approximately 80% of the total annual sediment discharged from the entire Pacific North-west (Karlin 1980), it can be said with confidence that these samples are representative of the material entering the Pacific Ocean within the study area. Samples were disaggregated and size-separated using standard techniques (Krissek 1982; Krissek & Scheidegger 1983). The 2-20/~m size fractions were freeze-dried and analysed using a quantitative X-ray diffraction technique (Gibbs 1967a). Results were presented by Krissek (1982) for the fluvial samples and by Krissek & Scheidegger (1983) for the marine samples 9 Elemental abundances of Si, AI, Fe, Ca, Na, K, Mg, Ti, Ba, Sr, Cu, and Mn were determined by atomic absorption spectrophotometry and calorimetric techniques (Krissek et al. 1980). End-member compositions representing the material derived from each source area for the two data sets (2-20 #m mineral and 2-20 pm chemical), were calculated using a dischargeweighted approach (Krissek 1982). Mineral data from 14 streams and chemical data from six streams were used in constructing the endmembers. Annual suspended sediment discharge values were taken from Karlin (1980). The end-member compositions were then used in linear mixing equations which include the weight fractions of each end-member as unknowns. Linear programming was used to calculate the weight fractions which estimate source area contributions to the slope sediments. The theory and detail of the linear programming algorithm used for this calculation has been discussed by Heath & Pisias (1979). Five mixing equations were used for the mineral calculations
Fine-grained sediment contributions on the Oregon and Washington slope 365 TABLE 1. River samples used to characterize material derived from the three major continental source areas of the Pacific Northwest of the US, arranged from north to south. Source area designations are: CR = Columbia River, K M = Klamath Mountains, CCR = California Coast Range River name Columbia River Sixes River Elk River Rogue River Hunter River Pistol River Chetco River Winchuck River Smith (CA) River Klamath River Redwood Creek Little River Mad River Eel River
and six mixing equations were used for the chemical data because Q-mode factor analysis of the mineral and chemical abundances indicated that five minerals and six elements accounted for most (> 80%) of the variance in the data sets. In a three end-member system, each combination of three equations from the five or six that are available provides a unique solution. Therefore, criteria must be provided for the selection of a 'best' solution. In the present study, two criteria were established: (1) all contributions must be non-negative, since negative contributions have no physical interpretation; and (2) the 'best' solution is the one which minimizes the residual (i.e. the sum, for all minerals or elements, of the absolute value of the difference between the calculated and the known abundance). As discussed by Heath & Pisias (1979), a major unknown in linear programming is the 'stability' of the calculated contributions, i.e. how the contributions respond to changes in the initial abundance data. If the contributions change rapidly with small variations in the initial data, then the modelling results must be viewed with caution. If the contributions change only slightly with small fluctuations in the initial data, then the results can be viewed with greater confidence. Because linear programming is a modelling and not a truly statistical technique, it is difficult to rigorously assign confidence intervals to the calculated contributions. The residual for these calculations (i.e. the portion of the initial data unaccounted for by the modelled composition) ranges from 10% to 15%. Therefore, it can be roughly estimated that the uncertainty associated
Source area drained CR KM KM KM KM KM KM KM KM KM CCR CCR CCR CCR
with the calculated contributions is + 15% (absolute). The uniformity of the chemical data may cause larger uncertainties for those results. Changes in end-member compositions away from each source area due to size-selective transport (Gibbs 1976b; Arcaro 1978) may also be a major source of error in the calculated contributions.
Data and results Mineral end-members (Columbia River, Klamath Mountains, and California Coast Range) for the 2-20 ~m fraction are presented in Fig. 2 (Krissek 1982). All other marine and fluvial mineral and chemical data obtained during this study are also tabulated in Krissek (1982). Material derived from the Columbia is rich in plagioclase feldspar, illite, and quartz, while California Coast Range sediments also contain chlorite; Klamath Mountain detritus includes a distinctive hornblende component. The calculated contributions are mapped in Fig. 3. The Columbia River contribution (Fig. 3(a)) is largest ( > 75%) to the north-north-west away from its mouth, with a weaker contribution southward on the Oregon margin. The southward influence has its maximum extent along the uppermost slope. The Klamath Mountain contribution (Fig. 3(b)) is strongest ( > 75%) immediately offshore from this source region (compare with Fig. 1), and decreases both downslope and alongslope. The California Coast Range contribution (Fig. 3(c)) is strongest (generally 25-50%) in a lobe-like feature extending northward along
366
L . A . Krissek
FIG. 2. Mineral composition of the discharge-weighted end-members in the 2-20/zm size fraction (from Krissek 1982). The California Coast Range material is chlorite-rich, and the Klamath Mountain sediments contain a distinctive hornblende signal. the middle to lower slope as far as 46.5 ~ N., with slightly lower contributions outside this feature. The abundance patterns of plagioclase and hornblende in the 2-20 ~m fraction are shown in Fig. 4 (Krissek & Scheidegger 1983) for comparison with the linear programming results. The Columbia River is the major plagioclase source (Fig. 4(a)) and plagioclase shows a large-scale dispersal pattern to the north-west similar to that exhibited in Fig. 3(a). South of 46 ~ N., however, the linear programming results provide a clearer and more detailed picture of Columbia River influence than the single mineral abundance pattern because equivalent feldspar concentrations can be contributed to this region from other source areas. Hornblende (Fig. 4(b)) is a distinctive tracer of the Klamath Mountain material, and provides an excellent indicator of Klamath Mountain influence on the upper slope close to the source (compare to Fig. 3(b)). Because hornblende is unique to the Klamath Mountains, its abundance provides a relatively detailed and sensitive indicator of the Klamath Mountain influence even at greater distances from the source. These two examples indicate that single mineral abundance patterns can provide relatively good indicators of source area influence when a single source dominates the sediment. In distal environments material from several sources is mixed, and only unique mineral tracers (e.g. hornblende for the Klamath Mountains) can indicate source area contributions. Non-unique tracers (e.g. feldspar) are poor indicators of source area influence in distal environments. In
areas where sediment is supplied by several sources (as on the central Oregon slope), the multivariate approach of linear programming provides better indicators of source area contributions than the univariate approach of mineral abundance patterns. The chemical compositions of the three major end-members in the 2-20 ~m fraction are listed in Table 2, and abundances for the Klamath Mountain and California Coast Range sources are normalized to the Columbia River source and plotted in Fig. 5. The two southerly sources show similar patterns relative to the Columbia River abundances with enrichment in Mg and Fe and depletion in Ca, Ti, and Sr. Mean elemental abundances for the 2-20 ~tm fraction of the marine samples are also given in Table 2, and the small standard deviations about these mean values indicate that the slope sediments have relatively uniform compositions throughout the study area. Results of a Q-mode factor analysis with elements normalized to equal means but before varimax rotation (Leinen & Pisias 1982) support this conclusion. The first principal factor identified by the analysis indicates that 96.6~ of the data variance is explained by the mean chemical composition. Only 3.4% of the variance is associated with changes from that mean value. As a result of subsequent varimax rotation in the Q-mode factor analysis, six elements were identified as best explaining the variability in the data set (Leinen & Pisias 1982). These elements were Fe, Ca, K, Mg, Ti, and Sr. The factor analysis also identified only two terrigenous-
Fine-grained sediment contributions on the Oregon and Washington slope 367
o
o
0
,.o +.a
0
o.~
2~e~
3~ ,
0
=.0
6~
FIG. 4. Mineral abundances in the 2-20/~m fraction of the slope sediments (from Krissek & Scheidegger 1983). (a) Feldspar abundances. Note similar pattern for the modelled Columbia River contributions (Fig. 3(a)). (b) Hornblende abundances. Note similar pattern for Klamath Mountain contributions (Fig. 3(b)).
FIG. 5. Chemical composition of the discharge-weighted end-members in the 2-20/~m size fraction normalized to the Columbia River source. Fe, Ca, K, Mg, Ti, and Sr were used in the linear programming modelling.
Fine-grained sediment contributions on the Oregon and Washington slope 369
+1
r~
+l
+~
Z~.~. N-
+l
C 3 ~
< ~
-
< @
<
Z <
+1
z~ e-i ~5
'
<
tr
~
N 9
.
" - +1
+1
z~
8~ Z~
~aN
88~'
370
L.A. Krissek
related factors in the data set. The first, which accounted for 44.5% of the data variance, was dominated by Fe and Mg and corresponded to the general composition of material derived from the southern source areas (Klamath Mountains + California Coast Range). The second factor, which accounted for 36.5% of the data variance, was dominated by Ca and Sr (probably associated with anorthite-rich plagioclase; Krissek & Scheidegger 1982), and corresponded to the general composition of material derived from the Columbia River. The third factor, which accounted for 17.9% of the data variance and was dominated by K, had highest loadings offshore and could not be related to any continental source area. Because of the results of the factor analysis and the data presented in Fig. 5, the 2-20/~m fraction was modelled by linear programming using only six elements (re, Ca, K, Mg, Ti, and Sr) and two end-members (a northern (Columbia River) and a southern (Klamath Mountains+California Coast Range) source). All elements were assigned equal weights so that the most abundant elements have the greatest influence of the model. Mean residual for the calculation was 6.5% and K consistently had the maximum residual. This reflects the importance of K in explaining the variance within the slope data set while K contents of the end-members are relatively uniform (Fig. 5). The calculated contributions show areal patterns similar to those from the mineral-calculated contributions. The Columbia River end-member is shown as an example (compare Figs 3(a) and 6). Highest Columbia River influence again occurs to the north-west of its mouth, with lower contributions extending southward on the slope. The magnitudes of the Columbia contributions calculated from the two data sets also agree reasonably well (Fig. 7). Mineral contributions tend to be higher than the chemical results, especially on the central and southern Oregon slope, but the reason for this is unknown. It is speculated that this difference arises because a Columbia-type mineralogy can be obtained by diluting the southern source mineralogies with some non-quantified component (opal, carbonate, or locally-important Oregon Coast Range material). The Columbia contribution may therefore be exaggerated whenever a nonquantified mineral component is present. Alternatively, the uniformity of the chemical data may cause those results to respond much more quickly and erratically to compositional changes, decreasing the calculated Columbia contribution. Such explanations, however, are purely speculative.
FIG. 6. Columbia River contribution calculated by linear programming of the chemical data. Note similarity of the pattern to Figs 3(a) and 4(a).
Discussion Processes
The processes controlling the dispersal of hemipelagic sediments on the Oregon-Washington continental margin have been discussed in detail by Krissek & Scheidegger (1983). California Coast Range material is transported almost due north along the slope, and contributes more than 25% as far north as 46030' N. The Klamath Mountain component is more geographically limited and moves offshore to the north-west and then alongslope to the north. Both contributions are moved largely by an eastern boundary undercurrent. The Columbia River influence is greatest ( > 75%) to the north away from its mouth, with a lesser contribution southward. This may reflect a site-dependent and seasonally-related divergence in transport direction.
Fine-grained sediment contributions on the Oregon and Washington slope 371 IOO-
/
COLUMBIA RIVER CONTRIBUTIONS
/" ~
2 - 2 0 pm
~
,o_
/ /
/ "/
-
CALCULATED60 FROM CHEMISTRY
0
/
/./
,. %
/
/
20
40
60
80
I00
% CALCULATED
FROM
MINERALOGY
F=G. 7. Comparison of the Columbia River contributions calculated using mineral and chemical data. Line shown has slope = 1 with uncertainty estimates of _+15?/o.Equation of the regression line is given. Agreement between the mineral contribution patterns (Figs 3 and 4), especially near each source, indicate that the linear programming results are consistent with the dispersal processes proposed by Krissek & Scheidegger (1983) and will not be discussed further here.
In the 2-20 pm fraction, the mineral data account for an average of 84.3% of the sediment (Krissek & Scheidegger 1983), further supporting the conclusion that other components exert only a minor influence within this size fraction. Sediment budgets
Relationships between chemical and mineral com-
position The data set presented here is one of the first to contain both elemental and quantitative mineral abundances for hemipelagic sediments. As such, it provides an excellent opportunity to examine the relationships between the chemical and mineral composition of fine-grained marine sediments. The relationship between contributions calculated using chemical data and those calculated using mineral data in the 2-20 l~m fraction appears reasonable (Fig. 7), indicating that the elemental abundances are dominated by mineral components which have been quantified. Nonquantified components may include carbonates, opal, non-diffracting or poorly crystalline detrital or authigenic phyllosilicates~ or non-diffracting or poorly crystalline hydrogenous iron oxides and hydroxyoxides (Lisitzin 1972: Cronan 1974).
The source area contributions calculated above can be used to construct a sediment budget for the slope within the study area. In addition to the estimated uncertainty of _+15% associated with the calculated contributions, other uncertainties also affect the parameters used in the sediment budget calculation. Sedimentation rates were obtained from an assortment of radiocarbon, biostratigraphic, and tephra-chronologic data, in some cases with only one age control point per core (Table 3). Although the range of values is reasonable when compared to examples from other continental slopes (Doyle et al. 1979; DeMaster 1981; Garrison 1981; Reimers 1981), these rate estimates bring their own uncertainty into the calculation. The discontinuous and nonuniform nature of sedimentation on the slope in both space and time (Kulm & Scheidegger 1979; Krissek & Scheidegger 1983), also introduces uncertainties into the budget results. From
L.A. Krissek
372
TABLE 3. Sedimentation rates on the Oregon and Washington continental slope Sedimentation Rate (cm/10 3 yr)
Number of rates available
Washington Upper Slope Washington Lower Slope
10-44 17-43
5 5
Barnard (1978) Barnard (1978)
Northern Oregon Upper Slope Northern Oregon Lower Slope
30-40 53
0 1
Estimated Kulm and Scheidegger (1979)
Central Oregon Upper Slope Central Oregon Lower Slope
20-30 25-50
0 0
Estimated Estimated
Southern Oregon Upper Slope Southern Oregon Lower Slope
l0 20-65
3 3
Kulm and Scheidegger (1979) Kulm and Scheidegger (1979)
Province
several 3.5 kHz seismic profiles, it can be estimated that as little as 50~o of the upper slope and 30~ of the lower slope may be subject to deposition. If these estimates are valid for the entire slope, the results of the budget calculation will overestimate the true total sediment accumulation by a factor of 2 to 3. The unit mass accumulation rate (g/cmZ/yr) used in the calculation may also vary both spatially and temporally with sediment grain-size and depositional process. Finally, the fluvial suspended sediment discharge data (Karlin 1980) used to estimate the effectiveness of the slope as a trap for fine-grained sediments are also subject to some question. These values are based on discharge records from two to nineteen years long, and may not include either geologically-significant very large-discharge events or anthropogenic effects. As a result of all these uncertainties, the 2-20 #m sediment budget presented here represents the best effort possible with the present data, but must be recognized as only an initial attempt. Available sedimentation rates were compiled from Barnard (1978) and Kulm & Scheidegger (1979). Rates for the northern Oregon upper and central Oregon upper and lower slopes were estimated by considering the rates known for adjacent environments at similar water depths and the location of each region relative to the major sediment sources. The rates are listed in Table 3. Mean wet bulk density of the surface sediments was 1.52 g/cm 3, and mean water content (weight of water/weight of solids) was 114~ (G.H. Keller, unpublished data 1982). These values yield a mean dry bulk density of the surface sediments of 0.78 g/cm 3. The accumulation of material from each source on the slope was then calculated by
Mi=
2 area of slope
S'pj'(D)'Xi'A
Reference
where M,-= total mass accumulating on the slope per year from source area i, S = sedimentation rate on the slope, pa = dry bulk density, D = weight fraction of the sediment in the size range of interest, Xi = contribution (weight fraction) from the source area i, and A = area of the slope with sedimentation rate S and source area contribution X~. The slope sediments contain an average of 40.5~ 2-20/~m material. The results of the budget calculation are given in Table 4. Contributions from only two sources to the 2-20/~m fraction can be calculated with the chemical data. These results show good agreement of the total mass contributed in the 2-20/~m fraction as calculated using the mineral and the chemical data, and suggest that the source areas decrease in importance from north to south (i.e. in the order Columbia River, Klamath Mountains, California Coast Range). When grouped as a southern vs. northern source, the combined influence of the Klamath Mountain and California Coast Range material is slightly greater than that of the Columbia River. The centralized location of the Columbia River within the study area and the dispersal mechanisms which control the transport of Columbia River sediments (Krissek & Scheidegger 1983) combine to produce a strong Columbia River influence on the Washington and the central and northern Oregon continental slopes. Material from the southern sources must be transported further to be included within the study area boundaries, thereby limiting the Klamath Mountain and California Coast Range influence on the Oregon-Washington continental slope. The estimates of fluvial input within the 2-20
Fine-grained sediment contributions on the Oregon and Washington slope 373 TABLE 4. Source area contributions to slope sediments, estimated discharge of suspended sediment from each source area, and estimated effectiveness of the continental slope as a trap for fine-grained sediments derived from each source area Contributions (1012 g/yr.) to the slope sediments % Retained on the slope Source area
Mineral results
Chemical results
Estimated fluvial input
2-20/~m Fraction Columbia River Klamath Mountains California Coast Range
1.53-3.05 1.21-1.94 0.72-1.21
1.52-2.69
6.01 4.22 10.45
1.93-3.21
~m size fraction (Table 4) were calculated using the annual suspended sediment discharge data of Karlin (1980) and the suspended sediment grainsize data compiled by Krissek (1982). Approximately 50% of the 2-20 /~m fluvial material discharged annually into the Pacific Ocean from streams within the study area is derived from the California Coast Range; only 7-12% of the California Coast Range input can be accounted for in the slope sediments. The remainder evidently either never enters the Oregon-Washington slope system or bypasses the system and is deposited in other areas (e.g. the Delgada Fan; Hein 1973). The Klamath Mountains contribute approximately 20% of the annual input of 2-20 #m material, and 28-46% of the Klamath material is deposited on the slope. The Columbia River contributes approximately 30% of the regional input, and 25-50% of this material remains within the system. It is interesting that both the Columbia River and the Klamath Mountains have between 25% and 50% of their 2-20 #m material trapped on the slope, since these two source areas lie predominantly or wholly within the boundaries of the study area. The California Coast Range is a more distant source, and its resultant influence on the slope (relative to its potential contribution) is smaller. The results of the 2-20/~m source area contribution and sediment budget calculations provide excellent examples of the formation of hemipelagic sediments from multiple sources. Finegrained sediment studies in the past (Biscaye 1965; Griffin et al. 1968; Kolla & Biscaye 1973; Kolla et al. 1976) have emphasized the compositional changes which occur away from a single source without providing information on the material which acts to dilute that signal. In the case of the Washington-Oregon continental slope, multiple sources are identifiable in the chemical and mineral data, and the terrigenous
Mineral results
25.4-50.7 28.7-46.0 6.9-11.6
Chemical results
25.3-44.8 13.2-30.7
components dominate the sediments. Therefore it is possible to describe the slope sediments as mixtures of material from the various source areas. In areas where a single source dominates the sediments (e.g. Columbia River influence on the Washington margin in the 2-20 ~m fraction) the consideration of abundance changes away from the source are probably sufficient to describe the processes of sediment formation. In cases where several sources contribute equally to the sediment (e.g. central Oregon slope in the 2-20 #m fraction), the consideration of single mineral abundance changes away from a source provide a limited picture of the processes of sediment formation. That picture may be determined by the mineral which is considered, and may describe only a small portion of the resultant sediments. Therefore, whenever possible, the supply of sediment from several source areas should be investigated, especially in continental margin environments. Such investigations demand multivariate data sets and source area samples whenever possible, but describe the processes of sediment formation much more fully than the single source approach.
Summary The application of linear programming to quantitative mineral and elemental abundances from each sample has made it possible to uniquely address several basic questions related to finegrained sedimentation in the marine environment. First, the effects of mixing sediment derived from several source areas on the formation of hemipelagic sediments were investigated and found to be especially important in regions of the slope away from a major source. Second, the mineral and chemical data were modelled independently, and their results indicate that the mineral data set provides more distinctive sedi-
374
L.A. Krissek
ment tracers. Finally, the source area contributions calculated during the modelling were used to construct a budget for the 2 - 2 0 / a m sediments on the slope. This budget indicates that proximal sources contribute m o r e sediment to the slope than distal sources, and that the slope traps a higher p r o p o r t i o n of the material discharged each year from proximal than from distal sources. ACKNOWLEDGEMENTS; The a u t h o r wishes to thank M a r k E. H o w e r for assistance with the
mineral and chemical analyses used in this study. Chi Meredith and Nicklas Pisias provided patient and valuable introductions to the world of linear p r o g r a m m i n g and factor analysis. K e n Scheidegger, Ross Heath, and Jack D y m o n d reviewed the manuscript and provided constructive critiques t h r o u g h o u t its m a n y stages. This work was funded by the N a t i o n a l Science F o u n d a t i o n under grant OCE-7819825 and by the N S F / Ocean Margin Drilling P r o g r a m u n d e r JOI Subcontract 26-81.
References ARCARO, N.P. 1978. The Control o( Lutite Mineralogy by Selective Transport, Late Pleistocene and Holocene Sediments of Northern Cascadia Basin-Juan de Fuca Abyssal Plain (Northeastern Pacific Ocean): A Test of Clay Mineral Size Dependency. Unpubl. M.S. Thesis, Lehigh University. BAKER, E.T. 1973. Distribution and composition of suspended sediment in the bottom waters of the Washington continental shelf and slope. J. sed. Petrol., 43, 812-21. BARNARD, W.D. 1978. The Washington continental slope: Quaternary tectonics and sedimentation. Marine Geol., 27, 79-114. BISCAYE, P.E. 1965. Mineralogy and sedimentation of recent deep-sea clay in the Atlantic Ocean and adjacent seas and oceans. Bull. geol. Soc. Am., 76, 803-32. COBLER, R. & DYMOND, J. 1980. Sediment trap experiment on the Galapagos spreading center, equatorial Pacific. Science, 209, 801-3. CRONAN, D.S. 1974. Authigenic minerals in deep-sea sediments. In: Goldberg, E.D. (ed.), The Sea, 5. Wiley-Interscience, New York. 491-525. DEMASTER,D.J. 1981. The supply and accumulation of silica in the marine environment. Geochim. cosmochim. Acta, 45, 1715-32. DOYLE, L.J., PmKEY,O.H. & WOO, C.C. 1979. Sedimentation on the eastern United States continental slope. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleont. Min. Spec. Pub. Tulsa. 27, 119-29. DUNCAN,J.R., KULM, L.D. & GRIGGS, G.B. 1970. Clay mineral composition of Late Pleistocene and Holocene sediments of Cascadia Basin, northeastern Pacific Ocean. J. Geol., 78, 213-21. DVMOND, J. & EKLUND,W. 1978. A microprobe study of metalliferous sediment components. Earth planet. Sci. Lett., 40, 243-51. , CORLlSS, J.B., COBLER, R., MURATLI, C.M., CHOU, C. & CONARD, R. 1980. Composition and origin of sediments recovered by deep drilling of sediment mounds, Galapagos spreading center. In: Rosendahl B.R., Hekinian et al. (eds), lnit. Repts. DSDP, 54. US Govt Print. Off., Washington, DC. 377-85. - 1981. The geochemistry of Nazca Plate surface sediments: an evaluation ofhydrothermal, biogenic, detrital and hydrogenous sources. In: Kulm, L.D. et
al. (eds), Nazca Plate: Crustal Formation and Andean Convergence. Geol. Soc. Am. Memoir, 154, Boulder, Colorado. 133-74. GARRISON, R.E. 1981. Pelagic and hemipelagic sedimentation in active margin basins. In: Douglas, R.E., Colburn, I.P. & Gorsline, D.S. (eds), Depositional Systems of Active Continental Margin Basins (short course notes), Pacific Section, Soc. econ. Paleont. Mini., Los Angeles. 15-38. GIBBS, R.J. 1967a. Quantitative X-ray diffraction analysis using clay mineral standards extracted from the samples to be analyzed. Clay Minerals, 7, 79-90. -1967b. The geochemistry of the Amazon River system: Part I. The factors that control the salinity and the composition and concentration of the suspended solids. Bull. geol. Soc. Am., 78, 1203-32. GRIFFIN, J., WINDOM, H. & GOLDBERG,E.D. 1968. The distribution of clay minerals in the world ocean. Deep Sea Res., 15, 433-59. GRIGGS, G.B. & HEIN, J.R. 1980. Sources, dispersal, and clay mineral composition of fine-grained sediment off central and northern California. J. Geol., 8 8 , 541-66. HEATH, G.R. & DYMOND,J. 1977. Genesis and transformation of metalliferous sediments from the East Pacific Rise, Bauer Deep, and Central Basin, northwest Nazca plate. Bull. geol. Soc. Am., 8 8 , 723-33. - & 1981. Metalliferous sediment deposition in time and space: East Pacific Rise and Bauer Basin, northern Nazca Plate. In: Kulm, L.D. et al. (eds), Nazca Plate: Crustal Formation and Andean Convergence, Geol. Soc. Am. Memoir, 154, Boulder, Colorado. 175-98. - - & PISIAS,N.G. 1979. A method for the quantitative estimation of clay minerals in North Pacific deep-sea sediments. Clay Minerals, 27, 175-84. HEIN, J.R. 1973. Increasing rate of movement with time between California and the Pacific Plate: From Delgada submarine fan source areas. J. geophys. Res., 78, 7752-62. KARLIN, R. 1980. Sediment sources and clay mineral abundances off the Oregon coast. J. sed. Petrol., 50, 543-60. KOLLA, V. & BISCAVE,P.E. 1973. Clay mineralogy and sedimentation in the East Indian Ocean. Deep Sea Res., 20, 727-38. --, HENDERSON, L. & BISCAYE, P.E. 1976. Clay
Fine-grained sediment contributions on the Oregon and Washington slope 375 mineralogy and sedimentation in the west Indian Ocean. Deep Sea Res., 23, 949-62. KRISSEK, L.A. 1982. Sources, Dispersal and Contributions of Fine-Grained Terrigenous Sediments on the Oregon and Washington Continental Slope. Unpubl. Ph.D. Thesis, Oregon State Univ. , SCHEIDEGGER, K.F. & KULM, L.D. 1980. Surface sediments of the Peru-Chile continental margin and the Nazca Plate. Bull. geol. Soc. Am., Part 1, 91, 321-31. -& -1983. The sedimentological significance of an eastern boundary undercurrent and its role in continental margin sedimentation. Marine Geol., in press. KULM, L.D. & SCHEIDEGGER, K.F. 1979. Quaternary sedimentation on the tectonically active Oregon continental slope. In: Doyle, L.F. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleo. Mini. Spec. Pub. Tulsa, 27, 247-63. LE1NEN, M. & PISIAS, N. 1982. An objective technique for determining end-member compositions and for
partitioning sediments according to their sources. Geochim. cosmochim. Acta (submitted). LISITZEN, A.P. 1972. Sedimentation in the world ocean. Spee. Publ. Soc. econ. Paleont. Min. Tulsa, 17, 218 PP. OLMSTEAD, D.L. 1972. Clay Mineralogy of the Washington Continental Slope. Unpubl. M.S. Thesis, Univ. of Washington. PINET, P.R. & MORGAN~W.P., JR. 1979. Implication of clay-provenance studies in two Georgia estuaries. J. sed. Petrol., 49, 575-80. REIMERS, C.E. 1981. Sedimentary Organic Matter: Distribution and Alteration Processes in the Coastal Upwelling Region Off Peru. Unpubl. Ph.D. Thesis, Oregon State University. RUSSELL, K.L. 1967. Clay Mineral Origin and Distribution on Astoria Fan. Unpubl. M.S. Thesis, Oregon State University. SPIGAI, J.J. 1971. Marine Geology of the Continental Margin off Southern Oregon. Unpubl. Ph.D. Thesis, Oregon State University.
L.A. KRISSEK, Department of Geology and Mineralogy, The Ohio State University, Columbus, Ohio 43210, USA.
Basin model for hemipelagic sedimentation in a tectonically active continental margin: Santa Barbara Basin, California Continental Borderland S.E. Thornton SUMMARY: A variety of Holocene sedimentary processes and depositional environments in Santa Barbara Basin may serve as a model for terrigenously-dominated hemipelagic basins devoid of submarine canyons: (1) suspensate transport; (2) turbidity current transport; (3) mass movement on the lower slope; (4) the imprint of current patterns on both fine-grained suspended load and bedload transport; and (5) preservation of resulting stratigraphy on the deep basin floor by anoxic water conditions which result in annual varves plus turbidites and flood suspensate layers. Silt patterns define a river-derived suspensate pathway driven by surface currents, which coincides with the distribution of 1969 flood layer sediments. A 'hemipelagic core' is present on the western portion of the basin floor, where terrigenous and biogenous input are most balanced. Mass-movement covers 13~ of the basin-slopes and is typified by mud-flows interpreted from seismic profiles and characteristic sedimentary structures in cores. A generalized vertical sequence of sedimentary structures derived from radiograph analysis may serve as a model for mud-flow deposits. Turbidites and flood suspensate layers have been traced from the middle slope to the basin floor on a canyonless slope and their distribution is controlled by surface current delivery of suspensates. Vertical sequences of turbidite sedimentary structures in grey-layer turbidites provide support for Piper's (1978) model for fine-grained turbidites. Varve counting in the central basin yields a sedimentation rate of 173 cm/1000 years and frequencies of 59 years for all grey-layers (flood + turbidite), and 120 years for grey-layer turbidites, which probably reflect roughly 50 and 100 year frequency floods. Fine-grained sedimentary processes in hemipelagic settings with high sedimentation rates and low slope gradients may be typified by: (1) mud-flows on the lower slope; (2) unchannelized turbidity currents and flood layer deposition controlled by surface currents and proximity to source; (3) laminated sediments in the deep basin if oxygen levels are sufficiently low, or preservation of only flood suspensate layers and turbidites if bottom waters are not anoxic; (4) sediment pathways definable from texture and coarse fraction analysis. Studies of sedimentary processes on continental margins have long concentrated on the turbidite canyon-fan-abyssal plain system (Gorsline 1978). As a consequence, research has focused on the predominantly sandy depositional system to the neglect of fine-grained marine sedimentary processes. In particular, continental margin slopes between submarine canyons or on slopes devoid of submarine canyons are pathways through which terrigenous fine-grained sediments travel to the deep sea, and deserve more study (Gorsline 1978). A variety of sedimentary processes can be anticipated to be important on fine-grained intercanyon slopes. Turbidity current transport need not be restricted to submarine canyons, and should be an important transporting process. M a s s - m o v e m e n t processes, particularly debrisflows, have been found on nearly all submarine slopes which have been studied (Embley 1976, 1980; Embley & Jacobi 1977; Embley & Morley 1980; Stanley & M a l d o n a d o 1981). As an example of the extent of mass-movement, Embley (1980) has estimated that at least 4 0 ~ of the
continental rise off eastern N o r t h America is covered by a veneer of m a s s - m o v e m e n t deposits. Because continental rise sediments represent a significant percentage of deep ocean sediments (McCave 1972), m a s s - m o v e m e n t must be considered a very important process in downslope transport on the lower slope and rise. Massm o v e m e n t features are found on all continental and insular slopes in the California Continental Borderland that have been studied ( N o m a r k 1974; Haner & Gorsline 1978; Crissman & Ploessel 1979; Field & Clark 1979; Nardin et al. 1979a; T h o r n t o n & Crissman 1979; Field & Edwards 1980). Consequently, m a s s - m o v e m e n t processes are an important sediment transport process on all submarine slopes and probably are important on fine-grained slopes. The specific types of m a s s - m o v e m e n t will depend on a variety of factors, including slope declivity, sedimentation rates, geotechnical factors such as consolidation state and shear strength, and triggering mechanisms (Field 1981). Because of the abundance of seismicity caused by transform margin tectonics in the California 377
378
S.E. Thornton
FIG. 1. Bathymetric chart of Santa Barbara Basin with Index map of the basin's location in the California Borderland. Depth in metres. Borderland and Santa Barbara Basin specifically (Lee et al. 1979; Yeats 1981), abundant triggering mechanisms are present to initiate slope failure. Current transport of fine-grained sediments should be detectable by textural patterns where a knowledge of current circulation exists. Tectonics (folding and faulting) and bioturbation would be forces that rework and modify already-deposited fine-grained sediments. In order to analyse the relative importance, specific dispersal pathways and depositional patterns of transport of fine-grained sediments across intercanyon or canyonless slopes to the basin floor, a single basin currently being infilled by fine-grained, hemipelagic sediments (compositional term implying mixture of lithogenic and biogenic fine sediments in the sense of Kuenen 1950) has been selected as a case study of
•[•.....•.......
9 .......
I.
i "'"
9 9
~1~
9
9
9 Ae
sedimentary processes acting on fine-grained clastics, Santa Barbara Basin (Fig. 1). Sixty-eight box cores, 25 piston cores (Fig. 2) and 600 km of seismic records were collected. To supplement this data base, 16 previously collected piston cores, 186 previously-collected box cores and approximately 2500 km of seismic data from McClelland Engineers and T. R. Nardin of the University of Southern California were also examined (line locations in Thornton 1981a). High-resolution seismic records were collected with a 3.5 kHz minisparker yielding vertical resolution of 0.75 m and penetration up to 50 msec two-way travel time (about 40 m). Subsampling of box cores followed the techniques of Edwards (1979). Point counts of sand constituents for the surface (0-2 cm) samples was accomplished by wet-sieving on a 63 micron sieve,
9 "...
Aoee ~A
.
-~o
A ~ AA
.
.
,
.
" " 9 9
9
~
:.
"
FXG. 2. Locations of box cores (circles) and piston cores (triangles) with the location of piston core 27875 referred to in the text indicated with an arrow. Box cores from supplementary data bases not shown.
Hemipelagic sedimentation in a tectonically active continental margin oven drying at 90~C and point counting by the grid method of at least 300 grains (Galehouse 1971a) on a micropalaeontological picking tray using a camel's hair brush for examination. Textural analysis was performed for 78 surface samples from box cores, 150 samples from 8 selected box cores for down-core analysis and 521 samples at selected intervals in 14 piston cores employing pipette analysis modified from the method of Galehouse (1971b). Piston core extrusion for subsampling and X-radiography was performed with great care to prevent disturbance of the sedimentary structures to produce a 1 cm thick slab from the centre of the cylindrical sample (Thornton 1981). X-radiography was performed on piston and box core slabs following the techniques and interpretive guidelines thoroughly covered by Bouma (1964, 1969) and avoiding interpretive pitfalls (Stow & Aksu 1978). Previously analysed surface samples used to supplement the data base were drawn from Day (1979), Gatto (1970) and Edwards & Gorsline (1978). All surface sample values for textural, organic carbon and calcium carbonate contents were entered into a computer data base and contoured using the SYMAP program (Dougenick & Sheehan 1977) with subsequent handsmoothing to reduce orthogonality in contours which is a result of contouring algorithm. Computer contouring was compared with previously contoured maps for similarity, which was quite close.
Setting and previous studies Santa Barbara Basin is the northernmost basin of the California Continental Borderland (Fig. l), a tectonic chequerboard of ridges and basins (Crowell 1974) with diverse sedimentary processes (Emery 1960; Gorsline 1978, 1980, 1981). Except for two very small submarine canyons on the western landward side of the basin (Fig. 1), the study area is devoid of submarine canyons. The basin lies between the mainland on the north, the Northern Channel Islands on the south, a sill on the west at 470 m and the Hueneme Sill on the east at 250 m. Elongate in an east-west direction due to the north-south compressional tectonics of the area (Yeats 1981), the basin has its deepest point, the basin centre, offset to the west at a maximum water depth of 590 m. A west-trending ridge, the Santa Clara Ridge and a west-trending trough, the Santa Clara Trough, are both offshore extensions of onshore structure in the Ventura Basin (Fischer 1976). Slopes comprise 96% of the 2300 square kilometre basin area
379
deeper than the shelf break, so sedimentation within the basin occurs predominantly on a slope setting. The fact that the Santa Barbara Basin is silled at the west and east has important implications. The deepest sill, at 470 m water depth on the west controls the level from which deep basin water may be derived. At this level, deep basin water can only be replaced by Pacific Intermediate Water within the oxygen minimum zone. This deep basin water, already low in oxygen, is further depleted of oxygen by bacterial uptake, to produce an area with oxygen contents of 0.1 ml/l or less, located approximately at the basin centre in water depths greater than 550 m (Emery & Hulsemann 1962). Because of the seasonality of rainfall and terrigenous sediment delivery to the basin, sediments cored in this anoxic zone are laminated (Hulsemann & Emery 1961; Soutar & Crill 1977). Emery (1960) was the first to recognize that these laminations were annual and could thus be termed non-glacial varves. More specifically, they should be termed post-glacial varves, ones caused by seasonal input of river-derived sediment (M6rner 1978). Two laminae constitute a varve, the darker consisting of detrital clay and silt transported by winter rains (November-March) and a lighter-coloured, diatom-rich lamina representing planktic productivity in the remainder of the year (Hulsemann & Emery 1961). Lamination preserved in this hemipelagic basin is thus caused by seasonal rainfall patterns and results from a combination of suspensate transport and settling through the water column. Processes thought to produce lamination in the deep sea are considerably different (Stow & Bowen 1978, 1980). Emery (1960) suggested that varve thicknesses might correlate with fluctuations in past rainfall and tree rings, a contention later tested by Soutar & Crill (1977). Four types of strata were recognized in the deep Santa Barbara Basin floor: (1) laminae; (2) disturbed laminae; (3) homogeneous sediment; and (4) turbidites (Hulsemann & Emery 1961). Grey silt and clay layers with fining-upwards textural patterns were interpreted as turbidites and correlated across the basin centre (Hulsemann & Emery 1961). Because the thickness and percentage of grey layers increase towards the north, a northern source was inferred (Hulsemann & Emery 1961). Using a set of very carefully collected, vented box core samples, Soutar and co-workers demonstrated a close correspondence between Pb-210 dates and varve counts (Koide et al. 1972; Soutar & Crill 1977). They found a mat-forming organism within the laminated zone which was partly responsible for density differences between
380
S.E. Thornton
dark 'winter' laminae and light 'summer' laminae, and established a statistical correlation between regional rainfall and varve thickness (Soutar & Crill 1977). Fish scales found in the same box cores reflect known historical estimates of fish populations and allowed estimates of fish biomass in the past (Soutar & Isaacs 1974). The palaeoclimatic information in the varved sediments has been utilized to study radiolaria (Pisias 1978, 1979), pollen assemblages (Heusser 1978) and organic carbon (Heath et al. 1977). Sediment trap experiments in this basin have revealed the role of pelleting in sedimentation and the fluxes of radioisotopes. Dunbar & Berger (1981) showed that faecal pellets constituted more than 60% and as much as 90% of trap-collected material, and that pellet transport can account for approximately one-half of the sediment flux to the basin floor, including the terrigenous supply. It should be noted, however, that the traps were deployed during the upwelling period (Pirie & Steller 1977), so biological activity within the surface mixed layer would have been enhanced and pelleting might well have been much greater than the remainder of the rainy season. Results from isotope studies of trap material are best explained by intensified scavenging and pelleting by organisms and by initial deposition of sediment on the shelf before resuspension and transport to the basin (Moore et al. 1981). The floods of 1969 transported a great deal of suspended sediment into the basin from the river point source (Ventura and Santa Clara Rivers) in the surface mixed layer, along density interfaces within the water column and along the bottom as a bottom nepheloid-layer (Drake 1971: Drake et al. 1972; Drake 1972). Flood sediment initially deposited on the shelf between Santa Barbara and Oxnard was subsequently resuspended by wave and current action to be transported into the deep basin as a measurable flood layer of reddish oxidized material (Drake et al. 1972). Similarities in the mineralogy of the grey-layers within piston cores of the central basin led Fleischer (1972) to conclude that they were flood layers, not turbidites. However, the source of sediment incorporated into turbidites would be the same as that of flood material, so similarity of mineralogy does not preclude a turbidite origin for the grey-layers (E. D. Drake, pers. comm.). To date, no single study has synthesized the basin-wide sedimentary processes to give a complete knowledge of transport pathways and processes. Rather, palaeoclimatic studies and sediment trap experiments have assumed that an anoxic, laminated central portion of the basin can be studied without regard to slope processes
except for suspended sediment transport. Because of the prevalence of sediment creep and other forms of mass-movement on the basin-slope (this study and Crissman & Ploessel 1979; Hein & Thornton 1979; Thornton & Crissman 1979; Richmond 1980; Thornton 1980, 1981a) this may have been a critical shortcoming of previous work. This study fills in this gap in previous studies: slope processes. In addition, sediment pathways for the whole basin can be delineated which reflect current transport. Combined knowledge of these slope processes and transport pathways will then lend further insight into the significance of palaeoclimatic studies.
Current transport, sources and sinks of fine-grained hemipelagic sediment To understand depositional processes in the basin as a whole in order to develop a basin model in a canyonless, transform margin setting, one must first synthesize present-day transport pathways which have likely persisted for approximately 10 000 years through the Holocene in the California Borderland. The surface samples used in this study truly reflect the present processes, because they represent very carefully collected 0-2 cm intervals in vented box cores, not subject to washing out of the fines. Computer-assisted contouring of percentage sand (Fig. 3) demonstrates that most of the basin deeper than 200 m contains less than 10% sand. As the shelf break for this basin lies at between 100-130 m, it is obvious that most of the sand is presently being trapped on the shelves, except in the area of the Hueneme Sill where sand constitutes 10-95~ of the bulk sediment. This anomaly can be readily explained by current winnowing of the fines, as the Anacapa Current moves across the sill to the west-north-west towards Coal Oil Point as a 200 m thick surface current, with the fastest current velocities recorded in the California Borderland of 25 cm/sec (Kolpack 1971; Drake et al. 1972; Pirie & Steller 1977; Edwards & Gorsline 1978: Thornton 1981b). This current winnowing results in the best-sorted sediment in the basin located in the area of the 90% isopleth sand along the Hueneme Sill (Fig. 3), with phi sorting values of less than 1.5 (Thornton 1981a). Bottom photography reveals asymmetric current ripples, current shadows behind organisms and larger particles and numerous current lineations (Edwards & Gorsline 1978). High-resolution seismic profiles reveal a large sand wave field covering 30 km 2 with individual waves having ampli-
Hemipelagic sedimentation in a tectonically active continental margin !'" Point I..... h c ~ ~ ' , I ~ ~
'
I
: ' . ", - ' . - " ' .
1120"W cool oil 9 9Point
~
o,,.,,r,.......
I le/o^klrq ~lii411NIO
I t~ o..
381
I
,o
ONTOUR INTERVAL I0% I
.
.
9
.
./,r_
.'-~
.
9
~t~ra
River Ventura Santa
FIG. 3. Computer-assisted contour map of percentage sand in box core tops (0-2 cm). Note the high percentage sand in the Hueneme Sill area. tudes up to 20 m and wavelengths of 150-300 m oriented at 160 ~ hugging the north slope of the Hueneme Sill (location shown later). It appears, then, that as the Anacapa Current crosses the Sill and touches the bottom, the greatest current shear would occur where the flow thins going into shallow water as it moves in a northerly direction across the sill (Fig. 4 for current location). This produces the sand wave field and winnows a greater area which results in the only sandy area of the basin. The majority of the basin floor is covered with hemipelagic sediment with less than 5"/O sand, most of which is silt (Fig. 5). This figure displays several diagnostic patterns of suspended sediment transport which are consistent with previous work. In the absence of significant sand, it appears that percentage silt is the diagnostic textural trend indicative of transport of terrigenous material in this basin, and elsewhere in the
JJ/';~Pt..'; Pt Argue~lo
'120~
California Borderland (D. S. Gorsline, pers. comm.). This observation may be great utility in other terrigenous slope and basin settings. Three sediment pathways or sinks which trace to sediment source are obvious from this silt contour and supplementary data. First of all, if one traces the 600/o silt isopleth in the eastern basin a large lineation exists, especially south of Santa Barbara within the 70% silt isopleth. This 70% silt isopleth leads directly to the major point source of terrigenous sediment within the California Borderland, the mouths of the Santa Clara and Ventura Rivers (Thornton 1981b). Thus, it appears that the 70% silt isopleth represents the mean pathway of fine-grained sediment transport as suspended load from the river point source. The outline of the 70% silt isopleth coincides with the location of the Anacapa Current (Fig. 4 and Thornton 1981 a) and also coincides closely to the mapped location of the 1969 flood layer (Drake
'
'
I ~;Pt. Conception ....... I ~ ~.......'..'.'.---.--...b.oa! uu ~'r. Santa Barbara
"qo I%
I
k
#
,
":
.V.R.
-
"<'><> ..
'~176176
FIG. 4. Surface currents in the area from Kolpack (1971).
S.E. Thornton
38z "
Point
I
on~:eptio..'.'"
....
.
.
9 .
"*
.
.__
'
\
. . . . .
"."
."
9
CoolOil
,_70....~-'~.-~~,,
.~'~;,o" iI "-~,s=~ , ~- - -' ~- - ~ X.
I
.
~.
9.,.,..........
~ L - - ~ P ~ - ; " ..... . ". . 9 o f-~/~__-.~'L~5~:~k-%;-,~L~.'..con,ouR,,rE,vA,
1%s'LT ,o to,.
i m m
o.,
~ , , o
.
I
9
" . . . . .
I "='~~.... I I
il20"W
" ' ' ' .
T,,.Z,,.x~l,~. ~
-~
-~..
,v.
3,'20_'
9
Cloro
9
r
~
,
"River
~._ "
40
50
FIG. 5. Computer-assisted contour map of percentage silt in box core tops (0-2 cm). average current trajectory of the Anacapa Current is responsible for transporting average winter sediment yield as well as flood year transport. OffPoint Conception, the 60 and 70~ isopleths delineate an area which, corroborated with current patterns (Fig. 4; Pirie & Stellar 1977; Thornton 1981b), indicates a south-eastward transport
1972) which extended beyond the nose of the 70~ silt isopleth in a south-westerly direction into the basin centre at 550 m. Furthermore, it closely follows the suspended sediment plumes (Fig. 6) following the smaller 1977-1978 floods (Thornton 198 lb). The flood layer location and the mean sediment transport direction of the fine suspensates should be roughly the same, as the same
--34~
~1~"~
Barbor,
Sonto
C oop,l_o n
Goleto
-i-I - , , ~ l
January
~
P~t
21,1978
Hueneme
~.~l~_
N
Pt Muau
.._z/
,.d~,.,,ij 9 ~"
"). ~Anacapa
9 k\
9
.
" .~.
"
9 .
'.
.~
"
I.
_
...~,,.-.-,, ". ; . < C . ~ . : - ~
.'~<-
I 9I
~
9 9 i 9
: . ,. . .. , , s0,,,
,.,~:~. ,2o~
9
.
~
9
.
9
.
~
.
.
~
.
.
9
.
9
:......,.. .::: :.-i ..i. .. i
,,9~ t
~ .
9 .... ".
.
.
.
FIG. 6. Plume originating from Santa Clara and Ventura Rivers and associated features interpreted from LANDSAT Band 5 imagery for January 21, 1978. Stippling concentration denotes relative intensity of turbidity patterns. From Thornton (1981).
Hemipelagic sedimentation in a tectonically active continental margin
trap intercalibration experiment (Dunbar & Berger 1981; Dymond et al. 1981; Moore et al. 1981). Most palaeoclimatic studies have been based on a single 10 m piston core located in the centre of the 550 m isobath and the sediment trap experiment was located just east of the piston core position (located on Fig. 8). It would appear, then, that a more appropriate position to study palaeoclimate would be in the centre of the hemipelagic core about 15 km east-south-east of the actual study spot. In this position, the true palaeoclimatic relationships would be found, as this is where the planktic influx is most important and preserved stratigraphically. To summarize, the transport directions and the location of the sink of hemipelagic sediment, Fig. 8 shows transport trajectories of river-derived silt, the current-winnowed area, the 'fan-derived' silt and the hemipelagic core. The displacement of the hemipelagic core to the west of the basin centre indicates that the overwhelming influx of terrigenous sediment from the river point source dilutes the planktic-biogenic component. Terrigenous input from the rivers, as reflected by the 70~ silt isopleth (Fig. 5) can most clearly be traced to the Ventura and Santa Clara Rivers and probably supplies roughly 90~ of the terrigenous fraction. The fan-derived silt source and a lesser influx of terrigenous material from the Northern Channel Islands accounts for the remaining 10~, in addition to coastal drainage along the northern mainland shelf (Fleisher 1972). This is in agreement with Fleischer's (1972) findings based on clay mineralogy and drainage area calculations which pointed to the Santa Clara River as the most important sediment source.
of fine sediment. Examination of suspended sediment plumes after the 1977-1978 floods shows a plume rounding Point Conception veering to the south-east in the same area as the 70~ silt isopleth (Thornton 1981b). This same plume was also detected with light transmissiometry and a thin flood layer was found after the 1969 floods in the area of the 609/0isopleth (Drake 1972). This sediment transport occurs on top of the long inactive Conception Fan which used to transport channelized turbidites into the basin centre during the Pleistocene, but has since moved to the north-west out of the Santa Barbara Basin as a result of tectonic upwarping of the western sill area (Fischer 1972, 1976), probably during the Pasadenan Orogeny 500 000 yrs BP. As a result this portion of the Conception Fan is now a suspended sediment transport area with a southeast trajectory, and only two very small submarine canyons remain whose influence is limited to the area just south-east of Point Conception (Fig. 5) where the 40~ silt isopleth veers subtly across the shelf edge. In the western portion of the basin centre is an area of 50% silt and 50~ clay, with very little sand centred around the 550 m isobath. Point counts of the little sand which exists shows that a maxima of planktic foraminifera constituting 40-50% of the sand fraction coincides with the location of the 50% silt isopleth (Fig. 7). One might term this area the 'hemipelagic core' of the basin, as terrigenous and biogenous input is most balanced in this area. The fact that this hemipelagic core is displaced to the west of the bathymetric basin centre is highly significant to the interpretation of palaeoclimatic studies which have previously been made, as well as the position of the sediment
-"'-~ -Pt. C o n c e p t i o n
~
~
9
"
"" 9 0 , 1 -
%PLANKTIC FORAMINIFERA
--:.. "~L'-.''"'.'
~
....- ""-
--
J
~
.'. ".''""-" : .
" i' ",:.. \
i
""
383
'
!
"-
O.
20
'
"- ..(
FIG. 7. Percentage planktic foraminifera in sand fractions of 0-2 cm box core samples determined by point counting. Remember sand constitutes less than 5% of the entire sediment in the area where greater than 10% of the sand consists of foraminifera.
384
S.E. Thornton
- - 2 f l O r a - _ _ :_---_~
0
IO
2Okra
-., "\ "'-
50%Silt ~'-_'~_"-~ CORE 5~176 . . . . ., ~.o . ,
-,1,_ ";2
. . . .
.
,"
.,,~.--"2.."~ ~ " ( r - ' ~
"
~
:--.
_.
....
loom-'-
ql
Santo Rosa I...,in
.
.
.
.
.
\
.
CURRENT- wIN N OW E D
i.S'N~
I::1
FIG. 8. Synthesis of sources, transport paths and transport processes in Santa Barbara Basin. Triangle to east of hemipelagic core is approximate location of core previously used for palaeoceanography and location of sediment trap deployment.
Mass movement Mass-movement on the slopes of Santa Barbara Basin is both extensive and varied in type. Figure 9 summarizes documented features of massmovement from this study, some of which had been previously reported (Crissman & Ploessel 1979; Ploessel et al. 1979). Interpretations of the type of mass-flow found are based on a combination of high-resolution seismic data (see Thornton 1981a for coverage), and sedimentary structures interpreted from radiographs of piston cores and box cores. The particular type of mass-movement found on a submarine slope can iPt
C
~
n v...~
" 9 ""
" " ""
"-,"Cqpilo% - ;~ .......
~i20 "'Naples ~ - .
Cool
depend on a number of factors (Field 1981). Here debris-flows, creep, gravity faulting and submarine slides (or disorganized slide zone; Field 1981) are all present, but debris-flows appear to be the most common. The three dimensional geometry of massmovement features can be best visualized from high resolution seismic sections. As an example of a debris-flow in this muddy basin, which should be termed a mud-flow (Nardin et al. 1979b), a typical seismic line across Flow A (Fig. 9) is shown in Fig. 10. In an area typified by an even, uniform slope surface (gradient of 1.2%, 0.56 ~) with regular, bottom-parallel acoustic reflectors,
SANTA
BARBARA
FAULTS. ~
MASS ~
"" ~
-~'~
~--.'-
"k +
~ 9
MOVEMENT
~ ; ' A 5 (,,., 9
AnacapoI . ~
~--'~
FIG. 9. Mass-movement features, faults and sand wave field. Arrow on sand wave field denotes predominant direction of crests.
Hemipelagic sedimentation in a tectonically active continental margin an extremely hummocky scarp area leads downslope to a smooth deposit (Fig. 10). The relief of the scarp area is highly variable, between 1-4 m, and the hummocky nature is discernible from the numerous discrete hyperbolic reflectors. The mud-flow deposit is quite smooth, by contrast, reaching a thickness of 8 m at the downslope end, where upfolding of a faint reflector indicates that the moving flow dragged underlying material along with it at the base. The morphology of the mud-flow deposit is identical to longitudinal profiles of subaerial debris-flow deposits (Johnson 1970). The morphology of the scarp and deposits is quite similar to mud-flows on the Mississippi Delta slope, where mud-flows originate in distinct channels and 'ooze' downslope to coalesce and form a single larger deposit (Prior & Coleman 1978; Prior et al. 1979). Discrete sampling at points within mass-movement scars and deposits provides information on the vertical sequence of sedimentary structures which may be synthesized into an idealized vertical sequence. An X-radiograph from a typical mud-flow core form Flow F (Fig. 9) depicts matrix-supported clasts and angular deformational structures overlying a laminated section (Fig. 11). The basal laminated section may be explained as either (a) a zone of layer by layer shear at the base of this mud flow (Hampton 1975) or (b) river derived laminations of clayeysilt. As this area of the Santa Clara Ridge is not within the anoxic laminated zone, rapid deposition would be necessary to preserve the lamination. Because of the lensoid nature of the lamination and the fact that this site is not within the main pathway of flood-derived sedimentation
385
(Drake 1972)~ it seems more probable that the lamination is caused by the mechanics of mudflow deposition (Hampton 1975, 1979). It is possible to generalize on the vertical sequence of mud-flows within this basin using a data base of 12 box cores which show sedimentary structures within the uppermost 60 cm and 8 piston cores with a maximum penetration of 5 m (Fig. 12). Dewatering structures are poorly-developed in the uppermost layer of mud-flow deposits with pipes and questionable dish structures similar to those in coarser sediments (Lowe & Lo Piccolo 1974) overlying ubiquitous lensoid deformational structures which have also been observed in Mississippi delta-front mud-flow deposits (Roberts 1980). Matrix-supported fabric with lineations in the matrix indicate intraclast shear and matrix mobility typical of inferred debris-flows in sedimentary rocks (Crowell 1957; Middleton & Hampton 1976; Nardin et al. 1979b) as well as marine sediments elsewhere (Embley 1976). Convoluted bedding, high angle deformational elements, minor faulting, inverted 'V' structures, minor folds and lensoid lamination constitute the remainder of the idealized mudflow vertical sequence of structures. This scheme may be typical of mud-flows in slope or rise deposits in fine-grained settings elsewhere where conditions are right for mass-movement. The most ubiquitous characteristics in this sequence are the lensoid deformational structures and matrix-supported, rounded mud clasts. It should be noted that, while mud-flows are the most common mass-movement type in this finegrained hemipelagic setting, other types of massmovement are present (Table 1).
FIG. 10. 3.5 kHz high-resolution profile across Flow A (Fig. 9) orientated north-north-west.
386
S.E. Thornton
FIG. 11. X-radiograph of piston core from deposit of Flow F (Fig. 9). Depth in centimetres.
Haemipelagic sedimentation in a tectonically active continental margin
387
Dewatering structures (pipes and dish ? structures), lensoid deformational structures. Matrix-supported, rounded clasts, 'floating' in matrix, with evidence of interclast matrix movement and shear (lineations in matrix).
Convoluted bedding (if original bedding is present). High-angle deformational elements. Minor faulting with displacements of centimetres. High-angle deformational structures, with inverted 'v' structures. Minor folds. [
~
~
Lensoid lamination% with evidence of 'necking' or boudinage. Laminae possibly produced by layer by layer shear.
FIG. 12. Idealized vertical sequence of sedimentary structures observed in Santa Barbara Basin mud-flow deposits.
TABLE l. Characteristics of mass-movement zones and Slope Area G. Flow Area (km 2) G
D A C F H
Gradient
200
highly variable 0.06-1.8 degrees 0.13-4.0% 26 2.00 degrees, 4.4% 18 0.56 degrees, 1.2% 15 2.34 degrees, 5.2% 11 1.5 degrees, 3.3% 10 1.45 degrees, 3.2%
Average deposit thickness 20 m
5 m? 5.8 m 3 m? 5 m? 19m
E
5
1.44 degrees, 3.2%
3m
I
4
1.10 degrees, 2.40/0
1m
B
3
1.2 degrees, 2.6%
2 m?
Type sediment creep, gravity faulting, reverse faulting mud flow mud ftow mud flow mud flow disorganized slide zone slump and mud flow slump and mud flow mud flow?
388
S.E. Thornton
Stratigraphy of the laminated zone: chronology, pathways and sedimentary structures within fine-grained turbidites and flood deposits The anoxic character of bottom waters below approximately 500 m in the basin (Emery & Hulsemann 1962) tbrtunately preserves sedimentary structures exquisitely and allows good sample control for analysing fine-grained turbidites (Hulsemann & Emery 1961). A typical X-radio-
graph of a piston core within the laminated zone (Fig. 13) displays the four characteristic strata found here: (1) olive-grey annual varves, the laminated section; (2) grey-layer turbidites; (3) homogeneous grey-layer flood deposits; (4) homogeneous olive-grey layers. This classification differs slightly from that of Hulsemann & Emery ( 1961 ) due to better resolution of sedimentary structures through X-radiography. In terms of thickness percentages, this core consists of: 43/o homogeneous olive-grey layers, 26% grey layers and 70','J'olaminated (varved) olive-grey sediment. A closer look at these features (Fig. 14) shows the excellent preservation of the annual varves, the
FIG. 13. X-radiograph of piston core 27875 with depth in centimetres. Dots denote grey-layers and 'H' denotes homogeneous olive-grey layers. Location in Fig. 2.
Hemipelagic sedimentation in a tectonically active continental margin
389
FIG. 14. X-radiograph of portion of core 27875 (Fig. 2) with depth in centimetres. Dot denotes grey-layer and H denotes homogeneous olive-grey layer. fine structure of the grey-layer turbidites and a typical homogeneous olive-grey layer. Greylayers are of two types: those possessing features of fine-grained turbidites (Piper 1978) and homogeneous grey-layers. The homogeneous greylayers may well be flood layer deposits similar to those found after the 1969 floods in the same area which reduced to a grey colour with burial (Drake et al. 1972; Drake 1972; Fleischer 1972) and show very similar thicknesses after accounting for compaction. Homogeneous grey-layers were all
1.5 cm or less in thickness, which may be related to original flood layer thickness. Flood-derived grey-layers were up to 2-3 cm thick on the sediment surface in the central basin floor of the laminated zone (Drake 1972). Assuming an initial water content of 80-85% typical of those measured in this study for the 0-2 cm interval in the same area, and a water content of 50~ at 2 m sediment depth, a 2 cm thick flood layer would compress under subsequent sedimentation, dewatering and consolidation to 1.12 cm thick. This is
39 0
S.E. Thornton
well within the range of measurements of this type of grey-layer in the piston cores. Turbidite greylayers which were thick enough to subsample showed fining-upwards textural patterns with increasing sorting and decreasing silt percentages upward (Fig. 15). The fact that the varves are annual allows absolute age determination, calculation of sedimentation rate and frequencies of characteristic deposits (Table 2). An absolute sedimentation rate of 173 cm/1000 yrs for piston core 27875 (Fig. 2) is in close agreement with corrected C-14 dating (Emery 1960). It is considerably lower than the sedimentation rate of 390-400 cm/1000 yrs determined by Koide et al. (1972) which did not account for compaction and was only measured for the uppermost 40 cm. Higher on the slope in the area of terrigenous influx, sedimen-
Mean 9 6 8
IAHF 278751 %Silt K-5 Sk Sorting 60 90 -I 0 I 2
FIG. 15. Textural characteristics of grey-layers (G) homogeneous olive-grey layers (H) and laminated secretions of core 27875 (Fig. 2).
tation rates of 200-250 cm/1000 yrs would be reasonable, but no dating is possible because of bioturbation and mass-movement. Frequencies of all grey-layers, flood and turbidites, are 1/59 yrs and 1/120 yrs for grey-layer turbidites. These two frequencies likely reflect the approximately 50 and 100 yr floods which occur in the area. An outline of the grey flood and turbidite layers is shown in Fig. 16 along with the outline of the laminated zone. The grey-layers are derived from the north-east and are transported to the southwest, reflected also in an increase in thickness and frequency of grey-layers previously reported (Hulsemann & Emery 1961). The grey-layers are essentially derived from a 'line source' (Gorsline 1978) caused probably by mass failure on the upper slope. Unfortunately, seismic data quality on the upper slope was poor. Where upper slope seismic records were of adequate quality, a sediment-thin zone of slope-parallel acoustic reflectors was evident. If mass-movement on the upper slope has produced these turbidites, the scars from whence they came are not obvious. Is it then reasonable to assume the turbidites were generated by mud-flows or slumps on the upper slope? Simple volume calculations for typical grey-layer turbidites point to the scale of massmovement from which they could have been generated. Mass-movement deposits on the lower slope have a volume on the order of 1 • 109m 3 which is of the same magnitude as the thicker correlated grey-layers which cover the entire laminated zone. However, it is unlikely that the lower slope mud-flows generated significant turbidites, unlike the theoretical considerations of Hampton (1972). This is because the volume of the deposits is roughly equivalent to that of the scars, leaving no large volume discrepancy unaccounted for. The long-term sediment delivery pathway from the riverine point source (Fig. 8) causes the sediment build-up which causes mass failure and turbidity current generation. The outline of the laminated zone, by contrast, is controlled by oxygen content of the bottom water, with the outline of the zone corresponding quite closely to the isopleth of less than 0.1 ml/l oxygen (Emery & Hulsemann 1962). Considerable insight can be gained by correlating the fine-grained turbidites across the laminated zone and classifying their sequence of structures in light of Piper's (1978) synthesis. Unfortunately, bioturbation and sediment creep make analysis of sedimentary structures in the area north-east of the laminated zone (Fig. 16) impossible, so one is only looking at the most distal portion of the flow within the laminated zone. Several different combinations of Piper's (1978) three zones can be found (Fig. 17). Of the
Hemipelagic sedimentation in a tectonically active continental margin
TABLE 2. Absolute
chronology ofgreyturbidite layers, greyfloodlayersandolive-greyturbidite
layers. Year
Type*
1925 1892 1795 1755 1742 1626 1622 1551 1539 1500 1434 1273 1215 1121 1049 908 883 811 803 789 780 775
ot
'-" . ' . ' .
39I
"." " ."" " ' " '
Year
ff f f t? ot f t f f t ot f t ~ f t ot t f f t t
745 741 718 666 649 616 605 603 599 547 477 397 382 334 309 256 136 l l l A.D. 115 B.C. 120 B.C. 183 B.C. 303 B.C.
Type* f f t f f t f f t t t t f t ot f t f t t t core base
* types: t = grey layer turbidite; f = grey flood layer; -- olive-grey turbidite
""""
"'" " 9 ". " . ' .
~t", L,~onceptlon , + ' - . .~.~,?y - "% Laminated~
" 70 w
I
.
.Santa~Barbara. .
u y l ~ , ~ 9
~
I
I .........
0
~"-'
10...nmi
~+
'
2b
'kin
ClaraR....~...
FIG. 16. Outline o f grey-layer locations and laminated zone with piston core locations as boxes. L a m i n a t e d zone outline is also based on box cores (location not shown here).
392
S.E. Thornton
Summary E2
-
'
El
A '.
'
"
EB El
B
C
E
F
Dc~
E~
El D
F1G. 17. Sedimentary structures in fine-grained turbidites observed in Santa Barbara Basin using Piper's (1978) terminology. grey-layers thicker than 1.5 cm, 80% had laminations varying in thickness from 1~, mm. The E1 or laminated interval consists of evenly spaced laminae or laminae decreasing in thickness upwards. At some levels, two sequences of upward-thinning laminae are found. The E2 or graded unit rarely shows laminated character in Santa Barbara Basin, yet shows textural finingupwards. The E3 or ungraded unit often contains floating foraminiferal tests which do not appear to have been implanted by burrowing. There is obviously a wide variability in turbidite structures in these clayey-silts and it is possible that we are seeing different portions of the flow. However, the hard evidence is that different sedimentary structures are present which should be explainable by fluid flow and the characteristics of deposition from the flow. Actual mechanics of fine-grained turbidite flow may follow the model of Stow & Bowen (1978, 1980) with the exception that flows are unchannelized by submarine canyons.
It appears that several characteristics of this fine-grained, hemipelagic basin may be more universally applicable to a basin model. Actual mean trajectories of terrigenous transport into the basin by way of suspended sediment transport may be discernible if the basin is sufficiently near the source. Perhaps a 'hemipelagic core' for a basin or a "hemipelagic line' for a basin-free margin may be definable where biogenic and terrigenous input is most balanced. Mud-flows may be the most commonly expected mass-movement feature and would be expected to be more frequent in areas with greater tectonic activity, as is the case with the California Continental Borderland. Other mass-movement features, such as creep, slumping and organized sliding should also be present. Unchannelized turbidity current deposits from a line source on the upper slope bring home a final point: submarine canyons are not a necessao, feature for transport of finegrained turbidity currents, but if present, might certainly channel the turbidity current. Turbidite frequency should reflect the effect of climate in delivering the sediment load to the basin entry point (Pilkey et al. 1980). ACKNOWLEDGEMENTS: This paper resulted from a dissertation under the direction of D. S. Gorsline. Critical reviews by S. K. Chough, D. J. W. Piper and D. J. Stanley are gratefully acknowledged. Financial support is gratefully acknowledged from the Tenneco Foundation, USC Sea Grant Program and N.S.F. grants OCE 76-00156, OCE 78-20453 and OCE 80-20290 to Gorsline, Douglas and Ku. Financial support for acquisition of L A N D S A T images was provided by the Department of Geological Sciences Graduate Research Fund.
References BOUMA, A.H. 1964. Notes on X-ray interpretation of
Islands Platform, Southern Cal~fi~rnia Borderland.
marine sediments. Marine Geol., 2, 278-309. 1969. Methods .for the Study of Sedimentary Structures. John Wiley, New York. 140-244. CRISSMAY, S.C. & PLOESSEL,M.R. 1979. Seismic evidence of mass movement in the Santa Barbara Basin, offshore southern California. Abs., Geol. Soc. Am. ann. mtgs., abstracts w. programs 11, no. 7, 407. CROWELL,J.C. 1957. The origin of pebbly mudstones. Bull. geol. Soc. Am., 68, 993-1010. -1974. Origin of late Cenozoic basins in southern California. In: Dickinson, W.R. (ed.), Tectonics and Sedimentation, Soc. econ. Paleont. Min. Spec. Pub., Tulsa, 22, 190-204. DAY, P.C. 1979. The Nature and Distribution of Surficial
M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 181 pp. DOUGENICK, J.A. & SHEEHAN, D.E. 1977. S Y M A P
Sediments on the Submerged Northern Channel
User's Reference Manual: Laboratory jor Computer Graphics and Spatial Analysis. Graduate School of
Design, Harvard Univ., Cambridge, Mass. 158 pp. DRAKE, D.E. 1971. Suspended sediment and thermal stratification in the Santa Barbara Channel, California. Deep Sea Res., 18, 763 9. 1972. Distribution and Transport of Suspended Matter, Santa Barbara Channel, Cal(fornia. Ph.D. Thesis, Univ. of So. Calif., Los Angeles, California. 358 pp. - - - , KOLPACK, R.L. & FISCHER, P.J. 1972. Sediment transport on the Santa Barbara-Oxnard shelf, Santa
Hemipelagic sedimentation in a tectonically active continental margin Barbara Channel, California. hi: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport: Process and Pattern. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 307-31. DUNBAR, R.B. & BERGER, W.H. 1981. Fecal pellet flux to modern bottom sediment of Santa Barbara Basin (California) based on sediment trapping. Bull. geol. Soc. Am., 92, 212-8. DYMOND, J., FISCHER, K., CLAUSON, M., COBBLER, R., GARDNER, W., SULLIVAN,L., SOUTAR, A., BERGER, W. & DUNBAR, R. 1981, A sediment trap intercomparison study in Santa Barbara Basin. Earth planet. Sci. Lett., 53, 409-18. EDWARDS, B.D. 1979. Animal Sediment Relationships in Dysaerobic Bathyal Environments: California Continental Borderland. Ph.D. Dissertation, Univ. So. Calif., Los Angeles, California. 291 pp. EDWARDS, B.D. & GORSLINE, D.S. 1978. New evidence of current winnowing on Heuneme Sill, California Continental Borderland (abs.) Bull. Am. Ass. Petrol. Geol., 62, 511. EMaLET, R.W. 1976. New evidence for occurence of debris flow deposits in the deep sea. Geology, 4, 371-4. EMBLEY, R.W. & JACOBI, R.D. 1977. Distribution and morphology of large submarine slides and slumps on Atlantic continental margins. Marine Geotech, 2, 205-8. EMBLEY, R.W. & MORLEY, J.J. 1980. Quarternary sedimentation and paleoenvironmental studies off Namibia (southwest Africa). Marine Geol., 36, 183-204. EMERY, K.O. 1960. The Sea off Southern California: A Modern Habitat for Petroleum. John Wiley, New York. 366 pp. EMERY, K.O. 1974. Pagoda structures in marine sediments. In: Kaplan, I.R. (ed.), Natural Gases in Marine Sediments. Plenum, New York. 309-17. EMERY, K.O. & HULSEMANN,J. 1962. The relationships of sediments, life and water in a marine basin. Deep Sea Res., 8, 165-80. FIELD, M.E. 1981. Sediment mass transport in basins: controls and patterns. In: Douglas, R.G., Colburn, I.P. & Gorsline, D.S. (eds), Depositional Systems of Active Continental Margin Basins. Pacific Section, Soc. econ. PaleD. Min., Short Course Notes. 61-83. FIELD, M.E. & CLARKE, S.H. JR. 1979. Small scale slumps and slides and their significance for basin slope processes, Southern California. In: Doyle, L.J. & Pilkey, D.H. Jr. (eds), Geology of Continental Slopes. Soc. econ. PaleD. Mon. Spec. Pub. 27. 223-30. FIELD, M.E. & EDWARDS, B.D. 1980. Large submarine slumps off Eureka, California (abs.). Bull. Am. Ass. Petrol. Geol., 65, 924-50. FISCHER, P.J, 1976. Late Neogene-Quaternary tectonics and depositional environments of the Santa Barbara Basin, California. In: Fritsche, A.E., Ter Best, H., Jr. & Wornardt, W.W. (eds), The Neogene Symposium. Pacific Section, Soc. econ. Paleont. Min. Spec, Pub. 33-52. FLEISCHER, P. 1972. Mineralogy and sedimentation history, Santa Barbara Basin, California. J. sed. Petrol., 42, 49-58.
393
GALEHOUSE, J.S. 1971a. Point counting. In: Carver, R.E. (ed.), Procedures in Sedimentary Petrology. Wiley-Interscience, New York. 385-408. 1971b. Sedimentary analysis. In: Carver, R.E. (ed.), Procedures in Sedimentary Petrology. WileyInterscience, New York. 69-94. GATTO, L.W. 1970. Sediment Distribution on the Shelf, Slope. and in Two Submarine Canyons of the Gaviota Area, Santa Barbara County, California. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 184 pp. GORSL1NE,D.S. 1978. Anatomy of margin basins. J. sed. Petrol., 48, 1055-68. - 1980. Depositional patterns of hemipelagic sediments in borderland basins on an active margin. In: Field, M.E., Bouma, A.H., Colburn, I.P., Douglas, R.G. & Ingle, J.C. (eds), Quaternary Depositional Environments of the Pacific Coast. Pacific Section, Soc. econ. PaleD. Min. 185-200. - 1981. Fine sediment transport and deposition in active margin basins. In: Douglas, R.G., Colburn, I.P. & Gorsline, D.S. (eds), Depositional Systems of Active Continental Margin Basins: Pacific Section, Soc. econ. PaleD. Min., Short Course Notes. HAMPTON, M.A. 1972. The role of subaqueous debris flow in generating turbidity currents. J. sed. Petrol., 42, 775-793. - 1975. Competence of fine-grained debris flows. J. sed. Petrol., 45, 834-44. - 1979. Buoyancy in debris flows. J. sed. Petrol., 49, 753-8. HANER, B.E. & GORSLINE, D.S. 1978. Processes and morphology of continental slope between Santa Monica and Dume submarine canyons, southern California. Marine Geol., 28, 77-87. HEATH, G.R., MOORE, T.C., JR. • DAUPHIN, J.P. 1977. Organic carbon in deep-sea sediments. In: Anderson, N.R. & Malahoff, A. (eds), The Fate of Fossil Fuel CO 2 in the Oceans. Plenum, New York. 616-8. HEIN, F.J. & THORNTON, S.E. 1979. Fine-grained mass-flow deposits in slope and slope apron settings: examples from the California Borderland. Abs., Geol. Soc, Am., ann. mtgs., abstracts w. programs 11, no. 7, 441-2. HEUSSER, L. 1978. Pollen in Santa Barbara Basin, California: a 12 000 yr. record. Bull. geol. Soe. Am., 89, 673-8. HULSEMANN, J. & EMERY, K.O. 1961. Stratification in recent sediments of the Santa Barbara Basin as controlled by organisms and water characteristics. J. Geol., 69, 279-90. JOHNSON, A.M. 1970. Physical Processes in Geology. Freeman, Cooper & Co., San Francisco. 571 pp. KOIDE, M., SOUTAR, A. & GOLDBERG, E.D. 1972. Marine geochronology with Pb-210: Earth planet. Sci. Lett., 14, 422-46. KOLPACK, R.L. 1971. Biologieal and Oceanographical Survey of the Santa Barbara Channel Oil Spill, III. Physical, Chemical and Geological Studies. Allan Hancock Foundation, Univ. of So. Calif., Los Angeles. 477 pp. KUENEN, PH.H. 1950. Marine Geology. John Wiley, New York. 337-8. LEE, W.H.K., YERKES, R.F. & SIMIRENKO, M. 1979.
394
S.E. Thornton
Recent earthquake activity and focal mechanisms in the western Transverse Ranges, California. USgeol. Surv. Circular 799-A. 37 pp. LOWE, D.R. & Lo PICCOLO, R.D. 1974. The characteristics and origins of dish and pillar structures. J. sed. Petrol., 44, 484-501. MCCAVE, I.N. 1972. Transport and escape of finegrained sediment from shelf areas. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport: Process and Pattern. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 225-48. MIDDLETON, G.V. & HAMPTON, M.A. 1976. Subaqueous sediment transport and deposition by sediment gravity flows. In: Stanley, D.J. & Swift, D.J.P. (eds), Marine Sediment Transport and Environmental Management. John Wiley, New York. 197-218. MOORE, W.S., BRULAND, K.W. & MICHEL, J. 1981. Fluxes of Uranium and Thorium series isotopes in the Santa Barbara Basin. Earth planet. Sci. Lett., 53, 400-9. N.-A. 1978. Varves and varved clays. In: Fairbridge, R.W. & Bourgeois, J. (eds), The Encyclopedia of Sedimentology. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 841-3. NARDIN, T.R., EDWARDS, B.D. & GORSLINE, D.S. 1979a. Santa Cruz Basin: dominance of slope processes in basin sedimentation. In: Doyle, L.J. & Pilkey, O.H., Jr. (eds), Geology of Continental Slopes. Soc. econ. PaleD. Min. Spec. Pub. 27, 209-22. --, HEIN, F.J., GORSLINE, D.S. & EDWARDS, B.D. 1979b. A review of mass movement processes, sediment and acoustic characteristics, and contrasts in slope and base-of-slope systems versus canyonfan-basin floor systems. In: Doyle, L.J. & Pilkey, O.H., Jr. (eds), Geology of Continental Slopes. Soc. econ. PaleD. Min. Spec. Pub. 27, 61-73. NAVLOR, M.A. 1981. The origin of inverse grading in muddy debris flow deposits--a review. J. Sed. Petrol., 50, 1111-16. NORMARK, W.R. 1974. Ranger submarine slide, northern Sebastian Viscaino Bay, Baja California, Mexico. Bull. geol. Soc. Am., 85, 781-784. PILKEY, O.H., LOCKER, S.D., & CLEARY, W.J. 1980. Comparison of sand-layer geometry on fiat floors of l0 modern depositional basins. Bull. Am. Ass. Petrol. Geol., 64, 841-65. PIPER, D.J.W. 1978. Turbidite muds and silts on deep sea fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. PIRIE, D.M. & STELLER, D.B. 1977. California Coastal Processes--LANDSAT II, Final Report. N.A.S.A. Goddard Space Flight Center, Greenbelt, Maryland. 146 pp. PISIAS, N.G. 1978. Palaeoceanography of the Santa Barbara Basin during the last 8 000 years. Quat. Res., 10, 366-84. -1979. Model for palaeoceanographic reconstructions of the California Current during the last 8 000 years. Quat. Res., 11, 373-86. M
6
R
N
E
R
,
PLOESSEL,M.R., CRISSMAN,S.C., RUDAT, J.H., SON, R., LEE, C.F., RANDALL, G.R. & NORTON, M.P. 1979. Summary of potential hazards and engineering constraints, proposed OCS lease sale no. 48, offshore southern California. Proc. l l th ann. Offshore Techn. Conf., Houston, Texas. PRIOR, D.B. & COLEMAN, J.M. 1978. Disintegrating retrogressive landslides on very-low angle subaqueous slopes, Mississippi Delta. Marine Geotech., 3, 37-60. --, COLEMAN, J.M. & GARRISON,L.E. 1979. Digitally acquired undistorted side-scan sonar images of submarine landslides, Mississippi River delta. Geology, 7, 423-5. RICHMOND, W.C. 1980. Geological hazards and constraints in the area of OCS oil and gas lease sale 48, southern California. US geol. Surv. Open-file Report. 82 pp. ROBERTS, H.H. 1980. Sediment characteristics of Mississippi River delta-front mud flow deposits. Trans. 30th ann. Gulf Coast Assoc. Geol. Soc. Mtg. 30, 485-96. SOUTAR, A. & CRILL, P.A. 1977. Sedimentation and climatic patterns in the Santa Barbara Basin during the 19th and 20th centuries. Bull. geol. Soc. Am., 88, 1161-72. & ISAACS,J.D. 1974. Abundance of pelagic fish in the 19th and 20th centuries as recorded in anaerobic sediment of the Californias. Fishery Bull., 72, 257-73. STANLEY, D.J. & MALDONADO, A. 1981. Depositional modes for fine-grained sediment in the western Hellenic Trench, eastern Mediterranean. Sedimentology, 28, 273-90. STOW, D.A.V. & AKSU, A.E. 1978. Disturbances in soft sediments due to piston coring. Marine Geol., 28, 135-44. & BOWEN, A.J. 1978. Origin of lamination in deep-sea, fine-grained sediments. Nature, 274, 324-8. & -1980. A physical model for transport and sorting of fine-grained sediment by turbidity currents. Sedimentology, 27, 31-46. THORNTON, S.E. 1980. The importance of slope-centered processes in Santa Barbara Basin, California Borderland. Abs. Bull. Am. Ass. Petrol. Geol., 64, 792-3. 1981a. Holocene Stratigraphy and Sedimentary Processes in Santa Barbara Basin: Influence of Tectonics, Ocean Circulation, Climate and Mass Movement. Ph.D. Dissertation, Univ. of So. California., Los Angeles, California. 351 pp. 1981b. Suspended sediment transport in surface waters of the California Current off southern California: 1977-1978 floods. Geo-Marine Letters, 1, 23-8. - & CRISSMAN, S. 1979. Spatial distribution of mass movement in Santa Barbara Basin: sedimentologic evidence. Abs., Geol. Soc. Am. ann. mtg., abstracts w. program II, no. 7, 527-8. YEATS, R.S. 1981. Quaternary flake tectonics of the California Transverse Ranges. Geology, 9, 16-20. -
-
-
-
-
-
-
-
S.E. THORNTON,Department of Geological Sciences, University of Southern California, Los Angeles, California 90007, USA. Present address: Shell Oil Co., Box 527, Houston, Texas 77001, USA.
Studies of fine-grained sediment transport processes and products in the California Continental Borderland D.S. Gorsline, R.L. Kolpack, H.A. Karl, D.E. Drake, P. Fleischer, S.E. Thornton, J.R. Schwalbach and C.E. Savrda SUMMARY: The California Continental Borderland contains a series of margin basins which are sediment traps at increasing distance from the principal terrigenous and biological sediment sources located in the northern coastal areas. Terrigenous sediments are contributed during a short winter season from small steep coastal drainages. Sand content can be as much as 30-40~o of the total contributed load. Silts and clays are sorted out of the river input by coastal wave action and distributed offshore by the seasonal ocean circulation systems. Suspended sediment moves to the basin floors via nepheloid plumes that are generated in the well mixed coastal waters and move offshore with the currents as surface, mid-depth and bottom turbid plumes. During times of low suspensate content in the dry season, lateral transport is dominant. At times of high sediment input and higher concentrations of fine suspensates in the coastal waters, cascading to successively deeper pycnoclines becomes an important process. Resuspension of fine sediments from the outer shelfsubstrate and the axes of submarine canyons by internal waves and tidally generated currents also contributes to the load of the bottom nepheloid plumes. Large semi-fixed seasonal eddies in the Borderland surface circulation hold the suspended load in the northern sector of the borderland for several days to weeks. Zooplankters have ample opportunity to filter out this material and aggregate it into faecal pellets which rapidly fall to the local basin floors. The removal is sufficientlyeffective so as to deposit a minimum of 60~o of the continental contribution within the northern inner zone of the borderland. High sedimentation rates in these basins promote mass-movement and also decrease the already low oxygen contents of basin floor water as a result of oxidation of the organic fraction of the suspended load. Primary sedimentation structures are preserved in these anaerobic basin floor environments. Mass-movements initiate mud turbidity currents that move to the central basin floors. Mass-movements include creep and sliding of thin sheets from low angle slopes. These effects are amplified following exceptional floods at intervals of around 30 years. Fine-grained detrital sediments constitute a significant portion of the bottom sediments in slope, basin and deep-sea deposits off southern California. Consequently, the processes that influence or control the transport and deposition of this material have been important factors in the sedimentary history of the continental margin deposits in the borderland. The broad framework of the California Borderland depositional system (Fig. 1) is controlled to a great extent by topography, sediment supply and water circulation (Emery 1960). Although the first two factors are reasonably well known, the associated influence of the water circulation system is complex and certainly represents a topic for major research effort. A direct approach to analysing the dispersal and deposition of fine-grained sediments involves measurements of the concentration and distribution of suspended particles to derive three-dimensional patterns with respect to seasonal and other climatic influences, sediment sources and rates of supply, biologic contributions, and the effects of current transport. Therefore, the marine geology
group at the University of Southern California has placed a strong emphasis on this method of studying the fine sediment transport to the margin basins of the California Borderland during the past 15 years (Rodolfo 1964, 1970; Fleischer 1970a; Drake 1971, 1972; Drake, Kolpack & Fischer 1972; Karl 1976; Gorsline 1981). This work includes analysis of satellite imagery, transmissometry, water sample filtration from various depths, analysis of suspended particle size and studies of accumulation rates of components over various time scales (Wildharber 1966; Rodolfo 1970; Gorsline & Prensky 1975; Pao 1977; Nardin 1981; Thornton 1981a). An excellent summary of the methods used and the processes involved has been compiled by Drake (1976). Many of his concepts about the settling of fine sediments have been substantiated by more recent work involving sediment traps. These studies and affiliated investigations have allowed us to develop gross sediment budgets (Schwalbach 1982) and to identify sources for the regional contributions of fine sediments throughout the area. For example, Fleischer's work 395
396
D.S. Gorsline et al. o
o
121 ,
o
119 I
117
I
I:..-",:'s. YNEZ' R.''I:.".'.
.
.
.
I .
.
.
9 9
.
9
.
3;
~B--A~ N
"~.'~'.~..'...- --..'i'" .'.- . .
~
- ~ C
R
~
~
.
'
.
9 . . 9
p
"
9
~i:~:"..:...".. ,~/~'.. .
'
.
:
-
.
~
.
.
:
u A ~
\
-
' "
"
33
9
.
-
:
~'~,
.
"
-:: .'... : . . ~:...:.'.. - . .
~
\ EST
, o 3 2 -
EA%T-, L~,.\ "tROUGH ~ "---'-',,SAN I//"% ) ! L____CORTES \ IV ~ ~ \
;ORTES
~,"",-"/---'-"~ ,.) CLEMENTE BASIN
BASIN
V
) BASIN
'N/ I
!
FIG. 1. Location map for the area off southern California.
(1970b) on the mineralogy of silt and clay fractions in several basins has shown that suspension transport in the California Current provides some input to the system from the area north of the borderland. A persistent approach to the problem also enabled us to conduct field sampling during a variety of conditions ranging from major events such as the flood of 1969 as well as at times of reduced sediment influx. In addition to a variety of benthic biological studies (e.g. Hartman 1955, 1960; Fauchald & Jones 1976), work by Savrda (1982) and Savrda et al. (1982) on the patterns and dimensions of burrow forms, tracks and trails has shown that these are correlative with sedimentologic and oceanographic factors. Almogi-Labin (1982) has also noted close association between detailed micropalaeontologic data and these same factors and trends. Douglas (1981) and his students have studied the parallelism in modern microfaunal data and the sedimentologic features. This is particularly important since environmental reconstructions of ancient fine-grained environments depends on palaeontologic data. In this regard, one can generalize that sands are dominated by physical factors, while clays and
silts can be said to be strongly influenced by biological factors. Studies by other workers, such as sediment trap collections from the Santa Barbara Basin (Dunbar & Berger 1981), have provided additional useful information about the system. Some of this work is evaluated in the context of the regional setting, topographic control, source and supply of sediments, climate and water circulation system to derive a synopsis of what is presently known about suspended sediment transport processes operative in the California Borderland. The sequence of discussion will commence with surface water plumes of suspended material, move to studies of fine-grained sediment transport on the shelf, transport of suspended sediments in submarine canyon axes, deep water turbidity data, and finally this information will be related to surficial fine-grained bottom sediments in the borderland.
Regional setting The coastal and offshore area of southern California is an active transform margin (Emery 1960;
Fine-grained sediment transport processes and products in California Crouch 1981) dominated by a series of basins with progressively increasing sill and basin floor depths at increasing distances from detrital sediment sources. The present basin arrangement and configurations appear to date from early Pliocene time (c. 3 my up, Ingle 1980; Nardin 1981). Maximum filling occurred first in the Los Angeles and Ventura Basins and then shifted to the next seaward row after the innermost basins filled to overflowing and renewed structural deformation began about 500 000 yrs up (Nardin 1981). Now, the present inner marine basins (see Fig. 1) are rapidly filling and capture much of the incoming terrestrial contribution; perhaps 6t.~o v/o/ of the total regional contribution is accumulated in the present four inner basins (Schwalbach 1982). Nardin (1981) has shown that the sedimentary history of the contemporary inner basins (Santa Barbara, Santa Monica, San Pedro and San Diego) is one of increasing sediment accumulation rate. In the summary of the fine sediment regimes outlined by Gorsline (in review), this area is one with relatively small annual sediment inputs of about 5-10 • 106 tons/yr in a typical decade, and moderate wave energy. About 60-70o/o of the sediment input is silt and clay-size material with sand and some gravel constituting the remaining portion. Locally, we have sometimes referred to the region as one of high wave energy. However, Paul Komar (pers. comm. with D. E. Drake) has noted that statistics show that southern California coasts have lower-third wave heights that are comparable to the Gulf Coast and well below those of northern California. Most of the rainfall in this semi-arid Mediterranean climate occurs during December-March. As a result of the typically steep, small coastal drainages of this active margin, sparse vegetation and the restricted periods of rainfall, most of the sediments are delivered directly to the wavedominated inshore zone by seasonal rivers. Very large floods with discharges of 100 • l06 tons/yr and probably more occur at intervals of a decade to a generation (Gorsline 1981). For example, since modern records were begun in the Los Angeles area in the late nineteenth century, historic floods occurred in 1875, 1884, 1889, 1892, 1916, 1938, and 1969. These major floods have almost always occurred in January and February. They deliver sediments into the system in amounts approaching those of the large continental river drainages and are analogs of major floods of greater magnitude that occur every 2-3 years in a large river system such as the Mississippi (Leopold et al. 1964). Water circulation in the California Borderland is dominated by the California Current system with its associated eddies and counter currents
397
(Reid et al. 1958; Hickey 1979). The dominant flow is southward along the western coast of the United States. However, south of Point Conception portions of the eastern part of the main current commence meandering and eventually form large eddies and rings. In the axis of the California Current along the west coast of North America to approximately the latitude of Point Conception, a line of large (200 km diameter) semi-fixed seasonal rings form the actual flow (Simpson et al., in press). South of Point Conception, portions of the eastern part of the main current meander and eventually form large eddies and rings within the borderland circulation system. A nearshore counter current, apparently derived from Pacific Intermediate Water, flows northward over the inner series of borderland basins from about the latitude of San Diego and north to Point Conception (Tsuchiya 1980; Kolpack 1971). Currents on the narrow mainland shelf are typically dominated by small eddies or circulation ceils that are controlled or influenced by shoreline configuration, submarine canyons, surface or internal waves and edge waves, and the northward-flowing counter current (Karl et al., in press; Shepard et al. 1979; Kolpack 1971, 1979; Tsuchiya 1980). Nearshore or littoral drift transport is predominantly toward the south along the mainland coast south of Point Conception as a result of incident wave action (Emery 1960). The combination of topography, water circulation, and climate produce episodic influxes of fine sediments that are transported offshore by a variety of local and regional currents.
Surface water turbid plumes During times of flooding, the surface expression of suspended sediment plumes emanating from a few point sources, such as the Santa Ynez, Santa Clara and Santa Ana Rivers, can be tracked for considerable distances by means of aerial photography, satellite imagery and measurements from surface vessels. Analysis of this information has provided considerable insight into the regional surface water circulation patterns and the pathways by which some fine sediments are transported to offshore basins. LANDSAT imagery for the winter of 1977-1978 was particularly useful because that wet year produced well developed turbid plumes. Thornton (1981b, this volume) has defined three distinct cells in the typical turbidity patterns over the northern borderland (Fig. 2). The imagery indicates that during the winter season of maxi-
398
D.S. Gorsline et al.
,
5~
f:e I Conr
~
(a)
~RBID4
T~Y
No,,,.9L,:,8NOS,T AFTER
FIG. 2. (a) Surface current pattern derived from turbidity plumes seen in LANDSAT imagery of January and February 1978 (Thornton 1981b). (b) Tracings of turbid plumes in surface water for January 1978 from LANDSAT imagery--compare with (a). (c) Cell positions of semi-fixed eddies in the Santa Barbara Channel based on LANDSAT imagery studies (Thornton 1981b; Davis 1980); also see Fig. 6 in Thornton, this volume.
-,
,.
,-,_ ""
THORNTON, 1978
9
(c)
-.-
"B"
".
"
I
Fine-grained sediment transport processes and products in California mum sediment discharge, the major gyres in the borderland trap and hold the surface plumes for periods ranging from a few days to a few weeks (Pirie & Stellar 1977; Thornton 1981b; Davis 1980; Fig. 2; also see Fig. 6 in Thornton, this volume). This entrapment, plus active biofiltration which occurs in the coastal waters, causes much of the suspended detrital matter in the surface plumes to be deposited in the inner margin basins (Fig. 3). The speed of particle removal has been studied by many marine planktonologists and suggests that the process operates over periods of days (e.g. Hofmann et al. 1981). Consequently, only a minor amount of this material is transported to the outer basins or is entrained into the main core of the California Current. Detrital suspended material from north of Point Conception can become incorporated in the California Current and may be held in the large-scale (200 km diameter) rings described by Simpson et al. (in press). These features would presumably provide surface zooplankton sufficient time to ingest the suspensates and incorporate the material in faecal pellets which settle out of the water column rapidly. Study of the subjacent deep-sea sediments would be an interesting research project. However, when the regional eastern Pacific oceanographic circulation changes in response to seasonal variations in the position and strength of the continental atmospheric high pressure cell (Hickey 1979), the various rings and eddies wane
~"~~~_~/'~
\
399
and shift and suspended particles then are transported along different pathways. Because the times of high surface water productivity need not coincide with times of maximum detrital input, the biogenic and detrital particles may follow different routes to depositional sites. Thus the accumulation loci of biogenic carbonate, silica and organic matter may not overlap the loci of maximum terrigenous accumulation rate in the margin basins or on the continental slopes. In this regard, it is interesting to review the earlier work of Sverdrup & Allen (1939) which shows the pattern of concentration of diatoms in the surface waters off southern California. Their diagrams and charts show high concentrations south of Point Conception and south of the northern Channel Islands as would be expected from knowledge of the upwelling occurring there. In addition, they show a concentration maximum in an area which we now know is in the axial region of the California Current offshore of Point Conception in the general position of the large seasonal semi-fixed axial rings described above. These areas of high productivity have traditionally been explained by the classic coastal upwelling mechanism. Certainly the times of maximum production are related to times of minimum wind stress. However, the interpretations supporting a classical upwelling mechanism, particularly in the area south of Point Conception, are still in debate. An alternative explanation may involve the complex eastern
TOTALIACCUMULATIONRATES'-mg c m - 2 yr -'
"':.:
"'"
20 (630:'" 9...
. i
FIG. 3. Accumulation rates of borderland sediments (Schwalbach 1982).
l
400
D.S. Gorsline et al.
frontal zone dynamics of the California Current at the western end of the Santa Barbara Channel rather than classic wind-driven upwelling from coastal sites. In any event, these examples illustrate the important interaction between surface water circulation patterns and the distribution of particulate matter in the underlying bottom sediments off southern California. Schwalbach (1982) has shown in his analysis of accumulation rates of the basin sediments of the borderland that the highest accumulation rates of biogenic material are found in the Santa Barbara Basin. Thus, it is evident that whatever drives the productivity processes in this area the rate of production is high. His data also show that the accumulation rates of biogenic material south of the northern Channel Islands are either lower or are distributed over a much larger area south of the area of high productivity. Certainly the LANDSAT data (Thornton 1981b; Davis 1980) show a more broadly developed set of gyres south of those islands as compared with the confined three-cell circulation in the Santa Barbara Channel. Schwalbach's data suggest that 60 to 75~o of sediment input from local sources to Santa Barbara Basin are captured by that basin whereas the inner basins further south (Santa Monica, San Pedro) capture perhaps half of the available local input and the remainder passes into basins further offshore or is carried south with the California Current. These three basins plus Santa Cruz Basin, capture about 60~ of the total coastal suspended sediment input into the borderland as noted earlier.
Subsurface shelf turbidity patterns: southern California shelves Ocean waters are stratified; typically a permanent or seasonal mixed surface layer is formed by wind mixing and overlies a deeper colder water mass. Thus, below a depth of a few metres, satellite imagery is not effective in recording turbid plume patterns. Transmissometry, water sample filtration and analysis of substrate characteristics are then the principal tools for field studies. A combined hydrographic, suspended sediment and bottom sediment investigation of the Santa Barbara Channel during 1969-1970 resulted in a time-series analysis of the behaviour of flood-derived sediments introduced to the system from a few point sources (Drake 1971; Kolpack 1971; Drake et al. 1972). Initial deposition of about 75~o of the local river-contributed silts and clays occurred on the adjacent shelf and slope. Portions of this material were then progressively trans-
ported to deeper water by wave and current reworking and nepheloid flow during the following several years. The distribution of suspended sediment concentration and light transmission values paralleled both the shelf circulation pattern and the progressive transport of fine bottom sediments across the mainland shelf at the eastern end of the channel. This information provided considerable insight into the mechanisms and pathways involved in fine sediment transport from source to ultimate depositional site in the margin basins. Mud turbidity currents and nepheloid flow probably were the principal transport processes. It should be noted, however, that the paths of fine suspended sediment transport and coarse bedload sediment transport from river mouth to deep sea may be quite different. Fine-grained sediments often remain in suspension for considerable periods of time and generally are transported in accordance with shelf water circulation patterns; whereas sands initially deposited near a river mouth tend to be transported in the longshore drift system. Canyons may capture both fractions but much of the fine suspension load can bypass canyons en route to offshore slopes and basins. In the central portion of the San Pedro Shelf the inshore waters shallower than the thermocline depth are well mixed and there is a significant amount of turbidity throughout the entire water column (Karl 1976). As these inshore waters move seaward, there is a turbidity layer bounded at its base by the regional thermocline (pycnocline). Particles tend to collect at that boundary owing to increased viscosity (decreasing settling velocity), and increasing density. Turbulence at the shear boundary between the surface layer and the deeper water layer also tend to hold falling particles and increase the suspended sediment concentration. Subsequently, particles settle through this layer to the bottom or flow along the shelf bottom from the base of the turbid mixed inshore zone. These layers then form either mid-water plumes as they move through deeper density discontinuities and/or continue along the bottom as a bottom nepheloid zone. These midwater particle concentration maxima have been observed at all sites sampled in the borderland (Drake 1971; Drake & Gorsline 1973). It is speculated that in some seasons internal waves impinge on the shelf surface in the vicinity of the thermocline intersection with the bottom and form internal surf, a surge which resuspends particles into the bottom nepheloid-layer. Cacchione & Southard (1974) discuss three models for internal wave shoaling; (1) formation of internal surf, (2) internal surges or bores, and (3)
Fine-grained sediment transport processes and products in Cali/ornia absorption of the impinging wave train. Which form is typical of a given shelf is a function of internal wave characteristics, shelf morphology and water density distributions. Off southern California, and probably off most narrow active margin shelves, the internal surf or surging seems to be most probable (Emery & Gunnarson 1973), and seasonal differences in the field data have been speculatively attributed to this mechanism (Fig. 4). Observations of suspended sediment concentrations on the San Pedro Shelf, obtained over a period of several years, have been reduced by Karl (1976)into winter and summer models (Fig. 5) showing the different plume positions, concentrations and interactions. McCave (1972) suggested that a cascading process from layer to layer may be important in plume maintenance and formation. Our data suggest that cascading becomes important only at high particulate concentrations above 3 mg/L and that at lower concentrations of suspended sediment, lateral
4oi
flow processes are dominant and turbulence at density discontinuities or near the bottom is strong enough to maintain suspension. As the various layers pass over the shelf-edge they typically encounter a shelf-edge or upper slope current that disperses the plumes along the contours of the slope. This deeper counter current flow is a typical part of models of coastal circulation in areas of upwelling. Karl et al. (in press) describe detailed current patterns and suspended sediment concentration patterns for early spring of 1978 on the San Pedro shelf off southern California. They believe that the significant changes observed in these patterns are the result of the onset of upwelling over the shelf or to an intrusion of cold water by advection from north of the area. The net result is an augmented seaward transport of fine sediment as contrasted with average conditions of shelf circulation. Drake (pers. comm.) has suggested that wave effects are a major factor in these shelf suspended sediment transport systems.
FIG. 4. Turbidity cross sections from Karl (1976) showing the probable influence of internal waves at the depth of thermocline interception with the bottom off San Pedro, California.
402
D.S. Gorsline et al.
D~pa'n
FIG. 5. Model of turbid layer development in typical winter and summer conditions for the southern California shelf (Karl 1976).
]It
In some areas the natural distribution of suspended particles is overwhelmed by the introduction of anthropogenic material from offshore sewage outfalls. For example, the distribution of suspended particles on either end of the San Pedro Shelf is much more complex owing to the discharge of about 99 000 tons/yr of suspended solids at the north-western end and about 36 000 tons/yr at the south-eastern end (SCCWRP 1980). Nevertheless, the relatively steady influx of this material permits one to use the material as a tracer to study processes operative in those areas. A similar situation exists in Santa Monica Bay where about 75 000 tons/yr are discharged from two outfalls located 5 and 7 km offshore. In that area the suspended particles derived from the effluent outfall can be traced over most of the central shelf and the distribution of material in the water column is markedly affected by internal and standing wave activity (Kolpack 1979; Fig. 6). The major transport of suspended material in Santa Monica Bay occurs in well-defined midwater conduits approximately normal to the coastline. As a consequence, there is only a relatively minor net accumulation of fine natural and anthropogenic material on the central portion of the Santa Monica Shelf.
Fine sediment transport in submarine canyons The model of submarine canyons as the pathways for mass-movements and turbidity currents is well known and well described. It is perhaps not as generally realized that these same pathways may be continuous conduits for the passage of suspended sediments to the deep-sea floor. As will be discussed, a canyon exerts a major influence on the water circulation of its shelf and also is the site of strong interactions by tidal, internal and surface waves. A submarine canyon is not a necessary element, of course, for across-shelf suspended transport systems. Thornton (198 la) has discussed the movement of fine-grained sediments into Santa Barbara Basin from a notably undissected shelf. Drake & Gorsline (1973) and Beer & Gorsline (1971) reported on the circulation of nepheloid plumes down canyon axes off southern California. In these canyons, the cyclic water flow over the bottom is tidally forced with a net downcanyon flow (Shepard et al. 1979). At peak flow, particles are reworked into suspension from the canyon floor. These particles commonly enter the bottom nepheloid-layer and thus are progressi-
Fine-grained sediment transport processes and products in California
FIG. 6. Three-dimensional representation of turbidity of a sewage plume in Santa Monica Bay (peak is discharge point) (a) and depth to the core of the plume from the water surface (b). After Kolpack (1979).
403
D . S . Gorsline et al.
404 120 ~ 30"
120 o 0 0 '
L
~
I ~-.~ 34 ~ 3 0 ' I -
\
1
.~o~'
'~
O~~s~
I
~.~.
5" ~
* ~q~v ,"
[.
.
i
5
. ....
.
SURFACE
nouticoi mi
.....
.....
.
.
-
.
""~. 34 ~
:~ k i l o m e t r e s
"C
"\119~ 30 '
..
"
SANTABARBARA
\.,, 9
I
TEMPERATURE
DECEMBER, ,969
156
*
/
\..[.
OXNARD
j
34 ~ 00' {-
S,N MIGUELISLANO ~
120 ~ 30"
k
~
120 ~ 0 0 '
SANTACRUZISLAND
~#/
34o00 '
119~ 3 0 '
FIG. 7. Surface water temperature in winter (Kolpack 1971) showing the intrusion of water over the sill at the east end of Santa Barbara Basin. vely moved down slope to the fan. However, once sediments in the canyon axes are brought into suspension, they also may be transported offshore within the water column (Kolpack 1979). As noted by Drake (1976), benthic infaunal reworking of surficial fine sediments can keep the material in a state of low bulk density which permits easier reworking by peak flows. Conversely, work reported by Richardson & Young (1980) suggests that this biological stirring can work in both directions; toward greater compaction or to lower bulk density, in any event, resuspension of the deposited fines is strongly influenced by sediment shear strength as well as particle threshold velocity characteristics, and sedimentation rate and infaunal bioturbation certainly affect this. Bennett & Nelson (1982) have reviewed the processes involved in massmovements on slopes and their paper is an excellent reference for these processes. As sea-level rises during a glacial eustatic cycle the rate of sea-level transgression will outstrip the headward erosion of a previously active submarine canyon head (Karl 1980). As the headward erosion falls behind transgression, and the shoreface drift transport ceases to be captured by the canyon (see Shepard & Dill 1966), the canyon will still influence the incoming wave trains. This influence will set up edge waves and offshore flow towards the canyon head that will continue to move suspended material from shore to canyon.
When the separation by the increasing depth of water removes this influence, the offshore flow ceases and the canyon is finally abandoned in the sedimentologic sense. Thus the stratigraphic results of canyon abandonment should be a marked decrease in grain-size in the canyon fill and a cessation of turbidite deposition in the fan and fan distributaries. Beer & Gorsline (1971) have shown that the heads of active canyons are the sites of active gyral motion in the overlying surface waters. This tends to pull shelf water to the canyon. Such an effect is probably driven by the tidally-forced cyclic flows described by Shepard et al. (1979) and is thus separate from the mechanism suggested by Karl. This may cause suspended load in shelf waters to continue to move toward canyons. However, the contribution is probably less significant than that from shore zone plumes; the stratigraphic record of inactive canyons would show very low accumulation rates if only the latter mechanism was operating. Gorsline (1980), from evidence of patterns of fine hemipelagic surface sediments in California Borderland basins, suggested that nepheloid flow from canyons imposes an important control on accumulation of these sediments. Thus, submarine canyons, when present, do have a significant effect on pathways of suspended sediment transport from coastal sources in addition to their contribution of turbidity currents and associated mass-movements.
