SEDIMENTARY FACIES ANALYSIS

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
 S P E C I A L P U B LI CA TI O N N U M B E R 22 OF THE

I N T E R N A T I O NA L A S S OCIA T I O N OF S ED I M E N T O L O GI S T S

Sedimentary Facies Analysis

A TRIBUTE TO THE RESEARCH AND TEACHING

OF HAROLD G. READING

EDITED BY A. GUY PLINT

b
Blackwell
Science
© 1995 The International Association DISTRIBUTORS
of Sedimentologists
Marston Book Services Ltd
and published for them by
PO Box 87
Blackwell Science Ltd
Oxford OX2 ODT
Editorial Offices:
(Orders: Tel: 01865 791155
Osney Mead, Oxford OX2 OEL
Fax: 01865 791927
25 John Street, London WClN 2BL
Telex: 837515)
23 Ainslie Place, Edinburgh EH3 6AJ
238 Main Street, Cambridge USA
Massachusetts 02142, USA Blackwell Science, Inc.
54 University Street, Carlton 238 Main Street
Victoria 3053, Australia Cambridge, MA 02142
(Orders: Tel: 800 215-1000
Other Editorial Offices: 617 876-7000
Arnette Blackwell SA Fax: 617 492-5263)
1, rue de Lille, 75007 Paris
Canada
France
Oxford University Press
70 Wynford Drive
Blackwell Wissenschafts-Verlag GmbH
Don Mills
KurfUrstendamm 57
Ontario M3C 119
10707 Berlin, Germany
(Orders: Tel: 416 441 2941)

Feldgasse 13, A-1238 Wien Australia

Austria Blackwell Science Pty Ltd
54 University Street
All rights reserved. No part of Carlton, Victoria 3053
this publication may be reproduced, (Orders: Tel: 03 347-0300
stored in a retrieval system, or Fax: 03 349-3016)
transmitted, in any form or by any
means, electronic, mechanical,
A catalogue record for this title
photocopying, recording or otherwise,
is available from the British Library
except as permitted by the UK
Copyright, Designs and Patents Act ISBN 0-86542-898-0
1988, without the prior permission
of the copyright owner.
Library of Congress
Cataloging-in-Publication Data
First published 1995
Sedimentary facies analysis: a tribute to
Set by Setrite Typesetters, Hong Kong the research and teaching
Printed and bound in Great Britain of Harold G. Reading/
at the Alden Press Limited, edited by A. Guy Plint.
Oxford and Northampton p. em.
(Special publication number 22 of the
International Association of Sedimentologists)
Includes bibliographical references
and index.
ISBN 0-86542-898-0
1. Facies (Geology)
2. Sedimentation and deposition. I. Plint, A. Guy.
II. Reading, H.G. [[[. Series: Special publication
of the International Association of Sedimentologists;
no. 22.
QE651.S43 1995
552' .5-dc20 94-30445
CIP
 To Harold

Photograph courtesy of Tim Barrett

We offer this collection of papers

as a token of our appreciation
for your friendship, guidance and inspiration.
In remembering your infectious enthusiasm, dedication
and sometimes daunting expectations,
we realize how deeply we were influenced
by your philosophy and attitude;
a gift that has, in no small measure, shaped the course
of our professional ]jves.

Your former students

 Contents

IX Preface

XI Harold G. Reading

xm Introduction

Clastic Facies Analysis

3 Alluvial palaeogeography of the Guaritas depositional sequence of southern Brazil

Paulo S. G. Paim

17 Sedimentology of a transgressive, estuarine sand complex: the Lower Cretaceous

Woburn Sands (Lower Greensand), southern England
Howard D. Johnson and Bruce K. Levell

47 An incised valley in the Cardium Formation at Ricinus, Alberta: reinterpretation as an

estuary fill
Roger G. Walker

75 Gravelly shoreface and beachface deposits

Bruce S. Hart and A. Guy Plint

101 The return of 'The Fan That Never Was': Westphalian turbidite systems in the Variscan
Culm Basin: Bude Formation (southwest England)
Robert V. Burne

137 Depositional controls on iron formation associations in Canada

Philip Fralick and Timothy J. Barrett

157 Facies models in volcanic terrains: time's arrow versus time's cycle
Geoffrey J. Orton

Tectonics and Sedimentation

197 Coarse-grained lacustrine fan-delta deposits (Pororari Group) of the northwestern

South Island, New Zealand: evidence for Mid-Cretaceous rifting
Malcolm G. Laird

VII
vm Contents

219 Sedimentation and tectonics of a synrift succession: Upper Jurassic alluvial fans and
palaeokarst at the late Cimmerian unconformity, western Cameros Basin, northern
Spain
Nigel H. Platt

237 The use of geochemical data in determining the provenance and tectonic setting of
ancient sedimetary successions: the Kalvag Melange, western Norwegian Caledonides
Rodmar Ravnas and Harald Fumes

265 Differential subsidence and preservation potential of shallow-water Tertiary sequences,

northern Gulf Coast Basin, USA
Marc B. Edwards

Sequence and Seismic Stratigraphy in Facies Analysis

285 Seismic-stratigraphical analysis of large-scale ridge-trough sedimentary structures in

the Late Miocene to Early Pliocene of the central North Sea
Joe Cartwright

305 Millstone Grit cyclicity revisited, II: sequence stratigraphy and sedimentary responses
to changes of relative sea-level
Ole J. Martinsen, John D. Collinson and Brian K. Holdsworth

Facies Analysis in Reservoir Sedimentology

331 Productive Middle East clastic oil and gas reservoirs: their depositional settings and
origins of their hydrocarbons
Ziad R. Beydoun

355 The evolution of Oligo-Miocene fluvial sand-body geometries and the effect on hydro­
carbon trapping: Widuri field, west Java Sea
Ray Young, W.E. Harmony and Thomas Budiyento

381 Index
 Preface

This book stands out in the series of Special Publi­ Bureau considered that this would be best done
cations of the International Association of Sedimen­ through a Special Publication on a subject in line
tologists. It is an acknowledgement of Harold with Harold's work (obvious topics were clastics,
Reading's commitment to lAS, for whom he has facies and depositional environments, sedimentation
been Publications Secretary, General Secretary and and tectonics).
President successively, over the last 30 years. It is therefore most appropriate that Guy Plint,
Harold has not only been source and inspiration as Editor chosen for this special publication, has
of many of the lAS policies and activities over this brought together a collection of original scientific
time, he has also been at the roots of 'facies sedimen­ papers authored by Harold Reading's students, or
tology' as an art in itself, and as a major tool in the students of theirs. To honour Harold Reading's own
broader field of geology. scientific scope, the subject chosen is broad: sedi­
More than providing his own personal contribution mentary facies analysis. The contributions contained
to this branch of the earth sciences, Harold created a in this Special Publication show to what extent facies
flourishing school of teaching and research. Harold's sedimentology, as fostered by Harold Reading, is
approach has burgeoned from Parks Road, Oxford, now established as a necessary basis to any under­
to become international not only through his standing of sedimentary rocks.
students, but also through 'his book'.
PETER HOMEWOOD
The Bureau of the lAS, taking up a suggestion by
/AS Publications Secretary
Robert Campbell of Blackwell Science, decided to
put together a scientific tribute to Harold. The