Fine-grained sediment transport processes and products in California
Turbidity patterns and fine sediment distribution patterns in the basins and over sills
ments from the eastern basin sill and movement of this material to the adjacent basin floor (Fig. 11). In addition, Edwards & Gorsline (1978) have shown that the sediment characteristics of the Hueneme Sill offsouthern California are residual and that bedforms are present attesting to the strong current. High resolution seismic profiling (3.5 kHz source) over the Hueneme Sill also shows appreciable scour and removal of old channel fill deposits as well as recording the presence of large sand waves (Thornton 1981a). Even during years of relatively low sediment input, the influence of sills on the particle load of deep water currents passing over the sills is striking. During 1982 we have sampled particle concentrations at closely spaced sampling points within the lower 100 m of the water column across several basin sills. Information from this survey shows that the particulate content not only increases downstream, but also that as currents pass over a sill, the scale of turbulence increases as the flow detaches at the sill and mixes the suspensates into a much thicker nearbottom nepheloid-layer. Emery & Hulsemann (1962) gave a classic description of the mode of fine sediment accumulation in margin basins using data from the anoxic Santa Barbara Basin off California. In that basin, low oxygen content bottom water originating from the Pacific Intermediate Water and oxygen minimum zone is further reduced in
The similarity between turbid plume flow and the general oceanographic circulation on the shelf in the eastern portion of Santa Barbara Channel is also true for a large part of the channel (Drake 1971; Kolpack 1971; Thornton 1981a). Examples of the coincidence of surface water turbidity and seasonal temperature patterns (Figs 7 and 8) as well as the surface water circulation pattern (Fig. 9 and Kolpack 1971) all outline the close association between suspended sediment transport and hydrographic features. Furthermore, the bottom sediment accumulation pattern recorded after the 1969 flood (Fig. 10) and discussed by Kolpack (1971) and Drake et al. (1972) also matches the overlying water flow (also see Fig. 2). The sediment layer shown in Fig. 10 accumulated progressively over two years as waves reworked the fine sediment initially deposited on the mainland shelves. Textural patterns in Santa Barbara Basin surface sediments are best explained by the same circulation model as shown by Thornton (1981a). It is evident that the mid-water and bottom water flow patterns are similar to the surface flows in the eastern half of the channel. Drake's deep transmissometer casts depict scour of sedi120 ~ 3 0 '
120 ~ 0 0 ' ~
1
SCAL[ s,.:~_j~___ ~c,=x~,,~_ ,o
34 ~ 00'
~
~ , ~ s
~.
SURFACE
-\
"~ ,onauticolmi.
kilometres
\
~ "~~ f
~ ~
~
~ ~ ~ ~
~ .
TURBIDITY
% TRANSMISSION 34 ~ 30' - ~ . 119 ~ 3 0 ' "r FEBRUARY,
/
METER
1970
SANTA 6ARBARA
---.,f8o,.'"'-'.,------~,, ~
,,
405
-
~
~\
"--5,, ',,o,?) C
,o_ <.
--%._,,/
60
"-.
",
-
(t,o_L 34000 ,
I
120 ~ 3 0 '
1
[
1
120000 '
_ I
L
t .....
2
. . . .
J
. . . . . -1
119 ~ 3 0 '
FIG. 8. Turbidity of the surface one metre of water showing the typical winter distribution (Drake 1972). Compare with temperature data in Fig. 7.
406
D.S. Gorsline et al.
120~
119"40"
9
o
40' I / :..}
SURFACE
l ~!:
I ~
CURRENTS
UNPINE fG H'TED
FEBRUARY, 1970 ( ~ = 2 5 % RECOVERY
. Clh ~ .
0'~ c t O l ' O "
~'~_
(Ik
~
4o'
DF~]FI" CAR DS
~
~...
~
5
~
~
B~.~Ix
.....~...:.
s=,~,~-
~
:
o
o
: :
,
,- ~"
....
___~ o o o ' ~ ~ 0O0 0'
-------"__..~~
$ N MIGUEL "~S-"NTA " ~
I 120~
~
~
ANACAPA
1
,
o
1
00'
1
120~
119~
"
FIG. 9. D a t a from K o l p a c k (1971) s h o w i n g drift card trajectories in winter. N o t e similarity to Figs 7 and 8.
,/o. 30
L I ~ I~
,ze-oo
~--
I
r
1
--
1
FLOOD
t--rrY-"'-I
.naulicalmi~CeRu,RYkilometres
~,, ~ . _ . . ~ - ~ ' ~ - ~ ~ "
p
9
.
.
9
.
.
p
/
.
*
~
SEDIMENT
..._~ . . a ~ .
.
"~.-,~
P
JUNE.J,70 . ......
"~.'*
SANTAISmOAItA
THICKNESS
"~
~~, ~ " ~ ~" , ~ ~ ~,~ , -,~'~ " , ~~_ ~
..
-;--?---...._~--~,.~ :,~] 1 i. . . . ~ _ _ _
p
9
~,.---.,~ ---,~.~
"-- ...
~
.
/~
"..
J FIo. 10. Pattern of sedimentation following 1969 flood reworking and transfer to the basin floor (1970) from Drake et al. (1972). See close association with surface circulation with an offset downcurrent due to particle settling lag.
Fine-grained sediment transport processes and products in California ,zo-~o
,zo.oo,
k. I
407
s,.T,o. LOCA,,ONS "~'~,
? -o~
5
71 ~.o
~ $CALt o
.
.
.
.
t
,0 n o u t i r
mi
.~
, .3o.
I
15sy ~5.s56 j2~^
I
,20"30'
15867
I
15865
I
15859
JULY, 1971
.
15854
~
~
I
,20"00
15850
15872 r
">...
,,9.1 o,
15873
15874
3_
15875
__
I
15876
15877 6 0 .-.__...~
100
200
300
400
500
FZG. 11. Top diagram shows station locations for profile in lower diagram, after Drake (1972). Lower diagram shows cross section east-west, along basin axis showing bottom and surface turbid plumes. Note that the bottom layer moves out of the section and then back through due to the gyral flow in the bottom water. oxygen content by decay of organics settling into the bottom waters. Douglas (1981) and Gorsline (1981) have described this influence of high sedimentation rates on the oxygen content of the basin bottom waters. As one moves northwestward through the inner basins of the borderland in the direction of the bottom water motion, the oxygen content decreases from dysaerobic levels in San Diego Trough and near anoxia in San Pedro Basin to anoxic (anaerobic) conditions
in Santa Monica Basin. In addition, as low oxygen content water moves into Santa Barbara Basin over its western sill from the core of the Northern Pacific Intermediate Water, high oxygen demand from organic debris produced in the surface water layer reduces the already depleted oxygen content of the bottom water. In central Santa Monica Basin (Malouta et al. 1981) and Santa Barbara Basin (Emery & Hulsemann 1962; Soutar & Crill 1977; Thornton 1981a)
408
D.S. Gorsline et al.
bottom waters are anaerobic and bioturbation of the substrate ceases (Savrda 1982; also see Figs 13 and 14 in Thornton, this volume). Savrda has shown that the ichnologic data can be related directly to depth and oxygen content, and that the resulting patterns match sedimentologic and environmental boundaries and patterns (Fig. 12). With anaerobic conditions, primary sedimentary structures are preserved. Resulting lamination demonstrates that the fine sediment input is seasonal, matching the climatic regime of the adjacent continental source area. Gorsline (1981) has described several models for formation of primary laminations showing that they can be produced by a variety of scenarios in which at least one component (terrigenous or biogenous) must vary seasonally (Fig. 13). Off southern California the important factor is the seasonal input of silts and clays during the winter rainy season. Soutar & Crill (1977) have shown that the thickness of the resulting terrigenous varves is related to annual rainfall. However, Thornton (1981a) and Drake et al. (1972) have shown that
thick layers can be the result of exceptional floods which can occur in otherwise average or dry years. For example, the floods of 1969 occurred over a two-week period in January-February, yet the total annual rainfall for 1969 was near the long term average. The resulting layer deposited in Santa Barbara Basin was over 3 cm thick as compared with more typical annual laminae thicknesses of 1-3 mm (Fig. 10). It is evident that varve thickness in anoxic basin mudstones can include both an annual rainfall measure and a flood influence. Thornton (1981a) has also reviewed the influence of high sedimentation rates on slope sediment stability. Major flood years will be followed by a period of slope instability and slumping. These slumps then introduce muddy or fine-grained turbidites into the basin floor (see Fig. 17 in Thornton, this volume). Thornton shows that in Santa Barbara Basin these events have periodicities of a few decades to as much as 1000 years. Shorter period frequencies are associated with turbidites from flood accumulations
FIG. 12. Bioturbation effects in the bottom sediments of Santa Monica Basin illustrating the close match in these effects with topography and bottom water condition. Basin floor lies within dysaerobic to anaerobic condition. I.--well laminated; II.--laminated with minor bioturbatiom lII.--lamination reduced to banding, slight bioturbation; IV.--faintly banded, bioturbation has destroyed laminae; V.---complete bioturbation; VI.--cross laminated, strong bioturbation; area between dashed line and zone IV is ephemeral pellet structures (Savrda 1982).
Fine-grained sediment transport processes and products in Calijornia
JANUARY
OgC~MBFR
4~ i
409
J
0
4, J ,
i= o'o',~[
IO~,o"I 0
N ~
I XO wol 0~ ~ 0 v
L~,- w x ~ i j
SEOIMENT
DEPOSIT
t
f~.\ //
FIG. 13. Models of laminae formation (Gorsline 198 l) illustrating that laminae can be formed by several sedimentation scenarios when two or more components are involved. Type III is typical of Santa Barbara Basin.
J
on shelves and slopes as in 1969; longer period frequencies are associated with times of major slope failures. These slopes require longer accumulation times to reach instability. Detailed study of laminae character will reveal grading and zonation that can assist in unravelling the story. Thornton (1981a) examined the coarse fractions of basin sediments and other textural and compositional data and constructed a map of sources (see Fig. 8 in Thornton, this volume). The effect of surface and deep water circulation can be seen in the permanent patterns of the slope sediments under the major current moving westward out of Hueneme Passage (east end of Santa Barbara Basin) and in the sorted and residual sediments of the sill and its adjacent slopes. Note that the pelagic-dominated core zone of the basin is displaced westward from the geographic centre by the circulation systems. Seismic stratigraphic analyses by Nardin (1981) of inner borderland basins shows that these surface patterns probably existed for much of the Late Pleistocene and Holocene. Malouta et al. (1981) have described an interesting pattern of central basin fine sediment accumulation rates that suggest that deposition of fine suspension load in the deep basin water is influenced by gyral currents in the deep basin waters caused by Coriolis effects on deep water circulation (Fig. 14). In that figure a dashed current arrow has been superimposed to show the effect. Similar effects probably occur in all basins or troughs and may influence organic carbon distributions. Organic debris has a low density
\k L
,
/- \
D
Iu j
)z t
{o o o g o ~ o
and will be strongly influenced even by low velocity currents of this type. These deep, slow circulations are not deep counter currents of the type described from the north-western Atlantic US rise by Heezen et al. (1966) and probably do not attain sufficient velocities to materially rework bottom fine sediments. It is more likely that these currents influence the primary deposition of fine sediments rather than resuspend bottom materials. Both are influenced by earth rotation but one is part of the major thermohaline ocean circulation system while the other more restricted basin type is probably driven by tidal forcing of bottom waters into marginal depression. Where sedimentation rates are high the stability of slope sediments is decreased and massmovements become common (Nardin et al. 1979; Field 1981). This is particularly evident in Santa Cruz Basin where tectonic stresses are apparently creating an active rift basin. Similar effects are also evident in Santa Barbara Basin. Overlying water circulation acts to hold suspended sediment over the Santa Cruz Basin where accumulation rates on the basin slopes are high enough to create metastability in deposits and they periodically slump to the slope base. This is shown in Fig. 15 in which it is evident that almost the entire perimeter of this basin is subject to mass-movement. Similar convergences of circulation and coastal sediment supply have produced an unstable slope in San Pedro Basin where a recent mass-movement (c. 100 to 500 yrs old) has been observed in some of our recent reflection profiling work. Figure 16 illustrates these examples of the concept that fine
4zo
D . S . G o r s l i n e et al.
-~.,-
P~
~RG~BON
( r--~,-'-x~
t cm / I 0 0 0 "r ~
~
~
) -'
CARBONATE
"•••UM
'Ib,.,r
~"
T
n
--
--"
57.5 I0 7.5 5
SE0,ME.T , .........
50
. 6 ~ , ~ ~ 7 ~ --
FIG. 14. Accumulation rates for contemporary hemipelagic sediments in Santa Monica Basin (Malouta et al. 1981). Note difference in patterns for total sediment as compared with carbonate and organics which are pelagic. Patterns are probably the result of gyral basin bottom water flow imprinted by nepheloid flow from coastal canyons from which detrital contributions come (see Fig. 1 for location).
cm /nooo yr \
_
7"='
'~
~
.. T
~
/
~
",o.~
Fine-grained sediment transport processes and products in California
4I I
FIG. 15. Slide and slump deposits in Santa Cruz Basin (Gorsline et al., in press.). Major elements are 80-100 km 2 in area. Ages are probably late Pleistocene or early Holocene.
~s Movement Feat ures~<,
D ~ Fan
D ol
,'5.
-,.9,"
o 33 ~
~---?
FIG. 16. Slump sequence in San Pedro Basin. The hummocky area is a blocky slide and slump: the smooth apron is formed by high concentration debris flows. Symbols D and T refer to surficial core features not discussed in this paper. Box cores show that thin mud turbidity currents also issue from the toe.
412
D.S. Gorsline et al.
sediment may follow several stages of transfer before reaching the ultimate depositional site in a basin floor or abyssal plain-slope apron. Hein (in review) has examined the micro-fabrics of clays from various slope and base-of-slope environments in Santa Barbara and Santa Cruz Basins and sees characteristic structures in areas of sliding. These micro-fabrics include swirl patterns, random chains and oriented aggregrates depending on whether the samples came from slumps, hemipelagic drapes or laminated muds.
bypassed. From the coast, all fines still in transport are increasingly acted upon by the zooplankton aggregating process. Much of the input is therefore passed rapidly to the underlying surface as pellets. Several surfaces may be receivers. The possible sites are canyons, shelves and slopes with some bypassing to more distant deep areas. Reworking of shallow shelf surfaces is indicated as is reworking from canyon walls and axes by wave and current action. Some bypasses as suspension load to the fans or deep floors. Major processes are indicated by SS--suspension transport, MM--mass-movement or high density flows; TC--turbidity currents. SS would also indicate nepheloid flow. Some additional input to the system beyond the coast comes from current advection, wind, and bottom erosion as indicated on the side of the diagram. All eventually moves to the deep ocean floor. Obviously some storage occurs in each level. This diagram can be used to show us where we lack quantitative data for the transport system. We have reasonably good data to show that the scale of the fine input is of the magnitude of about 10 x 106 tons/yr over a 1000 yr period. What the partitioning of this total may be within the system is our largest future problem. It is also evident that a major need is to examine the transport systems during times of stormy conditions since the great bulk of field measurements are during 'fair weather' conditions.
Summary of fine sediment transport In summary, we can follow the motion of fine sediment offthe California coast, and offmargins in general, in diagrammatic form (Fig. 17) beginning with river input and then tracing typical particle pathways to the eventual deep basin deposition site. We show three possible coastal areas (estuary, delta, beach) with a general indication of the effect. For example, an estuary is a sediment sink that also draws fine sediment from the marine side. It is thus a major storage area during its life. A delta collects fines but also passes some to the shelf as plumes (Gibbs 1981). Therefore, it serves as a filter. A beach (open shore) is a high energy zone and fines are virtually all
CONTINENTAL J SOURCE River
0
Estuary SINK STORAGE
Delta FILTER storage
Beach BYPASS
shelf
shelf
shelf
I PLANKTON 9
--~,
T
screen
9
shelf 4111~SS/k slope ~" 9 MM canyon basin walls411Daxis ~ floor
~
I
advection
~wind
bypass deep-sea SS
I erosion
/ FIG. 17. Diagrammatic representation of fine sediment distribution in the borderland and margins in general. SS--suspension transport (nepheloid flow); MM--mass-movement or high concentration flow; TC--turbidity currents.
Fan MM TC
basin floor
FINE SEDIMENT TRANSPORT SCHEME
Fine-grained sediment transport processes and products in California ACKNOWLEDGEMENTS: T h e s u p p o r t o f the B u r e a u o f L a n d M a n a g e m e n t a n d the N a t i o n a l Science Foundation (Grants OCE76-000156; OCE78-20453; OCE80-20290) is gratefully ackn o w l e d g e d for the w o r k since 1976. Studies by
413
Karl a n d D r a k e were s u p p o r t e d by N a t i o n a l Science F o u n d a t i o n G r a n t s G A - 4 0 0 4 9 a n d GA-22842. We also a c k n o w l e d g e with g r a t i t u d e the c o m m e n t s a n d reviews by Drs. R. G. D o u g l a s , R. B. D u t t o n , D. J. W. Piper a n d Tj. van Weering.
References ALMOGI-LAB1N, A. 1982. Planktic foraminiferal biofacies in the southern California Borderland. Geol. Soc. Am. Prog. Abst., 14, 431 pp. BEER, R.M. & GORSLINE, D.S. 1971. Distribution, composition and transport of suspended sediment in Redondo Submarine Canyon and vicinity (California). Marine Geol., 10, 153-75. BENNETT, R.H. & NELSEN, T.A. 1982. Seafloor characteristics and dynamics affecting geotechnical properties at shelf breaks. Bull. Am. Ass. Petrol. Geol., 65, 899 pp. CACCHIONE, D.A. & SOUTHARD, J.B. 1974. Incipient sediment movement by shoaling internal gravity waves. J. geophys. Res., 70, 2237--42. CROUCH, J.K. 1981. Northwest margin of California Continental Borderland; marine geology and tectonic evolution. Bull. Am. Ass. Petrol. Geol., 6 5 , 191-218. DAVIS, C.C. 1980. LA NDSA T Image Analysis of Circulation and Suspended Sediment Transport; Cal([brnia Continental Borderland. Unpubl. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 216 pp. DOUGLAS, R.G. 1981. Palaeoecology of continental margin basins--a modern case history from the Borderland of Southern California. In: Depositional Systems of Active Continental Margin Basins. Pacific Section Soc. econ. Paleo. Min. Short Course, 121-56. DRAKE, D.E. 1971. Suspended sediment and thermal stratification in the Santa Barbara Channel, California. Deep Sea Res., 18, 763-9. 1972. Distribution and transport of suspended matter. Santa Barbara Channel, California. unpubl. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 358 pp. 1976. Suspended sediment transport and mud deposition on continental shelves, h~: Stanley, D.J. & Swift, D.J.P. (eds.), Marine Sediment Transport and Environmental Management. John Wiley, New York. 127-58. -& GORSLINE, D.S. 1973. Distribution and transport of suspended particulate matter in Hueneme, Redondo, Newport and La Jolla Submarine Canyons, California. Bull. geol. Soc. Am., 84, 3949-68. --, KOLPACK, R.L. & FISCHER, P.J. 1972. Sediment transport on the Santa Barbara-Oxnard Shelf, Santa Barbara Channel, California. In: Swift, D., Duane, D. & Pilkey, O. (eds), Shelf Sediment Transport. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 307-31. DUNBAR, R.B. & BERGER,W.H. 1981. Fecal pellet flux to modern bottom sediment of Santa Barbara Basin (California) based on sediment trapping. Bull. geol. Soc. Am., 92, 212-18.
EDWARDS, B.D. & GORSLINE, D.S. 1978. New evidence of current winnowing activity on Hueneme Sill, California Continental Borderland. Bull. Am. Ass. Petrol. Geol., 62, 511 pp. EMERY, K.O. 1960. The Sea of Southern California. John Wiley, New York. 366 pp. -& GUNNARSON, e.G. 1973. Internal swash and surf. US Natl. Acad. Sci. Proc., 70, 2379-80. & HULSEMANN, J. 1962. The relationships of sediments, life and water in a marine basin. Deep Sea Res., 8, 165-80. FAUCHALD, C. & JONES, G.F. 1976. Variation in community structures of shelf, slope and basin macrofaunal communities of the southern California Bight. Bureau of Land Management Contract AA550-CT6-40, Washington, DC. 67 pp. FIELD, M.E. 1981. Deposition on Pacific shelf edge. Zone of contrasts. Bull Am. Ass. Petrol. Geol., 6 5 , 924 pp. FLEISCHER, P. 1970a. Mineralogy of Hemipelagic Basin Sediments, California Continental Borderland. Unpubl. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 208 pp. 1970b. Mineralogy and sedimentation history, Santa Barbara Basin, California. J. sed. Petrol., 42, 49-58. GIBBS, R.J. 1981. Sites of river-derived sedimentation in the ocean. Geology, 9, 77-80. GORSLINE, D.S. 1980. Deep-water sedimentologic conditions and models. Marine Geol., 38, 1-21. - 1981. Fine sediment transport and deposition in active margin basins. In: Depositional Systems of Active Margin Basins. Pacific Section Soc. econ. Palaeo. Min. Short Course. 39-60. , in press. Some thoughts on fine sediment production and transport. Sed. Geol. -& PRENSKY, S.E. 1975. Palaeoclimatic inferences for late Pleistocene and Holocene from California Borderland Basin sediments. Q. S., Roy. Soc. N.Z. 147-54. HARTMAN, O. 1955. Quantitative survey of the benthos of San Pedro Basin, southern California, Part I, preliminary results. Allan Hancock Pacific Expeditions, University Southern California, 19, 185 pp. - 1960. Quantitative survey of the benthos of San Pedro Basin, southern California; final results and conclusions. Allan Hancock Pacific Expeditions, University of Southern California, 19, 428 pp. HEEZEN, B.C., HOLLISTER,C. & RUDDIMAN,W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science, 152, 502-8. HEIN, F.J., in review. Controls on geotechnical properties of fine-grained mass movement deposits, California Continental Borderland. J. sed. Petrol. HICKEY, B.M. 1979. The California Current system--
4I 4
D.S. Gorsline et al.
hypotheses and facts. Progress in Oceanography, 8, 191-279. HOFMANN, E.E., KL1NCK, J.M. & PAFFENHOFER,G.A. 1981. Concentrations and vertical fluxes of zooplankton fecal pellets on a continental shelf. Marine Biol., 61,327-35. INGLE, J.C., JR. 1980. Cenozoic palaeobathymetry and depositional history of selected sequences within the southern California Continental Borderland. Cushman Fd. Spec. Publ., 19, 163-95. KARL, H.A. 1976. Processses Influencing Transportation and Deposition of Sediment on the Continental Shelfl Southern California. Unpubl. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 331 pp. 1980. Influence of San Gabriel Submarine Canyon on narrow-shelf sediment dynamics, southern California. Marine Geol., 34, 61-78. , DRAKE, D.C. & CACCHIONE, D.A., in press. Response of the suspended sediment transport system to continental shelf dynamics. Geo-Marine Letters. KOLPACK, R.L. 1971. Biological and oceanographical survey of the Santa Barbara Channel oil spill 1969-1970--v. I I, Physical, chemical and geological studies. Allan Hancock Foundation, University of Southern California, Los Angeles. 477 pp. 1979. Distribution of suspended particulate matter near sewage outfalls in Santa Monica Bay, California. In: Palmer, H.D. & Gross, M.G. (eds), Ocean Dumping and Marine Pollution. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 205-39. LEOPOLD, L.B., WOLMAN, M.G. & MILLER, J.P. 1964. Flm, ial Processes in Geomorphology. W.H. Freeman & Co., San Francisco. 522 pp. MALOUTA, D.N., GORSL1NE, D.S. & THORNTON, S.C. 1981. Processes and rates of basin filling in an active transform margin; Santa Monica Basin, California Borderland. J. sed. Petrol, 51, 1077-95. MCCAVE, I.N. 1972. Transport and escape of fine grained sediment from shelf areas. In: Swift, D., Duane, D. & Pilkey, O. (eds), Shelf Sediment Transport. Dowden, Hutchinson and Ross, Stroudsburg, Pa. 225-48. MOORE, D.G. 1969. Reflection profiling studies of the California Continental Borderland--structure and Quaternary turbidite basins. Geol. Soc. Am. Spec. Pap., 107, 142 pp. NARDIN, T.R. 1981. Seismic Stratigraphy of Santa Monica and San Pedro Basins, California Borderland--Late Neogene History of Sedimentation and Tectonics. Unpubl. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 295 pp. --, HE1N, F.J., GORSLINE, D.S. & EDWARDS, B.D. 1979. A review of mass movement processes, sediment and acoustic characteristics and contrasts in slope and base-of-slope systems versus canyon-fanbasin floor systems. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. PaleD. Min. Spec. Pub. 27, 61-73. PAO, G.A. 1977. Sedimentary History of California Continental Borderland Basins as Indicated by Organic Carbon Content. Unpubl. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. i 12 pp. PIPER, D.J.W. 1978. Turbidite muds and silts on -
-
-
-
-
deep-sea fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Trenches and Fans. Dowden, Hutchinson and Ross, Stroudsburg, Pa. 163-76. PIRIE, D.M. & STELLAR, D.B. 1977. California coastal processes--LA NDSA T H, Final Report. NASA Goddard Space Flight Center, Greenbelt, Maryland. 146 pp. REID, J.L. JR, RODEN, G.I. & WYLL1E,J.G. 1958. Studies of the California Current System. Calif. Coop. Fish. Investigation, Prog. Rpt. July 1, 1956:January 1, 1958.27-56. RICHARDSON, M.D. & YOUNG, D.K. 1980. Geoacoustic models and bioturbation. Marine Geol., 38, 205-18. RODOLFO, K.S. 1964. Suspended Sediment in Southern California Waters. Unpubl. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 91 pp. 1970. Annual suspended sediment supplied to the California Borderland by the southern California watershed. J. sed. Petrol., 40, 666-71. SAVRDA, C.E. 1982. Refinement of Euxinic Biofacies Models: California Continental Borderland. Unpubl. MS. Thesis, Univ. of So. Calif., Los Angeles, California. 180 pp. --, BOTTJER, D.J. & GORSLINE, D.S. 1982. Refinement of benthic basin biofacies models: biogenic sedimentary structures and macrofauna from Santa Monica Basin, California Continental Borderland. Geol. Soc. Am. Prog. Abst., 14, 608 pp. SCHWALBACH, J.R. 1982. Sediment Budget for the California Continental Borderland--Northern Region. Unpubl. M. S. Thesis, Univ. of So. Calif., Los Angeles, California. 190 pp. SHEPARD, F.P. & DILL, R.F. 1966. Submarine Canyons and other Sea Valleys. Rand McNally & Co., Chicago. 381 pp. --, MARSHALL, N.F., McLOUGHL1N, P.A. & SULLIVAN, G.G. 1979. Currents in Submarine Canyons and other Sea Valleys. Am. Ass. Petrol. Geol. Studies in Geology, 8, 173 pp. SIMPSON, J.J., DICKEY, T.D. & KOBLINSKY, C.J., in press. Mesoscale eddies in the California Current and their effects on climate. Am. Meteorological Soc. Intl. Conference on Meteorology and air-sea interaction of the coastal zone; 1982. The Hague, Netherlands. SOUTAR, A. & CRILL, P.A. 1977. Sedimentation and climatic patterns in the Santa Barbara Basin during the 19th and 20th centuries. Bull. geol. Soc. Am., 88, 1161-72. SOUTHERN CALIFORNIA COASTAL WATER RESEARCH PROGRAM 1980. Biennial report for the years 1979-1980. SCCWRP, Long Beach, California. 363 PP. SVERDRUP, I-t.U. & ALLEN, W.E. 1939. Distribution of diatoms in relation to the character of water masses off southern California in 1938. J. mar. Res., 2, 131-44. THORNTON, S.C. 1981a. Holocene Stratigraphy and Sedimentar)' Processes in Santa Barbara Basin-Influence of Tectonics, Oceanography, Climate and Mass Movement. Unpubl. Ph.D. Dissertation, of So. Calif., Los Angeles, California. 351 pp. 198lb. Suspended sediment transport in surface -
-
-
Fine-grained sediment transport processes and products in California waters of the California Current off southern California 1977-78 floods. Geo-Marine Letters, 1, 23-8. TSUCmVA, M. 1980. Inshore circulation in the southern California Bight, 1974--1977. Deep Sea Res., 27A, 99-118.
415
WILDHARBER, J.L. 1966. Suspended Sediment Over the Continental Shelf off Southern California. Unpubl. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 98 pp.
D.S. GORSLINE and R.L. KOLPACK, Department of Geological Sciences, University of Southern California, Los Angeles, California 90089 USA. H.A. KARL and D.E. DRAKE, US Geological Survey, 345 Middlefield Road, Menlo Park, California 94025 USA. S.E. THORNTONl, J.R. SCHWALBACH2, and C.E. SAVRDA,Department of Geological Sciences, University of Southern California, Los Angeles, California 90080-0741 USA. P. FLEISCHER, Sea Floor Division (Code 361), Naval Ocean Research & Development Activity, NSTL Station, Mississippi 39529 USA. 1. Now at Shell Oil Company, Box 527, Houston, Texas 77001, USA. 2. Now at EXXON Corp., Los Angeles, California, USA.
Fine-grained sediments associated with fan lobes: Santa Paula Creek, California R. Bourrouilh and D.S. Gorsline SUMMARY: Plio-Pleistocene slumps, debris-flow deposits, turbidites and hemipelagic sediments accumulated in the Ventura Basin to thicknesses of almost 2 kin. The section exposed along Santa Paula Creek, California, contains at least six units of thick sandstones and conglomerates that represent fan lobes deposited at times of either tectonic pulses or minor climatic changes. Sections exposed in Adams Creek about 5 km west of the Santa Paula section suggest that the Santa Paula section was located on the fringes of these lobes. Gross sedimentation rates for the entire section are of the order of I m/1000 yrs. This is similar to the accumulation rates in the contemporary Santa Barbara Basin. The bulk of the intervening mudstones, siltstones and fine sandstones are thin-bedded (less than 10 cm thick) and represent interdistributary and basin plain facies. Mudstones are often laminated and the relatively sparse evidence of bioturbation suggests that the basin bottom waters were anoxic. Open connection over sills is evidenced by the discovery of the skeletal parts of a whale. The section illustrates the structures and associations of fine-grained sediments in a deep-water turbidite basin of high accumulation rate.
Some of the largest Tertiary petroleum reservoirs of the southern California area are found in deep-water sands deposited in deep structurally controlled basins. These basins are the product of transform-margin tectonics operating during Neogene time (Crowell 1974, 1976; Crouch 1979), and are directly related to the development of the California Borderland which is the present expression of that margin (Crowell 1974; Blake et al. 1978; Nardin 1981). This evolution began about 30 my ago when the North American Plate met the East Pacific Rise (Atwater 1970; Atwater & Molnar 1973), segments of which progressively met the subduction zone and ceased to exist as an active spreading centre. This event caused major lateral motion to predominate and initiated the formation of the Californian Block (Anderson 1971), situated west of the San Andreas Fault Zone. In this paper, we are concerned with sedimentation in the Santa Barbara-Ventura Basin, one of a series of basins that evolved as transform motion continued. This east-west oriented basin was a long, deep trough (c. 1000 m according to palaeoecological data, Ingle 1978)which received sediments by various mass gravity movements, t.urbidity currents and pelagic settling. Several submarine canyons incised the northern slope of the basin and built small fan lobes towards the basin axis. Periods of higher sedimentation rates and introduction of sands and gravels alternated with periods of fine hemipelagic sedimentation onto what may have been an anoxic basin floor. Fine sedimentation also prevailed along the flanks of active lobes and in the overbank deposits associated with the main distributary chan-
nels. We have concentrated on the characteristics of the finer grained deposits in this paper. Studies of the coarse-grained fan lobes are reported in a paper in preparation.
Tectonic and sedimentary controls Tectonic controls on basin development in the southern California region includes (1) tectonic influence of the lateral displacement of the Californian Block along the sheared edge of the North American Plate in two main directions (a) north-west--south-east parallel to the San Andreas Fault system, and (b) north 90-110 ~ west, parallel to the Transverse Fault system which has possibly increased in importance with time during the period of interest; (2) tectonic control exerted by compressive and distensional forces generated both by transform tectonics and by basement structural effects (Crowell 1976). The Neogene and Recent basins of southern California, and the sediments deposited in them, have been further influenced by the style of tectonic activity which can occur both as discontinuous events (seismic activity) due to individual fractures and earthquakes, and as a more continuous process of regional compression or extension due to stresses exerted over long time periods as a result of large-scale plate motions or basement displacements (Fig. 1). Vertical forces generated by the above mentioned lateral forces and by isostatic adjustments due to loading and unloading of blocks are also present. In addition, it has been suggested that a trend of long term 417
4T8
R. Bourrouilh and D.S. Gorsline
~.-,,~ ..,..
~o
+I
~t
~
/'
/
t
.~z
9
o
X"
I
\
~ I I
/ _~
I
<~ ,,,
~, /
I
.
o~ ~,
',J~
'~/1 ~
-
Ic$ ~
~/
"
I~
~~-- ,,,'<<<>" II~~ ,, I ~" - / Y ; 7
L<s~ ....
~
_~.~_2~-
II ~s..~ ,.-. d
4" 4" 4' 4' u_
Q
%
1
4"
4" 4. 4" 4" 4" *I4" §
I,LI Z "r"
0
I,-,
c.) izl
=I
o II
<
~"
|!
.I-
b
~
4-
/
/
>.
II:/
LIJ
[/~ I y,,,,;-
zI
,
{ e..~Ill
,!I"
ili:
-
/
C
/ 6
i
Fine-grained sediments associated with fan lobes cooling was initiated with the demise of the overwhelmed spreading centre from about midMiocene times, which has caused a slow regional downward sinking (Pilger & Henyey 1979; Damon 1980). On a local basin scale, rift basin motions probably create both uplift and depression as in the example of Crowell's rift basins (1974). It is now well known that the transfer of sands from coastal and shelf sources to the deep basin floor occurs via submarine canyons, and that the fans built at the canyon mouths by these accumulations assume morphologies that are the result of such factors as sediment supply, texture of the supply, sea-level and shelf width, climate, basin relief and areal dimensions and tectonics (Gorsline & Emery 1959; Mutti 1977, 1979; Walker 1978; Normark 1978a; Gorsline 1980). Numerous models have been proposed (Haner 1971; Piper & Normark 1971; Middleton & Bouma 1973; Mutti & Walker 1973; Kruit et al. 1975; Mutti 1979; Nilsen 1980). In addition to this well-known process system, sands and muds on slopes may be moved to base-of-slope and basin floor by mass wasting processes of many scales; which, in turn, may generate turbidity currents of high mud concentrations from failure of slope deposits of high water content and low strength. Recently, Nardin et al. (1979) have discussed the processes and forms of such deposits and cite examples from the California Borderland. Bennett & Nelson (1983) have written a more recent and authoritative review which further discusses these processes and their effects on slopes. Other Borderland examples are given by Gorsline (1978), Haner & Gorsline (1979) and Field & Edwards (1980). Damuth & Embley (1979) have reported Atlantic examples as have a number of other authors (e.g. Embley 1976; Jacobi 1976). Almagor (1982) has discussed interesting large-scale gravity processes on the Israel margin. Further discussions of the slope environment and the dominant gravity processes that operate there are given in the symposium volume edited Doyle & Pilkey (1979). Nardin (1981) has discussed some mass-movements on fans in the inner California Borderland basins. All of these processes transfer sediments from land sources down the margin slopes to the deep ocean floor. The instability of slope deposits and the magnitude of the movements after failure will be primarily influenced by slope gradient and accumulation rate (sediment supply). Since the major regional slope generating processes are tectonic, the relation between tectonism and the types of processes responsible for basin sedimentation on an active transform margin is primary.
419
Climatic effects influence the processes through their influence on rate and texture of sediment supply, and upon sea-level position.
Structural setting and stratigraphy Ventura Basin
This basin (Fig. 1) is located north-west of Los Angeles and forms an east-west syncline within the Transverse Ranges province. The syncline is bordered to the north by the elevations of Sulfur Mountain and Santa Paula Ridge. The southern boundary is formed by the South Mountain and Oak Ridge positive elements. A double line of fractures borders the synclinc. The strike of these is essentially east-west and they have controlled the structural and sedimentologic evolution of the region since lower Miocene time. To the north arc the fractures of the Pine Mountain Fault, Santa Ynez and San Cayetano Thrusts; to the south are the Oak Ridge and Santa Susanna Fractures. These faults are at least Cretaceous to Eocene in age. They are situated exactly on the line of extension of the transform faults of the Murray Fracture Zone of the adjacent deep Pacific floor and the Garlock Fault on land. The sedimentation history of Ventura Basin is ancient and includes at least two successive cycles, First, from Cretaceous to Eocene time, rapid subsidence led to deposition of 7.6 km of marine sediments. The basin filled rapidly and by Eocene time was probably filled to sea-level. During Oligocene time, a period of continental deposition occurred that is represented by about 0.8 km of relatively coarse-grained sediments. Second, in Miocene to mid-Pleistocene time along the preceeding zone of subsidence, a new phase of south-north tensional subsidence occurred forming a basin commencing in lower Miocene time with an east-west axial trend (Crowell 1976). The depth of the basin is estimated from benthic microfaunal data to have been about 1500 m. Syn-sedimentary subsidence led to the accumulation of 8 to 9 km of sediments from Miocene to Pleistocene time (Fig. 2). During Miocene time, compressive forces produced folding on the southern border of the basin (Nagle & Parker 1971). In the Pliocene and early to middle Pleistocene time the basin was in tension (Crowell 1976). Pliocene and later sediments are slightly transgressive and discordant over the underlying Miocene section. The PlioPleistocene sediments are from 3-4 km thick (Jennings & Troxel 1954) and constitute probably
42o
R. Bourrouilh and D.S. Gorsline
o
E
~6
o
x~
Eo Z
>
+,a 0
0
g i
s e.i
Fine-grained sediments associated with .fan lobes the thickest accumulation of these sediments in the world. These are dominantly turbidites and deep water hemipelagic mudstones (Natland & Kuenen 1951; Crowell et al. 1965). The early to middle Pleistocene deposits are conformable on the Pliocene and are characterized by the same facies types. During the middle Pleistocene, compression caused by the Pasadenan Orogeny produced folds and faults. This phase of compression rejuvenated the faults bordering Oak Ridge (north strikes) and produced strong vertical movements on the San Cayetano Thrust. The deformed Plio-Pleistocene beds dip at high angles to the south-east. Santa Paula Creek Section (Plio-Pleistocene) Within the Ventura Basin succession, the classical section of Santa Paula Creek (Natland 1933; Natland & Kuenen 1951) exposes the deformed Miocene sediments below the tilted succession of Plio-Pleistocene mainly turbiditic beds. The younger beds strike west-north-west and they were folded by the Pasadenan Orogeny. The dip is 50'-60 -J to the south-east and their stratigraphic thickness is about 3.6 km. Sedimentation was apparently continuous in a shallowing marine basin (1500 to 300 m in depth) (Natland 1933; Blake 1979). The sediments are mainly mudstones, siltstones and sandstones, containing deep-water foraminifera (Natland 1933), and secondarily, intervals of coarser-grained thickerbedded conglomerates and sandstones (Crowell et al. 1966; Johnson 1978) (Fig. 3). Johnson (1978) examined about half (1.8 kin) of this section and has documented seven megasequences of coarse-grained sediments that occur at irregular intervals in the hemipelagic mudstones and thin-bedded fine-grained turbidites that constitute the deep basin floor deposits of the ancient Ventura Basin. These seven megasequences, of varying thicknesses, are formed of sandstones and siltstones, together with thickbedded conglomerates and pebbly mudstones in mega-sequences II and VI. Mega-sequences composed of medium to very thick layers of conglomerates and sandstones were interpreted by Johnson as channel fill assemblages overlying interdistributary or fan margin mudstones and thin-bedded distal turbidites. Blake (1979) has examined the microfossil assemblages in the (hemipelagic) mudstones and suggest that most of the section measured by Johnson is early to middle Pleistocene with possibly some very late Pliocene in the lowest measured mega-sequence. Blake's correlations would suggest that the section described in this paper includes perhaps 600 000 yrs of record, all
4 21
younger than 1.2 my BP (the age of the Bailey ash in other sections). The total section measured by Johnson may extend back to almost 2 my BP SO that the first six mega-sequences he described averaged about 100000 yrs in separation, although they are not that regularly spaced. The gross accumulation rate is thus about 1 m/1000 yrs which is comparable to those in the contemporary Santa Barbara Basin (Soutar & Cril11977; Schwalbach 1982).
Sediments Above the coarse conglomeratic beds that constitute the lower part of mega-sequence VI (Johnson 1978) the upper portion of the Santa Paula Creek section includes four zones (Fig. 3). These include: (1) 185 m of fine sediments consisting of fine-grained thin turbidites and muds; (2) 47 m of coarser deposits composed of conglomerates and sands (base of mega-sequence VII of Johnson 1978); (3) 796 m of fine-grained thin turbidites with mudstones, and (4) 70 m of sands and conglomerates which form the uppermost deposits of the section and are unconformably overlain by later Pleistocene shallow water and subaerial deposits. The microfaunal data, based on about a dozen evenly spaced samples over the measured section, indicate that this time is one of general shallowing from about 1500 to less than 300 m (Blake 1979). Zone 1 The 185 m of sediments of Zone 1 (Fig 4) are composed mainly of mudstones containing foraminifera indicating a depth greater than 270 m, and of sands and fine sands in thin beds with classic turbidite structures (Bouma 1962). These sands and granule sands commonly appear in millimetre-thick beds showing primary distal lamination separated by millimetre-thick mud layers. More rarely these beds attain thicknesses of a few centimetres. The coarser sandy beds and granule sands often lie over a base of granulesized particles in successively obliquely stratified beds. These units have thicknesses of up to 10 cm and represent a rapid deposition on the interdistributary areas or on the basin floor. Locally, ripple marks appear on the upper surface of these thicker sands, It appears that the currents which deposited these sands maintained the same average direction during the period of deposition, but varied in intensity so that internal erosion periodically modified the general deposition of the cross-laminae. Mudstones, which appear homogeneous in outcrop and have thicknesses of up to 30 cm,
422
R. Bourrouilh and D.S. Gorsline
FIG. 3. Measured stratigraphic section at Santa Paula Creek, California. Roman numerals identify sequences cited in the text. (Johnson 1978; R. Bourrouilh & D.S. Gorsline). Ornament shows alternation of coarse-grained (stippled) and fine-grained (horizontal shading) turbidite facies.
Fine-grained sediments associated with fan lobes
423
FIG. 4. Photograph of Zone 1 sediments. Laminated fine to medium-grained sands alternating with laminated muds. when X-radiographed, commonly show oblique lamination in units with individual thicknesses of 1 mm or more. Parallel lamination is characteristic of the finer mudstones. We interpret the parallel lamination of the fine silts and muds to be mainly the product of climatic effects rather than solely due to deposition from distal turbidites. These alternations could have been regular as seasonally deposited beds, or could also be the result of more exceptional climatic events such as major floods every decade or so. Similar features have been observed in Santa Barbara Basin (Drake et al. 1972) and are the result of major floods at intervals of 10-30 yrs. Other exceptional events could have been the reworking of the narrow shelves of the northern source lands by storm waves (Emery 1960), or simply the product of normal nepheloid transport as observed in Santa Barbara Basin today (Drake 1972; Soutar & Crill 1977). In some silts and very fine sands the
oblique stratification appears to be the product of waning turbidity current deposition, characteristic of the outer fan or fan fringe. Weak bottom currents could also have been active. These contributions from the oxygenated shelf or slope contrast with the anoxic basinal environment indicated by dark laminated mudstones with no evidence ofbioturbation. These organic-rich beds may act as source rocks for hydrocarbons, as several oil seeps are found in the area of the section. Zone 2
Sandwiched between the two thicker fine-grained sequences (Zones 1 and 3) there are 47 m of thick-bedded ( > 1 m) conglomerates and sandstones which we have called Zone 2 (Figs 3 and 5). Overlying a weakly scoured basal contact we identify six main phases of deposition. These
424
R. Bourrouilh and D.S. Gorsline
(a)
(b)
Fine-grained sediments associated with .fan lobes
425
(c) FIG. 5. Photographs of Zone 2 sediments. (a) General view of massive sandstones. (b) Detail of sediments (upper part of Zone 2) showing dark mudstones with plant debris (above the coin), overlain by alternating silty mudstones and mudstones parallel lamination, and occasional ripple formation, locally deformed by thixotrophy. (c) Fine-grained sediments between thick turbidite sandstones. include: (1) an 9.5 m thick fining and thinning upward unit from conglomerates through sandstones with shale rip-up clasts to dark organicrich mudstones; (2) a 5 m thickening-upward sandstone sequence; (3) 6.25 m of coarse sandstones and conglomerates; (4) another thickening-upward sequence comprising 12.5 m of sandstones with slumps towards the top; (5) 4.75 m of finer-grained parallel-laminated sandstones; and (6) a 9 m thick fining and thinning-upward sandstone-mudstone sequence. We interpret the 47 m thick Zone 2 as the relatively distal part of a coarse-grained submarine lobe. There was a sharp incursion of material over the fine-grained basinal or fan fringe sediments of Zone 1, and then a rather more gradual return to the similar facies of Zone 3. This sudden increase in the rate of supply and the grain-size of the material may have been caused by tectonic activity, climatic change, lowered sea-level or a shift in position of the submarine lobe. Work in progress suggests that the main axis of the lobe was several kilometres west of the Santa Paula Creek section and we may be looking at fluctuation in fan dimension in a peripheral section. If so, then the expansion of the larger fan is likely to have been influenced by tectonic or climatic
factors rather than by simple lobe or channel avulsion. Nardin (1981) has described the expansion and contraction of fan zones in the Hueneme-Mugu Fan in Santa Monica Basin in the California Borderland during Wisconsin time. Similar changes with respect to sea-level fluctuations may be the forcing function in the Santa Paula Creek section. Shallowing of the basin floor as evidenced by foraminiferal data would suggest that sedimentation was more important than tectonism. In that case, climatic/sea-level changes may be the driving force. Zone 3
This zone includes 796 m of thin-bedded turbidites and mudstones (Figs 3 and 6). Although the change from the preceeding facies of Zone 2 was progressive, it nevertheless took place over a relatively short interval. Depth decreased from 600 to 270 m up section (G. Blake, pers. comm.). At this time the basin was probably anoxic, with deposition of very dark fine-grained deposits rich in organic debris and plant matter, but was open across its sills to the ocean. A precise indication of this connection was found in the discovery of a skeleton of the whale Megaptere, about 570 m
426
R. Bourrouilh and D.S. Gorsline
(a)
(b) FIG. 6. Photographs of Zone 3 sediments. (a) Parallel and oblique lamination in fine-grained sandstones and mudstones. (b) Climbing ripple lamination and coarse basal layer.
Fine-grained sediments associated with fan lobes above the base of this third zone. Analysis of the position of the skeleton shows that the whale sank rapidly after death and lay on the sea floor on its back. It is not badly disassociated which suggests an absence both of bottom scavenger activity and of strong currents. The sediment accumulation rate was fast enough so that the remains were apparently rapidly buried. The species was migratory and lived in the open ocean and shows no post-depositional transport. The basin was probably much like the contemporary borderland silled basins. More than 80% of Zone 3 sediments are mudstones with thin intercalated sandy and silty layers. The sands and silts exhibit the classic turbidity current depositional structures over a much reduced interval; graded bedding passes to cross lamination and then to parallel lamination. Thicknesses of the sand and silt beds range from a few millimetres to a few centimetres and these are separated by thick mudstones. The silts display a variety of oblique lamination types: (1) oblique laminae characteristic of the limit of laminar flow; (2) laminae formed by traction and erosion transport; (3) laminae accompanied by thixotropic structures; (4) laminae formed by fluctuating flow characteristics. This oblique lamination is probably formed either by reworking of the substrate by periodic bottom currents or slope currents, or by turbidity current overbank flows from channels across the adjacent levees. The mudstones are also laminated and some show oblique lamination in X-radiographs. At certain horizons, coarser sediments are intercalated (Fig. 7). For example, about 240 m above the boundary with Zone 2, up to 20 thick diamictites were deposited in beds averaging 10-20 cm in thickness (maximum 90 cm). Most of these beds contain gravels and soft sediment clasts; some have slump structures and thixotropic features. Our interpretation is that these sediments were deposited in a silled open basin with a diminishing sediment supply. The rare thicker diamictites may have been generated by slumps on the upper slope and resedimented as debris-flows and turbidity currents. Zone 4
Zone 4 comprises 70 m of conglomerates, pebbly mudstones and sandstones with minor siltstones and mudstones (Figs 8 and 9). The transition from Zone 3 mudstones is relatively abrupt (Fig. 3). We have divided the zone into five sequences FIG. 7. Detailed section of part of Zone 3 (240 m above base of zone) showing coarser-grained interbeds.
427
428
R. Bourrouilh and D.S. Gorsline
o N
.o
06
Fine-grained sediments associated with fan lobes
429
.=.
o
~o
:Z
r,g3
0
0
0
0
.E
0 N
Y. o 0 9
o
430
R. Bourrouilh and D.S. Gorsline
including: (1) a 12 m thickening-up and coarsening-up conglomeratic sequence, the last bed attaining a thickness of 5 m, overlain by a finer-grained thickening-up turbiditic sequence ending with two conglomerate beds; (2) a 12 m thickening-up conglomeratic sequence containing rolled pebbles with long axes of up to 60 cm; (3) 18 m of sandstone and conglomeratic turbidites, with some pebbly mudstones containing dispersed pebble layers; (4) 12 m of sandstoneconglomeratic turbidites (several over 3 m thick) with interbedded mudstones; and (5) a 17-18 m thickening-up conglomeratic sequence with a maximum bed thickness of 1.5 m. In detail, the beds of Zone 4 show a number of interesting characteristics. The conglomerates are mostly mud-supported or, more rarely, sand-supported. The pebbles are of heterogeneous composition from mixed sources, rounded and elongate (up to 60 cm) perhaps indicating a period of fluvial transport. Both turbiditic and debris-flow sandstones and siltstones are identified, with sand dikes, sand and mud volcanoes, flutes, ripple marks and 'hard-grounds'. Fine-grained turbiditic mudstones and thicker-bedded pebbly mudstones also occur. Many of the beds are organicrich with much terrigenous plant debris, and this appears to have matured sufficiently in some sections to yield hydrocarbons and oil seeps. It appears that these sediments have been deposited more rapidly than those of Zone 2. The thickness of beds, reduced slumping, high rates of sediment supply and benthic fossil evidence indicates a shallowing, low-gradient environment in which sedimentation has outpaced tectonic changes. Water depths appear to have been about 200-250 m during this period. Channels may have been meandering in form. Neptunian sand dikes and sand or mud volcanoes may be indicative of seismic activity and shock wave remobilization of the debris and mud flows.
Conclusions The Ventura Basin was a narrow east-west trough bordered to north and south by fault systems (Fig. 10). During Plio-Pleistocene time prior to the Pasadenan orogeny, both the northern and southern margins appear to have been tectonically active leading to basin and downwarping (Natland & Kuenen 1951; Crowell et al. 1966; Crowell 1976). Both tectonic effects and sea-level fluctuations were probably responsible for the frequent generation of mass-movement and turbidity current flows. Data presented in this paper show that the Ventura basin was probably in open communication with the ocean over shallow sills, and with other basins to the east. The characteristic background sedimentation comprises mudstones and fine-grained turbidites. Within these finer grained deposits, thick-bedded sandstones and conglomerates occur at irregular intervals forming mega-sequences up to 50 m thick. These are made up of smaller-scale sequences that are both fining-upward, coarsening-upward and irregular. The episodic occurrence of deep water sands in the section, the lack of well-defined channels and the evidence of rapidly-accumulated poorlysorted sediments indicates rapid shifts in depositional locale, relatively steep slopes and rapid transfer to the basin floor. This may have been the result of tectonic events or shifts in submarine lobes or variations in sea-level related to climate. Several canyon sources may have been active. Work on other parts of the Ventura Basin further to the west (Hsu et al. 1980) suggests that several canyon-fan systems were present at intervals along the long east-west northern margin of the basin, rather than one or two large points of influx. The section exposed in Santa Paula Creek is probably offset by a few kilometres from the main part of the lobe or fan axis. N
~ ~ O ~ S A N i
C~IEL
-'~
FAUL
T
S
( ~ NARROW EAST'WEST TROtK]V~ BORDERED TO NORTH A ~ SOUTH BY FAULT SYSTEM
COM'~UNICATION WITH OCEAN OVER SHALLOW SILLS|
Ir"=::E ;~ " " - : ~ Il ~ ' ~ I1 ~'~t~.~'~l~..',w.~,,.\ rI ~ ~ ~ ~ [ ,L.._
,
~
~
-~.--~=~i
'* " - , ~ _ . . , = . ~ ~ _ ~ ; ~ C ' ~ - - ~ ~ z ! ~ ! _ ~
~
il
i
i
, ]' I' I '
I '
/J,
I
I
' I
l
I
S0PPOSI~
~oqO~d~
BASEVENT ~ P I N G
~.
/
not to scale
FIG. 10. Interpretation of Ventura Basin tectonic setting and evolution.
Fine-grained sediments associated with .fan lobes In Pliocene time the Ventura Basin may have been similar in form and pattern to the presentday Santa Barbara Basin (Thornton 1981, Fig. 42). The presence of slumps in the megasequences, and the evidence for rapid deposition show that both canyon-fan and slope-centred systems were operative (Gorsline 1978). Fan development may have been mainly as lobes rather than as fully developed fan systems (Mutti 1979). The sand content of the adjacent river basins may have been high and Mutti's low-efficiency fan model may be most applicable. It is likely then that each sequence represents the progradation of a new lobe as either transport paths shifted or as new cycles of sea-level or tectonism occurred. The periodicity of roughly 100 000 yrs between lobes suggests that tectonic or climatic change (sea-level change) are more likely than lobe shifts which might be expected to occur with greater frequency. Certainly at times, fine-grained sedimentation dominated the section and thus indicates times of either reduced supply or lower gradient. The strong influence of sedimentation rate is well documented in the contemporary California Borderland basins where the surface circulation controls the suspended sedimentation (Gorsline 1980; Thornton 1981) aided by biological aggregation and pelleting (Gorsline 1983). In the areas of high sedimentation the deposits accumulate faster than the tectonic rates can deform them and lenses or nearly horizontal and conformable sands and muds form the basin fill. In the contemporary borderland, the basins seaward of the inner zone of high sedimentation rate typically contain warped layers indicating that tectonism is operating at rates comparable to the accumulation rates and may exceed those rates.
431
Santa Paula Creek section represents a passage from times of high accumulation and tectonic activity to times of tectonic quiet and decreased sedimentation rate. The influence of sea-level change may also be present. The range of frequencies of important climatic cycles probably goes up to 105 yrs while the important tectonic cycles probably begin there and extend into periods of 106 yrs and more for large-scale global cycles. The rapid shoaling of the basin at the top of the Plio-Pleistocene section may represent dominance of sedimentation but in a regime of decreasing rates of both sedimentation and tectonism. This was ended by the renewed orogenic activity of the Pasadenan Orogeny of mid-Pleistocene time (c. 500 000 yrs BP). The fine-grained sediments in the basin were deposited by hemipelagic processes, turbidity currents and associated mass-flows. Thinlylaminated dark organic-rich shales represent periods of sedimentation in an anoxic basin, perhaps recording seasonal cycles or major floods. Pebbly mudstones and obliquelylaminated siltstones are undoubtedly the product of distal turbidity current deposition and massmovements. The mudstones with intercalated fine sand and silt turbidites less than 1 cm in thickness may be intrafan deposits of the lower fan or basin floor. The section illustrates the multiplicity of fine-grained sediment types and the broad range of transport processes that deliver them to the basin floor, ACKNOWLEDGEMENTS: We thank the National Science Foundation and the D G R S T for financial support. Dorrik Stow reviewed and revised an earlier version of the manuscript.
References ALMAGOR, G. 1982. Marine geotechnical studies of continental margins: a review--Part II. Appl. Ocean Res., 4, 130-50. ANDERSON, D.L. 1971. The San Andreas Fault. Sci. Am., 225, 53-68. ATWATER, T. 1970. Implications of plate tectonics for
the Cenozoic tectonic evolution of western North America. Bull. geol. Soe. Am., 81, 3513-36. - - & MOLNAR, P. 1973. Relative motion of the Pacific and North American Plates deduced from sea-floor spreading in the Atlantic, Indian and south Pacific Oceans. In: Kovach, R.L. & Nin, A. (eds), Proc. Conference on Tectonic Problems of the San Andreas Fault System. Stanford University Pubs. Geol. Sci., 13, 136-48. BENNETT, R.H. & NELSEN, T.A. 1983. Seafloor characteristics and dynamics affecting geotechnical properties at shelf breaks. In: Stanley, D.J. & Moore,
G.T. (eds), The Shelf Break: Critical Interface on Continental Margins. Soc. econ. Paleo. Min. Spec. Pub. 33, 333-55. BLAKE, G.H. 1979. Palaeobathymetric and Depositional History of Basinal Sediments, Santa Paula Creek, California. Unpub. Research Paper, Univ. of So. Calif., Los Angeles, California. 14 pp. BLAKE, M.C., CAMPBELL,R.H., DIBBLEE, T.W., HOWELL, D.G., NILSEN, T.H., NORMARK,W.R., VEDDER, J.C. & SILVER,E.A. 1978. Neogene basin formation in relation to Plate-Tectonic evolution of San Andreas Fault System, California. Bull. Am. Ass. Petrol. Geol., 62, 344-72. BOUMA, A.H. 1962. Sedimentology of some Flysch Deposits. Elsevier, Amsterdam. 168 pp. CONREV, B.L. 1959. Sedimentary History of the Lower Pliocene in the Los Angeles Basin, CaliJornia. Un-
432
R. Bourrouilh and D.S. Gorsline
pub. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 268 pp. CROUCH, J.K. 1979. Tectonic history of the outer part of the California Borderland. Geol. Soc. Am. Abst. Pros., 11, 407. CROWELL, J.C. 1974. Origin of late Cenozoic basins in California. In: Dickinson, W.R. (ed.), Tectonics and Sedimentation. Soc. econ. Paleo. Min. Spec. Pub. 22, 190-204. - 1976. Implications of crustal stretching and shortening of coastal Ventura Basin, California. In: Howell, D.G. (ed.), Aspects of the Geologic History of the California Continental Borderland. Pacific Section. Am. Ass. Petrol. Geol., Misc. Publ. 24, 365-82. --, HOPE, R.A., KAHLE, J.E., OVENSHINE, A.T. & SAMS, R.H. 1966. Deep-water sedimentary structures Pliocene Formation, Santa Paula Creek, Ventura Basin, California. Spec. Report 89, Cali[i Dir. Mines and Geol., 40 pp. DAMON, P.E. 1980. Continental uplift at convergent boundaries. In: Plateau Uplift: mode andmechanism. Lunar and Planetary Laboratory Topical Conference Proc. Flagstaff, Arizona. DAMUTH, J.E. & EMBLEY,R.W. 1979. Mass-wasting on the continental rise of eastern South America. Bull. Am. Ass. Petrol. Geol., 63, 438. DOYLE, L.J. & PXLKEY, O.H. 1979. Geology of Continental Slopes. Soc. econ. Paleo. Min. Spec. Pub., 27, 374 pp. DRAKE, D.E., KOLPACK, R.L. & FISCHER, P.J. 1972. Sediment transport on the Santa Barbara-Oxnard shelf, Santa Barbara Channel, California. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 307-30. 1972. Distribution and Transport of Suspended Matter, Santa Barbara Channel, California. Unpub. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 358 pp. EMBLEY,R.W. 1976. New evidence for the occurrence of debris flow in the deep sea. Geology, 4, 371-4. - 1980. The role of mass transport in the distribution and character of deep-ocean sediments with special reference to the North Atlantic. Marine Geol., 38, 23-50. EMERY, K.O. 1960. Sea Off Southern California. John Wiley, New York. 366 pp. FIELD, M.E. & EDWARDS, B.D. 1980. Slopes of the southern California Continental Borderland: a regime of mass transport. In: Colburn, I., Bouma, A., Field, M., Douglas, R. & Ingle, J. (eds), Quaternary Depositional Environments of the Pacific Coast. Pacific Section. Soc. econ. Paleo. Min. Symposium 4, 169-84. GORSLINE, D.S. 1978. Anatomy of margin basins. J. sed. Petrol., 48, 1055-68. -1980. Deep-water sedimentologic conditions and models. Marine Geol., 38, 1-21. - - & EMERY,K.O. 1959. Turbidity current deposits in San Pedro and Santa Monica Basins off southern California. Bull. geol. Soc. Am., 70, 279-88. HANER, B.E. 1971. Morphology and sediments of Redondo Submarine Fan, southern California. Bull. geol. Soc. Am., 82, 2413-32.
--
& GORSLINE, D.S. 1979. Processes and morphology of continental slope between Santa Monica and Dume Submarine Canyons, southern California. Marine Geol., 28, 77-87. HARTNETT, T.M. 1980. Vertical Sequence Analysis of Late Pliocene Pico Formation Sediments in Adams Canyon, Ventura County, California. Unpub. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 184 pp. HEIN, F.J. & WALKER,R.G. 1982. The Cambro-Ordovician Cap Enrage Formation, Quebec, Canada: a coarse-filled submarine channel. Sedimentology, 29, 309-30. Hsu, K.J., KELTY, K. & VALENTINE,J.W. 1980. Resedimented facies in Ventura Basin, California and model of longitudinal transport of turbidity currents. Bull. Am. Ass. Petrol. Geol., 64, 1034-51. INGLE, J.C. 1980. Cenozoic paleobathymetry and depositional history of selected sequences within the southern California Continental Borderland. Cushman Found. Spec. Publ., 19, 163-95. JACOBI,R.D. 1976. Sediment slides on the northwestern continental margin of Africa. Marine Geol., 22, 157-73. JENNINGS, C.W. & TROXEL, B.W. 1954. Ventura Basin. In: Jahns, R.H. (ed.), Geology of Southern California. Calif. Div. Mines and Geol. Bull. 170, Guide No. 2. 63 pp. JOHNSON, B.A. 1978. Vertical Sequence Analysis of a Deep-Sea Fan System, Santa Paula Creek, California. Unpub. M.S. Thesis Univ. of So. Calif., Los Angeles, California. 204 pp. KRUIT, C. 1975. Une excursion aux c6nes d'alluvions en eau profonde d'fige Tertiare pres de San Sebastian. International Sed. Congress. Guide pour l~xcursion Z-23, Nice, 75 pp. M1DDLETON, G.V. & BOUMA,A.H. 1973. Turbidites and Deep-water Sedimentation. Pacific Section, Soc. econ. Paleo. Min. Short Course Notes, 158 pp. M1DDLETON, G.V. & HAMPTON, M.A. 1972. SEDIMENT GRAVITYFLOWS:MECHANISMSOFFLOWANDDEPOSITION. In: Middleton, G.V. & Bouma, A.H. (eds), Turbidites and deep-water sedimentation. Pacific Section Soc. econ. Paleo. Min. Short Course Notes, 1-38. MUTTI, E. 1977. Distinctive thin-bedded turbidite facies and related depositional environments in the Miocene Hecho Group (south central Pyrenees, Spain). Sedimentology, 24, 107-31. - 1979. Turbidites et c6nes sous-marins profonds. In: Homewood, P. (ed.), Skdimentation Dbtritique (fluviatile, littorale et marine). 36me Cycle Romand en Sciences de la Terre, Univ. Fribourg, Switzerland. 353-419. & RICcI-LUccHI, F. 1972. Turbidites of the northern Apennines: Introduction to facies analysis. lnternl. Geol. Rev., 20, 125-66. -& WALKER, R.G. 1973. Turbidite facies and facies associations. In: Middleton, G.V. & Bouma, A.H. (eds), Turbidites and deep-water sedimentation. Pacific Section. Soc. econ. Paleo. Min., Short Course Notes. 119-58. NAGLE, H.E. & PARKER, E.S. 1971. Future oil and gas potential of onshore Ventura Basin, California. In:
Fine-grained sediments associated with .fan lobes Cram, I.H. (ed.), Future petroleum provinces of the U.S., Am. Ass. Petrol. Geol. Mem., 15, 254-97. NARDIN, T.R. 1981. Seismic Stratigraphy of Santa Monica and San Pedro Basins, California Continental Borderland." Late Neogene History of Sedimentation and Tectonics. Unpub. Ph.D. Dissertation. Univ. of So. Calif., Los Angeles, California. 295 pp. --, HEIN, F.J. & GORSL1NE, D.S. 1979. A review of mass movement processes, sediment and acoustic characteristics and contrasts in slope and base-ofslope systems versus canyon-fan-basin floor systems. In: Doyle, L.J. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleo. Min. Spec. Pub. 27, 61-73. NATLAND, M.L. 1933. Depth and temperature distribution of some Recent and fossil Foraminifera in the southern California region. Bull. Scripps Inst. Oceanography, La Jolla, California. - & KUENEN, PH.H. 1951. Sedimentary history of the Ventura Basin, California and the action of turbidity currents. In: Turbidity Currents and the Transportation of Coarse Sediment to Deep Water. Soc. econ. Paleo. Min. Spec. Publ. 2, 76-107. NILSEN, T.H. 1980. Modern and ancient submarine fans: discussion of papers by R.G. Walker and W.R. Normark. Bull. Am. Ass. Petrol. Geol. Bull., 64, 1094-101. NORMARK, W.R. 1978a. Fan valleys, channels and depositional lobes on modern submarine fans: characters for recognitions of sandy turbidite environments. Bull. Am. Ass. Petrol. Geol., 62, 912-31. - 1978b. Turbidites, time and again. Bull. Am. Ass. Petrol. Geol., 62, 549.
433
PILGER, R.H. & HENYEY, T.L. 1979. Pacific-North American Plate interaction and Neogene volcanism in coastal California. Tectonophysics, 57, 189-209. PIPER, D.J.W. & NORMARK, W.R. 1971. Re-examination of a Miocene deep-sea fan and fan valley, southern California. Bull. geol. Soc. Am., 82, 1823-30. SCHWALBACH, J.R. 1982. A Sediment Budget ,/'or the Northern Region of the California Continental Borderland. Unpubl. M.S. Thesis, Univ. of So. Calif., Los Angeles, California. 212 pp. SLOSSON, J.E. 1958. Lithofacies and Sedimentary Paleogeographic Analysis of the Los Angeles Repetto Basin. Unpubl. Ph.D. Dissertation, Univ. of So. Calif., Los Angeles, California. 128 pp. SOUTAR, A. & CRILL, P.A. 1977. Sedimentation and climatic patterns in the Santa Barbara Basin during the 19th and 20th centuries. Bull. geol. Soc. Am., 8 8 , 1161-72. THORNTON, S.E. 1981. Suspended sediment transport in surface waters of the California Current off southern California: 1977-78 floods. Geo-Marine Letters, 1, 23-8. WALKER, R.G. 1978. Deep-water facies and ancient submarine fans: models for exploration for stratigraphic traps. Bull. Am. Ass. Petrol. Geol., 62, 932-66. WINN, R.D. & DOTT, R.H. 1979. Deep water fan channel conglomerates of late Cretaceous age, southern Chile. Sedimentology, 26, 203-28.
R. BOURROUILH,Laboratoire de Ge61ogie, Universit6 de PAU, Avenue Philippon, F.64000 PAU, France. D.S. GORSLINE, Department of Geological Sciences, University of Southern California, Los Angeles, California 90007, USA.
Origin of varve-type lamination, graded claystones and limestone-shale 'couplets' in the lower Cretaceous of the western North Atlantic A.H.F. Robertson SUMMARY: The petrography and sedimentary structures of pelagic and hemipelagic sediments in the early Cretaceous interval of the Blake-Bahama Basin reveal a subtle interplay between terrigenous source materials, plankton productivity and diagenesis. Three specific sediment types drilled at Site 534A (Leg 76) of the DSDP are: (1) persistent, fine parallel lamination resembling varves; (2) thin, size-graded nannofossil claystones and 'black shales', and (3) 'couplets', composed of pale burrowed and dark laminated nannofossil radiolarian carbonates. The varve-type lamination reflects some combination of fluctuations of terrigenous plant material plus fine clastic input and plankton productivity. Both possibly reflect short periodicity (10s to 100s of years) climatic variations. Lithification took longer in the clay- and organic-rich laminae. Slow compaction then accentuated the lamination. By contrast, purer nannofossil oozes lithified more quickly and now show less compaction. The graded claystones and 'black shales' were mostly redeposited by turbidites from within, or near, the oxygen-minimum zone on the upper continental slope. Turbiditic input reduces the need for bottom water anoxicity. The dark-light couplets reflect wetter periods on land averaging 20-60 000 years, during which abundant plant material entered the Atlantic. Contrary to some recent suggestions, the early Cretaceous Atlantic need not have been poorly circulated and weakly oxygenated. Instead, both short and longer periodicity climatic fluctuations controlled input of mostly land-derived material. High rates of organic matter input, partly turbiditic, produced sub-surface sediment anoxia below a circulating relatively fertile early Cretaceous Atlantic Ocean. Many recent studies of depositional processes in the oceans have concentrated on particular mechanisms, for example, attempts to distinguish distal turbidites from contourites (Stow 1979). On the other hand, it is becoming increasingly clear that sedimentation in the deep oceans must reflect a complex interplay of variables which include source, oceanography, climate and diagenesis. This is particularly true of the abyssal basins adjacent to passive margins. Here the role of depositional and diagenetic processes in the lower Cretaceous interval of Blake-Bahama Basin in the Western North Atlantic is examined. Particularly, the lower Cretaceous deposition at Site 534A, drilled on DSDP leg 76, is discussed along with evidence from related drill sites (Fig. 1). During early Cretaceous times (Berriasian-Barremian) Site 534A was located above the calcium compensation depth (CCD) at water depths of c.3.5 km (back-tracking; Sheridan et al., 1982). Sediments accumulated in an abyssal basin around 70 km east of a major carbonate platform, now the deeper unexposed parts of the Blake Plateau. Site 534A shows an interplay between pelagic accumulation and both terrigenous and bioclastic input from the continental margin (Sheridan et al. 1982; Sheridan et al. 1984). Attention is focused on the nature and origin of
three specific sediment types which are widespread in the lower Cretaceous Atlantic and have been noted elsewhere in the oceans and on land: (1) persistent fine parallel lamination resembling varves; (2) thin graded nannofossil claystones and black shales, and (3) cyclical alterations of pale burrowed and dark laminated nannofossil carbonates ('couplets'). Explanation of these features depends on a combination of fluctuations of surface productivity, elevated terrestrial input, and gravity redeposition, plus the effects of differential diagenesis. Contrary to some earlier views there may be little need to invoke a well stratified poorly oxidized ocean, at least in the early Cretaceous Atlantic.
Regional setting Throughout the north-western Atlantic the lithologies of latest Jurassic (Tithonian) to early Cretaceous (Hauterivian) age have been named the Blake-Bahama Formation, which is predominantly composed of pelagic carbonates (Jansa et al. 1979). Similar successions extend into the eastern North Atlantic (Cape Verde Basin) (Lancelot et al. 1978) and into the western Tethys as the Maiolica (Bernoulli 1972). During Leg 76, 392 m of the Blake-Bahama Formation were drilled, 437
438
A.H.F. Robertson
BLAKE BAHAMA FORMATION e ~
j
4~176
M .3~
/ ~
J
/
40~
. 387: / ~ ; BERMUDA
t BAH_A.MA / 11 "" \ BASIN ] , , / t ~ t-i.391[323,-335,,]---'~I- .I
"N
I
\
I
L,o,
30~
'
-20~
20 ~
Be I
7r I
6r I
FIG. 1. DSDP sites where early Cretaceous pelagic limestones have been drilled in the western North Atlantic. Thickness of the Blake-Bahama Formation shown in brackets. Horizon B correlates approximately with the top of the Blake-Bahama limestones: the eastern boundary indicates pinch-out against basement, near magnetic anomaly M-11, interpreted as Valanginian (modified after Jansa et al. 1978). of which 267 m were recovered. A fuller account of the sedimentology of the succession at Site 534A is given in the Initial Report for Leg 76, including clay mineralogy, major element chemistry and carbon-oxygen isotopic analyses (Robertson & Bliefnick 1984). The discussion in this paper centres on the Valanginian, Hauterivian and Barremian, excluding the Aptian-Albian higher interval of the lower Cretaceous. One advantage of studying the pre-Aptian lower Cretaceous is that anoxicity in the sediment is only incipient, so that the various competing processes of input and diagenesis are highlighted. The generalized descriptions of DSDP cores published in the Initial Reports often obscure important sedimentary details, particularly small-scale sedimentary structures. Close examination of core intervals
can reveal a complex interplay of competing sedimentary processes.
Lithologies Using DSDP nomenclature the following lithologies are recognized in the lower Cretaceous interval: (l) nannofossil limestones; (2) marly nannofossil chalk; (3) nannofossil claystones; (4) black organic-rich claystones and shales; (5) both bioclastic and terrigenous redeposited elastics. The relative abundance of each of these lithologies is summarized in Fig. 2. The lower part of the succession, of early Berriasian age, is dominated by nannofossil limestones, which decrease upwards and disappear by the end of late Valanginian time, while marly nannofossil chalks
Varve-type lam&ation, graded claystones and limestone-shale 'couplets'
CORE NUMBER 47
0
50
439
100%
49
5o
il Barremian
52 53 54 55 56
...
'
57 / / / "4/"~''-'~:'b~" 5958 5a .'.~.".~.~~ L.__ 60 .... 62 63 64
,
tl Hauterivian ' ~]
i 9 "
"'.i ." .' ".
6867 69 7O 5b 71 72 73 74 75 76
Early Va langinian
78 79 5c 8O 82
83 84 85 86 87 88 89 9o 91
Va langinian
Late Berriasian 5d
Early Berriasian ~
Marly nannofossilchalk (Undifferentiated) Palelaminated Dark laminated
Quartzose,siltstoneand sandstone Sandylimestone Nannofossillimestone
Claystone
Black shale
FIG. 2. Summary of relative abundances of lithologies in the early Cretaceous interval drilled at Site 534A. The quartzose siltstones and fine-grained sandstones and sandy limestones were deposited by turbidity currents. Note the relative abundance of dark organic-rich claystones ('black shales'). Marly nannofossil chalks, both burrowed and laminated give way to nannofossil limestones in early Berriasian. Compiled from shipboard Visual Description Sheets.
44 ~
A.H.F. Robertson
become more abundant. The marly nannofossil chalks range from massive, to highly burrowed, relatively well-cemented, to finely laminated, more argillaceous, and softer facies. The finely laminated marly nannofossil chalks are present both as paler and darker intercalations. The darker laminated intervals are richer in organic matter, shown by Habib (1979) to be mostly terrestrial in origin. The nannofossil claystones exist as thin size-graded units ( < 5 cm) reaching a relative maximum in late Valanginian time. The black carbonaceous claystones and 'black shales' are present throughout the succession but become most abundant above lower Berriasian, where they locally reach 25% of the sediment volume. The redeposited clastics are restricted to the late Valanginian to Barremian, becoming dominant in early Hauterivian (Fig. 2). Similar lithologies, including varve-type lamination and alterations of nannofossil, chalk, claystone and limestone, have also been encountered in the lower Cretaceous interval of the Western North Atlantic on the Lower Continental Rise Hills (Site 105, Ewing & Hollister 1972), on the Blake-Bahama Outer Ridge (Site 101, Ewing & Hollister 1972), on the Bermuda Rise (Site 387, Tucholke et al. 1979) and in the Blake-Bahama Basin at Site 391, located 22 km NW of Site 534A (Benson et al. 1978). Additionally, a very similar succession is seen in the Eastern North Atlantic in the Cape Verde Basin (Site 367, Lancelot et al. 1978). Comparisons show that carbonate-claystone 'couplets' and varve-type lamination extended across the lower Cretaceous Atlantic from the lower Continental Rise, towards the Mid-Atlantic ridge, and again appear in the Eastern Atlantic basin. These features are obscured where gravitydeposited clastics dominated the succession, for example off Morocco. Any explanation of the couplets and varve-type lamination must explain their wide distribution throughout the ocean not closely related to a particular continental margin. We now focus, in turn, on the origin of: (1) the fine parallel varve-type lamination; (2) the thinbedded nannofossil claystones and black shales; (3) the dark-light nannofossil chalk 'couplets', and (4) the role ofdiagenesis. The descriptions are based on detailed observation of cores from Site 534A in the Blake-Bahama Basin, supplemented by regional comparisons.
Varve-type lamination Ubiquitous fine lamination in the marly nannofossil chalks (Fig. 2) first appears in the Berriasian and persists to Aptian time at both Sites 534A and 391 in the Blake-Bahama Basin. The finer lamina-
tion in both the paler and darker facies (Fig. 3(a), (b), (c)) was examined petrographically. In thin section the lamination in the paler nannofossil chalks could rarely be resolved at all. Thicker (to 1 ram) paler laminations are sometimes composed of slightly flattened radiolarian shells replaced by calcite, set in a micritic matrix (Fig. 4(d), (e)). Lamination in the darker facies tends to be more clearly defined. At high magnification, individual laminations are typically seen to comprise numerous minute flecks of translucent brownish organic matter, plus occasional pyrite and dolomite in a micritic matrix. In detail, the parallel lamination often exhibits a micro-lenticular or streaky texture (Fig. 4(a)). In some cases the organic matter is seen to be associated with relative enrichment of mica and fine quartzose silt. Radiolarian shells may either follow individual fine laminations (Fig. 4(e)) or may be scattered randomly. The fine lamination is hard to resolve even with the scanning electron microscope, but a well-defined platy fabric of preferentially orientated calcareous nannoplankton can sometimes be seen (Fig. 4(b)). This fabric contrasts strongly with the near-random fabric seen in the intercalated nannoplankton claystones which are much less recrystallized (see below, Fig.
4(c)). The varve-like lamination has been noted in other DSDP cores particularly at Site 391, 22 km NW of Site 534 in the Blake-Bahama Basin (Benson et al. 1978), at Site 387 on the Lower Bermuda Rise Hill (Tucholke et al. 1979), at Site 101 at the south end of the Blake Plateau outer ridge (Ewing & Hollister 1972), at Site 535 in the Florida Straits (Buffler et al., in press), and in the eastern North Atlantic at Site 367 in the Cape Verde Basin (Lancelot et al. 1978). In addition, varve-type sediments have also been drilled by the DSDP in mid-upper Cretaceous successions elsewhere, including the South Atlantic (Bolli et al. 1978) and around Mid-Pacific Mountains and the Southern Hess Rise (Dean & Claypool 1981). The fine lamination at Site 387 (e.g. Cores 46, 47) on the Bermuda Rise is particularly important as this site was located close to the spreading ridge during early Cretaceous time. On the Lower Continental Rise Hills (Site 105) lamination is disrupted by extensive syn-sedimentary deformation not seen in oceanward sites. At Site 391 Freeman & Enos (1978) were able to count up to 67 laminations per 1 cm of sediment thickness, equivalent to one lamination per 6 years on average. At Site 534A only up to 30 laminations per centimetre could be detected, equivalent to 1 lamination per 15 years, assuming sedimentation rates of 21 cm/my (Sheridan et al. 1984).
Varve-type lamination, graded claystones and limestone-shale 'couplets'
441
FIG. 3. Photographs of early Cretaceous sediment cores from Site 534A, Blake Bahama Basin. (a) Highly bioturbated marly nannofossil limestone with subordinate intercalations of organic-rich varve-type nannofossil claystone. The limestones are much harder than the organic-rich laminae. Core 7, Section 4, 53-88 cm (Valanginian). (b) Smooth alternation between paler and darker varve-type laminated marly nannofossil chalk. The 1-2 cm thick massive units with some burrowing (e.g. lower) are graded turbiditic nannofossil claystones. Note the absence of burrowing in the varve-type laminations. Core 72, Section 5, 84-118 cm (Valanginian). (c) Graded nannofossil claystones turbidites (massive dark grey) alternating with subordinate finely laminated sediment. From the interval of Late Valanginian to Hauterivian high terrigenous input. Core 61, Section 3, 21-55 cm.
Mode of deposition
TurbMites
Ideally, fine lamination in the ocean could relate to one or a c o m b i n a t i o n of: (1) B o u m a Tb, Td divisions of classicial turbidites; (2) contourites; (3) nepheloid-flow and drift deposits; (4) suspension-cascading of lutites; (5) variations in terrestrially-derived clastics, organics or nutrients; (6) plankton productivity variations; (7) diagenesis. Taking these alternatives in turn:
A turbiditic origin is easily ruled out; in the Bouma sequence, the laminations could only belong to the Td division, but neither the interturbidite Bouma Te nor other divisions are seen. Regionally, there is no sign of any obvious proximal-distal variations related to any particular source. Similar lamination is seen in the lower Cretaceous on both sides of the Atlantic and on the palaeo-ridge flanks.
442
A.H.F. Robertson
FIG. 4. Photomicrographs and scanning electron micrographs of early Cretaceous facies at Site 534A. (a) Typical well cemented marly nannofossil limestone. Note that the radiolaria are replaced by calcite but not flattened. Many tests were partly replaced by pyrite early in diagenesis. Core 79, Section 2, at 74-76 cm. (b) Marly nannofossil chalk. Calcareous nannofossils heavily overgrown with blocky calcite crystals. Core 71, Section 3 at 80-82 cm (c) Nannofossil claystone turbidites. Note the excellent preservation and random orientation of calcareous nannofossils. This contrasts sharply with the platy fabric and greater recrystallization of the intercalated marly nannofossil chalks. Core 48, Section 2, 185-167 cm. (d) Marly nannofossil chalk. The radiolarian shells have been replaced by calcite and flattened. Core 68, Section 5,55-58 cm. (e) Typical 'varve'-laminated marly nannofossil chalk. The white blobs are slightly flattened radiolaria replaced by calcite. Core 48, Section 3, 50-52 cm. (f) Laminated quartzose silt of turbiditic origin. In contrast to the 'varves' this lamination is defined by quartzose silt and is associated with coarser grained turbidites. Core 69, Section 7, 7-10 cm.
Varve-type lamination, graded claystones and limestone-shale 'couplets' 443 Contourites
Bottom currents, or specifically contour-currents, might rework sediment to form a fine lamination. Indeed in the Atlantic lower Cretaceous at least some bottom water motion is implied by the preferential orientation of grains in the marly nannofossil chalks. Reworking by bottom currents to form the fine lamination is, however, unlikely: the laminations are regular and laterally continuous at least within cores. Ripples, crosslamination, scours or other sedimentary structures normally associated with current deposits (e.g. Stow 1979) are not seen. Unlike the Holocene contourites of the Blake Outer Ridge (Heezen et al. 1966; Hollister & Heezen 1972), the lamination is not restricted to the lower continental rise or any other specific local setting, but instead occurs over wide areas of the abyssal plain on both sides of the Cretaceous Atlantic. The fine lamination also contrasts with drilled contourite deposits, for example Site 533 (Leg 76) on the Blake Outer Ridge (Sheridan et al., 1984). This is an almost featureless pile of muddy nannofossil ooze with few visible sedimentary structures. Also, the seismic signature of current-drift deposits appears to be that of hummocky reflectors, seen for example, in the Blake Outer Ridge (Site 533; Sheridan et al. 1982). By contrast, reflectors in the lower Cretaceous are basin-levelling, typical of turbidites and pelagic sediments. Lutite flow
McCave (1972) developed a model in which lutite flows (low-density, low-velocity turbidites) are not able to penetrate the ocean water density structure, but instead reach a density interface, then flow out over clearer water finally losing particles by settling. This could produce an episodic 'suspension cascade' and so conceivably produce a fine lamination. One problem pointed out by McCave is that the Coriolis force would deflect such low-density turbid currents parallel to the slope and thus fail to reach the deep ocean basins. Also this mechanism might be expected to produce graded partings of similar composition rather than the sharp compositional variations seen.
Terrestrial input
Could the lamination have formed in response to seasonal or longer term pulses of terrestrial organic matter input? It was noted that terrestrially-derived organic matter defines the darker varve-type lamination, with an occasional corresponding increase in terrigenous silt. Habib (1979,
1984) has described a rich flora of pollen grains, fern spores, woody tissue (tracheids) and epidermal cuticles, plus several types of amorphous matter, mostly derived from delta back-swamp areas. Where sedimentation rates of terrestrialderived material were high, as in the Valanginian and Barremian-Aptian interval at Site 534A, the floral material is well preserved. Habib also confirms that the distribution of carbonized wood largely controls sediment blackness, consistent with the abundance of 'black shales' in the late Valanginian at Site 534A (Fig. 2). In general, during early Cretaceous time, the North Atlantic was ringed by major deltas (e.g. Jansa & Wade 1975) during a time of gradually rising sea-level (Vail et al. 1980). In a model in which pulsed input of terrestrial, organic matter produced the fine lamination, it would be assumed that finely divided organics floated out to sea, were dispersed by currents, then settled to form an organic-rich layer. On the other hand most of the terrigenous clastics would have been trapped in the delta, coastal and adjacent carbonate platform system. A modern analogue would be the much smaller Gulf of California where Calvert (1966) considered seasonal input to be the main factor forming varve-type lamination. This model for the Atlantic lower Cretaceous requires variations in terrestrial input over decades rather than annually (rare floods?). The organic matter would have had to drift up to 400 km, being dispersed by currents, then settled to form discrete laminations. An alternative suggested by Gardner et al. (1978), is that fluctuations in the input of terrestrially-derived nutrients influenced surface productivity, but the absence of any clear correlation of varve-type lamination with upwelling along the continental margin is a problem.
Productivity
The lamination is strikingly like that observed both in lakes and along continental margins influenced by upwelling. For example, Thornton (this volume) records similar varve-type lamination in the modern Santa Barbara Basin on the California continental borderland. In the Miocene Monterey Formation (Bramlette 1946), varved diatomaceous sediment is also attributed to upwelling (Soutar et aL 1981), although the nature of lamination is much more varied than in the Cretaceous Atlantic. In the early Cretaceous Atlantic progressive sea-level rise (Vail et al. 1980) may have stimulated increased marginal upwelling producing fine lamination when faecal pellets settled on the sea floor.
444
A.H.F. Robertson
The main problems with upwelling as the cause for varved sediment are: (1) the frequent correlation of the lamination with black plant material (micritic material, Habib, 1984); (2) why upwelling should have occurred at all away from the continental margins in the middle of the equatorial palaeo-Atlantic. First, the relatively low frequency of lamination could be partly an artifact of repeated erosion during deposition of the interbedded turbidites, both pelagic and terrigenous. Alternatively, the low periodicity could be due to surface blooms persisting for periods of up to decades (Funnel, In: McCave 1979). Inter-annual variation in solar activity coupled through the atmosphere to the oceans may also influence oceanic circulation patterns and thus upwelling. For example, the 'Southern Oscillation' of the atmosphere leads to inter-annual climatic variation in the tropical Pacific Ocean (Julian & Chervin 1978; Philander 1979). During the early Cretaceous the Atlantic was open to the Western Tethys (e.g. Bernoulli 1972). Whether or not a connection with the Pacific existed is uncertain. It is probable that westward surface drift issued from the Tethys. Palaeo-latitudes where varve-type sediment existed ranged simultaneously from 12-25~ outside a possible zone of upwelling due to equatorial convergence. One possibility is that flow to the Pacific was blocked and surface currents were deflected to form nutrient-rich gyres in the western North Atlantic. In summary, the fine varve-type lamination has to be attributed to some combination of variations of input of plant material and surface plankton productivity. At present it is not possible to evaluate the relative roles of these two processes. The wide distribution of the lamination including, by mid-Cretaceous, both North and South Atlantic and parts of the Mediterranean, and the presence of detectable plant material right out to the early Cretaceous MidAtlantic ridge, point to an important terrestrial control, as in the modern Gulf of California. On the other hand, the association of lamination in the paler (non-organic-rich) pelagic carbonates with calcareous nannoplankton, radiolaria and faecal pellets, plus the close similarity of the lamination with other upwelling areas, all point to upwelling variations also as an important control. In either case it seems unnecessary to invoke a stagnant, well stratified, or oxygen-depleted Atlantic ocean during early Cretaceous time. Ultimately, both plankton productivity and plant input may relate to climatic changes on land caused by fluctuations in solar activity. The two variables may thus be linked rather than operating independently.
Graded marly chalks and 'black shales' Occurring independently of the varve-type lamination are thin graded marly nannofossil claystones and 'black shales', which range in colour from greenish, greyish to black. Close examination of sedimentary structures and composition show that most of this material is not indigenous pelagic sediment but has been redeposited by turbidity currents probably from the upper continental slope, a fact which has important implications for the origin of other Atlantic Cretaceous black shales. As shown in Fig. 2 the nannofossil claystones appear in the early Berriasian and persist throughout early Cretaceous time, becoming most abundant in the late Valanginian-Hauterivian. Black claystones and 'black shales' appear in late Berriasian, and then reach a relative peak in the late Valanginian. In the cores, the claystones are typically massive thin beds mostly up to 5 cm thick showing subtle grading over the basal 1-2 cm, a massive interior, and a top often with small-scale burrowing (Fig. 3(a), (c)). Throughout most of the succession the nannofossil claystones occur separately from the coarser turbiditic elastic input. The black claystones and 'black shales' are unburrowed. In thin section the nannofossil claystones comprise argillaceous micrite with scattered radiolaria. The basal several centimetre-thick graded intervals contain calcite-replaced or pyritized radiolaria, rare benthic foraminifera, minute shell fragments, fish debris and minor fine-grained quartzose silt. Higher in the succession, where the claystones are interbedded with terrigenous elastics, the quartzose content increases. The scanning electron microscope reveals well preserved calcareous nannoplankton with an open porous texture and little recrystallization compared to the more lithified nannofossil chalks and limestones. Smectites and mixed-layer clays dominate the clay mineralogy with some chlorite and palygorskite (Fig. 4(c)). The 'black shales' and black claystones show similar sedimentary structures and petrography but contain a higher volume of organic matter (up to 2.8~), mostly carbonized debris of terrestrial origin (Habib 1979). Black organic-rich sediment is not restricted to the graded beds, but is also seen in the darker intervals of varve-type lamination described above. Redeposition from the oxygen-minimum zone
As recently summarized by Dean & Claypool (1981), anoxicity in the ocean could be promoted
Varve-type lamination, graded claystones and limestone-shale 'couplets' by one or a combination of: (1) additional surface water productivity and thus extra oxygen demand to combust organic material; (2) increased rate of terrigenous organic matter input; (3) reduced bottom circulation; (4) increased overall sedimentation rate; (5) reduced solubility of oxygen in warmer waters. Three main models have been used to explain the Cretaceous Atlantic black shales (Schlanger & Jenkyns 1976; Fischer & Arthur 1978; Dean et al. 1978). The first model involves formation of a stratified ocean with anaerobic bottom water similar to the modern Black Sea. The second model produces organicrich sediment where the oxygen-minimum zone intersects the topography, the continental slope, or oceanic edifices. The third model, which has received less attention, requires high organic matter input leading to a high oxygen demand and hence anoxicity within the sediment, while the ambient sea water remains oxidizing. All three models could potentially play a role in the Atlantic Cretaceous. A currently popular view is that during the lower Cretaceous a decreased thermal gradient between equator and pole could have reduced thermohaline circulation and the content of dissolved oxygen in the bottom waters. High productivity both on land and at sea could have widened the oxygen-minimum zone and allowed more organic matter to reach the ocean floor without combustion. In a sluggishly circulating, poorly oxidized ocean, organic matter could have had a greater chance to reach the abyssal plain and thus promote subsurface anoxicity. Habib (1979) shows that sediment colour of the black-shale facies closely follows the distribution of carbonized wood debris, which in turn is related to terrigenous input. The interpretation of Cretaceous 'anoxic events' (Schlanger & Jenkyns 1976; Fischer & Arthur 1978; Jenkyns 1980) has, however, relied heavily on the recognition of reduced in situ organic-rich layers over wide areas of the ocean floor. Thus, it is significant that the present study has shown that much of the organic-rich sediment is not indigenous but has been redeposited. The graded nannofossil claystones and black claystones occur independently of the varve-type lamination attributed above to surface productivity variations. The obvious explanation of the graded claystones is that they are relatively fine-grained distal turbidites. Deposition from nepheloid-flow would appear to be ruled out by the bed-thickness (to 5 cm) and the presence of some silt-sized material. There is however a complete contrast in sedimentary structures and composition from the coarser clastic turbidites interbedded in the Valanginian-Hauterivian interval. This shows that the thin graded beds are
445
not merely the distal tail of coarser turbidites, but instead originally consisted almost entirely of fine-grained material. The excellent nannofossil preservation is consistent with a source well above the CCD, while the absence of coarser terrigenous or bioclastic material points to a transport route separate from the claystone turbidites. The obvious site is the upper continental slope, a potential source area of much pelagic and hemipelagic sediment. The upper continental slope is normally within the oxygen-minimum zone and is bypassed by coarser clastic sediment which is transported through the canyon systems and issues onto the continental rise from pointsources. By contrast, sediment redeposited from the upper continental slope would have come from 'line-sources'. Paull & Dillon (1980) argue that the present steep profile of the Blake escarpment is due mostly to Tertiary erosion by contour-flowing currents. Although little is known about the Blake Plateau in the Cretaceous, Enos & Freeman (1978) found no evidence of a barrier reef edge on the Blake Nose, suggesting the existence of a broader, less steeply sloping margin than at present. This would have provided space for accumulation of large volumes of muddy pelagic sediments on the upper continental slope, as in the modern windward Bahama margin (Mullins & Neumann 1979). Finer grained material on the upper continental slope deposited within the oxygen-minimum zone would have been bypassed by coarser grained shelf-derived and terrigenous material, which accumulated separately on the continental rise and beyond. Along modern upwelling continental margins, the upper continental slopes are characterized by a well-developed oxygen-minimum zone. For example, along the Ivory Coast and Ghana, the intermediate waters are oxygen-depleted (about 1.5 ml/1) but not anoxic. Organic matter survives below 2000 m because of high input relative to rates of combusion (e.g. Demaison & Moore 1980). In the Bay of Bengal, despite the absence of high surface productivity an anoxic layer of intermediate water is well-developed (yon Stackelberg 1972). There, in the oxygen-minimum zone, olive-grey muds are deposited which contrast with light-brown muds deposited in deeper oxidizing waters. Thus, summarizing, the modern analogies and the sedimentary features suggest that much of the turbiditic claystone and 'black shale' was redeposited from qine-sources' within or close to the oxygen-minimum zone on the upper continental slope. To the extent to which the organic matter is redeposited, the need for anoxic bottom waters is reduced. Only development of subsurface anoxi-
446
A.H.F. Robertson
city is needed to preserve the organic matter. With high rates of input of terrestrial organic matter, burial takes the sediment rapidly from the sediment-water interface of aerobic bacterial decay into the zone of anoxic bacterial sulphate reduction (cf., Curtis 1980). Recognition of turbidity-redeposition of much organic matter still leaves the black intervals of varve-type sediment as a possible bottom water anoxicity indicator. In a sluggishly-circulating warm relatively poorly oxidized ocean, terrestrially-derived organic matter, periodically floating out to sea, dispersed by currents, then settling would indeed stand a good chance of reaching the sea floor in sufficient abundance to promote sub-surface anoxia. However, if organic matter input is sufficiently great, as appears to have been the case in the early Cretaceous, then black organic-rich layers may form in a more normal oxidizing circulating ocean. The early Cretaceous Atlantic was open to the Western Tethys and the abundance of radiolaria and nannoplankton point to active upwelling in opposition to a stable stratified ocean at this time.
Cyclical alternations: 'couplets' Two distinct types of 'couplet' are present in the late Berriasian to Hauterivian successions at Site 534A and 391: (1) alternations of paler and darker varve-type laminated marly nannofossil chalks, and (2) cycles of laminated marly nannofossil chalks alternating with harder burrowed nannofossil chalks and limestones (Fig. 3(a),(b)). Darklight laminated cycles are c.10-30 cm thick but may be completely intergradational, or comprise rapid alternations on a scale of several centimetres down to individual laminations. The relative percentage of the darker and paler laminated facies is shown in Fig. 2. The pale harder burrowed carbonates either appear gradationally, with progressive breakdown of varvetype lamination, or may appear quite sharply above either paler or darker laminated marly nannofossil chalks (see below). Overall, the burrowed intervals in the Berriasian tend to be harder and better cemented, and contain less clay and organic matter than those in Valanginian time. Burrowed intercalations are scarce around the Late Valanginian-Hauterivian boundary when clastic input was at a maximum (Fig. 2, Fig.
3(c)). To test for cyclicity the transitions between different sediment types were determined from core photographs where contacts were undisturbed. The clearly gravity-redeposited claystones and clastic turbidites were excluded. For
351 measurements, by far the most abundant transition was from dark to pale laminated marly nannofossil chalk (27%) and vice versa (25%). The next most common transition was from pale laminated marly nannofossil chalk to burrowed nannofossil chalk or limestone (15%) and vice versa (16%). The least abundant transition was from dark laminated marly nannofossil chalk to pale burrowed chalks and limestones (8.5%) and vice versa (7.4%). Time intervals
During Valanginian time, average sedimentation rates of 21 m/my suggest that the burrowed limestones appeared on average every 20000100 000 yrs and persisted for 10 000-60 000 yrs. Alternations of pale and darker laminated chalk are much more frequent, often less than 10000 yrs down to 100s of years. The two sediment types are often completely intergradational. Two alternative models can be proposed: (1) changes in bottom water chemistry; (2) long period fluctuations in terrestrial organic matter input. Changing water chemistry
McCave (1979) noted at Site 387 on the Bermuda Rise that the dark-light 'couplets' tended to be asymmetrical with a gradual increase in organic matter content then a sudden switch to oxidized burrowed sediment. He suggested that the darklight 'couplets' could relate to periodic renewal of more oxidizing bottom water. Such bottom water change could have been triggered by the 19 000 yr precession or the 41000 yr obliquity-periodicity. An immediate problem is that although bottom water change could relate to local physical barriers, oceanwide bottom water replenishment is harder to envisage, particularly since similar couplets occur on both sides of the palaeo-Mid Atlantic Ridge. Also, the transition analysis of the Site 534A succession shows that by far the most common change is from paler to darker more organic-rich laminated marly nannofossil chalks, with only rare burrowed paler intervals above the dark laminated marly nannofossil chalks (8.5% of transitions) as noted by McCave (1979). Terrestrial organic input
Dean et al. (1978) and Gardner et al. (1978) relate the cycles of pale organic matter-poor and darker organic matter-rich nannofossil carbonates to pulsing of terrestrial organic matter with a periodicity ranging from 30 000 to 50 000 yrs. Wetter
Varve-type lamination, graded claystones and limestone-shale 'couplets' periods produced more organic matter and drier times less organic matter. Such climatic variations are consistent with the hot-humid circumNorth Atlantic climate inferred by Chamley (1979). Jansa et al. (1979) and Arthur & Natland (1979) note that the same cycles tend to match the orbital periodicity. At Site 534A the transition analysis is consistent with fluctuations of organic matter on a time-scale down to hundreds of years but extending up to c.100000 yrs. On average every 20 000-60 000 yrs organic matter decreased sufficiently for bottom sediment to become oxidizing and an infauna to develop (see below), thus forming the paler burrowed carbonates. This is consistent with there being twice as many transitions of pale-to-burrowed chalk (and vice versa), than dark-to-burrowed chalk. The latter transition could result from more sudden climatic fluctuations on land.
Role of diagenesis Diagenesis has also played an important role in producing the laminated burrowed couplets. There is a difference in the degree of lithification and the preservation of microfossils between the laminated and burrowed alternations. In the better-cemented burrowed paler intervals, with less organic-matter, cementation clearly took place early in diagenesis, prior to significant compaction. The radiolarian shells are almost all spherical with little evidence of flattening (Fig. 4(a)). By strong contrast, radiolarian shells are often more flattened in the softer clay-rich and organic-rich intervals (Fig. 4(d)). The higher content of clay and organic matter has inhibited calcite recrystallization. By contrast, scanning electron micrographs (Fig. 4(b)) show an advanced state of recrystallization of calcareous nannoplankton in the harder burrowed chalks. It is thus possible to propose a model (Fig. 5) in which fluctuations of organic-matter input control the amount of dissolved oxygen within the sediment during early diagenesis. During times of lower input of organic-matter, bottom water remained sufficiently oxidizing to promote a rich infauna, turbating and virtually destroying an original varve-type lamination. Early diagenetic lithification preserved the radiolaria with little compaction. By contrast, as the organic matter input increased burrowing stopped and the fine varve-type lamination was preserved. Clay and terrigenous organic matter go together (Habib 1979) so that the reducing layers also tended to be those in which diagenetic cementation was retarded. The clay content inhibited the recrystal-
447
lization of calcareous nannoplankton and the sediment remained soft, causing radiolarian shells to be flattened during slow compaction.
Depositional model The various types of fine-grained sediments in the late Cretaceous of the Blake-Bahama Basin can be understood in terms of a complex interplay of climatic, depositional and diagenetic processes. The overall model is illustrated in Fig. 6. First, there is the effect of varying source material. During early Cretaceous time the climate was hot with fluctuating humidity (Chamley 1979) with high terrigenous input from the source a r e a s of Palaeozoic fold mountains of SE USA. Rising sea-level during the early Cretaceous is reflected in increased input of shallow water carbonate material presumably from the Blake Plateau during late Valanginian, Hauterivian and Barremian time. Both the terrigenous and bioclastic material was presumably transported through a conventional canyon system to issue on to the continental rise from 'point-sources'. By contrast, clay and organic matter-rich pelagic carbonates accumulated within, or close to, the oxygen-minimum zone on the upper continental slope and were then redeposited from linear source areas to form the thin graded layers of nannofossil claystone and 'black-shale'. The various turbiditic influxes reworked and destroyed some of the varve-type lamination thus reducing the apparent frequency. Secondly, there are the oceanographical factors. The abundance of radiolaria in the pelagic chalks points to relatively fertile surface waters and hence to upwelling. The reason for such upwelling is unknown but could relate to inflow of nutrient-rich surface water from the Western Tethys (Fig. 7), or from an accentuated upwelling along the margin during a time of sea-level rise. Ocean water instability would possibly also have been triggered by inter-annual climatic fluctuations. During early Cretaceous time (pre-Albian) bottom currents in the North Atlantic could conceivably have been sluggish and relatively oxygen-depleted, but there is no evidence that deep waters were ever really static or anoxic. Anoxicity did, however, develop in the subsurface during the many times of enhanced terrestrialorganic matter input. Conversely during times of lower input bottom faunas flourished. Thirdly, diagenesis has played an important role in forming the laminated burrowed alterations (couplets). Organic matter-poor nannofossil oozes lithified early, while additional clay plus
448
A.H.F.
Robertson
o.o
o . ~_
0
"13 o
. o _~
~."-~-
-- X ~- 0 I
r-
o
5.0
o
Q;
co
c,'F.
-~.~_
.~_
o; O0
w~
o
0 ~
0 I
I
0 ~"
I
I
I
->
l
~
~o~ ~
+~'o
....
I I I
~
o.~,
r
p.-,
"~ o ~
=~
"F_ .c.
o~•
.90 . _ ~
I I
Q;
o
"~ .~ m
T
0
"o--
-> ~'-'
J
0
I
c~
.~ =_ o
E~
o~~ \ m
o~o
o
o
o~
cO -&
~.=_ ~ "~''~u
0
~.~
~.~
0
. ~ Ob
r" "0 '~1) 0
0 ~ - ~~ - - ~
.0"0 --
..~
/
-~ ~. ~E_
O. cO
F--o
U9 . ~
(1)
~o
=
o
co ex
e-
(1) .-I
=o~ ~ ~.~
e- .-~
_~o
0
~
e~o ~~ =~
".~ ~
Varve-type lamination, graded claystones and limestone-shale 'couplets' 449 e.-
0 :~
o~
~
o:
u
oo ,,~
E
o
N E E
J
i%-,_.
.,x,X.x.
.3
',~. 9
I
~ _o
,\
~
t~
.%.
~
9
o
<
og
c_
x
E o I-.
itl 0
+
+ +
4-
4-
o
+ at-
44-
+
k
_1
2 E o
m
,,6
A.H.F. Robertson
45o
" "."
:
9 .' ....." i..
9
"..-...
.'.'-
......
:.'."
,
" "
9
,1: t 9
*'.
i
!
! I I I
IIit
......
,
9".'.. : '... :.'.:.'." 9
9
9 :
.':'...'.h
9
o.
9
9 ~
: " '.".'..'. 4
9 9 "
"=
9
9. r
' ' I
m
O @
O
O
O g
O
. D
9 O
* O
9 I
O
9
D
. , _ . _ _ ,~'~ ~ 9 "" ' ,~ag~_~__
~
. . . . . .
~ ~ ' .
9 9"".
.317_ .