IX
 Harold G. Reading

Harold Reading was born in 1924 and, on leaving at Rijswijk, arranged for Harold to investigate the
school, joined the Indian Army. This early experience context of reported turbidites associated with English
left a lasting impression and undoubtedly contributed Carboniferous deltaics in the Pennines and in south­
to Harold's later concern for international cooper­ west England.
ation. He went up to Oxford in 1948, initially to read Cooperation with de Raaf and with Roger Walker,
Forestry, but his interests were diverted towards one of Harold's earliest research students, developed
geology and he graduated in that subject in 1951. a detailed appreciation of sedimentary structures
As an undergraduate, he visited North Norway to and their role in understanding processes, and led
investigate the Late Precambrian and Cambrian to the development of the style of facies analysis
stratigraphy of the Digermul Peninsula. This under­ exemplified by the 1965 classic paper on the Carbon­
graduate expedition not only shed significant new iferous cycles of North Devon. Thereafter, Harold's
light on the stratigraphy of the area but also sowed stable of research students grew rapidly as this volume
the seeds of a later rich sedimentological harvest. amply testifies. Until his retirement, it was unusual
Three years at Durham under K.C. Dunham led to a for him to have fewer than five or six doctoral
PhD with a project that involved mapping Carbon­ students at any one time, this in addition to a full
iferous Y oredale cycles across an area of bleak undergraduate teaching programme and responsi­
Pennine moorland. Although the main thrust of the bilities in college. This formidable work load was
study was stratigraphy and structure, the experience carried out with great conscientiousness but Harold
of Carboniferous cyclicity was to set a further pointer still had time to spare for external activities such
for the future. as his involvement with lAS and JAPEC. During
On completion of his PhD Harold joined Royal Harold's long career in Oxford, he only spent sus­
Dutch Shell and immediately found himself in the tained periods away on sabbatical on two occasions,
contrasting field conditions of Venezuela. This multi­ the first in Leyden in the mid-1960s and the second
national, multidisciplinary environment developed in Canada in 1972. The period in Holland led to
an appreciation of broader geological perspectives close cooperation with structural geologists working
and the pragmatic, though rigorous, approach to in the Cantabrians, an important extension of his
problem solving that has characterized Harold's interests outside Britain.
career. Of particular significance was a visit to Harold's earliest students developed his early
Venezuela by Ph. Keunen who was, at that time, interests, the Carboniferous deltas of Britain, and
actively promoting his pioneering work on turbidity the tillites, shallow-marine and fluvial sediments of
currents and their deposits. Kuenen's rigorous northern Norway. Later, the Lower Palaeozoic of
approach to understanding depositional processes Ireland and the Carboniferous of northern Spain
struck a chord with Harold, which was to be a were added. As students were attracted to Oxford
cornerstone of his approach to sedimentology. from around the world, the geographical spread
Harold returned to Oxford in 1957 as lecturer in grew. However, geographical diversification was not
geology, a position that he held until retirement in an end in itself but largely a result of Harold's
1991. When he took up his post, his teaching responsi­ curiosity about wider controls on sedimentation,
bilities included mapping and palaeontology and particularly the role of tectonics. He understood
stratigraphy. Sedimentology, as we know it, hardly very early the implications for sedimentology of
existed. Harold first revived his interests in northern Plate Tectonics, as exemplified by his pioneering
Norway through a further, largely stratigraphical paper with Andrew Mitchell. Curiosity about new
expedition to Digermul. Perhaps more importantly, geological ideas and the need to investigate their
he developed his interest in sedimentary process and implications for sedimentology and vice versa has
environments through a relationship with Shell. been a hallmark of Harold's geological thinking.
Maurits de Raaf, then Head of Geological Research By some standards, Harold has not been a prolific

XI
XII Harold G. Reading

author, although his papers are always thoughtful President have already been acknowledged by the
and stimulating. Published evidence of Harold's Association itself in the granting of Honorary
influence lies mainly in the rigour, originality and Membership to Harold. It is worth remembering
appreciation of the wider geological perspective that that it was in no small measure due to Harold's
characterize many of the publications of his research efforts that the Association changed from a largely
students and of second and third generation students. European organization to one of real international
Harold edited one Special Publication of the lAS stature. Harold's tireless efforts to meet and encour­
on-strike-slip mobile belts, but his most valued publi­ age sedimentologists of all ages and backgrounds
cation is the textbook Sedimentary Environments and around the world and his endless patience and
Facies, initially written largely by Harold's former diplomatic skill have been well rewarded in the
students and rigorously edited to reflect the high healthy Association that we enjoy today. Harold
standards he espouses. The 3rd edition currently has additionally been honoured by the Geological
occupies much of his 'retirement'. Society of London with the award of the Lyell Fund
Although this book is essentially a celebration and the Prestwich Medal and, most recently, by
of Harold's scientific influence, it is important, SEPM with the award of its prestigious Twenhofel
especially in a Special Publication of the lAS, to Medal.
acknowledge his enormous contribution to the
JOHN CoLLINSON
development of sedimentology internationally. His
Shrewsbury, UK
unstinting efforts on behalf of the lAS, as Publi­
cations Secretary, as General Secretary and as
 Introduction

This volume is a very personal compilation. Unlike from Harold's former graduate students and their
previous lAS Special Publications, it is not centred students and co-workers, but to impose no constraint
on a specific geological theme, and for that I make on topic, in order to illustrate the scope of Harold's
no apology. Instead, my intent was to illustrate, and knowledge, interest and vision. In consequence, the
celebrate, the breadth of interest, energy and inspi­ contents of this book are eclectic. The collection
ration that Harold Reading has brought to the field of papers serves to highlight the power of facies
of sedimentary geology. analysis, whether the rocks be volcanogenic, bio­
Few would deny the depth of Harold's influence genic, siliciclastic, or even 'catastrophic' (mega­
on sedimentology, world-wide. In part, this is due to olistoliths!), and of course reflect the scientific
his publications, in particular the enormously suc­ method fostered by Harold.
cessful Sedimentary Environments and Facies, It is particularly appropriate that, amongst the
unquestionably the cornerstone for all those who contributions, Ole Martinsen, John Collinson and
embark on sedimentary facies analysis! Equally Brian Holdsworth offer new interpretations of
important of course, has been his pivotal role in the Namurian deltaic rocks in the northern Pennines,
foundation and development of the lAS, a contri­ (upon which Harold cut his sedimentological teeth),
bution acknowledged recently with honourary mem­ but which, judging from referees comments, still
bership of that Association. provide fuel for heated debate! In similar vein,
His philosophy and attitude has of course travelled Bob Burne presents a review and discussion of the
with his graduate students, drawn from 13 countries depositional environment of the enigmatic Bude
on six continents. Because many of these students Formation (which Harold studied in the early
returned home upon completion of their work in 1960s), but which is still subject to sharply divergent
Oxford, and others now work and teach outside interpretations. In a salutory lesson to us all,
the UK, the approach Harold fostered during their Roger Walker shows how important it is, both to
graduate days has continued to spread. (He may not separate facts from interpretations, and to ques­
know this, but in a geneological sense, Harold tion one's cherished interpretation, when he boldly
is now a great-great grandfather to at least one reinterprets as an incised valley fill, rocks he pro­
young sedimentology student who doubtless is quite claimed a turbidite channel deposit just nine years
unaware of the history of the supervisory influence ago!
that has been passed down!) As Editor of this volume, I am indebted to the
Although initially conceived as a thematic volume following people whose thorough reviews served
with contributions to be invited from a panoply of to clarify the papers, and who made my job that
leading sedimentologists, two difficulties quickly much easier: Gail M. Ashley, Timothy R. Astin, T.
arose: first, just what was to be the theme? As Christopher Baldwin, Janok P. Bhattacharya,
Harold has been involved in so many areas of Charlie S. Bristow, H. Edward Clifton, Thomas C.
sedimentary geology, selection of any one topic Connally, Edward Cotter, William R. Dupre,
simply served to highlight gross neglect of another. Peter G. DeCelles, Frank G. Ethridge, Jill Eyers,
Secondly, it rapidly became apparent that numerous Stephen S. Flint, Edward C. Freshney, Robert L.
former students were anxious to pay their own Gawthorpe, Roland Goldring, Anthony J. Hamblin,
personal tribute, and whose contributions could, Alan P. Heward, Phillip R. Hill, Richard N. Hiscott,
alone, easily constitute a hefty volume! Of the 34 Richard S. Hyde, Elana L. Leithold, Peter J.
students whom Harold guided through doctoral McCabe, Kathleen M. Marsaglia, Franco Massari,
theses between 1961 and 1994, 16 have authored, or Gerrard V. Middleton, Robert A. Morton, George
co-authored papers in this volume. Postma, William C. Ross, Alastair H. Ruffell, Bruce
In keeping with the sentiment of this festschrift, I W. Sellwood, Gary A. Smith, Roger G. Walker,
took the decision to limit contributions to those James D. L. White, John A. Winchester and

xiii
XIV Introduction

Jonathon Wonham, plus two people who chose to clandestine spmt of this project, and provided
remain anonymous. essential intelligence on both Harold and his former
I am very grateful to Susan Sternberg, Edward graduate students.
Wates and Julie Elliott at Blackwell Science who
A. GuY PuNT
provided guidance at critical phases in the prep­
London, Ontario
aration of this book. I also thank Diana Relton
(Earth Sciences, Oxford) who entered into the
 Clastic Facies Analysis

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
Spec. Pubis int. Ass. Sediment. (1995) 22, 3-16

Alluvial palaeogeography of the Guaritas depositional sequence

of southern Brazil

P A U L O S. G. P AlM*
Earth Sciences Department , Oxford University, Parks Road,
Oxford OXI 3PR, UK

ABSTRACT

The Guaritas sequence is the uppermost stratigraphical level of the Camaqua Basin (southern Brazil)
and comprises an alluvial, deltaic and aeolian continental facies association up to 800 m thick. Facies
mapping of this unit has revealed a lateral association of tributary fans and trunk braided rivers
developed under semi-arid conditions.
Two main regions (lobes) of alluvial fan development can be discriminated and the source points of
both coincide with synforms in the nearby basement. This depositional system presents a normal down­
fan facies change. An anomalous lateral change of facies within the trunk river system is interpreted as
having been inherited from pre-existing alluvial fan deposits.
The main alluvial facies comprise trough cross-stratified (74%) and horizontally bedded (7%)
sandstones, massive (16%) and tabular cross-stratified (2%) orthoconglomerates, and massive mud­
stones (1%) .
Vertical aggradation of three-dimensional subaqueous dunes, followed by an upper flow regime
plane-bed phase, characterized the depositional events of the sandy areas of the alluvial system. Diffuse
gravel sheets and minor longitudinal and transverse bars were the main geomorphological features of
the gravelly alluvial reaches. Fine-grained sediments represent temporarily abandoned areas within the
braided channel network.

INTRODUCTION

The Guaritas depositional sequence constitutes the trending tectonic structure, in southern Brazil (Fig.
uppermost unit of the Camaqua Basin infilling and it 1}, and evolved during the latest phases of the
is an unconformity-bounded stratigraphical unit: it Brasilia no orogenic cycle (strike-slip basins of Brito­
overlies older deformed molasse strata (angular Neves & Cordani (1991}}.
unconformity) and is covered by Permian sedimen­ An extensional or transtensional event at the end
tary rocks of the Parana Basin. of the Brasiliano orogenic cycle, and the consequent
The Guaritas sequence, about 800 m thick, is formation of intermontane basins, has been pro­
almost always flat-lying, although, near to regional posed as the tectonic setting of the Camaqua Basin
faults some extensional reactivation has tilted the during the deposition of the Guaritas sequence
Guaritas deposits. The available radiometric dating, (Fragoso-Cesar et al., 1984, 1992; Beckel, 1990,
summarized in Soliani et al. (1984) and Fragoso­ 1992).
Cesar et al. (1984), indicates a Cambro-Ordovician In the past decade, the Guaritas depositional
age for the deposition of the Guaritas sequence. sequence has received attention from several authors
The Camaqua Basin is located in a NE-SW in terms of facies analysis and palaeoenvironmental
interpretation (Becker & Fernandes, 1982; Fragoso­
Cesar et al., 1984; Jost, 1984; Lavina et a!., 1985;
* Permanent address: UNISINOS- Departamento de
Geologia, Av. Unisinos 950, Sao Leopoldo RS, Caixa Beckel, 1990). Generally, these papers have indi­
Postal 275, CEP 93022-000, Brazil. cated continental sedimentation characterized by

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
3
4 P.S.G. Paim

Permo Triassic

�
L:...:!l
T Upper Vendian to Ordovician

BRAZIL � t-':(:�: .,
.. . Guaritas depositional sequence
>- .•
� >:<: older molasse sequences
c::

-� Middle to Upper Proterozoic

N

1· ;;; �
� granites

1
:f:f meta volcanic/sedimentary rocks

Archean to Lower Proterozoic

�
B.:::J
Study area

Br153
.....--_ Main roads
Scale (Km)
Major faults
- - - -
Ca9apava do Sui 6 12 18
6 3 0
2 Santana da Boa Vista

Fig. I. Location map and geological setting of the Camaqua Basin. Modified from DNPM/CPRM (1987).

alluvial fan and braided alluvial plain deposits, with terns of the Camaqua Basin was presented by Lavina
associated aeolian and lacustrine sediments. A semi­ et al. (1985). In this paper the alluvial facies were
arid environment has been proposed for the overall related to marginal alluvial fans (channel and debris
Guaritas sequence. flow deposits) associated with an axial braided
The most detailed study on the depositional sys- alluvial plain. Gravelly longitudinal bars and sandy
 Palaeogeography of the Guaritas sequence 5