PACIFIC
- - -
TETHYS
.9
0]-5,34~ ATLANTIC
I~
36"7/-2~
.,. ,.,.,,.. '
"
"
" " '
:!ii i iiii!iiiii:ii!iiiiiiii! i::iii i 9" . ' .
".'.'.'.
:..'."
" f f s D ~
-"~ - : B.''" //......
".'..'.
. " /P
" ." .'." ".'." '~:.-JJ 9149 9149
Fine
I). I--lamination ]~ .[ 9 9 ~
Surface currents
FIG. 7. Reconstruction of the central Atlantic in lower Cretaceous time (after Smith & Briden 1977), showing the locations of DSDP sites, the inferred surface circulation and probable distribution of varve-type lamination. organic matter inhibited cementation and accentuated the primary varve-type lamination. In conclusion, it is clear that pelagic deposition on ancient passive margins cannot be understood in terms of single depositional processes, but must take account of the complex interplay of competing factors including source, oceanography and diagenesis. The incidence of graded claystone and much of the 'black shales' can be explained in terms of gravity redeposition from higher on the continental margin. The dark-light couplets relate to varying inputs of terrestrial organic matter as climate changed over periods of tens of thousands of years. However, it remains hard to assess the relative importance of, on the one hand, surface productivity variations and, on the other, shorter period inter-annular variations in terres-
trial organic matter input, as factors in the formation of the distinctive varve-type lamination, seen across the lower Cretaceous Atlantic. Indeed, since climatic change on land may ultimately be linked with shifts in current circulation, possibly these two variables should be seen as operating in tandem and not independently.
ACKNOWLEDGEMENTS;This work stems from participation in DSDP Leg 76. D. Bliefnick co-authored the original description in the Leg 76 Initial Report. Helpful comments on the manuscript were received by D. Habib and L. Jansa. Mrs D. Baty drafted the diagrams. The manuscript was typed jointly by secretarial staff of the Grant Institute.
Varve-type lamination, graded claystones and limestone-shale 'couplets' 45~ References ARTHUR, M.A. & NATLAND,J.H. 1979. Carbonaceous sediments in the North and South Atlantic: the role of salinity in stable stratification of Earth Cretaceous basins. In: Talwani, M., Hay, W. & Ryan, W.B.F. (eds), Deep Drilling Results in the Atlantic Ocean: Continental margins and Palaeoenvironments. Maurice Ewing Series 3, Am. Geophys. Union, Washington, DC. 375-401. & SCHLANGER, S.O. 1979. Cretaceous 'oceanic anoxic events' as causal factors in development of reef-reservoired giant oil fields. Bull. Am. Ass. Petrol. Geol., 63, 870-85. BENSON, W.E., SHERIDAN, R.E., ENOS, P., FREER, T., GRADSTEIN, F., MURDMAA, I.O,, PASTOURET, L., SCHMIDT, R.R., STUERMER,D.H., WEAVER,F.M. & WORSTELL, P. 1978. Init. Repts. DSDP, 44. US Govt. Print. Off., Washington, DC. 1005 pp. BERNOULLI, D. 1972. North Atlantic and Mediterranean Mesozoic Facies: a comparison. In: Hollister, C.D., Ewing, J.I. et al. (eds), lnit. Repts. DSDP 11. US Govt. Print. Off., Washington, DC. 801-71. BOLLI, H.M., RYAN, W.B.G. et al. 1978. lnit. Repts. DSDP, 40. US Govt. Print. Off., Washington, DC. 1079 pp. BRAMLETTE, N.M. 1946. The Monterey Formation of California and the origin of its siliceous rocks. US geol. Survey Prof. Pap. 212, 57 pp. BUFFLER, R.T. SCLAGER,W. et al., in press. Init. Repts. DSDP, 77. US Govt. Print. Off., Washington, DC., in press. CALVERT, S.E. 1966. Origin of diatom-rich varved sediments of the Gulf of California. J. Geol, 76, 546-65. CHAMLEY,H. 1979. North Atlantic clay sedimentation and palaeoenvironment since the late Jurassic. In: M. Talwani, Hay, W. & Ryan, W.B.F. (eds), Deep Drilling Results in the Atlantic Ocean: Continental Margins and Palaeoenvironment. Maurice Ewing Series 3, Am. Geophys. Union, Washington, DC. 324-61. CURTIS, C.D. 1980. Diagenetic alteration in black shales. J. geol. Soc. Lond., 137, 189-94. DEAN, W.E., GARDNER,J.V., JANSA, L.F. & CEPEK, P. 1978. Cyclic sedimentation along the continental margin of northwest Africa, Leg 41, Deep Sea Drilling Project. In: Lancelot, Y., Seibold, E. et al. (eds), Init. Repts. DSDP, 41. US Govt. Print. Off., Washington, DC. 965-90. & CLAYPOOL, G.E. 1981. Origin of organic-carbon-rich Mid-Cretaceous limestones, Mid-Pacific Mountains and southern Hess Rise. In: Thiede, J., Vallier, T.L. et al. (eds), Init. Repts. DSDP, 42. US Govt. Print. Off., Washington, DC. 877-90. DEMA1SON,G.J. & MOORE, G.T. 1980. Anoxic environments and oil source bed genesis. Bull. Am. Ass. Petrol. Geol., 64, 1179-209. ENOS, P. & FREEMAN, T. 1978. Shallow-water limestones from the Blake Nose, Sites 390 and 392. In: Benson, W.E., Sheridan, R.E. et al. (eds), Init. Repts. DSDP, 44. US Govt. Print. Off., Washington, DC. 413-62. EWING, J.I. & HOLLISTER,C.D. 1972. Regional aspects
-
-
-
-
of deep-sea drilling in the Western North Atlantic. In: Hollister, C.D., Ewing, T.I. et al. (eds), Init. Repts. DSDP, 11. US Govt. Print. Off., Washington, DC. 951-73. FISCHER,A.D. & ARTHUR, M. 1978. Secular variation in the pelagic realm. In: Cook, H.E. & Enos, P. (eds), Deep-Water Carbonate Environments. Soc. econ. Paleo. Min. Spec. Pub. 25. 19-51. FREEMAN, T. & ENOS, P. 1978. Petrology of Upper Jurassic-Lower Cretaceous Limestones, Site 391. In: Benson, W,E., Sheridan, R.E. et al. (eds), Init. Repts. DSDP, 44. US Govt. Print. Off., Washington, DC. 463-75. GARDNER, J.W., DEAN, W.E. & JANSA, L. 1978. Sediments recovered from the Northwest African continental margin. In: Lancelot, Y., Seibold, E. et al. (eds), lnit. Repts. DSDP, 41. US Govt. Print. Off., Washington, DC. 1121-34. HABIB, D. 1979. Sedimentary origin of North Atlantic Cretaceous palynofacies. In: Talwani, M., Hay, W. & Ryan, W.B.F. (eds), Deep Drilling results in the Atlantic Ocean." Continental margins and Palaeoenvironments. Maurice Ewing Series 3, Am. Geophys. Union, Washington, DC. 420-37. --, 1984. Sedimentation rate-dependent distribution of organic matter in the North Atlantic JurassicCretaceous. lnit. Repts. DSDP, US Govt. Print. Off., Washington, DC. 781-94. HEEZEN, B.C., HOLLISTER, C.D. & RUDDIMAN, W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science, 152, 502-8. HOLLISTER,C.D. & HEEZEN,B.C. 1972. Geologic effects of ocean bottom currents: Western North Atlantic. In: Gordon, A.L. (ed.), Studies in Physical Oceanography, 2. Gordon & Breach, New York. 37-66. JANSA, L.F. & WADE, J.A. 1975. Geology of the continental margin off Nova Scotia and Newfoundland. Can. geol. Surv. Pap. 74-30, 2. 51-105. ~ - , ENOS, P., TUCHOLKE,B.E., GRADSTEIN, F.M. & SHERIDAN,R.E. 1979. Mesozoic-Cenozoic sedimentary formations of the North American Basin; Western North Atlantic. In: Talwani, M., Hay, W., Ryan, W.B.F. (eds), Deep Drilling Results in the Atlantic Ocean." Continental Margins and Palaeoenvironment. Maurice Ewing Series 3, Am. Geophys. Union, Washington, DC. 1-58. JENKYNS, H.C. 1980. Cretaceous anoxic events: from continents to oceans. J. geol. Soc. Lond., 137, 1 7 1 - 8 9 .
JULIAN, P.R. & CHERVIN, R.M. 1978. A study of the southern Oscillation and Walker circulation phenomenon. Mon. Wea. Rev., 106, 1433-51. LANCELOT, Y., SEIBOLD,E., DEAN, W.E., JANSA, L.F., EREMEEV,V., GARDNER,J., CEPEK, P., KRASHEN1NNIKOV, V., PFLAUMANN,V., JOHNSON, D., RANKIN, G. & TRABANT,P. 1978. lnit. Repts. DSDP, 41, US Govt. Print. Off., Washington, DC. 1259 pp. MCCAVE, I.N. 1972. Transport and escape of finegrained sediment from shelf areas. In: Swift, J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport: Processes and Pattern. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 225--48.
452 -
A.H.F. Robertson
1979. Depositional features of organic-carbonrich black and green mudstones at DSDP Sites 386 and 387, Western North Atlantic. In: Tucholke, B.E., Vogt, P.R. et al. (eds), Init. Repts. DSDP, 43. US Govt. Print. Off., Washington, DC. 411-17. MULLINS, H.T. & NEUMANN, A.C. 1979. Deep carbonate margin structure and sedimentation in the Northern Bahamas. Soc. econ. Paleo. Min., 27, 165-92. PAULL, C.K. & DILLON, W.P. 1980. Erosional origin of the Blake Escarpment: an alternative hypothesis. See also discussion. Geology, 8, 538-42, and discussion, Geology, 9, 339-40. PHILLANDER, S.G.H. 1979. Variability of the tropical oceans. Dyn. Atmos. Oceans, 3, 191-208. ROBERTSON, A.H.F. & BLIEFNICK,D.M., 1984. Sedimentology and origin of Lower Cretaceous pelagic carbonates and redeposited clastics, Blake-Bahama Formation DSDP Site 534A, Western equatorial Atlantic. Init. Repts. DSDP, 76. US Govt. Print. Off., Washington, DC. 795-828. SCHLANGER, S.O. & JENKVNS, H.C. 1976. Cretaceous oceanic anoxic events: causes and consequences. Geol. Mijnbouw, 55, 179-84. SHERIDAN,R.E. et al. 1982. Early history of the Atlantic Ocean and gas hydrates in the Blake Outer Ridge: Results from DSDP, Leg 76. Bull. geol. Soc. Am., 93, 876-85. -
--,
GRADSTE1N,F.M. et al., 1984. lnit. Repts. DSDP, 76. US Govt. Print. Off., Washington, DC. SMITH, A.G. & BRIDEN, J.C. 1977. Mesozoic and Cenozoic World Maps, Cambridge University Press, Cambridge. 63 pp. SOUTAR, A., JOHNSON, S.R. & BAUMGARTNER, T.M. 1981. In search of modern depositional analogies to the Monterey Formation. In: Garrison, R.E. & Douglas, R. (eds), The Monterey Formation and Related Siliceous Rocks of California. Spec. Publ. Pacific Section, Soc. econ. Paleo. Min. 123-49. STACKELBERG, U.V. 1972. Faziesverteilung in sedimenten des indischpakistanischen Kontinentalrandes (Arabisches Meer). 'Meteor' Forsch., Reihe C, 9, 1-173. STOW, D.A.V. 1979. Distinguishing between finegrained turbidites and contourites on the Nova Scotian deep water margin. Sedimentology, 26, 371-87. TUCHOLKE, B.E., VOGT, P.R., DEMARS,K., GALEHOUSE, J., HOUGHTON, R.L., KANEPS,A.G., KENDRICK,J., MCCAVE, I.N., MCNULTY, C.L., MURDAM, I.O., OKADO, H. • ROTHE, P. 1979. Init. Repts. DSDP, 43, US Govt. Print. Off., Washington, DC. VAIL, P.R., MITCHUM, R.M., SHIPLEY,T.H. & BUFFER, R.T. 1980. Unconformities in the North Atlantic. Phil. Trans. R. Soc. Lond., A294, 137-55.
A.H.F. ROBERTSON,Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, Scotland.
Hemipelagic chalks in a clastic submarine fan sequence: Miocene SW Turkey A.B. Hayward S U M M A R Y : The Miocene Salir Formation in SW Turkey is a 500 m thick dominantly terrigenous clastic succession deposited by turbidity currents and other mass-flow mechanisms on a series of small submarine fans. During intervals of non-clastic deposition hemipelagic chalk horizons up to 60 mm thick accumulated. The chalks comprise disseminated planktonic and small benthonic foraminifera dispersed in a micrite matrix with abundant poorly preserved coccolith plates and other nannoplankton fragments. Biogenic sand and silt-sized particles form up to 20% of the rock. Carbonate content ranges from 55% to 95%. The non-carbonate component comprises dispersed clay platelets and in some beds silt-sized particles. The chalks range from structureless homogeneous units, to beds which show a wide range of sedimentary structures which suggest reworking of the slowly deposited hemipelagic ooze by bottom currents of fluctuating velocity. The lateral and vertical distribution of the hemipelagic chalks reflects the area on the submarine fan on which they were deposited. Areas away from the main locus of deposition, for example interchannel areas in proximal sequences, and inactive mid-fan areas are characterized by frequent and often thick chalk beds. In active mid-fan and inner-fan channel areas hemipelagic chalks are scarce or absent.
Miocene clastic sediments of the Bey Daglari massif in SW Turkey (Fig. 1) document the emplacement of the allochthonous Antalya Complex unit onto an adjacent carbonate platform. The Miocene succession is approximately 700 m thick and comprises submarine fan deposits that pass upwards through shallow marine sediments into continental deposits. This paper focuses on the lower 500 m of the sequence, the Salir Formation of Burdigalian-Langhian (Lower Miocene) age. In the Salir Formation a thick sequence of dominantly clastic ophiolite-derived sediments deposited by turbidites and other mass-flow mechanisms is interbedded with thin chalk horizons originating from hemipelagic deposition above the carbonate compensation depth (CCD). The marked contrast in composition between the two sediment types enables the distinction between sediments originating from hemipelagic deposition and those from turbiditic deposition to be clearly made and allows sedimentary structures within sediments with an unambiguously hemipelagic origin to be described. Good palaeontological and sedimentological control allow the spatial and vertical distribution of the chalk horizons, which represent periods of non-turbiditic deposition to be related to turbidity current frequency and to the submarine fan sub-environment in which they were deposited.
Geological setting The Tauride Mountains of SW Turkey consist of a central para-autochthonous unit, the 'Tauride authochthon' (Brunn et al. 1970, 1971; Dumont et al. 1972), sandwiched between two allochthonous ophiolitic units, the Lycian Nappes (Brunn et al. 1970, 1971; Graciansky 1972; Poisson 1977) and the Antalya Complex (Robertson & Woodcock 1980) (Fig. 1); they form an extension of the Alpine orogenic belt into SW Turkey. The 'Tauride autochthon' comprises dominantly shallow marine sediments of a Liassic to lower Miocene (Aquitanian) carbonate platform with several non-sequences in the Lower Tertiary (Poisson 1977; Dumont et al. 1972). Miocene clastic sediments unconformably overlie the carbonate platform rocks. To the east, the Antalya Complex consists of five en echelon tectonic zones separated by steeply dipping structures with a strike-slip sense of movement. From west to east the tectonic zones record the transition from a Mesozoic carbonate platform (the Bey Daglari) across Mesozoic continental margin sediments into oceanic crust formed during the initial stages of continental rifting (Woodcock & Robertson 1982; Robertson & Woodcock 1981). Taken as a whole the Antalya Complex records the initiation, construction and later tectonic disruption of part of a 453
454
A.B. Hayward
Autochthonous Units Alluvium FIG. 1. Location map showing distribution of Miocene clastic sediments. Box encloses study area.
~
Miocene Clastic ~ Sediments Carbonate Platform ~
small Mesozoic-Cenozoic oceanic basin (Woodcock & Robertson 1980). The Miocene elastic sediments of the Salir Formation were derived exclusively from the Antalya Complex and record its emplacement from an easterly direction (Hayward & Robertson 1982).
The Salir Formation sedimentary facies Based on grain-size, palaeocurrent orientation and sediment type the Salir Formation sedimentary sequence can be subdivided into two facies associations. Proximal facies-association
The proximal facies-association, which is well exposed in the Akdere valley (Fig. 2), is
AllochthonousUnits Antalya Complex Lycian Nappes Eocene Flysch (Lycian Nappes)
dominated by conglomerates and coarse sandstones interbedded with subordinate mudstones and chalks. The following facies types are distinguished. (1) Clast and matrix supported conglomerates between 0.50 and 6 m thick (Fig. 2). The clasts comprise of moderate to well rounded limestone, dolerite, gabbro, chert and subordinate serpentinite and basalt rock fragments. Normal and inverse clast size grading, a-axes imbrication and the close association with turbiditic sandstones suggests that the conglomerates were deposited by mass-flows that varied between fully turbulent flows (normally graded clast supported conglomerate) and debris-flow (matrix supported disorganized conglomerate). (2) Grey-green to buffcoloured sandstones range from granule conglomerate to fine sand. The sandstones are generally poorly sorted; serpentinite, chert, dolerite, basalt and subordinate quartz and feldspar grains are cemented by
455
Hemipelagic chalks in a submarine fan sequence
i.
o
o
..:\
~
. .......~
., . . . .
9::
-.-).
.~-~:.~ ~ , t...
~
~
9. . . . . . . ,
'~
[ I; i~ I ~'-ko2 9
.
'
1. 4 "
. . .',,
'"""
~lOkm4"
I I I ~o. ""-[9-J~
i -'r-~ 9 - ,A
-.%,.,..~ [
[
)
2~
7 . : : . ,a ~.....:\
~;nl
..~ ~.t~
' ~
-L'l
..~
~_"'"~
--~'1-
9_ ..-L-~.. 9
.---:.1...,
U . -~ .
.,x
7 74
...,
,:_:,.~ :.-...-\
E
9. - - -..a
T. 7".I, . ~
m
cpc
4 Akcay
m
3 valley
cpc
10
m
2
Akdere
valley
Fro. 2. Generalized sedimentological logs illustrating the overall features of the Salir Formation sedimentation system. Inset map shows location of logs A and B are from the proximal 'inner-fan' sequence, C and D from the more distal mid-fan area. Key as in Fig. 3.
456
A.B. Hayward
carbonate. Bed thickness ranges from 0.05 to 2.80 m. Within the sandstones, parallel and ripple laminations are common. Flutes, grooves a n d rare gutter casts are present on the base of sandstone beds, along with load marks (ball and pillow structures). These features suggest deposition by turbidity currents. (3) Green mudstones. In thin section the mudstones are silty with a moderate to low clay content (60-25%). Carbonate content varies between 5% and 15%. The partially lithified brecciated nature in outcrop makes the recognition of any sedimentary structures difficult. Unit thickness ranges from 4-20 cm (average 5 cm), thicker units represent the amalgamation of several individual beds. In general appearance, the mudstones are faintly laminated to massive, with occasional evidence of grading from silt to clay. In thin section, silt laminae with sharp slightly scoured bases grade into clay over 2-3 cm. Weakly developed parallel and very low angle ripple laminations are sometimes present. X-ray diffraction indicates montmorillonite, illite, chlorite and serpentine are the main components of the clay fraction. The composition and sedimentary structures suggest deposition from low density turbidity currents similar to those described by Rupke (1975), Hesse (1975), Piper (1978), Stow & Shanmugam (1980) and Stow & Bowen (1980). (4) Beds of white chalk occur in association with the mudstones, they are described in detail below.
Distal facies-association
The proximal sedimentary succession (described above), exposed in the Akdere valley, is not exposed continuously across the Finike anticline (Fig. 2). However, identical ages, related sedimentary facies and palaeocurrent dispersal patterns indicate that the clastic sequence exposed in the Akcay valley can be correlated with the sequence to the east in the Akdere valley and was originally a continuous sedimentation system. Sequences in the west represent the development of a more distal facies association. The distal association is characterized by an absence of conglomerates (Fig. 2). Medium to coarse grained turbiditic sandstones with erosive scoured bases form poorly defined fining-upward sequences up to 5 m thick. Bundles of thick sandstones (>0.30 m) between 3 m and 15 m thick are restricted to the base and middle parts of the succession. Individual sandstone beds are between 0.30 and 3.20 cm thick, the majority exhibit well-developed Bouma cycles. Sandstones form approximately 50% of the total sequence
(Fig. 2) and are repeatedly interbedded with mudstones and chalks (Figs 3 and 4). Sedimentation system: summary
Taken as a whole the Salir Formation sedimentation system represents deposition on a series of small submarine fans passing from inner-fan to mid-fan regions from east to west. The inner-fan area (proximal facies-association) consists of amalgamated conglomeratesandstone units deposited in braided channelized areas; in interchannel areas, sequences of interbedded fine sandstones, mudstones and chalks were deposited. To the west, the mid-fan environment (distal facies-association) consists of amalgamated sandstone units deposited in shallow distributary channels and bundles of thick-bedded sandstones deposited on a series of depositional lobes. Further details of the sedimentation system are documented elsewhere (Hayward 1982; Hayward & Robertson 1982). Sedimentology of the chalk beds
General Description Chalk beds are present throughout both the proximal and distal sedimentary sequences. White to cream chalk horizons are between 5 and 60 mm thick, laterally continuous and well indurated. Bedding surfaces are either parallel or show a wave-like form with a wavelength of 10-15 cm (Figs 4 and 5). Locally, chalks show relief over underlying mudstone or sandstone beds. Tops of the chalks are generally sharp. Bases are sharp (planar or irregular) or transitional. Rarely, small mudstone rip-up clasts are present towards the base of beds.
Composition In thin section, planktonic foraminifera and rare benthonic foraminifera along with debris from both are dispersed in a matrix ofmicrite. In some cases, planktonic foraminifera (dominated by Praeorbulina sp) account for 15% by volume of the chalks. Biogenic sand and silt-sized particles form up to 20% of the rock. Carbonate content ranges from 55-95%. The non-carbonate component comprises dispersed clay platelets and, in some beds, silt-sized particles. Dispersed carbonaceous material is sometimes abundant
(~ 5%). SEM studies show the micrite matrix to be composed of equigranular calcite crystals with abundant poorly preserved coccolith plates and planktonic foraminifera fragments. Dispersed
Hemipelagic chalks in a submarine fan sequence
457
EP-
,
A-
,
~'6,,
~~ =
~
E o
~ o
~ o
~ 8 ~~ -
o~ o
u
~
,~=o
-_~ ~ ' o.=.--
-
E
u
~
~ ~ ~ ~ .~
E,':-
E
u
~
"c. o_'z,
~.o
~z
~;~
E u co
E,..2
, ~ , l ~ "~1 o~ ,-~ .~,,~-~, ~
<
. ~-~
i
"
Ii
I II ~
9
"
9.
I
I
I
I
.
~
iii
I
i
1
I
o
o
0
U
c~
/
. . . .
9 ..
,
I!,,,
/
FI6. 4. Hemipelagic chalk beds (white) interbedded with turbiditic mudstones and strongly channelized sandstone. Mid-fan area. Stick is 1 m long.
FIG. 5. Close-up of hemipelagic chalk beds (white) interbedded with turbiditic sandstones and mudstones. Note the variation in thickness of the chalks (related to the time between turbiditic events), the erosive base to the sandstone (arrowed) and the lenticular 'rippled' form of some of the chalk beds (R). Mid-fan area. Stick is 1 m long.
Hemipelagic chalks in a submarine f a n sequence clay platelets are found throughout the carbonate matrix.
459
Depositional processes for the chalks
Pelagic settling Bioturbation In the field there is no evidence of bioturbation within the chalk beds or the associated mudstones. Small grazing trails are only rarely present on the under surface of overlying mudstone horizons. In thin section areas of patchy inhomogeneous texture may indicate micro-bioturbation within the chalk beds.
Sedimentary structures Sedimentary structures within the chalk beds can be divided into two types: (1) Macro-structures delineated by some component of terrigenous material (Fig. 7): (a) lenticular siltstone/mudstone flasers between 0.5 and 1.0 cm thick; (b) asymmetric fading ripples (Fig. 7) with an amplitude of less than 1 cm; (c) small mudstone intraclasts up to 5 mm across are present at the base of some chalk beds. (2) In thin section and polished slabs, microstructures are seen as variations in grain-size and the alignment of the biogenic silt component of the chalks. Rarely, laminae are emphasized by the alignment of carbonaceous material. (a) Cross-laminated graded silt laminae (Fig. 8) are 5-10 mm thick and pass downcurrent over less than 10 cm into parallel-laminated silt laminae. Bases to these units are sharp and irregular with small scour marks. (b) Graded horizontal laminations are up to 4 mm thick (Fig. 8). Lower contacts are sharp, upper contacts are generally gradational to non-laminated chalk. Most laminae exhibit some form of normal grading; reverse grading is rarely present. (c) Pinch and swell silt laminations (Figs 6 and 8) are 1-3 mm thick and lenticular over 50 mm. Upper and lower surfaces are gradational. (d) Micrograded wispy silt laminations are up to 5 mm thick with irregular sharp bases and gradationai tops (Fig. 6). (e) Small scale cut-and-fill features are up to 10 mm across with graded fills and transitional tops (Figs 6 and 8) or more rarely non-graded homogeneous micrite fills. (f) Lags of winnowed foraminifera and nannofossil tests occur as laminae at the base of some chalk beds (Fig. 8). Within the chalks no predictable or idealized sequence of sedimentary structures is observed. In all cases, the sedimentary structures occur isolated within non-laminated structureless chalk (Fig. 6). Superimposed sedimentary structures separated by structureless chalk (Fig. 6) suggest periods of fluctuating current velocity.
The composition and texture of the chalk beds in contrast to the associated muds and sandstones indicates initial deposition from pelagic settling above the CCD. Similar chalk beds from a clastic dominated sequence have been documented both from the ancient (Weiler 1970; Hesse 1975) and from the recent (Rupke & Stanley 1974; Rupke 1975). In most of the examples documented (both recent and ancient) the pelagic (or hemipelagic) carbonate horizons have been extensively bioturbated and any primary sedimentary structures have been destroyed. In the Salir Formation, the general absence of bioturbation and the abundance of carbonaceous material in some chalk beds (normally destroyed under oxic conditions) suggests anoxic bottom conditions preventing the development of an abundant ichnofauna.
Bottom currents The variety of sedimentary structures (Figs 6, 7 and 8), some of which have been reported from sequences deposited by low density turbidity currents (e.g. Stow 1979 Stow & Shanmugam 1980) could argue for a turbiditic origin for the chalk horizons (e.g. Hesse 1975). However, several lines of evidence refute this, namely: (1) where sedimentary structures are present they do not comprise an entire chalk bed; (2) sedimentary structures that occur do not form any recognizable vertical sequence, as described from other fine-grained deep water deposits, e.g. Stow & Shanmugam (1980) (3) bases to beds are often gradational from the underlying mudstone or sandstone, and (4) there is no continuous grainsize grading through an individual chalk bed as would be expected from a single pulse of sedimentation originating from a turbidity current. The sedimentary structures may be divided into two types, those with a component of terrigenous material are the result of the mixing of 'hemipelagic' chalk with terrigenous sediment. Of these, flaser bedding is the most common (Figs 3 and 7). The presence of small flasers of silt and mud within a chalk bed indicates an intermittent current or current of fluctuating velocity. This structure is interpreted to represent the partial reworking of a pelagic chalk bed, with the incorporation of terrigenous material, by bottom current activity. Similar structures have been described in Recent sediments of the western Pacific (DSDP Leg 30). Here, flaser bedding was attributed to reworking of the sediments by
460
A.B. Hayward
(a)
(b) FIG. 6. Thin sections of hemipelagic chalk showing typical sedimentary structures. Dark flecks are scattered carbonaceous material. All the structures are delineated by the alignment and winnowing of biogenetic silt. Scale bars are 2 mm. (a) (1) Small scour (cut-and-fill feature) filled with coarser fill of winnowed biogenic silt. Dark line is formed by aligned carbonaceous fragments. Note distinct grading and two apparent pulses of sedimentation. (2) Pinch and swell silt laminations. (3) Micrograded wispy silt laminations. (b) (1) Couplets of discontinuous biogenic silt laminae, separated by structureless chalk. (2) Faint pinch and swell silt lamination passes laterally into structureless chalk. (3) Discontinuous silt laminations.
46I
Hemipelagic chalks in a submarine fan sequence E E
I::
fi
0
(/) 1,1(~o (91,1 <{Z rrv I,I o >--
1:
E
E E
o
0
I
I
o
o
_1 (./~ c 0
Q ~
~D
w ~-z xO
ILl(..)
_3
I
'1' II
ti '
i ii
:....
ii1 I'1
~i lil .: ~
I.I
I1 Jiil/ i m-I. .I.m
iii 0
I 0
,..o
0
i i
0
. ,,,,~
I'1 0
=
LLI
C9 ::D I"-"
>..
~2
Z W
Q)
~"a
m
a ILl
E
"
6~
462
A.B. Hayward
SEDIMENTARY STRUCTURE
SKETCH ~ 1 0
EXTERNAL CONTACTS
mm
AVERAGE THICKNESS
CONTINUITY
U. gradational Cross- l a m i nated graded silt laminae
3 - 5 mm
50 mm
2 - 4 mm
80 mm
I--3
50 mm
L. sharp
U. gradational Graded horizontal lamination
..,... :..,..,..~;.;~.... L. sharp
,.. ,.. ~ . . , ' . ' ., ~ . . 'o. , . ,.
U. gradational Pinch and swell silt lamination
,.:.......,~..:.:.1.'.
:.......
:
.' ".:
,"
~.'..'5.'..-
mm
L. gradational
U. gradational Micro-graded wispy silt lamination
%-..,;
9~
~
50 mm
L. sharp
U. gradational
Small c u t - a n d - f i l l structure
3-8 mm
5 0 mm
215
50 m m
L. sharp
U, gradational
Winnowed foraminifera debris lag 9. o
. o
. 9
.....,
",
-
... Q
Q
mm
L. sharp
FIG. 8. Summary table of micro-sedimentary structures visible in thin sections and delineated by alignment of winnowed biogenic silt.
Hemipelagic chalks in a submarine fan sequence bottom currents characterized by periodically varying velocity (Klein 1975). An alternative mechanism may be the introduction of several pulses of very finely dispersed terrigenous material from a very low density turbidite during normal 'hemipelagic' sedimentation. Small mudstone intraclasts towards the base of some chalk beds may be the result of reworking by bottom currents, of the partially lithified mudstone substrate. The range of sedimentary structures within the chalk beds with only a very minor component of terrigenous material, suggest reworking of slowly deposited hemipelagic ooze by some form of bottom current with fluctuating velocity. Particularly significant are the winnowed foraminifera lags, cut and fill features and biogenic silt laminations. Similar well-sorted concentrations of coarser microskeletons have been described from Recent pelagic sediments of the western Mediterranean (Rupke 1975) and from the Nova Scotian margin (Stow 1979). Table 1 compares diagnostic sedimentological features of muddy contourites, with the hemipelagic chalks of the Salir Formation. The following features which are common to both and may provide useful criteria for the distinction between a hemipelagic bottom current deposit and turbidite deposit in an ancient sedimentary sequence are: (a) generally homogeneous appearance in the field; (b) biogenic coarse lag concentrations; (c) laminae of winnowed biogenic tests; (d) a frequently high sand content of biogenic tests; (e)
463
composition consisting of a combination of biogenic and terrigenous material. The Salir Formation chalks represent deposition above the CCD, however these criteria apply equally well to diatomaceous or radiolarian sediments deposited below the CCD.
Lateral and vertical distribution of chalks Within the Salir Formation the chalks show considerable variation in bed thickness and in the proportion of the finer grained facies they represent (Fig. 3). Areas away from the main locus of deposition (e.g. interchannel areas, in proximal sequences, inactive mid-fan areas) are characterized by frequent and often thick chalk beds. In active mid-fan and inner-fan channel areas hemipelagic chalks are scarce or absent. From detailed measured sections it is possible to calculate upward transition probabilities between the facies conglomerate (cgl), sandstone (sst), mudstone (mdst) and chalk (ch). The transition between cgl/sst/mdst---,chalk expresses the relative sedimentation rate and may be an indicator of submarine fan sub-environment. A transition probability approaching 1.0 suggests long periods of time between successive turbidite events allowing hemipelagic chalk intervals to accumulate. A low transition probability is the result of the arrival of a turbidity current before a measurable hemipelagic chalk interval has
TABLE 1. Comparison of the sedimentary characteristics of muddy contourites (after Stow &
Lovell 1979) and pelagic chalks from the Salir Formation Muddy contourites
Pelagic chalks in the Salir Formation
- - bedding texture
not well defined homogenous
bioturbation, mottling - - burrows and pyrite mottles - - biogenic coarse lag concentrations - - primary silt/mud lamination
common common
mod. to well defined homogenous appearance in the field v. rare v. rare
common
common
rare
common
Texture
silty mud
silty chalk
Yes
Yes
combination of biogenic and terrigenous material
dominantly biogenic material with minor terrigenous component
Sedimentary characteristics
Structure -
-
-
-
-
-
frequently high sand content of biogenic tests
Composition
464
A.B. Hayward (o)
Facies Transition
for proximal
channel area
Transition probabilities for :
mdst/sst/cgl
=chalk 0"25
mdst/sst/cgl
:-mdst/sst/cgl 0"75
Preferred Transitions / cgl
....
.-" Tde
__=Tabce~.c
\
h ~ " 7
\\~'~md x st t
/
(based on 189 transitions) (b)
Facies Transitions for proximal interchannel fine grained sequence
Transition probabilities formdst/sst-
-~chalk 0"54
mdst/sst
a,mdst/sst 0"46
Preferred facies transitions glcgl/vcl ~ tde/s Ist
s
s
t
Tab ce
mdst .,~
ch ,
( b a s e d on 145 t r a n s i t i o n s ) FIG. 9. Facies transitions for the proximal sedimentary sequence, illustrating the difference in relative sedimentation rates between channel and interchannel sequences. Channel sequences show an overall fining-upward transition, (a) with chalks and mudstones deposited as the channel becomes abandoned. The fine-grained interchannel facies (b) shows an interbedded siltstone mudstone--chalk triplet with rarely the introduction of coarse sandstone (gl cgl/vc sst, Tabce) probably as a result of channel overflow. cgl--conglomerate gl cgl/vc sst--thin beds of granule conglomerate or very coarse sandstone Tabce---medium to thick bedded turbidite sandstone Tde/slst--thin bedded turbidite sandstone siltstone mdst--mudstone turbidite ch--hemipelagic chalk
465
Hemipelagic chalks in a submarine fan sequence formed, or in some cases due to the erosion of the chalk by the following turbidity current. In the Salir Formation, proximal channel fill sequences have a transition probability (Tp) of 0.25 (Fig. 9) compared with 0.50-0.80 for interchannel overbank areas. In the mid-fan area a similar relationship is seen, low net depositional
(a)
areas (e.g. abandoned lobes) have a Tp of between 0.6 and 0.75 compared with depositional lobes with Tp's of between 0.10 and 0.35 (Fig. 10). Little data exists on the distribution of pelagic sediments from modern clastic dominated sedimentary systems and tends to rely on cores of limited thickness. On the Laurentian fan and
Facies Transitions for Mid fan 'depositional lobe' (sequences with high sand:mudstone ratio)
Transition probabilities for: mdst/sst -----~chalk 0.10 rndst/sst
~-mdst/sst 0"90
Preferred facies transitions
Tbce _
"_ Tde/slst
ch
dst
(based on 116 transitions)
(b)
Facies Transitions for Mid fan area with low sedimentation rates (low sandstone: mudstone ratio)
Transition probabilities for: mdst/sst
=chalk 0-75
sst/mdst
= mdst/sst 0"25
Preferred facies transitions
Tbce
Tde
= mdst "
ch
(based on 272 transitions)
FIG. 10. Facies transitions for the mid-fan sequence illustrating difference between depositional 'lobe' and non-depositional (areas of low sedimentation rates) areas.
466
A.B. Hayward
Scotian Rise, Stow (1979) records variation over several hundred kilometres in texture and sorting in beds interpreted as bottom current deposits but no variation in thickness and or percentage of the succession is given. Variations in sedimentation rate and sediment texture have been recorded from the Nitanat submarine fan. Channels and immediately associated areas are characterized by a high suspended sediment load; by contrast, areas away from the main channel have a much lower sedimentation rate and, given the right conditions, these latter areas would be likely sites for the accumulation of thick pelagic sediments. In theory hemipelagic chalk beds represent periods of non-turbidite deposition or deposition following turbidity current activity. They will therefore reach their thickest development and greatest frequency in areas of low net and periodic deposition. On a submarine fan these areas are likely to be located in the inner-fan area. In the inner-fan area a turbidity current dumps its load over a limited area located within or close to the distributary channels open at the time. On other parts of the mid-fan and particularly in overbank areas away from the main inner-fan channels, long time intervals may occur during depositional events, allowing the accumulation of thick hemipelagic horizons. However, as the current travels distally it spreads out by dilution so that in the more distal mid-fan environments it covers a much larger area, restricting the thickness and frequency with which hemipelagic chalks can accumulate. Data from the Salir Formation (outlined above and Figs 9 and 10) fit well with this theoretical model.
Summary (1) The composition of the chalks indicates initial deposition from suspension settling. (2) Sedimentary structures within the chalks suggest reworking of the slowly deposited hemipelagic ooze by bottom currents of fluctuating
velocity. The main result was a winnowing and sorting of the coarser grained biogenic component of the chalks. (3) The similarity of the sedimentary structures preserved in the hemipelagic chalks with those described from fine-grained muddy turbidites (e.g. Stow & Shanmugam 1980) emphasizes the need for caution in attempting to distinguish turbidites from hemipelagic or bottom current deposits by sedimentary structures alone. As stated by Stow & Lovell (1979), a range of different but closely related mechanisms exist for the deposition of fine-grained sediment and it seems likely that a continuum may exist between dilute turbidity flows, bottom currents and hemipelagic settling. (4) Hemipelagic/bottom current deposits are characterized by: (a) a generally homogeneous field appearance; (b) a mixed composition of biogenic tests and terrigenous material, and (c) a high sand content of biogenic tests which frequently form laminae and coarse lag concentrations. (5) The lateral and vertical distribution of the hemipelagic chalks in many ways reflects the area on the submarine fan on which they were deposited. Areas away from the main locus of deposition (e.g. interchannel areas, in proximal sequences, inactive mid-fan areas) are characterized by frequent and often thick chalk beds. In active mid-fan and inner-fan channel areas hemipelagic chalks are scarce or absent. ACKNOWLEDGEMENTS: This work was carried out during the tenure of a Natural Environment Research Council grant, at the Grant Institute of Geology, University of Edinburgh. Logistal support in the field was provided by the Turkish Geological Survey (MTA). I am grateful to Maureen Fulton for drafting the diagrams. Drs. A.H.F. Robertson, D.A.V. Stow, D.J.W. Piper, C. Isaacs, K. Pickering, R.J. Bailey and J.P.B. Lovell made constructive comments on an earlier version of this manuscript.
References BOUMA, A.H. & HOLLISTER,C.D. 1973. Deep ocean basin sedimentation. In: Middleton, G.V. & Bouma, A.H. (eds), Turbidites and Deep Water Sedimentation. Soc. econ. Paleo. Min., Tulsa. 79-118. BRUNN, J.H., GRACIANSKY, P.C. DE, GUTNIC, M., JUTEAU,T., LEFEVRE,R., MARCOUX,J., MONOD,O. & POISSON, A. 1970. Structure majeurs et corr61ations stratigraphiques dans les Taurides occidentales. Bull. Soc. g~ol. Fr., 12(7), 515-51.
- - - , DUMONT, J.F., GRACIANSKY,P.C. DE, GUTNIC, M., JUTEAU, T., MARCOUX, J., MONOD, O. & POISSON, A. 1971. Outline of the geology of the western Taurides. In: Campbell, A.S. (ed.), Geology and History of Turkey. Petroleum Exploration Society of Libya, Tripoli. 225-57. DUMONT,J.F., GUTNIC,M., MARCOUX,M., MONOD,O. & POISSON, A. 1972. Essai de r6constitution d'un bassin Triassique ~iophiolites h la marge externe des
Hemipelagic chalks in a submarine fan sequence Taurides: le bassin pamphylien. C.R. Somm. S~anc. Soc. g~ol. Fr., 2, 73-4. GRACIANSKV, P.C., DE 1972. Recherches G~ologiques dans le Taurus Lycian Occidental. Thesis, University of South Paris, France. 571 pp. HAYWARD, A.B. 1982. Tertiary Ophiolite-related Sedimentation in S W Turkey. Thesis, University of Edinburgh, Scotland, 420 pp. - - - & ROBERTSON,A.H.F. 1982. Direction ofophiolite emplacement inferred from Cretaceous and Tertiary sediments of an adjacent autochthon, the Bey Daglari, southwest Turkey. Bull. geol. Soc. Am., 93, 68-75. HESSE,R. 1975. Turbiditic and non-turbiditic mudstone of Cretaceous flysch sections of the East Alps and other basins. Sedimentology, 22, 387-416. KLEIN, G. DE V. 1975. Resedimented pelagic carbonate and volcano-clastic sediments and sedimentary structures in Leg 30 DSDP cores from the western equatorial Pacific. Geology, 3, 39-42. PIPER, D.J.W. 1978. Turbidite muds and silts on deep sea fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. & BRISCO, C.D. 1975. Deep water continental margin sedimentation. Initial reports of the DeepSea Drilling Project Leg 28, Antarctica. In: Hayes, D.E., Frakes, L.A. et al. Init. Repts DSDP, 28, 727-55. POISSON, A. 1977. Recherches G~ologiques darts les Taurides Occidentales ( Turquie ). Thesis. University of South Paris, France. 795 pp. ROBERTSON, A.H.F. & WOODCOCK,N.H. 1980. Strikeslip related sedimentation in the Antalya Complex, SW Turkey. In: Ballance, P.F. & Reading, H.G.
-
-
467
(eds), Sedimentation in Oblique-Slip Mobile Zones. Blackwell, Oxford. 127-45. & WOODCOCK,N.H. 1981. Bilelyeri Group, Antalya Complex, SW Turkey: deposition on a Mesozoic passive continental margin. Sedimentology, 28, 381-99. RUPKE, N.A. 1975. Deposition of fine grained sediments in the abyssal environment of the AlgeroBalearic Basin, Western Mediterranean Sea. Sedimentology, 22, 95-109. RUDKE, N.A. & STANLEY,D.J. 1974. Distinctive properties of turbiditic and hemipelagic mud layers in the Algero-Balearic Basin, western Mediterranean Sea. Smiths. Contrib. Earth. Sci., 13, 40 pp. STow, D.A.V. 1979. Distinguishing between fine grained turbidites and contourites on the Nova Scotia deep water margin. Sedimentology, 26, 371-87. & BOWEN, A.J. 1980. A physical model for the transport and sorting of fine grained sediment by turbidity currents. Sedimentology 27, 31-46. & LOVELL,J.P.B. 1979. Contourites: their recognition in Modern and Ancient sediments. Earth Sci. Rev., 14, 251-91. SHANMUGAM,S. 1980. Sequence of structures in fine grained turbidites: comparison of recent deep sea and ancient flysch sediments. Sed. Geol., 25, 23-42. WEmER, Y. 1970. Mode of occurrence of pelites in the Kythrea flysch basin (Cyprus). J. sed. Petrol., 40, 1255-61. WOODCOCK,N.H. & ROBERTSON,A.H.F. 1982. Wrench and thrust tectonics along a Mesozoic-Cenozoic continental margin: Antalya Complex, SW Turkey. J. geol. Soc. Lond., 139, 147-63.
-
-
-
-
-
-
&
A.B. HAYWARD,Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, Scotland, U.K. Present address: BP, Britannic House, Moor Lane, London, UK.
Source rock potential of Bahamian Trough carbonates P.D. Crevello, J.W. Patton, T.W. Oesleby, W. Schlager and A. Droxler SUMMARY: The organic content of fine-grained carbonate muds deposited in the intraplatform basins and canyons of the Bahamas range between 0.1 and 2.6% total organic carbon (TOC). Geochemical analyses suggest some of these muds are a potential source for hydrocarbon. Interbedded with the carbonate muds are porous carbonate sands and gravels which can serve as local reservoirs or serve as conduits for hydrocarbons to migrate into carbonate slope and platform reservoir facies. In general, periplatform oozes are lean (generally less than 0.5 to 1% TOC) whereas green muds, very likely the products of turbidity currents, have higher organic contents (0.5 to more than 2.5% TOC), but values of TOC vary considerably both within and between study areas. Tongue of the Ocean and Exuma Sound are less than 80 km apart but their average TOC values differ markedly, 1% and 0.33% respectively. Higher sedimentation rates in the Tongue of the Ocean are thought, in part, to be responsible for better preservation of organic matter. In Tongue of the Ocean, cyclic variations in TOC are correlated with fluctuations of sea-level. High stands of sea-level are recorded by aragonite-rich muds with TOC on the average 0.5% higher than the calcitic periplatform oozes deposited during low stands of sea-level. Organic fractions of green muds have ~13C values ( - 16.4%o to - 13.9%o vs PDB) which are similar to platform carbonates, whereas lighter 613C values of periplatform oozes reflect mixing of organic material derived from platform and pelagic sources. Rock-Eval analysis shows that the periplatform oozes are relatively rich in oxygen; this, along with their low organic contents, results in a poor source rock potential. The green muds range between type II (mixed gas and oil potential) and III (gas potential) source rocks, which, through favourable maturation, should generate hydrocarbons. Gas chromatographic analysis of organic matter evolved at relatively low temperatures (300~ demonstrated that these immature sediments underwent vapourization and thermal cracking of relatively heat-sensitive or thermally labile material. Organic geochemical studies of carbonate rocks have recently demonstrated the important role that carbonates play in generating hydrocarbons (Jones 1980; Gardner 1980; Oehler 1980; Palacas 1980). Most of these studies, however, focused on the geochemical finger-printing of the kerogen (i.e. insoluble organic constituents) and the generated hydrocarbons. Our interest lies in understanding the factors contributing to the development of carbonate source rocks, primarily by studying their sedimentology and origin of organic constituents by palynologic and geochemical methods. Through this multi-disciplinary approach, our aim is to establish which conditions favour preservation of organic matter (the precursor to kerogen and source of hydrocarbons) and control its distribution within the sediments. The most ideal setting to study source type and controls on preservation is in a modern environment where sedimentary processes and stratigraphy have been thoroughly documented. The essentially pure carbonate province of the Great Bahama Banks, with its platform and trough configuration, lends itself to this type of study not only because of the numerous reports available on Bahamian Holocene and Pleistocene sedimentology, but also because the Bahamas replicate, in
many respects, certain ancient carbonate settings. Sediments deposited on broad platforms and in adjacent deep-water basins repeatedly occur throughout the geologic record and are similar to lithofacies reported from Bahamian sediments. Hence, the Bahamas have often been used as a standard for the interpretation of comparable ancient carbonate deposits. Similarly, our sourcerock model has potential application to ancient carbonate settings where basins adjacent to platforms generated hydrocarbons that presumably migrated into slope and platform reservoirs. Expanding our concepts on source-rock models is critical to understanding the minimum requirements for the development of carbonate source rocks. The sediments we examined range in age from Holocene to late Pliocene, lie within the upper 12.5 m of sea floor, and in terms of hydrocarbon maturity are considered very immature. Compared to many ancient settings with proven carbonate source rocks, sediments deposited in the deep troughs of the Bahamas are well oxygenated with dissolved oxygen levels of 3.3 to 5.8 ml/1 (Koczy et al. 1958). Surface trails on the sea floor and bioturbated sediments in core provide indirect evidence for aerobic conditions. Data obtained from this study, though, establishes that
469
47o
P.D. Crevello et al. PRO VIDENCE ~CHANNEL
2000 metre isobath
100
o G-17 KM
'DSDoP_ga '\ G-:
N
G-l(
GREAT
/,
t
"
G-28 BAHAMA G-13 G-89
G-53
BANK COLUMBUS G-15o
30 metre isobath
~ .
BASIN
FIG. 1. Index map of Great Bahama Bank and location of University of Miami sediment cores and DSDP site 98. P- 16 = P6907-016; P- 14 = P7102-014; G-I 3 = GS7507-013; G-53 = GS7507-053; G-89 = GS7507-089; G- 10 = GS7705-010; G- 14 = GS7705-014; G-I 7 = GS7705-017; G-28 = GS7705-028; G-34 = GS7705-034; G- 15 = GS7805-015. Dashed line defines toe-of-slope. carbonate sediments of good source-rock potential are preserved in the u p p e r 10 metres o f b a s i n a l s e d i m e n t s in certain B a h a m i a n troughs.
Methods S e d i m e n t s for o u r s o u r c e - r o c k study were selected f r o m a variety o f d e p o s i t i o n a l settings
a n d cover a range o f water d e p t h s (Figs 1 a n d 4; Table 1). Cores s a m p l e d i n c l u d e d five p i s t o n a n d gravity cores f r o m the cul de sac o f the T o n g u e o f the Ocean: o n e f r o m the lower slope (G-53), o n e f r o m the s a n d a p r o n o f the b a s i n - m a r g i n rise (G-13), a n d three f r o m the basin interior (G-89, G-28 a n d P-14). Cores f r o m this region c o n t a i n p e r i p l a t f o r m oozes a n d turbidites. T o evaluate the organic c o n t r i b u t i o n o f p u r e p e r i p l a t f o r m
TABLE 1.
Core No.
Latitude
Longitude
P6807-016 P7102-014 GS7507-013 GS7507-053 GS7507-089 GS7705-010 GS7705-014 GS7705-017 GS7705-028 GS7705-034 GS7805-015
24~ 23~ 23~ 23~ 23~ 24~ 24~ 25~ 23~ 25~ 22~
75~ 76~ 77~ 76~ 77~ 76~ 76~ 76~ 77~ 77~ 74~
I'W
Depth of Water Length of Core (m) (cm) 1847 1331 1385 886 1395 1483 1741 4016 1366 1935 2620
676 1257 121 277 145 123 595 850 862 1102 715
Source rock potential of Bahamian Trough carbonates ooze, away from the influence of turbidity currents, core G-34 was chosen because it was located on an isolated spur between two canyons on the lower slope of the axial valley of the Tongue of the Ocean. In Exuma Sound, two cores were from the basin interior (G- 14 and P- 16), and a third core (G-10) was from the lower slope. Representative samples of periplatform oozes were taken from a core from the basin interior of Columbus Basin (G-15). A final core (G-17) was retrieved from abyssal depths on Eleuthera Ridge. Only this core contains interbeds of hemipelagic clays. The sediments were collected during several University of Miami research cruises between 1969 to 1978. The cores are packed in air-tight core tubes and stored at 40~ and 95% relative humidity in the University of Miami cold-storage facility. We sampled the cores for our source-rock study in 1979 and 1980, carefully collecting only preserved, fresh-appearing sediments. The samples were sterilized with Zephiran chloride and shipped to Marathon Oil Company, Denver Research Laboratory for processing. Total organic carbon (TOC) percentages were determined by a Coulometrics TOC system 130 (Huffman Laboratories, Wheatridge, Colorado) and a Leco Corp WR-12 carbon determinator (Brown and Ruth Laboratories, Houston, Texas). In addition, Brown and Ruth performed the Rock-Eval analyses using the Geochem Rock-Eval II under normal operating conditions. Thermal vapourization experiments and 613C of the organic fraction were made at Marathon Oil Company. Thermovapourization gas chromatograms were made by heating dried, pulverized rock samples in a furnace held at 30OC in a stream of helium carrier gas for ten minutes. This material was passed through a dry-ice cooled trap, which was subsequently heated in the gas chromatograph oven to introduce it directly onto the gas chromatograph (GC) column. The GC column used was a Perkin-Elmer Dexsil-300 SCOT column with temperature programming from 30 to 325 ~ at 12~/min and a carrier gas flow of about 25 ml/min. 13C/12C ratios of the organic fraction were determined on a Nuclide RMS 6"-60: mass spectrometer. The organic material was converted to CO2 by combustion under oxygen with subsequent purification in a high-vacuum system. 13C/12C ratios are expressed in the usual 6-notation -1 F('3c/lzc) sample-] 6'3C, p e r m i l = / ~ ) f f D B J x103 relative to the PDB marine carbonate standard.
471
Depositional setting and basinal sedimentation Sediments deposited in Bahamian troughs consist mainly of lithoclastic-skeletal sands and gravels alternating with beds of lime muds (Rusnak & Nesteroff 1964; Bornhold & Pilkey 1971; Schlager & Chermak 1979; Crevello & Schlager 1980). The muds show signs of reworking by burrowing infauna (Fig. 2), while the carbonate sands and gravels are stratified, cross-laminated, and graded (Fig. 3). Because the basinal muds contain a significant volume of 'fines' thought to be shed from the platforms, Schlager & James (1978) termed these deposits periplatform ooze to distinguish them from true normal pelagic carbonate ooze. The noticeable influence that platform sedimentation has on basinal sedimentation was recognized very early from studies of the Tongue of the Ocean and Exuma Sound. Cyclic variations in the carbonate mineralogy of the periplatform oozes were interpreted by Supko (1963), Kier & Pilkey (1971) and Rucker (1968) to be the results of sea-level fluctuations. These authors suggested that the content of aragonite versus calcite in the ooze was controlled by the rate of production of aragonitic fines on the surrounding platforms. High input would correspond to bank flooding during high stands of Pleistocene sea-level. Lynts et al. (1973) and Droxler & Schlager (1982) emphasized the role of carbonate dissolution in controlling the aragonite cycles. Alternating with the perennial rain of fine suspended sediment are episodic pulses of sediment gravity flows that transport neritic carbonate and upper slope sands and gravels into the basins. In Exuma Sound, Crevello & Schlager (1980) estimated that coarse-grained deposits make up nearly 25% of the volume of basin-fill. Detailed high resolution seismic surveys and/or studies of closely spaced piston cores of the fiat-floored troughs of Columbus Basin (Bornhold & Pilkey 1971), Tongue of the Ocean (Schlager & Chermak 1979), and Exuma Sound (Crevello & Schlager 1980) led these authors to conclude that gravity-flow deposits of carbonate sediment have lobe-shaped geometries similar to comparable terrigenous deposits. More significantly, the latter two studies showed that carbonate platforms tend to provide a continuous source of sediment along a margin of reefs and sand shoals, resulting in a line source. Sediment gravity flows transport material down the slopes through multiple gullies and enter the basin in a manner reflecting the line source rather than a point source as described for classical deltacanyon-fan systems. In Tongue of the Ocean, the basin-rimming line
FIG. 2. Photographs of burrowed carbonate ooze from Bahamian Troughs. (a) and (b) Photographs of split cores from basin interior of Tongue of the Ocean. Bases of both cores are light grey to white carbonate ooze with burrows mostly less than 0.5 cm. Darker shades in photo are light olive green carbonate ooze (referred to as green mud in text) that contain or fill larger burrows which are up to 1 cm diameter. The olive green mud intervals in both cores are capped by light grey to white burrow mottled carbonate ooze. (c) Thin-section photomicrograph of carbonate ooze from Exuma Sound. The skeletal grains visible in the thin-section consist mainly of pteropods and planktic foraminifers. Note the burrow in centre of photo and the pelleted texture to the right of the burrow.
Source rock potential of Bahamian Trough carbonates
473
FIG. 3. Carbonate sands and gravels from Bahamian Troughs. (a) Carbonate breccia and graded, stratified gravel and sand deposited from a composite debris flow and turbidity current (described by Crevello & Schlager, Fig. 10, 1980). The top of layer is in left slab near 5 to 10 cm and base of layer in lower right slab. Large grains below 120 cm are mainly chalk lithoclasts. Comminuted chalk lithoclasts and platy fragments of the calcareous green alga Halimeda comprise most of the grains between 60-116 cm. Stratification occurs above 80 cm and overall the sequence is graded above 116 cm. (b) X-radiograph of graded carbonate sand with gravel clasts which shows complete Bouma sequence (from Crevello & Schlager, Fig. 6, 1980). Length of core is 37 cm. (c) Graded and laminated fine-grained carbonate sand and silt turbidite from basin interior of Columbus Basin.
474
P.D. Crevello et al.
source results in concentric facies belts (Fig. 4; Schlager & Chermak 1979). Coalescing sand and gravel deposits of turbidity currents and debris flows form an apron 15 to 25 km wide along the basin margin with muddier sediments, the tails of turbidity currents, dominating the basin interior. The concentric facies pattern is not as evident in Exuma Sound (Fig. 4; Crevello & Schlager 1980) because the eastern bordering platform (Cat Island Platform) source of sediment supply is insufficient to provide a line source comparable to the northern and north-western platform margin. Consequently, facies patterns are not concentric as in Tongue of the Ocean. Carbonate sediments deposited along open oceans and seaways adjacent to Little Bahama Bank show more variations in facies patterns than those described from the troughs (Mullins & Neumann 1979). Temporal variations in sediment input during periods of flooded and emergent platforms are indicated by sedimentation rates computed for Tongue of the Ocean and Exuma Sound (Fig. 5).
The data also show considerable variations in sedimentation rates within and between the two troughs. During the postglacial interval (younger than 40 000 yrs) and the latest interglacial interval (75-125 000 yrs), sedimentation rates in Tongue of the Ocean ranged from 20 to as much as 300 mm per 103 yrs, whereas glacial intervals show considerably reduced sedimentation rates, 10 to 60 mm per 103 troughs. Comparisons between the two troughs show that Exuma Sound has considerably lower sedimentation rates during both glacials and interglacials: 4 to 50 mm per 103 yrs.
Results of source rock evaluation Values of total organic carbon (Fig. 6) from the various basins range from less than 0.1 to nearly 2.6%. Sediments deposited in Tongue of the Ocean have the highest percentages of organic carbon, averaging 1.21% (0.12 to 2.58% range) for green muds and 0.7% (0.23 to 1.42~o range) for periplatform ooze. Oozes are also noticeably
FIG. 4. Sediment facies in two Bahamian troughs based on piston core and 3.5 kHz seismic studies of Schlager & Chermak (1979) for Tongue of the Ocean (upper illustration) and Crevello & Schlager (1980) for Exuma Sound (lower illustration). Large dots locate sediment cores that were analysed for geochemistry by this study. Refer to Fig. 1 for core numbers.
Source rock potential of Bahamian Trough carbonates S e d i m e n t a t i o n rate in m m / 1 0 0 0 yrs, 100
200
300 i
Ir[I
..............
C TM d a t i n g
o_
biostratigraphic dating
(3.
c
E co 100 T O N G U E of the O C E A N ---
EXUMA SOUND
FIG. 5. Sedimentation rates for Bahamian troughs. Rates for Tongue of the Ocean are from Kier & Pilkey (1971) and Rusnak (unpubl. data). Exuma Sound data from Lidz (1973) and Crevello (unpubl. data). richer in TOC in Tongue of the Ocean than the other basins. A noticeable decrease in organic carbon with depth (Fig. 6) occurs in Tongue of the Ocean. In pure carbonate basinal muds, sediment colour has been used as a rough index of organic richness. In Tongue of the Ocean, Schlager & Chermak (1979) recognized basin-wide colour intervals, which they used for correlating turbidite layers. They describe 'green' mud intervals alternating on a metre scale with 'white' intervals, and grossly characterize the green intervals as rich in organic material and the white as lean. Fig. 7 compares these basic colour intervals in Core P-14 with the TOC and relative sea-level curve based on carbonate mineralogy, which has been shown empirically to correlate approximately with the Quaternary sea-level curve as indicated by oxygen isotopes (Droxler & Schlager 1982). Aragonite-rich intervals correspond to times of high sea-level when the adjacent shallow-water platforms were flooded. Green sediment colour correlates distinctly with high aragonite content in the top three cycles. On the average, organiccarbon values are higher within the aragonitic intervals than the intervening calcitic intervals. The lowest aragonitic cycle (1000 to 1150 cm depth) lacks a corresponding green interval and samples are lower in TOC. The low TOC in the
475
base of the core might be related either to a decrease in the number of organic-rich turbidite muds reaching the basin interior relative to other times, or to changes in sediment or organic productivity on the upper slope and shelf. Core G-34 (Fig. 8) shows the variation in organic carbon ofperiplatform ooze deposited on a topographic high away from the influence of turbidity currents. Based on the detailed study of Core P- 14, only six sediment samples were necessary to test the TOC variation with sea-level. Values of TOC for periplatform oozes deposited during high stands of sea-level (i.e. aragonite-rich intervals) average 0.64% (0.41 to 0.97%; N = 3 ) versus 0.13% (0.1 to 0.17%; N = 3) during low stands. Carbon isotopes (613C) of the organic matter also display positive correlation with both TOC and percent aragonite, and a distinct separation occurs between the values corresponding to low stands ( - 2 2 . 3 to -25.1%o PDB) and high stands of sea-level ( - 18 to - 20.2%0 PDB). The correlation of 613C with sediment type strongly suggests two distinct sources for the organic matter. Isotopes of organic carbon in white periplatform oozes (low sea-level) approach typical values of pelagic carbonates ( - 2 5 to -28%0 PDB) indicating very little contribution from adjacent emergent platforms. In contrast, the green muds deposited during high stands of sea-level are in part shelf-derived because 513C values trend towards values typical for shelf carbonates in the Bahama-Florida area ( - 17 to - 13%o PDB; Crevello, unpublished). Figure 9 demonstrates further the correlation of sediment type with TOC and 613C. Green muds ( - 1 3 . 7 to -16.3%o PDB) are, with one exception, distinctly heavier than isotopic values of the white periplatform ooze having a variety of aragonite contents. To evaluate the source-rock potential of sediments analysed for this study, we employed two techniques commonly used to evaluate thermally mature source rocks: thermal vapourization/gas chromatography (GC) and Rock-Eval. The thermal vapourization/GC is a rapid method used to quantify the quality of source rocks. In Fig. 10 the thermal vapourization data, shown by the pyrolysate index, is plotted versus percent organic carbon, with the pyrolysate index being the normalized ratio of pyrolysate (thermally vapourized material) to percent organic carbon. Typical pyrolysate indices for good-quality mature source rocks lie between 8 and 12%. Several of the values for Core P-14 are above this good-quality value, suggesting sediments from Tongue of the Ocean have source-rock potential. The Rock-Eval method (Fig. 11) is a routine procedure performed on mature rocks used to
P.D. Crevello et al.
476
(a)
Om~ 0
0
l
9
9
(b)
0
0
~,, o o
200
-CO0
200
E
400
TONGUE of the OCEAN
0 o
E
~
600
o
o
o
~s 400
o
Periplatform ooze
oo
o
0
EXUMA SOUND
0
0
o o
o
Ck o
s 600 800
o I
"~
1000
Green mud
o ~_ Periplatform
i0
I
1. 210 ORGANIC C A R B O N , %
0.0
ooze
o
(d) o
I
I
)
I
I
0 0
1200 o
200
'
0.0
110
ORGANIC
(c)
0
'
2:0
CARBON,
'
3.0
0
40O E
o.,- Carbonate mud
0
200
COLUMBUS BASIN
O
="
O
O. 0
&
a
Marl
6OO
Periplatform ooze
E
N E PROVIDENCE CHANNEL
%
#
0
0 0
o
0
0"" Clay
e4-, e~
O
400
0
O
8OO
0 &
O O I
0.0
O
6OO
I
I
110 210 ORGANIC C A R B O N , %
a.0
o
0.1
' 1 i0 ' 2.'0 ' ORGANIC CARBON, %
FIG. 6. Organic carbon values determined by this study. (a), (b) and (c) are plots of percent total organic carbon of fine-grained carbonate muds and oozes from Bahamian troughs. Clays, marls, carbonate muds are shown in (d). Samples are plotted in depths below sea bottom.
Source rock potential of Bahamian Trough carbonates CORE P-14
Brown Olive
477
Aragonite, %
0
0i
o
i
50 1
i
100 J
9
_
o
White w/patches of tan & orange
onite
o
Olive
o
500 cm
o
9
Wh ite
Aragonite, %
Pale olive w/patches of white
0
~
0
25
o -,- Periplotform "
h
.E
210
Organic carbon, %
T
~
,
9
I
"
1
"
c
o O
t.
400 o
o O
500
o 0.0
~
0.5
L, L
1.0
1.5
-24 -22 -20 -18
Organic carbon, %
613C, ppt (PDB)
FIG. 8. Percent organic carbon, percent aragonite, and 613C of the organic fraction of the periplatform ooze of Core G-34. f
T
1
T ~
T
|
9
m m
-15
~
g~
3'0
FIG. 7. Comparison of sediment colour, sediment type (green mud or periplatform ooze), percent organic content, and percent aragonite in periplatform ooze. The aragonite curve down to 500 cm corresponds with the curve in core G-34 (Fig. 5) which is correlated by oxygen isotopes to glacial-interglacial stages and stands of sea-level. Below 500 cm the aragonite curve in P-14 does not correspond with G-34 and may be distorted by high turbidite influx. ~
,
I
o "e"
E 200
o
Ooze
1.'0
i
/,Approx,ma*e
~. 300
0.0
100
'
~Aragonite
100 .o
9o o
75
o
O
1000 cm White w/ patches of pale olive
50
'
9~ Green mud
9
mm
Q~ 0
03 0..
v
T O N G U E of the O C E A N
-20
9 Green m u d Periplatform ooze
Q
c).. (3.
9
Lt. b r o w n surface layer (high a r a g o n i t e ) e White mud (high a r a g o n i t e ) 9 White mud (low a r a g o n i t e )
Q
d O3
Q
I.---
0
oo |
EXUMA SOUND
-25
Periplatform ooze o
Lt. b r o w n m u d
m
1
1
1.0
2.0
0
0.0
J . ~
3.0
Organic carbon, % FIG. 9. 613C of fine-grained carbonate muds and oozes from Tongue of the Ocean and Exuma Sound plotted against total organic carbon. Note the periplatform oozes from Tongue of the Ocean are distinguished based on relative amounts of aragonite (high-aragonite light brown surface muds and white muds, and low-aragonite white muds). The separation by mineralogy was not done for Exuma Sound oozes.
478
P.D. Crevello et al. 30
o
u
Core P-14 9Green mud o Periplatform ooze o> 50% Aragonite o< 50% Aragonite
cO ..Q
20
o ,o_
o
~.2
o
|
10
9
om
o
|
5
qmo
121. o
9
o
0
9 o
015 110 1 ,'5 2'.0 Organic carbon, %
).0
FIG. 10. Thermal vapourization of sediments in Core P-14. Percent organic carbon is plotted against a pyrolysate ratio which normalizes total ppm evolved to percent organic carbon. Periplatform oozes are distinguished by mineralogy.
qualify the kerogen and the type of hydrocarbon the sample is expected to generate. The data show that the green muds and periplatform oozes plot along distinctly separate and nearly linear trends. The green muds contain more hydrogen and are not as oxidized as the periplatform oozes. Both sediment types fall between Type II (mixed oil and gas prone) and Type III (gas prone) source rocks (Tissot & Welte 1978). Type I would be considered an oil-prone source rock.
1000 9 Green mud
O
o
E
P e r i p l a t f o r m ooze
800
6 600
~ 4oo ~
mmn
~ 2oo
o
"1"
o II 0
200
400
600
800
1000
O x y g e n index (mg CO~/g Organic C)
F]6. 11. Rock-Eval analysis of green muds and periplatform oozes. Kerogen types I, II and III are generally used to define fields for oil-prone, oil- or gas-prone, and gas-prone kerogens respectively.
Discussion Source-rock evaluation
The most significant finding of this study has been the organic-rich sediments deposited on the sea bottom in the lower slope and basin of Tongue of the Ocean. Yet only 80 km away, sediments deposited in Exuma Sound and Columbus Basin are noticeably organic poor (Fig. 6). The generally accepted lower limit for potential source rocks is about 0.3 to 0.5% TOC. Below this TOC content there would be insufficient quantities of organics to produce significant hydrocarbon reserves. Organic carbon analyses of 53 chalks of Holocene-late Pleistocene to early Campanianlate Santonian age, penetrated by DSDP Site 98 between 0.3 to 348 m sub-bottom and analysed by the shipboard party, range between <0.I and 0.4% TOC (Hollister et al. 1972, 1060-62). These chalks were deposited in basin-floor and slope settings in 2700 to 3100 m of water depth (Schlager & Ginsburg 1981). The low TOC values are comparable to those of periplatform oozes we analysed from similar settings and depths (i.e. Core G-34, G-10 and G-17). Thermal vapourization analyses show that several samples from Tongue of the Ocean have source-rock potential. This can be explained by the high degree of immaturity of these sediments that, as of yet, have not undergone any type of thermal degradation. Perhaps with slightly deeper burial and elevated temperatures, the evolution of thermally labile components such as CO2, H20, NH4 and CH4 may bring these values closer in line with maturing source rocks. Another possibility for the high values, also related to organic immaturity, is that organic compounds are actually undergoing thermal cracking by pyrolysis. In either case, conclusions can still be made from these analyses. There do not appear to be any obvious correlations of high TOC values or sediment type with the pyrolysate index. In actuality, the data show that low and intermediate values of TOC, independent of the sediment type or source, have relatively good pyrolysate indices. Sediments with high TOC may not necessarily have the right types of organic material to produce hydrocarbons. Specifically, some green muds have a pyrolysis index of less than 5% with TOC values near 2%, suggesting that much of the organic material transported in from shallow water depths is some form of reworked or oxidized organic debris. As noted above, the amount of TOC present is also critical in making a good source rock. Depending on the types of organic materials present, the lower limit generally placed on carbonate rocks is in the
Source rock potential of Bahamian Trough carbonates range of 0.3 to 0.5% organic carbon. In this regard then, several of the samples shown in Fig. 10 with high pyrolysate indices also have values of TOC less than 0.5, making their source-rock potential marginal. Rock-Eval data (Fig. 11) show that the sediments are geochemically immature, and that the periplatform oozes are more oxidized than the green muds. Normal source-rock values lie below 200 on the oxygen index and most of these immature sediments plot above this value. During maturation of the organic material, the oxygen and hydrogen indices for each sample will shift with loss of CO2, by evolution of H20, or by CH2 (Tissot & Wette 1978). Hence, subsequent maturation will ultimately control the type of hydrocarbon to be generated (i.e. gas prone versus oil prone).
Sedimentation rates, neritic input and organic content Sedimentation rates are thought, in part, to influence the preservation of organic matter. As shown in Fig. 5, sedimentation rates of 10 to 300 mm per 103 years in Tongue of the Ocean contrast to lower rates of 4 to 50 mm per 103 years in Exuma Sound. Because the sediment-water interface is oxygenated, faster sedimentation would more rapidly bury the sediments beneath this aerobic surface layer (Demaison & Moore 1980), favouring preservation of organic matter. Also, rapid sedimentation would tend to prevent degradation by burrowing organisms. Some of the green layers are almost certainly the muddy tails of turbidites because they commonly follow on top of graded turbidite sands or form the basinward extension of graded beds (Schlager, unpublished). Other green intervals may represent periplatform ooze deposited by perennial rain of fines mixed with low density turbidity currents. In some cases, bioturbation of the green layers makes it difficult to differentiate these two depositional processes, and therefore we have not genetically classified the green muds. The green muds observed in Tongue of the Ocean have isotopic (613C) signatures similar to typical platform carbonates, leading us to believe that the higher sedimentation rates in Tongue of the Ocean reflect higher sediment input from the adjacent platform. This high input may well be related to the size of the source area. Of all Bahamian basins, Tongue of the Ocean is surrounded by the largest platforms, and the platforms to the east of the basin are probably the most significant source because of the predominant easterly winds. Potential platform source areas around Exuma Sound and Columbus Basin
479
are considerably smaller, particularly those east of these basins. Consequently, rates of high-stand sedimentation in these basins are relatively low; and organic-rich, olive-coloured sediments are absent. The difference between Tongue of the Ocean and the other basins may be enhanced by the effects of water depth. Southern Tongue of the Ocean, with a depth of 1100 to 1400 m, lies in the lowermost part of oxygen minimum zone (3.3 ml/l 02 between 500 to 800 m and steadily increasing to 5.8 ml/l 02 near 2000 m) (Koczy et al. 1958), whereas the flat floors of Exuma Sound and Columbus Basin (1500 to 1800 m and > 1800 m respectively) are distinctly below the midwater oxygen minimum.
Conclusions Burial of organic matter in Bahamian troughs is controlled by at least two processes that have varied in space and time: (1) organic input from planktic and shallow-platform sources, and (2) sedimentation rate. Yet another, but unresolved, factor that may also be important is water depth, i.e. position relative to the midwater oxygen minimum. Favourable combinations of these factors have led to deposition of sediments with distinct source-rock potential in certain areas and time intervals. Variation of source-rock potential in time appears to be tied to the Quaternary climatic cycles and consequent sea-level fluctuations. Conditions are favourable during interglacials when high carbonate production on the flooded platforms and perhaps low rates of carbonate dissolution make for high sedimentation rates and high input of organic matter from the platforms. Variations in space are controlled by the morphology and setting of the various troughs. Conditions are most favourable in the cul de sac of Tongue of the Ocean where a large source area of carbonate platforms leads to high sedimentation rates and high input of organic matter. With water depths of 1100 to 1400 m, this area may also be influenced by the midwater-oxygen minimum, favouring preservation of organic matter. Exuma Sound and Columbus Basin are considerably deeper and have smaller platform source areas. Consequently, sedimentation rates are lower, sediments are more highly oxidized and source-rock potential is low. ACKNOWLEDGEMENTS: This study conveys the results of a portion of a research project conducted on the source-rock/organic geochemistry of carbonate rocks. The project is under investi-
480
P.D. Crevello et al.
gation at M a r a t h o n Oil C o m p a n y , Denver Research Laboratory. We t h a n k M a r a t h o n for the o p p o r t u n i t y to c o n d u c t this study and for releasing the information for publication. We also t h a n k the following persons who assisted in various aspects of the data compilation, analysis
and preparation o f the manuscript: S.M. Andrews, L. Brinton, G . K . Guennel, H . M . Heck, W.H. L o h m a n , H. Maxwell, N. Neafus, C.L. Pedde, and D . D . Wallwey. W e are grateful to P.W. Choquette, H.T. Mullins, and D.A.V. Stow for their reviews of the manuscript.
References BORNHOLD, B.D. & PILKEY, O.H. 1971. Bioclastic turbidite sedimentation in Columbus Basin, Bahamas: Bull. geol. Soc. Am., 82, 1341-54. CREVELLO, P.D. & SCHLAGER, W. 1980. Carbonate debris sheets and turbidites. Exuma Sound, Bahamas. J. sed. Petrol., 50, 1121-48. DEMAISON,G.J. & MOORE, G.T. 1980. Anoxic environments and oil source beds. Bull. Am. Ass. Petrol. Geol., 64, 1179-209. DROXLER,A. 8~ SCHLAGER,W. 1982. Reverse pattern of glacial-interglacial sedimentation rates in periplatform carbonate ooze, Bahamas, as compared to the classic pattern of hemipelagic ooze. International Workshop on Fine-Grained Sediments, Deep-Water Processes and Environments, Program Abstracts, Halifax. GARDNER, W.C. 1980. Oils and source rocks of Niagaran reefs in the Michigan Basin. Bull. geol. Soc. Am., 12 (abs.), 431. HOLLISTER, C.D., EWING, J.I. et al. 1972. Init. Repts. DSDP, 11, 1159-61. JONES, R.W. 1980. Organic facies and tectonic setting of carbonate source rocks. Bull. geol. Soc. Am., 12 (abs.), 457. KIER, J.S. & PILKEY, O.H. 1971. The influence of sea level changes on sediment carbonate mineralogy, Tongue of the Ocean. Marine Geol., 11, 189-200. Koczv, F.F., CHEW, F., FEINSTEIN, A., RHIAN, E., RICHARD, J.D., SIEGLER,V.B. t~ WENNEKENS,M.P., 1958. Oceanographic survey of the Tongue of the Ocean. Univ. Miami Marine Lab. Tech. Rpt., 161 pp. LIDZ, B. 1973. Biostratigraphy of Neogene cores from Exuma Sound Diapirs, Bahama Islands. Bull. Am. Ass. Petrol. Geol., 57, 841-57. LYNTS, G.W., JUDD, J.B. & STEHMAN,C.F. 1973. Late Pleistocene history of Tongue of the Ocean, Bahamas. Bull. geol. Soc. Am., 84, 2665-84.
MULLINS, H.T. & NEUMANN, A.C. 1979. Deep carbonate bank margin structure and sedimentation in the northern Bahamas, In: Doyle, L. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleo. Min. Spec. Pub. No. 27, 165-92. OEHLER, J.H. 1980. Carbonate source rocks in the Jurassic Smackover Trend of Mississippi, Alabama, and Florida. Bull. geol. Soc. Am., 12 (abs.), 494. PALACAS,J. 1980. South Florida Basin, a prime example of carbonate source rocks of petroleum. Bull. geol. Soc. Am., 12 (abs.), 495. RUCKER,J.B. 1968. Carbonate mineralogy of sediments of Exuma Sound, Bahamas. J. sed. Petrol., 38, 68-72. RUSNAK, C.A. & NESTEROFF, W.D. 1964. Modern turbidites: terrigenous abyssal plain versus bioclastic basin. In: Miller, R.L. (ed.), Papers in Marine Geology. Macmillan & Co., New York. 488-507. SCHLAGER,W. & CHERMAK,A. 1979. Modern sediment facies of platform-basin transition, Tongue of the Ocean, Bahamas. In: Doyle, L. & Pilkey, O.H. (eds), Geology of Continental Slopes. Soc. econ. Paleo. Min. Spec. Pub. No. 27, 193-208. SCHLAGER, W. & JAMES, N.P. 1978. Low-magnesian calcite limestones forming at the deep-sea floor, Tongue of the Ocean, Bahamas. Sedimentology, 15, 675-702. SCHLAGER, W. & GINSBURG, R.N. 1981. Bahama carbonate platforms--The Deep and the Past. Marine Geol., 44, 1-24. SUPKO, P.R. 1963. A Quantitative X-ray Diffraction Method for the Mineralogy of Carbonate Sediments from the Tongue of the Ocean. Unpubl. Masters Thesis, University of Miami, Florida, US. 144 pp. TISSOT, B.P. & WELTE, D.H. 1978. Petroleum Formation and Occurrence. Springer-Verlag, New York. 538 PP.
P.D. CREVELLO, J.W. PATTON t~ T.W. OESLEBY, Marathon Oil Company, PO Box 269, Littleton, CO 80160, USA. W. SCHLAGER• A. DROXLER,University of Miami, Comparative Sedimentology Lab., Fisher Island Station, Miami, FL. 33139, USA.
Hemipelagic deposits in a Miocene basin, California: toward a model of lithologic variation and sequence C.M. Isaacs SUMMARY: Deep marine basins began to form along the continental margin of central California in late Oligocene time as a result of the initiation of a translational plate boundary. Deposition of transgressive shallow-marine sand (Vaqueros Formation) was followed by deposition of a thick ( > 500 m) massive clay unit (Rincon Shale). Rapid subsidence ( > 2000 m) coincident with rising sea-level created sediment-starved basins by late early Miocene time
(c. 18 Ma). Typical of many of these borderland basins was the Santa Barbara basin, located at least 50 km from the Miocene strandline. In this basin, between late early and latest Miocene time (18-5.5 Ma), was deposited a heterogeneous sequence of highly biogenous sediment (Monterey Formation) consisting of mixed diatomaceous, foraminiferal-coccolithic, and terrigenous debris generally rich in organic matter (mean 8%). Overlying latest Miocene and Pliocene sediments (Sisquoc Formation) contain increasingly abundant clay debris. The Miocene biogenous sequence reflects hemipelagic sedimentation from highly productive coastal waters during an extended period of generally low terrigenous influx. Accumulation of terrigenous debris decreased markedly in the early Miocene (from 11 to 1 g/cm2o103 yr) and was low throughout the middle Miocene, increasing somewhat at about 8 Ma and markedly after 5.5 Ma; this pattern is partly related to global sea-level changes but also resulted from high rates of subsidence and tectonic events. Biogenous silica accumulation peaked (mean 2.5 g/cm2" 103 yr) in the late early Miocene (18-15 Ma) and again in the late Miocene and early Pliocene (8-3.5 Ma) and was 5 to 25 times slower in the interval 15-11 Ma; these variations are thought to reflect primarily variations in the intensity of upwelling and associated production of diatoms. Accumulation rates of biogenous calcite gradually declined from a late early Miocene peak through the early late Miocene and virtually ceased in latest Miocene (8-5.5 Ma), probably due to dissolution. Massive stratification indicates that bottom waters were oxygenated until latest early Miocene time (c. ! 6 Ma), when expansion of the oxygen-minimumzone or tectonic formation of a sill within the oxygen-minimum zone produced marked lowering of oxygen levels resulting in the dominance of laminated stratification between 16 and 5.5 Ma. Abundance of organic matter is closely associated with the abundance of clay and of calcite but is inversely correlated with silica content. Organic matter is also about twice as abundant in massive or discontinuously laminated beds (indicating good to moderate oxygenation) than in associated beds with varve-like lamination (indicating minimal oxygenation). These relations indicate that grain-size outweighed the influence of low-oxygen bottom waters in preserving organic matter.
The Monterey Formation of California is a Miocene diatomaceous deposit, best known for its 'varved' diatomites and petroleum production from fractured reservoirs. Not as widely appreciated is the Formation's lithologic heterogeneity, which is especially marked in the western part of the Transverse and Coast Ranges. Here a significant thickness (about 100 m) of the Formation is carbonaceous marl averaging a mere 15% silica, and in several parts of the sequence the composition of rocks within 1 metre varies over the ranges 30-80% silica, 10-50% calcite, and 10-70% detrital minerals. The major objectives of this paper are: (1) to examine the origin of the sediment components in these fine-grained Miocene rocks, and (2) to explore the significance of the distinctive sequence of lithofacies. Deducing patterns in
palaeoclimate, palaeoproductivity, and upwelling is not the objective per se, except insofar as these parameters strongly influenced rock composition.
Geologic setting The Miocene and Pliocene rocks of the western Santa Barbara coastal area (Fig. 1) generally occur in a south-dipping homocline along the southern flank of the Santa Ynez Range which is bounded on the north by the Santa Ynez fault (Dibblee 1950, 1966). In this homocline, the Miocene and Pliocene sequence overlies a thick package of Cretaceous to Oligocene rocks which are largely marine, although Oligocene non-marine rocks are increasingly abundant toward the
48I
482
C.M.
Isaacs
122~
8 o
36 ~
.~:..:~\~ San Andreas U'-'-']
' "'i. "~ ".-',,
L
9 :..!
'"'~ \
PA
k
ClFIC
OCEAN :: :,
FIG. 1. Palaeogeography of the Santa Barbara basin. Top: non-palinspastic reconstruction of the palaeogeography of central California west of the San Andreas fault, showing relief at the beginning of the transgression in latest Oligocene and earliest Miocene time (c.25 Ma). Modified from Loel & Corey (1932). Bottom: non-palinspastic reconstruction of the Santa Barbara area in Miocene time, showing the position of the basin axis in early Miocene time and the middle Miocene strandline. Adapted from Carson (1965), Edwards (1971), Fischer (1976), and Clifton (1981). The dotted line 340 indicates the present coastline, and the open circles indicate the western Santa Barbara coastal area.
~
~..
r.-....
".', 1;0 kmt' "l "~:',
I~
33 ~
120 ~ i
'.'. Arguello Positive
~'" trough .::--.:. -
eastern part of the area. A major transgression began in latest Oligocene time with littoral deposits of sandstone (the Vaqueros Formation) followed by latest Oligocene and early Miocene clayrocks (Rincon Shale) interpreted as slope deposits, early Miocene through late Miocene marl and diatomite (the Monterey Formation) interpreted as basin deposits, and latest Miocene and Pliocene clay-rich diatomaceous mudrocks (Sisquoc Formation) interpreted as slope deposits (Ingle 1980, 1981a, b). These strata are capped in adjacent onshore and offshore areas by a thick sequence of Pliocene shale and sandstone (Dibblee 1950, 1966; Vedder et al. 1969) 9 Tectonically, the Miocene Santa Barbara basin was apparently a borderland-type basin resulting from the formation of a translational plate boundary in late Palaeogene time (Atwater 1970; Blake et al. 1978). As shown by palaeodepths of about 1000 m in earliest Miocene time (Edwards 1971; Ingle 1980), subsidence in the present Santa
":
~"":.
''
Strandline~.....
? "~ ,...~ ~ .,.~ ~ .
.
10 km
~,~olcanic . "".... ~
.
....... ".......i::; ""....~"~,Z ..... ":~--
East Santa Cruz Basin
ridge. . . .
....
Fault
Ynez Range was rapid in latest Oligocene and early Miocene time and totaled 2500-3000 m by the end of the early Miocene (c. 16 Ma). Early and middle Miocene volcanic deposits are widespread on the southern border of the basin, and a volcanic centre was also active in early Miocene time 10-15 km W N W of the Arguello Positive (Fig. 1). Although diatom productivity, at one time thought to be caused by volcanism (Taliaferro 1933), is now known to result from nutrient enrichment of surface water by upwelling (Calvert 1966), these volcanic and diatomaceous deposits may be co-genetic inasmuch as volcanic centres and sediment-starved basins may have resulted from the same tectonic controls. Palaeobathymetric analysis suggests that the early Miocene Santa Barbara basin was a narrow elongate basin with its axis more or less paralleling the present east-west trending coastline (Edwards 1971; Fischer 1976; Ingle 1980; Fig. 1). Whether the laminated rocks in the Monterey
Hemipelagic deposits in a Miocene basin, California were deposited in a basin-slope environment similar to settings in the present Gulf of California or in a basin-floor environment similar to settings in the present southern California borderland remains debatable (Donegan & Schrader 1981; Ingle 1980, 1981b; Pisciotto & Garrison 1981; Soutar et al. 1981; Summerhayes 1981). As presently situated (without palinspastic reconstruction), the western Santa Barbara coastal area is 25 to 50 km south-west of the nearest early Miocene (Rincon) strandline (Edwards 1971) and 70 to 115 km south-west of the main middle Miocene strandline (Fischer 1976; Clifton 1981). Well-documented oscillations of the strandline and controversy over tectonic movements make accurate reconstruction impossible, but actual distances from the strandline could have been considerably greater. Restoration of 1-45 km left-lateral movement along the Santa Ynez fault (Fig. 1), for example, increases the distance from the strandline. Alternately, reconstruction based on palaeomagnetic evidence of 90~ ~ clockwise rotation (Luyendyk et al. 1980) would place the present coastal area in Miocene time on the seaward side of a northsouth trending basin, at least 50 km seaward of a submarine ridge and probably at least 100 km off the main strandline with a submarine high (the Arguello Positive of Fischer 1976) to the west. In either case, the Monterey in the present western Santa Barbara coastal area was deposited at a considerable distance from the strandline and was protected from the influx of most coarse terrigenous debris by submarine highs flanking the basin.
Stratigraphy General characteristics of Miocene rocks in the coastal area west of Santa Barbara are summarized in Table 1 and Fig. 2. The Rincon Shale is described in more detail by Carson (1965) and Edwards (1971, 1972), the Monterey Formation by Bramlette (1946) and Isaacs (1981), the Sisquoc Formation by Wornhardt (1967), and the entire sequence by Dibblee (1950, 1966). Ages, although partly derived from provincial benthic foraminiferal stages, are based mainly on the coccolith and diatom zonation in Barron et al. (1981).
Influences on sediment composition The fine-grained sediments deposited in the Miocene Santa Barbara basin were complex deposits,
483
having significant components of planktic and benthic biogenous debris as well as terrigenous material and displaying evidence of a variety of bottom conditions. The following discussion of the influences on sediment composition emphasizes both detailed variations in closely associated beds and general changes through time. The main relations addressed are: (1) high abundance of silica deposited in lower Miocene strata and again in upper Miocene through lower Pliocene strata and low abundance in middle Miocene strata; (2) high abundance of calcite and organic matter in middle Miocene strata and disappearance of calcite in the interval 8 to 5.5 Ma; (3) changes in layering character from massive (through the early Miocene) to laminated; and (4) association between type of lamination and composition in middle and upper Miocene strata. The discussion is divided into the following sections: sediment accumulation rates, terrigenous accumulation, biogenous accumulation, layering, and synthesis.
Sediment accumulation rates
Estimated member-average rates of sediment accumulation are summarized in Table 2 and Fig. 2. These rates are low compared to average values for near-shore marine deposition but far exceed average rates of pelagic deposition, which for siliceous ooze ranges from 2 to 10 m/my and for calcareous ooze ranges from 3 to 60 m/my with an average of about 30 m/my (Berger 1974). Within the Monterey, rates of silica accumulation peaked in the lowermost (18-15 Ma) and uppermost (8-5.5 Ma) members and reached a minimum in the carbonaceous marl member (15-11 Ma). Accumulation of calcite decreased gradually through the early and middle Miocene, reached a minimum in the late part of the late Miocene (8-5.5 Ma), and then increased. Accumulation of terrigenous detrital debris was slow from late early Miocene through early late Miocene (18-8 Ma) and then increased markedly. The lower boundary of the Monterey is marked by a sharp reduction in the rate of accumulation of terrigenous debris (Fig. 2, Table 2). Whether or not this reduction was concomitant with increased silica accumulation is less clear. Total accumulation rates in the Rincon Shale are sufficiently high that rates of silica accumulation comparable to those in the Monterey could not be discerned on the basis of presently available data; average rates of accumulation in the lowest member of the Monterey would represent only 23~ silica in mudrocks of the Rincon. The local presence of as much as 30 m of distinctly siliceous strata in the upper part of the Rincon Shale
484
C.M. Isaacs
~
O0
~,..~
~q
~2
o
A
~.~ "~"~
~
~'-~
~9
0 -'s4
~
o
"~'~
~.,-~
.~ ~..~ o
~
~ ~
9
~.'~
~.~ ~ ~ ~
~ ~
i; o
~
-~'~
~~o
~
~
~ ~
~
~
"
~ ~
o
~
"~~
o
0
o
r
o"1
t~l
r
.~'~
~ t"q
m r~
,..s4 "~ o ~ N
o
0 O~
o.~ ~ ~
o'~
~~o
~
.~,o
~.-~ ~ "~,.~
N
~
~
~
~'~ ~o
~ , ~
~
.~. o~
~[.. ~
~o
& . ~ :~ :~ ~ o
~
o r~ "N
~
~
~ ~o ~
Hemipelagic deposits in a Miocene basin, California
485
0
=
I.Ll l-c~
rr
Z
-'
I,
~O
O
.~
I--
m
EO
.~
.c_ E
6
:::)
,-
F
fO
":"
,
0
~
o~
E L) d
'
~
~ o
":. ,'2_
I
'a
I
.
0
/,J
.~ fJ)
~L --
~ 7
I . . . . . . . . . . . . . .
8
-~ .
.
.
~
. . . . .
'
I
I
if)
("
I
I
~-- . . . . . .
i-
. . . . . . . .
e-.
I. . . . . . .
oO g 0"
NOIIVI~klO-I
,k3k131NOl~
8
=
.0
E i [= ~
T=
i E
~
I
8 .o
8
oo mm ~ o
~E o
~~.o ~ .o o
88
~
=
~
> (o ~>
"5 ~ ,.. _
.-
4,--'
~o 0 ,--'.W
O .-I p'-.
"' OOL I
I
I
.
I
i
T_
.NVIINOIN13(],
3NBOOlld
"3
!
i
.
I 3N3OOIB
.
.
!
T
.
NVINHOIAI 31V7
!
3N300IIAI "I/~
3N300IIAI ) ~ 7 U V 3
_=r 8~ 12.o O~
486
C.M. Isaacs
TABLE 2. Estimates of accumulation rates for the Miocene sequence in the Santa Barbara
coastal area. Values in g/cm 2.103 yr represent a range derived from the mean composition of each stratigraphic unit (Table 1), maximum and minimum thickness (adjusted for bulk density) of each unit, and the average time interval of each unit (Fig. 2;for data, see Isaacs 1983). Total accumulation rates are also given in m/my for general comparison with other areas; these values are based on an original porosity of 90~ (Hamilton 1976; Soutar et al. 1981) and would be halved based on an original porosity of 80% and doubled based on an original porosity of 95~ Total Formation Sisquoc Formation Monterey Formation Clayey-siliceous member Upper calcareoussiliceous and transitional marl-siliceous members Carbonaceous marl member Lower calcareoussiliceous member Rincon Shale
Total
Detritus
m/m.y,
Silica
Carbonate
Organic Matter
g/cm2" 103 yr
>750->3700
> 15->72
>9.1->45
>4.4->22
>0.7->3.6
>0.3-> 1.7
96-500
2.3-11
0.9-4.4
1.3-6.3
0-0.1
0.1-0.7
130-170
2.9-3.8
0.7-0.9
1.3-1.6
0.7-0.9
0.2-0.3
110-140
2.5-3.2
0.6-0.7
0.4-0.5
1.2-1.5
0.3-0.4
180-370
4.5-8.4
0.9-1.7
1.8-3.4
1.5-2.8
0.3-0.6
360-570
8.5-14
8.5-14
9
~
?
(Carson 1965; Edwards 1971) strongly suggests that silica was accumulating throughout early Miocene time. The base of the Monterey thus marked mainly a reduction in local terrigenous supply and not a sharp change in oceanographic conditions. The presence of diatomaceous strata of late Oligocene age in Baja California (Beal 1948; J. A. Barron pers. comm. 1982) similarly suggests that oceanographic conditions unique to the Miocene were not the major factor that resulted in high concentration of deposited diatomaceous silica in the Miocene. The upper boundary of the Monterey is also marked by a sharp increase in the rate of accumulation of terrigenous debris (Fig. 2, Table 2). Surprisingly, silica accumulation rates markedly increased at about the same time. Sisquoc strata, although generally rather nondescript mudrocks which are not strikingly siliceous, thus appear to represent the highest rates of silica deposition in the sequence. Because of the limits of age-dating, variations in composition among interbeds are somewhat more difficult to assess than variations among members in terms of accumulation rates. Detritus-rich rocks, for example, could result either from high terrigeno,us influx or from low biogenous influx. Because detritus-rich rocks tend also to be calcite-rich (e.g. Fig. 3), however, they
seem likely to have been deposited during periods of moderate productivity and slower overall accumulation (see section on planktic biogenous input below), a relation suggesting that terrigenous debris accumulated for the most part at a slow steady rate. If terrigenous debris is assumed to represent a constant background influx among closely associated beds, then short-term variations in accumulation rates can be estimated. Such estimates for the lower calcareous-siliceous member show that accumulation rates may have averaged as much as 20-40 g/cm2-103 yr in silica-rich beds, a value comparable to rates (20-30 g/cm 2.103 yr) calculated from laminae pairs interpreted as annual varves. In contrast, detritus-rich beds in the same member would have accumulated at rates as low as 2-3 g/cm 2. 103 yr.
Terrigenous accumulation The member-average rates of terrigenous accumulation show a marked decrease (from about 11 to about 1 g/cm 2.103 yr) at the base of the Monterey Formation, low rates ( < 1 g/cm 2.103 yr) well into the late Miocene, and a gradual increase after 8 Ma (base of the clayey-siliceous member) increasing markedly in the latest Miocene (to as much as 45 g/cm 2. 103 yr). The
H e m i p e l a g i c deposits in a M i o c e n e basin, California decrease in latest early Miocene time corresponds with a combination of regional subsidence (Vedder et al. 1969) and global rise in sea-level (Vail & Hardenbol 1979). Because low rates of terrigenous accumulation continued well into the late Miocene (to 8 Ma), irrespective of several episodes of global sea-level lowering (Vail & Hardenbol 1979), rates of terrigenous sedimentation through the early part of the late Miocene were apparently mainly influenced by continued subsidence, as indicated by water depths and sediment thicknesses (see also Ingle 1981b). Marked increase in terrigenous accumulation after about 8 Ma corresponds with a moderately low stand of sea-level (Vail & Hardenbol 1979) and a major unconformity in adjacent areas, whereas continued increase in terrigenous accumulation rates after about 5.5 Ma (during rising sea-level) probably resulted from regional tectonism (Vedder et al. 1969; Ingle 1980 and references therein).
Biogenous accumulation Biogenous debris was the principal component of most sediment in the Santa Barbara basin from late early Miocene (18 Ma) to latest Miocene time (5.5 Ma). The relative abundance of biogenous constituents, however, particularly calcite and silica, varied significantly through time on scales ranging from thousands to millions of years (e.g. Fig. 3). The following sections address these variations in terms of preservation, planktic biogenous input, benthic biogenous input, and organic matter.
Silica
Detrital Minerals
5o%
487
Preservation and dissolution Dissolution of biogenous debris on the sea floor significantly alters the composition of sediment accumulated in many areas of the ocean (Lisitsyn 1972; Berger 1974). In sediments above the carbonate compensation depth, however, preserved biogenous debris generally resembles the hard-shelled part of the biotic assemblage (e.g. Uschakova 1971), and this resemblance can be attributed to rapid deposition by faecal pellets (Schrader 1971; Honjo & Roman 1978; Dunbar & Berger 1981). In fact, some areas of rapid accumulation of diatomaceous sediment receive a surprisingly complete record of the productivity in overlying waters (Kozlova & Mukhina 1967). Biogenous material deposited in the Miocene Santa Barbara basin was undoubtedly much better preserved than is generally the case in more pelagic settings. Sedimentation rates, for example, were as much as two orders of magnitude greater than rates of biogenous sedimentation in pelagic settings (Berger 1974). In addition, the extensive depositional preservation of both nannofossils and diatom frustules in lower and middle Miocene Monterey strata together with the absence of 'residual' layers free of biogenous material suggests that sediment compositions are broadly representative of the sediment deposited. An important exception is the uppermost or clayey-siliceous member of the Monterey Formation and possibly much of the overlying Sisquoc Formation. Intense upwelling during deposition of these strata may have produced a planktic assemblage strongly dominated by diatoms, but the virtual exclusion of calcareous forms (< 0.1% carbonate or < 0.1 g/cm 2.103 yr) seems unlikely. Moreover, although benthic foraminifers are abundant in all modern depositional settings analogous to that of the Monterey despite oxygen levels as low as 0.15 ml/1 (Harman 1964; Phleger & Soutar 1973; Thornton 1981), virtually no calcareous benthic foraminifers are present in the clayey-siliceous member. However, tests of arenaceous foraminifers, which are not subject to total dissolution, are abundant throughout upper Miocene strata. These relations together indicate that the absence of preserved carbonate in uppermost Miocene strata resulted from dissolution, probably promoted by bottom water that aggressively dissolved calcareous shells (Berger 1974; van Andei et al. 1975; Ingle 1981a).
Carbonate -* A p a t i t e
FIG. 3. Composition of rocks in the lower calcareous-siliceous member of the Monterey Formation (weight ~, organic-matter-free basis). Note the wide range in proportions of calcite and silica.
Planktic biogenous input Living planktic assemblages are influenced by many oceanographic and climatic factors, but the factor most likely to have influenced production
488
C.M. Isaacs
of planktic material in the Miocene Santa Barbara sea is intensity of upwelling. Upwelling, which brings nutrient-rich subsurface waters to the photosynthetic layer, is closely related to total productivity. Both phytoplankton (such as diatoms and coccolithophores) and zooplankton (such as planktic foraminifers) have their highest absolute abundances in areas of major ocean upwelling (Hasle 1960; Lisitsyn 1972; Soutar et al. 1981). Of particular importance in influencing sediment composition, however, is the relative abundance of plankton groups under varying intensities of upwelling. Partly due to significantly varying growth rates (Braarud 1962; Eppley 1970) and varying motilities (e.g. Williams 1971), the relative abundance of plankton groups differs significantly between highly productive and sparsely productive surface waters. Diatoms, in particular, are unusually dominant in highly productive upwelling waters, whereas other plankton such as dinoflagellates in the Indian Ocean or coccolithophores in the Caribbean and Mediterranean Seas (Hasle & Smayda 1960; Bernard 1967; Guillard & Kilham 1977 and references therein) are relatively more abundant in poorly productive waters. In the Atlantic, for example, diatoms comprise less than 2.5% of the plankton in plankton-poor waters compared to at least 40~ and in some places over 80% of the plankton in productive regions (Guillard & Kilham 1977 and references therein). Surface water temperature has been indirectly related by several authors to lithofacies in the Monterey Formation (Pisciotto & Garrison 1981; Barron & Keller 1983). Surface temperature per se, however, is probably a negligible influence on the sediment composition as is obvious from the high concentrations of diatoms living in upwelling regions of the Antarctic, the Gulf of California, and the equatorial regions (Lisitsyn 1972; Guillard & Kilham 1977). Moreover, temperature fluctuations in the marginal eastern north Pacific (Ingle 1981b; Barron & Keller, in press) are not closely related to silica accumulation rates in upper Miocene and lower Pliocene strata of the Santa Barbara basin. Variations in silica and carbonate sedimentation during the Miocene most likely resulted, therefore, from variations in the intensity of upwelling. That broad variations in upwelling occurred and also paralleled broad sediment changes in the Santa Barbara area is indicated by Barron & Keller's (1983) synthesis of diatom assemblages in the north-eastern Pacific. They concluded that upwelling was strong at about 15 Ma (top of the silica-rich lower calcareous-siliceous member), generally weak from 15 to 12.5 Ma (silica-poor carbonaceous marl member),
increasing from 12.5 to 11 Ma (boundary between the silica-poor carbonaceous marl member and the moderately siliceous transitional marl-siliceous member), and generally strong from 11 to 5 Ma (generally silica-rich upper three members). Such variations may relate to varying dominance of the two eastern boundary currents of the North Pacific--the equatorward California Current which, because of wind stress, tends to pull away from the coastline and thus produces strong upwelling, and the poleward North Equatorial Current which tends to push toward the coastline and thus does not produce upwelling (Wooster & Reid 1963). During early Pliocene time, the convergence of these currents (which today is near Cedros Island off Baja California) ranged as far north as the California-Oregon border (Ingle 1981b). Shorter term variations in silica and carbonate deposition also seem likely to have resulted from variations in upwelling strength. Particularly in upper Miocene strata where 1 to 2 metre cycles are common, characteristics of the lamination suggest that carbonate-rich beds were deposited during intervals when the oxygen-minimum zone was either less intense or narrower than during deposition of silica-rich beds (see section on layering below). The greater abundance of clays in calcite-rich beds may also indicate slower accumulation, consistent with a lower overall productivity during periods of comparatively weak upwelling. Benthic biogenous input
Another factor that may account for some variations in silica and carbonate accumulation is the abundance of benthic calcareous foraminifers. Living benthic assemblages are undoubtedly influenced by some environmental factors such as bottom-water temperature that are quite different from those influencing living planktic assemblages. In addition, accumulation of benthic foraminiferal tests would be increased by moderate dissolution which preferentially dissolves planktic foraminifers (Kennett 1982). Although calcareous benthic foraminifers are common in the Monterey through lower upper Miocene strata (c.8 Ma), only scant data are available on the relative abundance of calcite in benthic foraminifers. Point counts indicate that only about 10~ of the calcite determined by chemical analysis is present as recognizable foraminiferal remains (both planktic and benthic) in the massive rocks of the lower calcareous-siliceous member; the majority of calcite in these strata is therefore probably present as nannofossils. Qualitative petrographic examination sug-
489
Hemipelagic deposits in a Miocene basin, California gests that benthic foraminifers may be somewhat more abundant in overlying laminated strata; because lamination indicates low-oxygen conditions, a proliferation of species tolerant of lowoxygen stress may have resulted from low competition (Phleger & Soutar 1973; Govean & Garrison 1981). Only in rare beds, however, do benthic calcareous foraminifers represent more than about 20% of calcite.
1o ,~ 9
9
~ 5 o <5
9
9
Organic matter Several relationships are apparent in the distribution of organic matter within the Monterey Formation: its abundance is strongly positively correlated with the abundance of detrital minerals, moderately positively correlated with calcite in the calcite-bearing part of the sequence, and strongly negatively correlated with the abundance of silica. These relationships hold both on a broad scale and in detail. The most silica-poor member (the carbonaceous marl member), for example, is also the member richest in organic matter (mean 13%, max 24%); and in individual beds throughout the Monterey sequence, abundance of organic matter closely follows abundance of detrital minerals (e.g. Fig. 4). The close association between detrital minerals and organic matter may be partly due to co-deposition, as a result of adsorption of organic matter on clay within the water column and co-transport to the sea floor within faecal pellets (Dunbar & Berger 1981). Preservation may also be important inasmuch as organic matter is generally preserved preferentially in fine-grained sediments (Hunt 1979 and references therein). This relationship suggests the observed association with clay and possibly also with nannofossil calcite rather than with the larger diatom frustules and foraminiferal tests. Bottom-water oxygen levels are generally regarded as the primary influence on preservation of organic matter in the Monterey (e.g. Summerhayes 1981), but evidence from the Santa Barbara area indicates that other influences were stronger. For example, mean organic matter contents of lower Miocene strata with massive bedding (indicating well-oxygenated bottom water) match those of upper Miocene strata with varve-like laminations (indicating very low oxygen levels in bottom water). Similarly, in the upper member of the Monterey, strata alternate between laminated silica-rich units and massive detritus-rich units which contain twice as abundant organic matter. Because significantly higher organic carbon contents are therefore present in the more oxygenated (massive) sediment, adsorption of organic matter on clay (and possibly on fine-
0
210
0
i
i
40
810
60
SILICA, wt % 10 z~
m
J, o
0
0
,
,
I
I
510
10 20 30 40 DETRITAL MINERALS, w t %
10 0
o
o
0
0
I 10
210
I t , 30 40 CALCITE, w t %
5
0
FIG. 4. Relationship between the abundance of organic matter and major sediment components in the lower part of the lower calcareous-siliceous member of the Monterey Formation. The correlation coefficientfor linear regression by the least squares method is -0. 8 for silica, + 0.8 for detrital minerals, and +0.6 for calcite. grained calcite) apparently outweighed variations in oxygen levels in preserving organic matter in the sediment.