subaqueous dunes and transverse bars were the main ALLUVIAL FACIES:
morphological elements attributed to the alluvial GENERAL FEATURES
palaeostreams.
Petrological studies by De Ros et al. (1994) To simplify terminology the lithofacies classifi­
on samples from alluvial and aeolian facies of the cation proposed by Miall (1977), as modified by
Guaritas sequence indicate the presence of: (i) fresh Miall (1978), Rust (1978) and Bromley (1991), was
feldspar and volcanic lithoclasts; (ii) aggregates of adopted. Table 1 presents the main characteristics of
hematite; (iii) oxidized grains; (iv) caliche (concen­ each sedimentary facies described in the field.
tric interlayering of calcite and iron oxide); and (v) The terminology and classification scheme pro­
silcretes. These early diagenetic features reflect arid posed by the SEPM (Society of Economic Paleon­
to semi-arid conditions during the deposition of the tologists and Mineralogists) Bedforms and Bedding
Guaritas sequence. Structures Research Symposium (Ashley, 1990) for
Basin-wide facies mapping of the Guaritas se­ description of large-scale flow-transverse bedforms
quence carried out by the author in 1988, reinforce (excluding antidunes) was adopted.
previous interpretations indicating intermittent vol­ The alluvial deposits (Table 1) are sand dominated
canic activity and an aeolian, alluvial and deltaic (facies S, 81% ) with a smaller amount of conglom­
facies association (Fig. 2). erates (facies G, 18% ) and an insignificant amount
Basin-scale changes of the alluvial facies charac­ of pelites (facies F, 1%). Facies S is composed
teristics suggest that an objective delineation of dis­ mainly of medium- to coarse-grained sandstones
tinct alluvial subenvironments is possible. These (41% ), with a significant proportion of pebbly to
alluvial subenvironments, as well as a brief descrip­ very coarse-grained (25%) and fine- to very fine­
tion and interpretation of the main alluvial facies, grained (15%) sandstones.
are the main subject of this paper. Facies G is composed of pebbles (9%) and
The alluvial deposits will be discussed in terms of granules (8% ) and minor amounts of cobbles (1% ).
their general features of texture, fabric, sedimentary A few boulders occur in the base of some conglom­
structures and palaeocurrent pattern on a basin­ erate beds, mainly near the eastern border of the
wide scale. Both the mean sedimentary facies charac­ Camaqua Basin.
teristics and the lateral facies changes within the The alluvial deposits are usually arranged in fining
alluvial system are described. upward cycles bounded by fifth-order surfaces (sensu
The data base includes 403 outcrop descriptions Miall, 1988). These cycles are 0.5-4 m thick and
distributed over an area of nearly 1600 km2 (see Fig. tens of metres in lateral extent (Fig. 3), both parallel
SA). This area was subdivided into 46 equal rec­ to and perpendicular to palaeoflow, and can be
tangles (8 x 9 km) and mean values of several par­ classified as laterally extensive to sheet-like deposits
ameters were calculated for each subdivision. The following the classification of Friend et al. (1979).
results of this approach are presented in Tables 1 & The proportion of the different textural classes
2 and summarized in Figs 5 & 6. This approach within the fining upward cycles changes laterally
involves comparison of values from different strati­ with increasing gravel content toward both margins.
graphical levels. The consistent results (see Fig. 5)
throughout the basin, with sampling at several
Conglomerates (G)
stratigraphical levels (Fig. 2), suggest that the
palaeoenvironments were more or less stationary Clast-supported conglomerates comprise around
throughout deposition of the Guaritas sequence. 18% of the alluvial facies and massive conglomerates
A detailed three-dimensional facies architecture are the most common (Table 1). Clast-supported
analysis (architectural elements approach of Allen conglomerates are a very common facies in the
(1983) and Miall (1985)), aiming to build up a local lowermost parts of the fining upward cycles.
alluvial model on a channel-fill scale, is part of my Massive clast-supported conglomerates (facies
ongoing studies and will be the subject of another Gm, Table 1) are the main lithotype of facies G and
publication. normal grading, clast orientation and imbrication
are their most conspicuous sedimentary features.
Facies Gp is characterized by gravels (mainly
pebbles) arranged in small- to large-scale, normally
isolated, sets of tabular cross-stratification. This
6 P.S.G. Paim

10km

\; ·>.1 Mainly alluvial facies Q Mainly eolian facies E:f=3-g Mainly deltaic facies

�
t:;;:.t1 Mainly volcanic rocks
o Pre-Guaritas
LA: j Permo Triassic
basement

Fig. 2. Three-dimensional view of Camaqua Basin and surrounding area (same region of Fig. 1): (A) topography and (B)
sketch of the Guaritas sequence facies.
 Palaeogeography of the Guaritas sequence 7

Table 1. Classification and relative percentage of the sedimentary lithofacies (lithofacies code adapted from Miall (1977,
1978) and Rust (1978))

Rock type Facies code Description Percentage

Conglomerates (G) Gm Massive or crudely bedded conglomerates (cobbles,

pebbles and granules) l6
Gp Small- to large-scale tabular cross-stratified
conglomerates (granules and pebbles) 2
Gt Small- to large-scale trough cross-stratified
conglomerates
Gms Massive, matrix-supported conglomerates (boulders
to granules dispersed in a muddy sand matrix)

Sandstones (S) St Small- to large-scale trough cross-stratified

sandstones (pebbly to very fine-grained) 74
Sh Horizontally bedded sandstones (medium to very
fine-grained) 7
Sp Medium to pebbly sandstone with small- to large­
scale planar cross-stratification

Mudstones (F) Fm Massive mudstones with mudcracks

Fl Laminated to rippled very fine sandstone to siltstone

Table 2. Relative percentage of trough cross-stratification also rare and occur, locally, near the eastern border
and horizontal lamination in each sandy textural class
of the Camaqua Basin. The main characteristic of
Sedimentary this facies is its chaotic arrangement of pebbles,
Texture Facies structures Percentage cobbles and, less commonly, boulders floating in a
muddy to sandy matrix.
Pebbly to very St Small scale 10
coarse grained Medium scale 49
Large scale 41 Sandstones
Sh 0
Trough cross-stratification (facies St, 74% ), in places
Coarse to St Small scale 10 disrupted and/or deformed by convolution, and
medium grained Medium scale 43 horizontal bedding (facies Sh, 7% ) are the main
Large scale 37 features of the alluvial sandy deposits (Figs 3 & 4).
Sh 10 Planar cross-stratification (facies Sp) is rare.
Facies St is characterized by very fine- to very
Fine to very fine St Small scale 15
coarse-grained sandstones with trough cross-bedding
grained Medium scale 37
Large scale 28 (Table 1). The cross-strata are predominantly of
medium to large scale in all textural classes, but the
Sh 20
proportion of small-scale trough cross-stratification
increases as sandstones become finer grained (Table
2). This facies is the most common in the fining
facies commonly occurs associated with facies Gm upward cycles.
(Fig. 3). Convolute bedding is common in trough cross­
Facies Gt is rare, finer grained than facies Gm and stratified sandstones (facies St). Within a single
Gp and characterized by small- to large-scale trough cross-stratified set, all gradations may occur from
cross-stratification (alternations of small pebbles and oversteep foresets, recumbent folding to intense
gravelly sands). This facies interfingers with facies deformation and even complete destruction of the
Gm and grades into facies St (Fig. 4). former bedding (facies Sm and Spo of Bromley,
Matrix-supported conglomerates (facies Gms) are 1991). Deformation near the top of the cross-
8 P.S.G. Paim

Fig. 3. Main alluvial lithofacies:

facies Gt, St, Sh, Spo and, in the
uppermost part of the picture, Fl,
Gm and Gp. Bar scale is 2 m long.

stratified set is commonly characterized by downcur­ nantly sandy and the suspension load, if deposited,
rent oversteepening of the cross-strata (Figs 3 & 4), was almost completely eroded by subsequent flood
and the intensity of convolution increases down the events. This type of stream commonly has a braided
slip-face. pattern characterized by low sinuosity and highly
Horizontal bedding (facies Sh) does not occur mobile channels (Collinson, 1986). The sheet-like
associated with pebbly and very coarse-grained sand­ geometry of the fining upward cycles enclosed by
stones and comprises 10% of the sedimentary struc­ fifth-order bounding surfaces suggests broad, shallow
tures of medium- to coarse-grained sandstones and channels.
20% of the fine- to very fine-grained sandstones In terms of the gravelly facies, the dominance of
(Table 2). This facies is often related to the upper­ clast-supported conglomerates (Table 1) is indicative
most parts of the fining upward alluvial cycles (Figs 3 of gravel deposition by strong tractive flows, whereas
& 4). the finer grained material (sand and mud) was still
Planar-tabular cross-stratified sandstones (facies being carried in suspension (Rust & Koster, 1984).
Sp) are not common in the Guaritas sequence alluv­ Thin beds of facies Gm associated with laterally
ial deposits (Table 1). They occur as small- to large­ extensive channels suggest the development of dif­
scale sets in pebbly to medium-grained sandstones fuse gravel sheets (Hein & Walker, 1977) by very
and are normally interlayered with facies St. extensive and shallow sheet-floods (Collinson, 1986).
Thicker deposits of facies Gm suggest deeper and
less ephemeral flows (Rust, 1978) causing more
Other facies
extensive vertical aggradation of gravel bars with
Massive mudstones are rare and commonly mud­ low depositional dips. These deposits commonly
cracks are their most conspicuous feature (facies have been associated with the development of longi­
Fm). Very fine-grained sandstones and siltstones tudinal and/or diagonal gravelly bars (Smith, 1970;
(facies Fl) are also, and can be either horizontal Rust, 1972, 1978; Miall, 1977, 1978; Rust & Koster,
(Fig. 3) or, more rarely, cross-laminated (Table 1). 1984; Collinson, 1986) under high water and sedi­
Both usually occur in the uppermost parts of the ment discharge (Hein & Walker, 1977).
fining upward alluvial cycles. Conglomerates with planar cross-stratification
(facies Gp) has been related to (i) two-dimensional
dune migration (transverse and/or linguoid gravel
Alluvial facies: summary of general features
bars of Hein & Walker (1977), Miall (1977) and
and interpretations
Middleton & Trujillo (1984)) as well as to (ii) later
The textural aspects (Table 1) suggest that the alluv­ modifications of longitudinal bars (Smith, 1970;
ial facies of the Guaritas sequence represent bedload Rust, 1978; Enyon & Walker, 1974) in modern
stream deposits in which the bedload was predomi- alluvial gravelly reaches. The frequent occurrence of
 ;;,o
!:)