Layering In Miocene and Pliocene sedimentary rocks of the
490
C.M.
Santa Barbara area, types of layering change markedly through the Miocene sequence and also display important variations closely associated with composition. The best known and most widely described type of layering is very thin pairs of light and dark laminae generally interpreted as varves. This type of layering, however, is abundant only in upper Miocene strata, where it represents less than 50~ of beds. Lamination in hemipelagic sediments is generally ascribed to seasonal variations such as influxes of detritus due to seasonal rains and influxes of diatom debris due to seasonal upwelling (Hulsemann & Emery 1961; Calvert 1964). Careful dating of recent laminae in the Santa Barbara basin and the Gulf of California has shown that laminae pairs in these areas represent annual layers (Soutar & Isaacs 1974; Soutar & Crill 1977; Soutar et al. 1981). Laminae are preserved where benthic fauna are excluded because of low-oxygen levels in bottom water, and laminae are disturbed or destroyed where generally or occasionally higher oxygen levels permit bioturbation by benthic fauna (Gorsline & Emery 1959; Calvert 1964; Govean & Garrison 1981). Although some other factors such as seasonality of input and accumulation rates may also affect the character of layering, massive layering is generally interpreted to indicate bioturbation by an active benthic fauna and thus deposition in a well-oxygenated environment, and laminated layering is generally interpreted to indicate total absence of bioturbation and thus deposition in low-oxygen bottom waters. Layering in the Monterey is divided into three general types (after Potter et al. 1980): (1) massive bedding or vague lamination, (2) discontinuous or uneven parallel lamination, termed irregular lamination, and (3) continuous even parallel lamination, termed regular or varve-like lamination (Figs 5 and 6). In massive and vaguely laminated beds, no individual layers are apparent and the beds are completely homogeneous although the bedding plane is usually discernible. Irregularly laminated beds have a prominent planar fabric but comparatively few continuous laminae (commonly 3-10 mm apart) and some wispy or markedly discontinuous layers. Regularly laminated beds have varve-like lamination, consisting of very thin (0.05-0.2 ram) pairs of light and dark laminae which are continuous and parallel. A fourth type of layering, typical of the laminated beds in the Sisquoc Formation, is widely-spaced (0.5-1 cm) uneven parallel lamination (Fig. 5). Through time, the distribution of layering features, indicated in Fig. 2, is divided into four phases: (1) massive layering to about 16 Ma; (2)
Isaacs
FIG. 5. Idealized layering types in Miocene rocks of the Santa Barbara coastal area. The uppermost type is found only in the Sisquoc Formation. laminated layering to about 8 Ma, with pronounced variation between irregula~ and varvelike lamination especially marked in the interval 11-8 Ma; (3) alternating varve-like lamination and massive layering 8 to 5.5 Ma; and (4) predominantly massive layering in younger Sisquoc strata. The first layering phase, which is marked by the monotonously massive bedding of the lower part of the lower calcareous-siliceous member, indicates deposition in well-oxygenated waters. Because faunal evidence suggests water depths of 1000-1500 m (Ingle 1981a, b), well below the deepest upper surface of the oxygen-minimum zone, these lowest Monterey strata must have been deposited below the range of the oxygenminimum zone, and the sea floor must have subsided through the oxygen-minimum zone during deposition of the older Rincon Shale. An important corollary of this conclusion is that waters of intermediate depth had access to the basin during early Miocene time. In latest early Miocene time (c. 16 Ma), layering changed gradually from massive to laminated, predominantly irregular lamination through middle Miocene time with increasingly abundant varve-like lamination in latest middle and late Miocene time (Fig. 2). Because significant deepening or shallowing is not indicated by benthic foraminiferal assemblages, the change in layering suggests either that the lower surface of the oxygen-minimum zone deepened or that tectonism caused formation of a basin sill within the oxygen-minimum zone, excluding oxygenated intermediate waters. Some evidence supports the
Hemipelagic deposits in a Miocene basin, California
491
FIG. 6. Examples of layering types in the Monterey Formation of the Santa Barbara coastal area. Varve-like or regular lamination (top), irregular lamination (middle), and massive layering (bottom). Thickness of strata in each view is about 1.5 cm.
492
C.M. Isaacs
latter possibility inasmuch as local tectonism in the latest early Miocene (c. 16 Ma) is indicated by a widespread zone of deformation in the Santa Barbara area and by a sand zone as thick as 115 m in the Santa Maria Valley oil field 50 km to the north. Throughout the second phase and particularly marked in latest middle and earliest late Miocene time, layering is closely and consistently related to bed composition (Fig. 7). Clay-rich silica-poor beds have the broadest, most discontinuous and uneven lamination, and clay-poor silica-rich beds have the best defined, most even, varve-like lamination. These marked relations between composition and layering indicate that a weak oxygen-minimum zone (represented by irregular lamination) accompanied periods of weak upwelling (represented by sparse silica) and that an intense oxygen-minimum zone (represented by varve-like lamination) accompanied periods of strong upwelling (represented by abundant silica). Judged from bed thicknesses and average accumulation rates, different productivity levels and bottom-water oxygen levels prevailed for long intervals (500 to 20 000 yrs). In the third phase, beginning in the middle of the late Miocene (c.8 Ma), cyclic variation in layering and composition is even more pronounced. Throughout the clayey-siliceous member, beds generally alternate between massive detritus-rich units and regularly laminated silica-rich units (Fig. 8). Some beds in detritusrich units are vaguely laminated and a few are even irregularly laminated, but none have varvelike lamination. Such alternations could be attributed to oscillating levels of the oxygen-minimum zone near its upper boundary (Govean & Garrison 1981) or to weakening of the oxygen-minimum zone caused, for example, by extremely strong upwelling which reduces total productivity and may result in flushing of silled basins with oxygenated intermediate water (Soutar et al. 1981). That layering differences represent more than differences in bottom water conditions is indicated not only by compositional differences but also by comparison of laminae thickness with overall sediment accumulation rates; interpreted as varves, laminae suggest that silica-rich beds were deposited at rates of 20-30 g/cm 3.103 yr, 2 to 10 times faster than member averages. Massive units thus seem likely to represent comparatively slow accumulation during periods of weak to moderate primary productivity with either a less intense or displaced oxygen-minimum zone; dissolution of biogenous material may also have been more intense. The last phase of layering types began in latest Miocene time with deposition of the Sisquoc
FIG. 7. Typical lithologic sequence in the upper calcareous-siliceous member of the Monterey Formation showing metre-thick laminated units grading upward from detrital-rich calcareous-siliceous rock to detrital-poor calcareous-siliceous rock. From Isaacs ( 1981).
Hemipelagic deposits in a Miocene basin, California
493
Formation. Layering in the Sisquoc is distinctive in being generally characterized by thick (about 1 mm) continuous laminae within broad (about 1 cm) homogeneous bands (Fig. 5); massive, spheroidally-weathering rocks are also abundant. That these changes are not solely related to the higher influx of terrigenous material is indicated by the varves presently forming in the Santa Barbara basin in predominantly terrigenous debris (Hulsemann & Emery 1961; Soutar et al. 1981; Thornton, this volume). Because evaluation of benthic assemblages in the Santa Barbara coastal area indicates a shallowing during deposition of the Sisquoc Formation to about 700 m (Ingle 1981a, b), deposition may have been above the oxygen-minimum zone or near its upper boundary. In addition, common evidence of soft sediment deformation and cross-lamination suggests sedimentation on an unstable slope as in the Holocene Santa Barbara Basin (Thornton 1981, this volume). Layering in most strata was thus probably mainly influenced by bottom currents, slope failures, and bioturbation with occasional intervals of strata with varve-like lamination deposited under low-oxygen bottom conditions.
Synthesis
FIG. 8. Typical lithologic sequence in the clayey-siliceous member of the Monterey Formation showing units of laminated clay-poor silica-rich rock (with thin interbeds of clay-rich siliceous shale) alternating with units of massive clay-rich siliceous mudstone. From Isaacs (1981).
The base of the Monterey Formation marks a major reduction in terrigenous supply, but accumulation of biogenous silica and calcite may have been increasing gradually for several million years before this time. Lower Monterey strata were apparently deposited below the lower surface of the oxygen-minimum zone, and sediments were homogenized by bioturbation. The most abundant component of the sediment was biogenous silica (mean 41%), but significant amounts of biogenous calcite (mean 33%), detrital minerals (mean 20%), and organic matter (mean 7%) were also deposited. Member-average accumulation rates ranged from 4.5 to 8.4 g/cm 2. 103 yr for the total sediment, and from 1.8 to 3.4 g/cm 2-103 yr for silica. Rocks with high silica contents probably represent intervals of particularly high upwelling and productivity; rocks with sparse silica probably represent intervals of reduced upwelling and slower overall accumulation rates resulting in a deposit rich in nannofossil calcite, clay, and organic matter. Gradually at about 16 Ma, bottom-water conditions changed probably due either to tectonic emplacement of a sill within the oxygenminimum zone or to deepening of the base of the oxygen-minimum zone. Accumulated sediments were laminated (but, for the most part, irregularly) and benthic foraminiferal remains were
494
C.M. Isaacs
abundant. At about 15 Ma, the sediment composition changed significantly, mainly due to a sharp reduction in silica accumulation rates (to 0.5 g/cm 2-10 3 yr). The interval from 15 to 11 Ma is inferred to represent a time of weak upwelling and slow sedimentation with the accumulation of unusually abundant organic matter (mean 13%) and the formation of apatite near the sedimentwater interface. Sporadic intervals of more intense upwelling associated with extremely lowoxygen bottom water increased markedly after 11 Ma, producing highly siliceous deposits with varve-like lamination. The increasing corrosiveness of bottom water dissolved an increasing amount of calcite culminating at about 8 Ma in the virtual absence of calcite. Also at about 8 Ma, terrigenous influx to the sediment increased, and upwelling oscillated between intense and moderate. During intense periods of upwelling, when silica accumulated rapidly, bottom waters were extremely low in oxygen and sediments retained varve-like lamination, whereas during periods of reduced upwelling bottom waters supported a more abundant benthic fauna and sediments were homogenized. At about 5.5 Ma, silica accumulation rates increased, but terrigenous accumulation rates increased even more markedly (at the base of the Sisquoc Formation). Regional tectonism or cessation of subsidence probably caused shallowing to above the oxygen-minimum zone or near its upper boundary, as most sediments are massive; preservation of calcite also increased. Resulting deposits are mainly clay-rich siliceous mudrocks with sparse foraminiferal remains.
Conclusions The following conclusions are suggested for the hemipelagic Miocene sequence in the Santa Barbara coastal area: (1) The marked lithologic boundary at the base of the Monterey Formation resulted from a sharp reduction in terrigenous supply, probably due to a combination of tectonic subsidence and eustatic sea-level rise in late Oligocene and early Miocene time. Silica was accumulating at rates characteristic of the highly siliceous facies of the Monterey (2-4 g/cm 2~103 yr) at least by late early Miocene time (r 18 Ma) and probably by middle early Miocene time (r Ma). (2) Until the late part of late Miocene time (c.8 Ma) when nearly all calcite was dissolved from the sediment, variations in silica and calcite abundance are principally related to primary variations in intensity of upwelling
of thousands or millions of years duration. Silica-rich rocks represent periods of particularly intense upwelling, and calcite-rich rocks represent periods of modest upwelling and comparatively slow deposition. (3) After latest early Miocene time (c.16 Ma) when sediments began accumulating in lowoxygen bottom waters, variations in layering features indicate that the oxygen-minimum zone varied considerably in intensity for periods of thousands or millions of years duration. Silica-rich sediment (which has varve-like lamination) was deposited during periods of intense low-oxygen conditions whereas sediment rich in calcite and clay (which has irregular lamination) was deposited during periods of moderately low-oxygen conditions. (4) Within the Monterey Formation, the abundance of organic matter is strongly correlated with the abundance of detritus and also with the abundance of calcite (into the late Miocene) and is inversely correlated with the abundance of silica. Relationships between organic matter content, composition, and layering indicate that grain-size outweighed the influence of low-oxygen bottom waters in preserving organic matter. (5) The lithologic boundary between the Monterey Formation and the overlying Sisquoc Formation apparently reflects a rapid increase in terrigenous supply and deposition in shallower, more oxygenated bottom waters as a result of regional tectonism. Rates of silica accumulation also increased (by a factor of 4) across the Monterey-Sisquoc boundary (r Ma), but not as rapidly as rates of terrigenous accumulation. ACKNOWLEDGMENTS: This paper is dedicated to the late John D. Isaacs of Scripps Institution of Oceanography whose challenging questions and observations about plankton abundances initiated this evaluation. I particularly thank James C. Ingle, Jr. of Stanford University for numerous stimulating discussions of the oceans and ocean sediments and Margaret A. Keller of the US Geological Survey for invaluable field discussions. John A. Barron, Alfred G. Fischer, Robert E. Garrison, Andrew Soutar, Scott E. Thornton, and John G. Vedder also provided many helpful comments and discussions. Preliminary versions of the manuscript were reviewed by Walter E. Dean, Robert E. Garrison, James C. Ingle, Jr., Margaret A. Keller, David K. Larue, Scott E. Thornton, and John G. Vedder. I also thank Jeanne A. Blank and Brigitta V. Fulop for
Hemipelagic deposits in a Miocene basin, California drafting the figures, M a r g a r e t A. Keller for supplemental sample processing, Barbara A. G r a c i a n o for c o m p u t e r p r o g r a m m i n g , Vicki A.
495
G e n n a i for c o m p u t e r processing, and D a v i d J.W. Piper and D o r r i k A.V. Stow for helpful editorial suggestions.
References ATWATER, T. 1970. Implications of plate tectonics for the Cenozoic tectonic evolution of North America. Bull. geol. Soc. Am., 81, 3513-56. BARRON,J.A. & KELLER,G., in press. Paleotemperature oscillations in the middle and late Miocene of the northeastern Pacific. Micropaleontology. , POORE, R.Z. & WOLFART, R. 1981. Biostratigraphic summary, Deep Sea Drilling Project Leg 63. In: Yeats, R.S., Haq, B.U. et al. (eds), Init. Repts. DSDP, 63, US Govt. Print. Off., Washington, DC 927-41. BEAL,C.H. 1948. Reconnaissance of the geology and oil possibilities of Baja California, Mexico. Mere. geol. Soc. Am., 31. BERGER,W.H. 1974. Deep-sea sedimentation. In: Burk, C.A. & Drake, C.D. (eds), The Geology of Continental Margins. Springer-Verlag, New York. 213-41. BERNARD, F. 1967. Research on phytoplankton and pelagic protozoa in the Mediterranean Sea from 1953 to 1966. Oceanogr. Mar. Biol. A. Rev., 5, 205-29. BLAKE, M.C. JR., CAMPBELL,R.H., DIBBLEE,Z.W. JR., HOWELL, D.G., NILSEN, T.H., NORMARK, W.R., VEDDER, J.G. & SILVER, E.A. 1978. Neogene basin formation in relation to plate-tectonic evolution of the San Andreas Fault System, California. Bull. Am. Ass. Petrol. Geol., 62, 344-72. BRAARUD, T. 1962. Species distributions in marine phytoplankton. J. Oceanogr. Soc. Japan, 20th anniversary volume. 628-49. BRAMLETTE, M.N. 1946. The Monterey Formation of California and the origin of its siliceous rocks. US geol. Surv. Prof. Pap., 212. CALVERT, S.E. 1964. Factors affecting the formation of laminated diatomaceous sediments in the Gulf of California. In: van Andel, T.H. & Shor, G.G. (eds), Marine Geology of the Gulf of California. Mem. Am. Ass. Petrol. Geol. 3, 311-30. - - . 1966. Accumulation of diatomaceous silica in the sediments of the Gulf of California. Bull. geol. Soc. Am., 77, 569-96. CARSON, C.M. 1965. The Rincon Formation. In: Weaver, D.G. (ed.), Western Santa Ynez Mountains, Santa Barbara County, California. Coast geol. Soc. and Pacific Section, Soc. econ. PaleD. Min., Bakersfield. 38-41. CLIFTON, H.E. 1981. Progradational sequences in Miocene shoreline deposits, southeastern Caliente Range, California. J. sed. Petrol., 51, 165-184. DIBBLEE, T.W. JR. 1950. Geology of the southwestern Santa Barbara County, California. Bull. Calif. Div. Mines Geol., 150. --. 1966. Geology of the central Santa Ynez Mountains, Santa Barbara County, California. Bull. Calif. Div. Mines Geol., 186. DONEGAN,D. St. SCHRADER,H. 1981. Modern analogues
of the Miocene diatomaceous Monterey Shale of California: evidence from sedimentologic and micropaleontologic study. In: Garrison, R.E. et al. (eds), The Monterey Formation and Related Siliceous Rocks of California. Pacific Section., Soc. econ. PaleD. Min., Bakersfield, 149-57. DUNBAR, R.B. & BERGER,W.H. 1981. Fecal pellet flux to modern bottom sediment of Santa Barbara Basin (California) based on sediment trapping. Bull. geol. Soc. Am., 92, pt. I, 212-8. EDWARDS, L.N. 1971. Geology of the Vaqueros and Rincon Formations, Santa Barbara Embayment, California. Unpubl. Ph.D. Thesis, Univ. California, Santa Barbara, California, USA. - - . 1972. Notes on the Vaqueros and Rincon Formations. In: Weaver, D.G. (ed.), Central Santa Ynez Mountains, Santa Barbara County, California. Pacific Sections, Am. Ass. Petrol. Geol. and Soc. econ. PaleD. Min. Bakersfield. 46--54. EPPLEY, R.W. 1970. Relationships of phytoplankton species distribution to the depth distribution of nitrates. Bull. Scripps Inst. Oceanogr., 17, 43-50. FISCHER,P.J. 1976. Late Neogene-Quaternary tectonics and depositional environments of the Santa Barbara Basin, California. In: Fritsche, A.E. et al. (eds), Neogene Symposium. Pacif. Sec., Soc. econ. PaleD. Min. Bakersfield, 33-52. GORSLINE, D.S. & EMERY,K.O. 1959. Turbidity current deposits in San Pedro and Santa Maria basins off southern California. Bull. geol. Soc. Am., 70, 279-88. GOVEAN, F.M. & GARRISON,R.E. 1981. Significance of laminated and massive diatomites in the upper part of the Monterey Formation, California. In: Garrison, R.E. et al. (eds), The Monterey Formation and Related Siliceous Rocks of California. Pacif. Sec., Soc. econ. PaleD. Min. Bakersfield. 181-98. GUILLARD, R.R.L. & KILHAM,P. 1977. The ecology of marine planktonic diatoms. In: Werner, D. (ed.), The Biology of Diatoms. Botan. Monog., 13, 372-469. Univ. Calif. Press, Berkeley, USA. HAMILTON, E.L. 1976. Variations in density and porosity with depth in deep-sea sediments. J. sed. Petrol., 72, 157-69. HARMAN, R.A. 1964. Distribution of foraminifera in the Santa Barbara Basin, California. Micropaleontology, 10, 81-96. HASLE, G.R. 1960. Plankton coccolithophorids from the Subantarctic and Equatorial Pacific. Nytt Mag. Bot., 8, 77-88. - • SMAYDA, T.J. 1960. The annual phytoplankton cycle at Drobak, Oslo-fjord. Nytt Mag. Bot., 8, 53-75. HONJO, S. & ROMAN,M.R. 1978. Marine copepod fecal pellets: production, preservation and sedimentation. J. mar. Res., 36, 45-57.
496
C.M. Isaacs
HUNT, J.M. 1979. Petroleum Geochemistry and Geology. W.H. Freeman, San Francisco, USA. HULSEMANN, J. & EMERY, K.O. 1961. Stratification in recent sediments of Santa Barbara basin as controlled by organisms and water character. J. Geol., 69, 279-90. INGLE, J.C. JR. 1980. Cenozoic paleobathymetry and depositional history of selected sequences within the southern California continental borderland. In: Sliter, W. (ed.), Studies in Micropaleontology. Cushman Foundation Foram. Res. Spec. Publ. 19. 163-95. - 1981a. Cenozoic depositional history of the northern continental borderland of southern California and the origin of associated Miocene diatomites. In: Isaacs, C.M. (ed.), Guide to the Monterey Formation in the California Coastal Area, Ventura to San Luis Obispo. Pacific Section, Am. Ass. Petrol. Geol., Bakersfield. 52, 1-8. - 1981b. Origin of Neogene diatomites around the north Pacific rim. In: Garrison, R.E. et al. (eds.), The Monterey Formation and Related Siliceous Rocks of California. Pacific Section, Soc. econ. Paleo. Min., Bakersfield. 159-79. ISAACS, C.M. 1981. Lithostratigraphy of the Monterey Formation, Goleta to Point Conception, Santa Barbara coast, California. In: Isaacs, C.M. (ed.), Guide to the Monterey Formation in the California Coastal Area, Ventura to San Luis Obispo. Pacific Section, Am. Ass. Petrol. Geol., Bakersfield. 52, 9-23. - 1983. Compositional variation and sequence in the Miocene Monterey Formation, Santa Barbara coastal area, California. In: Larue, D.K. & Steel, R.J. (eds), Cenozoic Marine Sedimentation, Pacific margin, USA. Pacific Section, Soc. econ. Paleo. Min., Bakersfield. 117-32. KENNETT, J.P. 1982. Marine Geology. Prentice-Hall, Englewood Cliffs, N.J., USA. KOSLOVA, O.G. & MUKHINA,V.V. 1967. Diatoms and silicoflagellates in suspension and floor sediments of the Pacific Ocean. Int. Geol. Rev., 9, 1322-42. LISITSYN, A.P. 1972. Sedimentation in the World Ocean. Spec. Publ. Soc. econ. Paleo. Min. 17. LOLL, W. & COREY, W.H. 1932. Vaqueros Formation, Lower Miocene of California. L Paleontology. Univ. Cal. Pubis Geol. Sci., 22, 31-40. LUYENDYK,B.P., KAMERLING,M.J. & TERRES,R. 1980. Geometric model for Neogene crustal rotations in southern California. Bull. geol. Soc. Am., 91, pt. 1, 211-17. PHLEGER, C.F. • SOUTAR, A. 1973. Production of benthic foraminifera in three east Pacific oxygen minima. Micropaleontology, 19, 110-15. PISr K.A. & GARRISON, R.E. 1981. Lithofacies and depositional environments of the Monterey Formation, California. In: Garrison, R.E. et al. (eds), The Monterey Formation and Related Siliceous Rocks of California. Pacific Section, Soc. econ. Paleo. Min., Bakersfield. 97-122.
POTTER, P.E., MAYNARD, J.B. & PRYOR, W.A. 1980. Sedimentology of Shale. Springer-Verlag, New York, USA. SCHRADER, H.J. 1971. Fecal pellets: role in sedimentation of pelagic diatoms. Science, 174, 55-7. SOUTAR, A. & ISAACS,J.D. 1974. Abundance of pelagic fish in the 19th and 20th centuries as recorded in anaerobic sediment of the Californias. Fishery Bull., 72, 257-73. -t~ CRILL, P.A. 1977. Sedimentation and climatic patterns in the Santa Barbara Basin during the 19th and 20th centuries. Bull. geol. Soc. Am., 88, 1161-72. --, JOHNSON, S.R. & BAUMGARTNER,T.R. 1981. In search of modern depositional analogs to the Monterey Formation. In: Garrison, R.E. et al. (eds), The Monterey Formation and Related Siliceous Rocks of California. Pacific Section, Soc. econ. Paleo. Min., Bakersfield. 123-47. SUMMERI-IAYES,C.P. 1981. Oceanographic controls on organic matter in the Miocene Monterey Formation, offshore California. In: Garrison, R.E. et al. (eds), The Monterey Formation and Related Siliceous Rocks of California. Pacific Section, Soc. econ. Paleo. Min., Bakersfield. 213-19. TALIAFERRO, N.L. 1933. The relation of volcanism to diatomaceous and associated siliceous sediments. Univ. Cal. Pubis Geol. Sci., 23, 1-56. THORNTON, S.E. 1981. Holocene Stratigraphy and Sedimentary Processes in Santa Barbara Basin." Influence of Tectonics, Ocean Circulation, Climate and Mass Movement. Unpubl. Ph.D. thesis. Univ. of So. California, Los Angeles, California. USCHAKOVA, M.G. 1971. Coccoliths in suspension and in the surface layer of sediment in the Pacific Region. In: Funnell, B.M. & Riedel, W.R. (eds), The Micropalaeontology of Oceans. Cambridge University Press, Cambridge, UK. 245-51. VAIL, P.R. & HARDENBOL,J. 1979. Sea-level changes during the Tertiary. Oceanus, 22, 71-9. VAN ANDEL, TJ.H., HEATH, G.R. & MOORE, T.C. JR. 1975. Cenozoic Tectonics, Sedimentation, and Paleoceanography of the Central Equatorial Pacific. Mem. geol. Soc. Am., 143. VEDDER, J.G., WAGNER, H . C . & S C H O E L L H A M E R , J . E . 1969. Geologic framework of the Santa Barbara Channel region. US geol. Surv. Prof. Pap., 679-A. WILLIAMS, D.B. 1971. The distribution of marine dinoflagellates in relation to physical and chemical conditions. In: Funnell, B.M. & Riedel, W.R. (eds), The Micropalaeontology of Oceans. Cambridge University Press, Cambridge, UK. 91-5. WOOSTER, W.S. & REID, J.L. 1963. Eastern boundary currents. In: Hill, M.N. (ed.), The Sea. 2 Wiley Interscience, New York, USA. 253-80. WORNHARDT, W.W. JR. 1967. Miocene and Pliocene marine diatoms from California. Occ. Pap. Calif. Acad. Sci., 63.
C.M. ISAACS,US Geological Survey, 345 Middlefield Road, Menlo Park, California 94025, USA.
Sapropels and organic-rich variants in the Mediterranean: sequence development and classification George C. Anastasakis and Daniel Jean Stanley S U M M A R Y : Sapropel is the term often applied in a general manner to dark, organic-rich
layers of late Quaternary age in the eastern Mediterranean. The DSDP Leg 42A studies have more precisely defined sapropel layers as those containing more than 2.0~oorganic carbon by weight, and sapropelic layers as those that contain between 0.5 and 2.0~oorganic carbon. The present study further focuses on extensive vertical and lateral petrological variations of these and associated organic-rich layers. Herein the different sediment lithofacies that form both sapropel and sapropelic sequences in radiocarbon-dated cores distributed throughout the basin are defined. Organic carbon and carbonate carbon content and their ratio are useful criteria; some sedimentary structures and textures also help define sediment types. The recognition of the dominant vertical successions of lithofacies that form typical sapropel and sapropelic sequences is important; these provide further evidence that favour marked episodic water mass stratification and associated stagnation. The lithofacies succession typically includes the following, from base up: greyish mud becoming darker upward and passing into a grey ooze which is separated from the overlying dark grey to black sapropel (or sapropelic) layer by a sharp contact; the latter layer, in turn, is topped by a thin grey ooze and an orange-brown oxidized layer. These sediments, poor in benthonic faunas, are interbedded within lighter, well-oxygenated open marine lithofacies. Sapropel and sapropelic sequences comprise sediment types which are largely hemipelagic in origin (suspension settling through the water column); variations within such vertical successions may be directly related to palaeoceanographic events. Regional study, however, reveals numerous variations displayed by these two sequences, and thus further classification of the different organic-rich layers is needed. Lithofacies members are sometimes absent, or poorly developed, or repeated as a result of transport processes (mostly gravity-flow mechanisms) and/or specific oceanographic conditions affecting a depositional site. On the basis of the above, we identify episapropel and episapropelic sequences. These are defined as highly irregular sequences which include sediment types that are locally introduced by sediment gravity-flows, commonly mud-carrying turbidity currents. A more accurate assessment of the diverse organic-rich sequences in cores across the basin helps us to better understand the nature of short-term stagnation-oxygenation cycles that affected the eastern Mediterranean, and may also shed light on some black shales of open marine origin preserved in the rock record.
Sapropel is the term applied, usually in a general manner, to dark organic-rich layers found interbedded in normal lighter-coloured open marine sediments in the eastern Mediterranean. It is the consensus of most workers that Mediterranean sapropels are the sedimentary expression of episodic basin-wide stagnation leading to the development of anoxic conditions. These deposits thus provide strong evidence that the palaeoceanographic conditions in this basin were significantly different from those at present. The occurrence of sapropel-type (reducing) sediment in the Mediterranean had been predicted by Bradley (1938), and these deposits were eventually recovered in cores collected during the 'Albatross' expedition of 1947-1948 (Kullenberg 1952). Several hypotheses have been proposed to account for the formation of these sediments under anoxic conditions: increased precipitation and fresh-water runoff during glacial episodes
(Bradley 1938; Kullenberg 1952); warming of surface waters (Van Straaten 1972); and an association with excess fresh-water input and current reversals (Huang & Stanley 1972; Stanley 1978). Many workers have suggested that, periodically, fresh-water input from the Black Sea formed a low-salinity layer over the basin which prevented oxygenation of the bottom layer (Olausson 1961; Ryan 1972; Thunell et al. 1977; Vergnaud-Grazzini et al. 1977; Stanley & Blanpied 1981). More recently, it has been proposed that excess fresh-water overflow from the River Nile triggered stagnation (Rossignoll-Strick et al. 1982). The exact relationship between palaeoclimatic and palaeoceanographic change and the development of anoxic conditions, however, remains unresolved. Moreover, deep-sea drilling (DSDP Leg 42A in 1975) recovered sapropels as old as middle Miocene, an indication that sporadic stagnation of the eastern Mediterranean 497
498
G.C. Anastasakis and D.J. Stanley
Basin occurred prior to the development of Northern Hemisphere glaciation in the late Pliocene (Kidd et al. 1978). There have been many studies of sapropels. The faunal and oxygen isotopic characteristics of these deposits have received the most attention (Parker 1958; Pastouret 1970; Herman 1972; Ryan 1972; Cita et al. 1973; Blanc et al. 1975; Thunell et al. 1977; Vergnaud-Grazzini et al. 1977; Kidd et al. 1978; Williams & Thunell 1979; Cita & Podenzani 1980; Mullineaux & Lohman 1981; Muerdter 1984). Aspects of their geochemistry, bulk mineralogy and clay mineralogy have been described (Nesteroff 1973; Sigl et al. 1978; Dominik & Mangini 1979; Mangini & Dominik 1979; Dominik & Stoffers 1979; Anastasakis 1981). Some attention has been paid to the development of Upper Quaternary sapropels in different parts of the basin (Maldonado & Stanley 1976; Anastakis 1981; Maldonado et al. 1981) and older sequences recovered during DSDP Legs 13 and 42A (Nesteroff 1973; Kidd et al. 1978). Review of the above-listed literature indicates that, most often, the term sapropel has been applied loosely to a diverse suite of organic-rich sediments. There is a need to more precisely define the different types of organic-rich deposits which have been termed sapropels. The objective of the present sedimentologic study is to provide a description and a classification of Upper Quaternary sapropels and variants in the eastern Mediterranean basin. This is a first step needed to better understand the conditions leading to their deposition.
Methods Piston and gravity cores (24 in number) containing Upper Quaternary sapropels serve as the basis for this study. Coring locations lie south of the Adriatic and Marmara seas (Fig. 1). Other cores, containing Upper Quaternary sapropels and organic-rich deposits, including 14 cores recovered in the vicinity of the central Hellenic Trench (Anastasakis 1981) and some from the Adriatic, Marmara and Black seas, were examined for comparative purposes. Core materials include those collected on several cruises by various organizations: N/O Ariane (1976 cruise); R/V Bannock (1981 cruise); R/V Chain (1975 cruise 119); R/V Conrad (1965 cruise 9); USNS Lynch (March 1972 cruise); R/V Pillsbury (1965 cruises 6507 and 6510); R/V Trident (1976 cruises 171, 172); and R/V Vema (1958 cruise 14). Some older sapropels recovered during DSDP Leg 42A (Fig. 1) also were examined (Anastasakis & Stanley, in
preparation). Organic-rich sections in all the above cores were X-radiographed and logged, and a total of 680 samples were selected for petrologic analysis (structures, texture, petrography and chemistry). Particular attention is paid herein to the Upper Quaternary sapropels identified and correlated in radiocarbon-dated cores distributed throughout the basin. The results presented are based largely on treatment of approximately 360 samples from the suite of 24 cores. Small bedforms and structures were revealed in X-radiographs, and organic and total carbon content was determined. The carbonate carbon content (given in percent by weight) was obtained by calculating the difference between total carbon and organic carbon. Carbon analysis was performed with a Leco combustion system. Details of the methods, listing of data and depiction of logs are presented in Anastasakis & Stanley (in preparation).
Lithofacies types forming sapropel sequences On the basis of examining numerous core samples we accept the premise that organic-rich lithofacies provide evidence of episodic, well-defined stratification and resulting stagnation as well as a return to fully oxygenated deep-water conditions (i.e. Ryan 1972). Proceeding one step further, we recognize a suite of distinct lithofacies that may be correlated with a series of transitional phases that occur from complete anoxia to oxygenation, and vice-versa (Stanley & Maldonado 1979). The typical complete eastern Mediterranean sapropel sequence sensu stricto, modified somewhat from that described by Maldonado & Stanley (1976), includes the following members from base to top (Figs 2, 3(a)): a greyish-greenish yellow mud passing into a yellowish-grey organic ooze becoming darker upward and separated from the overlying light olive-grey sapropelic or olive-grey sapropel sediment by a more or less pronounced sharp contact. The latter member, in turn, is topped by a thin light greenish-grey ooze and a pale yellowishorange to moderate brown oxidized layer. The basal greyish-greenish yellow mud and upper oxidized layer end-members are interbedded between lighter coloured, better-oxygenated sediments below and above the sequence. The greyish-pale greenish-yellow mud (5Y 8/4 to 10Y 8/2) contains a restricted range (between 0.12 to 0.230/o) of organic carbon (Fig. 2), but highly variable amounts of carbonate carbon. However, its carbonate carbon to organic carbon ratios are always high and restricted, between 26 and 37 (Fig. 2); these extreme values are generally
Sapropels and organic-rich variants in the Mediterranean , i
'~
l
20~
i
i 2~
i
i
i
2~
r
26~
499
Ip_~o;_, ~ 3~. P-6507-IB
, 32"
P.65~0, P-6507-6 .~6507-7G
L! ,>"(
% 10-3G
"-~ RC-9-176 TR-172-20
LY -Tr- 2
c
,
TR-172-19 t
~ ;
~-~,,Iw,-
,71-21 TR-171-19
TR-171-26(
;.._.
--._--
~ >p~
c,,b
CHN-IIg-
9
j
C,Ng,-~,
zoo --
~o
t
Fio. 1. General bathymetric map of the eastern Mediterranean and adjacent seas showing sample locations (depth in metres). The black dots indicate core (piston and gravity) position and asterisks show drilling sites of DSDP Leg 42A considered in the present study. References in text are made primarily to the cores south of the Adriatic (AS), Marmara (MS) and Black seas (BS). Cores are from the following cruises: AR = N/O Ariane (1976); BAN =N/O Bannock (1981 cruise); CHN = R/V Chain (1975 cruise 119); RC= R/V Conrad (1965 cruise 9); DSDP = Deep Sea Drilling Project (1975 Leg 42A); LY--USNS Lynch (March 1972 cruise); P = R/V Pillsbury (1965 cruises 6507 and 6510); TR= R/V Trident (1976 cruises 171 and 172); V= R/V Vema (1958 cruise 14).
reached by the end-members of this lithofacies. Some bioturbation structures are observed on X-radiographs, and both planktonic and benthonic forminifera are dominant in the coarse silt to sand fractions. This facies is shown as grey hemipelagic mud in Fig. 3. The yellowish-grey to light greenish-grey to pale olive (5Y 7/2 to 5G 8/1 to 10Y 4/2) organic ooze has a somewhat higher (0.18 to 0.49%) organic carbon (Fig. 2) and generally lower carbonate carbon content than the underlying greyish mud. It is always the most bioturbated unit in a core, the mottling in most cases ceasing approximately two centimetres below the contact with the overlying sapropel. The intensity of bioturbation in this lithofacies appears to have an effect on the organic carbon content; the lowest
values are observed in samples recovered from some intensely bioturbed core sections, and the highest values are found in the non-bioturbated section below the sapropel. The carbonate carbon to organic carbon ratios range between 7 and 26 (Fig. 2). The organic ooze often present above the sapropel is similar to this lithofacies, containing between 0.22 to 0.34~o organic carbon, values which are somewhat lower than the ooze underlying the sapropel. The carbonate carbon to organic carbon ratios lie in the same range as those from the organic ooze below the sapropel, although they never reach the extreme values. The carbonate carbon content of the ooze above the sapropel is generally higher than that of the ooze below it. The oozes always contain a large proportion of planktonic and benthonic fora-
50o
G.C. Anastasakis and D.J. Stanley
Co) s.n 4.0
.:"
3.o
.!" ..
2.0 ,.R "' "~ "~
SAPROPEL,SAPROPELIC
9
ORGANIC
OOZE
9
OXIDIZED
LAYER
x
GREY M U D
,
9
TR-171-21 TR-172-10 TR-172-19 TR-172-37 BAN-81-1OGC
LY-TT-2
'-'
o.'~ o.7 0.5
2
0.3
i
i
i
i
=
i
i
15
11
CARBONATE
i
i
i
19
i
i
23
CARBON/ORGANIC
i
27
i
i
31
i .
i
35
CARBON
(b) TR-171-27 T R - 171-28 TR-172-8 TR-172-20 CHN-IIg-6 CHN-II9-19 CHN-II9-31
Z= 9
,
II CARBONATE
FIG. 2. Graphic plots of organic carbon content versus the carbonate carbon/organic carbon ratio for numerous samples from cores widely distributed throughout the Eastern Mediterranean and South Aegean seas (core locations shown in Fig. 1). The samples plotted include all members of the sapropel (sensu lato) sequence. (a) cores showing primarily sapropel and sapropelic sequences; (b) cores showing sapropel-sapropelic and vaguely-defined episapropel-episapropelic sequences; (c) cores showing primarily episapropel and episapropelic sequences. The quadrants 1 to 4 cited in text and indicated in (a) also apply to (b) and (c).
9
,
15
,
.
,
19
,
9
,
23
CARBON/ORGANIC
9 ,
27
9 ,
i
i
i
31
CARBON
(c)
5.0 4.0 3.0
20
'.
TR-172-32 TR-172-33 TR-172-34 TR- 172-35 TR-172-36 AR-IO
9"?~
1.9 1.7
~ 1.5 ~u 1.3 ~ 1,1 0 ~ o.9 I).7 0.5
9
"lo
9
0.3
we
9
=.
9
0.1 i
&
3
i
!
11 CARBONATE
15
19
CARBON/ORGANIC
I
23
27
CARBON
31
35
I
Sapropels and organic-rich variants in the Mediterranean minifers. The ooze overlying the sapropel is only rarely bioturbated. The sapropel is greyish-olive to dark olive-grey or even greyish-olive green (10Y 4/2 to 5Y 3/2 to 5GY 3/2) and, as defined by Kidd et al. (1978), always contains more than 2~ organic carbon (but rarely ekceeds 7~, Fig. 2). The carbonate carbon content is generally lower than in the overlying ooze, a distinction which is not always observed between the sapropel and the underlying ooze. The carbonate carbon/organic carbon ratio is always less than 4; the lowest values (less than one) characterize the most organic-rich members (Fig. 2). The sapropel layers are devoid of benthic fossils and very rarely contain bioturbation structures (commonly Chondrites, Planolites, Zoophycos). The frequently observed lightcoloured micropatches, which resemble microbioturbation, most probably are escaping gas bubble structures with subsequent oxidation of the cavities. The sapropelic facies has a light olive-grey to greyish-olive (5Y 5/2 to 10Y 4/2) colour and, as defined by Kidd et al. (1978), contains between 0.5 and 2 ~ organic carbon (Fig. 2). The carbonate carbon content is generally lower than in the overlying ooze; as in the case of sapropels, this distinction is not always made with the underlying ooze. The carbonate carbon/organic carbon ratios range between two and nine (Fig. 2). Generally, the sapropelic deposits are devoid of benthic foraminifera and only less than 570 of them contain bioturbation structures (Chondrites, Planolites etc.). Light-coloured micropatches, also observed commonly, are possibly due to the oxidation of the space occupied by escaping gas microbubbles. The oxidized layer is pale yellowish-orange to moderate brown (10YR 8/6 to 5YR 4/4) and contains between 0.22 to 0.4570 organic carbon (Fig. 2). It always contains less carbonate carbon than the underlying (if present) ooze and the overlying sediment. Dark brown to black laminae are commonly inter-layered with orange-yellow laminae in oxidized layers that generally overlie sapropels. These dark brown laminae frequently contain abundant black material in the form of discrete, irregularly shaped particles and dark stains on microfaunal tests. X-ray diffraction analysis suggests that this material is poorly crystallized. It probably includes birnessite (Anastasakis 1981) and other iron monosulphide minerals may also be present [greigite (Fe3S4) and mackinawite (Fe Sx-1)]. The carbonate carbon to organic carbon ratios range between nine and 26 (Fig. 2), and in this respect are comparable to the oozes. The oxidized layers are not bioturbated.
501
Sequence developmentof organic-rich deposits Sapropel sequences The five above-described sediment types and their variants comprise the basic lithofacies constituents of a Mediterranean sapropel sequence. Complete sapropel sequences present a predictable succession of sediment types that, as will be discussed later, are the depositional expression of a distinct oceanographic condition affecting the hydrography of the basin. Sequences also frequently record the effects of depositional processes (Stanley & Maldonado 1979). A sapropel sequence as defined here must always contain a sapropel layer at least one centimetre thick, so that it can be clearly identified. From base up, a complete sapropel sequence (Fig. 3(a)) starts with a greyish-greenish mud passing upwards into an organic ooze that, in turn, is overlaid by the sapropel layer. The contact between organic ooze and sapropel is well pronounced. The sapropel layer with its relatively constant organic carbon content passes upward to an organic ooze and oxidized layer. A core section depicting this complete sequence is hemipelagic in the sense that it is deposited largely by settling through the water column (cf. Stanley & Maldonado 1979). Significant variations may exist with respect to the succession of sapropel lithofacies and organic carbon content. Three variants are depicted in Figs 3 and 4. In the case of the first (Fig. 3(b)), sapropel sediments overlie (with a sharp contact) organic ooze; organic carbon contents above the contact exceed 2 ~ and remain generally high until the uppermost sapropel section where organic carbon sharply decreases below 2~o (but remains over 0.5~). At this point the sapropel becomes a sapropelic sediment which is, in turn, separated from the organic ooze above it by a contact. This succession is noted in X-radiographs (Fig. 4b). A second variant (Fig. 3(c)) is characterized by an ooze overlain by a sapropelic layer. The contact between the two is generally well-defined and the sapropelic layer very rapidly passes upward into the sapropel sediment which records a high (over 270) organic carbon content. The sapropel is separated from the overlying ooze by a sharp contact as seen in Fig. 4(c). In the case of a third variant (Fig. 3(d)), the organic ooze is separated from the overlying sapropelic layer by a poorly-defined contact which, in turn, rapidly passes upward into a sapropel layer. Progressing upward, the sapropel becomes sapropelic and is separated from the organic ooze above it by a sharp contact (Fig. 4(d)).
G.C. A n a s t a s a k i s and D.J. Stanley
502
SAPROPEL SEQUENCE
rule, especially in areas on and in proximity to slopes. These include basinal areas near the base of slopes where shelf or upper slope sediments 4 accumulate, and intraoceanic depressions and 2 trenches surrounded by structural highs and steep 3 slopes that are isolated from shelf-derived sedi2 ments. We recall that a large part of the eastern I Mediterranean occupies a tectonically active area, one where sediment is readily dispersed by 1 ~ - [ ~ GREY HEMIP. MUD 4 ~ OXIDIZED LAYER gravity processes to deep-water environments (Stanley & Maldonado 1981). As a consequence, 2 ~ ORGANICOOZE 5~-[]- SAPROPELIC many organic-rich sequences provide some evidence of redeposition. Episapropel sequences 3 1 SAPROPEL typically display successions of reorganized sediFIG. 3. Schematic illustration of the (a) typical ment types with a plethora of sedimentary struceastern Mediterranean sapropel sequence (sensu tures. Many structures are indicative of high stricto) and (b)-(d) variants thereof. The numbers energy regimes, including emplacement by sedialong A refer to the basic sediment types-lithofacies ment gravity-flows such as turbidity currents. members of the typical eastern Mediterranean Nesteroff (1973) was the first author to call sequence. Actual examples are shown by the core attention to current-deposited and expanded X-radiographs in Fig. 4. sapropels recovered in the eastern Mediterranean during the DSDP Leg 13. Subsequent drilling There may be other differences observed in the during Leg 42A confirmed the common occursuccession of lithofacies members of a sapropel rence of current-emplaced sapropels (by gravitysequence. Although greyish mud and organic flows and possibly by bottom currents), and it has ooze are generally developed below the sapropel become obvious from the work of Kidd et al. layer, the succession above the sapropel may (1978) that considerably more detailed work on vary. In one case, the organic ooze passes upward such 'anomalous' sequence development is into an open marine facies that likely accumu- needed. We thus introduce the term episapropel to lated under well-oxygenated conditions (Figs 5(a) and 6A. In another case, the sapropel layer is group the many varieties of sequences which followed directly by an oxidized layer which then include gravity-emplaced lithofacies incorporpasses upward into sediments that presumably ated within and on the basic hemipelagic sedirecord fully oxygenated conditions (Fig. 5(b) and ment types. An episapropel sequence is generally 6B). It is noted that in the two sequences de- an expanded sapropel sequence that contains at scribed above the sapropel lithofacies show least a sapropel layer (with organic carbon > 2%). sapropel-sapropelic variations of the type This latter may be emplaced either by hemipelagic observed in Fig. 3(b), (c) and (d) are frequently suspension settling or by some form of gravirecorded. ty--(and possibly bottom current) flow, or a In all the sequence variants described above, coupling of the two mechanisms (Stanley 1983). the sediment lithofacies comprising the sapropel In one extreme case, an episapropel sequence sequence are essentially hemipelagic deposits. contains only a portion of the sapropel layer, and These do not display structures in X-radiographs this may display varve-like lamination (see Fig. indicative of physical processes such as those 5(c), numbers 3 and 12). This lamination is the produced by sediment gravity-flows and bottom result of transport processes and not annual currents (Figs. 4(a)-(d) and 6(a), (b)). Good cycles. At the other extreme, this sapropel layer is examples of these sapropel sequences occur on extensively interrupted by non-sapropelic layers the Mediterranean Ridge south-west of Crete. (Fig. 5(c), numbers 6 and 13). Episapropel However, on the basis of examination of cores sequences generally incorporate different succesthroughout the eastern Mediterranean, we find sions of sediment types; some of these facies are that the above complete sapropel sequences not as well-defined as those in the sapropel account for less than 10% of organic-rich sequences. Moreover, lithofacies members of the sequences. sapropel sequence are missing, or poorly-developed, or repeated as a result of the above-cited transport processes which result in redeposition Episapropel sequences and/or erosion. Although colours in split cores Sequences consisting of essentially hemipelagic suggest typical sapropel development, X-radiodeposits tend to be an exception rather than the graphs reveal that such sequences are, in fact, (a)
(b)
(c)
(d)
Sapropels and organic-rich variants in the Mediterranean
503
FIG. 4 X-radiographs of sapropel sequences of the type illustrated schematically in Fig. 3. (a)= sapropel sequence; (b)-(d)= variants. The different sediment types comprising these sapropel sequences are indicated by a number corresponding to the legend in Fig. 3. Core locations are shown in Fig. 1. Discussion in text.
expanded by the presence of several redeposited layers. Good examples of episapropels occur in the central Hellenic Trench. Sapropelic sequences
Another important variant is the sapropelic sequence which displays the same succession of sediment types as a sapropel sequence, but with a significant difference: a sapropelic layer (organic carbon content between 0.5 and 2%) rather than a sapropel layer is developed. Three sapropelic sequence variants are depicted in Fig. 7(a), (b) and (c). The typical complete sequence (Fig. 7(a)) starts with a greyish-pale greenish mud passing upwards into an organic ooze which is separated from the sapropelic layer by a vaguely defined
contact. The sapropelic layer is succeeded by an organic ooze and, on top of this, an oxidized layer (Figs 7(a) and 8(a)). The sapropelic sequence shown in Figs 7(b) and 8(b) is similar to 7(a), but here an organic ooze above the sapropelic layer passes upward directly into well-oxygenated sediments (Fig. 8(b)). In a third case (Figs 7(c) and 8(c)), the sequence is comparable to (a) but the sapropelic layer is directly overlain by an oxidized layer. All sediment types which constitute the various sapropelic sequences described above (Figs. 7(a) and 8(c)), are hemipelagic-suspension settling deposits. Good examples of sapropelic sequences occur locally in the south-eastern Mediterranean on the periphery of the Nile Cone, and on topographic highs less than 1200 metres in depth.
504
G.C. Anastasakis and D.J. Stanley
SAPROPEL SEQUENCE EPISAPROPELSEQUENCE (a)
(b)
(c) 7 I.,~.,I~
Irll Illl-1~
6
GREY HEMIP.MUD 81~I~ I EPISAPROPELIC=Td (OF BOUMASEQUENCE)
2[~/7/~ ORGANICOOZE
9~ !
EPISAPROPELIC,"ret
aI
SAPROPEL
10~
EPISAPROPEL;Ta-Tc
4~
OXIDIZED LAYER1 1 ~
51TIT! SAPROPELIC 6~
(OF BOUMASEQUENCE)
{OF BOUMASEQUENCE)
EPISAPROPEL:Td (OF BOUMASEQUENCE)
12~-ZI (OF EPISAPROPEL"Tet BOUMA SEQUENCE)
TURBIDITICSILT 1 3 ~
TURBIDITICSAND
7~1 TURBIDITICMUD Episapropelic sequences Often, pronounced variations of sapropelic sequences are observed, particularly with respect to assemblages of sediment types. Many of these have been emplaced by gravity-controlled mechanisms. An episapropelic sequence is defined herein as a generally expanded sapropelic sequence that contains at least one sapropelic layer (which may, but need not be, deposited by hemipelagic-suspension settling), together with a series of other layers, for the most part deposited by gravity-flow and/or bottom current transport (Figs 7(d) and 8(d)). Thus, episapropelic sequences are expanded, and the associated lithofacies which form them may be well-defined, or poorly-developed, or missing or repeated. Good examples of episapropelic sequences occur in the Nile Cone and in several of the small Aegean Sea basins. Both episapropel and episapropelic sequences are extremely variable (Figs 5(c) and 7(d)). Among the enigmatic variants are the ones that display numerous vertical variations such as varve-like couplets (Fig. 8(d)), numbers 8, 9 and Fig. 6(c), numbers 9, 10, 11) and marked compositional variations within the sapropel-sapropelic lithofacies. These structures most likely result from fairly rapid emplacement involving rather short dispersal. Moreover, high amounts of organic material in these varve-like structures within the sapropel and sapropelic lithofacies probably modified the settling behaviour of particles. These structures likely result from downs-
FIG. 5. Schematic illustration of (a), (b) variants of the eastern Mediterranean sapropel sequence and (c) of an episapropel sequence. The latter is a hypothetical example showing possible sediment types and attributes that have been observed in a number of cores. Actual examples are shown by the core X-radiographs in Fig. 6.
lope reworking and segregation of the organicrich particles into distinct layers (Stanley 1983) rather than from seasonal input of alternating organic-poor and organic-rich sediments (cf. McCoy 1974). This hypothesis is supported by examination of stratigraphically synchronous sapropel and sapropelic layers in areas where there is good radio-carbon dated sample control. Episapropel (> 2% organic carbon) and episapropelic (0.5-2% organic carbon) layers (Figs 5(c), numbers 8-12; Fig. 7(d), numbers 8, 9, 14) display sedimentary structures visible on X-radiographs such as vertical grading and distinct size-segregated bands. These structures are indicative of gravity-flow rather than suspension settling. The vertical succession of lithofacies types varies markedly from core site to core site, showing the importance of local transport events in the development of both episapropel and episapropelic sequences.
Sapropel sequences as palaeoceanographic indicators In their review, Demaison & Moore (1980) indicate that organic-rich layers are deposited under stagnant and also in fully oxygenated brackish to lacustrine waters, as well as in open marine waters under a wide range of conditions. Most workers, including ourselves, tend to accept the original definition of sapropel, that is, a sediment rich in organic matter formed under
Sapropels and organic-rich variants in the Mediterranean
5o5
FI~. 6 X-radiographs of sapropel sequences (a),(b) and an episapropelic sequence (c) of the type illustrated in Fig. 5. Numbers refer to sediment types identified in Fig. 5. Core locations are shown in Fig. 1. Discussion in text.
reducing conditions in a stagnant water body (Wasmund 1930; Pontoni6 1937). As indicated in the first part of this paper, the term is now used to include a wide range of organic-rich lithofacies. There generally is no implication as to the specific origin of the organic matter. Insofar as the Mediterranean is concerned, most workers believe that sapropels record palaeoceanographic
conditions, for the most part associated with oxygen-deficient conditions (Olauson 1961; Sigl et al. 1978). Anoxic sea floor environments in open marine basins in general occur in (a) areas of intense upwelling and (b) below stratified waters, that is, along and below a distinct thermocline, pycnocline or halocline (temperature, density and salinity boundaries, respectively) or combina-
G.C. Anastasakis and D.J. Stanley
506
SAPROPELIC SEQUENCE
(a)
(b)
(c)
EPISAPROPELIC SEQUENCE (d)
7 14 ,9 -6
--... ?
28 \1
2~
ORGANICOOZE
e~
TURBIDITIC MUD EPISAPROPELIC:Td
~.~
OXIDIZED LAYER9 ~
EPISAPROPELIC:Te'
sliT[- ~ SAPROPELIC 6~
13~
TURBIDITIC SILT 1 4 ~
(OF BOUMA SEQUENCE) {OF BOUMA SEQUENCE)
TURBIDITICSAND EPISAPROPELI BOUNA I SEQUENcE)T(I C,(oF -TC
tions thereof. The eastern Mediterranean sapropels likely were deposited below stratified waters as suggested by Bradley (1938), along and below a pycnocline-halocline boundary (Stanley 1978). Together, the structures, texture and composition of sapropel and sapropelic lithofacies should serve as reliable indicators of physical and chemical conditions under which they were deposited. An assessment of the development of the Mediterranean sapropel sequence is attempted by reviewing the succession of lithofacies types. We believe that the initiation of the stagnation phase is marked by deposition of the greyish mud associated with a decrease in the amount of the dissolved oxygen in the bottom waters. As this condition continues, life on the sea floor becomes unfavourable for most organisms. Only the most adaptable forms, including some benthonic foraminifera, remain; the enhanced preservation of organic matter in sediments (Fig. 2), due to reduced decay, attracts sediment-grazing (perhaps worm-like) organisms (see facies 2 in (a) and (b), Fig. 4). This is expressed by the transition of the greyish mud to grey ooze which always displays the greatest degree of bioturbation of any sediment type forming the sapropel sequence. Sometimes it appears that the ooze progressively grades upwards to a sapropelic layer (the protosapropel layer depicted by Maldonado & Stanley 1976). Closer inspection, however, reveals that a contact always separates the ooze from the overlying sapropelic layer (with organic carbon >0.5%). The absence, or scarcity, of bioturbation in the ooze immediately below the contact suggests that the concentration of dissolved oxygen in the bottom waters began to reach the tolerance limits for even the worm-like organisms
FIG. 7. Schematic illustration of sapropelic sequences (a--c) and episapropelic sequence (d) from the Eastern Mediterranean. The latter is a hypothetical example showing possible sediment types and attributes that have been observed in a number of cores. Actual examples are shown by the core X-radiographs in Fig. 8.
(upper part of facies 2 in (a),(b),(c) in Fig. 4, and (a) and (b) in Fig. 6). This attribute of the ooze thus signals the gradual establishment of anoxic conditions. The marked increase in organic carbon content in the sapropel and sapropelic lithofacies marks the phase of complete oxygen depletion. Marked oxygen minimum variations occurred laterally and with time during a single stagnation event. This is recorded by pronounced organic carbon variations within a sapropel lithofacies at any one location, and the concurrent deposition of both sapropel and sapropelic sequences in different parts of the basin. This latter is demonstrated by good stratigraphic control using numerous radiocarbon-dated samples. The transformation ofa sapropelic layer to a sapropel layer in time and space indicates that their deposition resulted from dissolved oxygen fluctuations rather than from differences in surface waters productivity. This conclusion is based on analysis of numerous time-equivalent organic-rich samples which show no marked time-space variation in the ratio of the total organic matter substance (including amorphous organic matter) to the organic matter of recognizable remains, mostly dinoflagellates visible in SEM photographs. The deposition of time-equivalent sapropel and sapropelic sequences at the same water depths in different parts of a basin probably indicate the presence of variable oxygen content within the oxygen-depleted water column. The accumulation of an ooze above the sapropel or sapropelic layer records a gradual return to oxic waters; the deposition of an oxidized layer directly above a sapropel or sapropelic layer would, on the other hand, indicate a more abrupt return of oxygenated waters. There may have been an increase
Sapropels and organic-rich variants in the Mediterranean
507
FIG. 8 X-radiographs of sapropelic (a)-(c) and episapropelic sequences (d) of the type illustrated in Fig. 7.
Numbers refer to sediment types identified in Fig. 7. Core locations are shown in Fig. 1. Discussion in text.
in organic productivity (perhaps up to three times that in normal oxic Mediterranean waters) during stagnation episodes in the eastern Mediterranean. This is suggested by the enhanced organic carbon content in the oxidized layer deposited just after the above stagnation events. Increased productivity alone, however, could not have induced oxygen depletion sufficiently to produce stagnation. The correlation among episapropel and episapropelic sequences and palaeoceanographic events is not clear, particularly in cases where sequences largely result from gravity-flow deposition. Those sequences which contain primarily hemipelagic lithofacies and only one or a few deposits emplaced by high energy mechanisms (gravity-flow, bottom currents) can shed some light on palaeoconditions. We recall that it is the hemipelagic lithofacies which are of most use in
interpreting palaeoceanography. The degree of relation between adjacent members is noted on the organic carbon versus carbonate carbon/organic carbon diagrams (Fig. 2). In some cases the episapropel and episapropelic layers (Fig. 5(c), numbers 5, 9, 12) show no sequence relation to the adjacent lithofacies members (Fig. 5(c), numbers 7, 13) of a sequence. In other cases, there may be good sequence continuity (Fig. 5(c), numbers 10, 11; Fig. 7(d), numbers 5, 9, 14). When, as in the case of sapropel sequences, there is continuous deposition of hemipelagic lithofacies (Fig. 2(a)), all samples plot nearlinearly and continuously within quadrant 1 (all sapropel and sapropelic types) and quadrant 4 (organic ooze, oxidized layer, grey mud). In cases where gravitative mechanisms have emplaced some layers in expanded episapropel and episapropelic sequences, some samples plot in other
508
G.C. Anastasakis and D.J. Stanley
portions of the graph. For example, the samples which plot to the right of a vertical line drawn perpendicular to the carbonate carbon/organic carbon ratio at 7 indicate gravity-emplaced episapropelic layers (quadrant 2, Fig. 2(b),(c). Samples which plot to the left of this line (at 7) and below the 0.5~ organic carbon line (quadrant 3, Fig. 2(c), also indicate gravity-transported layers interbedded within episapropel and episapropelic sequences. Organic carbon contents overlap, especially in adjacent members within a sapropel sequence, but carbonate carbon/organic carbon ratios of each specific lithofacies are clearly distinguished (Fig. 2 quadrants 1 & 4). As plotted the relation between organic carbon and carbonate carbon is most revealing because the carbonate-rich sediment that settles through the water column is directly influenced by two contrasting environments, the well-oxygenated water above the thermocline-halocline boundary and anoxic water below (Sigl et al. 1978). This results largely because of differences in oxygen and hydrogen sulphide content in the water column which results in dissolution of the carbonates as particles settle through the column below the thermocline-halocline boundary. The petrological-chemical classification presented in this study helps us to better interpret the organic-rich deposits which accumulated during the last major stagnation episode in the eastern Mediterranean at approximately 8000 yrs. BP (Ryan 1972). During this event a variety of sapropel-sapropelic and episapropel-episapropelic sequences of the type shown in Figs 3, 5 and 7 were deposited. There are sites in the basin where the upper-most and best-defined Holocene sapropel (or sapropelic) sequence is represented by only a few centimetres of sapropel (or sapropelic) layer. Time-equivalent episapropel (or episapropelic) sequences contain up to several metres of organic-rich sediment. Thus, a few centimetres of sapropel (or sapropelic) layer accumulated during the same stagnation time-span as the episapropel (or episapropelic) layer. The latter may be a hundred times thicker due to numerous interbedded episapropelic and non-sapropelic layers, some very fine-grained and without obvious sedimentary structure. In conclusion, recognition of sapropel-sapropelic and correlative episapropel-episapropelic sequences in the deeper basins of the eastern Mediterranean can be used to more precisely assess the causes and development of stagnation. Moreover, continuing investigation indicates that the Quaternary sapropels and their variants described in this study are comparable to older (upper Miocene to Quaternary) sequences reco-
vered during DSDP Leg 42A in this basin. From preliminary observations it would appear that systematic study of Mediterranean organic-rich deposits and their relation to oxic-anoxic cycles may bear on the interpretation of some open marine black shales in the rock record.
Summary (I) Detailed examination of late Quaternary organic-rich layers in radiocarbon-dated cores from the eastern Mediterranean show remarkable petrologic variations. The present study indicates that the term sapropel sequence s e n s u s t r i c t o in this basin should be reserved for the following succession (from base up) of lithofacies: grey hemipelagic mud, organic ooze, sapropel, organic ooze and oxidized layer. (2) The succession of types forming the sapropel sequence records the progressive oceanographic changes that influenced the basin: diminution of oxygen in the deeper water masses; further deterioration of oxygenation; complete anoxia throughout the basin; restoration of some oxygen in the water column; and rapid return of fully oxygenated conditions above the sea floor. (3) The five basic lithofacies types constituting the sapropel sequence, essentially of suspension settling origin, can be distinguished on the basis of physical and biogenic structures, textures and composition, including the organic carbon content. (4) The sapropel lithofacies is characterized by a high (over 2%) organic carbon content and low carbonate carbon/organic carbon ratio; the other three types (organic ooze, oxidized layer and grey hemipelagic mud) have lower organic carbon content (<0.5%) and show progressively higher carbonate carbon/organic carbon ratios. (5) Some organic-rich sequences, also emplaced by essentially suspension settling mechanisms, have a lower (0.5~ to 2.0%) organic carbon content. These, termed sapropelic, can be correlative with true sapropels, and record oxygen variations within the oxygen-depleted water mass over some parts of the basin. (6) Herein the variants of both sapropel and sapropelic sequences which display marked irregularities in the lithofacies sequence and are termed episapropel (organic carbon > 2%) and episapropelic (organic carbon, 0.5%-1.9%) are identified. These latter
Sapropels and organic-rich variants in the Mediterranean sequences, also correlative with true sapropels, generally occur on or near slopes and are expanded sequences that include flowemplaced layers such as mud turbidites. (7) Examination of open marine organic-rich layers throughout basins such as the Mediterranean reveal a large number of variants, as characterized by composition and sequence development, that lend themselves to classification. This investigation suggests that the petrologic-chemical approach is applicable to the study of dark organic-rich sediments in the adjacent Black, Adriatic, Marmara and Red Seas. (8) Recognition of the many variants and definition of their sequence development can be used to better correlate organic-rich layers throughout a basin and interpret the palaeoceanographic conditions (oxic-anoxic cycles) that lead to their origin. Facies analysis of these deposits may serve to better understand some open marine black shales in the ancient record.
509
ACKNOWLEDGMENTS: This study is based on cores available in the Smithsonian collection supplemented by core materials from the University of Rhode Island (1976 R/V T R I D E N T cruises 171 and 172), Woods Hole Oceanographic Institution (1975 R/V C H A I N cruise 119), Lamont-Doherty Geological Observatory of Columbia University (1958 R/V V E M A cruise 14 and 1965 R/V C O N R A D cruise 9; supported by ONR grant N00014-67-A-0108-0004 and NSF grant GA-35454; 1981 B A N N O C K cruise), and University of Paris VI (1976 N/O A R I A N E cruise). Our sincere appreciation is expressed to these organizations. We thank Mssrs. H. Sheng, and J. Nelen, who assisted in most laboratory operations and Dr. R. Stuckenrath for providing numerous radiocarbon dates. Drs. S.E. Calvert and J.W. Pierce reviewed the manuscript. This investigation, part of the Mediterranean Basin (MEDIBA) Project funded by Smithsonian Scholarly Studies grant 1233S201, was completed while the first author held a Smithsonian Postdoctoral Fellowship.
References ANASTASAKIS, G. 1981. Sedimentation and Shallow Structure of the South Cretan-Karpathos sector of the Hellenic Trench System. Unpubl. Ph.D. thesis. Keele University, England. 392 pp. BLANC, F., BLANC-VERNET, L., LAUREC, A., LECAMPION, J. & PASTOURET,L. 1975. Application palOo6cologique de la m6thode d'analyse factorielle en composantes principales: Interpr6tation des microfaunes de foraminif6res planctoniques Quaternaires en M6diterranOe. II. Etude des esp~ces de M6diterran6e Orientale. Palaeogeo., Palaeoelimac, Palaeocol., 18, 293-312. BRADLEY, W.H. 1938. Mediterranean sediments and Pleistocene sea-levels. Science, 88, 376-9. CITA, M.D., CHIERIC~, M.A., CIAMPO, G., ZEI, M., D'ONOFRIO, S., RYAN, W.B.F. & SCORZIELLO, R. 1973. The Quaternary record in the Tyrrhenian and Ionian Basins of the Mediterranean Sea. In: Ryan, W.B.F. & Hsii, K.J. et al., Init. Repts OSDP, 13. Natl. Science Found., Washington, DC. 1263-339. & PODENZANI, M. 1980. Destructive effects of oxygen starvation and ash falls on benthic life: A pilot study. Quat. Res., 13, 230-41. DEMAlSON,G.J. & MOORE,G.T. 1980. Anoxic environments and oil source bed genesis. Bull. Am. Ass. Petrol. Geol., 64, 1179-209. DOMINIK, J. & MANGINI, A. 1979. Late Quaternary sedimentation rate variations on the Mediterranean Ridge, as results from the Th-230 excess method. Sed. Geol., 23, 95-112. -& STOFFERS, P. 1979. The influence of Late Quaternary stagnations on clay sedimentationin the Eastern Mediterranean Sea. Geol. Rundsch., 68, 302-17.
-
-
HUANG, T.-C. & STANLEY,D.J. 1972. Western Alboran Sea: sediment dispersal, ponding and reversal of currents. In: Stanley, D.J. (ed.), The Mediterranean Sea--A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa.. 521-59. HERMAN, H. 1972. Quaternary eastern Mediterranean sediments: micropaleontology and climatic record. In: Stanley, D.J. (ed.), The Mediterranean Sea--A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa.. 129-47. KIDD, R.B., CITA, M.D. & RYAN, W.B.F. 1978. The stratigraphy of Eastern Mediterranean sapropel sequences recovered by DSDP Leg. 42A and their paleoenvironmental significance. In" Hsfi, K.S. Montadert, L. et al., Init. Repts. DSDP, 42. Part 1. Natl. Science Found., Washington, DC. 421-43. KULLENBERG, B. 1952. On the salinity of the water contained in marine sediments: Medd. Oceanogr. Inst. Goteborg, 21, 1-38. MALDONADO,A. & STANLEY,D.J. 1976. The Nile Cone: Submarine fan development by cyclic sedimentation. Marine Geol., 20, 27-40. --, KEELING, G. & ANASTASAKIS, G. 1981. Late Quaternary sedimentation in a zone of continental plate convergence--The Central Hellenic Trench System. Marine. Geol., 43, 83-110. MANGINI, A. & DOMIN1K, J. 1979. Late Quaternary sapropel on the Mediterranean Ridge: U-budget and evidence for low sedimentation rates. Sed. Geol., 23, 113-25. McCoY, F.W., JR. 1974. Late Quaternary Sedimentation in the Eastern Mediterranean Sea. Unpubl. Ph.D. Thesis, Harvard University, USA. 123 pp.
5IO
G.C. Anastasakis and D.J. Stanley
MULLINEAUX, L.S. & LOHMAN, G.P. 1981. Late Quaternary stagnations and recirculation of the Eastern Mediterranean: changes in the deep water recorded by fossil benthic foraminifera. J. Foram. Res., 11, 20-39. MUERDTER, D.R. 1984. Low salinity surface water incursions into the Strait of Sicily during sapropel intervals in the Late Quaternary. Marine Geol., (in press). NESTEROFF,W.D. 1973. Petrography and mineralogy of sapropels. In: Ryan, W.B.F., Hsii, K.J. et al., Init. Repts. OSDP, 13, Natl. Science Found., Washington, DC. 712-20. OLAUSSON, E. 1961. Studies of deep sea cores. Repts. Swedish Deep-Sea Exped., 8, 337-91. PARKER, F.L. 1958. Eastern Mediterranean foraminifera. Repts. Swedish Deep-Sea Exped. 1947-1948, 8, 217-83. PASTOURET, L. 1970. Etude s6dimentologique et pal6oclimatique de carottes pr61ev6es en M6diterran6e orientale. Tethys, 2, 227-66. PONTONII~,R. 1937. Die nomenklature der unterwasserablagerungen. Jb. preuss, geol. Landesanst. Bergakad., 58, 426-8. ROSSIGNOL-STRICK, M., NESTEROEF, W., OLIVE, P., VERGNAUD-GRAZZIN1, C. 1982. After the deluge: Mediterranean stagnation and sapropel formation. Nature, (Lond.), 295, 105-10. RYAN, W.B.F. 1972. Stratigraphy of Late Quaternary sediments in the Eastern Mediterranean. In: Stanley, D.J. (ed.), The Mediterranean Sea--A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 149-69. SIGL, W., CHAMLEY,H., FABRICIUS,F., D'ARGOUD, G., MI~LLER, J. 1978. Sedimentology and environmental conditions of sapropels. In: Hsfi, K.J., Montadert, L. et al., Init. Repts. DSOP, 42A. Natl. Science Found., Washington, DC. 445-65.
STANLEY, D.J. 1978. Ionian Sea sapropel distribution and Late Quaternary palaeoceanography in the Eastern Mediterranean. Nature, (Lond.), 274, 149-52. 1983. Fine parallel laminated deep-sea muds and coupled gravity flow-hemipelagic settling in the Mediterranean. Smith. Contr. Marine Sci., 19, 27 PP. -& BLANPIED, C. 1981. Late Quaternary water exchange between Eastern Mediterranean and the Black Sea. Nature, (Lond.), 285, 537-41. & MALDONADO,A. 1979. Levantine Sea-Nile Cone lithostratigraphic evolution: Quantitative analysis and correlation with paleoclimatic and eustatic oscillations in the Late Quaternary. Sed. Geol., 23, 37-65. -& MALDONADO,A. 1981. Depositional models for fine-grained sediment in the western Hellenic Trench, Eastern Mediterranean. Sedimentology, 28, 273-90. THUNELL, R.C., WILLIAMS,D.F. & KENNETT,J.P. 1977. Late Quaternary paleoclimatology, stratigraphy and sapropel history in Eastern Mediterranean deep sea sediments. Mar. Micropaleont., 2, 371-88. VERGNAUD-GRAZZINI,C., RYAN, W.B.F. & CITA, M.B. 1977. Stable isotope fractionation, climatic change and episodic stagnation in the Eastern Mediterranean during the Late Quaternary. Mar. Micropaleont., 2, 353-70. WASMUYD,E. 1930. Bitumen, sapropel and gyttja. Geol. For. Stockh. Forh., 52, 315-50. WILLIAMS, D.F. & THUNELL, R.C. 1979. Faunal and oxygen isotopic evidence for surface water salinity changes during sapropel formation in the Eastern Mediterranean. Sed. Geol., 23, 81-93.
-
-
G.C. ANASTASAKISAND D.J. STANLEY, Division of Sedimentology, NMNH, Smithsonian Institution, Washington, DC, 20560, USA.
The sedimentology of the Swedish Alum Shales A. Thickpenny SUMMARY: The middle and upper Cambrian (and locally lowermost Tremadoc) Alum Shale Formation of Scandinavia and the Baltic Sea area outcrops from Finnmark in north Norway to Bornholm in the south, and from Estonia westwards to the Oslo region. The Formation is generally underlain by shallow water sandstones but is locally unconformable on Precambrian continental basement; it is overlain by shallow marine glauconitic, phosphatic limestones. Detailed examination of borehole cores and some quarry sections from several of the Swedish Lower Palaeozoic outliers indicates the occurrence of five lithofacies: (1) Black laminated mudstone, rich in organic carbon, suggesting extreme anoxic conditions; (2) organically banded mudstone showing more variable organic input: (3) grey mudstone, laminated or unlaminated and lacking significant organic matter, suggesting more oxic conditions; (4) bituminous, fossiliferous limestone concretions also indicative of more oxic conditions, possibly above wavebase; and (5) glauconitic siltstones and sandstones, graded and transported, and quartz rich siltstones, mostly transported. Pyritic laminae may also occur within any of the above mudstone lithofacies. Sedimentation rates for some of the above facies are estimated and show values comparable with modern pelagic sedimentation, the lowest occurring within the bituminous limestones and progressively increasing to interlaminated black and grey mudstones. The sedimentation rates in combination with the lithological variation suggest less extreme reducing conditions in the middle Cambrian compared with the upper Cambrian, a factor important in the control of Uranium and other trace element anomalies found within the upper Cambrian. The Alum Shale Formation of Scandinavia and the Baltic Sea area (Fig. 1) is of early middle Cambrian to early Tremadoc age. Interest has been shown in the deposit for some 300 years since the salt alum was first mined. This century, the deposit has been opencast mined for oil and uranium which occur in high concentrations, and it is presently being assessed for trace metal extraction. The formation is part of a Cambrian and lower Ordovician relatively shallow marine succession deposited unconformably on Precambrian crystalline basement forming the western part of the Baltoscandian platform (Thorslund 1960). Outcrops can be conveniently separated into those of the Southern Swedish outliers where little structural deformation has occurred, and those of the Caledonide front and range, where structural style is dominated by thrust and nappe tectonics. Over much of the area, however, the stratigraphy of the Cambrian succession shows relatively little variation. Figure 2 is a generalized stratigraphic section for the Billingen-Falbygden area (Dahlman & Gee 1977). The Mickwitzia and Lingula sandstones are shallow marine deposits and the latter is locally conglomeratic, glauconitic and phosphatic in its upper part. They thicken southwards and westwards from central Sweden and in places are overlain by thin grey-green siltstones and shales that show variable transitions into the overlying Alum Shale. The limestones overlying the Alum Shale are sandy and
normally glauconitic and phosphatic. Thus, units above and below the Alum Shale are demonstrably of relatively shallow marine origin, and by inference so is the Alum Shale. The main Alum Shale lithology is black to dark brown organic-rich mudstone, with subordinate fossiliferous bituminous limestone concretions ('orsten' in Swedish). Minor intercalations of grey mudstone occur, mainly in the middle Cambrian, together with localized glauconitic sandstones and thicker packets of grey siltstones. The distribution of limestone concretions is highly variable although at two horizons in southern Sweden there are more continuous marker bands known as the Great Stinkstone and the Exporrecta Stinkstone (Fig. 2). Apart from local occurrences within the mudstones in SkS_ne (Scania), fossils are confined to the concretions. Using the concretions, a detailed biostratigraphy has been developed by Westergfird (1922, 1940, 1942, 1944a and b, 1947, 1948) with modification by Henningsmoen (1957). Figure 3 shows the zonation scheme, mainly after the SkS,ne section, adopted for the Alum Shale in several of the southern Swedish outliers. Because trilobites are preferentially preserved within discontinuous limestone concretions, their absence in any one section is not necessarily indicative of non-deposition or erosion, but probably due to their low preservation potential. However, the biostratigraphic control is generally good, allowing reliable sedimentolo5II
512
A.
Thickpenny
FIG. 1. Distribution of autochthonous and parautochthonous Cambrian deposits in Scandinavia and the Baltic area.
The sedimentology of the Swedish Alum Shales
FIG. 2. Generalized lower Palaeozoic stratigraphy of the Billingen area.
513
A. Thickpenny
5 I4
ALUM
SHALE
BIOSTRATIGRAPHY GEOGRAPHICAL AREA RANGES zll,I
FOSSIL
ZONES
oZ o~, .E~~
:0
--.,I "z
I~
mlk
m :O
T
t
Ceratopyge
Diclyonema flabelhforme Acerocare
q_
Peltura scarabaeoides
i
Peltura minor Prolopeltura praecursor
I
Leptoplastus ~ Eurycare Parabobna spinu&sa & Orusia Olenus & Hypagnoslus obesus Agnostus pisiformis Lejopyge laevigata Solenopleura brachymefopo (,3
Ptychognostus lundgreni d Goniognostus nothorsti Ptychognostus punctuosus
u) :5
.E
Hypagnostus parvifrons 01
Tomagnostus fissus PLy_c_b_~nostus atavus Ptychagnostus gibbus
~h
Eccoparodoxides pinus Eccoparadoxides insuloris FIG. 3. Middle and upper Cambrian Biostratigraphy.
_L_
e
:o
The sedimentology of the Swedish Alum Shales gical and geochemical correlation over large geographical areas. Previous geochemical studies have been confined mainly to the Billingen-Falbygden, Nfirke and Osterg6tland areas where a high Uranium peak has been identified in the P. scarabaeoides zone (Fig. 3) of the upper Cambrian (Dahlman 1962; Andersson 1974, 1977, 1979, 1980a, band c; Armands 1972; Edling 1974). Recent unpublished data of Sveriges Geologiska Unders6kning show similar uranium peaks together with other trace element anomalies in the J~imtland area of the Caledonide front (Fig. 1).
Stratigraphy Southern Swedish outliers
The Alum Shale in southern Sweden is wellknown stratigraphically and is relatively constant in thickness, generally 15-25 m, thickening in Skgme to 80-100 m. Variations in thickness and lithology are summarized in Fig. 4, which also includes representative columns from the Caledonides. In SkS.ne, P. paradoxissimus mudstones rest on lower Cambrian sandstones, locally silty or calcareous. Mudstones, locally sandy near the base of the formation and with rare stinkstones continue up into the Tremadoc, to be overlain by glauconitic phosphatic limestones. Organic carbon contents within the mudstones are lower than elsewhere in southern Sweden (Dahlman & Eklund 1953). The outliers ofV/isterg6tland (Halleberg-Hunneberg, Kinnekulle and Billingen-Falbygden) possess similar stratigraphy (Fig. 2)~ Billingen being the most studied area (Dahlman & Gee 1977). Deposition of Alum Shale commenced in P. paradoxissimas times (as in SkS,ne); the shale overlies lower Cambrian sandstones, conglomeratic at the top. The stinkstone content is much higher throughout, compared with SkS,ne, and thin Dictyonema shales are locally present at the top of the Alum Shale. A more detailed description is given in Andersson et al. (1983). (3sterg6tland, although in close proximity to Billingen, exhibits three major stratigraphic differences. The formation thins eastwards (Thorslund 1960) from 25 to 14 m, the P. paradoxissimus stage is developed as grey mudstone rather than Alum Shale and the Dictyonema shales are thicker and of greater extent. As the formation thins, stinkstones become more common, more conglomeratic and more phosphatic. In Nfirke the Alum Shale, 12-19 m thick, is further reduced in time span since the P. forchhammeri stage is
515
dominated by stinkstone and overlies grey and grey-green mudstone, and Ordovician limestones directly overlie the P. scarabaeoides zone mudstones. On (3land, the Alum Shale thins to < 1 m and is entirely up.per Cambrian in the North, while in southern (31and it reaches 24 m in thickness and about half is Tremadocian. The oldest stinkstone is of P. forchhammeri age and overlies non-Alum Shale lithologies (Westerghrd 1940). Several zones are apparently missing in the upper Cambrian, which contains progressively more stinkstone northwards. Recent drilling in central Gotland has identified Tremadoc and upper Cambrian black mudstones at depths of around 400 m (D.G. Gee, pers. comm. 1982). Stinkstones and some sandstones are subordinate lithologies in the sequence, which is up to 4.5 m thick. Swedish Caledonides
The stratigraphy of the Lower Palaeozoic succession in the Caledonides, compared to the southern Swedish outliers, is complicated by complex deformation especially in the incompetent Alum shales. However, both autochthonous and allochthonous sequences can be recognized (Gee et al. 1978) along the mountain front. Autochthonous late Precambrian and Lower Palaeozoic sediments are preserved along the eastern margin of the Caledonides between basement and overthrust allochthon. The Alum shale normally contains bituminous limestone concretions and varies slightly in thickness and lithology from north to south (Gee & Zachrisson 1979). In the southernmost Swedish Caledonides it rests on lower Cambrian shallow marine sandstones and is overlain by mid-Ordovician shales (Tegengren 1962). Northwards the sole thrust cuts out the Alum Shale as far as central J/imtland (Fig. 1) where an upper Cambrian black shale overlies middle Cambrian grey-green shales (Thorslund 1960). Uppermost Cambrian zones are missing suggesting a possible hiatus, since they are overlain by lower Ordovician phosphatic calcarenites. Local basement highs active during deposition have also been recognized (Thorslund 1940) on the basis of local thinning of the mudstone. In northern J/imtland about 15 m of middle and upper Cambrian black shales are preserved beneath the sole thrust. In south V/isterbotton, Lid6n (1910) recognized 40 m(?) of autochthonous upper Cambrian black shales overlying middle and lower Cambrian grey and green silts and shales. In northernmost Sweden (Tornetr/isk, Fig. 1) the autochthon reoccurs and is referred to as the 'Hyolithes' zone. Kullings (1964) estimate is of 40 m of probably middle Cambrian black
516
A. Thickpenny
d
,o ...a 0
.....,
E
o
,4
The sedimentology of the Swedish Alum Shales shales, lacking stinkstones, in the Tornetr/isk area. In Finnmark, northern Norway, Reading (1965) has described about 200 m of black shale of upper Cambrian or Tremadoc age, although further information as to detailed palaeontology and sedimentology is at present lacking. This is the thickest succession in Scandinavia and may imply either greater subsidence during deposition, or a small basinal area. Allochthonous late Precambrian and Lower Palaeozoic deposits are preserved in a relatively undisturbed state in northern and central JS.mtland, where a fairly detailed lithostratigraphy has been produced (Gee et al. 1974) including the Alum Shale equivalent, the Fj~illbr/inna Formation. Sediments immediately underlying the formation have yielded lower and middle Cambrian fauna and are shallow marine quartzites and shales. Black shales of the Fj/illbr~inna Formation are 50-55 m thick, thicken slightly westwards, and the base is often marked by a conglomerate (up to 1 m thick) with pyritized shale clasts (Gee et al. 1974). Stinkstones are present, particularly in the upper part, and non-bituminous limestones are locally present near the top. Tremadoc black phyllites are associated with pillow lavas and dolerite dykes, further west at the Norwegian border (Gee 1981). Black phyllites with the characteristic Alum Shale geochemistry are also found as far west as TommerS, s in central Norway and similar rocks, of probable Cambrian age, occur locally in the middle Allochthon which has been thrust from west of the lower Allochthon (Gee & Zachrisson 1979). Sedimentation rates in the Alum Shale
An important facet of the depositional environments for the Alum Shale is the relative sedimentation rates in the various successions. The method used for estimation of sedimentation rates is similar to that of Churkin et al. (1977). The SkS,ne section was selected to determine zonal lengths within the middle and upper Cambrian because it shows the least lithologic variability and contains fauna within the mudstones. The total time period of the Alum Shale (P. p a r a d o x i s s i m u s to D i c t y o n e m a ) is estimated to be c.23-25 my (mainly after Cowie & Cribb 1978) although the disparity in ages given by Gale et al. (1979), Harland et al. (1964), McKerrow et al. (1980) and Ross et al. (1978) for the Cambrian time-scale testify to the poor radiometric age control. An average thickness for each zone in Skhne was calculated from palaeontologically zoned boreholes (Westerggtrd 1944a) and the Alum
5I 7
Shale is assumed to have an almost constant rate of sedimentation since lithology varies little and there are no apparent breaks in succession: thus zone thickness is proportional to duration (Fig. 5). Measurements of the compaction ratio of mudstone to stinkstone at Ranstad Quarry Billingen, show that the mudstone is now about 20% of its original thickness (assuming that stinkstones are effectively uncompacted). Hence the original thickness of the mudstone may be estimated. Using the zonal durations derived from SkS,ne, sedimentation rates have been calculated for the two cores illustrated in Fig. 6, and the results are given in Table 1. Where possible, different lithofacies have been separated; the detailed palaeontology allowing this, is not shown in Fig. 6. The figures calculated, though by no means exact due to unknown variables, do give an estimate of sedimentation rates within the deposit.
Lithofacies Detailed examination of hand specimen and thin section material of borehole cores shows several lithofacies. Five lithotypes are recognized throughout the Alum shale, and two borehole core logs that show typical vertical distributions of the lithotypes are shown in Fig. 6. Black laminated mudstone
This lithotype consists of dark brown to black mudstone (grain-size < 10/~m) with a high content of total organic carbon (TOC), ranging from 5-7% in the middle Cambrian to 20~ locally in the upper Cambrian (Armands 1972). The organic matter is fine-grained and presumably of algal origin; it is present as compressed and discontinuous laminae separated by less organic rich areas. The characteristic fine parallel lamination is caused mainly be slight changes in organic matter content. Trilobites are almost invariably absent and there is no disturbance of laminae due to burrowing. Trace fossils are lacking. The mineralogy has been estimated by Armands (1972) from a borehole in the Billingen area. The dominant silicate minerals are illite (30~o in the upper Cambrian, 40% in the middle Cambrian), quartz (variable around 25%) and K-feldspar (5-15%). Chlorite, found in small quantities near the base of the formation is absent higher up the succession. The other main constituent is pyrite, forming 5-15% of the rock. The clay minerals form a fabric with the plates flattened parallel to lamination and constitute the
518
A. Thickpenny
NAME OF BOREHOLE
THICKNESS IN METRES
ANDRARUM GISLOVSH. TOSTERUP AVERAGE S. SANDBY ANDRARUM 1 2
ZONE D. flabelliforme Acerocare Peltura Lep toplas tus P. spinulosa Olenus A. pisiformis P forchhammeri P. paradoxissimus
9.6 7.6 20.6 1.3 7.5 12.7 6.1 12.8 11"7
7.4 2.3 10"5 11.4 4.5 5.1 10.8
8"7 8"7 11"5 2"2
16"5 4"6 20"4 1-1 5"5 6"8 3"9 3"8 12 "1
7.7 18"0 1.4 9.1 9"8 1"5 5"0 15"3
13 "1 8"0 17"6 1"7 8"2 10"2 4"0 6"7 12"4
DURATION (my)
3.7 2-2 5.0 0.5 2.3 2.9 1.1 1.9 3.5 ,,,
F1G. 5. Middle and upper Cambrian zonal thickness in Sk~ne (Scania) and estimated periods of duration.
Estimation o f sedimentation rates for different lithologies in two boreholes f r o m Ostergotland and Billingen
TABLE 1.
L YCKHEM BH Zone
Thickness now Original thickness
Dictyonema Total: 2.4 m
Sedimentation rates
Total: 11.2 m Orsten: 0.2 m Mudst.: 11 m % Orsten: 1.8
Mudst. + Orsten 3 m/my
Orsten: 0.2 m Mudst.: 2.2 m % Orsten: 8
Upper Cambrian (11.8 my)
Total: 5.7 m Orsten: 2.6 m Mudst.: 3.1 m % Orsten: 46
Total: 18.1 m Orsten: 2.6 m Mudst.: 15.5 m % Orsten: 14
Total: 1.3 mm/103 Great Stinkstone: 0.15 mm/103 yr
Middle Cambrian (5.4 my)
Total 6.1 m Orsten: 0.5 m Mudst.: 5.6 m % Orsten 8
Total: 28.5 m Orsten: 0.5 m Mudst.: 28 m ~ Orsten: 1.8
Mudst. + O r s t e n 5.3 mm/103 yr
(3.7 my)
= 3 mm/103 yr
Remaining mudst.: 3.4 mm/103 yr
DBH 16/74 Using the same method as outlined as above, the following results were obtained: Zone
Sedimentation rate
Peltura Great stinkstone A. pisiformis P. forchhammeri P. paradoxissimus
Mudstone + Orsten: 7.5 mm/10 3 yr 0.27 mm/103 yr 3 mm/10 3 yr Mudstone + Orsten: 16 mm/103 yr Mudstone + Orsten: 10 mm/103 yr
The sedimentology of the Swedish Alum Shales
519
FIG. 6. Sedimentological logs of the Alum shale in two bore-hole cores from Osterg6tland and Billingen.
520
A.
Thickpenny
dal in shape and up to 2 m in diameter. Often bedding is continuous from the concretions into adjacent shale beds, and may converge towards the margins (Henningsmoen 1974) due to concretion growth during sediment compaction. Concentric growth rings are also apparent in some examples (Henningsmoen 1974). Central parts of concretions are normally of arenitic grain-size, comprising uncompressed, fragmented trilobite remains and calcite crystals. The margins are often recrystallized to fibrous cone-in-cone calcite crystals, up to 10 cm in length. In many cases, the nodules appear to have amalgamated laterally and may show preferential growth along certain horizons giving a ridged appearance (Henningsmoen 1974). At several horizons, conglomeratelike layers have developed, though clasts are angular and do not appear to have been transported far. Organically-banded mudstone Concretions generally contain more than 80% calcium carbonate (Henningsmoen 1974). The The dark brown organically-banded mudstone has a similar mineralogy to the black laminated carbonate contains organic matter, which immudstone. However, there are lower TOC con- parts a brown colour and bituminous odour. The tents, always < 10% and often closer to 5%. Slight remaining matrix is of normal mudstone compovariation in TOC content causes colour banding sition. In thin-section, the majority of the carbonate consists of rounded sand-sized grains of in hand specimen. random orientation set in a poorly laminated matrix. Fossil fragments are distinguishable together with filamentous growths sub-parallel to Grey mudstone bedding, which also occur unattached in mudGrey mudstone occurs dominantly in the lower stones, as crystalline growth. At the margins there part of the Alum Shale formation (for example, are fewer carbonate grains and they appear to see Fig. 6). Two types of grey mudstone can be infill pores in the mudstone. The large fibrous recognized. The first is less common, occurring crystals have grown roughly perpendicular to the locally in the Tremadocian Dictyonema shales, concretion margins, and have apparently and consists of pale grey, non-calcareous deformed the surrounding mudstone as replacive laminated mudstone, with gradational but paral- cone-in-cone growth. lel upper and lower boundaries. The second type occurs mainly near the base of the Alum Shale and consists of grey-green mudstone, commonly Sandstones and siltstones with calcareous laminae, and is generally ex- Thin beds and laminae of sandstone or siltstone tremely friable and poorly laminated. Contacts occur locally near the base of the Alum Shale, and with interbedded darker laminae are generally may be glauconitic. 1-2 cm thick glauconitic uneven. A shelly fauna including bivalve and sandstone beds occur in core DBH 16/74; the brachiopod fragments may be present, for exam- glauconite is associated with quartz and both ple in the Lyckhem core (Fig. 6) where the minerals show a rounded appearance. Glauconite lithology grades downwards into grey-green grains are discrete and well-formed showing no mudstone lacking all darker beds. evidence of intermediate stages in formation. Beds are often graded from sharp, slightly erosive bases and show good sorting. The matrix is Bituminous limestone (stinkstone or orsten) mudstone. Other detrital minerals such as rutile, Bituminous limestone concretions occur chiefly staurolite and plagioclase are present in small as discontinuous or semi-continuous lenses in all amounts, and pyrite nodules are common. One parts of the Alum Shale, but more commonly in bed in core DBH 16/74 shows convolute-laminathe upper Cambrian. Locally, in southern tion of glauconitic sand and darker mudstone and Sweden, two continuous 'bands' occur; the Great may be a slide horizon. Several thin silty horizons are also visible in Stinkstone and the Exporrecta Stinkstone. The concretions are generally roughly ellipsoi- thin-section, consisting of quartz laminae, a few
extremely fine-grained groundmass. Quartz is also fine-grained and is distributed randomly throughout the sediment. K-Feldspar is partly detrital and partly authigenic, the latter form becoming dominant upwards through the formation. Fine-grained pyrite is ubiquitous in the groundmass but pyrite also occurs as scattered blebs and as laminae. Laminae often occur in 'packets', mainly in the upper Cambrian. Laminae are 1-2 mm thick and in packets 2-5 cm thick. The laminae generally show a gradational contact with surrounding mudstone, often showing pyrite concentrated towards the base and decreasing gradually upwards. In thin-section, the pyrite shows a fine-grained framboidai form. Diagenetic calcareous laminae, of similar thickness, may be associated with the pyrite laminae.
The sedimentology of the Swedish Alum Shales grains in thickness and with occasional glauconite grains, generally slightly coarser than the quartz. Thin non-glauconitic beds and laminae also occur near the base of the Alum Shale. Two types of laminae can be recognized. The first consists of pale < 1 cm thick quartz-rich laminae showing sharp contacts and a poorly-defined lamination, and has only been observed in core DBH 16/74. The second type is common throughout the lower part of the Alum Shale and consists of quartzrich, siltier layers, one to two grains thick, forming parallel lamination in the mudstone and only visible in thin section. Vertical and lateral distribution of lithofacies
Typical vertical variation of sediment types is shown by the sedimentary logs in Fig. 6, which shows a progressive transition from grey mudstone with fauna via interlaminated mudstones to black mudstone, probably reflecting a transition from oxic to anoxic conditions. The change is better shown in the Lyckhem borehole, where sedimentation rates can also be seen to roughly halve, from the grey mudstone dominated middle Cambrian to the predominantly black mudstone of the upper Cambrian. A similar, though less marked trend is visible in core DBH 16/74, with the additional features of glauconitic sandstones and quartz-rich siltstones. Bituminous limestone concretions occur predominantly in black mudstones but rarely in grey mudstone. A similar vertical facies change is reflected in many other cores from the southern Swedish outliers. Large-scale lateral facies changes are also present (Fig. 4). The upper Cambrian part of the formation varies least: slight thickness changes and the presence or absence of limestone concretions being the most common variables. The upper part of the middle Cambrian comprises mostly black mudstone, though conglomeratic limestones also occur, together with interlaminated black and grey mudstone. This is a common lithofacies association occurring as mmscale alternations of dark brown or black mudstone and medium-pale grey mudstone. Contacts between the two are variable, often sharp and straight at both boundaries, but lower contacts of the grey mudstone may be uneven and partly mixed with underlying black mudstone. Parallel lamination occurs locally within the grey mudstones and there is no obvious grading. Calcareous pods and laminae are more common where black and grey mudstone are interlaminated and where nodules of pyrite are present, they are generally confined to the grey laminae. The lower part of the middle Cambrian, however, shows considerable lateral facies variation with only
52I
local development of black mudstone (e.g. Skgme), while elsewhere greater thicknesses of non-Alum Shale lithology are developed (e.g. Gland). Tremadocian Dictyonemashales are variably developed in thickness and character; bituminous limestone may or may not be present and locally grey mudstone and siltier mudstone may occur. Where the Dictyonernashale is lacking, it is not clear whether it was eroded prior to subsequent limestone deposition, or whether it was never deposited. Evidence of erosion is also lacking in areas where parts of the upper Cambrian are not present (e.g. Nfirke).
Discussion Facies interpretation
The fine grain-size of the black mudstone and its lack of current induced sedimentary structures suggest deposition in quiet water below wavebase. High organic carbon contents, syngenetic pyrite and the absence of a fauna or trace fauna suggest anoxia at the sediment-water interface. The calculated sedimentation rates of 3 to 8 mm/103 years are similar to modern pelagic oozes (Berger 1974) although the depositional environment of the Alum Shale was on a shelf; a situation with no present-day analogues. The laminated mudstone facies was probably deposited below the aerobic/anaerobic boundary within the water column where the anoxic layer impinged on a large shelf. Thickness variations are slight in the upper Cambrian, where this lithofacies is dominant, suggesting there was no significant ponding. Slight variations in organic matter content are on a time-scale too large to be explained by annual variations and must be due to longer-term climatic changes. The occurrence of pyrite in laminae, mainly within the black mudstone facies, suggests they are primary sedimentary deposits, formed in a similar way to the mechanism described by Berner (1970). Silt laminae in the silt/sandstone facies have not been preferentially pyritized so it is unlikely that the pyritic laminae are related to grain-size variation. Controlling factors in pyrite formation are amounts of organic matter, reactive iron, sulphate and elemental sulphur. However, in the case of the Alum Shale, none of these factors have conclusive supporting evidence for controlling the production of pyrite laminae. The fact that pyritic packets are more common in the thinner sequences of Alum shale in Sweden may suggest slower sedimentation and/or decreased terrigenous input. Thus, the pyrite-producing reactions could take place with a continuing
522
A. T h i c k p e n n y
supply of organic matter, but in the absence of dilution by a terrigenous component. In addition, variations in elemental sulphur content may have been controlled by slight vertical movements of the oxic-anoxic boundary. The depositional environment of the dark brown mudstones was similar to that of the black mudstone. The variations in organic matter may be explained either by a variable input of organic matter and/or variations in detrital terrigenous input. For the Lyckhem borehole (Fig. 6), the estimated sedimentation rate (including stinkstones) is 5.3 mm/103 years for the middle Cambrian, where this lithology is most common, while the rate is only 3.4 mm/103 years for the upper Cambrian (excluding stinkstones). This suggests a less stable environment during the middle Cambrian. The poorly laminated grey-green mudstone (lithofacies probably indicates relatively normal oxic conditions. Poor lamination and uneven contacts are the result of burrowing, and the shelly fauna indicate the presence of macroscopic invertebrates requiring >1 ml O2/1 (Rhoads & Morse 1971). This lithology underlies the Alum Shale sensu stricto in the Lyckhem borehole, but interdigitates in the lower part of the 16/74 core, testifying to short periods of oxic conditions during transition to the typical Alum Shale environment. The isolated grey laminated mudstone occurrences within the Dictyonema shale indicate a variable organic carbon supply, without the establishment of fully oxic conditions. The gradational boundaries suggest that the process was not a sudden event and that variation occurred on the scale of hundreds of years, according to the calculated sedimentation rates (Table 1). The characteristic fine-scale alternations of the black and grey interlaminated facies association suggest variability in the input of organic matter and/or terrigenous material. Burrowing, evident at some levels in the grey mudstone shows that the bottom waters were not always toxic for a small soft-bodied fauna. This is supported by the local occurrence of calcareous macrofossil fragments within the mudstone. The interlaminated association suggests a dysaerobic environment (Rhoads & Morse 1971; Cluffet al. 1981) with alternating anoxic and more oxic conditions. Oxygen content was probably controlled by variations in the supply of organic matter. The origin of the bituminous limestone concretions has been explained as the remains of a once continuous limestone bed that was subsequently dissolved (Bjorlykke 1973), and as early stage concretions (Henningsmoen 1974). The evidence presented here suggests that the concretions are
similar to the early formed diagenetic concretions of Raiswell (1971), and not the remains of a previous limestone bed. The concretions appear to originate by nucleation on fossiliferous layers, possibly on the sea floor, followed by continued growth during early stages of compaction. This interpretation poses two major problems: (1) the concretions are highly fossiliferous and yet adjacent mudstones are apparently devoid of any fauna or traces of it, and (2) up to three trilobite zones occur in succession within the Great Stinkstone bed, indicating development over very long time periods. Hallam's model (1981) for the formation of Liassic concretions may explain the first problem, whereby slight undulations of the sea floor (on the horizontal scale of 10's of kilometres and the vertical scale of 10's of metres) occur on a broad epicontinental shelf. Intra-basinal highs are proposed as the sites of concretion formation, such that they penetrated through the probable permanent anoxic-oxic boundary within the water column allowing colonization by trilobite faunas, possibly adapted to low dissolved oxygen contents (Henningsmoen 1957). Winnowing and reworking could have concentrated the organic/ skeletal remains, although for the most part current structures are lacking. The extremely slow sedimentation rates would favour local lithification by calcite precipitation around concentrations of trilobite debris. Slight fluctuations of the oxic-anoxic boundary and/or subsidence of the highs would have destroyed the living trilobite fauna and brought the carbonate into more acidic conditions, allowing dissolution of much of the trilobite shell material except where concentrated or lithified. Growth and reprecipitation could have occurred around these nucleation sites giving the present fossil-rich concretions and the adjacent shale which is devoid of fossils. The second problem may be explained by considering the above model operating over long time periods where the environment was almost starved of sediment influx. Advanced lithification and coalescence of the incipient concretions may have occurred and later the whole horizon may have been modified by growth during compaction. That faunas are apparently in succession within the Great Stinkstone suggests progressive lithification upwards and low current and wave activity. However, the conglomeratic horizons in the Exporrecta stinkstone, for example, may suggest reworking, perhaps due to uplift to wavebase or to occasional storm action. They also testify to very early lithification. The most important consequence of this model is to show that conditions were oxic or semi-oxic for long periods of time over varying extents of
The sedimentology o f the Swedish Alum Shales the Alum Shale basin during the Upper Cambrian. Sedimentary evidence indicates that the coarser sandstone lithofacies and probably the siltier beds too, are transported material, possibly slumped in some cases. Odin & Matter (1981) suggest that glauconite generally occurs towards continental margins or on submarine highs where slow sedimentation rates or non-deposition and an ambient non-oxidizing micro-environment occur. Slight topographic highs with little terrigenous input probably developed in the Alum Shale depositional areas and could have favoured the accumulation of glauconite. Berner (1981) suggests glauconite is characteristic of the postoxic, non-sulphidic environment where low contents of organic matter were present, inhibiting sulphate reduction and pyrite production. The non-glauconitic siltstones of this lithofacies imply increased terrigenous input, although the processes by which this occurred are difficult to define. Wind transport and current activity (indicated by cross-lamination) are possible processes.
Alum shale depositional environment Discontinuous outcrops of Alum Shale at the present day suggest an original Alum shale depositional area of up to 2000 km north-south by 800 km east-west. Shortening by thrusting is estimated to be up to 500 km (Gee 1975), and several of the nappes contain black phyllites which are probable Alum Shale equivalents (Gee 1975). Therefore the basin may originally have extended west of the Norwegian coast. In the highest nappes, no black phyllites are recorded and metamorphosed sediments may, in part, be deeper water equivalents of the Alum Shale, tectonically transported eastwards (Gee 1975). Over the entire basinal area, the Alum shale is apparently overlain and underlain by shallow marine deposits with little variation in lithology and this evidence is used to suggest that the Alum Shale is also a relatively shallow water deposit. The size of the outcrop area suggests deposition in an epicontinental sea environment, which lacks modern analogues. Hallam (1981) has summarized likely oceanographic conditions in such a sea. Where water depth is shallow ( < 200 m) there is a tendency for stagnation away from the open ocean. This appears to be the environment of the Alum Shale, particularly as lithologies are remarkably constant over large areas and for long periods of time. General stratigraphic considerations suggest that contemporaneous sea-levels were high (Leggett et al. 1981). Consistent slow sedimentation in
523
a shallow water environment lends support to this idea, since it implies restricted detrital supply, one cause of which might be flooding of source areas. Redeposited sediment is rare, suggesting gentle sea-floor topography and the dominant process appears to have been sedimentation from suspension. Characteristically, the deposits of an epicontinental shelf sea show great lateral persistence (Hallam 1981). However, the lithofacies recognized here show variations in space and time. The vertical variation of lithofacies previously described, reflects a progressive cutting-off of sediment supply and expansion and stabilization of anoxic conditions in the Alum Shale sea. However, the occurrence of bituminous limestone concretions testifies to the likely presence of long ranging intra-basinal highs, penetrating the oxic layers of the upper water column. The sedimentation rates of the fossiliferous precursor sediments of the concretions is the slowest of all the lithofacies, and probably involves considerable pauses in sedimentation. Vertical facies variations not only indicate more variable sediment supply within lower parts of the Alum Shale, but also more rapid variations in the oxygen content of deeper waters. Lateral variation of lithofacies indicates a slow expansion of anoxic conditions through the middle and upper Cambrian reaching a peak roughly in the P. scarabaeoides zone. The overlying Dictyonema shales show a waning of these conditions since deposition appears to have become more localized and the sediments may contain silty laminae, indicating increased detrital supply. Since much of the original Alum Shale has probably been eroded it is not clear whether intra-shelf basins existed. However, if present, they would be candidates for more localized anoxic conditions. Similar, but thicker and less extreme facies trends also occur in Britain (Leggett 1980 and references therein), SE Newfoundland, Cape Bretton Island and New Brunswick (North 1971 and references therein). Anoxic conditions with the development of black mudstones often with concretionary limestones occur at certain preferred periods within the middle and upper Cambrian. Black or dark grey mudstones are generally interspersed with periods dominated by bioturbated mudstones and siltstones, with local development of sandstones. These facies variations demonstrate that the area of the probable SE margin of the Iapetus Ocean intermittently affected by anoxic conditions was of greater extent than the present Alum Shale outcrop. ACKNOWLEDGMENTS: I would like to acknow-
A. Thickpenny
524
ledge the help of the Swedish Geological Survey, in particular D r David Gee, and Aktiebolaget Svensk Alumskifferutreckling, in allowing examination and discussion of borehole core material. I would also like to a c k n o w l e d g e receipt of research assistantship support from Chalmers
University of Technology, G o t h e n b u r g during 1981-82. Finally, I w o u l d like to t h a n k Drs Jeremy Leggett, Kevin Pickering and D o r r i k Stow for critically reading the manuscript and A m a n d a H a r t l a n d for typing it.
References ANDERSSON,A. 1974. Unders6kningar av den 6verkambriska alunskiffer i fern b6rrkarnor fr~n Sydbillingens h6gplatfi. Ranstad Skifferaktiebolaget, Report TPM-IR-826, 1-12. 1977. Analys av alunskifferformationen i DBH 65/77 fr~n Ranstadsverkets dagbrott. Ranstad Skifferaktiebolaget, Report TPM-U-976, 1-9. 1979. Analyser av hela alunskifferformationen i DBH 16/74 fr~n Sydbillingen. Ranstad Skifferaktiebolaget, Report TPM-I106, I-3. 1980a. Borrningar 1975 av 15 k~irnor genom 6verkambrisk alunskiffer i Sydbillingen-Ranstadomr~tdet. Koordinatf6rtforteckning mfiktigheter och analyser. Ranstad Skifferaktiebolaget, Report TPM-1283, 1-5. 1980b. N/irkes tillg~nger av alunskiffer. Aktiebolaget Svenstk Alunskifferutveckling, Report ASA T9/80, 1-6. 1980c. Billingen-Falbygden-omr~tdets tillg~ngar av alunskiffer. Aktiebolaget svensk Alunskifferutveckling, Report ASA T58/80, 1-10. , DAHLMAN,B. & GEE, D.G., 1983. Kerogen and Uranium resources of the Cambrian Alum shales of the Billingen-Falbygden and N/irke areas, Sweden. Geol. F6r. Stockh. F6rh., 104, 197-209. ARMANDS, G. 1972. Geochemical study of Uranium, Molybdenum and Vanadium in a Swedish Alum shale. Stockh. Contr. Geol., 27, 1-148. ASKLUND,B. 8~;THORSLUND,P. 1935. Fj/illkedjerandens Bergbyggnad i Norra J/imtland och Angermanland. Sver. geol. Unders., ser. C 382, 1-110. BERGER,W.H. 1974. Deep sea sedimentation. In: Burk, C.A. & Drake, C.L. (eds), The Geology of Continental Margins. Springer-Verlag, New York. 213-41. BERNER,R.A. 1970. Sedimentary pyrite formation. Am. J. Sei., 1-23. 1981. A new geochemical classification of sedimentary environments. J. sed. Petrol., 51,359-65. BJORLYKKE,K. 1973. Origin of limestone nodules in the Lower Palaeozoic of the Oslo region. Norsk geol. Tiddskr., 53, 419-31. CHURKIN, M., CARTER, C. & JOHNSON, B.R. 1977. Subdivision of Ordovician and Silurian time scale using accumulation rates of graptolitic shale. Geology, 5, 452-6. CLUFF, R.M., REINBOLD,M.L. & LINEBACK,J.A. 1981. The New Albany Shale Group of Illinois. Illinois State Geol. Surv., cir. 518, 1-83. COWIE, J.W. & CgmB, S.J. 1978. The Cambrian system. In: Cohee, G.V., Glaessner, M.F. & Hedberg, H.D. (eds), Contributions to the Geological Time Scale. Am. Ass. Petrol. Geol. Studies in Geology 6, 355-62.
-
-
-
-
-
-
2 6 8 ,
-
-
DAHLMAN,B. 1962. Unders6kningar av Overkambrium reed hfinsyn till uranfordelning och sediment i mellersta och s6dra Sverige utforda vid Sveriges geologiska unders6kning. Sver. geol. Unders., internal report. Unpubl. DAHLMAN, B. & EKLUND, J. 1953. Sveriges uranf6rande alunskiffrar. Sver. geol. Unders., Unpubl. report. & GEE, D.G. 1977. Oversikt 6ver BillingenFalbygdens geologi. In: Betiinkande av Billingeutredningen, Billingen 4, exampel. Statens offentliga utredningar 1977: 47, Industri departementet, Stockholm. EDL1NG, B. 1974. Distribution of Uranium in Upper Cambrian Alum Shale from Ranstad, Billingen, V~istergotland. Publications from the Palaeontological Institute of the University of Uppsala, special volume 2, 1-118. GALE, N.H., BECKINSALE,R.D. & WADGE, A.W. 1979. A Rb-Sr whole rock isochron for the Stockdale Rhyolite of the English Lake District and a revised mid-Palaeozoic time-scale. J. geol. Soc. Lond., 136, 235-42. GEE, D.G. 1975. A tectonic model for the central part of the Scandinavian Caledonides. Am. J. Sci., 275, 465-515. 1980. Basement-cover relationships in the central Scandinavian Caledonides. F6r. Stockh. Fgrh., 102, 455-74. -1981. The Dictyonema-bearing phyllites at Nordaunevoll, eastern Trondelag, Norway. Norsk geol. Tiddskr, 61, 93-5. --, KARIS, L., KUMPULAINEN,R. 8r THELANDER,Y. 1974. A summary of Caledonian front stratigraphy, northern Jfimtland/southern VSsterbotten, central Swedish Caledonides. Geol. F6r. Stockh. F6rh., 96, 389-97. --, KUMPULAINEN,R. 8r THELANDER,T. 1978. The Thsj6 d6collement, Swedish Caledonides. Sver. geol. Unders., set. C742, 1-35. ZACHRISSON, E. 1979. The Caledonides in Sweden. Sver. geol. Unders., set. C 769, 5-48. HALLAM, A. 1981. Facies Interpretation and the Stratigraphic Record. Freeman, Oxford. HARLAND, W.B., SMITH, A.G. & WXLCOCK,B. (eds.). 1964. The Phanerozoic Time Scale. Q. J. geol. Soc. Lond., 120S. HENN1NGSMOEN,G. 1957. The trilobite family Olenidae, with description of Norwegian material and remarks on the Olenid and Tremadocian Series. Skr Norske Vidensk-Akad, Oslo. Mat. naturv. Kl., l, 1-303. 1974. A comment. Origin of limestone nodules in
-
-
-
-
The sedimentology of the Swedish Alum Shales the Lower Palaeozoic of the Oslo region. Norsk. geol. Tiddskr., 54, 401-12. KULLING, O. 1964. Oversikt 6ver norra Norrbottensfjfillens kaledonidberggrund. Sver. geol. Unders., ser. Ba 19, 1-166. LAMBERT, R. ST.J. 1971. The pre-Pleistocene Phanerozoic time-scale--a review. In: Harland, W.B. & Francis, E.H. (eds). The Phanerozoie Time-scale: a supplement. Spec. Publ. geol. Soc. Lond., 5, 9-31. LEGGETT, J.K. 1980. British Lower Palaeozoic black shales and their palaeo-oceanographic significance. J. geol. Sor Lond., 137, 139-56. LIDI~N, R. 1910. Kalkstensf6rekomster utefter inlandsbanan mellan Str6ms Vattudal och Pite /ilf. Sver. geol. Unders., ser. C 235, 1-45. MCKERROW, W.S., LAMBERT, R. ST. J. & CHAMBERLAIN, V.E. 1980. The Ordovician, Silurian and Devonian time scales. Earth planet. Sci. Lett., 51, 1-8. NORTH, F.K. 1971. The Cambrian of Canada and Alaska. In: Holland, C.H. (ed.), Cambrian of the New World. John Wiley, London. ODIN, G.S. & MATTER, A. 1981. De glauconarium origine. Sedimentology, 28, 611-43. RAISWELL, R. 1971. The growth of Cambrian and Liassic concretions. Sedimentology, 17, 147-71. READING, H.G. 1965. Eocambrian and Lower Palaeozoic geology of the Digermul peninsula, Tanafjord, Finnmark. Norg. geol. Unders., 234, 167-91. RHOADS, D.C. & MORSE, J.W. 1971. Evolutionary and ecologic significance of oxygen-deficient marine basins. Lethaia, 4, 413-28. Ross, R.J., NAESER,C.W., IZETT, G.A., WHITTINGTON, H.B., HUGHES,C.P., RICKARDS,R.B., ZALAS1EWICZ, J., SHELDON, P.R., JENKYNS,C.J., COCKS, L.R.M.,
525
BASSETT, M.G., TOGHILL, P., DEAN, W.T. & INGHAM, J.K. 1978. Fission-track dating of Lower Palaeozoic bentonites in British stratotypes. In: Zartman, R.E. (ed.), Short papers of the 4th International conference, Geochronology, Cosmochronology and Isotope Geology. US geol. Surv. Open-file Rep. 78-701,363-5. TEGENGREN, F.R. 1962. Vassbo blymalmfyndighet i Idre och dess geologiska inramning. Sver. geol. Unders., ser. C 586, 1-61. THORSLUND, P. 1940. On the Chasmops series of J~imtland and S6dermanland (Tv~iren). Sver. geol. Unders., ser. C 436, 1-191. 1960. The Cambro-Silurian. In: Magnusson, N.H., Thorslund, P., Brotzen, F., Asklund, B. & Kulling, O. Description to accompany the map of the pre-Quaternary rocks of Sweden. Sver. geol. Unders., ser. Ba 16, 69-110. WESTERG.~RD, A.H. 1922. Sveriges Olenidskiffer. Sver. geol. Unders., ser. C 18, 1-189. 1940. Nya djupborrningar genom/ildsta Ordovicium och kambrium i Osterg6tland och N/irke. Sver. geol. Unders., ser. C 437, 1-72. 1944a. Borrningar genom Sk'~nes Alunskiffer 1941-42. Sver. geol. Unders., ser. C 459, 3-38. 1944b. Borrningar genom Alunskifferlagret p'~ ()land och i Osterg6tland. Sver. geol. Unders., ser. C 463, 1-45. -1947. Supplementary notes on the Upper Cambrian trilobites of Sweden. Sver. geol. Unders., ser. C 489, 1-34. 1948. Non-agnostidean trilobites of the Middle Cambrian of Sweden. I. St, er. geol. Unders., ser. C 498, 1-32. -
-
-
-
A. THICKPENNY,Department of Geology, Royal School of Mines, Imperial College, London SW7 2BP.
Models for the deposition of Mesozoic-Cenozoic fine-grained organic-carbon-rich sediment in the deep sea M.A. Arthur, W.E. Dean and D.A.V. Stow SUMMARY: The widespread occurrence of organic-carbon-rich strata ('black shales') in certain portions of Jurassic, Cretaceous and Cenozoic sequences has been well-documented from Deep Sea Drilling Project sites in the Atlantic and Pacific Oceans and from sequences, now exposed on land, originally deposited in the Tethyan ocean. These ancient black shales usually have been explained by analogy with examples of modern deep-sea sediments in which organic matter locally is preserved by (1) increasing the supply of organic matter, (2) increasing the rate of sedimentation, and/or (3) decreasing the oxygen content of the bottom water. However, detailed examination of many black shales reveals characteristics that cannot be explained by simple local models, including: their approximate coincidence in time globally; their occurrence in a variety of different environments, including open oxygenated oceans, restricted basins, deep and shallow water; their interbedding with organic-carbonpoor strata which often dominate a so-called black shale sequence; their deposition by pelagic, hemipelagic, turbiditic and other processes; and the variations in type and amount of organic matter that occur even within the same sequence. A more complex model for the origin of black shales therefore appears most appropriate, in which the cyclic preservation of organic matter depends on the interplay of the three main variables, namely supply of organic matter, sedimentation rate, and deep-water oxygenation, each of which varies independently to some extent. The variation and relative importance of these parameters in individual basins and widespread black shale deposition in general are linked globally and temporally by changes in global sea-level, climate and related changes in oceanic circulation. An important and often overlooked factor for the supply of organic matter to deep-basin sediments is the frequency and magnitude of redepositional processes. The interplay of these variables is discussed in relation to the middle Cretaceous and Cenozoic organic-carbon-rich strata, in particular, which show marked differences in the relative importance of the different variables. Results of analyses of Deep Sea Drilling Project (DSDP) cores show that strata containing relatively high concentrations of organic matter (more than 2% organic carbon) are common at different stratigraphic horizons and in many parts of the world ocean. These strata are often loosely described as 'black shales' although they usually consist of interbedded rocks with varying contents of carbonate and/or biogenic silica and differ mainly in colour and/or concentration of organic matter. The most common lithologies are interbedded black or dark-olive shale or claystone, and lighter greenish-grey shale or claystone (i.e. interbedded 'black' and 'green' argillaceous rocks). The oldest occurrences of organic-carbon-rich strata in the present ocean basins are upper Jurassic (Callovian-Oxfordian) black claystones recovered at two DSDP Sites in the South Atlantic and one in the North Atlantic, and Tithonian-Neocomian deep-water marlstones and limestones that have been recovered from several DSDP sites in the North Atlantic. Blackshale deposition was widespread in the North and South Atlantic during the Barremian-Albian (100-115 my BP) and Cenomanian-Turonian (86-100 my ~p) at both continental-margin and
deep-basin sites (Arthur & Natland 1979; Jenkyns 1980; Weissert 1981) and locally during the Coniacian-Santonian (75-86 my BP). Cretaceous organic-carbon-rich strata in the Pacific are much more restricted in both time and space than in the Atlantic. They have been recovered from seven DSDP sites on the flanks of elevated volcanic plateaus and seamounts, mainly within very restricted stratigraphic intervals (Schlanger & Jenkyns 1976; Dean et al. 1981; Thiede et al. 1982). Organic-carbon-rich strata in the Pacific all occur within the same general middle Cretaceous time interval, but they are not strictly synchronous and occur in strata that range in age from 86-120 my SP. Middle Cretaceous strata recovered from deep basins of the Pacific generally are not rich in organic carbon, but a thin layer containing 5?/0organic carbon was recently recovered at the Cenomanian-Turonian boundary within a turbidite sequence in the eastern Mariana Basin (DSDP Site 585; Moberly et al. 1983). Organic-carbon-rich sediments of Eocene and Miocene age have been recovered at a few sites in the Atlantic (Arthur & von Rad 1979; McCave 1979a; Dean & Gardner 1982), but these strata are much more restricted in both time and space than those of Cretaceous age. In the Pacific, 527
M.A. Arthur, W.E. Dean and D.A.V. Stow
528
however, extensive organic-carbon-rich Miocene strata occur in both offshore and onshore localities. In this paper we first provide an outline of the main modes of preserving organic matter in modern deep-sea sediments, then describe the lithology, distribution and organic-carbon content of ancient Mesozoic and Cenozoic black shales, and finally discuss the elements of models for black shale deposition. Our emphasis throughout is on deep-water and fine-grained organic-carbon-rich sediments.
matter occurs during the dominantly oxic conditions of diagenesis in most pelagic deep-sea sediments (e.g. Heath et al. 1977; Demaison & Moore 1980). There are three main ways of preserving relatively high concentrations of organic matter in deep-sea sediments: the first is by increasing the supply of organic matter, the second is by increasing the rate of sedimentation, and the third is by decreasing the oxygen content of the water mass overlying the sediment. These processes usually but not always act in consort in modern marine environments.
Preservation of organic matter in modern deep-sea sediments
Organic matter supply
Most modern deep-sea environments are characterized by sediments that are relatively depleted in organic carbon (<0.5). At a water depth of 4000 m, less than one percent of the organic matter produced in surface waters survives to reach the sediment/water interface (Fig. 1; Suess 1980); a relatively small proportion of this organic matter is reactive in the sense of Toth & Lerman (1977). Therefore, not only are concentrations of organic matter low in the majority of deep-water sediments, but this organic matter is low in nitrogen and phosphorus, and is relatively oxidized; reactive organic matter is rapidly consumed by the benthic biomass (Mfiller & Suess 1979). Continued degradation of the organic CARBON FLUX PRODUCTION 0.001
O.O1 9
E
0.1 9
1.0 10
}:
.
e
v "1I-DILl C~
o
./
n" IJJ I--
9 100
1000
0.1
1
10
10,OO0 O0
% of Primary Production reaching Sed./H20 interface
FIG. 1. Proportion of organic carbon reaching the sediment/water interface at different depths as percent of primary productivity in surface waters. Data are results of sediment trap experiments compiled by Suess (1980). The data indicate an exponential decrease of organic matter accumulation with depth. but represent the situation for an oxygenated water column only.
Marine organic matter is supplied to the sediments from primary biological productivity in the surface waters; terrigenous organic matter is derived from land plants and transported to the sea mainly by rivers, but some organic materials are carried from land by strong prevailing winds (charcoal, spores, pollen, etc.). Areas of increased primary productivity occur where deep (below the thermocline, i.e. from below about 150-250 metres depth) nutrient-rich waters come to the surface along an equatorial belt, along northern and southern polar divergence zones and along the western margins of the major continents (e.g. Koblentz-Mishke et al. 1970; Demaison & Moore 1980). Increased supplies of terrigenous organic matter are constrained by fluvial discharge and by the climate and vegetation of the drainage area; equatorial regions therefore receive more organic matter than polar or temperate seas. Arid coasts within trade wind regions may supply little terrestrially derived material. Both marine and terrigenous organic matter can be supplied directly to deep-water sediments as a part of the pelagic 'rain' (or marine 'snow'; Shanks & Trent 1980) and by faecal pellet transport (e.g. Suess 1980; Dunbar & Berger 1981). Sinking rates of faecal pellets of up to 170 m per day and marine 'snow' of up to 95 m per day suggest that transfer of organic matter through the water column by these two processes is relatively rapid. However, sediment trap studies do suggest that much oxidation does occur in transit through oxic water masses by bacterial action or consumption by other organisms (e.g. Suess 1980; Fig. 1). In addition, and perhaps more commonly, the widespread downslope redeposition of organic-carbon-rich shelf and slope sediments is an important means of supply to deeper basinal sediments. This mode of transport includes supply from bottom nepheloid-layers as suggested by Summerhayes (1981b).
Deposition of organic-carbon-rich sediment
redeposited to the base of slope or basin floor also is rich in organic carbon. Such sediment, although redeposited, would probably retain many of the original compositional characteristics of the shallower sediment, including a high concentration of well-preserved organic matter.
There have been few systematic studies of the organic contents of redeposited sediments along modern continental margins, but it is known that widespread redeposition of slope sediment occurs, for example, in the area of high organic productivity off north-west Africa (Embley 1976; yon Rad et al. 1979). Slumps and slope-derived turbidites also are common along the California borderland (Nardin et al. 1979; Soutar et al. 1981), along the west coast of Mexico in the Gulf of California (Einsele & Kelts 1982), along the Peru-Chile Slope (Soutar et al. 1981), and in the Angola Basin off south-west Africa (Summerhayes et aL 1979; Embley & Morley 1980; Hay et aL 1982; Stow, in press). All of these regions have organic-carbon-rich slope sediments preserved as the result of upwelling and high biological productivity, and well-developed oxygen-minimum zones. More than likely, some of the sediment 100
i
. . . . . .
i
.
.
.
.
.
.
.
.
.
.
529
Sedimentation rate
Relatively high sedimentation rates apparently aid in the preservation of organic carbon in marine sediments because, up to a point, organic carbon contents and accumulation rates in modern sediments are positively correlated with bulk sediment accumulation rates (Heath et al. 1977; Miiller & Suess 1979; Fig. 2). Ibach (1982) suggested that this positive relationship between organic-carbon content and sedimentation rate also holds for Cenozoic and Mesozoic strata
.
.
.
.
.
.
.
.
C
,
.
i
ii,lll
<>
IdJ 10
=.
cr
, Z
1 >'
o ~, ~
01
O
rr 9
, o.oi
W.BALTIC
o PERU 9 OREGON
oool Ol
1
lO
10o
SEDIMENTATION
lO0O
RATE
o ARGENTINE BASIN
[cm(loooy)-I ] ,
i JJtllllt
L illqlll
ill~,.u
9 N.W. A F R I C A
o CENTRAL PACIFIC i ~Jl.Jl
A BLACK SEA
PRIM. PROD. RATE (gC m-2y -1)
300 t B 150 O-.J
, r2.. """
~ ") *
I
~mo
ORGANIC CARBON % 01 ! ................................. 01 1 10 100 1000 SEDIMENTATION RATE
[cm0oooy)-l]
FIG. 2. Relations among bulk sedimentation rate and (A) percent organic carbon, (B) primary productivity, and (C) organic carbon accumulation rates for modern marine sediments (Miiller & Suess 1979). The scatter about the best-fit regression line represents the effects of dysaerobic or anoxic conditions which further enhance preservation (e.g. Black Sea points), as well as no correction for differences in depth of deposition among other problems.
53o
M.A. Arthur, W.E. Dean and D.A.V. Stow
recovered at DSDP sites. The main reason for this relationship is probably that high sedimentation rates effectively bury organic matter more rapidly, removing it from zones of bioturbation, oxic decomposition, and sulphate reduction. Although organic-carbon contents and accumulation rates may be higher in high-sedimentationrate sequences, the organic matter itself may be somewhat more poorly preserved losing a greater proportion of H, P, N than that deposited under anoxic conditions. Hydrocarbon source potential is determined not only by the organic carbon content but also the characteristics of the organic matter.
Bottom-water oxygenation Oxygen concentrations of less than 0.5 ml/l and certainly 0.2 ml/l inhibit the activity of benthic metazoans (e.g. Rhodes & Morse 1971). Lack of bioturbation in turn limits the residence time of organic matter at the sediment/water interface and, therefore, the state of its oxic decomposition. Low-oxygen concentrations occur in both midwater oxygen-minimum zones and in silled stratified basins, and, where lowest (i.e. <0.5 ml/l), commonly lead to enrichment of organic carbon in underlying sediments. Most areas of modern oceans that are covered with organic-carbon-rich sediment are within well-developed oxygen-minimum zones (e.g. Demaison & Moore 1980). These zones originate partly because of the slow rate of sinking and advection of warm surface water in low latitudes (Wyrtki 1962). These warm-water masses are more poorly oxygenated than cold, deep waters that originate by sinking and advection at high latitudes because of the lower solubility of oxygen with increasing temperature (Fig. 3). Also, much of the decomposition of organic matter and concomitant consumption of dissolved oxygen takes place in the upper few hundred metres of the water column within and below the base of the photic zone (Suess 1980). In oxic environments the rate of organic matter decomposition decreases exponentially with depth (Fig. 1). Therefore, in regions of high biologic productivity, minimum concentrations of dissolved oxygen occur from about 100 m below the surface to about 1500 or 2000 m, but the extent of oxygen depletion varies (Fig. 4) depending on nutrient availability and surface water productivity, initial dissolved oxygen content of intermediate water masses, and rates of horizontal advection of oxygen. Bottom sediments within oxygen-minimum zones may preserve high concentrations of organic matter. For example, concentrations of organic carbon of up to 25% occur in sediments
OXYGEN
SATURATION
VALUES 3C
II
I
! I I 17 "l'-,,5.O,.i. ,,J.._[ I I
" -"
I "---"
tO O
~
""--.-4..-.. ~_
_z
~
""~" "~ " '~] ~ ~-- ~
2C
~
~
"-'-'--
"'~'-"~
15
,,~..~ - - . ~ - - '
6.0-
...._.~
a..
i
6.2-
-..__.]
6 . 8 - ~ 7.0,
32
--" 5.6~,
~
--5.8.:
~
-----,C
'"-"~ ---"~
~
~
33
"
---"
" ~ -
7 . 6 - ~ 7.8 . - -
0
-5.0
~
.___.
8.0 9- 8 . 2 8.4 9
~
" ' " ~ " - - "-----,,~ -; . . . . . ~
--'4--
I--
- 4.8 ~
-"
6.4.,, 6.6-
-4.6-
~
<
ILICr"
loi
I|11
""'":
I. . . . . . . .
(.9
ILl
i III
"~-..,~-4 ~
. . . .
25
I"--
<
mL/L
~
,
I
34
~
~
,.,___._
~ ,.
~ i
~ . ~
~
35
SALINITY
,_..._,~ ~
-~__,,"'-"-~
36
37
38
o/oo
FIG. 3. Theoretical oxygen solubility in surface seawater as a function of temperature, salinity, and modern atmospheric pO2 (after Wilde & Berry 1982). where the oxygen-minimum zone impinges on the inner continental shelf off south-west Africa (Calvert & Price 1971). However, organic-carbon enrichment in these sediments probably is due mainly to the high productivity which is centred over a very shallow shelf (< 150 m) (see later discussion). The Black Sea is an example of a silled anoxic basin (Fig. 4) with as much as 4 to 5~ organic carbon in bottom sediments (Shimkus & Trimonis 1974). The bottom water of the Black Sea is completely depleted in dissolved oxygen from 175 m to the bottom (about 2500 m). Anoxia is maintained by strong salinity stratification and relatively slow replenishment of bottom water, that is, slow replenishment in comparison to rates of oxygen consumption due to organic matter oxidation and sulphide oxidation near the point of bottom water sinking. However, the Black Sea is by no means the sluggish, stagnant basin implied by many authors; the entire deep-water mass is probably replenished every 700 yrs or less (Grasshoff 1975). Sulphate reduction occurs in the water column and results in dissolved hydrogen sulphide in the water column and a relatively high sulphur to carbon ratio in bottom sediments (Leventhal 1983). The Black Sea is a relatively
Deposition
of organic-carbon-rich
sediment
531
F1G. 4. Examples of modern marine settings in which dissolved oxygen is relatively depleted within intermediate or deep water masses. Shown are contours of dissolved oxygen with depth. (a) East-West section of Peru margin at 15~ (data from Burnett et al. 1981) showing an open-ocean oxygen-minimum zone; (b) East-West section across Guaymas Basin, Gulf of California, a narrow, semi-enclosed ocean basin with a well-developed oxygen-minimum zone; oxygen-depleted midwater masses have been advected into the basin; (c) Santa Barbara Basin off southern California (data from Emery 1960; Rittenberg et al. 1953); the basin is silled within the midwater oxygen-minimum zone to the west and therefore the deep waters of the basin are relatively depleted in oxygen initially; (d) Black Sea; a salinity stratified sea (data from Grasshoff 1975) that is anoxic below 175 m with sulphate reduction occurring in the water column; dotted contours below zero 02 level are H2S concentrations.
small silled basin with only moderate productivity in surface waters ( < 100 gC/m2/y) and unusual circulation conditions, but anoxia has persisted for the last 7000 yrs. About 5% of the organic carbon produced in surface waters is preserved in abyssal Black Sea sediments (Deuser 1974) as opposed to much less than 1~ in most oxygenated settings. The enhanced preservation is most certainly due to anoxia and the lack of benthic metazoan activity. It is not entirely clear whether such conditions could exist continuously in a basin as large as the Cretaceous Atlantic Ocean for periods of up to 10 or 15 my. The Mediterranean Sea is areally larger than the Black Sea, but apparently also has experienced periodic anoxic conditions since at least the Miocene (e.g. Ryan & Cita 1977; Kidd et al. 1978) possibly due to conditions similar to those in the Black Sea today. The Mediterranean is essentially a silled basin with respect to the Atlantic Ocean. The thin but laterally extensive sapropels ( > 2% organic carbon) of the Mediterranean contain mainly marine organic carbon with maximum concentrations of about 17% (Ryan 1972; Nester-
off 1973; Kidd et al. 1978). Many of the later Pleistocene sapropels probably are the result of increased runoff of glacial meltwater to the Mediterranean coupled with increased surface productivity and reduced rates of sinking of bottom water (Thunell et al. 1977; Williams et al. 1978; Jenkins & Williams, in press) or possibly due to the influence of increased fluvial runoff as the result of periodically increased monsoonal rainfall in north Africa (Rossignol-Strick et al. 1982). The Santa Barbara Basin is a restricted marginal basin offshore California as described by Rittenberg et al. (1953). Much of the basinal sediments in the Santa Barbara Basin are laminated and relatively rich in organic matter (Soutar et al. 1981) with contents of 3 - 5 ~ organic carbon in the deepest parts of the basin and < 1 on the topographic highs (Emery 1960). The only bottom water for the basin is derived from within a well-developed oxygen-minimum zone that impinges at sill depth (Fig. 4), which, coupled with high coastal productivity, explains why oxygen depletion extends to the basin bottom.
53Z
M.A.
Arthur,
W.E. Dean and D.A.V.
However, the basin circulation is not 'stagnant' and deep waters may be flushed out as frequently as every 20 yrs or so (e.g. Hfilsemann & Emery 1961; Soutar et al. 1981). Other examples of oxygen depleted basins are the thermally-stratified Cariaco Trench in the Caribbean Sea, and shallow-silled, salinity-stratified l]ords, such as the Gotland Basin of the Baltic Sea and Saanich Inlet, British Columbia. However, not all silled or restricted basins are characterized by oxygendepleted deep-water masses (e.g. the Red Sea; Grasshoff 1975). Sedimentary indicators of productivity and levels of bottom water dissolved oxygen
The three main factors that influence the accumulation of organic carbon: supply of organic matter; sedimentation rate; and concentration of dissolved oxygen in the water-column are commonly linked in some way in modern marine environments. It is frequently difficult to identify any one factor as the major cause of enhanced organic carbon preservation in a given ancient sedimentary environment, largely because the necessary unambiguous criteria do not exist or have not yet been formulated and tested. Primary sedimentary structures, mineralogy, biotic composition, and sediment geochemistry may, however, provide some constraints on interpretation. A basic matrix of dissolved-oxygen concentrations in bottom waters versus benthic biotic composition, primary sedimentary structures, and selected geochemical indicators in sediments is shown in Fig. 5. Such criteria are, of course, to be used with caution. In general, bioturbated sediments occur where bottom water dissolved oxygen levels are above 0.5 ml/l. Finely laminated or nonbioturbated sediments may indicate anaerobic ( < 0.2 ml/1) to anoxic conditions. Many organic-carbon-rich sediments deposited under anoxic conditions are therefore finely laminated. Lamination represents preservation of the record of seasonal variation in supply of two or more components or of currentcontrolled deposition. For example, annual varves preserved on the slope of the Gulf of California under an intense oxygen-minimum zone represent alternation of terrigenous clay minerals supplied by run-off during the rainy season and high productivity of diatoms and organic matter during seasonal upwelling (Calvert 1966b). The small diameter of burrows and predominance of shallow horizontal versus vertical feeding traces may indicate low-oxygen conditions at the sea floor which approach but do not go below 0.2 ml/l (e.g. Savrda et al. 1982). Even at 0.2 ml/l it is possible to find non-metozoan
Stow
organisms such as certain benthic foraminifers or serpulids (Calvert 1966b; Phleger & Soutar 1973) living on laminated sediments, and in certain settings benthic reworking of surface mud by pelagic crabs can be quite pronounced at dissolved oxygen levels of just above 0.25 ml/l (Soutar et al. 1981). The control of oxygen levels on activities of larger benthic metazoans is the main way in which the concentration of dissolved oxygen affects organic carbon preservation in sediments, and probably is the main reason for better preservation of organic matter in dysaerobic to anoxic depositional environments, regardless of the rates of primary productivity. Bioturbation increases the residence time of sedimentary particles, including organic matter, and thereby increases the chance for ingestion and oxidation of organic carbon. Higher sedimentation rates (Miiller & Suess 1979) act in the same manner to decrease the residence time of organic carbon in zones of oxic consumption and bacterial sulphate reduction. The presence of pyrite or other sulphide minerals in sediments does not necessarily imply anoxic conditions at the sea floor, but only that sufficient metabolizable organic matter remained during burial to promote anoxia, sulphate reduction within the sediment and sulphide mineralization (Goldhaber & Kaplan 1974; Berner 1977, 1978). However, Leventhal (1983) suggested that a weight ratio of reduced sulphur/carbon which is consistently above 0.4 (e.g. Goldhaber & Kaplan 1974) with a positive intercept of the best fit line to the data on the sulphur axis of a S/C plot may indicate sulphate reduction within an anoxic water column as in the modern Black Sea. Finely laminated, pyritic sediments rich in organic carbon and with high S/C ratios may indicate an anoxic depositional environment, but these criteria do not indicate the role of primary productivity in producing either the anoxic conditions or the relatively high (> 1%) concentration of organic carbon typical of those sediments. Large amounts of biogenic silica (mainly diatoms and radiolaria in modern sediments) commonly indicate high surface water productivity (Berger 1976), particularly where upwelling of cooler, nutrient-enriched intermediate water occurs. Other valuable criteria for evaluating productivity are relatively high accumulation rates of organic carbon and total phosphorous (biogenic debris as well as authigenic phosphate minerals; e.g. Burnett et al. 1980) in sediments. The data illustrated in Fig. 6 are for several modern and ancient environments compiled by Glenn & Arthur (in press; see sources of data within). The fields marked Peru and SW Africa represent regions of high primary productivity associated
Deposition of organic-carbon-rich sediment DISSOLVED
0 "
.u
=
I
I
0.2 '
-'
0.5 9
'
02 AT S E D I M E N T / W A T E R INTERFACE mL/L 1.0
~
533
PARAMETER
8.5
I
~ i
.-~ .=
~--
o
IH.'4"---
BENTHIC
Well oxygenated . . . .
ENVIRONMENT
="ii
1
i: ~ ~ : ] "~.-~~~ ..". " ~ - - - - - Diverse metazoans . . . . . =~ i = i ; ~ =Szs ." (calcareous/soft-bodied) o : o ~=]= .*" z : z JL~: ~o .." Tracks, trails, large burrow systems, : i~g= = .." : :J ~.." vertical and horizontal - ~
:
i
i
.~i.
=,"
low
.."
,, ~, ==.,~ ~='~ diversity ..."
o ~ =2"~:~ " .- ." ,,~ ~ = ~ . ~ ~ /son- .." - - r o= ~ '~ ~ t ~ j S = / .. =_ ,..~ ^ ~=.~ ., ,,, ,,,,- :%,. .."..
INFAUNA (Burrowing)
Diverse shelled and
BENTHIC
soft-bodied . . . . . metazoans
EPIFAUNA
4
(Fl~ain ~ i c z o n e )
:
2~
k l*" : Laminated : to Laminated} burrow- 4
BENTHIC
-~i
mottled
Generally > 1%orgCi (good preservation,} high Hydrogen ~ Indices) amount } depends on supply i d oA ~
Few primary sedimentary structures preserved
TEXTURE
(homogeneous to bioturbate) Generally < 1% orgC oxidized/refractory (structured to dark amorphous, with Hydrogen Index < 100 mg HC/gC) [May vary with productivity/sed rate]
9 "6 / [~
SEDIMENT
SEDIMENT ORGANIC CARBON CONTENT SEDIMENT
Reduced SIC < 0.4 (generally) (may vary with sedimentation rate) sulphate-reduction only below Sed/H20 interface
REDUCEDSULPHUR CONTENT
Shades of black,-shades of grey s brown, olive green- grey ~ olive-green ~ to tan and red - C o l o u r depends on CaCO3 content, Corg content (and type), presence of acid volatile sulfides, etc. GENERAL
CRITERIA
BOTTOM-WATER LEVELS
IN M U D D Y
FOR
RECOGNITION
DISSOLVED MARINE
SEDIMENT COLOUR
OF
OXYGEN
ENVIRONMENTS
It
certain species of polychaetes (e.g. Calvert, 1966) and/or calcareous or arenaceous benthic foraminifers (e.g. Phleger and Soutar, 1973) may inhabit surficial environments where benthic metazoans are excluded.
2~
degree of lamination in part depends on seasonal variations in clastic input a l t e r nating with higher productivity intervals, and on sedimentation rate.
3~
see text.
FIG. 5. Oxygen content at the sediment/water interface and generalized expected relationships between types of benthic organisms, primary sedimentary structures, sediment chemistry and mineralogy (sulphides), organic content and type (data and concepts from numerous sources).
M . A . Arthur, W.E. Dean and D . A . V . Stow
534
~E o "5 E
% v-
4 PERU 3
P 57
uJ I-.,< I:E z
2
P 39
o I._I
3 P.~ 1
1 n."
o ,10. o a.
//~1462
~CIR 189B
~
S . W ~
/ /.,ACK SEA
0.5
l/
(UNIT A)
CIR 17713. '~CIR 179B DSDP
1470
t ~1432
r 12302--1
1233e--4
12327-4 12337-4
/N.W. AFRICA/ 12329-4 12328-4 12347.4 12345-4 tDSDP 0
T 5
"1 10
T 15
20
~ 25
30
35
40
45
ORGANIC CARBON ACCUMULATION RATE (1() s rnol cm 2 ~1)
FIG. 6. Plot of total phosphorus accumulation rate vs. organic carbon accumulation rate for Holocene sediments from the Peruvian slope, the south-west African shelf (Namibia), the Black Sea abyssal plain and slope, the north-west African slope, and for average mid-Cretaceous 'black shales' from selected Atlantic DSDP Sites (see Glenn & Arthur, in press, for sources of data and calculations). Letters and numbers indicate individual cores. with coastal divergences with well-developed midwater oxygen-minimum zones. Note that the organic carbon and total phosphorous accumulation rates for upwelling settings are much higher than for the Black Sea, which is characterized by an anoxic water column but moderate to low productivity. Seasonal wind-driven upwelling, high nutrient supply and high rates of organic carbon production and deposition over the relatively shallow shelf off SW Africa (e.g. Calvert & Price 1971) probably are much more important than oxygen levels in controlling the accumulation of organic carbon because bottom water
oxygen levels rarely fall below 1 ml/1 on either the shelf or slope (with the exception of a few local areas of < 0.1 ml/1), and because the organic-carbon-rich sediments are not laminated. Lack of lamination may also be due to a lack of strong seasonal variations in supply of sediment components other than diatoms and organic carbon. On the Peruvian margin, some sediments with higher organic-carbon accumulation rates are associated with deposition under low-oxygen conditions ( < 0.02 ml/l) within the oxygen-minimum zone. There are some differences in phosphorus and
Deposition of organic-carbon-rich sediment
535
of supply of organic matter probably offset some of the effects of an oxygenated environment. Transfer of organic carbon to sediments is largely by faecal pellets (Bishop et al. 1978), and marine lipids are the only ones present off the arid coast of Namibia. Concentrations of dissolved oxygen in bottom waters off Peru are significantly lower than in bottom waters of most areas off Namibia, and this must aid in preservation of organic carbon. The higher phosphorus accumulation rates of several Peruvian cores (P-36 and P-42, Fig. 6) may indicate a supply of phosphorus in addition to that associated with organic carbon
organic carbon accumulation rates between offshore Peru and SW Africa (Namibia). Surface water nutrient supply and productivity are nearly the same in both settings (Koblentz-Mishke et al. 1970; Huntsman & Barber 1977). The results shown in Fig. 6 for SW Africa all came from sediment samples from the upper 50 cm or less in cores from shallower than 134 m on the Namibian shelf; the results for Peru are from cores at depths of 186-645 m on the outer shelf or upper slope. The accumulation rates of organic carbon are not too different in either setting. The shallow depth of the Namibian examples and the high rate
_M C
KEY M - MAESTRICHTIAN
(M)
c - 1 % - 30~
~.
_
_
~'~ (T,R)
~7
C - CAMPANIAN S - SANTONIAN
mat 3 - 2 0 %
~_
c
s
~cT " I - 3%(M'T'R)
C - CONIACIAN
c
u
M'~
T - TURONIAN
A ~](M)
C - CENOMANIAN A - APTIAN
A A
H
'
H - HAUTERIVIAN
v
I
V - VALANGINIAN
B
B - BERRIASIAN
L
L JUR- LATE JURASSIC
, IUF
~ GENERAL CORE
AI
k
BI
MI?
I
1-
A
(M,T)
-6
V
VI
SITE 364
JR
357 c.~ ]:::.: ~%
(M) 23% (M)
V
A
-j
HI
c,li:
SITE
356
COVERY
D- 6%
A
C l : u: "r (T,R)
V .-g
'~' _..A 1-19%
--,,m
-1%
.~wis
I
SITE
m II
C
B
B - BAFIREMIAN
T
i
TI:A
A
A - ALBIAN
IIII =~,
c
cmF
C
<M)
Z-~c ~
I
530 EDIMENT INDICATING SITE 363
t-.i ~.,i ?
.2: A il
OR MUDSTONE " LAMINATED M
TO SLIGHTLY BIOTURBATE
c -g-
2%=ORGANIC CARBON CONTENT ~ N T
B
-fi-
"7"
Corg TYPE
(M)= MARINE AMORPHOUS
uR ~
(R) =OXIDIZED RESIDUAL A A
SITE 511
E
~
3~
330
0.5%
/ ,,J
,so
"7" c A , I% (M)
A
H
B
SITE
c --
-g-,
9
-~
3,,
~-
s/"
4% (M) 92%
V
r-
M
) a~, ') j" ~j
"T
(T) = TERRIGENOUS
-5% (M,T)
(~ 1 - 3% (T,R)
?"
v i L JUR SITE 327
FIG. 7. Stratigraphic and core recovery columns for DSDP Sites from South Atlantic Ocean basins for the Mesozoic. Diagram illustrates intervals of 'black shale' deposition, generalized contents and types of organic matter.
536
M.A. Arthur, W.E. Dean and D.A.V. Stow
cW
-~ C
HI
I
T'' A
I
C T~ T ul ~ sl
M
~~ T C T T u ~s C
C
M - -
S
A ,~
A
i
--
T
T
u
m
S
C A
A
~-g-
B
T
C
__~ ~ B~
H__ v
H
S
~A
C u S T
AIZ
--~
i A
----...-7'-
--
C
H
C
~s=
~"
o-t8%
A
41J
--
J~
_ _ l.?L.
-_---
A __ --ILU
o-4~ M,T
B
0-2.9%
B N
~ roll
--
T.R
r.,
H
V z~ T,R
--
I V I-
B
SITE SITE 39O
7-
392 SITE 534
SITE KEY
391
M - MAESTRICHTIAN C - CAMPANIAN S - SANTONIAN
M
C - CONIACIAN T - TURONIAN
? H
C
i
C - CENOMANIAN A - ALBIAN
T
A - APTIAN
m
U
T s
B - BARREMIAN
M
H - HAUTERIVIAN V - VALANGINIAN B - BERRIASIAN L JUR-LATE JURASSIC 2%=ORGANIC CARBON CONTENT DOMINANT Corg TYPE] (M):MARINE AMORPHOUS (T)= TERRIGENOUS
M 1
?=
C~
C ? --~ A
~
C M
O"3- I3 t'lb R
T~ c_-
AY TL
_M
c~ S
r
~
C
C =.~-~
c
V _~
~
c~
S
C T
C
(R) =OXIDIZED RESIDUAL ERAL CORE ECOVERY GRAPHIC ITHOLOGY
- -
A A
KEY- GRAPHIC LITHOLOGY ~
~
CLAY or CLAYSIONE
~
CARBONATE OOZE
~
CHALK
~
LIMESTONE
J ~
~
B
C o r g - R I C H INTERVAL
ARGILLACEOUS LIMESTONE VOLCANOGE NIC SEDIMENTS SANDSTONE or CONGLOMERATE DOLOMITE
9 9 CHERT
H --
B
- -
--~
-C --
~
SITE 101
T
A A ~ B A
A
H
B A
V
-- ? B'~
y
-~
_..i
SITE98
A
H ~
__V~r
B ,"~
_i-~' ~
SITE 100
S,TE4
SITE 99A
FIG. 8. Stratigraphic and core recovery columns for DSDP Sites from the western North Atlantic Ocean basin for the Mesozoic (see Key for explanation of symbols).
Deposition of organic-carbon-rich sediment
537
---~ ..~, M :, :-'A, _ _
M T~-:
C
s c 0-15% T.R
M ~ 0-4% T.R
0-3% T.R
0-1.2% T,R
C
C
A
C
?
A
T
HI
A
C
'rT
T
u
__
C
A
0-15% M,T
v~ B .
.
B
V B
M C
?
B i Hm~ .
H
?
,q,~l
.
B
S
A ~ SITE 105
T
0-5% M,T
SITE 384
S C
SITE 385
T C
.
?
0-}4 T,M ( a v . : 1.4 n : 4 8 )
A
SITE 382
A M
I
SITE 387
B
C
H
S
V B
c
:)
T
-] C ~
SITE 386
A
C
A B
I
s c__
80 ~ I
T V B
A
D 9
m
I
~
SITE 5
60 ~
50 ~
!.
9 0-11% M
C I
70 ~ ,
L
4r 0-10% T,M o.
A B
/=9o: \
30~ ~ \392~ ;s3,~ (
H V
\ w
2 0~
387" ,3a6
0-0.8% T.R
391 980
^ 4-~~1l t0~0
o418
~ 9 9
B
SITE 417
SITE 144
FIG. 8 (cont.)
538
M.A. Arthur, W.E. Dean and D.A.V. Stow
M -C
1
H T
- - ' . i - - - ~ : 0.1-2.5%: T, R A ~ (~ n:lS)
~
B
?
H v a__
-g-
C~
~
T~
M
S C 0.1 & 10 ( n - 2 ) T.R
--
U
C
s
M"
S
~-u --
c~
B
C
H
.
A
A A
U
A
s
A~
B
T
B
C
C
A
H V
A
A
A
SITE 397
B
B V B
s-c__
J B
SITE 136
SITE 416
.1-4.8% :M,T,R (av.=l.5:n :25)
C SITE 140
A
.1-6.1%~ T,R fay =.82:n:33
4% M ( v.=l S n=3N'
0al-6
--
0-55 M
C
(av. 21, n=19)
I~. ~2o~)
T ~
L A
T
SITE 370
~ul
.av 34. n=]7j
H
4-21% M
(0v.-lO%1
A
A~__
A
B
B
H
-H g :-:
H I A T U S
:36.5%1
0-1.3%;T,R (av.: .56: n=24)
S
A
~,
T
C ~
' ~ - ~ " 5%block
~"
~-~:;;"'~(aw0.43:n=205) (l"coaly" bed
M
A
-T-
V
..5
SITE 138
C
B __-~_H ~ 0-1.4%:T
0.1-1%
_5
c
S
T
T
av :O. 56%
V
T
C
T
H
i A
A
H
H
C
T
:C
-e - ~
SITE 137
A
A
A - -
i
S C -T
A
C
H
C
H i 3~wh.~ n
V __B
SITE 369
c
?
0,9-1.3%: T i n : 10}
A A B
H V B _.a
SITE 367
SITE 368
SITE 415
FIG. 9. S t r a t i g r a p h i c a n d core recovery c o l u m n s for D S D P Sites f r o m the eastern a n d n o r t h e r n N o r t h Atlantic Ocean basins for the M e s o z o i c (see K e y on Fig. 8 for e x p l a n a t i o n o f symbols).
Deposition of organic-carbon-rich sediment
539
K. E ' ( M - MAESTRICHTIAN C - CAMPANIAN -. .....
S " SANTONIAN
,
c
i:--_--:.
s
-----
.....
C - CONIACIAN A T
TURONIAN
-
C - CENOMANIAN A - APTIAN H - HAUTERIVIAN
_M,.
0-2~ T M tav ca. : %i
<1% T.M.R
-- 5_<-: ca. A ~~----: I av. t,r
A - ALBIAN B - BARREMIAN
C ~--:: ~ 0-9% M __
T 0-II% M,T
V - VALANGINIAN
c~
I%
B - BERRIAS~AN L JUR - LATE JURASSIC 2%-ORGANIC CARBON CONTENT
__--AB"~s176 ~':''j 0-3~ t.M R
DOMINANT Co,g TYPE~
H
T === o-3.5~ -- ~ M T C
V
(M)- MARINE AMORPHOUS (T) - TERRIGENOUS
N"
(R) -O XIOIZED RESIDUAL NER AL CORE ECOVERY
B
GRAPHIC ITHOLOGY
SITE 551
s
SITE 550
398
C
~_
~
i~ ~
0-6%1 M R I0,,.:<~%
KEY- GRAPHIC LITHOLOGY
g-.
V
c, 1"
B__..,.q~
IM~
T
-- U
bITE 549
~
Corg -RICH INTERVAL
~
CLAY or CLAYSTONE
~
CAR~ONAIE OOZE
~
CHALK
~
LIMESTONE ARGILLACEOUS LIMESTONE VOLCANOGENIC SEDIMENIS SANDSTONE or CONGLOMERATE
T C ~
A
H -~
M~
I
H
I
CI H T
T
uS
U
SITE 401
0 . 6 - 1.8%
0.1-.3.3% T M R
)-~ (av co C.:,,
A B H V g
Ii
V B
135
SITE 4OO
50~
ss;~s49 .S%,o2~
550
U-115% av = 4
T~ n:o~
~9
SITE 548
400ll-401-
398e~
B
e 1 3 5 ~
H V
41~ 397~
~3~.~38 a 6 J
B
140e
.._1
SITE
'0~
T
C A
20~
A
s_A C
DOLOMITE 9 9 CHERt
~
ii 368~
SITE 402
367a
FIG.
9 (cont.)
0 ~'
540
M.A. Arthur, W.E. Dean and D.A.V. Stow
or other biogenic debris (e.g. fish debris) perhaps associated with detrital material which has a much higher rate of supply off Peru than off Namibia. High-velocity currents are common in many upweUing regions (e.g. Huyer 1976; Summerhayes et al. 1979), and these currents may cause local winnowing or erosion of muddy sediment and organic matter and thereby reduce the rate of accumulation and/or content of organic carbon. Primary productivity is higher (by a factor of 1.5 to 2) off NW Africa than in surface waters of the Black Sea, yet organic-carbon accumulation rates are significantly lower in sediments from the deeper NW African margin than in time-equivalent Black Sea strata. The suites of cores in each basin (Fig. 6) come from equivalent depths so that the difference in accumulation rates in large part must be related to levels of dissolved oxygen in bottom waters. Such an accumulation rate matrix (Fig. 6) when better defined and understood, may provide important constraints on interpretation of ancient 'black shales'.
Lithology and distribution of late Mesozoic deep-sea black shales Jurassic-early Cretaceous The first occurrences of organic-carbon-rich strata in the Atlantic Ocean basins that have been documented by DSDP drilling are of late Jurassic age. Callovian-Oxfordian black mudstones were recovered at Sites 330 and 511 on the Falkland Plateau (Fig. 7). Organic-carbon contents are as high as 4% (see review in Dean et al., in press) and the organic matter is largely terrestrial or residual marine (Gilbert, in press; v o n d e r Dick et al. 1983). Deposition occurred under marginally oxic to occasionally anoxic water masses. Site 534 in the western North Atlantic (Fig. 8) also recovered Callovian-Oxfordian strata (Sheridan et al. 1982) which consist of interbedded greenishblack and brown radiolarian-rich claystone, and some interbedded grey limestone. Organic carbon contents are as high as about 4%, but most values are < 1%. The organic matter consists largely of terrestrial or oxidized (residual) marine material. Although anoxic conditions at the sea floor at Site 534 may have existed periodically as evidenced by finely laminated intervals, the main source of organic carbon was terrestrial (Summerhayes & Masran, 1983). The Callovian-Oxfordian has
been suggested as an 'Oceanic Anoxic Event' by Jenkyns (1980) because organic-carbon-rich marine deposits of that age occur in Tethyan sequences and in other parts of Europe (Hallam & Bradshaw 1979; Morris 1979). Interbeds of laminated dark-olive to black marlstone within bioturbated white to light-grey limestone of Tithonian to Neocomian age have been recovered from many deep-basin DSDP sites in the eastern and western basins of the North Atlantic (Dean et al. 1977; Jansa et al. 1979; Dean & Gardner 1982). The laminated dark marlstone beds generally contain more than 5% organic carbon and one bed at Site 367 in the Cape Verde Basin off north-west Africa (Fig. 9) contains 33% organic carbon (Dean et al. 1977). Turbidity currents were active in the deposition of at least some parts of this stratigraphic unit (Lancelot et al. 1972; Jansa et al. 1979). Jansa et al. (1979) suggested that this unit was deposited in an oxygenated bathyal environment, and that some of the lamination may represent reworking by bottom currents. Similar strata have been described from Tethyan localities as well (Bernoulli & Jenkyns 1974; Weissert et al. 1979; Arthur & Premoli Silva 1982). The end of deposition of the Neocomian carbonates apparently was caused by a sudden rise in the carbonate compensation depth (CCD) during the Aptian (Arthur 1979; Thierstein 1979), and the remainder of the section in the North Atlantic through the Eocene is dominated by multicoloured clay-rich strata. At some DSDP sites in the North Atlantic, the Neocomian carbonates are overlain by the so-called middle Cretaceous (Aptian-Albian to Turonian) blackshale facies (Figs 8 and 9), but at some sites the two are separated by a unit of interbedded red and green claystones of late Aptian to early Albian age (Dean & Gardner 1982). Middle Cretaceous The middle Cretaceous black shale facies represents the main period of accumulation of organic matter in the Atlantic, and at most DSDP sites it consists of interbedded green and black clay-rich lithologies in which the black shale commonly accounts for less than 50% of the section (Figs 7-9; see reviews by Arthur 1979; Tucholke & Vogt 1979; McCave 1979b; Arthur & Natland 1979; Tissot et al. 1979,1980; de Graciansky et al. 1982; Dean & Gardner 1982; Dean et al., in press). Concentrations of organic carbon in the black lithologies range from 2~o to 37% and are commonly more than 5% (Lancelot, Seibold et al. 1977; Tissot et al. 1979, 1980; Summerhayes 1981b; Dean & Gardner 1982), whereas the
Deposition
of organic-carbon-rich
interbedded green lithologies usually contain less than 0.5~ organic carbon. The type of organic matter in the black lithologies varies in both time and space (see Figs 7-9; Tissot et al. 1979, 1980; Habib 1979, 1982; de Graciansky et al. 1982; Summerhayes 1981b). Marine organic matter (i.e. organic matter characterized by type II kerogen of Tissot et al. 1974) for the most part only occurs in Aptian-Albian organic-carbon-rich strata in the South Atlantic (Fig. 7) and off the coast of north-west Africa (Fig. 9; Tissot et al. 1979, 1980; Summerhayes 1981b; de Graciansky et al. 1982). Everywhere else in the North Atlantic (Figs 8 and 9) the Aptian-Albian organic-carbon-rich strata are characterized mainly by terrestrial organic matter and some admixtures of highly degraded, residual organic matter with the exception of a few beds at Sites 386, 417, and 418 that contain marine organic matter. The predominance of marine organic matter and higher concentrations of organic carbon in sites in the south-eastern North Atlantic and along the continental margin of North Africa where palaeolatitudes were typical of trade wind circulation and coastal upwelling, and at Site 144 offnorthern South America, is due to higher rates of production of organic matter and consequently higher fluxes to the sea bottom and more intense midwater oxygen deficits. Numerous authors previously have pointed out this relationship for Cretaceous examples (Einsele & Wiedmann 1975; Arthur & Natland 1979; Tissot et al. 1979, 1980; Jenkyns 1980; Summerhayes 1981b; Parrish & Curtis 1982). In spite of these differences in sources of organic matter, most of the organic-carbon-rich strata are interbedded with bioturbated, organic-carbon-poor strata with abundant evidence of having been deposited in an oxygenated bottom water environment. Cenomanian to possibly earliest Turonian black shales in nearly all North and South Atlantic basin sites (see Figs 7-9) contain higher amounts of organic carbon (up to 30~o) much of which is amorphous marine material (Tissot et al. 1979, 1980; Summerhayes 1981b). Strata of this age may record an oceanwide productivity/preservation event (Schlanger & Jenkyns 1976; Arthur & Schlanger 1979; Jenkyns 1980; Scholle & Arthur 1980; Summerhayes 1981b). The Cenomanian-Turonian organic carbon burial episode is the best evidence for what is termed an 'oceanic anoxic event' (Schlanger & Jenkyns 1976) because it is both widespread and fairly brief in duration in comparison to the widespread but much longer period of deposition of BarremianAlbian 'black shales' (see later discussion). Many middle Cretaceous black shales have
sediment
541
very fine fissile lamination, and commonly have been interpreted as pelagic sediments that have accumulated under anoxic conditions. However, detailed examination (McCave 1979b; Stow & Dean, in press) reveals a greater variety of sedimentary structures in many intervals including horizontal silt-mud lamination, micro-cross lamination, small-scale grading and micro-bioturbation (Fig. 10). The interbedded green and red claystones commonly have a greater degree and a larger scale of bioturbation which has obliterated many of the same primary sedimentary structures as found in the black shales. It appears that both turbiditic and pelagic processes have operated throughout the deposition of most black shales sequences and are not specific to any one lithology. In the western North Atlantic, mid-Cretaceous black shales were recovered at Sites 101 and 105 (Lancelot et al. 1972), 386 and 387 (McCave 1979b), 391 (Benson et al. 1978), and 417 and 418 (Donnelly et al. 1980) and 534 (Sheridan et al. 1982) (Fig. 8). These are interbedded with red, green and grey mudstones and marlstones and thin radiolarian-rich sand/silt layers, and commonly comprise less than 50~ of the section (Figs 8 and 10). The black shales may be laminated, massive, or bioturbated and are variably enriched in organic carbon from mixed terrigenous and marine sources, although much of the marine organic carbon is degraded. Fine-grained turbidites comprise parts (mainly less than 30~) of the mid-Cretaceous section at each site, and occur in both the black shales and the associated lithologies (Fig. 10). There are thin-bedded laminated siltstone and mudstone turbidites, mostly far-travelled from the North American continental margin, as well as thicker-bedded biogenic turbidites, probably derived locally from intrabasin highs. In the eastern North Atlantic offthe Euro-African continental margin, mid-Cretaceous black shales (Fig. 9) that showed evidence of turbidite deposition were recovered at Sites 367, 368 and 370 (Dean et al. 1977; Dean & Gardner 1982), 397 (Cornford 1979), 398 (Arthur 1979; de Graciansky & Chenet 1979; Habib 1979), 400 and 402 (de Graciansky et al. 1979) and 415 (Lancelot et al. 1980). The black shales again form part (usually less than 50~) of a more oxic section comprising varicoloured mudstones and marlstones. Most of the turbidites recognized are also fine-grained siltstone or mudstone, although at Sites 370, 397, 398 and 402 there are interbedded coarser-grained and thicker-bedded turbidites, debris-flow deposits and slumped units (Fig. 10) including redeposited shallow-water carbonate material. There is no consistent association of
(a)
(b)
(c)
FIG. 10. Examples of redeposited sediments and sedimentary structures associated with 'black shale' facies in DSDP drillholes. (a) 64-480, P20-3, 30-60 cm: Guaymas Basin slope, Gulf of California. Hydraulic piston core interval in finely laminated muddy diatom ooze. The white laminae are richer in diatoms, and the light-dark couplets represent annual varves. Note the even, planar nature of the laminations and general lack of irregular or cross-cutting laminations. Such continuous fine lamination is typical under an oxygen-minimum zone of relatively high sedimentation rate settings with seasonal sedimentation contrasts. Some uncomformities and apparent cross-bedding, however, were noted in these cores (see Schrader et al. 1981). (b) 41-368, 37-3, 4, 47-85 cm, and (c) 41-368, 30-1, 62-70 cm: upper Palaeocene to lower Eocene, Cape Verde Rise eastern North Atlantic. Dark intervals in dark-light intercalations are mud turbidites (see Dean et al. 1977). Note sharp basal contacts and general lack of burrowing at base of each unit. Here the mud turbidites are black to dark green and dark olive-green and contain as much as 3 ~ organic carbon, but usually less than 1~o organic carbon. Scale is in centimetres. Photograph in (c) is a closeup of a mud turbidite containing a pyrite nodule.
Deposition of organic-carbon-rich sediment
(d)
(e)
543
(f)
FIG. 10. Examples of redeposited sediments and sedimentary structures associated with 'black shale' facies in DSDP drillholes. (d) 62-465A, 40-1, 88-107 cm: Hess Rise, Pacific Ocean. Laminated, olive, organic-carbon-rich limestone of late Albian age. Similar beds in sequence contain up to 9% organic carbon. The limestone contains radiolarians and organic carbon with a high hydrogen index. Although high productivity may have characterized surface waters, and bottom waters may have been anoxic, these laminated limestones probably were redeposited, as shown by cross-bedding in photo (see Dean et al. 1981). Bar is 3 cm. (e) 75-530B, 8-1 and 2, 19-51 cm: Southern Angola Basin adjacent to Walvis Ridge, South Atlantic. These Pleistocene sediments contain up to 8% organic carbon, but were largely redeposited by debris-flows to deep water from highly productive shallower water settings on the nearby Walvis Ridge. Conglomeratic character of the soft mudclasts demonstrates redeposition, but some intervals are dark, fine-grained and laminated (note lower right: see Stow 1983). Similar debris-flow deposits at Site 530 also occur in the Pliocene and Miocene sections. Bar is 5 cm. (f) 41-369A, 40-2 and 3, 81-125 cm: Continental slope off Cape Bojador, north-west Africa. This Santonian continental slope sequence of argillaceous chalks probably was deposited within an oxygen-minimum zone. Some intervals are laminated, alternating with bioturbated zones. Organic carbon contents are as high as 6.4% and average about 1.5%. Much of the sequence, like the strata illustrated here, has been redeposited by slumps and debris-flows. Scale is in centimetres.
544
M.A. Arthur, W.E. Dean and D.A.V. Stow
(g)
(h)
(i)
FIG. 10. Examples of redeposited sediments and sedimentary structures associated with 'black shale' facies in DSDP drillholes. (g) 41-367, 28-3, 41-62 cm, and (h) 411-3367, 28-3, 64-79 cm: Cape Verde Basin, eastern North Atlantic. These photos are typical examples of interbedded light grey limestones and dark olive-grey to black marlstones found in lower Cretaceous (Neocomian) sequences on both sides of the Atlantic. The dark beds typically are finely laminated (note discontinuous light-coloured, coccolith-rich laminae), whereas adjacent light grey limestones are highly bioturbated. The laminated marlstones typically contain up to 3 or 4% organic carbon of marine origin. Such sequences constitute the primary evidence that the deep Atlantic was at least periodically anoxic at or above the sea floor. Bars in both photographs are 2 cm. (i) 43-387, 49-5, 100-119 cm: western North Atlantic basin. Cycles of Neocomian marlstone-limestone similar to those in (g) and (h), but in this sequence the light-coloured limestones also are finely laminated. However, very faint burrow mottles do occur in the limestones. Although the laminations may indicate deposition under anoxic conditions, they may also be due to the action of bottom currents. High sedimentation rates, therefore, may have limited benthic activity. Bar is 2 cm.
Deposition of organic-carbon-rich sediment
(j)
(k)
545
(i)
FIG. 10. Examples of redeposited sediments and sedimentary structures associated with 'black shale' facies in DSDP drillholes. (j) 75530-A, 98-2, 85-105 cm and (k) 75-530A, 104-3, 90-105 cm: Angola Basin adjacent to Walvis Ridge, South Atlantic. Examples of Albian black shale or mudstone facies which are homogeneous to faintly laminated and contain up to 15% organic carbon, largely of marine derivation. Numerous thin dark layers with sharp basal contacts and bioturbated upper parts are evident in photograph (j). These may represent thin redeposited units (perhaps from adjacent Walvis Ridge). A thin bioturbated zone in photograph (k) passes through the middle of the black unit suggesting that this unit may be composed of more than one 'event'. Note also pyrite (light-coloured) at boundaries of oxidized and reduced lithologies in both photographs. Bars in both photographs are 2 cm. (1) 47B-398D, 77-3, 42-67 cm: Vigo Seamount, north-east Atlantic. Middle Albian black mudstone unit containing about 1~ organic carbon, mainly of terrestrial or residual origin. Note the light-coloured layers which contain concentrations of calcareous nannofossils and siderite. Much of this unit may have been redeposited (see Arthur 1979). Bar is 2 cm.
546
M.A. Arthur, W.E. Dean and D.A.V. Stow
(m)
(n)
(o)
FIG. 10. Examples of redeposited sediments and sedimentary structures associated with 'black shale' facies in DSDP drillholes. (m) 47B-398D, 116--1, 97-118 cm: Vigo Seamount, north-east Atlantic. This unit of upper Aptian black to dark grey mudstone is similar to (1) above, except that there is abundant evidence of redeposition of terrigenous sand and carbonate material, including large slump units just above. Bar is 2 cm. (n) 47B-398D, 123-4, 0-24 cm: Vigo Seamount, north-east Atlantic. Dark grey to black homogeneous to laminated mudstone and calcareous mudstone of early Aptian age contains up to 2.5~ organic carbon, largely of terrestrial and/or residual marine origin. Terrigenous sand/silt turbidites are common, and it is likely that much, if not all of this 'black shale' unit was redeposited. Bar is 2 cm. (o) 47B-398D, 127-4, 84-105 cm: Vigo Seamount, north-east Atlantic. Black mudstones of late Barremian to early Aptian age in this interval contain up to 2% organic carbon, mainly of terrestrial origin. Note the thin-bedded, fine-grained turbidities that are totally unbioturbated. Fine laminations are exceptionally well-preserved. Either bottom waters were anoxic or high sedimentation rates inhibited a benthic fauna. Bar is 2 cm.
Deposition of organic-carbon-rich sediment
(P)
547
(q)
FIG. 10. Examples of rediposited sediments and sedimentary structures associated with 'black shale' facies in DSDP drillholes. (p) 41-370, 35-5, 5-16 cm: Moroccan Basin, eastern North Atlantic. Dark intervals within Barremian turbidites contain an average of about 1% organic carbon, mainly of terrestrial origin. Redeposited intervals dominate the entire sequence recovered in the Moroccan Basin. The turbidite shown here contains glauconite, dolomitized ooids and other shelf carbonate debris. Bar is 1 cm. (q) 41-370, 41-3, 28-46 cm: Moroccan Basin, eastern North Atlantic. Barremian terrigenous silt/sand turbidite similar to photograph (n). Bar is 2 cm.
548
M.A.
Arthur,
W.E. Dean and D.A.V.
turbidites with any one particular facies. For example, at Site 398 thin black mudstone turbidites grade upwards into biogenic-rich pelagites, whereas at Site 400 biogenic calcilutite turbidites grade up into black shales, parts of which may represent pelagic sedimentation. Of the eight DSDP Sites in the South Atlantic (Fig. 7; see Dean et al., in press, for review) that have recovered mid-Cretaceous black shales, at least three show clear evidence of interbedded turbidites. At Site 361 in the Cape Basin (Natland 1978; Arthur & Natland 1979) thick sandstone and pebbly sandstone turbidites and slumps occur within a sequence of green, grey and black mudstones. The coarse-grained turbidites commonly are rich in terrestrial organic debris and grade upward into thinly laminated black shales some of which have more marine organic matter. Both pelagite and turbidite interpretations of the black shales at Site 361 are possible. The black shales and associated facies in the Angola Basin at Sites 364 (Bolli et al. 1978) and 530 (Stow & Dean, in press) have a much more distal or basinal character. Black shale beds comprise from 5 to 50~o of the section at Site 530, and occur as numerous thin beds with variable organic carbon content that is dominantly of marine origin (Meyers et al., in press). Very fine-grained, thin-bedded turbidites comprise between 5 and 20% of the section at Site 530 and occur both in the black shales and in the associated facies. The few occurrences of middle Cretaceous organic-carbon-rich strata in the Pacific, except those on southern Hess Rise (Sites 465 and 466) appear to have been caused by short-lived events of oxygen depletion that were associated with volcanic activity and downslope sediment redeposition (Dean et al. 1981; Thiede et al. 1982). These strata appear to be the result of a coincidence of local tectonic factors acting to produce and preserve organic matter at times that were not strictly synchronous but were within the same general middle Cretaceous time period as Atlantic-Tethyan black shale occurrences. The organiccarbon-rich strata on Hess Rise also are of the same general age and are associated with volcanic activity and sediment redeposition but cover a longer time span (late Albian-early Turonian) as Hess Rise crossed under the highly productive equatorial divergence. At all Pacific sites, organic-carbon-rich strata probably accumulated on the flanks of volcanic edifaces within oxygen-minimum zones in relatively shallow water (several hundred metres). An increase in organic carbon concentrations also occurs within Coniacian-Santonian sequences at DSDP Sites in the eastern South Atlantic (Fig. 7; the so-called sapropels of Ryan
Stow
& Cita 1977; McCoy & Zimmerman 1977), the eastern North Atlantic and the Caribbean (Arthur & Natland 1979). Many of these organiccarbon-rich beds may be the result of higher rates of supply of organic matter related to upwelling and/or enhanced preservation under marginally oxic conditions. Redeposition from adjacent continental margins, however, cannot be ruled out as a cause. Organic-carbon-rich strata at Sites 369 and 144 may have been deposited within expanded and intensified oxygen-minimum zones associated with coastal divergences and upwelling. Cenozoic
Eocene sediments in both the eastern and western basins of the North Atlantic are relatively enriched in organic carbon. Dean et al. (1977) and Dean & Gardner (1982) described rhythmically interbedded black and green radiolarian clay layers of Eocene age from Site 367 in the Cape Verde Basin in the eastern North Atlantic (Fig. 10). The black clay layers contain up to 4 ~ organic carbon of marine origin. These beds commonly have sharp basal contacts with silt stringers and bioturbated tops, and are best interpreted as mud turbidites. McCave (1979a) proposed a similar interpretation for dark siliceous Eocene mudstones in the western North Atlantic (Site 386) that contain up to 2~o organic carbon. More than 1000 m of lower to middle Miocene sediment at Site 397 on the Cape Bajador upper continental rise was redeposited from the outer shelf and upper continental slope along the continental margin of north-west Africa (Arthur & v o n Rad 1979; Arthur et al. 1979). Darkcoloured clays at Site 397, interpreted as mud turbidites, and the matrix and clasts of thick debris-flow deposits contain up to 7~o organic carbon of marine origin (Cornford 1979) (Fig. 10). Benthic foraminifera and sedimentary structures (Arthur & von Rad 1979) all indicate downslope redeposition of these sediments. Organic-carbon-rich sediments in the southeastern Atlantic off south-west Africa (Calvert & Price 1971) extend over a wide area to depths of at least 1330 m, and over a time-span of late Miocene to Recent (Site 362, Bolli et al. 1978; Site 532, Hay et al. 1982; Gardner et al., in press). At Site 532 the concentration of organic carbon of marine origin is as much as 8~ in the dark green and olive siliceous clays of late Pliocene to early Pleistocene age, and less in the interbedded calcareous oozes, marls and muds (Gardner et al., in press; Meyers et al., in press). All the lithologies are pelagic or hemipelagic in origin, are thor-
Deposition of organic-carbon-rich sediment oughly bioturbated and accumulated relatively rapidly at rates of 2.5 to 6 cm/1000 yrs under an oxygen-minimum zone. The late Miocene to Recent section at Site 530 in the Angola Basin, immediately north of the Walvis Ridge in 4650 m water depth also is high in organic carbon with a maximum of 6% in lower Pleistocene sediments (Meyers et al., in press). The organic-carbon-rich facies corresponds to that on Walvis Ridge except that they are all redeposited as thick-bedded mud-ooze turbidites and debrites (Fig. 10; Stow, in press). The Walvis Ridge and Angola Basin sediments are mostly well-bioturbated, non-laminated and dark greenish-olive in colour, but with subsequent burial to greater depth they may come to more closely resemble the Mesozoic black shales in appearance. Neogene black shales or sapropels are common in the Mediterranean and Black Seas, and they have been interpreted, in part, as being of turbidite origin (e.g. Sites 378 and 380, Hsfi et al. 1978; Ross et al. 1978). DSDP drilling off the coast of southern California and in the Gulf of California recovered relatively organic-carbon-rich sequences of Miocene to Recent age (e.g. Schrader et al. 1980; Summerhayes 1981b; Gilbert & Summerhayes 1981, 1982). Most of these sequences contain marine organic matter or mixtures of marine and terrestrial organic matter, and represent deposition associated with upwelling and preservation of organic matter under intense oxygen-minimum zones. They are younger analogues or offshore equivalents of the Miocene Monterey Formation (Pisciotto & Garrison 1981).
Discussion By analogy with modern examples of organiccarbon-rich sediments we can assume that three main variables were significant for the preservation of organic matter in ancient black shales. These were: (1) variation in the supply of both marine and terrigenous organic matter from surface productivity (pelagic settling, faecal pellets), fluvial discharge, redepositional processes and settling from slope and bottom nepheloidlayers; (2) variation in the rate of sedimentation and hence rate of burial of organic matter; (3) variation in bottom water oxygenation within an oxygen-minimum zone or throughout the basin as a result of oceanographic and climatic conditions.
549
Various authors have usually stressed the particular significance of one or another of these variables from study of a black shale sequence in a particular locality or region, and have, to some extent, presented models that were too generalized or simplified. Basin deoxygenation and expansion and intensification of an oxygen-minimum layer have been proposed to explain the accumulation of the middle Cretaceous organic-carbon-rich strata (e.g. Schlanger & Jenkyns 1976; Ryan & Cita 1977; Fischer & Arthur 1977; Thiede & van Andel 1977; Arthur & Schlanger 1979). Both models require very low to zero concentrations of dissolved oxygen in part of the water column as a result of reduced advection of oxygenated water and/or increased supply of organic matter. Both models imply that reducing conditions in the sediments (and therefore the increased degree of preservation of organic matter) are largely the result of anoxic or near-anoxic conditions in the overlying waters. On the other hand, the importance of supply factors, namely increased productivity, terrigenous runoff and/or redeposition of organic carbon by turbidity currents, have been stressed by Dean et al. (1977), Cornford (1979), de Graciansky et al. (1979), Habib (1979, 1982), Welte et al. (1979) and Natland (1978) for more specific areas off north-west Europe and northwest and south-west Africa. Settling of finegrained terrigenous organic matter from possible bottom nepheloid-layers also has been proposed as a mechanism of transporting organic carbon to deep-sea environments (Summerhayes 1981b). However, it seems clear to us that there are important differences between black shales of different ages and places, and even between different beds within the same sequence. No simple model therefore will be adequate to explain what is in fact a very complex depositional system, and we must appeal to a more complex interplay of controls each of which may have varied independently to some extent. The chief characteristics of ancient deep-water black shales that must be explained by any model are the following: (1) the approximate coincidence in time of the main periods of black shale accumulation in many ocean basins within the Jurassic, Cretaceous and Cenozoic; (2) the range of environments in which black shales were deposited, from open ocean to restricted basin, and from hundreds of metres to a few kilometres water depth; (3) the formation of many black shales under poorly-oxygenated waters as evidenced by bioturbation characteristics in Cretaceous sequences, and in fully-oxygenated ocean
550
M.A.
Arthur,
W.E. Dean and D.A.V.
basins such as those deposited during the Eocene and Miocene; (4) the interbedding of organic-carbon-rich facies with organic-carbon-poor facies, in cycles with irregular periodicities that average 20 000 to 140000 yrs, (Dean et al. 1977; McCave 1979b; Arthur 1979; Dean et al., in press), and with a variable order of occurrence of the different facies (McCave 1979b; Stow & Dean, in press); (5) the variation in lithologies that are enriched in organic-carbon, including mudstones, marlstones, limestones, sandstones and diatomites, all of which have been loosely referred to as 'black shales' (e.g. See Ibach 1982); (6) the range of processes that have been responsible for the deposition of different black shales, including pelagic, hemipelagic (e.g. settling from nepheloid-layer) turbiditic and other mass-flow processes, and the fact that no one process is uniquely associated with any one of the associated facies (Stow & Dean, in press); (7) the different types of organic matter and amounts of total organic carbon in black shales even within the same sequence. Factors in the origin of Cretaceous black shales
To explain these characteristics for the middle Cretaceous black shales that are so widespread throughout the Atlantic and Tethyan ocean as well as on isolated plateaus and seamounts in the Pacific, we must consider the worldwide conditions of climate and ocean circulation that probably were influential. Global climate during the middle Cretaceous was warm, eustatic sea-levels were generally high, spreading rates were fast, pelagic sediments in the world ocean were accumulating rapidly, and oceanic surface and bottom water temperatures were high (Douglas & Savin 1975; Fischer & Arthur 1977; Berger 1979; Brass et al. 1982). Increased surface water temperatures would have had two main effects: first, thermohaline deepwater circulation, driven today by the sinking of cold, oxygen-rich surface waters in high latitudes, would have been less efficient and driven mainly by salinity differences; and second, warm, saline water, which would contain lower initial concentrations of dissolved oxygen (Fig. 3), and was created on restricted, evaporative low latitude shelves and in semi-isolated basins would become the main deep-water mass in the ocean (Roth 1978; Arthur & Natland, 1979; Wilde & Berry 1982; Brass et al. 1982). It is not clear what the relative rates of deep-water overturn would be under such conditions, although Brass et al.
Stow
(1982) argued that turnover rates could be as high or higher than those of today, and Barron & Washington (1982) suggested that oceanic circulation on the whole was not 'sluggish' as is commonly assumed because of warmer surface and bottom waters during the Cretaceous. In any case, bottom water circulation was sufficient to supply some oxygen to maintain oxidizing conditions in the deep basins of the Pacific but oxygen supply was inefficient enough to permit depletion of dissolved oxygen over a broad range of midwater depths in areas of high productivity of organic matter, possibly because of lower initial dissolved oxygen contents in intermediate and deep-water masses due to the higher temperature and salinity. Even in the Angola Basin, which was highly restricted during the middle Cretaceous, circulation was sufficient to provide oxygenated bottom waters most of the time. The accumulation of organic-carbon-rich sediments at many places in the world ocean at times that were not always strictly synchronous undoubtedly is the result of a coincidence of several factors acting to produce and preserve organic matter. During the middle Cretaceous, much of the world ocean may have been so poised that relatively small changes in the flux of organic matter and/or rates of deep-water circulation at any one place may have caused anoxia or nearanoxia within midwater oxygen-minimum zones and possibly throughout much of the bottom water mass under extreme conditions in more tectonically restricted basins. An expanded and intensified oxygen-minimum zone during much of the early and middle Cretaceous would explain the excellent preservation of organic carbon, and the increase in accumulation rate of organic carbon over much larger areas in continental slope environments, on flanks of high-standing features on the sea floor, and in deep-sea settings largely by redeposition. It is not inconceivable that intensified oxygen-minimum zones periodically extended as deep as 2500 to 3000 m; this would help to explain the presence of upper Aptian-lower Albian black shales containing well-preserved organic carbon on or near the crest of the Mid-Atlantic Ridge in the North Atlantic (Site 386, Tissot et al. 1980; Summerhayes 1981b; Sites 417 and 418, Deroo et al. 1980) where it is unlikely that continental-margin-derived turbidites could have been the supply of the organic matter. Oxygen-minimum zones probably were the most expanded and intensified in areas of pronounced upwelling and high productivity, such as off north-west Africa (e.g. Berger & yon Rad 1972; Arthur & Natland 1979; Thierstein 1979; Fig. 9), which would explain the higher overall contents, accumulation rates, and
Deposition of organic-carbon-rich sediment more marine character of organic matter in mid-Cretaceous sediments from the eastern North Atlantic basin. This explanation also probably accounts for equatorial Pacific mid-Cretaceous black shales that accumulated on the flanks of high-standing seamounts and plateaus. In the south-east Angola Basin, burrows commonly become smaller and less abundant going from a green claystone to an overlying black shale bed (DSDP Site 530; Stow & Dean, in press), which suggests that periodic deep-basin oxygen-deficient conditions probably occurred there. However, although bottom water anoxia or near anoxia may have aided in the preservation of
Geern "sSh~mf ~ lfikents ~s S 1 ~h ~d : l m~
~
";~ ~
Highprod......
-If Black sed..... ~ " ~ U
(o,~o, . . . . ~oo >~o~
~__~.~___L~'
organic matter it was not necessarily the only cause for accumulation of organic-carbon-rich strata. An increase in the relative amount of organic debris deposited was equally important. At different sites, there were differences in the relative amounts of organic debris derived from areas of increased productivity within the basin and preserved in sediments under oxygen-minima on continental margins and organic debris from areas with increased terrigenous supply. In many places, organic matter was redeposited to basinal sites by turbidity currents; in other places, there was relatively rapid accumulation of pelagic organic matter. The main variables that may have
02;>0 ~
elln
Oz=O
_'y_ . . . . . . . . . . . . . . . . . . . . . . . . . .
\
L a t e l i b Y a n to E a r l y
Slope'~ ~
C. . . . . . i . . . . d Coniacian
8....
A
02>0 ~
T rb dr U
~
"~s
~
~~
I
0
o,I I
.......
,org. . . . . . . ,oo> 2 ~ o , - - ~ \ \
? T r O,od.....
.--. o..o
' . ' . o . . . . . . '--
551
....o . . . . . . ,,,~
Gf...... ~ _ ~ (.......... 09) / I black sediment ~ .// -4000m I
'J
I
FIG. 11. Cartoons illustrating the interplay between organic carbon fluxes from terrigenous sources and from surface water productivity, and changes in their relative importance in sedimentary deposits as the result of changes in climate, sea-level, upwelling and oceanic fertility, and deep-water oxygen contents. Also emphasized are pathways for organic matter deposition, including by downslope movement (e.g. turbidites) from river mouths (terrigenous organic matter) or from within oxygen-minimum zones (well-preserved marine organic matter), pelagic settling (e.g. faecal pellets and/or marine snow), and settling from nepheloid-layers (both marine and terrigenous organic matter). (A) Late Albian-early Cenomanian or Coniacian-Santonian in North and South Atlantic: low to moderate productivity with 'normal' mid-water oxygen-minimum zone, oxygenated (i.e. > 1.0 ml/1) deep-water masses. (B) Late Cenomanian-early Turonian or possibly middle Aptian-middle Albian in South and North Atlantic. Higher rate of supply of organic carbon and greater extent of oxygen depletion in deep-water masses (after Dean et al., in press).
552
M.A. Arthur, W.E. Dean and D.A.V. Stow
operated during a period of time when less organic matter was accumulating and during a period of time when more organic matter was accumulating are summarized in Fig. 11. The widespread distribution of black shale and related organic-carbon-rich facies occurred within relatively restricted time periods during the Cretaceous. The term 'Oceanic Anoxic Events' (OAE) has been applied to these periods (Schlanger & Jenkyns 1976). Unfortunately, this term has been construed by others to indicate that anoxic conditions existed throughout the water column in all ocean basins at precisely the same time. The original intent of the term was to point out that there were periods of time (BarremianAlbian, Cenomanian-Turonian and possibly Coniacian-Santonian) when organic-carbon-rich marine facies were much more widespread in epicontinental seas and open ocean basins than they are today. Although single black shale beds probably cannot be correlated from basin to basin, packages of organic-carbon-rich facies, no matter what their amount and type of organic matter, generally are time equivalent. This emphasizes the possibility that the occurrences of organic-carbon-rich facies in different basins might be linked in some way, even though they may occur in different lithologies and depositional settings. We favour fluctuations in sea-level and climate as the dominant causes for variations in amount of organic matter supplied and preserved (e.g. Fischer & Arthur 1977). Higher global sea-level and warm, equable climates probably influenced rainfall, production of terrestrial vegetation and runoffto the oceans, thereby increasing the flux of terrigenous sediment and organic matter to ocean basins, particularly the relatively narrow North and South Atlantic Oceans during the Hauterivian through Albian (e.g. Sites 398; 400; 402; Habib 1979). The rapid rate of supply of terrigenous organic matter to a basin that was already poorly-oxygenated (but not necessarily anoxic) probably led to enhanced oxygen deficits in some basins, and allowed preservation of what marine organic matter was produced and supplied to basin deeps. Locally, particularly in shallow epicontinental seas and restricted ocean basins, high runoff periodically may have led to less saline surface waters and stable stratification of the water column, which, in turn, produced deeper water oxygen deficits (e.g. Ryan and Cita 1977; Thierstein & Berger 1978; Arthur & Natland 1979). This would be particularly true if the supply of terrigenous organic matter was high, or if autochthonous productivity increased, perhaps stimulated by nutrients supplied by runoff. McCave (1979b) and Hochuli & Kelts (1980)
suggested that periodic (every 50 ky or less) high productivity intervals could have produced the interbedded organic-carbon-rich and organiccarbon-poor facies so typical of the Cretaceous deep-water black shale sequences. In general, however, sea surface productivity during the Cretaceous probably was low overall in comparison to that of today (Roth 1978; Berger 1979; Thierstein 1979), and only in certain regions (e.g. offN.W. Africa) can high sea surface productivity be called upon to produce Hauterivian-Albian deep-water organic-carbon-rich facies. Marine organic matter appears to be more prevalent in strata of mid-Aptian to early Albian age, and of Cenomanian age in many AtlanticTethyan sequences. This might indicate periods of overall higher sea surface fertility and intensification of midwater oxygen minima related to higher sea-level stands coupled with more rapid production of warm, saline bottom water (Southam et al. 1982; Brass et al. 1982) or some other mechanism (Tucholke & Vogt 1979; Summerhayes 198 lb). A high sedimentation rate may have been a factor in the preservation of organic matter in some regions (e.g. Ibach 1982), but some of the lowest sedimentation rate intervals, which occur during higher sea-level stands (e.g. de Graciansky et al. 1981), have the greatest enrichment of organic matter. Figure 12 illustrates a compilation of organic-carbon contents (core averages) for the western North Atlantic Cretaceous sequences plotted against interval sedimentation rates (compare with Fig. 2, after Miiller & Suess 1979). The dotted line represents the approximate Miiller & Suess (1979) relationship for sediments from modern marine environments corrected for compaction of the older Cretaceous strata. The scatter of points is high, and many of the points lie above the line, suggesting that the rate of organic carbon supply and/or preservation under lowoxygen conditions was at times more important than bulk sedimentation rate in causing enhanced organic carbon accumulation in Cretaceous ocean basins. Even so, most Cretaceous deepwater organic carbon accumulation rates were, on average, much slower than modern shallower water organic carbon accumulation rates (Fig. 6). Factors in the origin of Cenozoic organic-carbonrich strata
The oceanographic and climatic conditions under which the Cenozoic black shales were deposited in the Atlantic are very different from those in the Cretaceous. The ocean basin was wide and unrestricted by the Eocene so that basin floor deoxygenation is not likely and there are no character-
Deposition of organic-carbon-rich sediment 7-
553
SEDIMENTATION RATE vs Corg AVERAGE CRETACEOUS WESTERN NORTH ATLANTIC
SITE O SITE 3 8 6 [] SITE 105
s
A r~
FZ ILl 0 nLU n
SITE
101A
V SITE 391 SITE 387 [] SITES 417D.418A&B
4
_~p2
3 0
0 Lu .r w u.i ,r
m 2
0
**
[]
.
IB O
O o
a
0
[]
v
9
.
M-S relationship assuming ~ 50% compaction ~
,
=
9
*V
[] []
V
A
v
r~2
7 I
"7
A I
~
v I
10
,o
l
go
INTERVAL SEDIMENTATION RATE (mm 1000y-')
F16. 12. Plot of interval sedimentation rate (mm 1000 y - l) versus average organic carbon content by core (or core section) in DSDP Sites from the western North Atlantic. The points represent compilation of data from the Hauterivian through Cenomanian of Sites 386, 387, 101, 105, and 417, 418 (data in Arthur & Dean, in press). The dashed line is the 'best-fit' regression line of Mfiller & Suess (1979) for recent marine sediments adjusted for 50~ compaction of Cretaceous strata. Note large scatter of points (see text for discussion).
istics of the organic-carbon-rich facies that would suggest that such an event occurred. Global climate had begun to deteriorate near the beginning of the Tertiary so that by the late Eocene there was a fairly marked temperature differential between the poles and equator. There is considerable evidence now to infer that modern deep thermohaline circulation began at this time, with the formation of cold dense water at high southern latitudes and its northward flow into the Atlantic as Antarctic Bottom Water (Berggren & Hollister 1977; Kennett 1977; Tucholke & Vogt 1979; Shor & Poore 1978; Moore et al. 1978). Initially upwelling of older nutrient-rich waters forced by the deep circulation changes would have led to increased productivity, particularly along the continental margins, and hence to an increased supply of organic matter to the sediments. Redeposition of this material to the deeper
ocean basins (e.g. at DSDP Sites 367, 370 and 386) was most likely by turbidity currents and other mass-flow processes. In contrast to the above possibility, Brass et al. (1982) suggested that the production of warm saline bottom water was linked to eustatic sea-level rise and transgression of shelf areas. An early to middle Eocene sea-level highstand may have increased production of bottom water and increased deep-water turnover rates thereby stimulating biologic productivity, which could have led to intensification of oxygen-minimum zones (Southam et al. 1982). Kelts & Arthur (1981; and references within) further suggested that the widespread Eocene cherts of the deep North Atlantic ('Horizon A') may be explained by increased productivity of siliceous organisms on the margins and subsequent downslope redeposition. The middle to late Miocene to Pliocene
554
M.A. Arthur, W.E. Dean and D.A.V. Stow
organic-carbon-rich facies (mainly biogenic siliceous sediments) in the Atlantic (Diester-Haas & Schrader 1978; Gardner et al., in press) and Pacific (Ingle 1981; Summerhayes 1981a) also appear to be related to vigorous ocean circulation, upwelling and prolific diatom productivity. This was in response to the continued deterioration of global climate and mid-Miocene buildup of the Antarctic ice cap (Savin 1977; Kennett 1977). Antarctic Bottom Water, Arctic Bottom Water and Norwegian Sea Overflow Water all provided deep vigorous circulation at this time (e.g. Shor & Poore 1978). Organic-carbon-rich sediments accumulated under upwelling zones in the open, oxygenated north-east and south-east Atlantic, and all around the margin of the north Pacific which apparently was more fertile than the Atlantic because of an estuarine-type circulation (Berger 1970). A widespread period of tectonism in the circum-Pacific led to the formation of silled marginal basins which became deoxygenated where the midwater oxygen-minimum zone impinged on the bottom at sill depth (Ingle 1981). Lowered sea-level also may have helped to lower the base of the oxygen-minimum zone (Summerhayes 1981a), and coastal divergences may have moved seaward, off the shelf edge, and provided a more direct source of organic matter to deeper oceanic settings (e.g. Gardner et al., in press).
Conclusions From the above discussion it is apparent that the origin of interbedded more- and less-reduced lithologies with variable amounts of organic matter and variable amounts of pelagic, hemipe-
lagic, and terrestrial sediment is complex and probably is not due to any one simple process although events in different ocean basins may be linked by a common factor, such as changes in global sea-level. The main variables are supply of organic matter from land and from surface water productivity, sedimentation rate, and oxygen concentration in bottom waters. These variables are greatly influenced by climatic, oceanographic, geographic and tectonic factors. The supply of organic matter in turn determines the thickness and intensity of a midwater oxygen-minimum zone. Surface water productivity is determined, at least in part, by the intensity of upwelling of nutrient-rich water from within the upper part of the oxygen-minimum zone. Another important factor in the accumulation of deep-water black shales is the frequency and magnitude of sediment redepositional events. The middle Cretaceous black shales owe their origin to an interplay of these three main variables. At this time, the oceans were poised at relatively low oxygen levels which periodically tipped in favour of organic matter preservation, largely due to increased supply of organic matter from surface productivity or terrigenous input, commonly aided by downslope resedimentation. By contrast, the Cenozoic oceans witnessed an increasingly vigorous thermohaline circulation, possibly higher overall fertility, widespread upwelling and enhanced productivity. Organicmatter was preserved both in restricted marginal basins and in open ocean sites; redeposition was locally important. Each region and even each bed within a black shale sequence is likely to have its own particular combination of factors that led to organic-matter preservation, so that generalized models may not always be applicable.
References M.A. 1979. North Atlantic Cretaceous black shales: the record at Site 398 and a brief comparison with other occurrences. In: Sibuet, J.-C., Ryan, W.B.F. et al. (eds), lnit. Repts. DSDP, 47, part 2. US Govt. Print. Off., Washington, DC. 719-51. & N A T L A N D , J.H. 1979. Carbonaceous sediments in North and South Atlantic: the role of salinity in stable stratification of early Cretaceous basins. In: Talwani, M., Hay, W.W. & Ryan, W.B.F. (eds),
reef-reservoired giant oilfields: Bull. Am. Ass. Petrol.
ARTHUR,
- -
Deep Drilling Results in the Atlantic Ocean: Continental Margins and Paleoenvironment. Maurice
--
Ewing Series 3, Am. Geophys. Union, Washington, DC. 375-401. & SCHLANGER, S.O. 1979. Cretaceous 'oceanic anoxic events' as causal factors in development of
Geol., 63, 870-85.
--
- -
& PREMOLI SILVA, I. 1982. 2. Development of widespread organic carbon-rich strata in the Mediterranean Tethys. In: Schlanger, S.O. & Cita, M.B. (eds), Nature o f Cretaceous carbon-rich facies. Academic Press, London. 7-54. & VON R A D , U. 1979. Early Neogene base-of-slope sediment at Site 397, DSDP Leg 47A: sequential evolution of gravitative mass transport processes and redeposition along the northwest African passive margin. In: von Rad, U., Ryan, W.B.F., et al. (eds), Init. Repts. DSDP, 47A. US Govt. Print. Off., Washington, DC. 569-614. , VON RAD, U., CORNFORD, C., McCoY, F. &
Deposition of organic-carbon-rich sediment SARNTHEIN, M. 1979. Evolution and sedimentary history of the Cape Bojador continental margin, Northwest Africa. In: von Rad, U., Ryan, W.B.F., et al. (eds), Init. Repts. DSDP, 47A. US Govt. Print. Off., Washington, DC. 773-815. BARKER, P.F., DALZIEL,I.W.D. et al. 1977. Init. Repts. DSDP, 36. US Govt. Print. Off., Washington, DC. 1079 pp. BARRON, E.J. & WASHINGTON,W.M. 1982. Cretaceous climate: a comparison of atmospheric simulations with the geologic record: Palaeogeogr., Palaeoclimatol., Palaeoecol., 40, 103-33. BARTON, E.D., HUYER, A. & SMITH, R.L. 1977. Temporal variation observed in the hydrographic regime near Cabo Corviero in the northwest African upwelling region, February to April, 1974. Deep Sea Res., 24, 7-24. BENON, W.E., SHERIDAN, R.E. et al. 1978. Init. Repts. DSDP, 44. US Govt. Print. Off., Washington, DC. 1005 pp. BERGER, W.H. 1970. Biogenous deep-sea sediments: fractionation by deep-sea circulation. Bull. geol. Soc. Am., 81, 1385-1402. -1976. Biogenous deep-sea sediments: production, preservation and interpretation. In: Riley, J.P. & Chester, R. (eds), Chemical Oceanography, 5, 2nd Edition. Academic Press, London. 265-388. 1979. Impact of Deep Sea Drilling on paleoceanography, In" Talwani, M., Hay, W.W. & Ryan, W.B.F. (eds), Results of Deep Drilling in the Atlantic Ocean: Continental Margins and Paleoenvironment: Maurice Ewing Series 3, Am. Geophys. Union. 297-344. , DIESTER-HAASS, L. & KILLINGLEY, J.S. 1978. Upwelling off northwest Africa: The Holocene decrease as seen in carbon isotopes and sedimentologic indicators. Oceanologica Aeta, 1, 3-7. -& VON RAD, U. 1972. Cretaceous and Cenozoic sediments from the Atlantic Ocean. In: Hayes, D.E., Pimm, A.C. et al., Init. Repts. DSDP, 14. US Govt. Print. Off., Washington, DC. 787-953. BERGGREN, W.A. & HOLLISTER, C.D. 1977. Plate tectonics and paleocirculation-commotion in the ocean. Tectonophysies, 38, 11-48. BERNER, R.A. 1977. Stoichiometric models for nutrient regeneration in anoxic sediments. Limnol. Oeeanog., 22, 781-6. 1978. Sulfate reduction and the rate of deposition of marine sediments. Earth planet. Sci. Lett., 37, 492-8. BERNOtJLLI, D. 1972. North Atlantic and Mediterranean Mesozoic facies: a comparison. In: Hollister, C.D., Ewing, J.I. et al., Init. Repts. DSDP, 11. US Govt. Print. Off., Washington, DC. 801-72. & JENKYNS, H.C. 1974. Alpine, Mediterranean, and central Atlantic Mesozoic facies in relation to the early evolution of the Tethys. In: Dott, R.H., Jr. & Shaver, R.H. (eds), Modern and Ancient Geosynclinal Sedimentation. Soc. econ. PaleD. Min. Spec. Pub. 19, 129-60. BISHOP, J.K., KETTEN,D.R. & EDMOND,J.M. 1978. The chemistry, biology and vertical flux of particulate material from the upper 400 m of the Cape Basin in
-
-
-
-
555
the southeast Atlantic Ocean, Deep Sea Res., 25, 1
1
2
1
-
6
1
.