�
�
-§
�
-.:;,
�
"'

C)
§
;::.
s
"'
"'
.E
"'
"'
Fig. 4. Detailed view of Fig. 3 (enlargement of its lower part): facies Gt, St, Sh, Spo and thin tabular beds of Gm. Bar scale is 2 m long. ;:s
'"'
"'

'.0
10 P.S.G. Paim

isolated sets of facies Gp within deposits of facies stress acting on a liquefied sand bed and caused by
Gm could be explained more easily by the second current drag (Allen & Banks, 1972; Doe & Dott,
hypothesis. 1980; Owen, 1987) or by the movement of large
Trough cross-stratified conglomerates are rare bedforms over an unconsolidated substrate during
(facies Gt) and have been associated with (i) three­ high-flow stages (Plint, 1983).
dimensional dune migration, as observed by Horizontal bedding (facies Sh) occurs most often
Fahnestock & Bradley (1973) and Galay & Neill in the finest fraction of the sandy deposits (Table 2).
(1967), and (ii) channel scour-and-fill structures This textural control, associated with the occurrence
(Miall, 1977; Middleton & Trujillo, 1984). The same of parting lineation and scattered small pebbles and
criteria previously used to interpret facies Gp can granules near the base of the horizontally bedded
also be applied in this case: the solitary nature of this sets, indicates its origin as an upper flow regime
facies suggests the deposition of gravel in depressions bedform.
around diffuse gravel sheets. Deposits with the same characteristics of facies Sh
Matrix-supported conglomerates (facies Gms) are usually have been linked to an upper flow regime
also rare and represent mud- and debris-flow deposits phase developed during flood stages on the channel
commonly associated with an alluvial fan setting floor (McKee et al., 1967; Williams, 1971; Miall,
(Blackwelder, 1928; Bull, 1963; Hooke, 1967; 1977) or under the influence of high-velocity and low
Rust & Koster, 1984; Collinson, 1986; Blair & depth flows on the top of sand-flats (Cant and
MacPherson, 1992). Walker, 1978; Miall, 1977; Collinson, 1986). The
Sandy sediments constitute the majority of the common occurrence of this facies (Sh) on the upper­
Guaritas alluvial deposits (Table 1) and are exten­ most parts of the fining upward cycles supports an
sively dominated by facies St (Table 2). Trough interpretation involving upper flow regime currents
cross-stratified sandstones have been related almost reworking the top of the previous alluvial deposits.
invariably to migration of three-dimensional dunes Planar-tabular cross-stratified sandstones (facies
(e.g. Collinson, 1970; Williams, 1971; Harms et al., Sp) are rare. Within alluvial settings this facies
1975; Miall, 1977; Rust, 1978). In braided alluvial commonly has been related to slip-face advance of
settings these bedforms usually have been associated two-dimensional dunes (transverse -linguoid or
with in-channel deposition (Cant & Walker, 1976, lobate bars of Collinson (1970, 1986), Smith
1978; Cant, 1978; Walker & Cant, 1984). Such (1970), Williams (1971), Asquith & Cramer (1975),
repetitive sand deposits commonly are considered as Miall (1977), Cant & Walker (1978) and Cant (1978);
flood-stage bedforms (Williams, 1971) and are larger or sand waves and straight-crested megaripples
in deep channels (Cant, 1978). of Smith (1970), Collinson (1986) and Miall
The association of facies St with the lower and (1978)). Smith (1970) related the origin of the two­
middle part of sheet-like fining-upward cycles dimensional dunes to the development of 'deltas' in
suggests this facies could be related to flood stage in pre-existing channel-floor depressions, whereas Cant
shallow channels. Subcritical climbing trough cross­ & Walker (1978) related them to flow expansion
strata (facies St) indicate subaqueous dune aggra­ at channel junctions or places where the channels
dation. Sporadic lateral accretion of these bedforms widen.
is indicated by inclined planes (first-order bounding Facies Fm and Fl are not common in the alluvial
surfaces of Miall (1988)) dipping perpendicular to system of the Guaritas sequence (Table 1). Their
the dune migration direction (Paim, 1994). rarity and generally lenticular geometry (Fig. 3) are
The absence of third-order surfaces (except the suggestive of waning flood deposits settling on to
rare occurrence of lateral accretion surfaces) associ­ temporarily abandoned areas of the braided system
ated with the subcritical climbing of the trough (Cant, 1978; Cant & Walker, 1978; Miall, 1978).
cross-bedded sets (facies St) suggests rapid depo­ In general, diffuse gravel sheets and longitudinal/
sition of a sandy load, transported by traction plus diagonal bars were the main geomorphological
suspension, without macroform (sensu Jackson, elements of the gravelly reaches, whereas sub­
1975) development. aqueous three-dimensional dunes characterized the
Deformation of trough cross-stratified sandstones sandy portions of the Guaritas alluvial system.
is a very conspicuous feature of the Guaritas sandy The predominance of vertical aggradation of dunes
alluvial facies. Recumbent folding in cross-bedded instead of downstream and/or lateral accretion of
sandstones commonly has been attributed to shear more stable sandy accumulations (e.g. sand-flats)
 Palaeogeography of the Guaritas sequence 11

suggests a highly variable hydrological character and 282°, with a correlation coefficient of 0. 86), reflecting
predominance of the upper part of lower flow regime a sedimentary input towards the basin axis;
conditions within the channels. 2 from the basin axis to the western border, palaeo­
Debris-flow and sheet-flood deposits suggest the currents were predominantly parallel to the struc­
presence of alluvial fans within the alluvial system as tural trend (general mean vector of 211 with a o,