BOLLI, H.M., RYAN, W.B.F. et al. 1978. Init. Repts. DSDP, 40. US Govt. Print. Off., Washington, DC. 1079 pp. BRASS, G.W., SOUTHAM,J.R. & PETERSON,W.H. 1982. Warm saline bottom water in the ancient ocean. Nature, 296, 620-3. BREMNER, J.M. 1980. Physical parameters of the diatomaceous mud belt off South West Africa. Marine Geol., 34, M67-M76. BRONGERSMA-SANDERS,M. 1971. Origin of major cyclicity of evaporites and bituminous rocks: an actualistic model. Marine Geol., 11, 123-44. BURNETT, W.C., VEEH, H.H. & SOUTAR, A. 1980. U-series oceanographic, and sedimentary evidence in support of recent formation of phosphorite nodules off Peru. In: Bentor, Y.K. (ed.), Marine Phosphorites. Soc. econ. PaleD. Min. Spec. Pub. 29, 61-71. CALVERT, S.E. 1964. Factors affecting distribution of laminated diatomaceous sediments in Gulf of California. In van Andel, T.H. & Shot, G.G., (eds), Marine Geology of the Gulf of California. Am. Ass. Petrol. Geol. Mere. 3, 311-30. 1966a. Accumulation of diatomaceous silica in the sediments of the Gulf of California. Bull. geol. Soc. Am., 77, 569-96. 1966b. Origin of diatom-rich, varved sediments from the Gulf of California. J. Geol., 74, 546-65. & PRICE, N.B. 1971. Upwelling and nutrient regeneration in the Benguela Current, October 1968. Deep Sea Res., 18, 505-23. CORNFORD, C. 1979. Organic deposition at a continental rise: organic geochemical interpretation and synthesis at Site 397, Eastern North Atlantic. In: von Rad, U., Ryan, W.B.F. et al., Init. Repts. DSDP, 47A. US Govt. Print. Off., Washington, DC. 503-10. DEAN, W.E., A~TnUR, M.A. & STOW, D.A.V. 1984. Origin and geochemistry of Cretaceous deep-sea black shales and multicolored claystones, with emphasis on DSDP Site 530, southern Angola Basin. In: Hay, W.W., Sibuet, J.C. et al., Init. Repts. DSDP, 75. US Govt. Print. Off., Washington, DC, in press. & GARDNER, J.V. 1982. Origin and geochemistry of redox cycles of Jurassic to Eocene age, Cape Verde Basin (DSDP Site 367), continental margin of northwest Africa. In: Schlanger, S.O. & Cita, M.B. (eds), Nature and Origin of Cretaceous Organic Carbon-Rich Facies. Academic Press, London. 55-78. - - , GARDNER,J.V., JANSA, L.F., CEPEK, P. & SEmOLD, E. 1977. Cyclic sedimentation along the continental margin of northwest Africa. In: Lancelot, Y., Seibold, E. et al., Init. Repts. DSDP, 41. US Govt. Print. Off. Washington, DC. 965-86. --, TH1EDE, J. & CLAYPOOL, G.E. 1981. Origin of organic carbon-rich mid-Cretaceous limestones, Mid-Pacific Mountains and southern Hess Rise. In: Vallier, T.L., Thiede, J. et al., Init. Repts DSDP, 62. US Govt. Print. Off. Washington, DC. 877-90. DEGENS, E.T. & MOPPER, K. 1976. Factors controlling
-
-
-
-
-
-
-
-
556
M.A. Arthur, W.E. Dean and D.A.V. Stow
the distribution of early diagenesis of organic matter in marine sediments. In: Riley, J.P. & Chester, R. (eds), Chemical Oceanography, 6, 2nd Edition. Academic Press, New York. 59-113. -& Ross, D.A. (eds) 1974. The Black Sea-Geology, Chemistry and Biology. Am. Ass. Petrol. Geol. Mem., 20, 633 pp. -& STOFFERS,P. 1976. Stratified waters as a key to the past. Nature, 263, 22-7. DEMAISON, G.J. & MOORE, G.T. 1980. Anoxic environments and oil source bed genesis. Bull. Am. Ass. Petrol. Geol., 64, 1179-209. DEROO, G., HERBIN, J.P., ROUCACHE, J. & T1SSOT, B. 1980. Organic geochemistry of Cretaceous sediments at DSDP Holes 417D (Leg 51), 418A (Leg 52), and 418B (Leg 53) in the Western North Atlantic. In: Donnelly, T.W., Francheteau et al., Init. Repts. DSDP, 51-53, part 2. US Govt. Print. Off., Washington, DC. 737-46. , HERBIN, J.P., ROUCACHE, J., BOUDON, J.P., ROBERT, P., JARDINE, S. & MARESTANGE,P. 1982. Geochemistry and optical study of organic matter in some Pleistocene and Pliocene sediments from DSDP/IPOD Site 474 to 481 of Leg 64, Gulf of California. In: Curray, J.R., Moore, D.S. et al. lnit. Repts. DSDP, 64. US Govt. Print. Off., Washington, DC. 855-64. DEUSER, W.G. 1974. Evolution of anoxic conditions in the Black Sea during the Holocene. In: Degens, E.T. & Ross, D.A. (eds), The Black Sea-Geology, Chemistry and Biology. Am. Ass. Petrol. Geol. Mem., 20, 133-6. DE GRACIANSKY, P.C., AUFFRET, G.A., DUPEUBLE, P., MONTADERT, L. & MULLER, C. 1979. Interpretation of depositional environments of the Aptian/Albian black shales on the north margin of the Bay of Biscay (DSDP Sites 400 and 402). In: Montadert, L., Roberts, D.G. et al., Init. Repts. DSDP, 48. US Govt. Print. Off., Washington, DC. 877-907. DE GRACIANSKY,P.C. et al. 1982. Les formations d'fige Cr&ac6 de l'Atlantique nord et leur mati&e organique: palaeog6ographie et milieux de d6pft. Rdv Inst. Francais Pdtrdle, 37, 275-336. & CHENET, P. 1979. Sedimentological study of Cores 138 to 56 (upper Hauterivian to middle Cenomanian): an attempt at reconstruction of paleoenvironments. In: Sibuet, J.-C., Ryan, W.B.F. et al., Init. Repts. DSDP, 47B. US Govt. Print. Off., Washington, DC. 403-18. DIESTER-HAASS, L. & SCHRADER, H.J. 1979. Neogene coastal upwelling history of northwest and southwest Africa. Marine Geol., 29, 39-53. DONNELLY, T.W., FRANCHETEAU,J. et al. 1980. Init. Repts DSDP, 51-53, part 1. US Govt. Print. Off. Washington DC. 351 pp. DOUGLAS, R.G. & SAVIN, S.M. 1975. Oxygen and carbon isotope analyses of Tertiary and Cretaceous microfossils from Shatsky Rise and other sites in the North Pacific Ocean. In: Larson, R.L., Moberly, R.L. et al., Init. Repts. DSDP, 32. US Govt. Print. Off., Washington, DC. 509-21. DSDP LEG 76 SCIENTIFICPARTY 1981. Challenger drills at sites off east coast. Geotimes, 26, 23-5. DUNBAR, R.B. & BERGER,W.H. 1981. Fecal pellet flux -
-
to modern bottom sediment of Santa Barbara Basin (California) based on sediment trapping. Bull. geol. Soc. Am., 92, 212-8. EINSELE,G. & KELTS, K. 1982. Pliocene and Quaternary mud turbidites in the Gulf of California: Sedimentology, mass physical properties and significance. In: Curray, J.R., Moore, D.G. et al., Init. Repts. DSDP, 64. US Govt. Print. Off., Washington, DC. 511-28. - & WIEDMANN, J. 1975. Faunal and sedimentological evidence for upwelling in the Upper Cretaceous coastal basin of Tarfaya, Morocco. lOth International Congress on Sedimentology, Nice, Theme 1. 67-72. EMBLEY, R.W. 1976. New evidence for the occurrence of debris flow deposits in the deep sea. Geology, 4, 371-4. & MORLEY, J.J. 1980. Quaternary sedimentation and paleoenvironmental studies of Namibia (SW Africa). Marine Geol., 36, 183-204. EMERY, K.O. 1960. The Sea off Southern California: A Modern Habitat o f Petroleum. John Wiley, New York. 366 pp. FISCHER, A.G. & ARTHUR, M.A. 1977. Secular variations in the pelagic realm. In: Cook, H.E. & Enos, P. (eds), Deep Water Carbonate Environments. Soc. econ. Paleon. Min. Spec. Pub. 25, 19-50. GARDNER, J.V., DEAN, W.E. & WILSON, C. 1983. Carbonate and organic-carbon cycles and the history of upwelling DSDP Site 532, Walvis Ridge, South Atlantic Ocean. In: Hay, W.W., Sibuet, J.-C., et al., lnit. Repts. DSDP, 75. US Govt. Print. Off., Washington, DC., in press. GASKELL, S.J., MORRIS, R.J., EGLINTON,G. & CALVERT, S.E. 1975. The geochemistry of a recent marine sediment off northwest Africa. An assessment of source of input and early diagenesis. Deep Sea Res., 22, 777-89. GILBERT, D. 1984. Organic facies variations in the Mesozoic South Atlantic. In: Hay, W.W., Sibuet, J.-C. et al., lnit. Repts. DSDP, 75. US Govt. Print. Off., Washington, DC., (in press). GILBERT, D. & SUMMERHAYES,C.P. 1981. Distribution of organic matter in sediments along the California continental margin. In: Yeats, R.S., Haq, B.U. et al., lnit. Repts. DSDP, 63. US Govt. Print. Off., Washington, DC. 757-61. -1982. Organic facies and hydrocarbon potential in the Gulf of California. In: Curray, J.R., Moore, D.S. et al., Init. Repts. DSDP, 64. US Govt. Print. Off., Washington, DC. 865-70. GLENN, C.R. & ARTHUR, M.A. 1983. Sedimentary and geochemical indicators of productivity and oxygen contents in modern and ancient basins, I. The Holocene Black Sea as the "type" anoxic basin. Chemical Geol., (in press). GOLDHABER, M.B. & KAPLAN, I.R. 1974. The sulfur cycle. In: Goldberg, E.D. (ed.), The Sea, 5, Marine Chemistry. John Wiley, New York. 569-655. GRASSHOFF,K. 1975. The hydrochemistry of landlocked basins and fjords. In: Riley, J.P. & Chester, R. (eds), Chemical Oceanography, 2, 2nd Edition. Academic Press, New York. 456-597. HABIB, D. 1979. Sedimentary origin of North Atlantic Cretaceous palynofacies. In: Talwani, M., Hay, W. -
-
Deposition of organic-carbon-rich sediment
557
continents to oceans: J. geol. Soc., London, 137, 171-88. KELTS, K. & ARTHUR, M.A. 1981. Turbidites after ten years of Deep Sea Drilling-wringing out the mop?. In: Warme, J.E., Douglas, R.G. & Winterer, E.L. (eds), The Deep Sea Drilling Project: A Decade o f Progress. Soc. econ. PaleD. Min. Spec. Pub. 32, 91-1127. KENDR1CK, J.W. 1979. Geochemical studies of black clays from Leg 43, Deep Sea Drilling Project. In: Tucholke, B. Vogt, P. et al., Init. Repts. DSDP, 43. US Govt. Print. Off., Washington, DC. 633--42. KENNETT, J.P. 1977. Cenozoic evolution of Antarctic glaciation, the circum-Antarctic Ocean, and their impact on global paleoceanography. J. geophys. Res., 82, 3843-60. KIDD, R.B., C1TA, M.B. & RYAN, W.B.F. 1978. Stratigraphy of Eastern Mediterranean sapropel sequences recovered during DSDP Leg 42A and their paleoenvironmental significance. In: Hsfi, K., Montadert, L. et al., Init. Repts. DSDP, 42, part 1. US Govt. Print. Off., Washington, DC. 421-43. KOBLENTZ-MISHKE, O.J., VOLKOVINSKY, V.V. & KABANOVA, J.G. 1970. Plankton primary production of the world ocean. In: Wooster, W.S. (ed.), Scientific Exploration of the South Pacific. National Academy of Science, Washington, DC. 183-93. LANCELOT, Y., HATHAWAY, J.C. & HOLLISTER, C.D. 1972. Lithology of sediments from the western North Atlantic. Leg 1l, Deep Sea Drilling Project. In: Hollister, C.D., Ewing, J.I. et al., Init. Repts. DSDP, l l . US Govt. Print. Off., Washington, DC. 901-50. - - , SEIBOLD,E. et al. 1977. Init. Repts. DSDP, 41. US Govt. Print. Off., Washington, DC. 1259 pp. - - , WINTERER,E.L. et al. 1980. Init. Repts. DSD P, 50. US Govt. Print. Off., Washington, DC. 868 pp. LEVENTHAL, J.S. 1983. Relationship between organic carbon and sulfide sulfur in euxinic sediments. Geochim. cosmochim. Acta, 46, 133-8. MCCAVE, I.N. 1979a. Diagnosis of turbidites at Sites 386 and 387 by particle-counter size analysis of the silt (2-4/~m) fraction. In: Tucholke, B., Vogt, P.R. et al., Init. Repts. DSDP, 43. US Govt. Print. Off., Washington, DC. 395-405. - 1979b. Depositional features of organic black and green mudstones at DSDP Sites 386 and 387, 159-80. western North Atlantic. In: Tucholke, B.E., Vogt, JANSA, L.F., GARONER, J.V. & DEAN, W.E. 1977. P.R. et aL, lnit. Repts. DSDP, 43. US Govt. Print. Mesozoic sequences of the central North Atlantic. Off., Washington, DC. 411-6. In: Lancelot, Y., Seibold, E. et al., Init. Repts. McCoY, F.W. & ZIMMERMAN,H.B. 1977. A history of DSDP, 41. US Govt. Print. Off., Washington, DC. sediment lithofacies in the South Atlantic Ocean. In 991-1031. Perch-Nielson, K., Supko, P. et al., Init. Repts. , ENOS, P., TUCHOLKE, B.E., GRADSTEIN, F.M. & DSDP, 39. US Govt. Print. Off., Washington, DC. SHERIDAN, R.E. 1979. Mesozoic-Cenozoic sedimen1047-79. tary formations of the North Atlantic basin; western North Atlantic. In: Talwani, M., Hay, W.W. & MEYERS, P.A., BRASSELL, S.C. & HUE, A.Y. 1984. Geochemistry of organic carbon in South Atlantic Ryan, W.B.F. (eds), Deep Drilling Results in the sediments from Deep Sea Drilling Project Leg 75. In: Atlantic Ocean: Continental Margins and PaleoenHay, W.W., Sibuet, J.-C. et al., Init. Repts. DSDP, vironment. Maurice Ewing Series 3, Am. Geophys. 75. US Govt. Print. Off., Washington, DC. (in Union, Washington DC. 1-57. press). JENKINS, J.A. & WILLIAMS, D.F. 1983. A model for Pleistocene anoxic events in the Eastern Mediter- MOBERLY, R.L. & SCHLANGER, S.O. et al. 1983. The Mesozoic superocean. Nature, 302, 381. ranean. Bull. geol. Soc. Am., (in press). MOORE, J.C., JR., VAN ANDEL., T.H., SANCETTA,C. & JENKYNS, H.C. 1980. Cretaceous anoxic events: from
& Ryan, W.B.F. (eds), Deep Drilling Results in the Atlantic Ocean." Continental Margins and Paleoenvironments. Maurice Ewing Series 3, Am. Geophys. Union, Washington DC. 420-37. 1982. Sedimentary supply origin of Cretaceous black shales. In: Schlanger, S.O. & Cita, M.B. (eds), Nature and Origin of Cretaceous Carbon-rich Facies. Academic Press, London. 113-28. HALLAM, A. & BRADSHAW, M.J. 1979. Bituminous shales and oolitic ironstones as indicators of transgressions and regressions. J. geol. Soc. Lond., 136, 157-64. HAY, W.W., S1BUET, J.-C., et al. 1982. Sedimentation and accumulation of organic carbon in the Angola Basin and on Walvis Ridge: preliminary results of Deep Sea Drilling Project Leg 75. Bull. geol. Soc. Am., 93, 1038-50. HEATH, G.R., MOORE, T.C. & DAUPHIN, J.P. 1977. Organic carbon in deep-sea sediments. In: Anderson, N.R. & Malahoff, A., (eds), The Fate of Fossil Fuel C02 in the Oceans. Plenum Press, New York. 605-25. HsO, K.J., MONTADERT, L. et al. 1978. Init. Repts. DSDP, 42A. US Govt. Print. Off., Washington, DC. 1249 pp. HULSEMANN, J. & EMERY, K.O. 1961. Stratification in recent sediments of Santa Barbara Basin as controlled by organisms and water character. J. Geol., 69, 279-90. HUNTSMAN, S.A. & BARBER, R.T. 1977. Primary production off northwest Africa: the relationship to wind and nutrient conditions. Deep Sea Res., 24, 25-33. HUYER, A. 1976. A comparison of upwelling events in two locations: Oregon and northwest Africa. J. mar. Res., 34, 531-46. IBACIq, L.E.J. 1982. Relationship between sedimentation rate and total organic carbon content in ancient marine sediments. Bull. Am. Ass. Petrol. Geol, 66, 170-88. INGLE, J.C. 1981. Origin of Neogene diatomites around the North Pacific rim. In: Garrison, R.E., Douglas, R.G., Pisciotto, K.E., Isaacs, C.M. & Ingle, J.C. (eds), The Monterey Formation and Related Siliceous Rocks of California. Pacific Section, Soc. econ. PaleD. Min., Spec. Pub.
558
M.A. Arthur, W.E. Dean and D.A.V. Stow
PISIAS, N. 1978. Cenozoic hiatuses in pelagic sediments: Micropaleontology, 14, 113-38. MORRIS, K.A. 1979. A classification of Jurassic marine shale sequences: an example from the Toarcian (Lower Jurassic) of Great Britain: Palaeogeogr., Palaeoclimatol., Palaeoecol., 26, 117-20. MULLER, P.S. & SUESS,E. 1979. Productivity, sedimentation rate, and sedimentary organic carbon content in the oceans. Deep Sea Res., 26, 1347-62. NARDIN, T.R., EDWARDS,B.D. & GORSLINE,D.S. 1979. Santa Cruz Basin, California Borderland: Dominance of slope processes in basin sedimentation. In: Doyle, L.J. & Pilkey, O.H., Jr. (eds), Geology of Continental Slopes. Soc. econ. PaleD. Min. Spec. Pub. 27, 209-22. NATLAND, J.H. 1978. Composition, provenance, and diagenesis of Cretaceous clastic sediments drilled on the Atlantic continental rise off southern Africa, DSDP Site 361. In: Bolli, H., Ryan, W.B.F. et al., Init. Repts. DSDP, 40. US Govt. Print. Off., Washington, DC. 1025-62. NESTEROFF,W.D. 1973. Petrography and mineralogy of sapropels. In: Ryan, W.B.F., Hsii, K.J. et al., lnit. Repts. DSDP, 13. US Govt. Print. Off., Washington, DC. 713-20. PARRISH, J.T. & CURTIS, R.L. 1982. Atmospheric circulation, upwelling, and organic-rich rocks in the Mesozoic and Cenozoic eras. Palaeogeogr., Palaeoclimatol., Palaeoecol., 40, 31-66. , ZIEGLER, A.M. & SCOTESE, C.R. 1982. Rainfall patterns and the distribution of coals and evaporites in the Mesozoic and Cenozoic. Palaeogeogr., Palaeoclimatol., Palaeoecol., 40, 67-10 I. PETERS, K.E. & StMONEIT, B.R.T. 1982. Rock-eval pyrolysis of Quaternary sediments from DSDPIPOD Leg 64, Sites 479 and 480, Gulf of California. In: Curray, J.R., Moore, D.S. et al., Init. Repts. DSDP, 64. US Govt. Print. Off., Washington, DC. 925-32. PHLEGER, F.P. & SOUTAR, A. 1973. Production of benthic foraminifera in three east Pacific oxygen minima. Micropaleontology, 19, 110-5. PISCIOTTO, K.A. & GARRISON, R.E. 1981. Lithofacies and depositional environments of the Monterey Formation, California. In: Garrison, R.E., Douglas, R.G., Pisciotto, K.E., Isaacs, C.M. & Ingle, J.C. (eds), The Monterey Formation and Related Siliceous Rocks of California. Soc. econ. PaleD. Min. Spec. Pub., 97-122. RHOADS, D.C. & MORSE, J.W. 1971. Evolutionary and ecological significance of oxygen-deficient marine basins. Lethaia, 4, 413-28. RHODES, D.C. 1973. The influence of bottom-feeding benthos on water turbidity and nutrient recycling. Am. J. Sci., 273, 1-22. RITTENBERG, S.C., EMERY, K.O. & ORR, W.L. 1953. Regeneration of nutrients in sediments of marine basins. Deep Sea Res., 3, 23-45. Ross, D.A., NEPROCHNOV, S. et al. 1978. Init. Repts. DSDP, 42, part 2. US Govt. Print. Off., Washington, DC. 1244 pp. ROSSIGNOL-STRICK, M., NESTEROFF, W., OLIVE, P. & VERGRNAUD-GRAZZINI,C. 1982. After the deluge:
Mediterranean stagnation and sapropel formation. Nature, 295, 1105-1110. ROTH, P.H. 1978. Cretaceous nannoplankton biostratigraphy and oceanography of the northwestern Atlantic Ocean. In: Benson, W.E., Sheridan, R.E. et al., Init. Repts. DSDP, 44. US Govt. Print. Off., Washington, DC. 731-52. RYAN, C.B.F. 1972. Stratigraphy of late Quaternary sediments in the eastern Mediterranean. In: Stanley, D.J. (ed.), The Mediterranean Sea--A Natural Sedimentation Laboratory. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 149-69. RYAN, C.B.F. & CITA, M.B. 1977. Ignorance concerning episodes of ocean-wide stagnation. Mar. Geol., 23, 197-215. SAVRDA, C.E., BOTTJER, D.J., & GORSLINE,D.S. 1982. Refinement of benthic basin biofacies models: biogenic sedimentary structures and macrofauna from Santa Monica Basin, California Continental Borderland. Geol. Soc. Am. Prog. Abst., 14, 608. SCHLANGER, S.O. • JENKYNS, H.C. 1976. Cretaceous oceanic anoxic events-causes and consequences. Geologic en Mijnbouw, 55, 179-84. SCHOLLE, P.A. & ARTHUR, M.A. 1980. Carbon isotopic fluctuations in pelagic limestones; potential stratigraphic and petroleum exploration tool. Bull. Am. Ass. Petrol. Geol., 64, 67-87. SCHRADER, H. et al. 1980. Laminated diatomaceous sediments from the Guaymas Basin Slope (Central Gulf of California): 250,000-year climate record. Science, 207, 1207-9. SHANKS, A.L. & TRENT, J.D. 1980. Marine snow: sinking rates and potential role in vertical flux. Deep Sea Res., 27, 137-44. SHERIDAN, R.E., GRADSTEIN, F.M. et al. 1982. Early history of the Atlantic Ocean and gas hydrates on the Blake Outer Ridge: results of the Deep Sea Drilling Project Leg 76. Bull. geol. Soc. Am., 93, 876-85. SHIMKUS, K.M. & TRIMONIS, E.S. 1974. Modern sedimentation in the Black Sea. In: Degens, E.T. & Ross, D.A. (eds), The Black Sea: Geology, Chemistry, and Biology. Am. Ass. Petrol. Geol. Mem. 20, 249-78. SHOR, A.N. & POORE, R.Z. 1978. Bottom currents and ice rafting in the North Atlantic: interpretation of Neogene depositional environments of Leg 49 cones. In: Luydendyk, B.P., Cann, J.R. et al. Init. Repts. DSDP, 49, US Govt. Print. Off., Washington, DC. SIESSER, W.G. 1980. Late Miocene origin of the Benguela upwelling system off northern Namibia. Science, 208, 183-285. SOUTAR. A., JOHNSON, S.R. & BAUMGARTNER, T.R. 1981. In search of modern analogs to the Monterey Formation. In: Garrison, R.E., Douglas, R.G., Pisciotto, K.E., Isaacs, C.M. & Ingle, J.C. (eds.), The Monterey Formation and Related Siliceous Rocks of California. Soc. econ. Paleo. Min., Pacific Section, Spec. Pub., 123-48. SOUTHAM, J.R., PETERSON,W.H. & BRASS,G.W. 1982. Dynamics of anoxia. Palaeogeogr., Palaeoclimatol., Palaeoecol., 40, 183-98.
Deposition of organic-carbon-rich sediment
-
Sxow, D.A.V. 1984. Turbidite facies, associations and sequences in the Angola Basin. In: Hay, W.W., Sibuet, J.-C. et al., Init. Repts. DSDP, 75. US Govt. Print. Off., Washington, DC., (in press). & DEAN, W.E. 1984. Middle Cretaceous Black Shales at Site 530 in the southeastern Angola Basin. In: Hay, W.W., Sibuet, J.-C. et al., lnit. Repts. DSDP, 75. US Govt. Print. Off., Washington, DC., (in press). SUESS, E. 1980. Particulate organic carbon flux in the oceans-Surface productivity and oxygen utilization. Nature, 288, 260-3. SUMMERHAYES,C.P. 1981a. Oceanographic controls on organic matter in the Miocene Monterey Formation, offshore California. In: Garrison, R.E., Douglas, R.G., Pisciotto, K.E., Isaacs, C.M. & Ingle, J.C. (eds), The Monterey Formation and Related Siliceous Rocks. Pacific Section, Soc. econ. Paleo. Min. Spec. Pub., 213-9. 1981b. Organic facies of middle Cretaceous black shales in the deep North Atlantic. Bull. Am. Ass. Petrol. Geol., 65, 2364-80. --, BORNHOLD, B.D. & EMBLEY, R.W. 1979. Surficial slides and slumps on the continental slope and rise off South West Africa: A reconnaissance study. Marine Geol., 31,265-77. --, MASRAN, P. 1983. Organic facies of Cretaceous and Jurassic sediments from DSDP Site 534 in The Blake Bahama Basin, Western North Atlantic. In: Sheridan, R.E., Gradstein, F. et al., Init. Repts. DSDP, 76. US Govt. Print. Off., Washington, DC., 469-81. THIEDE, J., DEAN, W.E. & CLAYPOOL, G.E. 1982. Oxygen deficient depositional paleoenvironments in the mid-Cretaceous tropical and subtropical central Pacific Ocean. In: Schlanger, S.O. & Cita, M.B. (eds), Nature and Origin of Cretaceous Organic Carbon-Rich Facies. Academic Press, London. 55-78. -t~ VAN ANDEL, T.H. 1977. The paleoenvironment of anaerobic sediments in the late Mesozoic South Atlantic Ocean. Earth planet. Sci. Lett., 33, 301-9. THIERSTEIN, H.R. 1979. Paleoceanographic implications of organic carbon and carbonate distribution in Mesozoic deep sea sediments. In: Talwani, M., Hay, W.W. & Ryan, W.B.F. (eds), Deep Drilling Results in the Atlantic Ocean." Continental Margins and Paleoenvironment. Maurice Ewing Series 3, Am. Geophys Union, Washington, DC. 249-74. & BERGER, W.H. 1978. Injection events in Earth history. Nature, 276, 461-6. THUNELL, R.C., WILLIAMS, D.F. KENNETT, J.P. 1977. Late Quaternary paleoclimatology stratigraphy and sapropel history in eastern Mediterranean deep-sea sediments. Mar. Micropaleont., 2, 371-88. TISSOT, B. 1979. Effects on prolific petroleum source rocks and major coal deposits caused by sealevel changes. Nature, 277, 463-5. , DEMAISON, G., MASSON, P., DELTEIL, J.R. & COMBAZ,A. 1980. Paleoenvironment and petroleum potential of middle Cretaceous black shales in -
-
-
-
-
559
Atlantic basins. Bull. Am. Ass. Petrol. Geol., 64, 2051-63. - - , DEROO,G. & HERBIN,J.P. 1979. Organic matter in Cretaceous sediments of the North Atlantic: contributions to sedimentology and paleogeography. In: Talwani, M., Hay, W.W. & Ryan, W.B.F. (eds), Deep Drilling in the Atlantic Ocean: Continental Margins and Paleoenvironment. Maurice Ewing Series 3, Am. Geophys. Union, Washington, DC. 362-74. --, DURAND, B., ESPITALII~ • COMBAZ, A. 1974. Influence of nature and diagenesis of organic matter in formation of petroleum. Bull. Am. Ass. Petrol. Geol., 58, 499-506. & WELTE, D.H. 1978. Petroleum Formation and Occurrences. Springer-Verlag, New York. 538 PP. TOTH, D.J. & LERMAN,A. 1977. Organic matter reactivity and sedimentation rates in the ocean. Am. J. Sci., 277, 465-85. TtJCHOLKE, B.E. & VOGT, P.R. 1979. Western North Atlantic: sedimentary evolution and aspects of tectonic history. In: Tucholke, B.E., Vogt, P.R. et al., Init. Repts. DSDP, 43. US Govt. Print. Off., Washington, DC. 791-825. VON DER DICK, H., RULLKOTTER, J. & WELTE, D.H. 1983. Content, type, and thermal evolution of organic matter in sediments from the eastern Falkland Plateau, Deep Sea Drilling Project, Leg 71, In: Ludwig, W.J., Krasheninnikov, V. et al., Init. Repts. DSDP, 71. US Govt. Print. Off., Washington, DC. 1015-1132. VON RAD, U., CEPEK, P., YON STACKELBERG, U., WISSMAN, G. & ZOBEL, B. 1979. Cretaceous and Tertiary sediments from the northwest African slope (dredges and cores supplementing DSDP results): Marine Geol., 29, 213-312. WEISSERT, H. 1981. The environment of deposition of black shales in the Early Cretaceous: an ongoing controversy. In: Warme, J.E., Douglas, R.G. & Winterer, E.L. (eds), The Deep Sea Drilling Project: a Decade of Progress. Soc. econ. Paleo. Min. Spec. Pub. 32, 547-60. - - , MCKENZIE,J. & HOCHULI,P. 1979. Cyclic anoxic events in the Early Cretaceous Tethys Ocean. Geology, 7, 147-51. WELTE, D.H., CORNFORD,G. & RULLKOTTER,J. 1979. Hydrocarbon source rocks in deep sea sediments. Proceedings Offshore Technology Conference, Houston. 457-61. WIEDMANN, J., BUTT, A. d~. EINSELE, G. 1978. Vergleich von Marokkanischen Kreide-Kustenuafschlussen und Tiefsee bohrungen (DSDP): Stratigraphie, Paleoenvironment and Subsidenz, an einem passiven Kontinentalrand. Geologisches Rundschau, 67, 454-508. WILDE, P. & BERRY,W.B.N. 1982. Progressive ventilation of the oceans-potential for return to anoxic conditions in the post-Paleozoic. In: Schlanger, S.O. & Cita, M.B. (eds), Nature and Origin of Cretaceous Carbon-Rich Facies. Academic Press, London. 209-24. WILLIAMS,D.F., THUNELL,R.C. & KENNETT,J.P. 1978. Periodic fresh-water flooding and stagnation of the
560
M.A. Arthur, W.E. Dean and D.A.V. Stow
eastern Mediterranean Sea during the late Quaternary. Science, 201, 252--4.
WYRTKI,K. 1962. The oxygen minima in relation to ocean circulation. Deep Sea Res., 9, 11-23.
M.A. ARTHUR1, Department of Geology, University of South Carolina, Columbia, South Carolina 29208, USA. W.E. DENY,US Geological Survey, Denver, Colorado, 80225, USA. D.A.V. STOW, Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh, Scotland EH9 3JW. 1 Present address: Graduate School of Oceanography University of Rhode Island, Narragansett, R.I. 02882, USA.
Plasticity and compaction characteristics of the Quaternary sediments penetrated on the Guatemalan Transect DSDP Leg 67 R.W. Faas S U M MARY: The plasticity characteristics of the Quaternary sediments of the Guatemalan continental margin were determined from five sites drilled on Leg 67 of the Deep Sea Drilling Project. Sixty-four samples were analysed from various marine environments, including the Cocos oceanic plate, Middle America Trench, and the trench lower slope to mid-slope of the Guatemalan continental slope. Quaternary sediments on the Guatemalan continental slope are generally mixed siliceous and calcareous hemipelagic silty-clays and clayey-silts, frequently interbedded with volcanic ash. In the Middle America Trench, similar sediments alternate with sandy turbidites. Siliceous hemipelagic silty-clays and clayey-silts form the uppermost 170 metres of the Cocos Plate. The sediments are generally classified as 'organic clays of medium to high plasticity, containing micaceous sands and silts,' with 14~oclassed as 'inorganic clays of medium to high plasticity.' Plasticity (measured by the Plasticity Index, It,) decreases downslope to the trench, then increases sharply in the Cocos Plate sediments. Organic matter is greatest highest on the slope (Site 496) and decreases seaward, averaging 1.14~o at Site 495 on the Cocos Plate. Sediment compaction varies, according to position along the traverse and to the dominant sediment type. Greatest compaction (44.5~) occurs at Site 494 at the base of the lower trench slope. Least compaction (18.6%) is found at Site 495 on the Cocos Plate. Compaction of the trench and mid-slope sediments ranges between 31-38.5%. High sedimentation rates in Quaternary sediments are due, in part, to sediment gravity-flows which depend upon rheological properties, i.e. sediment plasticity. Mudflows and cohesive debris-flows appear to be significant downslope transport mechanisms in these highly plastic sediments. The goals of the Guatemalan Transect were to study a convergent margin where accretion and imbrication may have been continuous during most of Tertiary time; to strengthen the tie between onshore and offshore geology; and to study modern sediment sequences in the Middle America Trench and continental slope. The objective of this research was to evaluate the usefulness of the Atterberg limits of the Quaternary sediments in predicting modes of downslope subaqueous sediment transport; and to determine the gravitational compaction characteristics of the sediments accumulating on the Guatemalan continental slope, Middle America Trench, and on the Cocos oceanic plate. (Fig. 1). Gravitational compaction is defined as the expulsion of pore fluids and pore volume thereby decreases in a sedimentary column as a result of normal and shear-compressional stresses due to the overburden load (Riecke I I I & Chilingarian 1974).
Methods Quaternary sediments---general The Quaternary sediments of the Guatemalan continental margin form a seaward thinning
mantle of hemipelagic muds, containing significant quantities of siliceous and calcareous fossil remains. Sediment thickness varies from greater than 200 m (Site 4 9 6 ~ u p p e r trench slope), to approximately 100 m (Site 498--lower trench slope). Within the Middle America Trench, sediments are interbedded pelagic diatomaceous muds and fine sand, and muddy turbidites. Sediment thickness varies, ranging from about 80 m (Site 500) to 120 m (Site 499). On the oceanic plate, diatomaceous hemipelagic muds thin to about 60 m at Site 495. Sedimentation rates vary from 300 m/my in the trench (Site 499) to 37 m/my on the oceanic plate (Site 495). The greatest accumulation occurs in the trench as sediment is transferred to it from higher elevations on the slope. Dominant clay minerals are illite, montmorillonite, and kaolinite; the latter two are most abundant. Illite is missing from Site 495 on the ocean plate. As more interest is focused on active continental margins and their deposits, it is obvious that many are areas of downslope sediment transport. Such transport may be rapid and episodic, slow and pervasive, or some combination of the two. Lowe (1979) has indicated the importance of rheology in distinguishing the mass transport
563
564
R. W. Faas
~,~
........... ~,..-i-':i;..
91~oo' '.. .'
9 '
9 .
9
. .'.
9o~oo' 9
San
/,
.
.
~
.
. .
.
I
/
..,. ~
' '.'.::::!....;
""'"--
I I
,----
....
I /
.
/"
,,'" I
Jose'
,..
-
"'"",i7.'.;'.':....
"
" ' ' : ' " : " {.J..,7;;:.7.-..!:.... ;...
I
I I
\ \
N I I /
\\
I I I
I
/
13
:.~...'....
\
~ 00'
0o 03
Middle
America
Trench
Axis
I~ 03
~D 03
FIG. 1. Location map of Guatemalan continental margin and location of all D S D P Leg 67 drill sites.
565
Plasticity and compaction characteristics mechanisms as fluidal flows or debris-flows by their fluid and plastic behaviour respectively. If this concept is valid, the initial step in the study of submarine mass transport mechanisms should be the determination of the expected rheological behaviour of the sediments. This can be readily accomplished by analysing the sediments for their Atterberg limits.
Atterberg limits Atterberg limits (Casagrande 1948) are based on the concept that a sediment is a two-phase system of inorganic particles and water. The rheological behaviour state of such a mixture changes from fluid to plastic to solid as the mixture loses water (Fig. 2). The water content at which a sediment/ water mixture ceases to behave as a liquid is its liquid limit (Wc). Similarly, plastic behaviour of such a system ceases at its plastic limit (Wp) and it passes through a semi-solid state, determined by its shrinkage limit (ws). Each limit is the empirical measure of the water content of the mixture at that point where its behaviour changes from one state to another. The plasticity index (Ix) is a measure of the range of water content through which the mixture exhibits plastic behaviour
Analytical procedure The primary source of data for this study was provided by wet bulk density profiles of the cores retrieved during Leg 67 drilling. Wet bulk density is measured continuously on all cores, using the Gamma Ray Attenuation Porosity Evaluator (GRAPE) (Evans & Cotterell 1970). Each core is slowly passed through the GRAPE and its wet bulk density is recorded. Prior to each analysis, the GRAPE is calibrated to a distilled water (1.00 Mg/m 3) and aluminum (2.60 Mg/m 3) standard. A description of the GRAPE method and results has been presented by Boyce (1975). Atterberg limit tests were performed on board the Glomar Challenger immediately after the cores were sampled. A sediment sample was placed in a preweighed aluminum moisture dish, weighed on a triple-beam balance, and placed in the drying oven at 110~ for drying overnight. At the same time, liquid limits were determined. Half of the sample was used to determine its water content; the other half was used for the plastic limit analysis. Liquid limits were performed according to the simplified procedure described by Lambe (1951) and calculated as follows
Z~=wL-w,,
0.121
w
N
CONSI5TENCY 5 TA TE5 LIQUID wL Liquid limit
PLASTIC
Wp Plas'fic limit SEMI-SOLID wS Shrinkage limit
SOLI O
FIG. 2. Sediment consistency states and relation to Atterberg limits.
where wN=the water content of the sediment sample which closes the groove in N blows in the standard liquid limit cup. Values of liquid limit (Table 1) represent an average of three separate determinations of the same sample. The plastic limit is defined as the water content of the remoulded sediment as it passes from a plastic to a brittle condition. The test involves rolling the sediment over a frosted glass plate and forming a thread 1/16" in diameter. When the thread no longer holds together, the plastic limit has been reached, and the water content is determined. The procedure is fully discussed in Lambe (1951). Samples were also taken for organic carbon analysis and particle size distribution. Some of the carbon analyses were performed on ship, others were done later at the University of Oklahoma. All were analysed in a Leco combustion apparatus. Grain-size analysis was carried out at the DSDP laboratory in La Jolla, California. Shear strength was measured perpendicular to the bedding at one-metre intervals, on the split core with the hand-held CL-600 Torvane. The Torvane was rotated at a rate designed to reach failure in about 10 seconds with constant loading.
566
R. W. Faas
.+~
.~o 0 0 0 0 0 0 0 0
9,,,.+
-
.~._~..~._~
~ ~
0 0 0 0
000000000
~ .=
-~+.~.~
~
oo
.-
o
"-d "-d 9,.,.+ .,-~
.,--~
mm
o+-~i---,+~
',R.
',,R. ~.
t,~ ,~o ur ~0 tr ,~-
--.~
...+ .,,.+
& d ~ d d d d d
d
d ,--: d d d ,--: ,--: d d
d
++,..,+
..,.,+ P,,,,.
<
~:::b22212--
~ ,,4 t-z -.: o6 ,-: ,H ,d t-z ~,d ,--: ~
~
I
~
+
I
I
I t l
I
I
II
+
Plasticity and compaction characteristics
~
I
I ~
~
I ~
~
~
~
.~.~ ~
~
~
~
0
I~-~
0
0
0
0
.~.~.~.~ <
~ ~
<
~
~ .
ug ',~-
eq
-
~
~
0
0
0
~
.
. . . . .
~.~ .~ .~ .~ .~
~
I ~ -~
~ ~
~ <
<<
- ~
~
0000000000000000
567
-
~
~
-
-
-
-
~
.~ ~.~ .~
~
~
~
-
-
2 2 d ~ d d d d d d d d d
dd
~
m <
0
I
I
I /
I
I
I
I
I l [
I
I
I
I
I
~
I
I
I /
I / /
I /
I
I |
I
568
R. W. Faas
0 0 0 0 0 0 0 0 0 0 0 ~
i~-~
Z<<
~.~ 9
~,~ ~
~ ~ ~~
.~,~.~
, ~ ~
d M ~ d d M
d d ~
~d
g
0
,.,..-4 ,.m
=
~:
(:~
E~ ~
x x
~..o=
~, = = ~ ~ >-,"=~'~__, -~ "~-ca "~ ~ .--'- ,. ~ c~ ~'B ~'.n.. >
~-~ ~ ~ ~ , _ _ , ~
~ ~ :.-~ I ~ e ' ~.,-. ~ II II II <
I I I
I I
I
~
I [
I
I I
e ~ - = . <~< ; o o [/3
Plasticity and compaction characteristics
Plasticity and compaction Plasticity characteristics By determining the Atterberg limits, a sediment can be classified into various degrees of high, medium, and low plasticity. This classification (Casagrande 1948) utilizes the relationship between the liquid limit (wL) and the plasticity index (Ip) of a sediment to specify its position within the total range of plastic behaviour. The A-line (Fig. 3) is an empirical boundary between inorganic clays (above the line) and organic clays and inorganic silts (below the line). Figure 3 shows the position of each sample analysed on the Casagrande Plasticity Chart. Table 1 indicates that the majority are classified as 'organic clays of medium to high plasticity, containing micaceous sands and silts,' (OH soils as defined by the Unified Soil Classification (Wagner 1957)). Those defined as 'inorganic clays of medium to high plasticity' are primarily from Site 497, and Table 1 indicates that these sediments comprise a layer extending to a depth of 40 m.
Gravitational compaction characteristics Gravitational compaction of a sediment occurs due to closer particle packing, reduction in pore volume, and increase in bulk density. Hamilton (1976) presented theoretical relationships of bulk density and porosity with depth for various types
569
of normally compacting marine sediments and developed equations to predict values of bulk density and porosity at selected depths. It should be possible to estimate the qualitative state of compaction by plotting any of these parameters against depth. Undercompacted sediments should show lower values of bulk density and greater pore volume than predicted, with the reverse being observed in overcompacted sediments. Inasmuch as changes in volume and density relationships ultimately depend upon overburden pressure, graphs relating overburden pressure to depth in the sediment column should also reveal compaction trends. Zones of under-or overcompacted sediments should deviate from the trend established for a normally compacting lithologically similar sediment, with undercompaction indicated where the overburden pressure plots below the normal compaction line and over-compaction indicated where overburden pressure falls above the normal compaction line. A normally compacting sediment is one in which the rate of drainage of interstitial water equals the rate of sediment accumulation, and overburden pressure is only a function of the sediment skeleton--the lithostatic load of Bishop (1979). Theoretical compaction curves for the commonly-occurring marine sediments (Fig. 4) were constructed from Hamilton's data (1960; 1976). Overburden pressure was calculated according to the following:
140 Inorganic
X UJ a zm
of medium
/o
100
~>--
8O
0m
60
F-CO <~ _J Ct.
clays
to high plasticity
120
o/~-
o
~o
g~l
0
495
4o Z~
20--
,J, 20
40
Organic clays of medium to high plasticity and micaceous or
9 496 [3
497
,~
499
d i a t o m a c e o u s fine sands and silts
I
u
!
I
I
I
I
60
80
100
120
140
160
180
1 200
LIQUID LIMIT FIG. 3. Casagrande Plasticity Chart showing plasticity characteristics of all samples analysed.
570
R. W. Faas (kPa) 196.13 (kg/cm 2) 2.0
397.16 4.0
Overburden
Pressure
588.28 6.0 r
784.37 8.0 1
980.46 10.0 f
1176.56 12.0 i
1372.93 14.0 '"1
100
A 150 vE J=
==
r E 200 P e l a g ~
,~ 25o
300
350
F]c. 4. Theoretical compaction curves for normally compacting common marine sediments.
Dial
I
400
Overburden pressure (a~,) = (P,,.s - p,,, ) d where pws = wet bulk density of the sediment (Mg/m 3) Ps,, = sea water density (assume 1.025 Mg/m 3) d = depth increment (m) The curves show markedly different profiles, all following the same general pattern. Each shows an increasing response to overburden pressure with the water-rich biogenic ooze showing lesser increases, probably due to a change in behaviour between clay-rich and granular material. After an initial nearly linear steep slope to about 100 m, the curves become concave upward as overburden pressure increases with depth. The compaction curves for terrigenous sediment and diatomaceous oozes represent the two extremes of expected marine sediment types. All other types (e.g. diatomaceous muds) lie within the limiting envelope as defined by the upper and lower curves.
Data analysis In an earlier report (Faas 1982), each site drilled on Leg 67 was analysed to determine its compaction characteristics, according to the method presented above. Inasmuch as this paper is primarily concerned with the presentation of a
I
technique rather than a geologic synthesis, only Site 499 (Middle America Trench) will be described in detail. However, a generalized summary of compaction states of all Leg 67 sites will also be presented, each having been analysed in the following fashion. Site 499
The upper Quaternary consists of trench-filling sediments, primarily biogenic muds and turbidite sands. The lower sediments are hemipdagic diatomaceous muds to 200 m. Chalks occur below 200 m in lower Pliocene sediments. The compaction curve (Fig. 5) follows that of a normallycompacting diatom ooze to 35 m, then turns sharply upward and parallels the curve of a terrestrial sediment to 150 m. Below this depth, the compaction curve suggests a change in lithology to pelagic sediments, beginning in the lower Quaternary and extending into the lower Miocene. Figure 6 indicates a unit of low bulk density (diatom mud) extends from the mudline to 33 m. The turbidites (33 m to 125 m) contain shallowwater fossils and wood fragments. Geophysical records show localized regions where reflections pass laterally into diffuse zones, indicating an abundance of gas in the sediments. Methane
Plasticity and compaction characteristics
571
Hole 4 9 9 Lithology
Age
Overburden Pressure (kPa) 200
=6
400
600
800
1000
1200
l
I
I
g=
1400
upper =~==== Quaternary g'= o~
==6
E
o-E
m
ee e
~176
lower Quaternary upper Pliocene
T
i
==~==o
"a,
lower Miocene
300
I
I
I
I
"tO
I
""~ p
I
FIG. 5. Compaction curve--Site 499 content of the sediments is quite high (65-75%--shipboard analyses, 1979), suggesting that microbial decomposition of organic matter is responsible for their gassy nature. GRAPE density shows unusual variability, shear strength is low (10 15 kPa), and the interval is believed to be composed of alternating over-and undercompacted (turbidite-hemipelagic) sedimentary units. The hemipelagic diatomaceous muds (125-200 m) represent a change toward lower density sediments (Fig. 5) and the GRAPE density profile (Fig. 6(a)). Water content increases significantly through the interval (Fig. 6(b)), but shear strength remains relatively constant at 30 kPa (Fig. 6(c)). Lithified lower Miocene chalks occur at 200 m and show an abrupt increase in GRAPE density (Fig. 6(a)) and a decrease in water content (Fig. 6(b)). The interval between 125 m and 200 m, confined between units of greater density above and below, is interpreted to be undercompacted.
General compaction patterns Figure 7 is a composite compaction diagram of all the sites drilled on Leg 67. The theoretical compaction curves for terrigenous sediment and diatomaceous ooze are plotted and form a limit-
ing envelope (T and D) within which all individual compaction curves are located. Sites 494, 500, and 499 lie toward the terrigenous curve and indicate general overcompaction, with Site 494 showing the greatest compaction. Site 497 appears generally normally-compacted, lying between the terrigenous and diatomaceous ooze curves. Site 496 lies toward the diatomaceous ooze curve and appears generally undercompacted. Both sites are located on the Guatemalan Continental Slope. Site 495, on the Cocos Plate, shows various forms of compaction behaviour.
Decompaction analysis If conclusions drawn from the preceding data are correct, they should be substantiated by a 'decompaction' analysis. This is performed by expanding the sedimentary section back to its original depositional void ratio (void ratio, e, is defined as the volume of the voids divided by the volume of the solids). Hamilton (1976) presents the analysis in detail and suggests the proper initial void ratio (ei) to use for the commonly occurring marine sediments. If the sediment was a mixture of two pure sediment types (e.g. hemipelagic diatomaceous mud), the initial void ratio
R. W. Faas
57 2
Hole 499 Core o
Water Content (% dry w t . ) 60 ,.
- - i
80 i
100 ,
Shear Strength (kPa)
GRAPE Density (Mg/m 3)
120 140 i 9 ,
1.3
1.4 ,
.,
2 I
1.5 ,
1.6 ,
1.7 ,
o
lO
20 i
3O i
60 70
5O
40 i
i
9
3
e e
9
4
I I
9
5
-II
9
5i
.
-7i
"
9 Average value/core
I I
A Maximum value/core
I
i 9
!): I
I I I
8 --m
9
9
11 i
v ~100 r E
I 12
13" 14 I
""
I
15"
\
16
\ \
150 /
/
/
18 i
\
1-;n
I I I I
-|
I
,Z
22
200
23 i
\
9e
24 9
9 A
B
C
FI6.6. Mass physical properties--site 499. (a) Bulk Density (b) Water Content (c) Shear Strength
selected was the average of the two void ratios for each type. Sandy turbidites were considered as terrigenous sediments with an initial void ratio of 2.57 (Table 1). Decompaction was done in two ways. In the first, variable intervals were decompacted following lithologic changes as indicated in Initial Core Descriptions for Leg 67. For example, in Site 496, the first 17 m of diatomaceous mud was decompacted, followed by 47 m of hemipelagic mud. In this way, total compaction and compaction per lithologic unit for each site can be determined.
Table 2 shows the effect of lithology on compaction. Both 'terrigenous sandy mud' and 'diatom ooze' show maximum compaction, while 'turbidites', 'pelagic clay', and 'calcareous sediment' compact to a lesser degree. In the second, each 9.5 m core was decompacted using the average measured void ratio (ep) for the core (in some cases, as many as six measurements formed the average; in other cores, only a single measurement was made). Where gaps due to poor recovery occurred, the average measured void ratio for the previous core was
Plasticity and compaction characterh3tics
573
Overburden Pressure (kPa) 200
400
600
I
I
800
1000
1200
1400
!
i
i
.I
1800
1600 !
50
100
E
150i
gQ E
200
S 494
250
3OO
497
,%.
350
195
495
1
1
I
I
D
I
1
FIG. 7. Generalized compaction diagram for all Leg 67 sites. extrapolated through the missing interval. Initial void ratios were chosen which seemed to best correspond to the generalized unit stratigraphy and were maintained throughout the unit. This approach de-emphasized specific lithologic control of compaction but provided an overall view of total site compaction. Table 3 presents the
compaction values determined for each site by both techniques.
Discussion Interpretation
of decompaction
analyses
The analysis revealed three general compaction
TABLE 2. Decompaction values for lithologies at each Site, Leg 67
Sediment Type
494
Terrigenous sandy mud
66
495
496
497
Average % 500 Compaction
53.2 30.5
49.9
11.3 18.9 20.8 17.3 12.8 20.2 47.6 19.8
Calcareous sediment (nanno) Pelagic clay
41.1 13.5 15.7 30.2 45.5 19.7
Diatom ooze
57
Turbidites
499
21.2 27.6
35.9 18.8 54.0 57.0 49.1 32.7 36.8
43.4
31.1 21.5
26.3
574
R. W. Faas
patterns among all the sites (Fig. 8). Sites 496 and 497 on the continental slope compacted between 31-38%, with greater compaction occurring at Site 496, located close to the head of the San Jose sea valley. Site 495, on the oceanic plate, showed least compaction (11-18.6%, depending upon technique), attributable to low sediment loading rates. Greatest compaction was observed at Site 494, located at the base of the trench slope. The column was composed of diatomaceous muds, terrigenous sandy mud and mudstones, and a basal pelagic clay. Both trench sites (499 and 500) showed significant but not unusual compaction. Both are characterized by a thick turbidite sequence interbedded with hemipelagic sediments. Decompaction according to specific lithologies is difficult, but indicated similar behaviour at the two sites, with Site 499 (43~o compaction) comparing closely to Site 494 on the inner trench slope. Decompaction analysis generally followed the compaction trends shown in Fig. 7, with a few exceptions. In Fig. 7, Site 496 appears least compacted. Decompaction analysis shows it to be 38% compacted and Site 495 the least compacted (11-18.6%). Site 494 still indicates overcompaction (44.5%).
Implications of plasticity for sediment transport and compaction The Quaternary sediments at Sites 496 and 497, located on the Guatemalan continental slope are characterized by high plasticity indices which decrease toward the Middle America Trench, (Fig. 9). This signifies a trenchward decrease in plasticity, likely attributable to increasing amounts of sandy material, i.e. sandy turbidites, probably transported down the San Jose submarine canyon for redistribution in the trench. The low plasticity of Site 499 trench-fill sediments is clearly seen in Fig. 3. Hein & Gorsline (1981) show that slump
deposits, mudflows, thick and thin turbidites, and liquified flows along the California Borderland characteristically possess high plasticity indices and vary from unstable to moderately unstable. Field (1981) shows an increase in the number of sediment-flow deposits, the internal mechanical behaviour of which are plastic, occurring along the California Borderland in areas where sedimentation rates are high. Because of the high plasticity indices of the continental slope sediments (Sites 496 and 497), they should be particularly effective in generating mudflows, cohesive debris-flows, or cohesive flows, either through failure due to excessive sediment loading or, more likely, through tectonic activity. Site 494, located on the inner wall of the lower trench slope, provides ample evidence of downslope mass-movement in the Pleistocene and Quaternary sediments. The upper 130 m contain abundant shallow water foraminifera and plant debris which, due to their unsorted nature, were probably transported by mudflows or cohesive debris-flows rather than by turbidites. The deposits also include micritic limestone pebbles, fragments of pelecypod valves, a rounded cobble of coarse-grained quartz diorite, and abundant Miocene and Eocene siltstone clasts. The plasticity index increases at Site 495 on the Cocos Plate where it reaches extraordinary values (Fig. 9). A greater hemipelagic element dominates these sediments and the increase may be due to this fact. However, organic matter content is low and this usually is reflected in lower liquid and plastic limits. Seed et al. (1964) have shown that the addition of illite to a montmorillonite/ kaolinite mixture lowers the plasticity index of the mixture through the tendency ofmontmorillonite to form interstratified or compound particles with illite particles. The mixture would take on the plasticity characteristics of an illite/kaolinite mixture, which has lower plasticity than a montmorillonite/kaolinite mixture. Inasmuch as illite has not been identified in the Quaternary sediments at
TABLE 3. Percent decompaction for both techniques--Leg 67 drill Sites Technique Site 496 497 499A 500 495 494
Specific General 38.5 31.0 35.5 37.7 18.6 44.5
37.8 31.0 43.0 34.0 11.0 44.0
Location Continental Slope Continental Slope Middle America Trench Middle America Trench Cocos Oceanic Plate Inner Trench Wall
Plasticity and compaction characteristics
pntu sno~.wol.,a
I
575
~uo.pnw pue pnlN
i Ael::)J
~t
I
pntu ,noeolulo),eI.Q
I
:g
,|
euo),$pnm pue PnlN plOI
>lleqopue ezooe~=uoq,eoJ
pnw sno~oewo)e!a
/
•l••qo
@
.=-
II
II
~. o
azoo eieuoq~e~
pn~ snoeoemme!a
IIIt-qOUai/ }
~lleq0
8
c~
r
to
s~.uatu!pesII!pqoueJ/
_1
I
pnm o!6elad!UJaH o
sJ.uelu!pesiI!}-qoueJ.L
lPnw o,,elad!'",H I
c~ o I
pnw i!.otouueu pue ,,,m.e!Q
I o~
pues pue pnlN
i
.n,,,,.o,ouu.u.u.,-o.,o
O
pn,,, I!SSojouueupue ,-m.e!a
]
pnm ^pu.s
~
~
I
r
....
. , . I pmu ApueS
pmu l!SSOjOUUeUpue wole!Q I
I
I
I
I
. i
~
1
(w) ql.deO mo~oq-qn$
l
I
I
I
576 :s 9
4
R. W. Faas
3"0 2-5 2"0 1"5 1.0 0-5
9 ---..-.._. 9 ,-...._.__
I00
.o
80
0
9
!
|
/.7
I 9
/J
i
/
/o
II
/
60
$ (66"4)
S/j
9
(81.6)
9 (68.1)
40
(57"7)
20
(58-5)
I 496
1 494 + 494A
I 497
Slope
Inner trench wall
I 499 4500
I 495
Trench axis
Ocean plate
F16.9. Plasticity index and Total Organic Content (TOC) profile along the Guatemalan transect. Site 495, its elevated plasticity may reflect this lack of illite. This, plus the fact that Site 495 averages almost 60% clay-sized particles (Table 1) may be the reason why Site 495 sediments possess such high plasticity. Site 495 is located on a horst block approximately 2000 m above the floor of the Middle America Trench. It shows overcompaction in the upper 50 m and undercompaction from 50 m to 200 m. This may reflect a direct lithologic control inasmuch as Pleistocene hemipelagic silty muds extend from the mudline to 57 m sub-bottom, followed by Pliocene and lower middle Miocene water-rich diatomaceous silty mud to 180 m sub-bottom.
Summary and conclusions The study of the physical properties of sediment cores obtained by rotary drilling techniques is always hampered by the obvious problems associated with core disturbance, particularly in the younger, unlithified sedimentary section. H o w -
ever, it is possible to obtain useful information about depositional processes through the use of Atterberg limits which yield information on the rheological characteristics of the sediments while severely disturbed (remoulded). It is also possible to evaluate the various states of compaction in these disturbed sediments. A technique which utilizes the relationship between overburden pressure versus depth for actual marine sediments and compares the compaction curves with theoretical compaction curves for various types of normally compacting, commonly occurring marine sediments has proven to be useful in assessing the compaction states of the sediments drilled on the Guatemalan Transect on DSDP Leg 67. This evaluation is qualitative and should not be considered equivalent to consolidation tests performed on undisturbed cores of high quality. Continental slope sediments show the same degree of compaction as those in the Middle America Trench. However, they tend to follow a compaction curve intermediate between pelagic clays and diatom ooze. This pattern may arise from the presence of bacterially generated methane, or from decomposition of a gas hydrate zone
Plasticity and compaction characteristics of unknown thickness which occurs at about 300 m (von Huene et al. 1980). The overall effect is to reduce the lithostatic load in the upper 300 m and create zones of undercompaction in the relatively impermeable fine-grained sediments. The trench sediments follow a compaction curve intermediate between pelagic clays and terrigenous sediments. This results from the mixing of pelagic sediments with terrestrially-derived turbidites. The compaction behaviour of the Cocos Plate sediments is controlled by sedimentation rate, sediment type, and tectonic events. Slowly accumulating hemipelagic sediments appear overcompacted. The water-rich siliceous oozes below the impermeable hemipelagic clays show relative undercompaction due to impeded drainage. Plasticity of the Quaternary sediments decreases from the upper continental slope to the lower trench slope and reaches its lowest value in the trench-fill turbidites within the Middle America Trench. Plasticity increases significantly in the
577
hemipelagic sediments of the oceanic plate. Sedimentation within the trench is high (300 m/my) and indicates that significant downslope massmovement is occurring. The high plasticity of the upper slope and ocean plate sediments suggests that the dominant processes are mudflows and cohesive debris flows as these processes depend upon plastic internal mechanical behaviour and can move long distances over very low angle slopes. ACKNOWLEDGEMENTS: I thank Bill Harrison and Joe Curiale for providing organic matter analyses of many of the samples. Grain-size and additional organic matter analyses were done at the DSDP Laboratory in La Jolla, California. Various drafts of the manuscript were reviewed by Richard Bennett, Kate Moran, Jim Booth and Paul Carlson and I have benefited from their comments. Finally, I thank the National Science Foundation for its continued support of the Deep Sea Drilling Program.
References BISHOP, R.S. 1979. Calculated compaction states of thick abnormally pressured shales. Bull. Am. Ass. Petrol. Geol., 63, 918-33. BOYCE, R.E. 1975. Definitions and laboratory techniques of compressional sound velocity parameters and wet water content, wet-bulk density, and porosity parameters by gravimetric and gamma ray attenuation techniques. In: S.O. Schlanger & E.D. Jackson et al., Init. Repts. DSDP, 33. US Govt. Print. Off., Washington, DC. CASAGRANDE,A. 1948. Classification and identification of soils. Trans. Am. Soc. cir. Eng., 113, 901-31. EVANS, H.B. & COTTERELL, C.H. 1970. Gamma ray attenuation density scanner. In: M.N.A. Peterson et al., Init. Repts. DSDP, 2. US Govt. Print. Off., Washington, DC. FAAS, R.W. 1982. Gravitational compaction patterns determined from sediment cores on the Guatemalan Transect: Continental slope, Middle America Trench and Cocos Plate, on D.S.D.P. Leg 67. In: J. Aubouin., R. von Huene, et al., Init. Repts. DSDP, 67, US Govt. Print. Off., Washington, DC. 617-38. FIELD, M.E. 1981. Sediment mass-transport in basins: Controls and patterns. In: R.G. Douglas, Colburn, I.P. & Gorsline D.S. (eds), Depositional System of Active Continental Margin Basins. Pacific Section, Soc. econ. Paleo. Min., Short course notes, San Francisco. 61-83. HAMILTON,E.L. 1960. Ocean basin ages and amounts of
original sediments. J. sed. Petrol., 30, 3, 370-9. 1976. Variations of density and porosity with depth in deep sea sediments. J. sed. Petrol., 46, 2, 280-300. HEIN, F.J. & GORSLINE,D.S. 1981. Geotechnical aspects of fine-grained mass flow deposits: California Continental Borderland. Geo-Marine Letters, 1, 1-5. LAMBE, T.W. 1951. Soil Testing for Engineers. John Wiley, New York. LOWE,D.R. 1979. Sediment gravity flows: their classification and some problems of application to natural flows and deposits. Soc. econ. Paleo. Min. Spec. Pub. 27, 75-82. RIEKE, III, H.H. & CHILLINGARIANG.V. 1974. Compaction of Argillaceous Sediments. Elsevier Scientific Publishing Co. 424 pp. SEED, H.B., WOODWARD,R.J. & LUNDGREN,R. 1964. Clay mineralogical aspects of the Atterberg limits. J. Soil Mech. Div., A.S.C.E., 90, SM4, 107-31. YON HUENE, R., AUBOUIN,J. et al. 1980. Leg 67: The Deep Sea Drilling Project Mid-America Trench transect off Guatemala. Bull. geol. Soc. Am., 91, l, 421-32. WAGNER, A.A. 1957. The use of the Unified Soil Classification system by the Bureau of Reclamation. Proceedings of the 4th International Conference on Soil Mechanics and Foundation Engineering (London), 1. 125 pp. --.
R.W. FAAS,Department of Geology, Lafayette College, Easton, Pennsylvania, USA.
Fabric of muds and shales: an overview C.F. Moon and C.W. Hurst S U M M A R Y : This contribution essentially constitutes an attempt to update a previous review of clay sediment microstructure written over a decade ago. Recent developments in this field are considered and a scheme has been developed in which the fabric of argillaceous sediments is traced from its inception during sedimentation and its subsequent modification at and below the depositional interface. It is held throughout that the primary control in the determination of clay fabric is the chemical environment that prevails during and after the deposition of the sediment. Consideration of the evidence from a variety of sources suggests that fissility in mudrocks is the result of an anisotropic microfabric in which clay aggregates are preferentially aligned. This microstructure is thought to result from the dispersing action of organic compounds common in certain sedimentary environments; a brief discussion of these compounds is included. Other factors are also taken into account which have been offered as the cause of differing mudrock fabrics.
The microfabric of clay sediments and mudrocks has for long been regarded as the 'b6te noire' of sedimentary geology. This is due, in no small part, to the inherent problems associated with their investigation and foremost among these is the extremely small size of clay particles. Clarke (1924) estimated that argillaceous sediments account for 80% of the global rock column and if this is so, it is rather surprising that more attention has not been paid to their microscopical characteristics (cf. sandstones). In the past twenty years, various authors have made concerted attempts to unravel the nature and origins of the fabric of clay sediments. The results of a number of thorough investigations from a variety of disciplines have revealed that certain common factors exist in the genesis of microfabric. In contrast, these same studies have cast doubt on a number of traditional assumptions that have previously been adhered to. There is little doubt that clay sediments are complex materials. Sub-microscopic particle size is merely one aspect of their nature; the complexity is greatly increased by the electrochemistry of clay particles. Their behaviour is difficult to predict in sedimentological terms and has largely been realized by deduction; it is only now that a pattern of behaviour of the origin and modification of clay microstructure is becoming clearer. This has been made much easier in recent years by use of the scanning electron microscope (SEM). A decade ago, one of the authors was responsible for a review (Moon 1972) which outlined the studies and conclusions that had emerged on the topic of clay microstructure to that date. This work is now sadly lacking since much relevant information has appeared in the interim period. The present contribution is an attempt to update the earlier paper and to include various observa-
tions made by the authors which have a direct bearing on more recent published discussions. This contribution traces the development of microfabric throughout the history of the sediment from its deposition through burial. Initially, the structure ofsedimenting clays is discussed and this is followed by a study of the fabric of uncompacted sediments. Compacted clays are then investigated, with particular attention to the controversial topic of fissility and a review of the effects of shear forces on argillaceous deposits. It will be observed throughout this discussion that much evidence for microstructural aspects has been supplied by sources in engineering, ceramics and pedology in addition to sedimentology. The authors have strived to assimilate the various approaches and synthesize the results into a coherent scheme from which clay microstructure can be seen to develop in geological environments.
The microfabric of sedimenting clay Under this heading are considered dilute clay suspensions in which clay mineral particles are settling, i.e. moving predominately under the influence of gravity towards the depositional interface. As a direct result of the practical difficulties involved in the observation of suspended clay particles, this topic is open to the widest speculation. Few authors have reported any specific findings but certain generalities do emerge from their work to date. Due to the extremely complex electrochemistry of clay particles and their immediate surroundings, the chemistry of the depositional environment is critical in deciding the spatial arrange579
580
C.F. Moon and C. W. Hurst
ment that the particles will assume during settling and deposition. The major chemical controls are pH, Eh and the presence or absence of surface active ions (Goodwin 1971). In addition, it would seem to be logical that the volume concentration of clay is an important factor. It is widely recognized that clay particles will form aggregates in electrolytic solutions by the mechanism of flocculation (Van Olphen 1963). Whitehouse et al. (1960) in an exhaustive study of clay deposition write: 'Clay minerals are not transported and do not settle as single solid grains or in irreversible states of aggregation in saline water. The settling unit is a thermodynamically reversible assembly of solid grains or threads, in association with occluded water and water of hydration, and has been termed a 'coacervate'.' This then, suggests that although flocs may initially form when clays are introduced into a saline environment, they may dissociate themselves by dispersion if the environment alters; a change in environment may occur at, or above the depositional interface. Whitehouse et al. (1960) conclude that the flocculation characteristics of a sediment depends on its mineralogy. However, this has subsequently been challenged by Kranck (1975) who attaches more importance to the particle size mode. Sherman (1953) has shown that flocs are able to form in freshwater environments. Probably Van Olphen (1963) has contributed more to the understanding of clay particle association than any other author and his hypothesis suggested a number of possible particle arrangements. All of these involve the edge-to-edge (EE), edge-to-face (EF) and face-to-face (FF) interactions of platy-clay crystals, and it appears that electrolyte concentration decides which of these modes will result. Subsequently, Rieke & Chilingarian (1974) have claimed to have observed four modes of aggregation in 'thick' clay-water suspensions. In two of these, their 'flocculateddispersed' (sic) and 'flocculated with EF contacts' models are not single particle structures but are packets of particles with FF interactions. Unfortunately, the authors do not explain h o w these have been observed, neither do they mention the volume concentration of clay which is critical in this context. An important step in the study of suspended clay aggregates was made by Schweitzer & Jennings (1971) who, by means of a complex lightscattering technique indirectly observed montmorillonite associations in dilute suspensions. In a detailed analysis of their results they conclude that the most likely arrangement is of the EF
cardhouse type which compares most favourably with one of Van Olphen's models. As a rider to this conclusion, Flegmann et al. (1969) have noted that at values above 7 the pH of a clay suspension may give rise to EE flocculation (pH of sea-water ~ 8). It does seem that the available evidence, although limited, suggests that single-plate, cardhouse-type structures are present in sedimenting clays under normal conditions ofpH and salinity. These structures have perhaps a dozen individual particles arranged in a simple EF array. In recent years, an interesting development has emerged which demonstrates that aggregation may not be solely due to electrochemical bonding. Ernissee & Abbott (1975), Zabawa (1978) and McCall (1979) have shown by electron microscopy that suspended sediments of clay size may be agglomerated by faunal secretions. The significance of this in microstructural studies cannot be underestimated.
The microfabric of freshly sedimented clay In this context the expression 'freshly sedimented' refers to that material which has immediately settled and has not been subjected to any degree of compaction by superincumbent sediments; it lies at the depositional interface (or mudline). The fabric of a sedimenting clay is difficult to observe directly and the same is true of a fresh sediment, but by studying material that has suffered only slight burial, certain inferences can be made and conclusions drawn by the extrapolation of results. Many models of clay microstructure have emerged from various sources, notably from the study of clays by soil engineers who have contributed much in this direction. It must be stated that observations made on engineering clay soils are, in many cases not directly applicable to a study of fresh clay sediments but if treated with caution, certain common factors emerge. Karl Terzaghi (1925) suggested a variety of clay microfabrics which he termed 'strukturarten'. Foremost among these was the 'honeycomb' structure which was later adopted by Casagrande (1932) although the latter author admitted: 'We have not succeeded, so far, in reproducing this natural structure in the laboratory'. It is indeed a tribute to Terzaghi that more recent authors have directly observed fabrics bearing a strong resemblance to his model (Mattiat 1969; Sergeyev et al. 1980). Rosenqvist (1953) thought of marine clay as having a very open structure due to flocculation
Fabric o f muds a n d shales." an overview and chemical influences were taken up by Lambe in 1958. Lambe considered that clay sediment microstructure is primarily dependent upon electrolyte concentration. Three fabric types emerge from his thesis: salt-flocculated, non salt-flocculated and the dispersed structure. The first two of these are characterized by a predominance of EF particle contacts and were nominated 'cardhouse' structures. These were considered three-dimensionally by Tan (1959). The cardhouse structure assumed great significance in later years and was rigorously adopted by Rosenqvist (1962). A certain amount of support for the existence of the cardhouse structure was provided by the author via transmission electron microscopy (TEM), but the 'salt leaching theory' of quick clay behaviour has now been largely abandoned (see Conlon 1966; Cabrera & Smalley 1973; Moon 1979). A modified version of the cardhouse structure was postulated by Von Engelhardt & Gaida (1963) in their 'aggregate structure' which is held to form in electrolyte-rich solutions. A significant development in microstructural studies emerged from the Australian authors Aylmore & Quirk who introduced the domain concept. In 1960 they wrote: 'A domain was envisaged as a microscopic or submicroscopic region within which the clay particles are in parallel array. These groups of oriented crystals, or domains~ are randomly placed with respect to one another throughout the clay matrix; that is, the domains are in turbulent array.' Domains in this context are also referred to by the authors as turbostratic groups and two major types have been identified by subsequent workers (Fig. 1). Aylmore & Quirk's conclusions emerged by way of their TEM studies using carbon replicas and it was at this time that the scanning electron microscope (SEM) was being taken up in earnest by sedimentologists and soil researchers. The direct observation of fine texture in threedimensions radically changed the whole approach to the investigation of clay microstructure. Smart (1967, 1969) was one author in the vanguard of SEM investigations and produced a comprehensive account of particle associations. His observations demonstrated that both singleplate and turbostratic structures could be present
I
|
~
|
a
I
•
b
I
I
I
I
I
c
Fro. 1. Domain structures. (a) 'bookhouse' (Sloane & Kell 1966). (b) and (c)'stepped face-to-face variants' (Smalley & Cabrera 1969).
581
in freshly sedimented clay. Another authority who carefully investigated microstructure at this time was O'Brien (1971) who has contributed a wealth of information by his research. By way of an SEM study of laboratory prepared, flocculated clays, this author observed randomly orientated domains which could perhaps, be termed a random domain structure. The domains exhibited the stepped FF structure of Smalley & Cabrera (1969) (Fig. 1). A study by Mattiat (1969) utilized freeze-dried laboratory prepared kaolinite and illite sediments for an SEM investigation. Among the fabrics recognized was the 'kartenhausgefuge' which does appear to be composed of domains in cardhouse arrangement. The evidence therefore, would seem to point to the existence of domains in freshly deposited, uncompacted clays and this is given some support by Bowles et al. (1969) who studied natural sediments (by TEM) and identified domains in a clay which had been subjected to a low overburden pressure of -,~50kPa. The conclusion here is that domains are an original feature of the clays and hence present before compaction. Similarly, Barden (1972) offered the view that single plate arrays are only present in dilute clay suspensions; in a dense clay system, only domain structures in some arrangement exist. The years following the studies of O'Brien and of Barden represent something of a hiatus in clay microfabric research. It was not until much later in the decade that an excellent and highly comprehensive work by Bennett and his co-workers (1977, i 981) provided a detailed investigation of previous work, techniques and the results of their own observations on Mississippi Delta and DSDP samples. Of paramount importance is the conclusion that sediments subjected to very shallow burial (1.4 m): 'reveal numerous randomly arranged domains in EE and EF contact producing a relatively high porosity sediment.' In general terms the authors showed that the microfabric of the DSDP samples approximated to that proposed by Pusch (1966) for quick clays and which consists of flocs with linking chains. The Mississippi Delta deposits, however, were similar to the flocculated (i.e. random domain) model of Moon (1972). The former of their conclusions echoes somewhat the words of Burnham (1970) who identified domains in a Recent marine alluvium sampled at a depth of only 1.6 m. Both natural and artificial clay sediments have been closely investigated by Osipov & Sokolov (1978) and the microstructure of Recent uncompacted deposits was found to be a type of honeycomb arrangement: ' . . . of foliated clay particle microaggregates
582
C.F. Moon and C. W. Hurst
interacting face-to-edge, less frequently faceto-face...' i.e. the random domain structure. Other types of natural sediments from Lake Maracaibo have been the subject of study by Hyne et al. (1979). It must be admitted that their results are to a certain extent enigmatic in that non-marine delta lev6e samples exhibit domainlike fabrics but the authors claim that higher salinity lake sediments have a 'cardhouse or honeycombed microstructure'. The manner of deposition would appear to have significance in fabric studies and yet this topic has received scant attention. An example is provided by O'Brien et al. (1980) who made a careful SEM examination of the clay microstructure of turbidites and hemipelagic silts. Their observations reveal that the hemipelagites have a greater degree of preferred orientation of clay particles than the turbiditic sediments. This is attributed to a slower sedimentation rate in the dispersed state as opposed to a more rapid and flocculated deposition which results from turbidity flows. On reflection it emerges that domain structures are usually present in freshly deposited clay sediments; the weight of the available evidence is in favour of this. If the conclusions of the previous section are taken into account then there must be a transition from single-plate to domain structures at, or very near to, the depositional interface. It is presently uncertain why this change occurs; possibly it is due to a sudden change in the clay mineral volume concentration as more particles are brought into closer contact. Whatever the reason, one feels that it must be an electrochemical mechanism that is operative.
The microfabric of compacted clay: beyond the mudline Under this heading some discussion of terminology is necessary. The term 'shale' is firmly entrenched in the geological literature and has been so for many years (Tourtelot 1960). A cursory glance at a number of geological texts reveals that even today, some confusion surrounds the precise definition of the word. The main contention involves the use of the descriptors 'laminated' and 'fissile', i.e.: shale: ' . . . has well-marked bedding plane fissility (primarily due to the orientation of the clay particles parallel to the bedding planes).' (Whitten & Brooks 1972) shale: ' . . . laminated and fissile.' (Pettijohn 1975) shale: ' . . . laminated clayey rock.' (Tourtelot 1960).
In an attempt to rationalize the terminology, Potter et al. (1980) insist on specificity in stating that shale must have > 33% clay size material and have laminae < l0 mm thick. On the subject of fissility, this is defined by the authors simply as a type of parting having a thickness between 0.5 and 1.00 mm. Hence, it is possible for a non-fissile shale to exist. Spears (1980) has considered this very problem of classification. In this discussion, the authors will be content to regard a shale as an indurated clay rock readily splitting into thin laminae along the bedding planes. Consolidated clay rocks not fitting this description will be referred to as mudstones. The consideration of compacted clay presented here attempts to trace the influence of increasing axial pressure on a clay deposit and consider the associated aspects of diagenesis. In connection with diagnetic changes, one must include, of course, the chemical transformations that occur. As this review concerns only microstructural changes it is not proposed that any more than a mention be made of this. The subject has been dealt with at length by von Engelhardt & Gaida (1963), Weaver & Beck (1971) and Chilingarian et al. (1973). It is appreciated here that there is a complex interrelationship between the chemical and mechanical changes that take place during the diagenesis of clays and muds. An important point for discussion concerns the reason why some clay sediments become lithified to shale, whilst others are converted to mudstone. Certain authors have assumed that the development of fissility is merely a function of a high overburden pressure; Baldwin (1971) has suggested that fissility develops in clays at 350 m depth. The evidence available is against this premise, since many sediment cores sampled at depth show fissile shales to be found a b o v e non-fissile mudrocks (White 1961). Another mechanism other than pressure must operate in the formation of shale as opposed to a mudstone and it emerges that environmental factors are relevant in this context. The domain structure has been repeatedly observed in clays subjected to compaction and among the first authorities to identify these structural units were Sloane & Kell (1966) whose 'bookhouse' structure was readily adopted in the literature (Fig. 1). Much later work has proved with little doubt the existence of domains in clay deposits compacted both naturally and artificially (Barden et al. 1969; Barden & Sides 1970, 1971). A pertinent point has been made by Blackmore & Miller (1961) who found that as compaction of a montmorillonite clay proceeds, or as overburden depth increases, a greater number of particles is incorporated into each
Fabric o f muds and shales: an overview domain. In contrast to these findings, however, McConnachie (1974) found that this was not the case in kaolinite slurries under test. Hesling (1970) in direct contrast, appears to consider only single-particle structures in compacted clays and maintains that compactive pressure increases particle parallelism. From his specific surface measurements the author concludes that at about 1000 m depth there is a dramatic decrease in specific surface which he attributes to particle fusion. Of prime concern in this section is the reason for the development of fissility. An obvious assumption would be that the existence of an anisotropic fabric would produce weak planes of splitting as in slate. The temptation, therefore, is to conclude that fissility in shales is due to the alignment of clay flakes in a preferred direction: 'fissility in shales is usually associated with a parallel arrangement of the micaceous clay p a r t i c l e s . . . ' (Ingram 1953). If this is so, this must by necessity be modified in the light of the domain model. Examination of clay shales by SEM is an easy matter in comparison with sediments having a high moisture content. Since the moisture content of indurated rocks is low there is a minimal chance of fabric disruption due to drying or water replacement. O'Brien (1968a, 1968b) has observed well-developed particle alignment in addition to the identification of domains. The present authors have similarly noted a high degree of preferred orientation in all shale samples studied (Fig. 2). Non-fissile mudstones do not appear to display any degree of preferred orientation (Fig. 2(e)) and so it does seem that fissility is related to particle (domain) orientation. It must be admitted that domains are not easy to recognize in highly compacted mudrocks, especially shales; the very process of compaction pushes domains so close together that they merge into one another. Logically, domains should be present in compacted clays even though difficult to recognize essentially because they are present in fresh, uncompacted clay sediments. Smart (1967) has mapped domain distributions in an artificially prepared kaolinite sediment which was consolidated under a pressure of ,-~550kPa. In addition, McConnachie (1974) identified domains in all of the kaolinite samples which were compressed to a maximum of 105kPa and comments: ' . . . in the experiments described here, domains were present from the outset.' Another comprehensive study of artifical compaction, both isotropic and anisotropic, has been undertaken by Krizek et al. (1975)in which kaolinite slurries were prepared with flocculating
58 3
and dispersing additives. Among the conclusions is the observation that anisotropic compaction (i.e. with lateral restraint) tends to produce a 'highly oriented fabric' and that dispersion produces greater orientation than flocculation. There was a suggestion of domains in some samples. By careful assimilation of the various conclusions above, it is apparent that: (1) there is a distinct connection between fissility and clay particle orientation; (2) clay particles are orientated to form domains which display parallel alignment in shales but not so in mudstones; (3) there is a suggestion that for some clay mineral species, increasing pressure produces an increase in the number of particles per domain. The above speculations have been discussed by Byers (1974) who offers an alternative explanation for the absence offissility in some mudrocks. This author, in making a careful study of a Devonian argillaceous horizon, found a lateral transition from shale to mudstone. Close inspection revealed that the mudstone was strongly bioturbated and Byers' conclusion is that fissility is due simply to a lack of benthic infauna. In essence, mudstones display no fissility because they have been re-worked and hence a biogenic origin is envisaged for their microstructure rather than a physico-chemical mechanism. As a rider, the author adds that the azoic shales did display an orientated fabric which is presumably responsible for fissility. The role of bioturbation has also been given some importance in clay sediments lacking fissility by Barrows (1980). In any study of fissility and mudrock microstructure reference must be made to Spears' (1976) study of Coal Measures sediments. Following an investigation of weathered and unweathered shales from the same sequence, Spears concluded that fissility is primarily due to the weathering of organic-rich laminations. The explanation offered is that on weathering, the laminations absorb a large quantity of water which causes swelling and induces splitting along these laminations. In support of this, it has been reported that organic-rich layers in mudrocks are characterized by strong preferred orientation (Odom 1967). To some extent, the importance that Spears places upon weathering is a development of the conclusions of Ingram (1953). This constitutes a reasonable approach to the problem but logically one would expect a fissile structure to result from the weathering of a rock with a parallel, orientated microfabric basically because weathering will proceed more rapidly in the plane of particle orientation. The higher weathering rate is the
584
C.F. Moon and C. W. Hurst
(a)
(b)
(c)
(d)
(e) FIG. 2. (a) Fissility of non-marine shale, Lower Coal Measures. Scale bar lmm. (b) Microstructure of shale shown in Fig. 2. Scale bar 20/~m. (c) Microstructure of Ludlovian (upper Silurian) shale. (d) Microstructure of Stockdale Shale, Llandoverian, (lower Silurian). Note disruption of fabric by authigenic framboidal pyrite. Scale bar 20#m. (e) Microstructure of mudstone (Ashgillian, upper Ordovician). Scale bar 20pm.
Fabric o f muds and shales." an overview result of a much greater permeability in this direction. The weathering hypothesis was later taken up by Curtis et al. (1980) who observed, as perhaps would be expected, that equant quartz grains in mudrocks have a disruptive effect on microstructure. On the subject of fissility, however, they state rather paradoxically: ' . . . it may therefore be that clay orientation has nothing to do with fissility.' This is in direct contrast to the conclusions of a great variety of other authorities (e.g. O'Brien 1968; Barrows 1980).
The influence of non-vertical forces: the effects of shear In some environments, sediments may be subjected to shearing forces prior to complete consolidation, an obvious example being gravity sliding on submarine slopes. Some investigations have been made to assess the influence of shear and the development of shear planes upon the microstructure of clays. These studies have in the main, been performed by workers in the field of soil mechanics who have sheared plastic clays (i.e. at moisture contents below the liquid limit) to failure and have studied the induced failure planes microscopically in an attempt to gain a better understanding of the failure mechanism of clay soils. Many of these investigations have immediate relevance to the study of fine-grained sediments and possibly to tectonic mechanisms. A brief review of the more pertinent literature is given below and this is augmented by certain of the present authors' observations. A large number of investigations have successfully demonstrated that in a random domain fabric, shear strains of varying magnitude cause an alignment of particles parallel or sub-parallel to the plane of shear. At very large strains more radical effects on fabric are seen, but in all cases varying degrees of rotation of domains result. Weymouth & Williamson (1953) studied the effect of extrusion on a plastic clay by forcing the material out of a large diameter cylinder into a smaller one arranged concentrically. The fabric was studied by the optical microscopy of peels and amongst their conclusions is found: 'Diagonal shear joints crossed the fabric of the extruded cylinder. They contained clay layers having (001) at a small acute angle to the shearing surfaces.' (Presumably, since optical methods were employed, it was domains rather than particles which were identified.) A comprehensive account of shear-induced
585
microstructural changes has been provided by Morgenstern & Tchalenko (1967). After subjecting laboratory-prepared kaolin samples to direct shear, a variety of features were apparent. Among these were the Reidel shears, or tension fractures noted by Weymouth & Williamson above. Also recognized were kink-bands containing zones of orientated domains and 'compression textures' along shear surfaces in which two bands of domains orientated parallel to the shear plane enclose another aligned zone with particles/ domains set at 40~ ~ to these bands (Fig. 3). Tchalenko (1968) recognized similar structures in natural samples of London Clay. The effect of shear has also been studied by Clark (1970) who, in accordance with other authors maintains that distortion as opposed to simple compression is more significant in the development of preferred orientation. Clark suggested that preferred orientation resulting from the shear stresses is primarily due to the breakdown of flocs. This may have some importance in that it is possible that domains are destroyed by shear. Extensive use of the SEM has been made by Barden et al. (1969) whose results lent much support to the conclusions of Morgenstern & Tchalenko, especially in that they observed shear zones to have a tripartite structure, i.e. the compression texture. Similar observations have been made by Foster & De (1971) and a more recent, quantitative analysis has been undertaken by Tovey & Wong (1980). In addition to parallel alignment, other features may be identified which seem to be produced at large shear strains. Included here are compression textures and also
FIG. 3. Diagrammatic representation of 'compression texture' (Morgenstern & Tchalenko 1967).
586
C.F. Moon and C. W. Hurst
o
,.o
0
0"~ N.~
"~_, ~1
~'~
6~
Fabric of muds and shales." an overview 'fabric rolls'. The authors have identified both of these features in sheared clays and they are shown in Fig. 4. Both particle arrangements are reminiscent of shear-induced megascopic features familiar in structural geology. It may be reasonable to consider the compression texture to be a form of imbricate or schuppen structure whilst the fabric rolls could be regarded as drag folds (Badgley 1965). Apart from the work of Tchalenko (1968) few authors have paid attention to the identification of shear structures in natural sediments. Foremost in this context are the more recent investigations by Bohlke & Bennett (1980) in relation to a study of pro-delta crusts. The crusts are typified by a high shear strength and a TEM study showed them to have a dense domain microfabric with a high degree of preferred orientation quite unlike the sub- or superjacent sediments. The authors suggest that the existence of the crusts and the origin of their fabric can be explained by shear deformation resulting from mud flow sliding and slumping on an unstable pro-delta slope. As a direct corollary to the above, a rather radical approach has been advanced by Davies and his co-workers (Davies & Cave 1976; Davies 1980). The postulate is that primary clay flake orientation is held responsible for cleavage in slates and by way of an elegant argument, Davies (1980) attributes slaty cleavage to a parallel clay mineral fabric produced by shear slippage in unconsolidated clays. The shear component is the result of the viscous flow of sediment on a slope (or due to an 'uneven superincumbent load'). The author writes: '. . . . . . the restriction of commercial quality slate to a few, comparatively thin stratigraphic horizons (in North Wales) could imply that the clay structure on these horizons became particularly meta-stable and more susceptible to shear and consequent reorientation of the clay flakes.' It can be inferred from the foregoing that to a certain extent, in high water content sediments, parallel alignment is easily produced by shear forces. At lower moisture contents below the liquid limit, other more complex microstructural features are the result. As a postscript to this section, mention is made of the effect of shear forces upon sedimenting muds. Stow & Bowen (1978) have invoked 'shear sorting' as a mechanism in the development of lamination in fine-grained turbidites. During turbid flow, clay flocs are intermixed with silt particles of similar diameter and hence similar settling velocity. Upon reaching the flow boundary layer, however, shear stresses cause the flocs to disaggregate which in turn reduces their set-
587
tling velocity and segregation of clay and silt produces a laminated sediment. In this context, one is reminded of Clark's (1970) conclusions.
Synthesis and conclusions The previous sections attempt to trace the origin and development of particle arrangements in clays. It remains to marshal the available information into a logical scheme and provide some explanation of the observations and conclusions made to date. It is apparent that several authors have observed a parallel fabric in fissile mudrocks. Conversely, a random domain structure is seen to be present in non-fissile clay sediments. There appears a distinct and recognizable correlation between fissility and domain orientation. Preferred orientation of platy particles produces an anisotropic fabric which is not uncommon in other rocks, e.g. flaggy sandstones and slates, which are typified by strength anisotropy manifested by planes of splitting. The authors can see little reason why this does not apply to mudrocks, at least, in the majority of cases. If this is accepted, then a mechanism must be found in which preferred orientation is developed in some mudrocks and not in others. It is the authors' contention that this is mainly the result of the environment of deposition. Especially, the chemical conditions that exist during and immediately after, the sedimentation of an argillaceous deposit. The detailed chemistry involved is at present obscure but the variables concerned must include electrolyte concentration, pH and Eh. Rheologists (e.g. Flegmann et al. 1969; Goodwin 1971) have demonstrated that the manner in which clay particles associate depends to some extent upon pH. Colloid chemistry has highlighted the equal importance of electrolyte concentration (Van Olphen 1963) but perhaps it is oxidation potential that has not received the attention it otherwise deserves. The basic premise here is that in organic-rich conditions (marine or non-marine) there will exist the required dispersing agents which allow particles to settle individually and produce a parallel, orientated microfabric. This in turn will result in an anisotropic strength responsible for fissility. It is not assumed that this is the sole reason for fissility but the authors consider it to be one of great importance. One must also bear in mind the speculations of Spears (1976) and Hallam (1980) who attribute fissility to localized bands of organic material. Various authors have noted the association between organic content and shales (Ingram
C.F. M o o n a n d C. W. H u r s t
588
1953; O'Brien 1968b; Leventhal & Shaw 1980). The organic material in these cases manifests itself in the dark colour of the rock and Ingram (1953) has observed that shales tend to be dark whilst mudstones are usually of a light colour. Indeed, black shales are commonplace rocks and regarded as a distinct facies in their own right. It is well known that certain organic ions in very low concentrations are capable of dispersing (or peptizing) clay minerals (Van Olphen 1963) and a number of these organic compounds are common in a variety of depositional environments. Probably the most studied of these is the anoxic basin in which bottom waters become oxygen-depleted due to the aerobic decomposition of planktonic organisms and produce a reducing environment. Classic examples are
l]ords (Strem 1955; Grasshof 1975), fjord-like basins such as the Saanich Inlet (Brown et al. 1972) and other ocean deeps for example the Santa Barbara Basin (Degens et al. 1961) and the Cariaco Trench (Deuser 1975). In addition to these is the barred basin typified by the Black Sea (Degens & Stoffers 1980). An examination of the hydrochemistry of these areas (Deuser 1975) indicates that an anoxic zone exists at a depth which is rich in organic compounds. The transition from oxic to anoxic conditions may coincide with the mudline or it may be located within the overlying water column. It proposed that clay sediments entering the anoxic zone have the charges on their particle surfaces neutralized by organic compounds and a dispersed (domain?) deposit results. Following compaction and induration, a parallel orientated
o
+
.<
~................. .7.............
5
~
I
o
• ,,,
~ ~ / ~' - ~~ J~ / /
I
.~.
\.7~
I \
,'~ _\
//
\
/ \---, \ / ,
............ I \> z / o
_./
I ~ /l . . . .m. .u d - -J-- - ~~ - ~k._i___ <,._--____/_~__ line dora "
LITH
~ ~
~
~
..7.~-~---I ~--~-~'
~
FICATION
~ ~~:~~_____-~-.:
_ -~_-~
@
~___~__-~_
~ ~ - - ~ ~
~
~_"
_ ~ . ~ ~ ~ f
I '
MUDSTONE (non-fissi le)
~
---
-~_
---~_.
~ _
~-.
SHALE (fissile)
FIG. 5. Schema depicting the suggested major microstructural changes in clay sediments.
Fabric o f muds and shales." an overview domain structure is formed conducive to the development of fissility. In 'normal' marine conditions, no dispersion occurs, the sediment remains flocculated producing a random domain structure and hence induration results in a nonfissile rock. An idealized scheme of the proposed process is shown in Fig. 5. It is probable that this is a major mechanism in the formation of fissile mudrocks and this was loosely suggested by Marr (1925) in his study of the Stockdale Shales of Cumbria. Although at present, the necessary chemical environment is not common in the global ocean, there is evidence that it has been more prevalent in the past, for instance in the Cretaceous (Jenkyns 1980). Degens & Stoffers (1976) have written: 'The well-mixed oxygenated ocean of today seems not to be the model for the past 600 my.' Consideration must be given to those shales that do not have a marine origin, an example being the non-marine shales of the Coal Measures. These sediments are customarily thought of as having been laid down in a paludal environment fed by rivers draining a densely vegetated hinterland and hence with a high organic content. It can be surmised that shallow basins developed anoxic zones as a direct result of these organicrich conditions. Of interest here is Twenhofel's (1939) distinction between humic and bituminous black shales. The former are characterized by coal-like carbonaceous constituents derived from the more advanced plants whilst bituminous shales are non-coaly and have organic material originating from fatty algae and spores. Twenhofel observes: "The humic black shales seem to have accumulated in waters of poor circulation in which bacterial or biochemical decomposition was very limited. The bituminous black shales accumulated, in general, in less stagnant and more open water, but waters that did not have good circulation.' Pertinent to this discussion is a brief consideration of the nature of the organic compounds that may exist in such environments and which could induce the dispersion of clay minerals. Van Olphen (1963) treats the interaction of clay minerals and organic compounds at length and distinguishes between compounds of low to moderate molecular weight and macromolecular compounds. The latter group when dissolved in water, represent hydrophilic colloids and require special consideration. Among the former are discussed organic anions, organic cations and polar groups. Moore (1969) has divided the organic matter present in marine sediments as follows:
589
(1) original and altered proteins and their decomposition products: amines, amino acids and their complexes; (2) carbohydrates including cellulose and sugars; (3) lignin; (4) pigments including chlorophyll-like compounds, carotenoids and porphyrins; (5) lipids derived from phytoplankton. At this point, therefore, it is necessary to attempt to equate the organic compounds available in reducing sedimentary environments and those which are capable of clay mineral dispersion. Of Van Olphen's list of clay-reactive compounds, the organic cations and anions appear to be the most significant in marine environments and amino groups emerge as common constituents of sea water. Of the amino acids, Degens (1965) is a source of much relevant information. He observed that as microbial activity increases with depth so do the amino acids and in reducing conditions, Recent sediments contain a high concentration of free amino acids. As an illustration, sediments of the Santa Barbara Basin are shown to have a free amino acid content of ~ 200 pg/g. Degens envisages that hydrolysed plankton protein produces free amino acids which are redistributed along the mudline and adsorbed and re-sorbed on clay minerals. Hare (1969) has re-iterated this point. 'The water-sediment interface shows the highest concentration of amino acid material . . . containing as much as three to four orders of magnitude more amino acid material (free and combined) than the overlying water.' Sugars (e.g. pentoses and hexoses) also receive attention by Degens (1965) and this author has identified high concentrations of these carbohydrates at the depositional interface. In a similar way to amino acids, the free sugars may become redistributed and preserved in reducing conditions whereas in oxidizing environments, they are eliminated biochemically by microbes and the benthos. In anoxic surroundings the free sugars, after prolonged interaction with clay minerals, should form stable complexes. The dispersing properties of sugar polymers (polysaccharides) have been demonstrated by Bloomfield (1956) who has studied the deflocculation of kaolinite by vegetable extracts of trees. No mention has so far been made of humic acids and their related compounds. These chemically ill-defined substances are present in a variety of environments and although the majority of these compounds are regarded as allochthonous, there is a strong suggestion that the humic material in some deep-sea sediments is autochthonous in origin (protein derived?) (Moore
59o
C.F. Moon and C. W. Hurst
1969). The effect of humic acids and their derivatives on clay minerals is generally known: 'The peptizing effects of humus on clays is important in sedimentary geology. The peptizing action is followed by the formation of adsorbed humus coatings around the clay particles which in turn retard the flocculating effects of the electrolytes on the clays.' (Swain 1963). If humic acids are known to exist in deep-sea sediments and also in eutrophic lakes and marshes, by the mechanism of dispersion/peptization it is possible that they could give rise to both humic and bituminous shales. By virtue of their frequent occurrence, it is the dark grey and black shales that have received most attention here. In any discussion of shale fabric some attention must be paid to those of other colours such as the red and green shales. As many authors have noted (Pettijohn 1975) with the exception of red beds, quantitative information on these rocks is scarce. McBride (1974) has identified a colour progression from red to green to grey based on the ratio of ferric/ferrous iron content. It should be stated, however, that Potter et al (1980) have shown that the percentage of organic carbon and FeS2 also have a modifying effect on shale colour. In general, a high Fe3+/Fe 2+ ratio produces red colours whilst green shales are associated with low ratios. It follows that in most cases it is oxidation state that controls colour. It is usually accepted that the high Fe 3+ content of red shales is attributable to oxidizing continental conditions. If this is so, in accordance with the present thesis, clay sediments deposited in oxidizing/aerobic environments should not display fissility since there would not exist the required organic dispersants. It is suggested here that fissility in red and green shales is due to the dispersion of clay particles by inorganic compounds present in (arid?) continental environments as opposed to the organic dispersants of
black shale formations. Possible inorganic agents which would bring this about are phosphates or evaporites (given the appropriate pH conditions). As a caveat to this, it should be remembered that the continental Permian Marl Slate of England is currently thought of as having been laid down in an anaerobic/euxinic environment (Smith 1979). On the basis of the above, it becomes apparent that: (1) there is a distinct connection between dispersing agents and the origin of fissility in mudrocks which is usually due to the parallel orientation of clay particle domains; (2) certain organic compounds found in a variety of environments furnish the required peptizing properties e.g. amino acids, saccharides, lipids or lignin derivatives are possible contenders. The authors feel that much further research is needed into the aspects listed above both by making rigourous SEM and TEM studies of mudrock fabrics and by careful laboratory modelling of clay deposition and compaction. The relationship between organic matter and muds and its origin in shales also merits close study. In short, the authors can do no better than to emphasize the timely conclusions of O'Brien (1981) who stresses the importance of developing a standard method for the quantification of mudrock microfabric (i.e. through directional analysis techniques). This must surely be of prime concern to any future work concerned with the elucidation of the microstructure of clays, muds and shales. ACKNOWLEDGEMENTS: The authors would like to express their thanks to Dr D.B. Lewis and Mr G. Ingleton for SEM assistance. Also to Jane Tory for carefully reading the manuscript, and to Drs. O'Brien and Stow for their careful reviews. Some of the work was carried out whilst one of the authors (CWH) held the tenure of a Research Assistantship at Sheffield City Polytechnic.
References AYLMORE, L.A.G. & QUIRK, J.P. 1960. Domain or turbostratic structure of clays. Nature, Lond., 187,
1046-8. BADGLEY, P.C. 1965. Structural and Tectonic Principles.
Harper & Row, New York. 521 pp. BALDWIN, B. 1971. Ways of deciphering compacted sediments. J. sed. Petrol., 41, 293-301. BARDEN, L. 1972. The relation of soil structure to the engineering geology of clay soil. Q. J. eng. Geol. Lond., 5, 85-102. -8/- SIDES, G.R. 1970. Engineering behaviour and
the structure of compacted clay. Proc. Am. Soc. cir. Engrs., 96, SM4, 1171-200. -& SIDES, G.R. 1971. The microstructure of dispersed and flocculated samples of kaolinite, illite and montmorillonite. Can. Geotech. J., 8, 391-9. --, SIDES, G.R. & KARUNARATNE, J.P. 1969. A microscopic examination of aspects of clay structure. Proc. SE Asian Conf. Soil Eng. 2nd. Bangkok. 67-72. BARROWS, M.H. 1980. Scanning electron microscope studies of samples from the New Albany Shale
Fabric of muds and shales." An overview Group. In: SEM/1980/I, SEM Inc., A.M.F. O'Hare IL60666. 579-85. BENNETT, R.H., BRYANT, W.R. & KELLER, G.H. 1977. Clay fabric and geotechnical properties of selected submarine cores from the Mississippi Delta. NOAA Prof. Paper 9, 86 pp. , BRYANT,W.R. & KELLER,G.H. 1981. Clay fabric of selected submarine sediments: fundamental properties and models. J. sed. Petrol., 51, 217-32. BLACKMORE, A.V. & MILLER, R.D. 1961. Tactoid size and osmotic swelling in calcium montmorillonite. Soil Sci. Soc. Am. Proc., 25, 169-73. BLOOMFIELD,C. 1956. The deflocculation of kaolinite by aqueous leaf extracts: the role ofcertain constituents of the extracts. Trans. 6th. Int. Cong. Soil Sci., Paris 4, 27-32. BOHLKE, B.M. & BENNETT, R.H. 1980. Mississippi prodelta crusts: a clay fabric and geotechnical analysis. Mar. Geotech., 4, 55-82. BOWLES, F., BRYANT, W.R. & WALLIN, C. 1969. Microstructure of unconsolidated and consolidated marine sediments. J. sed. Petrol., 39, 1546-51. BROWN, F.S., BAEDECKER, M.J., NISSENBAUM, A. & KAPLAN, I.R. 1972. Early diagenesis in a reducing fjord, Saanich Inlet, British Columbia--Ill: Changes in organic constituents of sediment. Geochim. cosmochim. Acta, 36, 1185-203. BURNHAM,C.P. 1970. The micromorphology of argillaceous sediments: particularly calcareous clays and siltstones. In: Osmond, D.A. & Bullock, P. (eds), Micromorphological Techniques and Applications. Soil Surv. Tech. Monogr. 2, 83-96. BYERS, C.W. 1974. Shale fissility: relation to bioturbation. Sedimentolog3,, 21,479-84. CABRERA, J.G. & SMALLEY, I.J. 1973. Quickclays as products of glacial action: A new approach to their nature, geology, distribution and geotechnical properties. Eng. Geol. Amsterdam, 7, 115-33. CASAGRANDE, A. 1932. The structure of clay and its importance in foundation engineering. J. Boston Soc. cir. Eng., 19, 168-221. CHIL1NGARIAN, G.V., SAWABINI, C.T. & RIEKE, H.H. 1973. Effect of compaction on chemistry of solutions expelled from montmorillonite clay saturated in sea water. Sedimentology, 20, 391-8. CLARK, B.R. 1970. Mechanical formation of preferred orientation in clays. Am. J. Sci., 269, 250-66. CLARKE, F.W. 1924. The data of geochemistry. Bull. US geol. SurF., 770, 841 pp. CONLON, R.J. 1966. Landslide on the Toulnoustoc River, Quebec. Can. Geotech. J., 3, 113-44. CURTIS, C.D., L1PSHIE,S.R. & PEARSON,M.J. 1980. Clay orientation in some Upper Carboniferous mudrocks, its relationship to quartz content and some inferences about fissility, porosity and compactional history. Sedimentology, 27, 333-9. DAVIES, W. 1980. On slaty cleavage and structural development in lc~wer Palaeozoic Wales. Univ. Coll. Wales, Aberystwyth Dept. Geol. Pub., 8, 43 pp. & CAVE, R. 1976. Folding and cleavage determined during sedimentation. Sed. Geol., 15, 89-133. DEGENS, E.T. 1965. Geochemistry of Sediments: A Brief Survey. Prentice-Hall, New Jersey. 342 pp. -
-
591
• STOFFERS, P. 1976. Stratified waters as a key to the past. Nature, Lond., 263, 22-7. & STOFFERS, P. 1980. Environmental e v e n t s recorded in Quaternary sediments of the Black Sea. J. geol. Soc. Lond., 137, 131-8. --, PRASHNOWSKY,A., EMERY, K.O. & PIMENTA, J. 1961. Organic materials in recent and ancient sediments Part II: Amino acids in marine sediments of Santa Barbara Basin, California. Neues Jahrb. Geol. Palaeontol. Monatshefte, 8, 413-26. DEUSER, W.G. 1975. Reducing environments. In: Riley, J.P. & Skirrow, G. (eds), Chemical Oceanography, 3. 2nd Edition. Academic Press, London. 1-38. EISMA, E. & JURG, J.W. 1969. Fundamental aspects of the generation of petroleum. In: Eglinton, G. & Murphy, M.T.J. (eds), Organic Geochemistry: Methods and Results. Springer-Verlag, Berlin. 676--98. ERNISSEE,J.J. & AaBOTT, W.H. 1975. Binding of mineral grains by a species of Thalassiosira. Nova Hedwigia, 53, 241-8. FLEGMANN, A.W., GOODWIN, J.W. & OTTEWILL, R.H. 1969. Rheological studies on kaolinite suspensions. Proc. Brit. Ceram. Soc., 13, 31-45. FOSTER, R.H. & DE, P.K. 1971. Optical and electron microscopic investigation of shear induced structures in lightly consolidated (soft) and heavily consolidated (hard) kaolinite. Clays Clay Min., 19, 31-47. GOODWIN, J.W. 1971. Rheological studies on the dispersion of kaolinite suspensions. Trans. J. Brit. Ceram. Soc., 70, 65-70. GRASSHOE, K. 1975. The hydrochemistry of landlocked basins and fjords. In: Riley, J.P. & Skirrow, G. (eds), Chemical Oceanography, 2, 2nd edition. Academic Press, London. 456-598. HALLAM,A. 1980. Black shales. J. geol. Soc. Lond., 137, 123-4. HARE, P.E. 1969. Geochemistry of proteins, peptides and amino acids. In: Eglinton, G. & Murphy, M.T.J. (eds), Organic Geochemistry: Methods and Results. Springer-Verlag, Berlin. 438-63. HESLING, D. 1970. Micro-fabrics of shales and their rearrangement by compaction. Sedimentology, 15, 247-60. HYNE, N.J., LAIDIG, L.W. & COOPER, W.A. 1979. Prodelta sedimentation on a lacustrine delta by clay mineral flocculation. J. sed. Petrol., 49, 1209-16. INGRAM, R.L. 1953. Fissility of mudrocks. Bull. geol. Soc. Am., 64, 869-78. JENKYNS, H.C. 1980. Cretaceous anoxic events: from continents to oceans. J. geol. Soc. Lond., 137, 171-88. KRANCK, K. 1975. Sediment deposition from flocculated suspensions. Sedimentology, 22, Ili-23. KRIZEK, R.J., EDIL, T.B. & OZAYDIN, I.K. 1975. Preparation and identification of clay samples with controlled fabric. Eng. Geol. Amsterdam, 9, 13-38. LAMBE, T.W. 1958. The structure of compacted clay. Proc. Am. Soc. cir. Eng., 94, SM4, 85-106. LEVENTHAL,J.S. & SHAW, V.E. 1980. Organic matter in Appalachian Devonian Black Shale: I. Comparison of techniques to measure organic carbon, II, Short
--
592
C.F. Moon and C.W. Hurst
range organic content variations. J. sed. Petrol., 50, POSCH, R. 1966. Quickclay microstructure. Eng. Geol. Amsterdam, 1,433-43. 77-81. MARR, J.E. 1925. Conditions of deposition of the RIEKE, H.H. & CHIL1NGAR1AN,G.V. 1974. Compaction of Argillaceous Sediments. Elsevier, Amsterdam. Stockdale Shales of the Lake District. Q. J. geol. 424 pp. Soc. Lond., 81, 113-36. MATTIAT, B. 1969. Eine Methode zur elektronenmik- ROSENQVIST,I. TH. 1953. Considerations on the sensitivity of Norwegian quick clays. Geotechnique, Lond., roskopischen Untersuchung des Mikrogefuges in 3, 195-200. tonigen sedimenten. Geol. Jahrb. Hannover, 88, 1962. The influence of physico-chemical factors 87-111. upon the mechanical properties of clays. Clays Clay MCBRIDE, E.F. 1974. Significance of colour in red, Min. Proc. Nat. Conf. Clays Clay Miner., 9, 12-27. green, purple, olive, brown and grey beds of Difunta Group, NW Mexico. J. sed. Petrol., 44, RUBEY, W.W. 1931. Lithologic studies of fine-grained Upper Cretaceous sedimentary rocks of the Black 760-73. Hills region. US geol. Surv. Prof. Pap. 165, 54 pp. MCCALL, P.L. 1979. The effect of deposit feeding oligochaetes on particle size and settling velocity of SCHWEITZER,J. & JENNINGS,B.R. 1971. The association of montmorillonite studied by light scattering in Lake Erie sediments. J. sed. Petrol., 49, 813-8. electric fields. J. Colloid Interface Sci., 37, 443-57. MCCONNACHIE,I. 1974. Fabric changes in consolidated SERGEYEV,Y.M., GRABOWSKA-OLSZEWSKA,B., OSIPOV, kaolin. Geotechnique, Lond., 24, 207-22. V.I., SOKOLOV, V.N. & KOLOMENSKI,Y.N. 1980. MEADE, R.H. 1964. Removal of water and rearrangeThe classification of microstructure of clay soils. J. ment of particles during the compaction of clayey Microsc., 120, 237-60. sediments--A review. Geol. Surv. Am. Prof. Paper SHERMAN, I. 1953. Flocculent structure of sediment 497-B, 1-23. suspended in Lake Mead. Trans. Am. Geophys. MOON, C.F. 1972. The microstructure of clay sediUnion, 34, 394-406. ments. Earth Sci. Rev., 8, 303-21. 1979. Concerning the relation between energy and SLOANE, R.L. & KELL, T.F. 1966. The fabric of mechanically compacted kaolinite. Clays Clay cementation in quick clay disturbance. Eng. Geol. Min. Proc. Nat. Conf. Clays Clay Miner., 14, Amsterdam, 14, 197-207. 289-96. MOORE, L.R. 1969. Geomicrobiology and geomicrobiological attack on sedimented organic matter. In: SMALLEY, I.J. & CABRERA,J.G. 1969. Particle association in compacted kaolinite. Nature, Lond., 222, Eglinton, G. & Murphy, M.T.J. (eds), Organic 80-1. Geochemistry: Methods and Results. Springer-VerSMART, P. 1967. Particle arrangements in kaolin. Clays lag, Berlin. 265-303. Clay Min. Proc. Nat. Conf. Clays Clay Miner., 15, MORGENSTERN,N.R. & TCHALENKO,J.S. 1967. Optical determination of preferred orientation in clays and 241-53. 1969. Soil structure in the electron microscope. its application to the study of microstructure in Proc. Int. Conf. Solid Mech. Eng. Design 1,249-55. consolidated kaolin, I & II. Proc. R. Soc. London SMITH, D.B. 1979. Rapid marine transgressions and A300, 218-34. regressions of the Upper Permian Zechstein Sea. J. O'BRmN, N.R. 1968a. Clay fabric of very fissile Paleogeol. Soc. Lond., 136, 155-6. zoic gray shales. Marit. Sed. Halifax, 4, 104-5, 1968b. Electron microscope study of a black shale SPEARS,D.A. 1976. The fissility of some Carboniferous shales. Sedimentology, 23, 721-5. fabric. Naturwiss., 55, 490-1. 1980. Towards a classification of shales. J. geol. 1970. The fabric of shale--an electron microscope Soc. Lond., 137, 139-56. study. Sedimentology, 15, 229-46. 1971. Fabric of kaolinite and illite floccules. Clays STow, D.A.V. & BOWEN, A.J. 1978. Origin of lamination in deep sea, fine-grained sediments. Nature, Clay Miner., 19, 353-9. Lond., 274, 324-8. 1981. SEM study of shale fabric--a review. In." O'Hare, A.M.F. (ed.), SEM/1981/1, SEM Inc., STROM, K.H. 1955. Land-locked waters and the deposition of black muds. In: Trask, P.D. (ed.), Recent Illinois, 569-576. Marine Sediments. Soc. econ. Paleo. Min., Spec. , NAKAZAWA,K. & TOKUHASm, S. 1980. Use of Publ. 4, 356-72. clay fabric to distinguish turbiditic and hemipelagic SWAIN, F.M. 1963. Geochemistry of humus. In: Breger, siltstones and silts. Sedimentology, 27, 47-61. I.A. (ed.), Organic Geochemistry. Pergamon Press, ODOM,E. 1967. Clay fabric and its relation to structural New York. 81-147. properties in mid-continent Pennsylvanian sediTAN, T.K. 1959. Structure mechanics of clays. Sci. ments. J. sed. Petrol., 37, 610-23. OsIPOV, V.I. & SOKOLOV,V.N. 1978. Microstructure of Sinica, 8, 83-96. recent clay sediments examined by scanning electron TCHALENKO, J.S. 1968. The microstructure of London Clay. Q. J. engrg. Geol. Lond., 1, 155-68. microscope. In: Whalley, W.B. (ed.), Scanning Electron Microscopy in the Study of Sediments. Geo- TERZAGHI,K. 1925. Erdbaumechanik aufBodenphysikao lischer Grundlage. Liepzig, Vienna. 399 pp. Abstracts, Norwich, U.K. 29--40. PETTIJOHN, F.J. 1975. Sedimentary Rocks. 3rd edn. TOURTELOT, H.A. 1960. Origin and use of the word 'shale'. Am. J. Sci., 258, 355-43. Harper & Row, New York. 628 pp. POTTER, P.E., MAYNARD, J.B. & PRYOR, W.A. 1980. TOVEV, N.K. & WONG, K.Y. 1980. The microfabric of deformed Kaolin. J. Microsc., 120, 329-42. Sedimentology of Shale. Springer-Verlag, New TWENHOFEL, W.H. 1939. Environments of origin of York. 310 pp. -
-
-
-
-
-
-
-
-
-
-
-
-
-
Fabric of muds and shales. An overview black shales. Bull. Am. Ass. Petrol. Geol. 23, 1178-98.
VAN OLPHEN, H. 1963. An Introduction to Clay Colloid Chemistry. Interscience, New York. 301 pp. VON ENGELHARDT,W. & GAIDA, K.H. 1963. Concentration changes of pore solutions during the compaction of clay sediments. J. sed. Petrol., 33, 919-30. WEAVER, C.E. & BECK, K.C. 1971. Clay water diagenesis: How mud becomes gneiss. Geol. Soc. Am. Spec. Pap., 134, 96 pp. WEISS, A. 1969. Organic derivatives of clay minerals, zeolites and related minerals. In: Eglinton, G. & Murphy, M.T.J. (eds), Organic Geochemistry: Methods and Results. Springer-Verlag, Berlin. 737-81. WEYMOUTH, J.H. & WILL1AMSON, W.O. 1953. The
593
effects of extrusion and some other processes on the microstructure of clay. Am. J. Sci., 251, 89-108. WHITE, W.A. 1961. Colloid phenomena in the sedimentation of argillaceous rocks. J. sed. Petrol., 31, 560-70. WmTEHOUSE, U.G., JEFFREY, L.M. & DEBBRECHT,J.D. 1960. Differential settling tendencies of clay minerals in saline waters. Clays Clay Min. Proc. Nat. Conf. Clays Clay Miner., 7, 1-79. WHITTEN, D.G.A. & BROOKS,J.R.V. 1972. The Penguin Dictionary of Geology. Penguin Books, Middlesex. 516 pp. ZABAWA, C.F. 1978. Microstructure of agglomerated suspended sediments in Northern Chesapeake Bay estuary. Science, 202, 49-51.
C.F. MOON & C.W. HURST, Mudrock Research Project, Department of Civil Engineering, Sheffield City Polytechnic, Sheffield S1 1WB, England.
Bioturbation in deep-sea fine-grained sediments: influence of sediment texture, turbidite frequency and rates of environmental change A. Wetzel S U M M A R Y: In the deep sea, burrowing infauna can adapt to variation in sediment texture
within < 100-300 years. When there is continuous sedimentation, a continuous bioturbation process also takes place and the biogenic trace association found in the fossil record may be dominated by deeply penetrating burrows. A low-diversity trace fossil association (Zoophycos ichnofacies) is typical. On the other hand, when turbidites interrupt the continuous bioturbation process, a highly diverse trace fossil association is often found in the fossil state that also contains highly organized near-surface traces (Nereites ichnofacies). This biotope is usually controlled by the non-turbiditic sedimentation rate and the recurrence time of turbidites. Biogenic trace assemblages are mainly controlled by the nature of sedimentation and other environmental conditions rather than by water depth. Therefore, the widely accepted use of trace fossils as palaeobathymetric indicators in the deep sea may be restricted.
Most environmental factors that influence macrobenthic infauna, for example, grain-size, sedimentation rate and benthic food content, are all somehow related to water depth (Tait 1971). Furthermore, Seilacher (1967) found that trace fossil associations (ichnocoenoses) are not randomly composed and proposed a depth zonation. He distinguished six bathymetry-controlled trace fossil communities, so-called 'ichnofacies'. In this paper, only the two deepest ichnofacies, the Zoophycos ichnofacies (slope to abyssal plain) and the Nereites ichnofacies (abyssal plain) are discussed. A number of authors have since described modern as well as fossil deep-sea ichnocoenoses based on Seilacher's (1967) depth-controlled model (e.g. Crimes 1977; Ekdale & Berger 1978; Berger et al. 1979) and based on Deep Sea Drilling Project cores (e.g. Chamberlain 1975: Ekdale 1977). Results of these investigations are not entirely consistent with the ichnofacies model (Ekdale & Berger 1978), and it seems especially important to distinguish between turbidite and non-turbidite ichnocoenoses. In this paper, an attempt is made to quantify ichnofaunal response to sedimentation processes that influence sediment texture. Two aspects are of main interest: (1) the timing of changes in sediment texture and reaction of infauna and (2) variations of biogenic trace diversity.
Methods and study areas Over 100 box cores (up to 50 cm diameter) and gravity cores (10 cm diameter, up to 11 m in length) were used in the present study. The cores
were cut longitudinally into 8 mm thick slices for X-radiography. Serial sections and stereo-pair sections were used to enhance three-dimensional representation of certain biogenic structures. In most cases trace fossils were identified at the generic level only (after Hfintzschel 1965, 1975), although with some, further subdivision ofichnogenera was possible (method of Wetzel 1979). Core suites were studied from three contrasting areas (Fig. 1): (a) the slope to abyssal plain off N W Africa, (b) the Sulu Sea Basin from the slope to abyssal plain, and (c) the Central Pacific Ocean abyssal plain. N W African continental margin
In an area between 12~ and 28~ 86 sediment cores retrieved from 56 sampling stations have been studied. A comprehensive report on this area is given by Seibold et al. (1976) and Sarnthein et al. (1982). The sediments investigated are up to 250000 yrs in age, and show the strong influence of climatic and oceanographic variation. In the north, the trade winds lead to coastal upwelling with increased organic productivity, and the Harmattan winds transport aeolian dust seawards from arid areas. In the south, rivers are the main source of clastic material supplied to the ocean. The sedimentation rate decreases from the slope (30-40 cm/1000 yrs) to the abyssal plain (2-3 cm/1000 yrs). Detailed stratigraphic evaluations based on O16/O TM and C 14 determinations have been carried out by Koopmann (1981). During glacial periods, the Saharan desert expanded both north- and southwards; furthermore, increased intensity of the trade and Hat- .. mattan winds provided a greater supply of 595
596
A. Wetzel
117~ 11
119~
118 ~ I
120 ~
121 ~
122 ~
123 ~
C2 10
r o210 207 O208 0217 o218
I
9' e219
0232
220o 'le
% \ o222 o223 224~=
7el
0
C7
117 o
118 ~
11'9~
120 ~
121 ~
122 ~
123 ~
(b) HAWAII
:.
II
j ! Ctorion froctur(
i+1
Ctipperton froc~ r e zone
42 9 29
q,76
36
6~
so 4o 3o 20
48 q 151 e
lo
150 a 149" 148 c 147~ 146~ 145 c 1&4~
(c) FIG. I. Areas investigated. (a) NW African continental margin, (b) Sulu Sea, and (c) Central Pacific Ocean. Core numbers refer to catalogue of cores of the Geological Institute, Kiel University. Cores discussed or figured in text are marked with a large dot.
598
A. Wetzel
organic matter (via upwelling) and of aeolian dust. Simultaneously, sediment input from rivers decreased. At the end of glacial periods during the most rapid rise of sea-level, mass-movements on the slope, including submarine landslides and turbidity currents, became more frequent. The surface sediments off NW Africa contain between 0.5 and 4% Corg but show lower amounts downcore. All the deposits are completely bioturbated by infaunal animals, except for turbidites and slide masses. Within the sediments, biogenic trace diversity is characterized by two features: (1) it is higher in sediments accumulated during glacial periods, and (2) it decreases from the slope towards the deep sea (Wetzel 1981).
tion, biogenic trace diversity increases with water depth. Central Pacific Ocean
Six cores of Quaternary to Tertiary age have been studied from the central Pacific. The sediments accumulated very slowly ( < 1 cm/1000 yrs) and do not contain any turbidites. These biogenic oozes and red clays are well-oxygenated. The organic-carbon content decreases from 0.4% at surface to < 0.1% deeper within sediment. The deposits are totally bioturbated. A comprehensive report about this area is given by Mfiller (1975).
Sulu Sea Basin
A comprehensive report of the Sulu Sea and its sediments is given by Exon et al. (1981). For this study, 23 sediment cores from 18 sampling stations have been used (Fig. 1). The Sulu Sea Basin is surrounded by shelves and islands. Water exchange with the ocean is restricted and hence, below a wind-mixed surface layer, the deeper waters are relatively low in oxygen ( < 1 ml O2/1). The slope and rise fall steeply to a maximum basin depth of about 5000 m. Down to 500 m water depth, hard grounds and biodetrital or terrigeneous sediments predominate. Between 500 and 4000 m there are muds and oozes of different compositions. The sedimentation rate in these areas varies between 5 and 20 cm/1000 yrs. In the basin plain, pelagic and turbiditic sediments alternate, sedimentation rates are up to 100--200 cm/1000 yrs where turbidites are dominant. The surface sediments contain between 1 and 4% Corg, decreasing to lower values deeper within the deposits. The slope and rise sediments are totally bioturbated whereas the more rapidly deposited basin plain sediments show only 10-30% bioturbation. In contrast to this decrease in the degree ofbioturba-
Model of the bioturbation process In general, two types of biogenic structures have been distinguished: (1) biogenic traces (trace 'fossils') show a definite shape and sharp contours which allow classification in terms of palaeontological nomenclature; (2) biodeformational structures have no sharp contours and are not characterized by a distinct shape, but simply destroy formerly existing structures. They represent a lesser degree of behavioural specialization than do biogenic traces (Wetzel 1981). In Fig. 2 only the biogenic traces found preserved within the sediments are shown. For evaluation of the ecological significance of a trace fossil community, the interdependencies between the various types of burrows within the sediment have to be considered. The preservation of biogenic traces is mainly controlled by two factors: (1) displacement of previous structures by deeper penetrating burrows and (2) destruction of burrows by compaction or diagenetically formed minerals. In this context only the first case will be considered.
FIG. 2. Commonly-occurring biogenic traces in the sediments investigated. Planolites, seldom branched, straight or gently curved, inclined or horizontally-oriented tubes. The five types shown differ in sediment fill and diameter. Miinsteria, seldom branched, steeply to slightly inclined tubes showing a crescentic back-fill structure consisting of coarse-grained particles, type A 7-12 mm and type B 2-6 mm in diameter. Scolicia, a bilaterally symmetrical burrow with typical internal structure consisting of pinnate, 'watch-glass' shaped lamellae composed of coarser material. Skolithos, straight unbranched tubes more or less perpendicular to the bedding. Chondrites, a three dimensional tube system~ downwards ramifying at angles of 30::-60'~:,five types (A to E) distinguished on basis of diameter and sediment fill. Hehninthopsis, an irregularly curved tube system with some blind ending protrusions, filled with finer-grained particles. 'Mycellia' (sensu Blanpied & Bellaiche 1981), very thin, often branched burrow systems, type A 0.6-1.0 mm and type B 0.1-0.5 mm in diameter.
Bioturbation in deep-sea.fine-grained sediments
599
?Protopaleodictyon, a quite variable meandering trace with ramifications forming irregular faint polygonal patterns.
Thalassinoides, a tube system consisting of horizontal networks connected to the surface by vertical shafts, swellings occur at points of branching and elsewhere. Trichichnus, a threadlike and rarely branching cylindrical burrow up to 20 cm in length. Lophoctenium, resembles bunches of closely spaced inwardly bent 'twigs' with comblike branches that join to form a main axis; general view and detail. Phycosiphon, consists of small U-shaped spreiten which are arranged in an antler-like system, mostly oblique to the bedding, type A 0.8-1.5 mm and type B 0.2-0.5 mm marginal tube diameter. Teichichnus, resembles stacked roof gutters with pipe on top. Zoophycos, spreiten structures surrounding a central axis in distinct levels or coilings. These burrows show a regular backfill structure that appears as crescentic en-echelons in vertical sections.
6oo
A. Wetzel
In continuously accumulating, oxygenated deep-sea sediments, a tiering within the bioturbated zone can be observed (Wetzel 1979), and several bioturbational levels can be distinguished by their inventory of biogenic traces (Fig. 3). Under steady-state conditions the bioturbated zone is moving upward corresponding to the sediment buildup. In this way, shallow burrows are destroyed by deeper-penetrating ones, such as Zoophycos, which then predominate in the fossil record. Consequently, the chance for a biogenic trace to be preserved mainly depends on the burrowing rate within deeper levels. This process has been modelled by Werner & Wetzel (1982). The distribution of biogenic traces with sediment depth reflects physical and behavioural adaptations of organisms to (1) the mechanical properties of the deposits and (2) the sparser and less decomposable benthic food content deeper in sediment. Therefore, the depth within sediment and the thickness of each level of bioturbation depends on environmental conditions, such as the oxygen content of the bottom and pore water, the mechanical properties of the substratum (sediment texture and composition), and the benthic food content. The benthic food content depends on primary production, sedimentation rate, and water depth and decreases with both lowering of primary production and sedimentation rate, and increasing water depth (Miiller & Suess 1979; Suess
1980). Non-turbidite sedimentation rate normally decreases with distance from the coast. Consequently, the thickness of the bioturbated zone and the thickness of each bioturbation level decreases towards the deep sea. In extremely slowly deposited sediments (< 1-2 cm/1000 yrs) individual levels may be condensed, forming a surface layer and a deeper level (Wetzel 1981). The zonation into bioturbational levels is welldeveloped in slope and rise sediments off NW Africa (Fig. 3) where five levels have been distinguished (Wetzel 1981). (1) The top layer is homogenized by epifaunal and small infaunal organisms, it may contain the highly organized burrows known as graphoglyptids (e.g. Palaeodictyon, Fuchs 1895) that are adapted to fine-grained deep-sea sediments limited in nutrients (Seilacher 1977b), and also some simpler forms (classification of Seilacher 1977 a). (2) The Scolicia level is dominated by ScOlicia, sometimes associated with vertically oriented tubes and/or U-shaped burrows. (3) The Planolites level contains Planolites type A, B, and/or C. (4) The Chondrites-Helminthopsis-Lophoctenium level is defined by these burrows. Additionally, Planolites type D and E, Teichichnus, and Thalassinoides may occur. (5) The Zoophycos level is defined by this burrow; sometimes Trichichnus may be associated.
FIG. 3. Schematic diagram of vertical distribution of biogenic traces within the burrowed zone. This ichnocoenose is typical for water depths of 2000-3500 m off NW Africa. See text for discussion.
Bioturbation in deep-sea fine-grained sediments However, these five bioturbational levels are not always well-developed, because the burrowing activity of the infaunal animals is controlled by environmental conditions that change with time. In this way, burrowing within a certain level may be favoured, whereas the burrowing rate within other levels may become restricted.
Environmental and ichnofaunai changes Interrelationships between ichnofaunas and their environment have previously been described qualitatively (Crimes 1977; Wetzel 1981), but attempts at quantitative evaluations have been rare. This is partly due to difficulties in measuring environmental parameters in the deep sea. For instance, absolute data of oxygen content on the modern sea floor are sparse and most of them have not been determined in connection with the associated biogenic traces (Wetzel 1983). Variations of benthic food content are also difficult to quantify, because most trace-producing animals
are unknown, and hence little is known about the chemical compounds used by them. Therefore, only variations of sediment texture which can more easily be determined with respect to the timing of bioturbation will be considered here. The two aspects of particular interest are, firstly, the variation of ichnocoenoses within the same ichnofacies and, secondly, ichnofacies changes. For this purpose, two processes responsible for changes of sediment texture have been studied: the deposition of aeolian dust and of turbidites. The influence of aeolian dust deposition has been investigated in sediments from the NW African continental margin, while turbidites occur in both the Sulu Sea Basin and off N W Africa. Aeolian dust
A higher input of aeolian dust leads to a coarsergrained sediment (silt-sand), dilution of carbonate and, perhaps, an increased deposition rate (Fig. 4). Consequently, ichnofauna adapted to fine-grained substrata, e.g. Planolites burrows, biogenic sedimentary structures 20 (schematic)
age
groin size distribution [%] carbonate content [%] 50
0
1.2 I . ~ . , . ,
0
.,.1.,
.,i~
t\', ]/', 1/
10 ....
I ....
./
/
/
3--
x
\
/
/
sedimentation rate [cmll.000 years] 0 10 20
--12.000--
i!
o
years B.P
I
1
/./
11
I ....
i
"-
a'
2-
I ....
6oi
--13500--
I
i
--18.000--
~ .......
\
I 9 < 2 ;Jm on carbonate § 2-20;Jm free base x> 20,urn
--22000-C: Chor~drites T:Teichichnus P: Planolites S:
Z:Zoophycos
Scolicia
FIG. 4. Ichnofaunal changes due to variations in sediment texture, Core 347, N.W. African margin, 2576 m depth. See text for discussion.
602
A. Wetzel
are replaced by Scolicia traces which preferentially burrow in the coarser sediments. These changes in ichnofauna occur within the Zoophycos ichnofacies. The adaptation of infauna to a coarser grainsize needs a certain time interval. In the case of adequate time for adaptation, ichnofauna may change in correspondence with the environmental variations. Off N W Africa, a criterion for fully adapted ichnofaunal change is that the organisms are able to maintain about 100~ bioturbation.
Consequently, the threshold rate of change in sediment parameter, below which the fauna can fully adapt to this change, can be evaluated if the degree of bioturbation is slightly, but definitively lowered after the environmental changes have taken place (Fig. 5). In the deep sea o f f N W Africa, the ichnofaunal change occurs within 3-6 cm of sediment at an average sedimentation rate of about 20 cm/1000 yrs (Fig. 4). Considering the vertical mixing effect by Scolicia-producing organisms (Wetzel 1981)
FIG. 5. Partially unburrowed aeolian dust deposits off NW Africa, Core 239 (119-134 cm depth) X-radiograph radiograph (negative) scale 3 cm; (a) Scolicia, (b) Helminthopsis, (c) Zoophycos, (d) undisturbed silt layer.
Bioturbation in deep-sea fine-grained sediments these changes have been taking place in an absolutely shorter (about 1/3) time interval. Thus, the coarsening of sediment texture due to aeolian dust input might have taken place in less than 200 yrs at an average sedimentation rate of about 20 cm/1000 yrs. During this time the median grain-size (of non-carbonate components) increases from 5 #m to 35 #m (Fig. 4). As shown in Fig. 5 such aeolian dust layers are not always completely bioturbated. Thus, the threshold rate for environmental and ichnofaunal changes appears to have been reached. Consequently, the time interval necessary for adaptation to a coarser-grained substratum can be estimated at 100-200 yrs.
Turbidites The influence of turbidites on grain-size and other parameters is markedly different from that of aeolian dust in that: (1) turbidites are deposited very rapidly; (2) a comparatively thick sediment layer is built-up during a short time; (3) turbidites normally contain a coarser and/or wider grainsize spectrum than aeolian dust deposits, and (4) a turbidity current might have eroded the surface sediment layer more or less effectively. Therefore, in a biotope influenced by turbidite deposition, rather different behavioural and physical adaptations of infaunal organisms are found. The biogenic traces occurring in turbidites and associated sediments appear to be preferentially adapted to fine-grained sediments, to a wide grain-size spectrum, or to coarse-grained deposits. The relationship between these three burrow categories within an ichnocoenose is determined by the recurrence time of turbidity current events and the non-turbiditic sedimentation rate.
Burrows adapted to fine-grained sediments The deep-burrowing forms such as Chondrites, Planolites, Teichichnus, and Zoophycos are all adapted to fine-grained sediments. So too are the graphoglyptid traces that show a highly-developed feeding behaviour and characterize the Nereites ichnofacies (Seilacher 1967). These burrows are commonly mucus-lined hollow smalldiameter tube systems. Highly-developed forms such as Palaeodictyon, have multiple openings at the sediment surface for ventilation and are therefore restricted to the upper layer. Because graphoglyptids are hollow, they can easily be destroyed by compaction and therefore require special conditions for preservation (Seilacher 1977 a). They are commonly restricted to areas where the erosional effects of turbidity currents are low, or where bioturbation and
603
tiering of organisms within the sea floor sediments is reduced. Besides their infilling by turbidites, in slightly compacted sediments two cases can be identified: (1) The non-turbiditic sediment accumulation is low; in which case the sediment build-up between two turbidite events will be too small to establish a second bioturbational level below the top level (about 3-5 cm, Berger et al. 1979; Wetzel 1979) and graphoglyptids may predominate. (2) The non-turbiditic sediment accumulation rate is high; in which case graphoglyptids will co-occur with biogenic traces of deeper levels and, in some cases, those adapted to a wider grain-size spectrum. The sedimentation rate is sufficiently high (>40 cm/1000 yrs) that the underlying levels are only incompletely burrowed and hence biogenic surface traces are also found deeper within the sediment. However, observations on such a deep-sea ichnocoenose are still lacking.
Burrows adapted to a wide grain-size spectrum In sedimentary sequences where turbidites are common, biogenic traces occur that are adapted to a wider grain-size spectrum, for example, Miinsteria, Phycosiphon, Scolicia, and Thalassinoides (~ Granularia). Because these forms are normally lacking in continuously accumulated fine-grained deep-sea sediments, they may represent forms transported from the shelf or the slope to deeper environments (Crimes 1977; Wetzel 1981). A significant occurrence of these burrows mainly depends on the recurrence time of turbidite events rather than on the spacing between such layers. This has beendeduced from comparisons between the deposits off NW Africa and the Sulu Sea. Off NW Africa, thick (2-36 cm) turbidites are found separated by non-turbidite intervals of 3-20 cm thick that represent periods of at least 1500 yrs (assuming a sedimentation rate of about 2 cm/1000 yrs). Burrows adapted to a wide grain-size spectrum are sparse. This can be also estimated by calculation of population density based on the relation between frequency of burrows and the sediment volume studied (Wetzel & Werner 1981). In the Sulu Sea Deep, 2-4cm thick turbidites are more frequent and the spacing between them is 10-60 cm. Based on absolute age determinations (Erlenkeuser; Kiel, unpubl) the recurrence time of turbidity currents is 100-700 yrs, or about 170 yrs on average. Within these deposits, Miinsteria, Phycosiphon, Thalassinoides, and Zoophycos frequently occur (~>1 burrow m-Z; Fig. 6). However, they do not occur so commonly where the turbidites have a rela-
6o4
A. Wetzel
8~
eq'~
S~
~o
I~
,,~,,,~ " , ~
.-~
s
~g .~ 9 ~,.~
Bioturbation in deep-sea fine-grained sediments tively long recurrence interval as in the examples off N W Africa. With respect to these observations, two aspects have to be considered in some more detail: establishing and maintaining of a formerly exogenous fauna adapted to a wide grain-size spectrum. The critical population density to establish an exogenous fauna in the deep-sea may be reached when the imported organisms could reproduce in the deeper environment. Because a turbidity current derives animals from a small source area and spreads them over a large depositional area, only a low population density of exogenous organisms can be established by any one turbidity current. Therefore, the probability of reproduction will be raised when turbidity currents repeatedly occur during the life-time of the imported organisms. Assuming a life-time of 50-100 yrs for deep-sea macrobenthos, as Turekian et al. (1975) determined for some deep-sea animals, a turbidity current frequency of 20-200 events/1000 yrs seems to be adequate to establish a reproducing, but formerly exogenous, fauna adapted to a wide grain-size spectrum. On the other hand, to maintain a fauna as discussed above, a lower frequency of turbidity currents may be sufficient. Based on the observations on the Sulu Sea deposits, a frequency of 0.5 to 2 turbidity currents/1000 yrs seems to be adequate if the sedimentation rate of finegrained, non-turbiditic material does not exceed 50 cm/1000 yrs.
605
Environmental changes and ichnofauna diversity As shown above, the recurrence time of turbidity currents and the non-turbiditic sedimentation rate are two important controls in the composition of ichnofauna in deep-water environments. Besides a quantification of the boundary conditions for the various groups of burrows, the effects of these parameters on biogenic trace communities should be expressed by diversity values. Sediments o f f N W Africa and offthe Sulu Sea will be compared and, to complete the environmental and bioturbational trends from the slope to deep sea, deposits of the Central Pacific Ocean are also discussed (Fig. 1 and 7). In upper slope sediments of both areas, diversity is high, and biodeformational structures are replaced by more specialized biogenic traces (Wetzel 1981). In middle-slope to rise sediments a nearly constant number of ichnogenera has been observed, although in the Sulu Sea the ichnofauna within this depth range is less diverse than o f f N W Africa. This may be due to the low oxygen content and/or to the less variable climatic conditions in the Holocene sediments of the Sulu Sea. Towards the deep sea, both coastal upwelling and fluvial and/or aeolian sediment input are less and the biotope becomes increasingly uniform with distance from coast. In such an environment an optimal fractionation into ecological niches can take place due to a long-term adaptation of
wafer degree of biofurbotion (%) number of biogenic depth NW-Africo/PQcific Sulu Sea trace types preserved (km) 0 50 100 0 50 100 0 5 10 15 |11111111 :::::::::::::::::::::::::::::
1.0
!ii!iiiiiiii!iiiiiiiiiiiii
1.5
iiiiiiiiiiiiiiiiiiiiii
2.0 2.5
I
3.0 3.5 /,.0 4.5 5.0 m
~
I
/ Pocifi/
Sulu Sea
FIG. 7. Degree of bioturbation and diversity of preserved biogenic trace types in sediments off NW Africa, in the Central Pacific Ocean, and in the Sulu Sea Basin. See text for discussion.
6o6
A. Wetzel
the fauna (Seilacher 1977a). In fact, abyssal surface sediments contain a comparatively high number of ichnogenera if grapholyptids occur (Kitchell 1978; Ekdale 1980). However, biogenic trace diversity may be secondarily reduced due to destruction of these shallow burrows by deeper penetrating ones. In this way a monotonous low-diversity ichnofauna may reach the fossil record, dominated by Chondrites, Planolites, (Teichichnus), and Zoophycos, and sometimes with 'composite burrows' of Helminthoida/Helminthopsis (Chamberlain 1975). This type of ichnocoenose has been observed in DSDP cores (e.g. Chamberlain 1975; Ekdale 1977) as well as in modern sediments (e.g. Ekdale 1978; Berger et al. 1979), and corresponds to the Zoophycos ichnofacies (Seilacher 1967). It is typical for continuously accumulating deep-sea deposits without major changes of environment and sediment texture. It probably also occurs in rise to slope sediments accumulating under similarly uniform environmental conditions. Although the morphology of Zoophycos varies widely, a depth zonation of different forms has not been observed (Wetzel & Werner 1981). In areas of extremely slow accumulation, the bioturbated zone becomes very thin as organic matter cannot be buried deeply within the sediment (M/iller & Suess 1979). In this case, Zoophycos may also be absent, because there is no food available in the sediment depth where Zoophycos normally occurs (Wetzel & Werner 1981). On the other hand, Zoophycos is also lacking in sediments deposited more rapidly in water depth above 2000-1000 m off NW Africa and the Sulu Sea. Thus, in both extremely slowly accumulating and upper slope deposits the index burrow of the Zoophycos ichnofacies is absent making an ichnological interpretation difficult. In the Sulu Sea Deep, the high frequency of turbidites prevents complete bioturbation and steady state sedimentation so that shallow-penetrating biogenic traces may also be preserved and a high-diversity ichnofauna may reach the fossil record. Such highly diverse, graphoglyptid-bearing ichnocoenoses belong to the Nereites ichnofacies which therefore is typical for discontinuous sedimentation due to turbidites. This comparison shows that different ichnofacies may occur in the same range of water depth. Their occurrence seems to depend mainly on the presence or absence ofturbidites. Thus, the use of ichnofacies for palaeobathymetric analyses is restricted in abyssal plain sediments. Assuming sufficient supply of benthic food and oxygen, medium burrow diversities (10_+ 5 biogenic trace types) occur abundantly in slope to rise sediments, with the exception of the upper
slope and canyons. In the deep sea, both extremely low diversity values (~< 5 biogenic trace types) occur in totally bioturbated deposits and very high values ( >~ 15 biogenic trace types) occur in incompletely reworked sediments containing turbidites. Variations or organic matter and/or oxygen content, and of sediment texture may influence these values.
Conclusions (1) Biogenic trace assemblages can be best interpreted using a tier model with a variable number of distinctive bioturbation levels. Thus, the preservation potential of a biogenic trace mainly depends on the burrowing rate at deeper levels within the sediment where buried shallow burrows are replaced by deeper-penetrating burrows. (2) The response of ichnofauna to environmental changes requires a certain although very short time interval. An adaptation to a coarser sediment texture due to input of aeolian dust (coarse silt) takes place within < 100-200 yrs in rise to abyssal plain sediments. This value also seems to be applicable to other environmental changes, if the 'new' ichnofauna occurs latently. (3) Highly-developed near-surface biogenic traces, so-called graphoglyptids, can best reach the fossil record in turbidite environments. Infauna adapted to a wide grain-size spectrum seems to be imported from shelf to slope sediments, because these deposits represent a similar grain-size spectrum even if turbidites are intercalated. It can be assumed that an abundant occurrence of these burrows in the deep sea depends on recurrence time of turbidites rather than of the spacing between individual coarsegrained layers. For a significant and permanent occurrence of new, perhaps exogenous, (ichno-) faunal elements, an autochthonous reproduction of these seems to be necessary. This can be best achieved if repeated 'import' events occur during the life-time of such organisms. To establish a fauna adapted to a wide grain-size spectrum, a value of 20-200 turbidites/1000 yrs has been assumed. To maintain this fauna 0.5-2 turbidites/1000 yrs. seem to be adequate. (4) Biogenic trace diversity profiles from slope to the abyssal plain show medium diversities for middle slope to rise sediments, 10_+5 burrow types; variations in the intensity of coastal upwelling or of fluvial or aeolian sediment input, may raise or lower this number. In abyssal plain sediments, extreme values of ~<5 and >~15 burrow types have been found. They correspond to two basic types of bioturbation: (a) Low-diversity Zoophycos ichnofacies (slope to abyssal plain);
Bioturbation in deep-sea fine-grained sediments under stable environmental conditions, slowly and continuously-accumulating sediments are completely reworked by organisms. Thus, the fossil record is d o m i n a t e d by deeply-penetrating biogenic traces, such as Chondrites, Planolites, (Teichichnus), and Zoophycos. (b) High-diversity Nereites ichnofacies (rise to abyssal plain?); in turbidite-dominated sequences the continuous bioturbation process is interrupted by coarse sediment layers preventing total bioturbation. Thus, near-surface biogenic traces may also reach the fossil record. (5) Consequently, the use of biogenic traces for bathymetric interpretation of the deep-sea sediments is restricted.
607
ACKNOWLEDGEMENTS: The investigations on the sediment cores have been carried out in the laboratories of the Geologisches Institut der UniversitS.t Kiel. E. Seibold (now Bonn) and F. Werner (Kiel) advised and contributed valuable help and stimulating discussion during investigations. P. Ballance (Auckland) and M. Pye (Britoil, Glasgow) critically reviewed the manuscript, and D. Stow (Edinburgh) made helpful c o m m e n t s on the text and carefully improved the English. Financial support for an earlier phase of these investigations was given from the Deutsche Forschungsgemeinschaft. All these contributions are gratefully acknowledged.
References BERGER, W.H., EKDALE,A.A. & BRYANT, P.P. 1979. Selective preservation of burrows in deep-sea carbonates. Marine Geol., 32, 205-30. BLANPIED, C. & BELLAICHE,G. 1981. Bioturbation on the Pelagian platform: ichnofacies variations as paleoclimatic indicators. Marine Geol., 43, M49-M57. BROMLEY,R.G. & ASGAARD,U. 1975. Sediment structures produced by a spatangoid echinoid: a problem of preservation. Bull. geol. Soc. Denmark, 24, 261-81. CHAMBERLAIN,C.K. 1975. Trace fossils in DSDP cores of the Pacific. J. Paleontol., 49, 1074-96. CRIMES,T.P. 1977. Trace fossils of an Eocene deep-sea sand fan, northern Spain. In: Crimes, T.P. & Harper, J.C. (eds), Trace Fossils 2. Geol. J., Spec. Iss., 9, 71-90. EKDALE,A.A. 1977. Abyssal trace fossils in the worldwide Deep Sea Drilling Project cores. In: Crimes, T.P. & Harper, J.C. (eds), Trace Fossils 2. Geol. J., Spec. Iss., 9, 163-82. -1980. Graphoglyptid burrows in modern deep-sea sediments. Science, 207, 304-6. -& BERGER, W.H. 1978. Deep-sea ichnofacies: modern organism traces on and in pelagic carbonates of the western equatorial Pacific. Palaeogeogr., Palaeoclimatol., Palaeoecol., 23, 263-78. EXON, N.F., HAAKE,F.-W., HARTMANN,M., KOGLER, F.C., M(ILLER, P.J. & WHITICAR,M.J. 1981. Morphology, water characteristics and sedimentation in the silled Sulu Sea, Southeast Asia. Marine Geol., 39, 165-95. FUCHS, T. 1895. Studien fiber Fukoiden and Hieroglyphen. Denkschr. Akad. Wiss. Wien, 62, 369-448. HANTZSCHEL, W. 1965. Vestigia Invertebratorum et Problematiea; Fossilium Catalogus I." Animalia pars 108. W. Junk, s'Gravenhage. 142 pp. -1975. Trace fossils and problematica. In: Teichert, C. (ed.), Treatise on Invertebrate Paleontology; Part
W. Miscellanea, Supplement 1. Geol. Soc. Am., New York and Univ. Kansas Press, Lawrence, XXXI, 269 pp. K1TCHELL, J.A., K1TCHELL, J.F., JOHNSON, G.L., & HUNKINS, K.L. 1978. Abyssal traces and megafauna: comparison of productivity, diversity and density in the Arctic and Antarctic. Paleobiology, 4, 171-80. KOOPMANN, B. 1981. Sedimentation yon Saharastaub im subtropischen schen Nordatlantik wfihrend der letzten 25.000 Jahre. 'Meteor' Forsch.-Erge., C, 35, 23-59. MULLER, P.J. 1975. Diagenese stickstoffhaltiger organischer Substanzen in oxischen und anoxischen marinen Sedimenten. 'Meteor' Forsch.-Erg., C, 22, 1-60.
--
& SUESS,E. 1979. Productivity, sedimentation rate and sedimentary organic matter in the oceans. I. Organic carbon preservation. Deep-Sea Res., 26, 1347-62. SARNTHEIN, M., THIEDE, J., PFLAUMANN,U., ERLENKEUSER, H., F1]TTERER, D., KOOPMANN,B., LANGE, H. & SEIBOLD, E. 1982. Atmospheric and oceanic circulation patterns offNorthwest Africa during the past 25 million years. In: von Rad, U., Hinz, K., Sarnthein, M. & Seibold, E. (eds), Geology of the Northwest African Continental Margin. SpringerVerlag, Berlin, Heidelberg. 545-604. SEIBOLD, E., DIESTER-HAASS,L., F~TTERER, D., HARTMANN, M., KOGLER, F.-C., LANGE, H., MULLER, P.J., PFLAUMANN,U., SCHRADER,H.J. & SUESS, E. 1976. Late Quaternary sedimentation off the western Sahara. An. Acad. bras. Cient., 48, supl., 287-96. SEILACHER,A. 1967. Bathymetry of trace fossils. Marine Geol., 5, 413-28. 1977 a. Pattern analysis of Paleodictyon and related trace fossils. In: Crimes, T.P. & Harper, J.C. (eds), Trace fossils 2. Geol. J., Spec. Iss., 9,289-334. - 1977 b. Evolution of trace fossil communities. In:
6o8
A. Wetzel
Hallam, A. (ed.), Patterns of evolution. Developments in Paleontology and Stratigraphy, 5. Elsevier, Amsterdam. 359-76. SUESS, E. 1980. Particulate organic carbon flux in the oceans--surface productivity and oxygen utilization. Nature, 288, 260-3. TAIT, R.V. 1971. Meeres6kologie. Thieme, Stuttgart. 305 pp. TUREK1AN,K.K., COCHRAN,J.K., KHAKAR,D.P., CERRATO, R.M., VAISNYS, J.R., SANDERS, H.L., GRASSLE,J.F. & ALLEN,J.A. 1975. Slow growth rate of a deep-sea clam determined by 228Ra chronology. Proc. Nat. Acad. Sei. USA, 72, 2829-32. WERNER, F. 8/; WETZEL, A. 1982. Interpretation of biogenic structures in oceanic sediments. Bull. Inst. Geol. Bassin d'Aquitaine, 31, 275-88. WETZEL, A. 1979. Bioturbation in spdtquartiiren Tiefwasser-Sedimenten vor N W Afrika. Ph.D. Thesis, Univ. Kiel. 111 pp.
-
-
-
-
- -
- -
1981. Okologische und stratigraphische Bedeutung biogener Gef/ige in quart/iren Sedimenten am NW-afrikanischen Kontinentalrand. 'Meteor' Forsch.-Erg., C, 34, 1-47. 1983. Biogenic sedimentary structures in a modern upwelling area: the NW African continental margin. In: Thiede, J. & Suess, E. (eds), Coastal Upwelling: and Its Sediment Record, Part B, Sedimentary Records of Ancient Coastal Upwelling. Plenum Press, New York. 123-44. • WERNER, F. 1980. Biogene Sedimentgef/.ige. In: Werner, F. Fahrtbericht "Meteor'-Reise 53 C/DOstatlantik vor Marokko. Unpubl. Report, Geologisch-Palfiontologisches Institut der Universit/it Kiel. 20-2. & WERNER, F. 1981. Morphology and ecological significance of Zoophycos in deep-sea sediments off NW Africa. Palaeogeogr., Palaeoclimatol., Palaeoecol., 32, 185-212.
A. WETZEL, Geologisches Institut der Universit/it, Sigwartstrasse 10, D7400 T/ibingen, West Germany.
Deep-water fine-grained sediments: facies models D.A.V. Stow and D.J.W. Piper S U M M A R Y: Based on a large amount of published data and stimulated by the papers and discussion at the International Workshop on Fine-Grained Sediments held in Halifax, Canada in August 1982, we have attempted a synthesis of deep-water fine-grained sediment facies. Three main facies groups related to depositional processes can be identified: turbidites, contourites and pelagites/hemipelagites. There is a continuum between the different processes and hence a continuum between facies. Nevertheless, it is possible to define several distinct facies models within each of these groups on the basis of sedimentary structures, texture and composition, and to provisionally interpret these in terms of depositional hydrodynamics. Patterns of horizontal and vertical facies distribution can be related to depositional subenvironments. There is much variability within and departure from the facies models we propose, and many interesting and problematic areas of research remain in the quest for better understanding of deep-water fine-grained sediments.
Only five years ago it was fair to say that fine-grained sediments were grossly understudied by comparison with, for example, sands and gravels. However, in the last few years, a large amount of data has appeared on both modern and ancient clays, muds and silts and their biogenic equivalents (e.g. Potter et al. 1980). A variety of fine-grained facies types have been recognized on the basis of size, grade and internal organization of beds (e.g. Piper 1978; Stow 1984a; in press), and new questions have emerged regarding their origin, processes and environments of deposition. However, there is still need for a general synthesis or organization of these data into composite facies models. Such models may then serve as a framework to guide description, a standard against which to compare new information, a predictive tool in new geological settings and as a basis for hydrodynamic interpretation (Walker 1975). There have been several recent attempts to construct facies models for deep sea sediments, including turbidite muds and silts (Piper 1978; Stow & Shanmugam 1980), 'homogeneous' muds (Stanley 1981), carbonate turbidites (Hesse 1975; Stow, Wezel et at., this volume), muddy contourites (Stow & Lowell 1979; Stow 1982; Faug6res et al. 1984; Gonthier et al. this volume), hemipelagities and pelagites (Hoffert 1980; Einsele 1982; Jenkyns, in press; Thornton, this volume; Hill, this volume). There have been similar attempts to collect together the diverse data on more shallowwater fine-grained sediments (e.g. tidal deposits, Ginsburg 1975; deltaic deposits, Coleman 1976; shelf deposits, Shepard et al. 1960): these are not considered further here. In this paper we attempt, therefore, to collate information on deep-water fine-grained sediments from our own studies, from a growing body of literature, and from the papers and
discussion at the International Workshop on Fine-Grained Sediments (Halifax, Canada, 1982) from which this volume has been edited. We recognize and describe various facies models within three main facies groups: fine-grained turbidites, muddy contourites, and pelagites/ hemipelagites; and briefly discuss their interpretation in terms of depositional processes. We then consider the distribution of these facies in different environmental settings and some of the post-depositional changes that they undergo. Our discussion highlights some of the interesting and problematic areas of interpretation. There is by no means a concensus view on the description and interpretation of fine-grained sediments. Many readers will feel we are premature in defining some of the facies models and are guilty of generalization in the discussion of their horizontal and vertical distribution in the deep sea. However, we believe that some degree of synthesis and simplification is necessary and that the conclusions are useful. Certainly, they should be the subject of rigorous scrutiny by future research.
Processes and facies There appears to be a continuum of processes operating in the deep sea (e.g. Walker 1978; Stow 1984a). These include resedimentation (mass gravity) processes, normal bottom currents and pelagic settling (Fig. 1). There is close interaction and overlap between these processes both during transport and during the final stages of deposition. Any single resedimentation event, for example, may be initiated by the slumping of unstable slope sediments and then, by mixing with seawater, evolve into a debris flow, a high concentration turbidity current and, finally, a low-concent6II
D.A.V. Stow and D.J.W. Piper
612 TIME AND/OR SPACE
ROCKFALL
CREEP
SLUMP
DEBRITE
TURBIDITES + ASSOCIATED FACIES Collr l e medium fine
CONTOURITE
PELAGITE HEMIPELAGITE
ROCKFALL
~ SL,DE~~
CREEP
SLUMP REMOLDING SPILLOVER ~
LIOUEFACTION
---.___
1 SHELF S USPE NSION
~
FLUID TURBUL EN C E
L~176176
SURFACE-WATER PRODUCTIVITY AEOLIAN INPUT
"'2EP20~;2rA::: E" + SURFACE SUSPENSION ~, CURRENTS
FLOW
INITIATION
LONG-DISTANCE
DEPOSITIONAL
FACIES
MODELS
TRANSPORT
FIG. 1. Diagram showing the main processes of flow initiation and long distance transport in the deep sea and the depositional facies models related to each distinct process. In nature, there is a continuum of processes and facies. (Modified after Walker 1978). ration turbidity current. Deflection of the lowvelocity dilute tail of a turbidity current by a regional contour-following bottom current will cause the downslope flow to grade imperceptibly into an alongslope bottom current. Pelagic sediment settling vertically through the water column in the open ocean may be deflected slightly by the action of very weak bottom currents. The mechanisms involved in erosion, transport and deposition of fine-grained sediments within this process continuum are discussed more fully in Section 1 of this volume (papers by McCave, Gorsline, Kranck, Eittreim, all this volume). We recognize that there is also a facies continuum in the deep sea but that it is nevertheless possible to distinguish facies types that are the result of a distinct turbidity current, bottom current and pelagic processes. Our facies models are, therefore, closely related to the depositional process (Fig. 1). We have not attempted to include facies models for slumps, sediment creep deposits or debris-flow deposits ('debrites'), although these commonly involve fine-grained sediments (but see Naylor 1980, 1981; Thornton, this volume; Stow 1984 a,b, in press.). Several other facies types have been proposed in the literature, resulting from processes such as turbid layer flows (Moore 1969; Stanley and Maldonado 1981), suspension cascading
(McCave 1972), unifite flows (Stanley 1981), nepheloid layers (Biscaye & Eittreim 1977), canyon currents (Drake et al. 1978) and so on. These do not, however, differ significantly either in character or in the depositional process from the broader categories that we suggest and do not appear valid to us as distinct facies. They do serve to indicate the range of mechanisms that may exist within our general process groups. Black shales may be deposited by a variety of processes and so are not considered as a separate facies in this context. However, several of the papers in this volume describe organic-rich sediments of different types and deposited by different processes (e.g. Crevello et al.; Isaacs; Anastasakis & Stanley; Thickpenny, all this volume). Arthur et al. (also this volume) have attempted to synthesize a more generalized model for the black shale sedimentation. Neither do we include a detailed discussion of the acoustic character of deep-sea sediments and the recognition of acoustic facies: useful summaries are given by Damuth (1975, 1978), Jacobi (1982) and Nardin et al. (1979).
Fine-grained turbidites Fine-grained turbidites are made up of material
Deep-water fine-grained sediments. facies models dominantly in the silt and clay size grades (i.e. over 50~ less than 63 /~m grain size). They are widespread in the deep sea and, volumetrically, the most important of our facies groups (Piper 1978). They occur as very thin to very thick beds that have been deposited rapidly (from a few hours to a few days) from a single resedimentation event. The chief criteria that can be used to distinguish them from other facies in the deep sea which they may resemble, include: (1) a regular vertical sequence of sedimentary structures commonly associated with a positive grading; (2) the presence of sedimentary structures indicating rapid deposition, with bioturbation restricted to the tops of beds; (3) compositional, textural or other features which indicate that they are exotic to their depositional environment. With high resolution acoustic profiling systems, fine-grained turbidites show good penetration and appear well stratified, in contrast to the hard, often irregular, reflective bottom found in coarser turbidites. Fine-grained turbidites occur in a variety of environments that can be distinguished in seismic reflection profiles, such as levees and ponded basin plains. They may pass laterally into coarser turbidites that often fill channels and appear as discontinuous strong reflectors in seismic reflection profiles. In Bouma's (1962) classical sequence for sandmud turbidites, all fine material was classed as a featureless E division. Later work distinguished between turbiditic and pelagic mud (Kuenen 1964; Van der Lingen 1969) and Piper (1978) further subdivided the turbidite mud into El, E2 and E3 structural divisions. Fine-grained turbidites that occur alone are, in some cases, the distal equivalents of thicker-bedded, coarser-grained turbidites, but they may also occur without the proximal sand-mud 'parents'. On the basis of grain size, internal organization and composition we recognize four distinctive facies within this group: silt turbidites, mud turbidites, biogenic turbidites and disorganized turbidites. Each of these can best be described in terms of a separate facies model. Silt turbidites
6I 3
Well-documented examples include those of the Aleutian Trench (Piper 1973), the Indus fan (Jipa & Kidd 1974), the Nile cone (Maldonado & Stanley 1976), the Antarctic continental rise (Piper & Brisco 1975), the Zaire fan (van Weering & van Iperen, this volume) and the south-east Angola Basin (Stow 1984c). Similarly, many ancient slope, fan and basin plain successions comprise interbedded siltstone turbidites and mudstones (e.g. Lundegard et al. 1980; Pickering, this volume; Stow, Wezel et al., this volume). Facies model (Fig. 2)
Silt turbidites commonly exhibit the same suite of structures (Piper 1978; Kelts & Arthur 1981) as the thicker-bedded, classical sandy turbidites described by Bouma (1962). A complete sequence from top to bottom would be (Fig. 2): F hemipelagic or pelagic sediment, biogenic, bioturbated E mud, graded, commonly bioturbated D fine silt, parallel-laminated alternating silt and minor clay, often with synsedimentary deformational structures, graded C medium silt, cross-laminated, rarely convolute, graded
PELAGITE
TURBIDITE MUD ___ graded + laminated
D
TURBIDITE SILT graded, fine, parallel-laminated
C
.graded, medium cross-laminated
B
_+ graded, medium parallel-laminated
In distal turbidite environments, silt beds ( > 70~ I massive, medium-coarse, silt-sized particles) are more abundant than sands E A poor or no-grading and commonly occur as thin or medium-bedded ~' sharp -F scoured base turbidites. Early work demonstrated a change from sands to silts moving basinwards from fans to abyssal plains (e.g. Horn et al. 1971; Normark FIG. 2. Silt turbidite facies model. Structural divi& Piper 1972), and there have since been many sions follow Bouma (1962) as modified by van der descriptions of silt turbidites from the deep sea. Lingen (1969)
6I 4
D.A.V. Stow and D.J.W. Piper
FIG. 3. Photographs of fine-grained turbidite facies. Width of sections approximately 7 cm. (a) Silt turbidites, upper Cretaceous, DSDP Site 530A, Angola Basin. (b) Silt and mud turbidites, upper Jurassic, Brae slope apron, North Sea. (c) Mud and ooze turbidites (3 core sections), Plio-Pleistocene, DSDP Site 530B, Angola Basin. (d) Calcirudite-calcilutite turbidite, upper Cretaceous, DSDP Site 530A, Angola Basin. (e) Disorganized silty-mud turbidite, Pleistocene, DSDP Site 530B, Angola Basin. (f) Silt laminated mud turbidites, Cambro-Ordovician, Halifax Formation, Nova Scotia (g) Calcarenite-calcilutite turbidite, upper Cretaceous, Scaglia Rossa Formation, Italy.
D e e p - w a t e r f i n e - g r a i n e d sediments." f a c i e s m o d e l s B medium silt, parallel-laminated, graded A medium-coarse silt or sandy silt, massive, poor or no grading, some floating clasts, minor scouring at base The interpretation and origin of this sequence, in terms of flow decline during deposition from a single turbidity current event, we suggest is similar to that proposed for the Bouma sequence in sandy turbidites (Harms & Fahnestock 1965). However, the complete sequence is very rarely found (e.g. Fig. 3). More commonly, base-cut-out sequences occur comprising CDE and DE divisions (graded sorted silts and laminated silts respectively of Piper, 1978) in beds from 1 to 10 cm in thickness. Thicker-bedded and coarser silt top-cut-out (AB and B) or mid-cut-out (AE) sequences occur more rarely (the ungraded massive silts of Piper, 1978).
615
Facies model
From this large amount of data it is possible to synthesize an ideal facies model for mud turbidites (Fig. 4). Piper (1978) proposed the subdivision of Bouma's (1962) E division in to three parts, giving from top to bottom: F hemipelagic or pelagic sediment E3 ungraded mud E2 graded mud E1 laminated mud D laminated sand and silt Apart from the topmost E3 division, mud turbidites commonly show slight but distinct positive grading in both grain size and composition. Textural grading is best observed as a progressive decrease in maximum or mean size through successive silt laminae (El division) (Piper 1972b), or as a decrease in silt to clay ratio through the E2 division. A wide range of compositional grading has been observed, including Mud turbidites an upwards increase in micas, various clay Deep-sea terrigenous successions are commonly mineral species and organic carbon, and upward dominated by muds and, in many settings, decrease in heavy minerals, quartz and foraminibetween 50% and 80% of this is ofturbidite origin fera (Rupke & Stanley 1975). Carbonate is known (e.g. Hesse 1975; Piper 1978). The features that to show either an increase or decrease depending characterize mud turbidites are subtle and fre- on the type and grain size of the carbonate and quently overlooked, especially when the turbi- associated components. Colour changes comdites are interbedded within a thick monotonous monly mirror the textural and compositional hemipelagic mud sequence. However, clear exam- grading and provide the simplest method for ples have been described from both the east and visual identification of mud turbidites. Stow (1977) and Stow & Shanmugam (1980) west Mediterranean (Rupke & Stanley 1974; Bartolini et al. 1975; Got, this volume), the have shown that there is a further hierarchy of Cascadia Channel (Griggs & Kulm 1970) and structures within the graded laminated mud and Astoria fan (Nelson 1976) in the eastern Pacific, silt of Piper's scheme, many of which can be the Angola Basin (Stow 1984c), the Northwest helpful as diagnostic criteria in describing mud Atlantic Mid-Ocean Channel (Chough & Hesse turbidites. The complete sequence from top to 1980), the Laurentian Fan (Stow 1981) and the bottom is as follows (Fig. 4): Cap Ferret fan in the north-east Atlantic (Cremer P pelagite or hemipelagite, bioturbated 1981, 1983). T8 turbidite (+ part pelagite), microbioturIn ancient sequences exposed on land, weatherbated ing and fracturing often make it impossible to T7 ungraded mud, occasionally with silt discern characteristic structures in mudstones pseudonodules and shales. Where preservation has been good T6 graded mud, often with dispersed silt lenses and rocks have been suitably smoothed, as on Ts wispy convolute silt laminae in mud wave-cut platforms, the same features as deT4 indistinct, discontinuous silt laminae in scribed for modern mud turbidites may be even mud more clearly demonstrated. Examples range from T3 thin, regular, continuous parallel silt the Precambrian slope facies of northern Norway laminae in mud (Pickering 1982a, this volume), through succesT2 thin, irregular, slightly lenticular silt sions from the Cambro-Ordovician (Piper 1972a; laminae in mud, often with low-amplitude Stow, Wezel, et al., this volume), Silurian (Piper climbing ripples 1972b), Devonian-Carboniferous (Hall & StanTI thick mud layer, often with thin convolute ley 1973; Lundegard et al. 1980), Triassic (Hicks silt laminae 1981) and Cretaceous (Ingersoll 1977) to the To thick, basal, lenticular, silt lamina, often well-known Cenozoic turbidite formations of the with fading-ripple top, microlaminated inItalian Apennines (e.g. Mutti 1977; Mutti et al. terior and scoured, load-cast base 1978; Ricci Lucchi 1978, 1981). The complete sequence is interpreted as a single
616
D.A.V. Stow and D.J.W. Piper
P
PELAGITE/HEMIPELAGITE
1-8
bioturbation, microbioturbation
F --
E3
Tr
E2
T6
UNGRADED TURBIDITE MUD
,~
GRADED TURBIDITE MUD
~ h,~
-I- silt lenses
~ . . . . .. .~176
,,
,~176
... ..~176 .~176176
T~
9. o - .
.~
..........
wispy silt. laminae
o...-o.
indistinct silt laminae
la3
.~176 ........
El
T3
regular parallel silt laminae
T2
irregular~ thin lenticular silt laminae
Tt
convolute silt laminae
To
thick basal lenticular silt lamina
5,,}-r'~a3 .~cr (_9t--
FIG. 4. Mud turbidite facies model. Structural divisions after Piper (1978) and Stow (1977). depositional unit from a large fine-grained turbidity current (e.g. Stow & Bowen 1980). Internal lamination in the basal layer (To) and migrating ripple lamination indicate periods of tractional movement during deposition of the coarser silt grains. The silt ripples with muddy troughs (fading ripples) are deposited from turbidity currents containing a high proportion of claysized material. This commonly also gives rise to a thick mud layer (TI), with convolute silt laminae formed either by loading into the soft mud or by incipient ripple development. At a lower current velocity and with less silt available the very thin laminae of low-amplitude, long wavelength ripples (T2) are formed. Continued waning of the current velocity as the flow passes results in the overall positive grading observed (T3-T4) in which the alternation of silt and mud laminae is believed to be due to the depositional sorting of silt grains from clay flocs caused by increased shear in the bottom boundary layer (Stow & Bowen 1978, 1980; but see Hesse & Chough 1980). The more homogeneous mud unit (T7) at the top of the sequence comprises the finest silt and clay which
was not so effectively sorted into distinct laminae because of the lack of silt and the very fine grain size.
Variability Although a standard structural sequence can be identified, a wide range of variations is possible and, as with the Bouma sequence for sandy turbidites, the complete set of divisions is rarely present in any one bed (Figs 3 and 5). In many cases the beds are relatively thin ( < 10 cm), averaging 2-5 cm, and comprise only the upper parts of the sequence (EzE3 or T4-Ts, base-cut-out units), or lower parts of the sequence E 1 or T0-T4, top-cut-out units). In the extreme base-cut-out case we have a mud or clay turbidite with no trace of silt lamination, whereas the top-cut-out case is gradational to a silt turbidite. In other cases, only the middle parts of the sequence occur (E2 or Tz-Ts), or the middle parts are cut out to give an E IE3 (T07s) bed analogous to a Bouma AE turbidite. Repetition of these different bed types in any one area can lead to the development of a
Deep-water fine-grained sediments." facies models .-..-
617
+..'.:.. : : . . : . . - : . : : - .
-. ~...
.......
i"'
- .
.
.
.
.
.
T6
.
. . .....
.
..., .....,
.. .. ....
9. . , ,
"5""''57....
.......... ,. .....
...
..........
TZ,Z;. . . . . . .
I2
FACIES3-
... . . . . . . . .
FACIESl
_.--
~ ' J ~b~O~t~qctc~
T~
IDEAL SEQUENCE
""~" 17
CO?,
FACIES4 - ~ ~ c)UT S,o
T6
r-7 _.___.
T3
FIG. 5. Examples of variability in mud turbidites: typical base-cut-out and top-cut-out sequences (after Stow, Alam & Piper, this volume). distinctive facies, such as thick lenticularly laminated silts and thin parallel-silt-laminated mud, in which individual turbidite units may not be readily distinguished. Thin-bedded turbidites of these various kinds have been widely reported from both modern and ancient successions (e.g. Piper 1978; Nelson et al. 1978; Stow & Shanmugam 1980; Kelts & Arthur 1981; Hill 1981 and this volume; Chough, this volume; van Weering & van Iperen, this volume). In some cases, the thick basal laminae of a graded laminated unit is preceded by one or two thinner, finer grained silt and mud laminae. These appear to be precursors of the main depositional event. In other cases, there is variation in the sequence of To to T8 divisions through a single unit, perhaps related to an instability in the flow conditions. Both medium and thick-bedded mud turbidites have also been described, commonly 10-50 cm thick but ranging up to several metres (Fig. 3) (e.g. Rupke & Stanley 1974; Piper 1978; Blandpied & Stanley 1980; Stanley 1981; Stow 1984c). These show the same sequence of structures (El-E3, T0-Ts) but each division is vertically expanded. They probably result either from large muddy turbidity currents derived entirely from fine-grained sediments (e.g. from a slump on a
muddy upper slope), or from the ponding of a muddy turbidite tail in an enclosed basin (Wezel 1973; Bowen et al. 1984). Stanley (1983) suggests that slow (hemipelagic) settling from detached turbidity currents in a well stratified basin is an important process in the accumulation of beds that he describes as unifites. Similar slow settling from thick dilute turbidity currents may be important elsewhere (e.g. Stow & Bowen 1980). However, there are generally no diagnostic features that allow recognition of depositional processes involving turbidity current detachment. We recognize that the upper parts of mud turbidites may have been deposited by a variety of turbidity-current-related processes.
Biogenic turbidites Biogenic pelagic sediments are very widespread in the open ocean and, where terrigenous input is minimal, along the continental margin. In areas of topographic relief and/or tectonic activity, such as mid-ocean ridges, seamounts and other submarine highs, resedimentation of pelagic siliceous and calcareous oozes occurs via slumping, debris flows and turbidity currents (e.g. Kelts & Arthur 1981). Off carbonate platforms and reef margins, resedimented biogenic material derived
618
D.A.V. Stow and D.J.W. Piper
from shallow water is also common. Calcirudites, calcarenites, calcidebrites and carbonate slump deposits have been described from many slope and basinal systems (e.g. McIlreath & James 1979). As with terrigenous mud turbidites, finegrained carbonate turbidites are not always readily distinguished from associated pelagic and hemipelagic facies, except where deposition has occurred below the carbonate compensation depth (Hesse 1975). Several authors besides Hesse have encountered similar problems in making clear facies distinctions (e.g. Wilson 1969; Carrasco 1977; Cook & Taylor 1977; Enos 1977; Reinhardt 1977; Homewood & Winkler 1977; Stow, Wezel et al. this volume; Faugeres et al. this volume; Heath & Mullins, this volume). In other cases, more definitive interpretations of carbonate turbidites have been made for a variety of Palaeozoic to Recent limestones (e.g. Van Andel & Komar 1969; Thomson & Thomasson 1969; Davies 1977; Anatra et al. 1980; Kennedy 1980; Kelts & Arthur 1981; Faugeres et al. 1982). Siliceous turbidites rich in diatoms, radiolarians, sponge spicules and other siliceous organisms are less well known than the carbonate equivalents. However, they have been encountered at several DSDP sites in the North Atlantic (Beall & Fisher 1969; Peterson et al. 1970;
Kagami 1979; McCave 1979) and in the Gulf of California (Curray et al. 1980).'A few ancient examples have also been documented from the Mediterranean area (Nisbet & Price 1974; Kalin et al. 1979), Japan (Imoto & Fukutomi 1975) and North America (Folk & McBride 1978). There is not always a clear compositional distinction between the different biogenic turbidites so that all mixtures of carbonate, silica and clay materials can occur. In addition to relatively pure carbonate and siliceous ooze turbidites, Stow (1984c) describes marl, 'sarl' (siliceousclayey) and 'smarl' (siliceous-calcareous-clayey) biogenic turbidites from the south-east Angola Basin (terminology of Dean et al. 1984). Facies m o d e l
Because there are close similarities between the numerous descriptions of calcareous and siliceous fine-grained turbidites and because many of them comprise various admixtures of these components, it seems appropriate at this stage to propose a unified facies model for biogenic turbidites irrespective of composition. We have used a modified Piper (1978) sequence for mud turbidites, including an E/F division to emphasize the extremely gradational transition from turbidite to pelagite that commonly occurs (Fig. 6).
PELAGITE
E/F
BIOGENIC TURBIDITE-PELAGITE reverse-graded bioturbated
E3
BIOGENIC MUD ungraded bioturbation increases upward
E2
FIG. 6. Biogenic turbidite facies model. Structural divisions introduced here, modified from Piper (1978) and Stow, Wezel et al., (this volume).
BIOGENIC MUD slightly-graded isolated burrows
BIOGENIC MUD-SILT graded parallel diffuse lamination broad scoured sharp base
Deep-water fine-grained sediments: facies models The detailed structural divisions of the Stow (1977) sequence have not yet been recognized in biogenic turbidites, whereas the twofold unifitehemipelagite division of Stanley (1981) appears too simple. The complete sequence from top to bottom is: F hemipelagic or pelagic sediment, bioturbated E/F mixed turbidite and pelagite, reversegraded, extensively bioturbated E3 ungraded, very fine-grained, very homogeneous, isolated burrows E2 slightly graded, fine-grained to very finegrained, otherwise homogeneous, rare isolated burrows E1 graded, laminated, laminae horizontal and indistinct or diffuse, alternation of coarser (biogenic) and finer (clay-biogenic) material, thicker basal lamina of sandy-silty biogenic debris with sharp scoured base. Textural grading is most evident through the graded laminated division (El), but is very slight through the E2 and absent in the E3 divisions, which are commonly very fine-grained. The background pelagic or hemipelagic sediment, comprising large unbroken forams, diatoms and other planktonic organisms, is often coarser grained than the E2-E3 turbidite divisions. The gradual upward transition from turbidite to pelagite, therefore, shows a marked reverse grading. Where the components are of mixed types, compositional grading is the most striking. There has been little experimental work on the relative hydraulic equivalence of different biogenic particles (e.g. Berger & Piper 1972) but empirical data from descriptions of biogenic turbidites suggests that forams and shell debris alternate with terrigeneous clay-rich laminae in the lower divisions, whereas diatoms and the finest carbonate material (e.g. nannofossils) are concentrated towards the top. This sequence of structures and associated grading can be interpreted in terms of deposition from a waning turbidity current, in the same way as for terrigenous mud turbidites. The coarsest material deposits first as a thin layer at the base. Shear-sorting together with a waning current velocity results in the graded laminated division (Piper 1972b; Stow & Bowen 1978, 1980), in which the rather diffuse lamination indicates a poor separation of silt-sized biogenics from claysized material. This may be, in part, a function of the floc composition, with floc strength less than in clay-rich sediment. The absence of structures and poor or absent grading in the E2-E3 divisions, the extended transitional division, E/F, and
619
the extent of burrowing into the turbidite, all appear to suggest that there is an even less clear distinction between the processes responsible for deposition of the turbidite and pelagite facies in biogenic than in terrigenous systems (Heath & MuUins; Crevello et al., both this volume). Stow, Wezel et al. (this volume) suggest that finegrained 'pure' carbonate carried into a basin by a turbidity current is more likely to disperse into the water column than the equivalent terrigenous material, perhaps because it is less prone to forming strongly-bound flocs. Thus, the finegrained low-concentration tails of carbonatecharged turbidity currents mix with surface and mid-water suspensions derived in situ or from the basin margins and settle slowly through the water column. The biogenic turbidite will therefore grade imperceptibly upwards into pelagite. Whether or not a similar reasoning applies to siliceous biogenic turbidites is not certain. However, it seems less common that both the background pelagic sedimentation and the resedimenting turbiditic material are of equivalent siliceous composition, so that there is mostly a marked compositional distinction between the pelagite and turbidite facies. Variabilit),,
Examples of variability of biogenic turbidites are illustrated in Fig. 3. Thick-bedded (commonly > 1 m), fine-grained, biogenic turbidites showing the complete structural sequence (E 1-E/F) occur in two main settings. They are found as relatively proximal deposits on platform slopes and ridge flanks where the source material is entirely finegrained (e.g. Stow 1984c). Thinner-bedded basecut-out turbidites (E2-E/F, E3-E/F and E/F) occur more distally (e.g. Bartolini et al. 1975). They also occur as more distal deposits in small ponded basins (e.g. Wezel 1973; Stanley 1981) where the original turbidity current may have already deposited its coarser load. There are many examples of thick fine-grained caps to even thicker (up to 10 m) coarse-grained biogenic debrites and turbidites. Complete gradational sequences do occur (e.g. Faug6res et al. 1982; Gonthier et al. 1982; Stow 1984b,c), but more commonly the basal units are variously missing so that a thin to thick fine-grained bed overlies a much reduced biogenic sandy or gravelly layer. The contact between the two is often sharp or apparently erosive. The other main variable besides structural sequence and bed thickness is composition. Carbonate mud turbidities may be rich in nannofossils (e.g. Kennedy 1980), resedimented periplatform ooze of mixed composition (e.g. Crevello &
D.A.V. Stow and D.J.W. Piper
620
Schlager 1980; Heath & Mullins, this volume), or mixed planktonics and indeterminate lime mud (Stow, Wezel et al., this volume). Every gradation of mixed carbonate-elastic mud turbidites occurs (see examples in Cook & Enos 1977; McIlreath & James 1978), with the characteristic features of the facies changing accordingly. Gradations between siliceous and calcareous biogenic turbidites are also well-known (e.g. Kelts & Arthur 1981; Stow 1984c). Almost pure diatomaceous turbidites have been described from the Gulf of California (Curray et al. 1979), radiolarian chert turbidites from Greece (Nisbet & Price 1974), and mixed siliceous turbidites from many parts of the world (e.g. Kelts & Arthur 1981; Chough, this volume).
Disorganized turbidites Disorganized turbidites (Fig. 7) show only poor diffuse positive grading with bioturbation, if present, concentrated towards the tops of beds. Otherwise they are structureless. They may be dominantly silt grade, clay grade or mixed and have variable admixtures of biogenic and terrigenous material. However, separation into distinct laminae of different grain size or composition has not occurred and the sorting is therefore
~Z~'~
~
poor. They have a sharp, often scoured base and a gradational, indistinct top, occurring in beds from 5 cm to several metres in thickness. Disorganized fine-grained turbidites of this sort are relatively widespread, occurring both as the dominant turbidite type in an area and as isolated examples within a suite of more organized thin-bedded turbidites (Fig. 3). However, their interpretation is not always clear. In some cases they may be a part of one of the other turbidite sequences, such as the massive silts and ungraded muds described by Piper (1978). In other cases they may result from the ponding of large turbidity currents in small basins (e.g. Wezel 1973; Stanley 1981). In proximal slope or shelfedge settings, disorganized turbidites may be deposited by turbidity currents that have not matured sufficiently during flow to allow some internal sorting of their loads (e.g. Ballance et al., this volume; Hill, this volume). In more distal settings, where the associated turbidites are wellorganized (e.g. van Weering & van Iperen, this volume), very rapid deposition perhaps due to a local topographic effect may cause deposition of isolated disorganized beds. Chough & Hesse (1980) interpret such thin poorly-sorted layers, intercalated with thinly laminated fine-grained turbidites on the levees of the Northwest Atlantic Mid-Ocean Channel, as resulting from spill-over of the head rather than body of a turbidity current.
PELAGITE
Contourites u
poor or diffuse grading bioturbefion near top sfrucfureless poor sorting
rarefloatingclasts sharp+ scouredbase
~r A
m FIG. 7. Disorganized turbidite faciesmodel (this paper).
Contourites are a volumetrically significant facies in the present-day deep sea, where they can occur in enormous sediment drifts that are equivalent in size to some larger submarine fans, but they have proved extremely difficult to identify in the ancient record. This is perhaps because they have a low preservation potential, since almost all are developed on oceanic crust. Contourites are sediments deposited or reworked by currents that are persistent in time and space and flow mainly along-slope in relatively deep water. Deposition is controlled by bottom currents driven by thermohaline circulation or, more rarely, by wind-forced surface currents, rather than by gravity-driven turbidity currents. Johnson et al. (1980) and Halfman & Johnson (this volume) have extended the definition of the facies to include lacustrine contourites from Lake Superior. The chief sedimentary criteria that can be used to distinguish contourites from other deep sea facies include: (1) an irregular vertical arrangement of facies types and structures with both negative and
Deep-water fine-grained positive grading, but no regular structural sequence; (2) evidence for more or less continuous bioturbation that has kept pace with deposition, but with the 'ghosts' of current-induced structures remaining; (3) compositional, textural or other features that indicate a combined in situ and exotic origin. For present-day examples there are a range of non-sedimentary criteria that provide important additional evidence for a contourite interpretation. These include the measurement of bottom currents and nepheloid layers, the identification of large and small-scale current-induced morphological features, and certain seismic and stratigraphic characteristics. The acoustic character of contourites using high-resolution profiling systems is dependent on grain size variation and on the presence of intermediate-scale bedforms. In seismic reflection profiles, both acoustically transparent and well-laminated contourite drifts have been described (e.g. Scrutton & Stow 1984). The nature of drift growth and the positive accumulation features (large-scale mud waves, sediment drifts) are diagnostic. There are other bottom currents in the deep sea, such as the up and down-canyon currents measured by Shepard et al. (1979), and these may deposit sediments with very similar characteristics to the contourites we describe. There have been very few descriptions of sediments that can be related clearly to bottom currents of this type, so we have not considered them further in this paper. According to Heezen et al. (1966) and Hollister & Heezen (1972), the thin laminated silt and fine sands on the continental rise off eastern North America are the typical contourite facies. Subsequent work on this margin has shown that the distinction between contourites and interbedded facies is not so clear cut (Stanley et al. 1979; Stow 1979a; Shor et al., this volume). Piper & Brisco (1975) describe a different set of characteristics for silty contourites on the Antarctic margin. Based largely on evidence from coring and drilling the major sediment drifts in the North Atlantic, Stow & Lovell (1979) and Stow (1982) have characterized two main contourite facies: (1) muddy contourites, that result from deposition by bottom currents; and (2) sandy contourites, that result from a combination of deposition and reworking by bottom currents. To these we might add a third, less common facies type: (3) gravel-lag contourites, that result from erosion and winnowing of fines by powerful bottom currents (Carter & Schafer 1983).
sediments." facies models
621
Faug~res et al. (1984) and Gonthier et al. (this volume), from a study of the Faro Drift contourires in the Gulf of Cadiz, point out that there is in fact a continuum of contourite facies from the very fine clay grade to the coarse sand and gravel grades. They recognize an additional intermediate mottled silt and mud contourite facies as well as a silty-sandy facies (rather than a strictly sandy facies). We describe two contourite facies models in this paper, one for muddy contourites and one for silty or fine sandy contourites, and then discuss their occurrence in irregular vertical sequences.
Muddy contourites Muddy contourites include a range of deposits at the fine end of the grain size spectrum, and appear to be the dominant contourite facies in the deep sea, making up over 75% of the main contourite drifts. They have been described from some 15 separate drifts in the North Atlantic (see reviews by Stow & Lovell 1979; Stow & Holbrook, this volume) and show a consistent set of sedimentary characteristics (Fig. 8) that are similar in many ways to hemipelagic and pelagic sediments. There have been no specific identifications of muddy contourites in ancient rocks, although several authors have interpreted ancient slope-rise sequences where they believe bottom currents to have been active (e.g. Bein & Weiler 1976; Anketell & Lovell 1976). On first appearance the facies is monotonous, homogeneous and structureless, but closer inspection reveals important structural features. There are no very distinct beds, but both positive and negative grading between more clayey and more silty-sandy sediment is commonly present over a few centimetres to a few tens of centimetres. The contacts between clay-rich and siltrich parts are mostly gradational, less commonly sharp and erosive. Primary lamination of either silt or clay is rare, and may be distinct and planar or indistinct and wavy. Where not laminated, the muds are thoroughly bioturbated. The silt fraction is often concentrated in irregular pockets and lenses that show some horizontal alignment, but that appear to have been broken up by bioturbation (Faug~res et al. 1984). Texturally, these mud contourites are siltyclays with up to about 10 or 15% sand fraction and a mean size that varies from less than 5 ~tm to about 40 ~m. They are mostly poorly sorted. The shape of the grain size curves (Rivi~re 1977) varies from logarithmic with slight hyperbolic tendency for the finer-grained samples to parabolic for the more silty samples (Gonthier et al., this volume).
D.A.V. Stow and D.J.W. Piper
622
d
~ - ~,r ~'ll ~ _ ~
""
_
Structure
'~ - -
9
,.-w~.)~ ]
_ 6 ~i|
irregular horizontal lense% layers + wispy
|
rare parallel lamination
--G-
_
[email protected]~Ts 9 E)
w Ior
D oI.-
bedding poor or absent
"_~
. ~
,,,.
.
,, ,~,'-.;
, ' . . " ,~,,--.~ ,' . ~ : ,~.~
9 -. ~ -
"~."""~E~-, ~'. ,.
Z tJ
cm
Texture dominantly silty mud 0-15% sand sized poor to moderate sorting irregular vertical variation
__~
,. ,,. L.....-"v ) I0
mostly homogeneous Jr thoroughly bioturbated~
~..~.'~.. "9- ' . " ~ ' - c . - ~.. ~/
r 1
Composition mixed biogenic-terrigenous mixed fragmented-whole biogenics mixed planktonics-deep benthonics
o
FIG. 8. Muddy contourite facies model (after Stow & Lovell 1978; Stow 1982).
In terms of composition, many muddy contourites are mixtures of biogenic and terrigenous material in variable proportions depending on the distance from land, source of supply and so on (Laughton et al. 1972; McCave et al. 1980; Stow & Holbrook, this volume). The biogenic fraction includes calcareous and siliceous planktonic and benthonic forms. Some are broken and ironstained. The benthonics are usually m situ deeperwater forms rather than shallow-water derived species. The terrigenous fraction is commonly dominated by fine quartz and clays. Other contourites may have a more or less pure pelagic composition (Shor & Poore 1978; Kidd, pers. comm. 1983) and some appear to be dominated by volcanogenic material from mid-ocean ridges and seamounts (Taylor et al. 1975; Van Stackelberg et al. 1979).
Silty-sandy contourites Silt and fine sand contourites appear to be less abundant in the present deep sea as distinct beds. Many of the muddy contourites described from contourite drifts contain intervals that are more sandy and silty, some of which might be termed sandy muddy silts. In other cases, cleaner sandy and silty beds have been documented (e.g. McCave et al. 1980; Faug~res et al. 1979, 1984; Gonthier et al., this volume). The origin of the sharply-defined silt laminae and cross-laminated fine sands on the continental rise off North America remains enigmatic (e.g. Hollister & Heezen 1972; Stow 1979; Shor et al., this volume), and so we have not included the features of these deposits in our synthesis of a silty-sandy contourite facies model (Fig. 9). Most of the ancient
623
Deep-water fine-grained sediments."facies models n
M UD
I ~'~'---~:='I I -'~ ~"~-C'~'~
bioturbated laminated
a w
cross-laminated
SILT
c~
bioturbated
silt mottles + lenses irregular horizontal alignment
MOTTLED SILT + MUD
w --> I--
bioturbated
CO
0(3_
massive
irregular sandy pockets
SANDY SILT
bioturbated contacts
sharp to gradational
silt mottles~ lenses + irregular layers bioturbated contacts variable
MOTTLE O SILT -I- M U D
----
tm LI.I a
> l-
oE
'Ot
MUD
-F . ~-- ~_~a~--
,~, ~
-
1
W
bioturbated
0 FIG. 9. Typical arrangement of contourite facies in negatively to positively graded couplets (after Gonthier et al., this volume; Faug6res et al., 1984).
sandy contourites that have been described have been identified on the basis of the Heezen-Hollister criteria (see reviews in Stow & Lovell 1979; Lovell & Stow 1981), and may in fact not be contourites, Silty or sandy contourites occur in irregular beds from 1 cm to about 20 cm in thickness. The top and bottom contacts of beds may be sharp and relatively flat, erosional or completely gradational. In many cases they display no primary structures apart from irregular concentrations of coarser material and slight positive or negative grading. In other cases, primarycross-lamination and parallel lamination has been preserved, but no consistent sequence of structures has been observed. Secondary bioturbation, continuous throughout the beds, is the most common feature, ranging from large distinct burrows to small irregular mottling (e.g. Chough & Hesse, in press), Medium to coarse-grained silts with up to 4 0 ~ sand and less than about 10% clay are the most common, but fine sandy contourites are also
known. They are mostly moderately well-sorted but with a distinct fine tail evident on grain size curves. According to Gonthier et al. (this volume) the shapes of the curves tends towards parabolic or to a combination of hyperbolic (coarse tail) and parabolic (fine tail). Compositionally, silty-sand contourites are similar to muddy contourites in comprising mixed biogenic-terrigenous material (Faug6res et al. 1984). The larger biogenic grains are commonly fragmented and iron-stained, and there is clearly less clay material. More or less pure foraminiferal contourite sands have also been described (Faug6res et al. 1979; Shor et al. 1980).
Contourite 'sequence' There is no regular sequence of structures within contourites as there is in the various turbidite models described. However, a distinctive feature ofcontourite successions appears to be the presence of both negatively-graded sequences in
624
D.A.V. Stow and D.J.W. Piper
which the grain size increases, and positivelygraded sequences in which the grain size decreases upwards (Gradstein et al. 1982; Faug6res et al. 1984; Gonthier et al., this volume). Both sequences may be from about 10 to 100 cm thick and occur separately or as a combined negativepositive unit (Fig. 9) which shows, from top to bottom: - - homogeneous mud facies --mottled silt and mud facies - - mottled facies with silt layers - - silt-sand facies --mottled facies with silt layers --mottled silt and mud facies --homogeneous mud facies There is considerable variability in this sequence as it is observed in different contourite successions, particularly in terms of its thickness, its completeness and its symmetry. Such variations in grain size of the facies and in the associated structural and compositional characteristics can be interpreted in several different ways. Faug6res et al. (1984) relate the Faro Drift sequences to long-term variation in velocity of the transporting current. A complete negative-positive sequence represents a gradual increase, a maximum and then gradual decrease in the average current velocity at a given site. The time-scale for deposition of the sequence would be of the order, say of 1000 to 30 000 yrs, and the mean velocities might vary between about 5 and 25 cm/s (Gonthier et al., this volume). Alternatively, such sequences might reflect variation in the grain size and/or biogenic content of material supplied to the system. We emphasize, again, both the subtlety of the sedimentary features observed in contourites and their variability (Fig. 10a-c). On the one hand they are similar to those of some indistinct fine-grained turbidites, and on the other hand they are almost indistinguishable from hemipelagite characteristics. This fact underlines the existence of a process-continuum in which flow velocity, concentration and frequency can all vary, together with sediment supply.
Pelagites and hemipelagites The third major facies group of the deep sea comprises the pelagic and hemipelagic sediments (Hsu & Jenkyns 1974; Jenkyns 1978, in press; Einsele & Seilacher 1982). These are widespread throughout the world's oceans and widely recognized in both modern and ancient successions. They have been deposited primarily by slow settling through the water column in the absence of any s u b s t a n t i a l bottom current or turbidity
current activity. However, many of the processes proposed for the accumulation of hemipelagites involve current-induced slow advection of suspended sediment (Drake et al. 1978). Hemipelagic sediments accumulate on continental margins and in other settings not far removed from terrigenous sediment sources. They comprise a mixture of indigenous biogenic material and silt and clay size terrigenous detritus. They accumulate slowly and are thus intensely bioturbated. True pelagic sediments accumulate in the open ocean and comprise principally skeletal parts of plankton with some admixture of very fine silt and clay, much of which has reached the open ocean by aeolian transport. The proportion of terrigenous material may be increased by dissolution of the biogenic components. Bioturbation is ubiquitous except in anoxic basins. Textural and compositional criteria can be used to distinguish four types ofhemipelagite/pelagite facies (modified after Berger 1974): (l) pelagic ooze, with > 75% biogenics; (2) muddy pelagic ooze ('arl'), with 25-75% biogenics and a terrigenous component predominantly of clay; (3) pelagic clay, with <25% biogenics and > 60% clay in the terrigenous fraction; and (4) hemipelagites, with >5% biogenics and a terrigenous component with > 40% silt. Other minor facies include the purely chemogenic sediments that are composed almost entirely of authigenic minerals, such as ferromanganese nodules and phosphorites. These are not considered further here. The chief distinguishing features of pelagites and hemipelagites include: (1) evidence for low or very low rates of sedimentation and continuous bioturbation (except in anoxic basins); (2) no primary sedimentary structures or other evidence of current-controlled deposition; (3) a mainly uniform composition within any one succession, that may show a regular cyclicity related to climatic or other controls; (4) a variable biogenic component mainly of planktonic tests, a very fine grained, often far-travelled, terrigenous component, and commonly a significant authigenic component. Seismic reflection profiling shows a more or less uniform, acoustically transparent sediment drape over bottom irregularities although there may be a thickening in basins (Moore 1969). Stratigraphically, pelagites and hemipelagites commonly show continuous deposition, although accumulation rates may be as low as one metre per million years and hiatuses may be present.
Deep-water fine-grained sediments."facies models
625
FIG. 10. Photographs of contourite, pelagite and hemipelagite facies. Section widths about 7cm. (a) Silty contourite, Pleistocene, Faro Drift, Gulf of Cadiz. (b) Mottled silt and mud contourite, Pleistocene, Faro Drift, Gulf of Cadiz. (c) Muddy contourite, Plio-Pleistocene, Blake-Bahama Outer Ridge, western North Atlantic. (d) Bioturbated pelagite-hemipelagite, upper Cretaceous, Scaglia Rossa Formation, Italy. (e) Organic-carbon-rich siliceous pelagite (diatomite), Miocene, Monterey Formation, California. (f) Alternating black shade and marlstone, mainly hemipelagic (part turbiditic), Cretaceous, DSDP Site 530A, Angola Basin. (g) Bioturbated siliceous/calcareous pelagite, Plio-Pleistocene, DSDP Site 532, Walvis Ridge.
626
D . A . V . S t o w a n d D . J . W . Piper
Pelagic ooze
Pelagic oozes are most typical of the open ocean basins far from a terrigenous source. They have been the subject of much study over the past 125 years and hence their sedimentary characteristics are well known. Some important syntheses are to be found in the volumes by Hsu & Jenkyns (1974) and Cook & Enos (1977), and in papers by Arrhenius (1963) and Jenkyns (1978 and in press). Specific examples in this volume where pelagic sediments are described as part of the succession, include those of the western and central North Atlantic (Robertson), the north-western Mediterranean (Monaco & Mear), the Miocene of southern Turkey (Hayward) and the Cretaceous-Tertiary of the Apennines (Stow, Wezel et al.).
N r-l-
.J-I
o
I
massive no primary structures bioturbated sandy-silty-clayey bedding poor or absent > 7 0 % calcareous tests
~la-.-L_ _~-I I
-', '~L
I "L- --'-~'---,~
massive no primary structures except rare ash layers bioturbated silty-clay < I 0 % biogenics
Facies model (Fig. 11) One of the chief characteristics of pelagic oozes is their very slow rate of accumulation, commonly from less than 1 mm to 10 mm/1000 yrs, although this can be an order of magnitude higher under zones of upwelling. They are usually, therefore, thoroughly homogenized by bioturbation, and without any primary current-induced structures. A variety of burrow types may be preserved with different assemblages or ichnofacies dependent on different environmental factors, such as water depth, grain size, sedimentation rate and redox conditions (Seilacher 1967; Werner & Wetzel 1982; Wetzel, this volume). Some of the main diagnostic trace fossils are Zoophycos, Chon-
drites, Planolites, Scolicia, Trichichnus, Teichichnus and Lophoctenium. Several tiers of trace fossil assemblages are commonly superimposed on one another (Werner & Wetzel 1982). The grain size of pelagic oozes is largely dependent on the composition of the biogenic fraction. Coccolith plates are very small (clay size), whereas some foram-rich or diatom-rich oozes may have a mean grain size that is in the silt range. The terrigenous component is mainly clay-sized. Full grain size analyses, however, are rarely carried out on pelagic oozes as the hydraulic equivalence of biogenic particles is not wellknown and hence interpretation of the grain size distribution would be difficult. Pelagic oozes are composed dominantly (> 75%) of the tests of planktonic organisms, either calcareous (coccoliths, forams, pteropods) or siliceous (radiolarians, diatoms, silicoflagellates) or a mixture of both (Berger 1974). The other components (Lisitzin 1972) can include very fine-grained terrigenous material (principally quartz, feldspars and clays), volcanogenic debris (palagonite and derived clay minerals),
ferromanganese nodules massive no primary structures bioturbated sandy-silty-clayey bedding poor or absent
> 70%
FIG. 11. Pelagite facies models. Schematically interbedded calcareous ooze, red clay and siliceous ooze, with ferromanganese nodules developed at ooze-clay contact.
authigenic minerals (such as phosphate, barite, zeolites, ferromanganese nodules and coatings), and rare extraterrestrial material. Under normally oxic conditions the organic-carbon content is extremely low, but under anoxic conditions pelagic black shales can contain over 20% organic carbon (Isaacs 1981; this volume; Arthur et al., this volume). The characteristics, composition and distribution of the different types of pelagic ooze are dependent on a number of interacting variables which we will not discuss at length (but see Berger 1970; Lisitzin 1972; Broecker 1974). These include: the water depth and the corresponding carbonate compensation depth; the source and supply of terrigenous and volcanogenic material; surface water productivity and the supply of biogenic material; surface currents and bottom circulation patterns; climate and basin physiography; and physiochemical conditions. Pelagic sediments are perhaps more affected by this range of
D e e p - w a t e r f i n e - g r a i n e d sediments." f a c i e s models different controls than are turbidites or contourites because of their two principal attributes: (1) they are composed mainly of biogenic CaCO3 and/or SiO2 both of which are soluble in sea water; and (2) they settle relatively slowly through the water column and are buried only slowly, and hence are exposed to external factors for a relatively long time period, either in the water column and/or on the sea-floor. The actual processes of settling as single grains or as larger flocs and pellets are discussed at more length in papers by Gorsline, McCave, and Eittreim (all this volume).
Muddy pelagic ooze There is a continuum of facies from pelagic ooze with > 75% biogenic material to pelagic clay with < 25% biogenics. Muddy pelagic ooze (or 'arl', terminology of Dean et al. 1984) is a relatively common intermediate sediment type, with characteristics intermediate between an ooze and a clay. It differs from true hemipelagic sediment in having a dominantly clay-sized rather than siltsized terrigenous component, and in being an open-ocean rather than continental margin facies.
627
Clay minerals are the dominant components, whereas quartz, feldspar and other terrigenous materials are very minor (Arrhenius 1963). Authigenic components include the zeolites, ferromanganese minerals (goethite, micronodules) together with some clays and feldspars. Biogenic material is often very scarce, but a complete gradation exists with muddy pelagic oozes ( > 25% biogenics). Volcanogenic material, essentially palagonite, is present in very variable amounts. In certain environments especially close to mid-ocean ridges or immediately overlying ocean floor basalts, pelagic clays can be highly enriched in a variety of metals and trace elements. Hoffert (1980) has identified four types of pelagic clay based on slight differences in their mineralogical and chemical compositions. These are associated with (1) siliceous oozes, (2) calcareous oozes, (3) volcanogenic material, and (4) none of the above, but with a mixed composition. It is by alteration and dissolution of the different biogenic or volcanogenic components of these other oceanic sediments, together with authigenesis and diagenesis of new minerals, that pelagic clays are formed in a process somewhat analogous to pedogenesis on land.
Hemipelagites Pelagic clays Pelagic clays, also known as red clays, brown clays and abyssal clays, accumulate in the deepest, and most remote parts of the ocean basins. They are particularly well represented in parts of the Pacific Ocean, resulting from dissolution of biogenics and aeolian transport of dust. Their sedimentary characteristics are relatively well known (e.g. Arrhenius 1963; Griffin et al. 1968; Hsu & Jenkyns 1974; Jenkyns 1978; Hoffert 1980). Ancient examples have also been documented (e.g. Audley-Charles 1965).
Facies model (Fig. 11) Pelagic clays have one of the lowest sedimentation rates of all the pelagic/hemipelagic sediments, commonly less than 1 ram/1000 yrs but ranging up to about 7.5 mm/1000 yrs. The rate of growth of ferromanganese nodules and crusts, however, is one or two orders of magnitude lower. As with pelagic oozes, pelagic clays are usually well-oxygenated and thoroughly bioturbated. The trace fossil assemblages are those adapted to the deepest water, and finest grain size. Texturally, they are very fine-grained, clay or fine-silt sized, and poorly to moderately well sorted with an even distribution of grain sizes over a small size range.
Hemipelagic sediments are more typical of marginal oceanic settings where there is a ready supply of terrigenous material. In published descriptions of sediment sequences they are frequently referred to as the background, normal, ubiquitous or interbedded facies, and are often not described in any great detail although they may comprise the greater part of a given sequence. At high latitudes, for example in the Arctic Ocean and Baffin Bay, ice-rafting is a major contributor to fine-grained hemipelagic sediments. Amongst the large volume of literature that simply refer to hemipelagic sediments in this vein there are, however, some authors who have documented the facies characteristics in rather more detail. Particularly useful descriptions of modern hemipelagites include those of Rupke (1975) and Stanley & Maldonado (1979) from the western and eastern Mediterranean respectively, Stanley et al. (1972) and Hill (1981) from Nova Scotian margin, Moore (1974) and Kolla et al. (1980) from the Indian Ocean, and various contributions by Gorsline and colleagues (e.g. Gorsline 1978, 1981) from the basins of the California Borderland. In this volume there are a number of papers dealing, in part, with hemipelagic facies in the modern deep sea (e.g. Monaco & Mear, Chough, Krissek, Thornton, Gorsline et
628
D.A.V. Stow and D.J.W. Piper
al., Faug6res et al., Isaacs, and Auffret et al.; all
this volume). Although more difficult to identify in ancient rocks there have been a number of papers describing hemipelagites from inferred slope and basinal settings (e.g. Hesse 1975; Piper et al. 1976; Ingersoll 1978; Hicks 1981; Pickering 1982b). Papers in this volume by BaUance et al., Bourrouilh & Gorsline, and Pickering also describe ancient hemipelagites. Facies model (Fig. 12)
We are therefore able to summarize the chief sedimentary characteristics ofhemipelagites (Fig. 12). They are commonly homogeneous and structureless, with bedding poorly defined or absent except when it has been accentuated by burial and diagenesis (see below). There are no primary current-induced sedimentary structures such as lamination, ripples or erosional contacts, although a depositional lamination may be preserved under anoxic conditions (e.g. Isaacs, this volume; Thornton, this volume).
Under normal oxic conditions, however, bioturbation is ubiquitous and thorough, often resulting in a completely homogenized sediment with a mottled aspect. Burrow traces are also often preserved, with the same major trace fossil assemblages represented as for the pelagic oozes described above. These similarly depend on water depth, grain size, sedimentation rate and redox state (Werner & Wetzel 1982; Wetzel, this volume). Iron sulphide filaments (Mycelia) and mottles are also a common feature ofhemipelagic sediments. Texturally, hemipelagic sediments are silty clays with 1 to 15% of sand of mainly biogenic origin. They are poorly sorted and show no systematic grading apart from that associated with the compositional cyclicity discussed below. The shape of the grain size cumulative curves (Rivi6re 1977) appears almost uniform or logarithmic, although there may be irregularities due to a typical biogenic input or rare exotic terrigenous material (e.g. ice-rafted debris). Apart from the mixed biogenic-terrigenous aspect of hemipelagites, it is difficult to generalize
Structure massive no primary structures thoroughly bioturbated bedding poor or absent limestone-marlstone cycles
z" '
o_ nn m
Texture
.,=
sandy silty mud very poorly sorted size variation related to biogenic content
I
Composition mixed biogenic-terrigenous mixed planktonics- benthonics biogenic- rich/clay- rich cycles common ice-rafted input at high latitudes
12. Hemipelagite facies model. Rhythmic alternation of biogenicrich and clay-rich intervals. FIG.
o
-
Deep-water fine-grained sediments."facies models further about their composition. The biogenic input may be slight or dominant, calcareous or siliceous depending on surface productivity, carbonate compensation depth, etc. It is commonly of mixed planktonic and in situ (deep-water) benthonic origin. The terrigenous components are mostly uniform in any given area and show very gradual compositional trends towards the source area. Their nature can, of course, be highly varied depending on tectonic, climatic and other factors, and the sediments may also contain a minor to significant fraction of far-travelled (wind-blown, ice rafted, etc.) terrigenous debris. Seismic-reflection profiles of hemipelagic sediments frequently show a draped morphology over bathymetric irregularities suggesting that they were deposited by vertical settling through the water column. Observations of oceanographic processes and of regional patterns of sediment distribution, however, suggest that other rather more complex processes may also have an effect on hemipelagic deposition (e.g. Moore 1969; Damuth & Kumar 1975; Drake et al. 1978; Stanley & Kelling 1978; Karl et al. 1983; Hill & Bowen 1983). In particular, slow lateral displacement of fine suspended sediments in mid-water and bottom water nepheloid layers; sediment dispersion by up and down-canyon normal currents; resuspension and slow diffusion at the shelf break; and sediment creep over gentle slopes all appear to be important processes. These are discussed at more length in papers by Gorsline, McCave, Eittreim, Thornton and Gorsline et al. (all this volume).
Hemipelagite-pelagite cycles Examples of some hemipelagic and pelagic sediments are shown in Fig. 10d-g. An important feature of many, but not all, hemipelagic and pelagic successions is a cyclic alternation of biogenic-rich (pelagite) and clay-rich (hemipelagite or pelagic clay) beds (Fig. 12) (e.g. Dean et al. 1981; Einsele 1982; Einsele & Seilacher 1982; Stow, Wezel et al., this volume). Most of the cyclic sequences that have been described from both modern and ancient successions are limestone-marl cycles with the primary variation being the relative proportions of carbonate and clay, although similar cycles are also known from biogenic siliceous sequences. Other compositional (e.g. organic carbon, trace elements) and textural (grain size) characteristics may mirror the limestone-marl cycle or may vary independently. There is a close correspondence worldwide in the order of magnitude of both Quaternary-Recent and older cycles. They are commonly between 10 and 100 cm thick, with a
629
sedimentation rate of 0.5 to 3 cm/1000 yrs and a periodicity of 20 000 to 100 000 yrs. Possible causes for cyclic variation in the relative abundance of CaCO3 or biogenic silica include: (1) variation in the rate of production of CaCO3 or SiO2 by planktonic organisms; (2) variation in the rate of input of CaCO3 or SiO2 via turbidity currents; (3) variation in the amount of dissolution of CaCO3 or SiO2 either shortly after sedimentation or at depth after burial; and (4) dilution of the biogenic component by variation in the non-carbonate (terrigenous) input. All of these processes have been shown to be important in different cases. The close relationship of the periodicity to the earth's orbital changes and climatic variation suggest that climate is most often the ultimate control, with sea-level changes and fluctuation in the carbonate compensation depth also being important (e.g. Fischer & Arthur 1977). There are rather fewer examples of similar cycles in siliceous biogenic sediments (e.g. Garrison & Fischer 1969; Barrett 1982). Some of these probably have a similar origin to the limestonemarl cycles, but others were more clearly controlled by turbidity current input of the siliceous material (e.g. Nisbet & Price 1974; Folk & McBride 1978).
Environments and facies distribution The various fine-grained facies outlined above are not distributed randomly through the deep sea but occur preferentially in certain environments. They are commonly closely associated and interbedded with each other and with the range of coarser grained facies also found in deep water. Based largely on their occurrence at the present day, we outline here some aspects both of their generalized distribution on a global scale and of their more detailed horizontal distribution within slope, fan and basin plain settings. We then describe their occurrence in characteristic vertical sequences in the ancient record. Our intention here is to highlight only the main aspects of facies distribution and not to attempt rigorous documentation. We believe that more detailed consideration will be a fruitful area for further research.
Global distribution The principal environments of turbidite deposition (both coarse-grained and fine-grained) are the very large areas of continental slopes and rises and oceanic abyssal plains (Fig. 13). Also impor-
630
D.A.V. Stow and D.J.W. Piper e~
.,..a
~A
.=_
t~
O
~,,,~
8 Q
0
O
Deep-water fine-grained sediments: facies models tant are the slopes and floors of smaller marginal seas and land-locked basins. Turbidites are dominant, therefore, around the margins of the continents and particularly those trailing-edge margins that receive most sediment from continental drainage systems and glacial erosion, including much of the Atlantic Ocean, north Indian Ocean, Mediterranean Sea and the circum-Antarctic margin (Inman & Nordstrom 1971). Large amounts of sediment draining from south-east Asia is mostly trapped in back-arc basins or on broad continental shelves of the western Pacific. There is generally less sediment supplied to the coasts along collision margins, although the tectonically unstable and metastable slopes in these areas are important sites of resedimentation via slumping, debris flows and muddy turbidity currents (e.g. Gorsline et al., this volume). Silt turbidites are associated with channels and other conduits across the slope, and are commonly fed to the channel distributary regions on the lower slope (e.g. mid and lower fan lobes) and to more proximal parts of abyssal plains (Piper 1978). Mud turbidites are more widespread, occurring in levee and interchannel areas of the slope and throughout the basin plain. Biogenic turbidites are restricted to regions off carbonate banks and reefs, upwelling zones or seamounts and oceanic ridges, where a biogenic source is available. The distribution of disorganized turbidites is less clearly understood; they occur both in proximal and distal regions of turbidite sedimentation. The chief environments of contourite deposition are closely related to the deep-water thermohaline circulation pattern of the ocean basins and, in particular, to the higher velocity bottom currents that occur principally on the western margins of basins or by the acceleration of flow through restricted passageways. In the North Atlantic, for example (Fig. 14), a number of large sediment drifts can be identified that are made up almost entirely of muddy and silty contourites, and of pelagic or muddy pelagic oozes that have been moulded by bottom currents during deposition. These occur as isolated elongate mounds both in the middle of the ocean and parallel to or projecting from the continental rise, and as irregular dome-shaped mounds near seamounts and other topographic highs. Coarser-grained sand and gravel lag contourites occur in association with muddy contourites in the Straits of Gibraltar and in the western Labrador Sea. On the continental rise of eastern North America contourites are closely interbedded with finegrained turbidite and hemipelagite facies. Pelagic and hemipelagic sediments are inter-
631
bedded to varying degrees with both turbidite and contourite facies wherever they occur. Elsewhere in the deep sea they are the principal facies. Calcareous pelagic oozes are dominant in the Atlantic Ocean and over the shallower mid-ocean ridge parts of the other oceans. Siliceous pelagic oozes occur mainly in two high latitude circumpolar bands, a Pacific equatorial band and under upwelling zones on the western boundaries of the continents. Pelagic clays are confined to the very deepest and central parts of the oceans, in particular the Pacific. Hemipelagites are most abundant near the continents, except where masked by a major input of resedimented facies. At very high latitudes, there is an ice-rafted component within the other facies which, if dominant, gives an ice-rafted hemipelagite facies. Slope aprons, submarine fans and basin plains: horizontal distribution
Slope aprons are morphologically heterogeneous and have a correspondingly complex and irregular distribution of facies (Fig. 15) (Bouma et al. 1978; Doyle & Pilkey 1979; Stow 1984a). On normal terrigenous-supplied slopes there is commonly a mud-line (Stanley & Wear t978) that separates the shallow, higher-energy, sandy shelf facies from the mainly fine-grained slope sediments. Some sand spillover occur along the shelf break as well as the funnelling of coarser sediments down canyons or gullies to isolated depositional lobes. Areas of sediment creep and slumping give rise to resedimented slump and debrite masses and, in some cases, to mud turbidites. On the open slope there is an interbedding of and gradation between fine-grained turbidites, contourites, hemipelagites and pelagites, with the turbidites being more common near channels, the contourites occurring in areas of bottom current activity, and the pelagites increasing in abundance distally. Very rapid progradation occurs off deltas, particularly at times of lowered sea-level. Fluctuations in river discharge may produce graded beds resembling turbidites (see discussion in Stow, Alam & Piper, this volume). Sediment facies include both turbidites and hemipelagites accumulated from suspension fall-out. Both may be little bioturbated because of high rates of deposition. Similar facies occur in high latitudes off active ice margins. At the base of slopes there may be either a smooth gradual transition to the basin plain facies or areas of more positive construction. Isolated channel-fed lobes comprise silt and fine sand turbidites (Piper 1978), slump/debris-flowfed lobes comprise mud turbidites (Stow 1984c),
632
D.A.V.
Stow and D.J.W.
Piper
FIG. 14. Present-day deep-water circulation in the North Atlantic. Areas of bottom-water formation shown in wide stipple. Main contourite drifts shown in close stipple. Contour at 2000 m; mid-ocean ridge in heavy dashes. (After Stow 1982). and contourite mounds or drifts may be constructed parallel to the slope contours. Alongslope trends of sediment characteristics are observed in these drift deposits (Gonthier et al., this volume). Other papers in this volume that discuss facies distribution on modern clastic slopes include those by Hill, McGregor et al., Ballance et al., Krissek, Got, and Gorsline et al (all this volume). Variations on these facies distributions occur on slopes that are controlled by active faulting or diapirism. Resedimented and pelagic biogenic facies are dominant on carbonate slopes (e.g. McIlreath & James 1978; Mullins & Neumann 1979; Heath & Mullins, this volume; Faug6res et al., this volume) and on the flanks of oceanic ridges and seamounts (e.g. Gonthier et al. 1982).
The distribution of facies on submarine fans has received rather more attention in the past than that on slopes (e.g. Mutti & Ricci Lucchi 1972; Walker & Mutti 1973; Walker 1978). Several important papers have addressed specifically the question of distribution of the fine-grained facies (Piper 1978; Nelsen et al. 1978; Stow 1981; Cremer 1981, 1983) so that a relatively clear picture has emerged (Fig. 16). Van Weering & van Iperen (this volume) and Hayward (this volume) also deal specifically with fine-grained facies on deep-sea fans. There is both an elongate and concentric distribution of coarse to fine-grained facies on most fans. Slumps and slides are mostly confined to the slope, upper fan and channel margins. Debrites may be more widespread (Nardin et al.
Deep-water fine-grained sediments: facies models IRREGULAR SLOPE
SMOOTH SLOPE MIXED FACIES
r"
SLUMP MASS
MIXED FACIES
DISORGANIZED TURBIDITES
63 3
DEBRITE MASS
\ ~'9 .'.':..:'. :.. -9 9149 9 9 : ':,, \al__ ~.~. ..-.-......:....-. _ ,~-" _9. . . . . . . . . . . . . . ' . - . .j_ _1_ i~9." ." .'." -.'- . " . . . .a_ .L 9 ..
9
9
,-
.
.
."
-.
|
9
::::. : J
"
...-.... ..-.: : .:. :... ::.. 9 : 9 ;.
~
- ."
9 .'...'.
'
"...,"
.-
".
1
~-~_
"
~
. J .
/ m
9
9 : .... 9149149 9 .
- 9
9 o 9
.
9
tiiii!ii!iiiii 1 MASSIVE SILT
~
TURBIDITE
.Jl_
" . . .
9
9
..-",._,~,, .a.,;~
.,,
.