well as flashy discharge due to sporadic, but torren­ correlation coefficient of 0.96).
tial, rainy seasons.
The above interpretations together suggest an
Pattern of textural dispersion
alluvial drainage developed under semi-arid con­
ditions (large discharge fluctuations) with alter­ The alluvial deposits of the Guaritas sequence are
nation of flood events and dry seasons. These composed primarily of sandstones (mainly facies St
conclusions are reinforced by the aeolian associ­ and Sh), minor conglomerates (mainly facies Gm
ation (Lavina et al., 198S) and by petrographical and Gp) and trace amounts of fine-grained sediments
evidence related to early diagenetic processes (facies Fm and Fl), as has been described in the
(De Ros et a/., 1994). previous section. In this paper three types of alluvial
deposits are distinguished: sandy (:2: 70% sand­
stones); mixed (70-30% sandstone); and conglom­
ALLUVIAL FACIES: eratic ( ::::: 30% sandstone).
LATERAL CHANGES Figure SC shows the percentage of sandstone (rela­
tive to conglomerate) through the entire basin and
The previous section describes the pattern of alluvial illustrates a gradual decrease from sand dominated
sedimentation in terms of mean regional values and, alluvial deposits along the basin axis (axial alluvial
in this way, reflects the major features of the alluvial sedimentation), to mixed alluvial deposits toward
deposit. In the following section, spatial variation in both basin margins (marginal alluvial sedimen­
some sedimentary features is described and, when tation). Likewise, Fig. SD presents a plan view of
possible, interpreted. To achieve this, the mean the spatial changes of the percentages of coarser
values, per unit area, of several sedimentary par­ grained sediments (conglomerates plus pebbly and
ameters were calculated using the outcrop locations very coarse-grained sandstones) relative to finer
and grid presented in Fig. SA. grained sediments (coarse to very fine sandstones).
A pattern quite similar to the former (Fig. SC) can
be seen. Clearly, the facies St and Sh are gradually
replaced by facies Gm towards both basin borders.
Palaeocurrent pattern
In both cases (Figs SC & SD) the only exception
The pattern of sediment transport within the entire to the general pattern of sediment distribution is a
Camaqua Basin was calculated using the grid and NW -SE trending intrusion of coarse material in the
outcrops shown in Fig. SA. southeast region of the basin.
In order to eliminate problems associated with the
analysis of several types and scales of sedimentary
Alluvial facies: interpretation of lateral changes
features (Miall, 1977) mean vectors were calculated
only from trough cross-stratification. By using only Figure 6 presents an interpretation of the alluvial
one rank of sedimentary features, difficulties related palaeogeography of the Guaritas sequence based on
to vector magnitude were eliminated (Allen, 1963; the lateral variations of the textural and palaeocur­
Miall, 1974). In addition, dunes seem to be associ­ rent data. This figure was constructed according to
ated with high-stage flow and, consequently, should the following considerations.
be good indicators of the true downstream direction The palaeocurrents suggest the coexistence of
(Miall, 1977). two distinct alluvial subenvironments (Fig. SB) with
The distribution of the palaeocurrent vector means almost orthogonal mean sedimentary transport pat­
(Fig. SB) indicates two major dispersal compart­ terns (282° versus 211°).
ments within the alluvial system: 1 The first dispersal system (282°), developed in the
1 from the eastern border to the basin axis the eastern part of the basin, is characterized by the
sedimentary transport was almost perpendicular to highest palaeocurrent vector dispersion and by palaeo­
the regional tectonic trend (a general mean vector of flow almost perpendicular to the tectonic trend of
12 P. S.G. Paim

B 12 - Mean vector and number of readings per area

Boundary between tributary alluvial fan system

and trunk braided river system

50 � Mean vector and number of readings of both alluvial

systems

\
A
5
3 46 ---

,514 / -36 1
60
44 ti 99
;; I
26 --
101 41 \ "4
/ 100 </:: 800
! 64 ""'-89 14
/ 32 /

3 ""'-
N
...... 26

c
§ >90
80-89
D
H 30-39
��29
70-79 0 �

Fig. 5. Lateral changes within the alluvial system: (A) grid and location of alluvial outcrops used to calculate palaeocurrent
and textural mean values; (B) palaeocurrent mean values per area; (C) percentage of sandstone relative to conglomerate;
and (D) percentage of coarser grained sediments (gravel plus pebbly to very coarse-grained sand) relative to finer grained
sediments (coarse to very fine-grained sand plus mud).

the basin. This subenvironment is interpreted as a the western portion of the basin, is interpreted as a
tributary alluvial fan (sensu Rust & Koster, 1984). trunk braided river (sensu Rust & Koster, 1984).
2 The second dispersal system (211°), represented As Collinson (1986) stated, 'it is sometimes poss­
by palaeoflow parallel to the basin axis, by low ible to identify individual fans by the establishment
palaeocurrent vector disperson, and characteristic of of a radial pattern of palaeocurrents over an area'.
 Palaeogeograph y of the Guaritas sequence 13

Gm n Sandy alluvial deposiiS

Gt
o Mixed alluvial deposiiS

St Tributary fan streams

Sh
- Braided trunk river streams
Spo
Sm Reworked tributary fans
1 - Mudstones
--- Boundary between trunk rivers and tributary fans
2 - Sandstones
3 - Conglomerates
A Plan view of the alluvial palaeogeography
B Western margin trunk river facies association
(reworked alluvial fans)
C Axial sandy trunk river facies associalion
D Eastern margin proximal alluvial fan facies
association
E Summary of the alluvial facies

Fig. 6. Alluvial palaeogeography and lateral facies changes: (A) plan view of the alluvial subenvironments; (B) and (D)
vertical profiles of marginal facies association; (C) vertical profile of axial facies association; and (E) ideal vertical
arrangement of the main facies. Facies code from Miall (1978); Bromley (1991).

Here, the tributary fan mean vectors (Fig. 5B) of easily erodible metapelites) in the nearby base­
indicate the coalescence of two main fan lobes (Fig. ment (Fig. 2). The southern lobe was more important
6A). The northern fan has a radius of 15 km, whereas than the others in terms of sedimentary input, as can
the southern fan has a radius of 20 km. These dimen­ be deduced both from it having penetrated furthest
sions are comparable to the size of recent examples into the basin (Fig. 5B) and from the major intrusion
of semi-ariel and arid alluvial fans documented by of coarsest material from southeast to northwest
Heward (1978). The point source of both lobes (Figs 5C & 5D) in the southeast region of the basin.
coincides with structural lows (synforms composed Comparison of the grain size distribution (Figs 5C
14 P. S.G. Paim