'
[~=~=~f TURBIDITES
/ I'~('
SILTY CONTOURITES
k,,,~
MUD TURBIDITES 4.- PELAGITES
MUDDY CONTOURITES CHANNEL TO LOBE
k,.,,~~ CONTOURITE DRIFT
FIG. 15. Schematic distribution of fine-grained sediment facies across a typical slope and rise. Note irregular distribution of mixed facies types 9 1979). Coarse-grained turbidites are transported through the channel system and deposited along their length and on the sandy lobes that spread out at their terminations. Fine-grained turbidites are transported either down channels and then laterally by overflow onto the levees and interchannel areas, or as thick unconfined low-density
flows (Bowen et al. 1984). They commonly show both a down-fan and away-from-channel evolution of textural, structural and compositional features. Hemipelagites, pelagites and, in some cases, contourites are interbedded with the resedimented facies in areas of lower energy. Basin plains are the ultimate trap for deep-sea
634
D.A.V.
Stow and D.J.W.
MUD TURBIDITES + PELAGITES ACROSS FAN
Piper SILT TURBIDITES DISTAL CHANNEL + LOBE
iii1
~
I.'.'.'..'.::'3
sea-level
.
5.-:.:. ;..--~" ..."
,,at.
.at.
: . . . . . :-:.:.:
SLUMPS UPPER SLOPE
HEMIPELAGITES q- DISORGANIZED TURBIDITES MID-SLOPE
DOWN-FAN TO ABYSSAL PLAIN
FIG. 16. Schematic distribution of fine-grained sediment facies across a typical large deep-sea fan. Note systematic down-fan and across-fan distribution of mainly turbiditic facies. sediments that have been eroded and transported by currents or that have settled slowly through the water column (Gorsline 1978; Pilkey et al. 1980; Stow, 1984a). They vary widely in their areal extent, depth and shape, including for example, the tiny plains of slope basins and the major oceanic abyssal plains. A single basin plain may be fed from several sources, including channels and fans, the surrounding slopes and surface waters. All sediment facies types are represented in basin plains (Fig. 17), depending very much on
morphological, tectonic, sedimentary and sealevel controls, and the fine-grained facies commonly predominate. Both centripetal and longitudinal facies distributions are observed; turbidites, contourites and hemipelagites/pelagites are often intimately interbedded. Several papers in this volume outline specific examples of basinal sedimentation, including those by Auffret et al., Got, Chough, Thornton, Gorsline et al., Robertson and Crevello et al. (all this volume). Where a basin is fully enclosed and circulation
Deep-water fine-grained sediments."facies models /-..~.
\ "~
'~
~'~'-~-~-~"
~.~"~.
635
_ see-level
. . .:.... 2 - : . )
..
.... ......-._::.-.:_-..
. ..". "!i:i!i:i!i! ~...
:!vif
\ i \ I,, \~
,,
\
hediiPoerl;aginteSed turbidites
.....
.":::::'".'?.'2":':.':'.'."."_.._
::::::::::::::::::::::::::::
BI.AS!NFRINGE ] ~ ' ~ ~ . . ~ silt-I- mud BASIN PLAIN turbidites thin silt + thick mud turbidites hemipelogites
_V
V t'.!i~
, IRREGULAR SLOPE slump deposits hemipelagites debrites
FIG. 17. Schematic distribution of fine-grained sediment facies within a relatively small enclosed marginal basin. Note irregular to concentric distribution of mixed facies types.
DEEP-WATER SLOPE DEPOSITS "2/" ~
}
Upper slope
/
Slump scars Lower slope
~
Channel-fill
sandstones
DEEP-WATER FAN DEPOSITS Inner fan
Channel fills; including olisthostromes
Middle fan Channel-fill sequences
DEEP-WATER PLAIN DEPOSITS ........... , .... ~ ......... ..........
T ..T ........... ............
,
~
............
Fine/very fine sandstones, siltstones of marked lateral continuity
9~
....... .~....[' . . -'r.........
,:~i?. ~ i o oli~i~o ~
Olisthostromes and slump deposits
T
--rThickening/ coarsening upwards sequences Outer fan Medium-tofine-graded sandstones of good lateral continuity
"'~ ....... ~' ..'....'. ~- ~." :.-:f. .=P.. , , . . . . . . ~ . . . ~ . . . . . . . . . . . ~ ~ I'J
50 m 1
FIG. 18. Typical facies associations for coarse-grained and fine-grained sediments in slope, fan and basin plain environments. (After Mutti & Ricci Lucchi 1972).
636
D.A.V. Stow and D.J.W. Piper
restricted, periodic anoxic bottom conditions can lead to the accumulation of organic-rich sediments (black shale facies) (e.g. Arthur et al., this volume). Facies associations & vertical sequences In ancient record it is often possible to recognize slope, fan and basin plain palaeoenvironments on the basis of the vertical and/or lateral association
I
,
. ,
of the characteristic facies outlined above. Thus, Mutti & Ricci Lucchi (1972) proposed facies associations for each of these three main environments (Fig. 18). Finer-grained facies make up a large part of these facies associations. Within these associations there are commonly smaller scale vertical sequences of beds that appear characteristic of different subenvironments or morphological elements in the deep sea (Fig. 19). For example, fining-upwards sequences
e,l,ee MID-FAN/SLOPE CHANNELS
PROXIMAL CANYON OR C H A N N E L
D E B R I S - SLUMP MASSES
CONTOURITE MOUND
DISTRIBUTARY CHANNELS
SANDY LOBE
SILTY-SANDY DISTAL LOBES
MUD-FILLED CHANNELS
PROXIMAL MUD L O B E
DISTAL SILTMUD L O B E
mud turbldites thin silt lurbidltes
6
send turbldltes
=c::mum
I
p e b b l y sand + gravel "turbidites"
I debrltes
PROXIMAL LEVEE
DISTAL LEVEE
INTERCHANNEL
OPEN S L O P E
slumps contourlte slits bl&ck s h l l e s
pelagites § ~emlpelagltes LEGEND
BASE-OF-SLOPE BASIN WEDGE
OR
OVER-SUPPLIED BASIN
UNDER-SUPPLIED BASIN
RESTRICTED BASIN
FIG. 19. Typical vertical facies sequences for various deep-water sub-environments (or morphological elements). Note variation in scale and in grading characteristics of the different sequences. (From Stow 1984a).
Deep-water fine-grained sedimen ts: facies models of turbidites are often interpreted as representing channel- or canyon-fill deposits, whereas coarsening-upwards sequences are taken as indicative of lobe deposits (Ricci Lucchi 1975; Rupke 1977; Walker 1978). Vertical sequences of predominantly fine-grained facies have been described by Shanmugam (1980); Stow et al. (1982); and Stow (1984a) among others. In this volume, papers by Hill; Picketing; Stow, Alam & Piper; Bourrouilh & Gorsline; Hayward; and Isaacs (all this volume) give further specific examples of sequences of a variety of facies in different environments. Canyons and channels can be filled by muddy sediments rather than coarse-grained facies, although much of the fill comprises slumps and debrites as well as mud turbidites and hemipelagites in a rather chaotic vertical sequence (e.g. Coleman et al. 1983). Stow (1984c) has interpreted thickening-upward followed by thinning upward sequences ofbiogenic-mud turbidites and debrites from the south-east Angola Basin as representing progradational-regradational submarine fan facies. The more distal parts of fan lobes or finer-grained terminal lobes on large muddy fans appear to show symmetrical vertical sequences of silt and mud turbidites (e.g. Pickering 1982b). Irregular or relatively uniform sequences of interbedded turbidites, contourites and hemipelagites/pelagites appear to be more characteristic of levees, interchannel, open slope and some basinal environments. These commonly differ in the relative proportions of the main facies present. More purely contourite sequences occur through contourite drifts (e.g. Gonthier et al., this volume; Stow & Holbrook, this volume), and these apparently show an irregular alternation of sandy, silty and muddy contourite facies, although very little data are yet available. Thick sequences of regularly-spaced pelagite-hemipelagite alternations characterize many low energy slope and basinal environments.
Discussion In this section we highlight briefly some interesting and problematical areas, mostly concerning detailed characteristics of the deep-water finegrained facies outlined above. These are some of the areas of current research and where future advances are likely to be made. Silt-mud lamination Silt-laminated muds are a very common deep-sea facies and occur repeatedly in the geological
637
record. However, their mode of emplacement and the process of lamina formation is not fully understood, and many different mechanisms have been proposed (see reviews by Stow & Bowen 1978, 1980). In many, perhaps most, cases, the lamination is current-induced. Velocity fluctuation within a single current (Lombard 1963), reflection of a turbidity current from the walls of a small basin (Van Andel & Komar 1969), the quasi-cyclic bursting process in boundary layers (Hesse & Chough 1980), and a series of small distinct flows or suspension clouds (Dzulynski & Radomski 1955) have all been suggested as the prime cause of such lamination. Perhaps more widely applicable are the processes of congregational sorting described by Piper (1972b), and of shear sorting of clay flocs from silt grains during final deposition through the base of the bottom boundary layer (Stow & Bowen 1978). This latter mechanism has received some support from experimental work with mud-water suspensions (see Kranck, this volume). Although these kinds of processes would apply equally to bottom currents, the association of abundant laminae with bottom currents as proposed by Hollister & Heezen (1972) has not subsequently been confirmed (see e.g. Hill, this volume; Shor et al., this volume). The laminae that have been described from contourites of sediment drifts are irregular or wispy and discontinuous (Piper & Brisco 1975; Stow 1982; Gonthier et al. this volume). Those of turbidites are more regular, distinct and continuous (e.g. Hill; Stow, Piper & Alam; Pickering; all this volume). In anoxic environments that cannot support a burrowing infauna or epifauna, pelagic and hemipelagic sediments can also preserve a primary lamination. This lamination is not currentinduced but results from periodic changes in the type of sediments supplied. The laminae are thus more truly varves, though not necessarily of seasonal periodicity (e.g. Isaacs, this volume; Robertson, this volume). Arthur et al. (this volume) show that lamination in black shales is in some cases turbiditic, in other cases pelagic and in other cases is better described as a fissility rather than true silt-mud alternation.
Mud fabric & fissility The fabric of sands and gravels has long been used as a diagnostic of depositional process and flow direction and silt fabric in fine-grained sediments can be used in a similar manner (Piper 1972a,b; Stow 1979b). However, there has been very much less work of an equivalent nature on the fabric of muds and shales (see reviews by
63 8
D.A.V.
Stow and D.J.W.
Moon 1972; Bennett et al. 1977; O'Brien 1981; Moon & Hurst, this volume). O'Brien (1981) and Moon & Hurst (this volume) agree that the development of fissility in shales is associated with an original orientation of clay flakes parallel to bedding. However, it remains uncertain whether the parallel orientation results from deposition of clay in the dispersed state as a result of an organic or inorganic deflocculant, or from deposition of oriented floc domains and subsequent mechanical reorientation on burial. Moon & Hurst (this volume) further discuss the importance of shale fissility in the generation and primary migration of hydrocarbons from organic-rich black shales. Geotechnical properties The study of sea floor soil mechanics has lagged behind that of its terrestrial counterpart, but has recently gained great impetus from the offshore oil, gas and minerals industries. In an important synthesis, Bennett et al. (1977) review the work on clay fabric and methodology and begin to relate different clay fabric types to selected geotechnical properties, depth of burial and laboratory consolidation loads. They suggest that for smectite or illite-rich muddy sediments, low void ratios result from high density packing of oriented clay particles, whereas high void ratios develop in sediments having non-oriented chain-like domains of clay particles. The possible relationship of geotechnical properties to particular fine-grained facies deposited by different deep-sea processes has been investigated by Hein & Gorsline (1981) for muddy sediments in various borderland basins off California (see also Gorsline et al., this volume). In a study of various hemipelagic sediments from the Guatemalan margin, including the slope, trench and Cocos oceanic plate environments, Faas (1982 and this volume) identifies marked regional trends in geotechnicai properties. He notes high plasticity in the upper slope mudflow and debris flow deposits, together with high organic-carbon content, and a decrease downslope to lowest values in the trench-fill turbidites. The siliceous hemipelagites of the Cocos plate also have a relatively high plasticity index and are relatively overcompacted compared to oozes. Bioturbation Trace fossil and bioturbation assemblages (ichnofacies) are very common in fine-grained sediments throughout the deep sea and are useful for sedimentological and environmental interpreta-
Piper
tion (Seilacher 1967, 1978; Crimes & Harper 1977; Werner & Wetzel 1982; Wetzel 1982 and this volume). Various attempts have been made to summarize the distribution and character of ichnofacies assemblages in relation to bathymetry, sediment type, ecological stress and biotic diversity and density (e.g. Potter et al., 1980, p. 44). Whereas early work suggested a Zoophycos slope assemblage and a Nereites basinal assemblage, the situation is now seen to be more complex with at least five different assemblages found in the deep sea (Werner & Wetzel 1982). Although there is not a simple relationship between sediment facies and ichnofacies, it appears that in many cases the three main facies groups do have a characteristic suite of burrows and/or degree of bioturbation. Pelagites and hemipelagites are very thoroughly bioturbated and may show a wide range of burrow types with several superimposed tiers being present. Contourites show almost complete bioturbation but insufficient to have destroyed all primary structures. Different burrow types characterize the more muddy and more silty contourite facies. Turbidites tend to be markedly less bioturbated with a concentration of burrows towards the tops of individual beds and a still more restricted diversity of burrow types. Superimposed tiers of burrows may also be present where there has been sufficient time between successive flows. Turbidity currents can apparently introduce an exotic shallow-water burrowing fauna into a deepwater setting (Wetzel, this volume). Ichnofacies assemblages are also proving to be an important characteristic for understanding the depositional environment and origin of black shales. In particular, the nature of the burrows in the interbedded non organic-rich facies and in the transition zone to black shales appear related to bottom water oxygenation (Byers 1977; Ekdale 1980; Stow & Dean 1984). Bioturbation is presumably absent from many black shales because of the anoxic bottom conditions during sedimentation. However, there are many examples of apparently oxic environments and sediments that still lack bioturbation. The other factors that deter burrowing organisms are still poorly known. Textures The analysis of grain size and other textural attributes of fine-grained sediments can provide very useful information on the processes and mechanics of deposition (e.g. Piper 1973; Rivi6re 1977; McCave, this volume; Kranck, this volume). However, textural analyses are more difficult to perform on consolidated material, and
Deep-water fine-grained sediments."facies models hence comparison of ancient and modern sediments is difficult. There is commonly a clear textural distinction between current-deposited and pelagic-settled sediments (e.g. Passega 1964). This can be seen as a sorting parameter, on a plot of coarser one percentile against median diameter, or in terms of the shape of the cumulative grain-size distribution curve. Using Rivi6re's (1977) terminology, the grain-size curves tend towards logarithmic for pelagic/hemipelagic sediments, hyperbolic (coarse tail) for muddy contourites and parabolic (fine tail) for turbidites and silty or fine sandy contourites. The types of grain-size sequence or grading are also diagnostic of the different facies: distinct positive grading and positive grading through grouped silt laminae are characteristic of turbidites; more irregular, alternating positivenegative grading is characteristic of contourites; and negative grading is most pronounced through the turbidite-pelagite transition in biogenic turbidites. Grain size as analysed in the laboratory differs significantly from grain size during deposition in two respects: (1) the clay fraction is analysed in the dispersed state but usually deposited as flocs of various sizes; and (2) the correspondence between grain size and hydraulic equivalence of biogenic and terrigenous material is poorly known. The measurement of grain size of suspended sediment in sea water allows greater insight into the depositional state of fine-grained material (e.g. McCave, this volume; Eittreim, this volume). Kranck (this volume) attempts to distinguish between grain-size populations that have been deposited as flocs and those that have settled in the dispersed state. There has been little serious attempt to determine the hydraulic equivalence of biogenic material (but see Berger & Piper 1972).
Composition and colour There are many different aspects to the composition of fine-grained sediments, some of which have received much attention and others less attention (see, for example, Potter et al. 1980; Tissot & Welte 1978). The biogenic content is clearly important for both biostratigraphic and palaeoecological information. Clay mineralogy has been much used in palaeoenvironmental and provenance studies. Inorganic geochemistry is beginning to be used more for provenance study. Organic geochemistry is vital for our understanding of hydrocarbon source-rock deposition and potential. What seems to be lacking most in our understanding and interpretation of compositional characteristics is better use of a combined
639
approach involving a number of separate compositional studies on the same suite of rocks. Although in individual cases compositional criteria are extremely valuable in distinguishing between turbidite, contourite and pelagite facies (e.g. Piper 1978; Stow & Lovell 1979; Spears & Amin 1981), such criteria are not readily generalized as so much depends on the local source and supply, competition from other components, and so on. The colour of fine-grained sediments is closely related to composition and has been a topic of long-standing interest, but one in which relatively little advance has been made in our understanding of the origin ofcolour and hence in using it as a diagnostic feature (but see McBride 1974; Pettijohn 1975; Potter et al. 1980). Potter et al. (1980) present a useful synthesis of what we know so far and have attempted to quantify the relationship of shale colour to both carbon content and the oxidation state of iron.
Diagenesis Of prime importance to all studies of ancient fine-grained sediments is the question of diagenesis. The twin processes of mechanical and chemical diagenesis can so profoundly alter the structural, textural, compositional, colour and fabric characteristics of deep-sea facies, that their use as criteria for distinction and for interpreting depositional history must necessarily change. The clay-rich beds in a pelagite-hemipelagite sequence can be reduced to thin shale partings by carbonate dissolution (e.g. Fischer & Arthur 1977). Primary sedimentary structures can be destroyed and deformational structures introduced during compaction (e.g. Rieke & Chilingarian 1974). Clay minerals undergo progressive alteration with increased burial pressures and temperatures (e.g. Perry & Hower 1970). Organic carbon is subject to progressive maturation into hydrocarbons with increasing temperature (e.g. Tissot & Welte 1978). Fluids released and mobilized during these various processes migrate through the formation and cause further chemical changes (e.g. Neglia 1979). These and numerous other effects of diagenesis we do not discuss here, and were not discussed at any length at the Fine-Grained Sediments Workshop in Halifax. We would like simply to emphasize the importance of considering diagenetic history in any study of fine-grained sediments in the deep sea. One of the key areas of research in the near future must surely be the development of diagenetic models, for both fine-grained and coarse-grained sediments, to complement the facies models we have outlined in this paper.
640
D.A.V. Stow and D.J.W. Piper
Conclusions (1) The wealth of data that exists and the depths of our knowledge on fine-grained sediments has increased dramatically in the past five to ten years. The holding of a workshop on just the deeper water fine-grained sediments and the editing of this volume has enabled us to take stock of where we have reached and where we should be going in this field of research. In this paper we have attempted a synthesis of facies models and have drawn heavily on other papers in the volume. (2) We identify three broad facies groups in the deep sea: turbidites, contourites and pelagites-hemipelagites. These are related primarily to the processes of deposition by, respectively, gravity-driven turbidity currents, thermohaline/wind-driven normal bottom currents, and vertical settling through the water column. Within each of these groups we describe several distinct facies models: turbidites (1) silt turbidites (2) mud turbidites (3) biogenic turbidites (4) disorganized turbidites contourites (1) muddy contourites (2) silty/fine sandy contourites pelagites-hemipelagites (1) pelagic oozes (2) muddy pelagic oozes (3) pelagic clays (4) hemipelagites These facies models, based on a large amount of observational data on both modern and ancient sediments, are intended as a summary of principal sedimentary characteristics and as a basis for interpretation of the depositional processes involved. Neither the complete sets of characteristics nor the full
sequences of sedimentary structures will necessarily be observed in every section examined and we have tried to show, to some extent, the variability to be expected in each case. (3) We have also given a brief summary of the hydrodynamic interpretation for each facies model. However, it should be emphasized that there appears to be a continuum between the various processes that operate in the deep sea and hence there will be a continuum of resultant facies deposited. (4) The different fine-grained facies show certain characteristic patterns of occurrence both horizontally over the surface of the sea floor and vertically in boreholes or in ancient rock successions. They occur interbedded with and adjacent to each other and to coarser-grained facies. These patterns of distribution, which we have briefly described, can be used to help identify depositional palaeoenvironments. (5) Finally, in discussion, we attempt to do little more than highlight some of the interesting and problematical areas of current research in fine-grained deep-water sediments. We mention, in particular, certain aspects of facies characteristics including: silt-mud lamination, fabric and fissility, geotechnical properties, bioturbation, textures, composition and colour, and diagenesis. ACKNOWLEDGEMENTS:In a synthesis paper of this kind we have many people to thank for their helpful and stimulating discussion. In particular, we thank all the participants in the International Workshop on Fine-Grained Sediments held at Halifax, Canada in 1982. DAVS acknowledges financial support from the Natural Environment Research Council, UK, and technical and secretarial support at the Grant Institute of Geology. We are very grateful for the helpful comments of our reviewers, Ali Aksu, Donn Gorsline, Phil Hill and Julian Holbrook.
References ANATRA, S., ACKERMANN,T. & HOMEWOOD,P. 1980. Les facies de l'Ultrahev6tique du Montsalvens (Pr6alpes Externes) et de la r6gion d'Anzeinde (Pr6alpes Internes). Ecol. geol. Heh'., 73/1,283-92. ANKETELL,J.M. & LOVELL,J.P.B. 1976. Upper Llandoverian Grogal Sandstones and Aberystwyth Grits in the New Quay area, Central Wales: a possible upwards transition from contourites to turbidites. Geol. J., 11, 107-8. ARRHENIUS,G. 1963. Pelagic sediments. In: Hill, M.N. (ed.), The Sea. Interscience, New York. 3, 655-727. AUDLEY-CHARLES,M.G. 1965. A geochemical study of
Cretaceous ferromanganiferous sedimentary rocks from Timor. Geochim. cosmochim. Acta., 29, 1153-73. BARRETT, T.J. 1982. Jurassic bedded cherts from the North Apennines, Italy: dyscyclic sedimentation in the deep pelagic realm. In: Einsele, G. & Seilacher, A. (eds), Cyclic and Event Stratification. SpringerVerlag, Berlin. 389-403. BARTOLINI, C., BERLATO,S. & BORTOLOTTI,V. 1975. Upper Miocene shallow-water turbidites from west Tuscany. Sed. Geol., 14, 77-122. BEALL, A.O. & FISCHER, A.G. 1969. Sedimentology.
Deep-water fine-grained sediments."facies models Init. Repts. DSDP, 1, US Govt. Print. Office, Washington, DC. 521-93. BEIN, A. & WEILER, Y. 1976. The Cretaceous Talme Yafe Formation: a contour current shaped sedimentary prism of calcareous detritus at the continental margin of the Arabian Craton. Sedimentology, 23, 511-32. BENNETT, R.H., LAMBERT, D.N. & HULBERT, M.H. 1977. Geotechnical properties of a submarine slide area on the US continental slope northeast of Wilmington Canyon. Mar. Geotech., 2, 245-61. BERGER, W.H. 1970. Biogenic deep sea sediments: fractionation by deep sea circulation. Bull. geol. Soc. Am., 81, 1385-401. 1974. Deep-sea sedimentation. In: Burk, C.A. & Drake, C.D., (eds), The Geology of Continental Margins. Springer-Verlag, New York. 213-4 1. & PIPER, D.J.W. 1972. Planktonic Foraminifera: differential settling, dissolution and redeposition. Limnol. Oceanog., 17, 275-87. BISCAYE, P.E. & EITTREIM, S.L. 1977. Suspended particulate loads and transports in the nepheloid layer of the abyssal Atlantic Ocean. Marine Geol., 23, 155-72. BLANPIED, C. & STANLEY, D.J. 1981. Uniform mud (unifite) deposition in the Hellenic Trench, eastern Mediterranean. Smiths. Contrib. Marine Sci., 13, 40 PP. BOUMA, A.H. 1962. Sedimentology of Some Flysch Deposits. Elsevier, Amsterdam. 168 pp. , MOORE, G.T. & COLEMAN, J.M. (eds). 1978. Framework, facies and oil-trapping characteristics of the upper continental margin. Am. Ass. Petrol. Geol. Studies in Geology, 7. BOWEN, A.J., NORMARK,W.R. & PIPER, D.J.W. 1984. Modelling of turbidity currents on Navy submarine fan, California Continental Borderland. Sedimentology, (in press). BROECKER, W.S. 1974. Chemical Oceanography. Harcourt Brace Jovanovich, Inc., New York. 214 PP. BYERS, C.W. 1977. Biofacies patterns in euxinic basins: a general model. Soc. econ. Palaeo. Min. Spec. Pub., 25, 5-18. CARRASCO-V, B. 1977. Albian sedimentation of submarine authochthonous and allochthonous carbonates, east edge of the Valles-San Luis Potosi platform, Mexico. Soc. econ. Palaeo. Min. Spec. Pub., 25, 263-72. CARTER,L. & SCI-IAFER,C.T. 1983. Interaction of the Western Boundary Undercurrent with the continental margin off Newfoundland. Sedimentology, 30, 751-78. CrIOUGH, S.K. & HESSE, R. 1980. The Northwest Atlantic Mid-Ocean Channel of the Labrador Sea" III head spill vs. body spill deposits from turbidity currents on natural levees. J. sed. Petrol., 50, 227-34. & HESSE,R. (in press). Contourites from the Eirik Ridge, south of Greenland. Sed. Geol. COLEMAN, J.C. 1976. Deltas: Processes of Deposition and Models for Exploration. Continuing Education Publ. Co., Inc., Champaign. 102 pp. --, PRIOR, D.B. & LINDSAY, J.F. 1983. Deltaic -
-
-
-
-
-
641
influences on shelf-edge instability processes. Soc. econ. Palaeo. Min. Spec. Pub., 33, 121-38. COOK, H.E. & ENOS, P. (eds.). 1977. Deep-water carbonate environments. Soc. econ. Palaeo. Min. Spec. Pub., 25. & TAYLOR,M.E. 1977. Comparison of continental slope and shelf environments in the Upper Cambrian and lowest Ordovician of Nevada. Soc. econ. Palaeo. Min. Spec. Pub., 25, 51-82. CREMER, M. 1981. Distribution des turbidites sur l'6ventail du Canyon du Cap-Ferret. Bull. Inst. Geol. Basin d'Aquitaine, Bordeaux., 30, 51-69. -1983. Approches SOdimentologiques et G~ophysique des Accumulations Turbiditiques. Th6se, Doctorat d'Etat, Universit6 Bordeaux I. CREVELLO, P.D. & SCHLAGER, W. 1980. Carbonate debris sheets and turbidites, Exuma Sound, Bahamas. J. sed. Petrol., 50, 1121-48. CRIMES, T.P. & HARPER,J.C. (eds). 1977. Trace fossils. Geol. J. Spec. Issue, 9, 351 pp. CURRAY, J.R. & MOORE, D.G. et al. 1979. Drilling the Gulf of California, DSDP leg 64. Geotimes, 24(7), 1 8 - 2 0 .
DAMUTH, J.E. 1975. Echo-character of the western equatorial Atlantic floor and its relationship to the dispersal and distribution of terrigenous sediments. Marine Geol., 18, 17-45. 1978. Echo character of the Norwegian-Greenland Sea: relationship to Quaternary sedimentation. Marine Geol., 28, 1-36. 1980. Use of high frequency (3.5-12kHz) echograms in the study of near-bottom sedimentation processes in the deep-sea: a review. Marine Geol., 38, 51-75. & KUMAR,N. 1975. Late Quaternary depositional processes on the continental rise of the western equatorial Atlantic: comparison with the western North Atlantic and implications for reservoir rock distribution. Bull. Am. Ass. Petrol. Geol., 59, 2172-81. DAVIES, G.R. 1977. Turbidites, debris sheets and truncation structures in upper Palaeozoic deepwater carbonates of the Sverdrup Basin, Arctic Archipelago. Soc. econ. Palaeo. Min. Spec. Pub., 25, 221-48. DEAN, W.E., GARDNER,J.V. & CEPEK,P. 1981. Tertiary carbonate dissolution cycles on the Sierra Leone Rise, eastern equatorial Atlantic Ocean. Marine Geol., 39, 81-101. - - . , STOW, D.A.V., BARRON,E. & SCHALLREUTER,R. 1984. A revised sediment classification for siliceousbiogenic, calcareous-biogenic and non-biogenic compounds. Init. Repts. DSDP, 75. US Govt. Print. Office, Washington DC. DOYLE, L.J. & PILKEY, O.H. (eds). 1979. Geology of continental slopes. Soc. econ. Palaeo. Min. Spec. Pub., 27. DRAKE, D.E., HATCHER, P.G. & KELLER, G.H. 1978. Suspended particulate matter and mud deposition in Upper Hudson submarine canyon. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 33-41. DZULYNSKI,S. & RADOMSKI,A. 1955. Origin of groove
642
D.A.V. Stow and D.J.W. Piper
casts in the light of turbidity current hypothesis. Acta Geol. Pologne., 5, 47-56. EINSELE, G. 1982. Limestone-marl cycles (peridotites): diagnosis, significance, causes--a review. In: Einsele, G. & Seilacher, A. (eds), Cyclic and Event Stratification. Springer-Verlag, Berlin. 8-53. - & S E I L A C H E R , A. (eds). 1982. Cyclic and Event Stratification. Springer-Verlag, Berlin. 536 pp. EKDALE, A.A. 1980. Trace fossils and stagnation of deep-sea basins. (Abstr.). Bull. Am. Ass. Petrol. Geol., 64, 704. ENOS, P. 1977. Tamabru limestone of the Pozo Rico trend, Cretaceous, Mexico. Soc. econ. Palaeo. Min. Spec. Pub., 25, 273-14. FAAS, R.W. 1982. Gravitational compaction patterns determined from sediment cores on the Guatemalan Transect: continental slope to Middle America Trench to Cocos Plate, on DSDP leg 67. In: Auboin, J., von Huene, R. et al. Initl. Repts. DSDP, 67, US Govt. Print. Office, Washington, DC. 617-38. FAUGI~RES, J-C. et al. 1979. Evolution de la s6dimentation profonde au Quaternaire r6cent dans le Bassin nord-atlantique: corps s6dimentaire et s6dimentation ubiquiste. Bull. Soc. g~ol. France., 21, 585-601. , GAYET, J., GONTHIER, E., POUT1ERS,J. ~; NYANG, I. 1982. La Dorsale M6dio-Atlantique entre 43 ~ et 56~ facies et dynamique s6dimentaire dans plusieurs types d'environements au Quaternaire R6cent. Bull. Inst. Geol. Bassin d'Aquitaine, Bordeaux., 31, 195-215. , STOW, D.A.V. & GONTHIER, E. (1984) Contourite drift moulded by deep Mediterranean outflow. Geology, (in press). FISCHER, A.G. & ARTHUR, M.A. 1977. Secular variations in the pelagic realm. Soc. econ. Palaeo. MiD. Spec. Pub., 25, 19-50. FOLK, R.L. & MCBRIDE, E.F. 1978. Radiolarites and their relation to subjacent 'oceanic crust' in Liguria, Italy. J. sed. Petrol., 48, 1069-102. GARRISON, R.E. & FISCHER, A.G. 1969. Deep-water limestones and radiolarites of the Alpine Jurassic. Soc. econ. Palaeo. Min. Spec. Pub., 14, 20-55. GINSBURG, R.N. 1975. (ed.). Tidal Deposits: ,4 Casebook of Recent Examples and Fossil Counterparts. Springer-Verlag, Berlin. 428 pp. GONTHIER, E., CREMER,M., FAUGERES,J.C. & NIANG, I. 1982. Turbidites et contourites sur la marge sud-arabique (Sukra). Reunion Ann. Sci. de la Terre., Paris. 284 pp. GORSLINE, D.S. 1978. Anatomy of margin basins. J. sed. Petrol., 48, 1055-68. - 1981. Fine sediment transport and deposition in active margin basins. In: Douglas, R.G., Colburn, I.P. & Gorsline, D.S. (eds), Depositional systems of active continental margins. Pacific Section, Soc. econ. Palaeo. Min., Short Course Notes, Notes, Los Angeles, 1981.39-59. GRADSTEIN, F.M., SHERIDAN, R.E. et al. 1981. Leg 76: Western North Atlantic Ocean. Joides Journal, VII. 29-35. GRIFFIN, J.J., WlNDON, H. & GOLDBERG,E.D. 1968. The distribution of clay minerals in the world ocean. Deep Sea Res., 15, 433-59.
GRIGGS, G.B. & KULM, C.D. 1970. Sedimentation in Cascadia Deep Sea Channel. Bull. geol. Soc. Am., 81, 1361-84. HALL, B.A. & STANLEY, D.J. 1973. Levee-bounded submarine base-of-slope channels in the Lower Devonian Seboomook Formation, northern Maine. Bull. geol. Soc. Am., 84, 2101-10. HARMS, J.C. & FAHNESTOCK,R.K. 1965. Stratification, bed forms and flow phenomena (with an example from the Rio Grande). Soc. econ. Palaeo. MiD. Spec. Pub., 12, 84-115. HEEZEN, B.C., HOLLISTER, C.D. & RUDDIMAN, W.F. 1966. Shaping of the continental rise by deep geostrophic contour currents. Science., 152, 502-8. HEIN, F.J. & GORSLINE,D.S. 1981. Geotechnical aspects of fine-grained mass flow deposits: California continental borderland. Geo-Marine Letters, 1, 1-5. HESSE, R. 1975. Turbiditic and non-turbiditic mudstone of Cretaceous flysch sections of the East Alps and other basins. Sedimentology, 22, 387-416. -• CHOUGH, S.K. 1980. The Northwest Atlantic Mid-Ocean Channel of the Labrador Sea: II Deposition of parallel-laminated levee muds from the viscous sub-layer of low density turbidity currents. Sedimentology, 27, 697-71 I. HICKS, D.M. 1981. Deep-sea fan sediments in the Torlesse zone, Lake Chau, South Canterbury, New Zealand. N Z J. Geol. Geophys., 24, 209-30. HILL, P.R. 1981. Detailed Morphology and Late Quaternary Sedimentation on the Nova Scotia Slope South of Halifax. Unpublished Ph.D. Thesis, Dalhousie Univ. Halifax, NS, Canada. 331 pp. - & BOWEN, A.J. 1983. Modern sediment dynamics at the shelf-slope boundary off Nova Scotia. Soc. econ. Palaeo. MiD. Spec. Pub., 33, 265-78. HOFFERT, M. 1980. Les 'Argiles Rouges des Grands Fonds' clans le Pacifique Centre-est: Authigenkse, Transport, Diagen~se. Thesis, Universit6 Louis Pasteur de Strasbourg, M6moire No. 61,231 pp. HOLLISTER,C.D. & HEEZEN, B.C. 1972. Geologic effects of ocean bottom currents: western North Atlantic. In: Gordon, A.L. (ed.), Studies in Physical Oceanography. Gordon Breach Science Pubs. 2, 37-66. HOMEWOOD, P. & WINKLER, W. 1977. Les calcaires d6tritiques et noduleux du Malm des M6dianes Plastiques dans le Pr~alpes fribourgeoises. Bull. Soc. Frib. Sc. Nat., 66, 116-40. HORN, D.R., EWING, M., HORN, B.M. & DELACH,M.N. 1971. Turbidites of the Hatteras and Sohm Abyssal Plains, Western North Atlantic. Marine Geol., 11, 287-323. Hsu, K.J. & JENKYNS, H.C. (eds.). 1974. Pelagic Sediments: on land and under the sea. Spec. Publ. Int. Ass. Sediment., 1,447 pp. IMOTO, N. & FUKUTOMI, M. 1975. Genesis of bedded cherts in the Tamba Belt, southwest Japan. Assoc. Geol. Collabor. Japan Monogr., 19, 3542. INGERSOLL, R.V. 1977. Submarine fan facies of the upper Cretaceous great valley sequence, north and central California. Sed. Geol., 21, 205-30. INMAN, D.L. & NORDSTROM,C.E. 1971. On the tectonic and morphologic classification of coasts, d. Geol., 79, 1-21.
Deep-water fine-grained sediments."facies models ISAACS, C.M. 1981. Lithostratigraphy of the Monterey Formation, Goleta to Point Conception, Santa Barbara Coast, California. Am. Ass. Petrol. Geol. Field Guide, 4, 9-24. JACOBI, R.D. 1982. Microphysiography of the SE North Atlantic and its implication for the distribution of near bottom processes and related sedimentary facies. Bull. Inst. Geol. Bassin d'Aquitaine., 31, 31-46. JENKYNS, H.C. 1978. Pelagic environments. In: Reading, H.G. (ed.) Sedimentary Environments and Facies. Blackwell Scientific Publications, Oxford. 314-71. 1984. Pelagic environments. In: Reading, H.G. (ed.), Sedimentary Environments and Facies. 2nd edition. Blackwell Scientific Publications. Oxford. (In press). JIPA, D. & KIDD, R.B. 1974. Sedimentation of coarser grained interbeds in the Arabian Sea and sedimentation processes of the Indus Cone. In: Whitmarsh, R.E., Weser, O.E., Ross, D.A. et al., Initl. Repts. DSDP, 23, US Govt. Print. Office, Washington, DC. 471-92. JOHNSON, T.C., CARLSON, T.W. & EVANS, J.E. 1980. Contourites in Lake Superior. Geology, 8, 437-41. KAGAMI, H. 1979. Transformation of opaline silica in sediments from Bay of Biscay and Rockall Bank. In: Montadert, L., Roberts, D.G. et al. (eds), Init. Repts. DSDP, 48, 757-64. US Govt. Print. Office, Washington, DC. KALIN, O., PATACA, E. & RENE, O. 1979. Jurassic pelagic deposits from southeastern Tuscany: aspects of sedimentation and new biostratigraphic data. Ecol. Geol. HelL,., 72/3, 715-62. KARL,H.A., CARLSON,P.R. & CACCHIONE,D.A. 1983. Factors that influence sediment transport at the shelf break. Soc. econ. Palaeo. Min. Spec. Pub., 33, 219-32. KELTS, K. & ARTHUR, M.A. 1981. Turbidites after ten years of deep-sea drilling--wringing out the mop? Soc. econ. Palaeo. Min. Spec. Pub., 32, 91-127. KENNEDY,W.J. 1980. Aspects of chalk sedimentation in the southern Norwegian offshore. In: The Sedimentation of North Sea Reservoir Rocks, Norwegian Petroleum Society, Geilo. KOLLA, V., EITTREIM,S., SULLIVAN,L., KOSTECKI,J.A. & BURCKLE,L.H. 1980. Current-controlled abyssal microtopography and sedimentation in the Mozambique Basin, Southwest Indian Ocean. Marine Geol., 34, 171-206. KUENEN, PH.H. 1964. Deep-sea sands and ancient turbidites. In: Bouma, A.H. & Brouwer, A. (eds). Turbidites--Developments in Sedimentology, 3. Elsevier. 3-33. LAUGHTON, A.S., BERGGREN, W.A. et al. 1972. Initl. Repts. DSDP, 12. US Govt. Print. Office, Washington, DC. LISlTZIN, A.P. 1972. Sedimentation in the world ocean. Soc. econ. Palaeo. Min. Spec. Pub., 17, 218 pp. LOMBARD, A. 1963. Laminites--a structure of flyschtype sediments. J. sed. Petrol., 33, 14-22. LOVELL, J.B.P. & STOW, D.A.V. 1981. Identification of ancient sandy contourites. Geology, 9, 347-9. LUNDEGARD, P.D., SAMUELS, N.D. & PRYOR, W.A. -
-
643
1980. Sedimentology, petrology and gas potential of the Brallier Formation--Upper Devonian turbidite facies of the central and southern Apalachians. US Dept. Energy Report DOE/METC/5201-5,220 pp. MALDONADO,A. & STANLEY,D.J. 1976. The Nile Cone: submarine fan development by cyclic sedimentation. Mar. Geol., 20, 27-40. MCBRIDE, E.F. 1974. Significance of colour in red, green, purple, olive, brown and grey beds of Difunta Group, NW Mexico. J. sed. Petrol., 44, 760-8. MCCAVE, I.N. 1972. Transport and escape of finegrained sediment from shelf areas. In: Swift, D.J.P., Duane, D.B. & Pilkey, O.H. (eds), Shelf Sediment Transport: Process and Pattern., 10, 225-48. 1979. Diagnosis of turbidites at Sites 386 and 387 by particle-counter size analysis of the silt (2-40/~m) fraction. In: Tucholke, B.E., Vogt, P.R. et al. Initl. Repts. DSDP., 43, 395-405. --, LONSDALE,P.F., HOLLISTER, C.D. & GARDNER, W.D. 1980. Sediment transport over the Hatton and Gardar contourite drifts. J. sed. Petrol., 50, 1049-62. MCILREATH,I.A. & JAMES,N.P. 1978. Facies models 13. Carbonate slopes. Geoscience Canada., 5, 189-99. MOON, C.F. 1972. The microstructure of clay sediments. Earth Sci. Rev., 8, 303-321. MOORE, C. 1974. Turbidites and terrigenous muds: DSDP leg 25. In: Simpson, E.S.W., Schlich, R. et al., Initl. Repts. DSDP., 25, 441-79. MOORE, D.G. 1969. Reflection profiling studies of the Californian continental borderland: structure and Quaternary turbidite basins. Geol. Soc. Am. Spec. Pap., 107, 1-142. MULUNS, H.T. & NEUMANN, A.C. 1979. Deep carbonate bank margin structure and sedimentation in the northern Bahamas. In: Doyle, L.J. & Pilkey, O.H. (eds). Geology of Continental Slopes., Soc. econ. Palaeo. Min. Spec. Publ., 27, 165-92. MUTTI, E. 1977. Distinctive thin-bedded turbidite facies and related depositional environments in the Eocene Hecho Group (South-central Pyrenees, Spain). Sedimentology., 24, 107-31. -& R i c o Luccm, F. 1972. Le torbiditi dell' Appenino settentrionale: introduzione all'analisi di facies. Mere. Soc. geol. ltal., 11, 161-99. - - , NILSEN,T.H. & RIccl LUCCHI,F. 1978. Outer fan depositional lobes of the Laga formation (Upper Miocene and Lower Pliocene), East-Central Italy. In: Stanley, D.J. & Kelling, G. (eds), Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 210-23. NARDIN, T.R., HE1N, F.J., GORSLINE,D.S. & EDWARDS, B.D. 1979. A review of mass movement processes, sediment and acoustic characteristics, and contrasts in slope and base-of-slope systems versus canyonfan-basin floor system. Soc. econ. Palaeo. Min. Spec. Pub., 27, 61-73. NAYLOR, M.A. 1980. Origin of inverse grading in muddy debris flow deposits--a review. J. sed. Petrol., 50, I I I 1-6. 1981. Debris flow (olistostromes) and slumping on a distal passive continental margin: the Palombini limestone-shale sequence of the northern Apennines. Sedimentology, 28, 837-52.
-
-
-
-
644
-
-
-
D.A.V. Stow and D.J.W. Piper
NEGLIA, S. 1979. Migration of fluids in sedimentary basins. Bull. Ass. Am. Petrol. Geol., 63, 573-97. NELSON, C.H. 1976. Late Pleistocene and Holocene depositional trends, processes, and history of Astoria deep sea fan, North East Pacific. Marine Geol., 20, 129-73. , NORMARK,W.R., BOUMA,A.H. & CARLSON,P.R. 1978. Thin-bedded turbidites in modern submarine canyons and fans. In: Stanley, D.J. & Kelling, G. (eds.). Sedimentation in submarine canyons,fans and trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 177-89. NISBET, E.G. & PRICE, I. 1974. Siliceous turbidites: bedded cherts as redeposited, ocean ridge-derived sediments. Spec. Publ. Int. Ass. Sediment., 1, 351-66. NORMARK, W.R. & PIPER, D.J.W. 1972. Sediments and growth pattern of Navy deep sea fan, San Clemente Basin, California Borderlands. J. Geol., 80, 198-223. O'BRIEN, N.R. 1981. SEM study of shale fabric--a review. In: O'Hare, A.M.F. (ed.) Scanning Electron Microscopy 1981 I. SEM Inc., Chicago. 569-75. , NAKAZAWA, K. & TOKUHASHI, S. 1980. Use of clay fabric to distinguish turbiditic and hemipelagic siltstones and silts. Sedimentology, 27, 47-61. PASSEGA, A. 1964. Grain size representation by CM patterns as a geological tool. J. sed. Petrol., 34, 830-47. PERRY, E.A. & HOWER, J. 1970. Burial diagenesis in Gulf Coast pelitic sediments. Clays Clay Minerals., 18, 165-77. PETERSON, M.N.A. & EDGAR, N.T. et al. 1970. lnitl. Repts. DSDP, 2. US Govt. Print. Office, Washington, DC. PETT1JOHN, F.J. 1975. Sedimentary Rocks. 3rd edn. Harper & Row, New York. 628 pp. PICKERING, K.T. 1982a. The Kongstjord Formation--a late PreCambrian submarine fan in NE Finnmark, N Norway. Norges geol. Unders, 367, 77-104. 1982b. Middle-fan deposits from the Late PreCambrian Kangsfjord Formation submarine fan, NE Finnmark, N Norway. Sed. Geol., 33, 79-110. PILKEY, O.H., LOCKER, S.D. & CLEARY, W.J. 1980. Comparison of sand-layer geometry on flat floors of 10 modern depositional basins. Bull. Am. Ass. Petrol. Geol., 64, 841-56. PXPER, D.J.W. 1972a. Sediments of the Middle Cambrian Burgess Shale, Canada. Lethaia., 5, 169-75. 1972b. Turbidite origin of some laminated mudstones. Geol. Mag., 109, 115-26. 1973. The sedimentology of silt turbidites from the Gulf of Alaska. In: Kulm, L.D., von Huene, R. et al., Initl. Repts. DSDP, 18. US Govt. Print. Office, Washington, DC. 847-67. 1978. Turbidites muds and silts on deep-sea fans and abyssal plains. In: Stanley, D.J. & Kelling, G. (eds). Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson & Ross, Stroudsburg, Pa. 163-76. & BRISCO, C.D. 1975. Deep water continental margin sedimentation, DSDP leg 28, Antarctica. In: Hayes, D.E., Frakes, L.A. et al. Initl. Repts. DSDP, 28, US Govt. Print. Off., Washington, DC. 727-55. -
-
-
- - - , NORMARK,W.R. & INGLE,J.C. 1976. The Rio Dell Formation: a Plio-Pleistocene basin slope deposit in Northern California. Sedimentology, 23, 309-28. POTTER, P.E., MAYNARD, J.B. & PRYOR, W.A. 1980. Sedimentology of shale. Springer-Verlag, New York. 303 pp. REINHARD'r, J. 1977. Cambrian off-shelf sedimentation, central Appalachians. Soc. econ. Palaeo. Min. Spec. Pub., 25, 83-112. R i c o LUCCHI, F. 1975. Miocene paleogeography and basin analysis in the Periadriatic Apennines. Reprinted from Geology of Italy, C. Squyres, (ed.). Pet. Explo. Soc. Libya., 111. 1978. Turbidite dispersal in a Miocene deep-sea plain: the Marroso-Arenacea of the Northern Apennines. Geol. Mijnbouw., 57, 559-76. -1981. The Miocene Marroso-Arenacea turbidites, Romagna and Umbria Apennines. IAS 2nd E U R MJG., Excursion Guidebook. RIEKE, H.H. & CHILINGARIAN,G.V. 1974. Compaction of Argillaceous Sediments. Elsevier, Amsterdam. 424 pp. RIVIERE, A. 1977. MOthodes Granulom~triques: Techniques et Interpretation. Masson, Paris. 170 pp. RUPKE, N.A. 1975. Deposition of fine-grained sediments in the abyssal environment of the AlgeroBalearic Basin, western Mediterranean Sea. Sedimentology, 22, 95-109. -1977. Growth of an ancient deep-sea fan. J. Geol., 85, 725-44. & STANLEY, D.J. 1974. Distinctive properties of turbiditic and hemipelagic mud layers in the AlgeroBalearic Basin, western Mediterranean Sea. Smiths. Contrib. Earth. Sci., 13, 40 pp. SCRU'rTON, R.A. & STow, D.A.V. (1984). Seismic evidence for early Tertiary bottom current controlled deposition on the Charlie Gibbs Fracture Zone. Marine Geol., in press. SEILACHER,A. 1967. Bathymetry of trace fossils. Marine Geol., 5, 413-29. 1978. Use of trace fossil assemblages for recognising depositional environments. In: Basan, P.B. (ed.), Trace Fossil Concepts. Soc. econ. Palaeo. Min. Short Course No. 5, Oklahoma. SHANMUGAM, G. 1980. Rhythms in deep sea, finegrained turbidite and debris-flow sequences, Middle Ordovician, eastern Tennessee. Sedimentology, 27, 419-32. SHEPARD, F.C., PHLEGER, F.B. & ANDEL, TJ. H. VAN. (eds) 1960. Recent Sediments, Northwest Gulf of Mexico. Am. Ass. Petrol. Geol. Mere., Tulsa, Oklahoma. 394 pp. --, MARSHALL, N.F. McLOUGHL1N, P.A. & SULLIVAN, G.G. 1979. Currents in submarine canyons and other sea valleys. Am. Ass. Petrol. Geol. Studies in Geology. 8, 179. SHOR, A.N. & POORE, R.Z. 1978. Bottom currents in ice rafting in the North Atlantic: interpretation of Neogene depositional environments of Leg 49 cores. In: Luyendyk, B.P., Cann, J.R. et aL Initl. Repts. DSDP, 49. US Govt. Print. Off., Washington, DC. 859-72. --, LONSDALE, P., HOLLISTER, C.D. & SPENCER, D. 1980. Charlie-Gibbs fracture zone: bottom-water -
-
-
-
-
-
Deep-water fine-grained sediments." facies models transport and its geologic effects. Deep Sea Res., 27A, 325-45. SPEARS, D.A. & AMIN, M.A. 1981. A mineralogical and geochemical study ofturbidite sandstones and interbedded shales, Man Tor, Derbyshire, UK. Clays Clay Minerals., 16, 333-45. STANLEY, D.J. 1981. Unifites: structureless muds of gravity-flow origin in Mediterranean basins. GeoMarine Letters., 1, 77-83. 1983. Parallel-laminated deep-sea muds and coupled gravity flow--hemipelagic settling in the Mediterranean. Smiths. Contrib. Marine Sci., 19, 19 pp. -& KEELING, G. (eds). 1978. Sedimentation in submarine canyons, fans and trenches. Dowden, Hutchinson & Ross, Stroudsborg, Pa. & MALDONAOO, A. 1979. Levantine Sea-Nile Cone lithostratigraphic evolution: quantitative analyses and correlation with paleoclimatic and eustatic oscillations in the late Quaternary. Sed. Geol., 23, 37-65. -- & -1981. Depositional models for fine-grained sediment in the western Hellevic Trench, eastern Mediterranean. Sedimentology, 28, 273-90. & WEAR, C.M. 1978. The 'mud-line': an erosiondeposition boundary on the upper continental slope. Marine Geol., 28, M 19-M29. --, SWIFT, D.J.P., SILVERBERG,N., JAMES, N.P. & SUTTON, R.G. 1972. Late Quaternary progradation and sand 'spillover' on the outer continental margin offNova Scotia, southeast Canada. Smiths. Contrib. Earth Sci., 8, 88 pp. STOW, D.A.V. 1977. Late Quaternary Stratigraphy and Sedimentation on the Nova Scotian Outer Continental Margin. Unpub. Ph.D. Thesis, Dalhousie University. 1979a. Distinguishing between fine-grained turbidites and contourites on the Nova Scotian deep water margin. Sedimentology, 26, 371-87. 1979b. Method for silt and sand fabric analysis in deep-sea cores. J. sed. Petrol., 39, 627-30. 1981. Laurentian Fan: morphology, sediments, processes and growth pattern. Bull. Am. Ass. Petrol. Geol., 65, 375-93. 1982. Bottom currents and contourites in the North Atlantic. Bull. Inst. Geol. Bassin d'Aquitaine., 31, 151-66. 1984a. Deep-sea clastics: where are we and where are we going? In: Brenchley, P.J. & Williams, B.P.J. (eds), Sedimentology: Recent Developments and Applied Aspects. Geol. Soc. Lond. Spec. Pub. (in press). 1984b. Anatomy of debris flow deposits. In: Hay, W.W., Sibuet, J.C. et al., Initl. Repts.,DSDP, 75. US Govt. Print. Office, Washington, DC. 1984c. Turbidite facies, associations and sequences in the south-eastern Angola Basin. In: Hay, W.W., Sibuet, J.C. et al., Initl. Repts. DSDP, 75. US Govt. Print. Office, Washington, DC. (in press). Deep-sea clastics. In: Reading, H.G. (ed.), Sedimentary Environments and Facies. 2nd edition. Blackwell Scientific Publications, Oxford. & BOWEN, A.J. 1978. Origin of lamination in -
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
645
deep-sea fine-grained sediments. Nature, 274, 324-28. & -1980. A physical model for the transport and sorting of fine-grained sediments by turbidity currents. Sedimentology, 27, 31-46. & DEAN, W.E. 1984. Middle Cretaceous black shales at Site 530 in the southeastern Angola Basin. In: Hay, W.W., Sibuet, J.C. et al., Initl. Repts. DSDP, 75. US Govt. Print. Office, Washington, DC. & LOVELL,J.P.B. 1979. Contourites: their recognition in modern and ancient sediments. Earth Sci. Reviews, 14, 251-91. & SHANMUGAM,G. 1980. Sequence of structures in fine-grained turbidites: comparison of recent deepsea and ancient flysch sediments. Sed. Geol., 25, 23-42. --, BisHoP, C.D. & MILES, S.J. 1982. Sedimentology of the Brae Oil Field, North Sea: fan models and controls. J. Petrol. Geol., 5, 129-48. TAYLOR, P.T., STANLEY, B.J., SIMKIN, T. & JAHN, W. 1975. Gillis seamount: detailed bathymetry and modification by bottom currents. Marine Geol., 19, 139-57. THOMSON, A.F. & THOMASSON, M.R. 1969. Shallow to deep water facies development in the Dimple Limestone (Lower Pennsylvanian), Marathon Region, Texas. Soc. econ. Palaeo. Min. Spec. Pub., 14, 57-78. TISSOT, B.P. & WELTE, D.H. 1978. Petroleum Formation and Occurrence. Springer-Verlag, Berlin. VAN ANDEL, T.H. & KOMAR, P.D. 1969. Ponded sediments of the mid-Atlantic ridge between 22 ~+23 ~ North latitude. Bull. geol. Soc. Am., 80, 1163-90. VAN DER LINGEN, G.J. 1969. The turbidite problem. N Z J. Geol. Geophys., 12, 7-50. VON STACKELBER6,U., VON RAD, U. & ZOBEL, B. 1979. Asymmetric sedimentation around Great Meteor Seamount (North Atlantic). Marine Geol., 33, 117-32. WALKER, R.G. 1975. Upper Cretaceous resedimented conglomerates at Wheeler Gorge, California: description and field guide. J. sed. Petrol., 45, 105-12. 1978. Deep water sandstone facies and ancient submarine fans: Models for exploration for stratigraphic traps. Bull. Am. Ass. Petrol. Geol., 62, 932-66. & MUTrl, E. 1973. Turbidite facies and facies associations. In: Middleton, G.V. & Bouma, A.H. Turbidites and deep water sedimentation. Pacific Section, Soc. econ. Palaeo. Min., Short Course Notes, Pacific Section, Anaheim. -
-
-
-
-
1
1
9
-
5
8
.
WERNER, F. & WEa'ZEL, A. 1982. Interpretation of biogenic structures in oceanic sediments. Bull. Inst. Geol. Bassin d'Aquitaine., 31,275-88. WE'rZEL, A. 1982. Biogenic sedimentary structures in a modern upwelling area: the NW African continental margin. In: Suess, E. & Thiede, J. (eds), Coastal Upwelling: Its Sediment Record. Plenum Press, New York. 123-44. WEZEL, F.C. 1973. Diacronismo degli eventi geologicio-
646
D.A.V. Stow and D.J.W. Piper
ligo-miocenici nelle Maghrebidi. Revista Mineraria Siciliana, Anno XXIV. N 142-144, 219-32. WILSON, J.L. 1969. Microfacies and sedimentary struc-
tures in 'deeper water' lime mudstones. Soc. econ. Palaeo. Min. Spec. Pub., 14, 4-19.
D.A.V. STow, Grant Institute of Geology, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, Scotland. D.J.W. PIr'ER,Atlantic Geoscience Center, Geological Survey of Canada, Bedford Institute of Oceanography, PO Box 1006, Dartmouth, Nova Scotia, Canada B2Y 4A2.
Index Abyssal basins 437-50 hills, GLORIA survey 147-51 see also Basins Accretion, Tertiary, Guatemalan Transect 563 see also Compaction Accumulation rates biogenous 487-94 Cretaceous period 417-31 Santa Barbara 400, 485 sediments 483-94 terrigenous, Miocene 486-7 versus tectonic rates 431 Acoustic facies, interpretation of GLORIA 145-51 methods, scattering 72 stratigraphy 187 see also Echo-sounders Acritarchs, fossil 344 Actinolite, turbidite, Ulleung Basin 194 Active rift basins, Santa Barbara, Santa Cruz 409 Advection, in augmented transport 401 Aeolian dust sediment input 601,606 transport 17-19 turbidites 161-2 Aegean Sea see Mediterranean, East Africa, SW, shelf, organic carbon 534 see also Cape Verde Basin Aggregates, sediment, properties 48 Aggregation and transport in suspension 46-55 Airgun profiles 1 9 3 - 5 Albian black shales 545 see also Black shales strata, Atlantic-Tethyan 552-3 Algae, calcareous 199 Allochthonous sequences 512, 515 Alpine piston core see Core analysis Alum shales, Swedish, sedimentology 511-25 biostratigraphy 514 Cambrian deposits 512 depositional environment 523 discussion 521-4 lithofacies 517-21 stratigraphy 513, 515-17 Aluminium, in surface sediments 364, 369 Amino acids, proteins, recent sediments 589-90 Amphiboles 127, 178 Anacapa Current, California 381 Anaerobic conditions see Oxygen content Analytical methods see Methods, analytical Angola basin Cretaceous 550 Zaire sediments 95-112 Anisotropy, magnetic susceptibility see Magnetic susceptibility, anisotropy Anoxicity see Oxygen content Apatite, content of calcareous siltstone 333, 484 Aptian strata, Atlantic-Tethyan 552
Aragonite and green mud, Bahamian 475 Little Bahama bank 200-5 Argillaceous sediments, microstructure 579-90 see also names of components: Clays; Marls; Mudrock; Shales; Siltstones Ash falls, periodicity of volcanic events 18 Ash layers lower and upper Tephra 172, 175 Ulleung basin 187, 191 Asterosoma 338-9 Atlantic, central reconstruction, Lower Cretaceous 450 see also North Atlantic; South Atlantic Atomic absorption spectrophotometry 364 Attapulgite, calcareous sediments 210, 212 see also Illite Atterberg limits, sediments 565-9 Autochthonous sequences, Swedish Caledonides 512, 515-16 Bahamas, fine-grained carbonates 199-207 conclusions 207 global CO2 budget 207 methods 200-2 mineralogy 202--4 processes 204-7 setting 202 Bahamian Trough carbonates, source rock potential 469-80 basin sediments 471-4 carbon content, analysis 474-8 depositional setting 471-4 discussionconclusions 478-9 methods 470-1 Baltic Sea, Cambrian deposits 511-12 Barents Sea Group outcrop 343 Barium in slope sediments 364, 369 Barremian turbidites, organic carbon 547, 552 Basin(s) deep marine, translational plate boundary 481-94 marginal, schematic distribution 635 plains, definition 633 schematic distribution, sediments 6 3 4 - 7 and sills, sediment distribution 405-12 slopes, prodeltas, Norway 343-6 Basnaering formation, Norway 344-5, 356, 3 5 7 Bathymetric mapping (SEABEAM) 150, 156 Bathymetry-controlled trace fossils 595 B/itsl]ord formation, Norway 344 Bedforms, wavy contourites, Lake Superior 293 Benthic boundary layer 71 deep water indicators 337 food content, deep sea 600-6 Biodeformational structures 598 Biofiltration see Zooplankton Bioflocculation, in aggregation 46-52
647
648
Index
Biogenic detritus 77 indicators 117-18 see also Copper; Manganese mudrock, terminology 10 sediments 19 traces, use for bathymetric interpretation 598-606 see also Bioturbation turbidites, summary 617-20 Bioturbation and dissolved oxygen 532-3, 548-9 fine-grained sediments 595-607 biogenic traces 599, 600 conclusions 606-7 environment and ichnofauna 601-5 methods and study areas 595-8 model 598-601 North Atlantic contourites 249-50 Nova Scotia muds 313, 316 in rise canyon sediments 325-7 summary and discussion 638 and trace fossils 338-9 Birnessite (iron sulphide) in sapropels 501 Bituminous limestone 511,515-23 mudstone, Swedish Alum Shales 520 shales see Black shale Bivalves, Nuculid 337 Black laminated mudstone, Swedish Alum Shales 517-19 Black Sea, organic carbon 530, 534 Black shales 588-90 Albian 545 bacterial sulphate reduction 446 and claystones 437-50 model for deposition 527-40 Neogene see Sapropels organic carbon content 527-54 Pacific Mid-Cretaceous 551 sequences, bioturbation, lamination 540-8 South Atlantic 535, 540-9 Blake-Bahama basin, W. North Atlantic 437-50 'black shales' 444-6 'couplets' 446-7 depositional model 447-50 diagenesis 447 lithologies 438-40 marly chalks 444-6 regional setting 437-8 varve-type lamination 440-4 fo~xnation, lithology 437-50 sediments, diagenesis 448 Blake plateau, Bahamas, carbonate sediments 200 Bottom currents activity, lake 293-306 North Atlantic 245-6 Bottom-water oxygenation, inhibitory levels 530 see also Oxygen content Bouma sequence and episapropels 504 Bahamian trough 473 Blake-Bahama basin 441 E and F divisions 209
modified silt facies model 613 Obock Trough 218-19 partial, Halifax Formation 127 Salir Formation 456 sandy turbidites, standard 4 silt turbidites 613 turbidite 87,112,347-9 by X-radiography 328 Boundary-layer, benthic 71 Box cores, sampling method 295 see also Core analysis; DSDP sites Breccias, intraformational 350, 354 Burrowing see also Biogenic traces: Bioturbation indicator of oxidizing sediment, varve-type lamination 447 Calcarenites 227-39 Calcareous sediments New Zealand 331-41 Obock Trough 209-21,331 Calcilutites, Scaglia Rossa 223-39 bedding 228-30 bioturbation 234 burrowing 234 colour 232 composition 231-3 discussion 234-9 structure 234 texture 234 Calcirudite, Scaglia Rossa 231-2, 236 Calcite abundance in benthic foraminifers 488-9 calcareous sediments, NZ. 336-7 calcilutites, depositional model 231-6 classification 9 lithifaction and coalescence 522-3 magnesian, Bahama Bank 199, 202-7 Calcium analysis 364, 366, 369, 370 compensation depth 437, 445 see also Carbonate compensation California Coast Range 363-70 California Continental Borderland, fine-grained sediment transport 395-415 fine sediment transport, canyons 402-4 offshore sewage transport 402 regional setting 396-7 subsurface shelf turbidity patterns 400-2 surface water turbid plumes 397-400 transport scheme 412 turbidity, patterns, basins and sills 405-12 see also Santa Barbara Basin: Santa Paula Creek Cambrian stratigraphy, Alum shale 511-23 Canyons, submarine thalwegs 325-6, 328 transport and sedimentation 400, 402-4, 419 petit-Rhine canyon 121 US mid Atlantic region 319-29 Zaire canyon 95-8 Cape Verde Basin, Southern long-range sidescan sonar 145-51 acoustic facies interpretation 149-51 sonograph coverage 147-9 seismic and sediment facies 153-67
Index Cape Verde Basin, Southern (cont.) Aeolian turbidites 161 correlation between pattern and morphology 160-1 hemipelagic sediments 161 hiatuses 163-5 lithostratigraphy 156-9 processes 156-9 regional setting 155 Carbon-14 analysis 595 Nova Scotian muds 314 Carbon dioxide, global budget 205 Carbon rich strata, organic see Organic carbon Carbonates 9, 199-207, 209-21 Bahamian trough 469-80 muds, organic carbon 469-79 radiolarian nanofossils 437, 440-50 sedimentation and upwelling 488 see also Calcilutites: Calcite Carbonate compensation depth (CCD) 453,463, 540 see also Calcium compensation Casagrande Plasticity chart 569 Cascade-feeding process 171-2, 181 see also Mudflows Cascading, and particle concentration 395, 401 Chalk 10, 28 hemipelagic, Salir Formation 453-66 marly nanofossil 438-50 Challenger, Glomar
3-4
Chemical composition see named elements Cherts 227-39, 454, 484 Chlorite (clay) 19, 337, 365-6, 444, 456 see also Illite; Kaolinite; Montmorillonite Chloritoids 178 Chondrites 501 Chromatograms, gas 471 Chronology, turbidites, Santa Barbara Basin 391 Circulation deep water 245 and sediment movement 395-412 Classification, of fine-grained sediments 6-12 Clastic(s) bioclastic and terrigenous, Cretaceous 438-50 rocks, terminology 6-12 turbidites see Turbidites Clay 5, 19 clayballs 326-7 claystones and black shale 437-50 compacted, microfabric 582-5 freshly sedimenting 579-82 sediments, major microstructural changes 588 see also Chlorite; Illite; Kaolinite; Laterite; Montmorillonite; Mudrock Coacervates, in flocculation 580 Coarse-grained sediments, formation of fans 343 Coccolith(s) 206, 210, 331,338 and diatom zonation 483 ooze 3 Cocos oceanic plate 563-77 Cohesive material, erosion, transport, deposition 40-1 Colorimetry 364 Columbia river, Oregon-Washington 363-70
Columbus basin see Bahamian Trough Compaction artificial 583 characteristics, Quaternary sediments 563-71 and erosion 41-2 Conglomerates, Salir Formation 454--66 Ventura basin 421-30 Conglomeratic limestones, Cambrian 516 Continental margin, tectonically active, sedimentation 377-92 Continental shelf, effect of waves, currents 340 Continental slope Norway 343 Nova Scotia 311-16 sedimentation pattern and sea-level 325 US mid-Atlantic 319-20 Continental source area, hemipelagic sediment formation 363 Contour currents 293-306 Nova Scotian mud turbidites 317 Contourite(s) 122, 245-54, 257-72, 293-306 Blake-Bahama Formation 441,443 and chalks 463 facies, comparison with turbidite and hemipelagite 291,316-17 comparison with turbidites 349 criteria 620-4, 631 destruction by bioturbation 290 history 5 lamination 290 mottled silts, facies 282-4 muds, homogeneous, facies 284-7 muddy 5, 12 North Atlantic, comparison with sandy contourites 253, 275-91 North Atlantic, generalized plot 252 sandy 5, 12 North Atlantic, comparison with muddy contourites 253, 275-91 summary 620-4 muddy 621-2 'sequence' 623-4 silty-sandy 622-3 and turbidites, criteria 293 turbidite, hemipelagite, discussion 31 6-17 see also Turbidites Convoluted bedding 385-7 Copper atomic absorption spectrometry 117-18 pelitic sediments 115-25 in surface sediments 364, 369 Core analysis sites Blake-Bahama Formation 438 Cape Verde Basin 153 Central Pacific Ocean 597 Eastern Mediterranean 499 Faro Drift, Gulf of Cadiz 276 Great Bahama Bank 470 Guatemalan Transect 564 Lake Superior 294 Little Bahama Bank 200-1 North-west Africa, continental margin 596 Nova Scotia continental margin 258 Nova Scotian Slope 312
649
650
Index
Core analysis sites (cont.) Oregon-Washington Slope 364 Santa Barbara Basin 378 Sulu Sea, Borneo 597 Ulleung Basin, Japan 187 United States mid Atlantic continental rise 321 West Hellenic Arc margin 170 Zaire river deep sea fan 99 Cores, box, gravity 595 Coriolis effects, deep-water circulation 317, 409-10, 443 Coulometrics, TOC system 471,474-9 Coulter Counter, in deep sea suspensions 73, 75 'Couplets', radiolarian carbonates 437, 440-4 see also Varve-type lamination Cretaceous, Lower, Central Atlantic reconstruction 450 middle, climate 550 Cucullea, calcareous sediments 331,335 Current transport see Transport Currents bottom 245-7, 252-4 lake 293-306 pelagic depositional 275 turbidity 275, 317 Cyclicity in sediments, test for 446 Debris-flows 565 see also Mudflow Debrites 247-8 Decompaction analysis 571-3 Deep Sea Drilling Project see DSDP Deep-water circulation 409-10 fine-grained sediments, historical outline 3-6 terminology 12 morphology, terminology 11 Deformation see Slides; Slumps; Soft sediments Delta and pro-delta deposits, N. Norway 343,359-60 Density discontinuities, and thermoclines 400-1 Density profiles, wet bulk 565 Deposition fine-grained sediments 59-61 field measurements 61 laboratory measurements 59-61 theory 59-61 related to erosion and transport 36 Depositional sorting, in laminations 347, 349 Destabilization of sediment by biological material 45-6 Detritus biogenic 77 terrigenous 77 Diagenesis Blake-Bahama sediments 441,448-50 clays and muds 582 deep sea sediments 528 summary 639-40 Diapirs 112, 121 Diatom(s) benthic effect on sedimentation stabilization 44-5 productivity, Atlantic Miocene 554 concentration patterns 399 and volcanism 482
Diatomaceous muds, pelagic 563 Diatomaceous ooze 10 Diatomite 10, 28, 427, 484 Dictyonema shales 523 Discontinuous lamination, 490-4 Discordance, erosional-depositional 343, 355 Disorganized turbidites, summary 620 Distortion, in shear orientation of clay 585-7 Dolerite 454, 517 Dolomite 336-7, 493 Domains, microstructural studies, definition 581 Down-core analysis 303-6 Downslope mass movement 574-7 see also Mass flows; Slides; Slumps Drift card trajectories, surface currents, Santa Barbara 406 DSDP cores Atlantic sites 527, 535-9, 544-54, 551 Blake-Bahama basin 437-50 Japan basin 185 North Atlantic 246-8 Pacific sites 527, 542-3, 548-54 slides, criteria 343 West Hellenic Arc 170-7 East Cape, North Island, New Zealand 331-41 Echo-sounders, profiles 150, 156, 258 see also seismic profiles Elements, in surface sediments 364, 369 see also names of individual elements Elzone, particle analyser, electronic 295-306 en echelon tectonic zones 453 Enteromorpha, effect on sedimentation stabilization 44-5 Entrapment, in deposition patterns 399 Epidote 178, 194 Episapropels 502-9 Erosion-deposition model, lake sediment 301-3 Erosion, fine-grained sediments 36-46 biological influences 44-6 and compaction 41-2 critical condition 36-41 rate, physico-chemical parameters 42-4 equation 42-4 related to shear stress 42-4 Erosion, stoss-side 348-9 Erosional-depositional discordant surfaces, Norway 343, 351,355 Exuma Sound see Bahamian Trough 'Fabric rolls', in sheared clay 587 Facies models: deep water fine-grained sediments 611-40 contourites 620-4 discussion--conclusion 637-40 environment and facies distribution 629-37 fine-grained turbidites 612-20 pelagites and hemipelagites 624-9 processes--facies 611-40 Facies analysis, turbidite 106 Faecal pellets, zooplankton 395, 399-400, 412, 431 ocean deposits 443-4 sinking rates 528 Fan, deep sea, schematic distribution 634
651
Index Fan deposits clastic submarine 343-60, 453-66 Laurentian 85-91 Lower Nfieringselva member, Norway 359 Zaire 95-112 Faro Drift, Gulf of Cadiz contourites 275-91 comparison, turbidites, hemipelagites 291 discussion 290 oceanography and bathymetry 275-8 sediment facies 278-87 vertical sequence 287-9 Fault, palaeotransform 344 Fecal pellets see Faecal pellets Feldspar 231,336, 454 Ferrohastingsite 194 Fine-grained sediments, depositional processes deposition rates, biological productivity 22, 28 bioturbation 23, 28 dilution 23 oceanic dispersal systems 22 pelletization 22 sea-level position 22 tectonics 21 depositional sites, oceanic, deep sea and basin floors 27 shelves 26 slopes 27 marine, accumulation rate 61 aggregation 46-54 cohesion 35-6 critical deposition stress 61 distribution with depth 57 erosion 36-44, 62-3 historical outline 3-6 nepheloid layers 57-8 particle settling velocity 62 particle size distribution 49, 50-4, 62 radionuclides accumulation 61 settling velocity 54-6 shear stress 60 stabilization and destabilization 44-6 stratification 57 terminology 12 transport 46, 71,412 vertical flux 56-7 origins, bulk properties 20 cycles 20, 21 discharges, human influence 21 transfer to deep water, debris flows 25 mass movement 26 nepheloids 23 plumes 23 resuspension 24 turbidity currents 25 Fissility, in shales and muds 582-5, 587-90, 637-9 Flaser bedding, as current indicator 459-63 Floc-deposited sediments see Turbidites Floc size distributions 52 Flocculation clay 580-90 coacervates 580 Floods amplification of mass movements 397, 400, 408-9
suspended sediment 380-3, 390 Flow initiation, and long distance transport 612 Fluid shear 47-8, 60-2 Fluidal flows, Atterberg limits 565 Fluorite 200 Fluorometric titration 98 Fluvial discharge, terrigenous matter 528, 549 Fluvio-deltaic deposits Lower Nfieringselva model 359 mineral signatures 364, 372-3 and nepheloid layers 343, 349 post glacial varves 379 Flux, vertical and horizontal 56-7 Flysch deposits 136-42, 339-40 Foraminifera 260, 331 benthic 453 Foraminiferal deposits 199, 209-10, 214, 219 in calcilutite 231,237 in contourites 282, 283, 287 evidence of sea depth 421,425 Forward light scattering, suspended matter 72-3 Fossil communities, trace 595 Freeze drying, surface sediment samples 364 Garnet 178, 194 Gas chromatograms 471 Glaciation, Pleistocene-Holocene 115, 119 Glauconitic sandstones 511,520--3, 547 Glomar Challenger drill ship 1968-1983 5 GLORIA (long range sidescan sonar system) 145-51,319-24 Goldenville Formation, Nova Scotia shales 127--42 Grading 9-11 indicator of turbidity current flow 349 positive and negative 279, 287 see also Depositional sorting; Grain size; Stratification Grain size analysis, factor analysis, quartz sand 325-6 frequency curves, N. Atlantic contourite 281 pelitic sediments 115-19 Santa Barbara basin 378-9 summary 639 X-ray diffraction 364-5 see also Coulter Counter discussion 10, 11 distribution curves, Faro Drift 281 N. Atlantic contourites 251 GRAPE, wet bulk density profiles 565, 571 Graphoglyptid traces 603, 606 see also Nereites
Gravity cores 295 Gravity faulting, submarine slopes 384, 387 Grey mudstone, Swedish Alum Shales 520 Guatemalan Transect, Quaternary sediments 563-77 data analysis 570-3 decompaction values 573-5 discussion and summary 573-7 location DSDP sites 564 methods 563-8 plasticity and compaction 569-70 Gulf of Cadiz see Faro Drift Gullies, on basin slope 343, 356, 359-60 see also Canyons
652 Gully system, dendritic
Index 319, 322
Halifax Formation, Nova Scotia shales 127--42 alga 473 Hallam's model, Liassic concretions 522 Halloysite 336-7 Harmattan winds 595 'HEBBLE' definition 25 method, measurement ofs.p.m. 79 Hellenic Arc margin, west 169-96 composition 175-9 distribution 172-5 general lithology 172-5 morphology 169-70 processes 179-80 stratigraphy 170-2 structure 169-70 texture 175-9 thickness 179 Hemipelagic chalks, fan sequence, Miocene SW Turkey 453-66 depositional process, chalks 459-63 geological setting and sedimentary facies 453-9 lateral and vertical distribution, chalks 463-6 Salir Formation 453 Hemipelagic cores, in palaeoclimate studies 383 Hemipelagic sediments basin model 377-92 Cape Verde Basin 153-66 dispersal 370-2 formation on continental margin 363 Miocene 527-54 Monterey Formation, California 481-95 from multiple sources 373 recent, Bahamas 200 reworked 209 Santa Barbara Basin 377-9 Ulleung basin 191 Wiirmian grey muds 121 Hemipelagic settling definition 311 turbidity currents 431 and turbiditic muds 453 Hemipelagites 12, 20, 115-18, 121 comparison with contourite 291,316-17 criteria 624-9 cycles 629 gradation with contourites 254 muddy pelagic ooze 627 Obock Trough 209 and pelagites, summary 624-9 turbidite, contourite discussion 316-17 see also Pelagites Heyes, North and South Canyons see United States mid-Atlantic region Hiatuses, Southern Cape Verde Basin 163 Hjulstr6m's curve 59-60, 289 Holocene features California Borderland 380 Nova Scotian muds 311 Sulu Sea 605 Horizontal flux, particles 56-7 Hornblende 178, 194, 365-8 Halimeda,
Humic acids 589-90 Hydrocarbons, sources of 469-79 see also Organic carbon Hydrographic features see Basins and sills Hydrography Angola Basin 9 6 - 7 ocean currents 155, 165 see also Circulation; Turbidity currents Ice-rafting 295 Ichnofacies (ichnocoenoses) trace fossil communities 595, 600, 603-7 Ichnofauna, changes 601-7 Illite in calcilutites, Scaglia Rossa 23 California Coast Range 366 definition 19 East Cape, New Zealand 336-7 Guatemalan Transect 563, 574 Hellenic Arc margin 176-8 NW Mediterranean 116 Obock Trough 210-12 and plasticity index 574 ratios, other clay minerals 116-17 Zaire 95 see also Chlorite; Kaolinite; Montmorillonite; Smectite Imbricate structure 587 Imbrication, tertiary 563 Inter canyon slopes see Slopes Intermediate water masses see Oxygen content, oxygen-minimum Internal waves, and resuspension of sediments 395, 397, 400~ Intra-slide beds, Norway 351 Ionian Basin, characteristic clay minerals 176-9 Isopleths, sand and silt, from box core sampling 378-83 Iron 122, 364-6, 369-70 measurement, atomic absorption spectrometry 117-18 pelitic sediments 115 red/green shales 590 Japan
see
Ulleung Basin
Kaolinite artificial compaction 583 in calcilutites, Scaglia Rossa 231-2 definition 19 East Cape, New Zealand 336-7 Guatemalan Transect 563, 574 Hellenic Arc margin 176-8 Obock Trough 212 ratios, other clay minerals 116 Zaire 95 Kernels, coagulation aggregation 46-51 breakup 51-4 Kerogens in Bahamian carbonates 469 type II, Tissot 541 see also Hydrocarbons: Organic carbon Keweenaw Peninsula see Lake Superior
Index Kongsfjord Formation 344-60 Kithira basin 173-9 Klamath mountains, Oregon 363-70 Korea see Ulleung Basin Laconia system, West Hellenic Arc basin 169-80 Lacustrine contourites, Lake Superior 620 Lake Superior, contourite texture 293-307 core-top analysis 295-306 down-core analysis 303-6 methods 295 Sonar records 298 Laminated black marlstone, North Atlantic 540-54 Lamination, Ingrams system 9 bioturbated 282, 408-9 see also Varve-type lamination Laminites 9 terminology 12 Lamont measurement, forward light scatter 73, 76-7 LANDSAT see Satellite imagery Lateral fining, contourites 293 Laterite (clay) 19 Laurentian fan, turbidite sampling 85-91 Laurentian Ice Sheet, retreat 295, 303 Lavas, pillow 517 Lead- 210 measurement 327-8 profiles 293 Leco Corp WR-12 carbon determinator 471 Lensoid deformations, mudflows 385 Light-scattering, forward 73 Limestone bituminous 511, 515-23 nanofossil 438-50 Linear programming, conversion of XRD results 363-71 Lithifaction, Great Stinkstone, Sweden 522 Lithostatic load, Bishop 569 Lithostratigraphy see Stratigraphy Long-range sidescan sonar system see GLORIA Lophocterium and N. Atlantic contourites 279 Lutites Angola Basin 112 definition 9 suspension cascading, fine laminations 441,443 see also Clay; Mudrock Lyngbya, effect on sedimentation stabilization 44-5 Magnesian calcite 199, 202-7 Magnesite, West Hellenic Arc 178 Magnesium, Oregon-Washington sediments
364,
369-70
Magnetic signature 161,164 Magnetic susceptibility, anisotropy (AMS) Cape Verde Basin 161, 163-4 North Atlantic contourites 250 Nova Scotia continental rise contourites core samples, results 264-71 discussion 271-2 Holocene sedimentation 259-61 methods 261-4 Magnetite 261
257-73
653
Manganese, measurement, atomic absorption spectrometry 117-18, 122 in surface sediments 364, 369 Marine sediments, chemical and mineral relationships 371 Markov sequence analysis, non-turbidite association 316-17 turbidite association 315 Marl, definition 9, l0 Marlstone laminated black, N. Atlantic 540-54 Scaglia Rossa 225-7 Mass-flow processes, in seismically active basins 185, 188 Mass gravity transport, definition 12 Mass movements fine-grained sediments 395-412, 417-31 processes 377 Ulleung basin 195 Santa Barbara Basin 384-7 scars, volume calculations 385, 391-2 tectonically active margin 384-7 Mass wasting, slope, events leading to 360 Massive bedding, vague lamination 490-4 Matapan Trench 169-80 Mediterranean, Eastern 497-508 Mediterranean, North-west margin, during late Quaternary 115-25 basin plain 121-2 composition of < 40 #m fraction 115 continental rise 121 continental slope 119-21 dynamic sequence 122 geochemical criteria 117 granulometry of < 40 ~m fraction 115-16 mineralogical criteria 116-17 palaeoclimatic sequence 122 ungraded sequences 122-4 Mediterranean outflow, general effects 276 Mediterranean, Western 169-80 Mesozoic-Cenozoic carbon-rich sediment, model for deposition 527 bottom-water oxygenation 530-2 discussion 550-4 DSDP analysis 527-8, 535-9 iithology, black shales 540-9 organic matter, preservation 528-9 productivity indicators 532-3 sedimentation rate 529-30 Metamorphism, extent in continuous coastal exposure 344 Methodology, principal text list 8 Methods, analytical atomic absorption spectrophotometry 364 calorimetry 364 coulometrics, Total Organic Carbon 471,474 Coulter Counter 73, 75 electronic particle analysis (Elzone) 295-306 freeze drying 364 GRAPE, wet bulk density 565, 571 Lamont measurement, forward light 72 lead-210 measurement 293, 327 Leco Corp WR-12 carbon determinator 471 linear programming 363
654
Index
Methods, analytical (cont.) Markov sequence 315-17 mineral abundance data 365-70 pipette analysis 295 Q mode factor analysis 363 radiocarbon dating (C14) 325-6, 595 Rock Eval, GEOCHEM 471,475 scanning electron microscopy (SEM) 204, 206 sediment trap measurements 75-7 thermovaporization gas chromatography 471,475 transmissometry 405, 407 viscometer measurement 40 wet bulk density profiles 565, 571 X-ray diffractometry 202, 335, 364 Methods, statistical end member composition 364-5 linear programming 364-5 see also specific headings Mica, enrichment 440 Micrite, in chalk beds 456 Microfabric of muds and shales see Muds; Shales Middle America Trench analysis 570-1,576 sediments 563 Mineral abundance data, from XRD results by linear programming 363, 369-71 Mineral composition, East Cape sediments, New Zealand 336-7 Miocene-Pleistocene, sediment accumulation, Ventura Basin 419-20 Miocene sequence, Santa Barbara 484 Mississippi Delta, deposits 581 Monterey Formation, California 481-94 Montmorillonite (clay) 19, 336-7, 456, 563, 574 Mud(s) beds, under Newtonian fluid flows, behaviour 35-69 calcareous 247-54 and fissility summary 637-8 and shales, fabric 579-93 major structural changes, clay sediments 588 microfabric, compacted clay 582-5 sedimenting clay 579-82 shear, effects of 585-7 synthesis and conclusion 587-90 turbidites, ideal facies model 615-17 Mudflow, deposition 384-92 Mudrocks and anoxic conditions 523 biogenic 9, 10 black, Cambrian 516, 517-20 calcareous sandy 331-41 classification 9 parallel-laminated 343-8 scanning electron microscopy (SEM) 335 structureless 343-9 see also Clay; Silt; Siltstone; Slate
Nepheloid flow 441 layer, benthic 124 circulation 247, 252, 340-1 deep ocean 57-8 Pacific and Atlantic 77-9 plumes 395, 400-12 Nephelometer, Lamont, SPM sensing 72-3, 76-9 Nereites 339 ichnofacies 595, 603, 606-7 New Zealand, calcareous sediments, upper slope to shelf 331-41 debris flow deposits 339 flora and fauna 337-8 interpretation and discussion 339-40 lithological descriptions 332-5 palaeocurrents 339 stratigraphy and facies 331-2 trace fossils 338-9 X-ray diffraction mineralogy 335-7 Normal bottom currents 12 North Atlantic contourites (overview) 245-54 bioturbation 249 bottom circulation 245-6 continental rise contourites 248 facies 254 grain size distribution 251 processes 252--4 sediment drifts 246-8 sedimentary characteristics 248-9 see also Faro Drift; Nova Scotia Slope Eastern and Northern DSDP sites 538-9 sedimentation rate, cretaceous 551-3 Western, DSDP sites 536--7 North-west Africa margin 595-6, 600-7 Norway, sediment transport deposition, late Cambrian 343-60 basin slopes and prodeltas 343 discussion 360 erosional-depositional discordant surfaces 355-6 LNM depositional model 359 Lower N/ieringselva Member (LNM) Finnmark, facies 344-50 soft-sediment deformation 350-2 synthesis, palaeocurrent data 356 Nova Scotia, continental rise, contourite, anisotropy of magnetic susceptibility 257-73 see also Magnetic susceptibility, anisotropy Halifax Formation, shales and slates 127-42 slope muds 311-17 discussion 31 6-17 physiography 311-12 sediments 312-16 stratigraphy 312 Nucleation kernels, in concretions 522 Nuculid bivalves, calcareous siltstones 337
Nfieringselva Member, Lower, (LNM) Norway 343-60 Namibian Shelf, organic carbon 534-5 Nanoplankton, calcareous 338 National Oceanic and Atmospheric Administration 319-20
Obock Trough (Gulf of Aden) late Quaternary calcareous clayey-silty muds 209-22 comparison of sequences 219-22 general characteristics 209-14 sequence interpretation 219 turbidites, characteristics 214-19
Index Obock Trough (cont.) frequency and origin 221 Oil, Alum Shale Formation 511 see also Hydrocarbons; Organic matter; Total organic carbon Oil seeps 417, 430 Olistrosome, rose diagrams 339 Olivine 194 Ooze, classification 10 pelagic 12 periplatform 199-207, 239 Optical methods, deep sea suspensions 73 Ophiolite, derived sediments 453 Oregon-Washington continental slope, fine-grained sediments 363-75 continental source area contributions 363-4 data and results 365-70 discussion--summary 370-4 sampling and analytical methods 364-5 Organic carbon vs carbonate carbon, Aegean seas 500 NW Africa sediments 598 -rich strata, Atlantic Ocean distribution 540-54 model for deposition 527-40 Sulu Sea sediments 597-8 Organic matter carbonates, generated hydrocarbons 469-71 and detrital minerals, Monterey formation 489-94 marine sediments, classification 589-90 terrigenous 528 'Orsten' see Bituminous limestone Orthoclase 231 Overburden pressure and compaction trends 569-70 see also Atterberg limits Oxygen content and decay 405-8 deep sea, bottom water 527-54 general criteria 533 and laminations 489-94 mud sediments 377-92 oxygen-minimum zones, black shale deposition 530, 549-51 varve-type lamination 444-6 and organic carbon 521-3 and pelagic settling 459 reducing environment 588 and sapropels 497-509 Oxygen- 18 determinations 595 Oxygenation of carbonate rocks 469-78 Pacific North-west, US continental slope see Oregon-Washington Pacific oceanographic circulation, Eastern 399 Palaeoclimate and anoxic conditions 497, 505 Palaeocurrents, data, LNM, Norway 356 East Cape, New Zealand 339 Palaeozoic, Lower, stratigraphy 513, 516, 519 Palygorskite 444 Parallel-laminated mud, Ulleung Basin 191 Particle analyser, electronic 295-306 Particle analysis, statistical treatment 295-305 Particles, aggregation and size 46-51 Particulate matter see Suspended particulate matter
655
Pelagic carbonates, Blake-Bahama Basin 437-50 clay, definition 12 deposition, Blake-Bahama Basin 437-50 diatomaceous muds 563, 569 ooze 12, 626 sedimentation, 'black shale' 527-54 history 3 Pelagites criteria 624-5 cumulative frequency curves 115-17 gradation with contourites 254 and organic-rich sediments, Western North Atlantic 437 Scaglia Rossa limestones 237 see also Calcilutites; Hemipelagite Pelites definition 9 NW Mediterranean margin, late Quaternary 115-25 palaeoclimatic sequences 122 West Mediterranean Basin 121-2 see also Mudrocks; Pelagites Pelleting, in sedimentation 380 Peloponnesus see Hellenic Arc Periplatform oozes, sources of hydrocarbons 202-7, 471-9 Peruvian margin, organic carbon 534 Petroleum production, diatomaceous deposits 481 see also Hydrocarbons; Organic carbon Phosphorus accumulation rate, organic carbon 534 Phyllites, black 517, 523 Piper's sequence, turbidites 109, 316, 613 biogenic turbidites 618 Pipette analysis, grain size 295 Piston core analysis see Core analysis Plagioclase 231,336-7, 365-8 Plane lamination, Obock Trough 214 Planktic analysis, East Cape sediments, New Zealand 337 Plankton productivity 444 Planolites 501 Plasticity Casagrande chart 569 indices, Guatemalan Transect 574-7 Plate tectonics, history and discussion 5 Plumes, surface water 396-400 Pohutu formation, calcareous mudstones 331-2 Ponding, in synsedimentary reworking of deposits 122-4 Porcelanite 484 Pore size, volume, and plasticity 569 Post glacial clays, lake 293-306 Potassium 364, 366, 369-70 Primary production, and benthic food content 600--1 surface waters 528 Pro-delta, and delta deposits 343, 359 lobes 119-23 Providence Channel sediments, North-west 199 Pteropods, source ofaragonite 199, 210 Pycnoclines, and sediment movement 400 Pyrite, black laminated mudstone, Swedish Alum 517, 519-23
656 Pyrolysate index in TOC Pyroxene 194
Index 475-9
Q-mode factor analysis 363 Quartz East Cape, New Zealand 336-7 laminae 517, 520-3 Oregon continental slope 365-6 sands, US mid-Atlantic region 327 Quartzose silt 440, 442 Quaternary sedimentation, NW Mediterranean 115-24 Question Set Approach, Potter's 6, 7 Radioactive waste disposal, sea bed, preliminary study 145-51 Radiocarbon dating indicator of sediment accumulation 325-6 sapropels 498-509 Radiolarian carbonates, nanofossil 437, 440-50 clay, black and green 548 ooze 12 Rainfall, and varve lamination 380 Reducing sediments see Sapropels Resedimentation processes 12 Ripple-markings 343, 348-50, 3 5 6 - 7 Ripples asymmetric 380 interference 349 River discharge 17, 18, 21 Rivers, sediment load 95-112 see also Fluvio-deltaic Rock-Eval analysis, GEOCHEM 471,475 Rose diagrams 339 Salinity, dependence of erosion rate 45 Salir Formation, SW Turkey 453-5 Salt alum 511-23 Sampling methods anisotropy, magnetic susceptibility 261 box cores 295 gravity cores 295 piston cores see Core analysis seismic reflection profiles see Seismic profiles Shipek grab sampler 200 sidescan sonar image, GLORIA 319-24 San Andreas fault 482 Sand(s) content and compaction 4I-2 fractions, Faro Drift 283 quartz 326-7 shelf, as tracers 327-9 -silt-clay, terminology 9 spillover 341 waves, Hueneme Sill 405 Sandstones 129-42 calcareous, muddy 336-7 siltstones, Swedish Alum Shales 520 wave rippled fine-grained 343-50 Santa Barbara Basin, California Continental Borderland 377-94 current transport, fine-grained sediment 380-3 hemipelagic sedimentation 377-9
mass movement 384-7 setting and previous studies 379-80 stratigraphy 388-92 hemipelagic deposits 481-95 geology 481-3 layering 489-93 lithologic sequences 492-3 palaeogeography 482 sediment composition 483-9 stratigraphy 483 synthesis 493-4 Santa Cruz Basin Fault 482 Santa Monica Basin 396 Santa Paula Creek, California fan-lobes and fine-grained sediments 417-33 conclusions 430-1 sediments, zones 1-4, 421-30 structural setting, stratigraphy, Ventura Basin 419-21 tectonic and sedimentary controls 417-19 Santa Ynez Fault 481-2 Sapropels 172-8 marine sediments, Eastern Mediterranean 497-508 definition 497 episapropels 502 lithofacies 498-501 methods 498 organic rich deposits 501-4 as palaeoceanographic indicators 504-8 petrological classification 508 X-radiographs 505, 507 Mediterranean 531 Black Sea 549 Eastern N. Atlantic 548 Sarl 9, 10 Satellite imagery (LANDSAT), turbid plumes 397-400 Scaglia Bianca, central Italy 225 Variegata 225 Scaglia Rossa limestones, calcilutites 223-41 broader implication 239 characteristics 228 bedding 228-30 bioturbation and burrowing 234 colour 232-3 composition 231-2 structure and texture 234 depositional model 237-9 facies 227-8 interpretation 234-6 stratigraphy 223-7 turbidite and pelagite characteristics 237 Scattering measurements, forward suspended matter 72-3 Schuppen structure see Imbricate Sea-level changes and tectonic effects 359 effect on sedimentation patterns 325 Sea MARC 1 (mid range sidescan sonar system) 319 SEABEAM, bathymetric mapping 150, 156 SEABED cruise see Seismic survey Seabed Working Group, radioactive waste disposal 145
Index Sediment(s) accumulation, by lead-210 dating 327-8 measurement, by radiocarbon dating 325-6 budgets 371-3 draping 351,354-5 drifts, North Atlantic 246-54 mass gravity processes 169 slides, slumps see Slides; Slumps transport, down-canyon 329 implications of plasticity 574 processes, East Cape, New Zealand 340 US mid-Atlantic, canyons 319 mass wasting 329 trap measurements, Western Atlantic basins 75-7 /water interface, oxygen levels 533 see also Contourites; Fine-grained sediments; Turbidites Sedimentary sequences, reconstruction 115-24 structures, graphic symbols 11 Sedimentation rates Alum shales 517-23 and accumulation rate 532, 548-9 carbon-14 dating 252 deep sea 598-607 equation 372 glacial and inter-glacial 474, 477 and mass wasting 360 Mesozoic-Cenozoic strata 529-30 organic carbon content, North Atlantic 551 Swedish Alum Shales 517, 518 Seismic activity, sediments 417, 430 airgun profiles 170 reflection profiles 175, 185, 193, 209, 212, 319-20 in bathymetric maps 293, 294 Faro Drift 277 Ulleung basin 190 and sediment coring survey 153-66 SEM, calcareous sediments 332-5, 337 Serpentinite 454, 456 Settling velocity, particles 55-6 Sewage plume 402, 403 Shales Alum see Alum shales black see Black shales definition, Potter Question Set 7 Halifax Formation 129-42 and pressure 582 successions 6-7 terminology 8-10 see also Argillite; Clay; Claystone; Silt; Siltstone; Slate Shear-compressional stress 563 'Shear sorting' in laminated turbidites 587 Shear stress, related to erosion rate 42-4 Shearing forces, clays, effect on microstructure 585-7 Shelf sediments, dispersal by turbidity flows 327 Shipek grab sampler 200 Sidescan Sonar images 293-303 GLORIA 319-24 mid-range (Sea MARC 1) 319 Silica (silicon) 231-39, 364, 369 rich sediment, and high productivity 481-94
657
see also Chlorite; Illite; Pyrite; Quartz Siliceous turbidites 618-19 Siliciclastic sediments 9 Silicoflagellates 97 Sillimanite 194 Silt, classification 20 -sand-clay, terminology 9 turbidites, summary 613-15 Siltite see Siltstone Siltstone 129-42 calcareous sandy 331-41 current rippled 343-9 SEM 333 Sisquoc Formation, Miocene layering 490 Slide(s) diagnostic features in DSDP cores 343 folds, overturning, Lower N~ieringselva Member 351 sheets 352, 353 from soft sediment deformation 344-60 Slope(s) aprons 343 definition 631 devoid of canyons, gullies 377 failure, triggering mechanisms 378 mass wasting, events leading to 360 sediment, redeposition 528-9 -to-shelf sequences see New Zealand, calcareous sediments Slump scars 174, 187, 191-3 Slump packets 331 Slump-slides, Lower Nfieringselva Member 351-2 Ulleung basin 191--4 Slumping, in canyon erosion 328 sea floor 112 Smarl 9, 10 Smectite 116-17, 122, 176-8, 210, 212, 231,365, 444 see also Illite Sodium 364, 369 Soft-sediment deformation breccias, intraformational 350, 352 fluidization structures, Norway 350, 352 slides 343, 350, 358-60 synsedimentary faults 351,355 Sohm Abyssal Plain, turbidite sampling 85-91 Soil Classification, Unified 569 Sonar, images, sidescan 319-24 studies, sidescan 258 Sonograph, long range sidescan (GLORIA) 145-51 South Atlantic, Western, DSDP sites 535 South East Atlantic Ocean, surface current patterns 97 Spectrophotometry, atomic absorption 364 Spinel 178 Stabilization of sediment by biological material 44-6 Stagnation, and sapropels 497-509 Statistical analysis, particle analysis 295-305 Statistical methods, computer contouring, SYMAP 379-80 Staurolite 194 Stinkstone see Bituminous limestone Stokes'-deposited sediments 83-91 Storms, effect of carbonate transport 203
658
Index
Storms (cont.) frequency and intensity, Lake Superior evidence 305-6 Stow sequence, turbidites 619 Stratification terminology 10-11 Stratigraphic correlation, Southern Cape Verde 158 Stratigraphy Cape Verde Basin southern 156-9 Scaglia Rossa limestones 223-7 West Hellenic Arc basin 170-2 Strontium 364, 366, 3 6 9 - 7 0 Subarkoses 346 Sugars, in marine sediments 589 Sulphate reduction, organic carbon sediments 530, 532 Sulu Sea, bioturbation features 597-603, 604-5 Surf, internal 400-1 Surface currents, Santa Barbara Basin 381 Suspended particulate matter measurement, forward light scattering 71, 72 gravimetric 73 particle counting 75 particle size, spectra 77 variations, regional and temporal 72 Suspended sediment, continental slope, annual input 363 Swedish Alum Shales, sedimentology see Alum shales Swedish Caledonides succession 515 Swedish Deep Sea Expedition 71 Symbols, graphic 11 Systemic sedimentology, history 3 Tectonic(s) activity, Santa Barbara Basin 377-92 controls, and volcanic centres 482 factors, carbon-rich strata, Pacific 548 Pacific, Miocene 554 stresses, rift basin 409 transform-margin 417 zones, en echelon 453 Telegraph cables, corrosion studies 276 Tephra upper and lower, ash layers 172-5 volcanic 331-8 Terebellina, New Zealand and California 339 Terminology 6-12 descriptive 6-10 interpretative 10-12 Terrestrial organic matter, periodicity 446-7 sedimentation, Miocene 527-54 Terrigenous detritus 77 nepheloid settling 549 sedimentation rates 390-2 silt as indicator 381 Tethyan sequence, organic carbon 540, 550 Tethys sea, and Atlantic 444 Textural analysis see Grain size Thalweg, canyon 325-6, 328 Thermovaporization gas chromatograms 471,475-8 Titanium, California Borderland 364, 366, 369-70 TOC (total organic carbon) carbonate muds 469-79 profile and Plasticity Index, Guatemalan
Transect 576 Tongue of the Ocean see Bahamian Trough Trace fossils, calcareous slope sediments 338-9 Trace metals, Alum shales 511, 515 Transform motion see Tectonics Transition probability, in chalk/turbidite deposition 463-6 Transmissometer casts, deep water and surface water 405, 407 Transport California Coast Range 363-70 current velocity 288-91,319 down slope continental rise 377 long distance, and flow initiation 612 processes, fine-grained sediments 395-412 related to erosion 35-63 aggregate strengths and sizes 48-51 and aggregation 46-8 biological influences 44 breakup kernels 51 compaction 41 conclusion 61-3 deposition 59-61 erosion 36 rate 42 Newtonian fluid flows 35 settling velocity 55-6 vertical and horizontal flux 56-8 Trilobites 511,522 Trough sediments, Lake Superior 295, 301-6 Turbidite(s) analysis, Angola Basin 98-112 NW Mediterranean 116-24 Bouma standard sequence 4, 613 chronology, Santa Barbara Basin 390-91 and contourites, distinguishing criteria 293, 316-17 criteria 612-17 deposition, and aeolian dust 603 model 83, 84 facies 12 comparison with contourite 316-17 floc-deposited 83-91 global distribution 630 grain-size characteristics 83-92 analytical methods 87 discussion 90-1 material 85 muddy, sedimentary core structures 86, 87-8, 90-1 results 87 settling model 84-5 ichnocoenoses 595-607 interbedded 293 muddy, carbon rich 527-54, 563 see also Black shales pelagite, end-members 223, 237 revolution, history 4-5 and sapropels 504-9 SE Atlantic Ocean, climatic zonation 100 sedimentation, principal areas of world 630 sequences, Nova Scotian muds 312-16 shear sorting 83-91 summary, biogenic, 617-20
Index Turbidite(s) (cont.) disorganized 620 mud 615-17 silt 613-15 Stow sequence 619 type sequence 218, 220 X-radiography 98, 102, 104-5, 107-8 see also Contourites Turbidity currents cause of graded bedding 4 deposition processes, California Borderland 405-12 Obock Trough 214-21 Scaglia Rossa 234-9 effects of diapirs, Cape Verde 163 frequency and chalk horizons 453 transport, Santa Barbara Basin 377-9 see also Mass flow Turbidity patterns, basins and sills 405-12 subsurface shelf 400-2 Turkey, SW, Salir Formation 453-66 'Type sequence' see Turbidite sequence Tyvofjell Formation, Norway 344 Ulleung Basin, Sea of Japan, turbidites and associated deposits 185-94 acoustic stratigraphy 187-8 geological setting 185-6 hemipelagic settlements 191-4 physiography 186-7 processes 195-6 provenance 194-5 turbidites 188-91 Umbro-Marchean Apennines see Scaglia Rossa limestones Ungraded deposits see Contourites Unified Soil Classification see Soil classification Unifites, Angola Basin muds 109-11 terminology 12 United States mid-Atlantic region, canyons, late Quaternary 319-29 see also Wilmington Canyon Upwelling, classic mechanism 399-400 intensity, in planktic production 488-94 of nutrients 447 and productivity, deep sea 532-41,548-9 Uranium 511,515 Vague lamination (massive bedding) 490-2 Varve-type lamination Blake-Bahama Basin 437, 440-4 criteria 637 Monterey Formation 491 post glacial, Santa Barbara 379-80 Santa Barbara deposits 485, 4 9 0 - 2 Vegetation changes, in interpreting sediments 303 Ventura basin, California see California Continental Borderland tertiary deposits 420 VERTEX program, sediment trap arrays 76-7 Vertical flux, particles 56-7
659
Viscometer measurements, cohesive material, critical erosion stress, 40-1 Volcanic activity, carbon rich strata 548 events, active margins, periodicity 18 sources 17,18 tephras 172-5, 331-8 Water chemistry, varves and cyclical alternations 446 content of sediment 37-40 depth, and wave height, sandy shelves 340 trace fossil indicators 595-607 sampling, objectives and methods 73-5 Wave height and water depth, sandy shelves 340 Waves, effect on sedimentation 331,340-1 see also Internal waves; Sand waves Western North Atlantic see Blake Bahama basin Wet bulk density profiles 565, 571 Wilmington Canyon, canyons, late Quaternary 319-29 continental slope 319-20 discussion 329 physiography 320-4 processes 328 sediments 325-8 stratigraphy 325 see also Canyons Wind transport see Aeolian transport Winnowing, current 380 mechanisms, shelf to slope 340-1 X-radiography core analysis 325, 326, 379, 385-92 piston cores 379, 388, 389 turbidite 98, 102, 104-5, 107-8 sapropels 498-509 X-ray diffraction end member compositions, using discharge-weighting 364-70 mineralogy 335-7 techniques 200-1 Zaire deep-sea fan 95-112 conclusions 112 geological setting 95 grain size analysis 101, 109-11 hydrography 96 methods 97 study area 96 X-radiography 102, 104-8 Zaire River, suspended matter load 98, 99 Zephiran chloride, sterilisant 471 Zonation, bioturbational levels 600 Zoophycos
in biogenic traces, fine grained sediment 599-604 in Mediterranean sapropels 501 stratigraphy, Nova Scotia 312 Zooplankton, biofiltration of faecal pellets 395, 399-400, 412, 431