& 50) and palaeocurrent mean vectors (Fig. 5B) CONCLUSIONS

reveals some obvious relationships as well as some
discrepancies. The alluvial deposits of the Guaritas sequence reflect
The gradual decrease of grain size, from both a lateral association of tributary alluvial fans and
basin margins towards the basin axis (Figs 5C & 50) trunk braided rivers. The alluvial fans show a down­
can be interpreted as a consequence of lateral alluv­ stream decrease in mean grain size whereas the
ial fan input. This matches the palaeoflow data of trunk rivers present no longitudinal variation in
the eastern side of the basin, but not that on the texture. Instead, the trunk rivers show a lateral
western side (Fig. 5B). grain-size change that is interpreted to have been
The textural, palaeocurrent and facies data sum­ inherited from a hypothetical alluvial fan system fed
marized in Fig. 6A demonstrate that the grain-size from the western margin of the basin.
distribution within the trunk braided river system The alluvial fans are dominated by water­
does not show a downcurrent fining, which is a very laid deposits and comprise two main lobes. The
common characteristic of many braided alluvial source points of both alluvial fan lobes coincide with
environments (e.g. Smith, 1970; Miall, 1977, 1978; structural lows, suggesting control by basement
Rust, 1978; Collinson, 1986). Instead, the trunk topography.
rivers present a lateral change from sand-dominated Semi-arid conditions during the alluvial depo­
deposits near the basin axis to mixed deposits sition are suggested by: sheet-flow and debris-flow
towards the western basin border. Such a discrepancy deposits; petrological evidence, such as fresh feld­
can be related to a dominant alluvial input from the spar and volcanic lithoclasts, interstitial hematite
eastern border (tributary alluvial fans) causing the aggregates, caliche and silcretes; and the association
development of a trunk braided river system on the with aeolian facies (although aeolian facies are com­
western side of the Camaqua Basin. The emplace­ mon in early Palaeozoic sequences because of the
ment of trunk rivers in the western region could absence of land vegetation regardless of climate).
cause a major remoulding of the alluvial fan deposits Diffuse gravel sheets and longitudinal bars were
on the western border without erasing the down-fan the main geomorphological elements in the alluvial
fining. Some of these previous alluvial fan deposits gravelly reaches. Subaqueous dune aggradation, fol­
could be preserved and thus could explain some lowed by partial reworking of the deposit by upper
atypical palaeocurrent readings made near the west­ flow regime currents, characterized the sandy
ern margin, which point to a southeasterly directed reaches.
sediment discharge. An idealized channel-fill succession is typified,
Transitions between purely sandy or gravelly from base to top, by: (i) a horizontal to slightly
reaches were the norm inside the alluvial system undulatory erosional surface; (ii) gravel deposits,
of the Guaritas sequence. The facies are usually representing diffuse gravel sheets and longitudinal
arranged as fining upward cycles, bounded by fifth­ bars (Gm), locally with avalanche faces (Gp); and
order bounding surfaces, with the major facies super­ (iii) sandy deposits, consisting mainly of three­
imposed in the following order: Gm-Gp-St-Sh­ dimensional subaqueous dunes (St), rare two­
Fm. dimensional dunes (Sp), and plane beds (Sh), on the
This characteristic vertical arrangement, reflecting top.
the proportion of each facies (Table 1), is illustrated
in a summary (idealized) vertical profile (Fig. 6E)
incorporating the mean values of the principal facies ACKNOWLEDGEMENTS
observed throughout the basin. Comparison of
the summary profile (Fig. 6E) with actual sections Thjs study was carried out during the tenure of a
around the basin, and with the regional textural postgraduate scholarship awarded by the Research
variation (Figs 5C & 50), facilitates the identification Council of the Brazilian Government (CNPq), and
of some common marginal and axial facies associ­ forms part of the author's D .Phil. thesis at the
ations, summarized in Figure 6: profiles B and D University of Oxford, England, written under
typify common marginal facies associations, whereas the supervision of Dr H.G. Reading. Field-work
profile C represents the axial part of the basin. costs were supported by CNPq (Grant 413321/
88-6), Universidade do Vale do Rio dos Sinos
(UNISINOS), and Company of Research of Mineral
 Palaeogeography of the Guaritas sequence 15

Resources (CPRM) of the Brazilian Government. sandy braided river sedimentation: comparison of the
The author wishes to thank H. G. Reading and H.C. South Saskatchewan river and the Battery Point Forma­
Jenkyns for criticism and revision of an earlier version tion. In: Fluvial Sedimentology (Ed. Miall, A.D.). Can.
Soc. petrol. Geol., Calgary, Memoir 5, pp. 627-639.
of the manuscript. Later reviews by G. Plint, G.V. CANT, D.J. & WALKER, R.G. (1976) Development of a
Middleton, A.P. Hamblin and P.A. Allen have braided-fluvial facies model for the Devonian Battery
enabled me to make several very useful improve­ Point, Quebec. Can. J. Earth Sci. , 13, 102-119.
ments to the paper. CANT, D.J. & WALKER, R.G. (1978) Fluvial process and
facies sequence in the sandy braided South Saskatchewan
river, Canada. Sedimentology, 25, 625-648.
CoLLINSON, J.D. (1970) Bedforms of the Tana river,
RE FERENCES Norway. Geogr. Annaler, 52A, 31-55.
COLLINSON, J.D. (1986) Alluvial sediments. In: Sedimen­

ALLEN, J.R.L. (1963) The classification of cross-stratified tary Environments and Facies (Ed. Reading, H.G.).

units, with notes on their origin. Sedimentology, 2, pp. 20-62. Blackwell Scientific Publications, Oxford.
93-114. DERos, L.F., MORAD,S. & PAlM,P.S.G. (1994) The role
ALLEN, J.R.L. (1983) Studies in fluviatile sedimentation:
of detrital composition and climate on the diagenetic
bars, bar-complexes and sandstone sheets (low-sinuosity evolution of continental molasse: evidence from the
braided streams) in the Brownstones (L. Devonian), Cambro-Ordovician Guaritas sequence, southern Brazil
Sediment. Geo/., 92, 197-228.
Welsh borders. Sediment. Geol., 33, 237-293.
ALLEN, J.R.L. & BANKS, N.L. (1972) An interpret­
DNPM/CPRM (1987) Cachoeira do Sui- F6lha SH.
atiOn and analys1s of recumbent-folded deformed cross­ 22-Y-A. Projeto mapas metalogeneticos e de previsao
bedding. Sedimentology, 19, 257-283. de recursos minerais, texto e mapas. Escala 1:250000.
ASHLEY, G.M. (1990) Classification of large-scale sub­
Brasflia. MME/DNPM/CPRM, 13 pp.
aqueous bedforms: a new look at an old problem. DoE, T.W. & Dorr, R.H.,J R. (1980) Genetic significance
1. sediment. Petrol., 60(1), 160-172. of deformed cross bedding- with examples from the
AsQUITH, G.B. & CRAMER, S.L. (1975) Transverse braid
Navajo and Weber sandstones of Utah. J. sediment.
Petrol., 50(3), 793-812.
bars in the Upper Triassic Trujillo sandstone of the
Texas Panhandle. J. Geol., 83, 657-661. EYNON, G. & WALKER, R.G. (1974) Facies relationships in

BECKEL, J. (1990) Metalogenia del Cu, Pb y Zn en

Pleistocene outwash gravels,Southern Ontario: a model
Ia Cuenca de Camaquii durante e/ Cicio Orogenico
for bar growth in braided rivers. Sedimentology, 21,
Brasiliano- Rio Grande del Sur (Brasil). Unpublished
43-70.
PhD thesis, Universidad de Salamanca,274 pp. FAHNESTOCK, R.K. & BRADLEY, W.C. (1973) Knik and

BECKEL, J. (1992) Evolus;ao geotect6nica da Bacia do

Mantanuska rivers, Alaska: a contrast in braiding. In:
Fluvial Geomorphology (Ed. Morisawa, M.). State
Camagua, RS- Proposta de classificas;ao dentro de um
conceito mobilista. J Workshop s6bre as bacias molassicas University of New York, Binghamton, Proc. 4th
Geomorph. Symp., pp. 220-250.
brasilianas, Boletim Especial Resumos Expandidos,
FRAGOSO-CESAR, A.R.S., LAVINA, E.L., PAlM, P.S.G. &
pp. 1-5. Sao Leopoldo.
FACC!Nl, U.F. (1984) A antefossa molassica brasiliana do
BECKER, M. R. & FERNANDES, L.A.D. (1982) Caracter­
izas;ao faciol6gica de uma sequencia vulcano-sedimentar cinturao Dom Feliciano no escudo do Rio Grande do
eo-Paleoz6ica na f6lha Passo do Tigre ( RS). Acta geo/. Sui. Congress. Bras. Geol., 33 (Anais 7), 3272-3283.
Leopold. , 13(6), 287-321.
FRAGOSO-CESAR, A.R.S., MACHADO, R., SAYEG, H.S.

BLACKWELDER, E. (1928) Mudflow as a geologic agent

& FAMBRJNJ, G.L. (1992) Bacias orogenicas do ciclo
in semi-arid mountains. Geol. Soc. Am. Bull., 39, brasiliano no Rio Grande do Sui e Uruguai. J Work­
shop s6bre bacias moltissicas brasilianas, Boletim
465-483.
BLAIR, T.C. & McPHERSON, J.G. (1992) The Trollheim Especial Resumos Expandidos, pp. 47-53.
FRIEND, P.F., SLATER, M.J. & WILLIAMS, R.C. (1979)
alluvial fan and facies model revisited. Geol. Soc. Am.
Bull., 104, 762-769.
Vertical and lateral building of river sandstone bodies,
BRITO NEVES, B.B. & CoRDANl, U.G. (1991) Tectonic
Ebro Basin,Spain. J. geo/. Soc. London, 136,39-46.
GALAY, V.J. & NEILL, C.R. (1967) Discussion of 'No­
evolution of South America during the Late Protero­
zoic. Precam. Res. , 53, 23-40. menclature for bed forms in alluvial channels'. J. Hydr.
Div. A mer. Soc. Civ. Eng. , 93, 130-133.
BROMLEY, M.H. (1991) Variation in fluvial style as revealed
HARMS, J.C.,SOUTHARD, J.B., SPEARING, D.R. & WALKER,
by architectural elements, Kayenta Formation, Mesa
Creek, Colorado, USA: evidence for both ephem­ R.G. (1975) Depositional Environments as Interpreted
from Primary Sedimentary Structures and Stratification
eral and perennial fluvial processes. In: The Three­
dimensional Facies Architecture of Terrigenous Clastic
Sequences. Society of Economic Paleontologists and
Sediments and its Implication for Hydrocarbon Discovery
Mineralogists, Tulsa,Short Course 2.
HElN, F.J. & WALKER, R.G. (1977) Bar evolution and
a_nd Recovery (Eds Miall, A.D. & Tyler, N.). Soc.
Econ. Paleontol. Mineral. Spec. Publication, Concepts development of stratification in the gravelly braided
Sedimentol. Paleontol. Tulsa, 3, 94-102. Kicking Horse River, British Columbia. Can. J. Earth
Sci. , 14, 562-570.
BuLL, W.B. (1963) Alluvial fan deposits in Western Fresno
HEWARD, A.P. (1978) Alluvial fan sequence and mega­
County, California. J. Geol., 71, 423-251.
CANT, D.J. (1978) Development of a facies model for
sequence models: with examples from Westphalian
16 P.S.G. Paim

D- Stephanian B coalfields, Northern Spain. In: Fluvial Steel, R.J.). Can. Soc. petrol. Geol.,Calgary, Memoir
Sedimentology (Ed. Miall, A.D.). Can. Soc. petrol. 10, 189-201.
Geol.,Calgary, Memoir 5,597-604. OwEN, G. (1987) Deformation process in unconsolidated
HooKE, R. L E B. (1967) Process and arid-region alluvial sands. In: Deformation of Sediments and Sedimentaty
fans. J. Geol., 75, 438-460. Rocks (Eds Jones, M.E. & Preston, R.M.F.). Geol.
JACKSON, R.G., II (1975) Hierarchical attributes and a Soc. London, Spec. Pub!., No. 29, 11-24. Blackwell
unifying model of bed forms composed of cohesionless Scientific Publications, Oxford.
material and produced by shearing flow. Geol. Soc. Am. PAlM, P.S.G. (1994) Depositional Systems and Palaeogeo­
Bull., 86,1523-1533. graphical Evolution of the Camaqua and Santa Barbara
losT,H. (1984) Sedimentacao e vulcanismo durante o ciclo Basins, Brazil. Unpub. D.Phil. thesis, University of
Brasiliano no Rio Grande do Sui: Uma revisao. Congres. Oxford,277 pp.
Bras. Geol., 33, 3421-3457. PuNT, A. G. (1983) Sandy fluvial point bar sediments from
LAVINA, E.L., FACC!Nl, U.F., PAlM, P.S.G. & FRAGOSO­ the Middle Eocene of Dorset, England. In: Modern
CESAR, A.R.S. (1985) Ambientes de sedimenta�ao and Ancient Fluvial Systems (Eds Collinson, J.D. &
da Bacia do Camaqua, RS. Acta geol. Leopold., 21, Lewin,J.). Spec. Pub!. int. Ass. Sediment. No. 6, 19-
185-227. 33. Blackwell Scientific Publications, Oxford.
McKEE, E.D., CROSBY, E.J. & BERYHILL, H.L. (1967) RusT,B.R. (1972) Structure and process in a braided river.
Flood deposits, Bijou Creek, June (1965) J. sediment. Sedimentology, 18,221-245.
Petrol. , 37, 829-851. RusT, B.R. (1978) Depositional model for braided alluv­
MIALL, A.D. (1974) Palaeocurrent analysis of alluvial sedi­ ium. In: Fluvial Sedimentology (Ed. Miall, A. D.). Can.
ments: a discussion of directional variance and vector Soc. petrol. Geol., Calgary, Memoir 5, 605-625.
magnitude. J. sediment. Petrol. , 44(4), 1174-1185. RusT, B.R. & KosTER, E.H. (1984) Coarse alluvial
MIALL, A.D. (1977) A review of the braided-river depo­ deposits. In: Facies Models (Ed. Walker,R.G.). Geosci.
sitional environment. Earth Sci. Rev. , 13, 1-62. Can., Reprint Ser. 1, 71-89.
MIALL, A.D. (1978) Lithofacies types and vertical profile SMITH, N.D. (1970) The braided stream depositional
models in braided river deposits: a summary. In: Fluvial environment: comparison of the Platte River with some
Sedimentology (Ed. Miall, A.D.). Can. Soc. petrol. Silurian clastic rocks,North Central Appalachians. Geol.
Geol., Calgary, Memoir 5, 597-604. Soc. Am. Bull. , 81,2993-3014.
MIALL, A.D. (1985) Architectural-element analysis: a new SOLIANI, E., JR.,FRAGOSO-CESAR, A. R.S.,TEIXEIRA, W. &
method of facies analysis applied to fluvial deposits. KAWASHITA, K. (1984) Panorama geocronol6gico da
Earth Sci. Rev. , 22, 261-308. por�ao meridional do escudo Atlantica. Congress. Bras.
MIALL, A.D. (1988) Architectural elements and bounding Geol., 33, 2435-2449.
surfaces in fluvial deposits: anatomy of the Kayenta WALKER, R.G. & CANT, D.J. (1984) Sandy fluvial systems.
Formation (Lower Jurassic), Southwest Colorado. Sedi­ In: Facies Models (Ed. Walker, R.G.). Geosci. Can.,
ment. Geol., 55,233-262. Reprint Ser. 1,71-89 (2nd Edn).
MIDDLETON, L.T. & TR UJ IL LO,A.P. (1984) Sedimentology WILLIAMS, G.E. (1971) Flood deposits of the sand-bed
and depositional setting of the Upper Proterozoic Scan­ ephemeral streams of central Australia. Sedimentology,
lan Conglomerate, Central Arizona. In: Sedimentology 17, 1-40.
of Gravels and Conglomerates (Eds Koster, E.H. &
Spec. Pubis int. Ass. Sediment. (1995) 22, 17-46

Sedimentology of a transgressive, estuarine sand complex:

the Lower Cretaceous Woburn Sands (Lower Greensand),
southern England

H O W A R D D . J O H N S O N* and B R U C E K . L E V E L L t
* Department of Geology, Imperial College of Science, Technology and Medicine,
Prince Consort Road, London SW7 2BP, UK; and
t Shell UK Exploration and Production Ltd, Shell-Mex House, Strand,
London WC2R ODX, UK

ABSTRACT

A sedimentological investigation of the Lower Cretaceous Woburn Sands of southern England has been
used to develop a depositional model for a transgressive estuarine (or embayment) sand complex.
The Woburn Sands average 70 m in thickness, but are over 100 m thick in places, and infill a NE-SW
trending trough 25-30 km wide, which cuts into the western end of the NW-SE trending London­
Brabant land mass. Initially, this trough was of limited extent to the northeast but opened out into a
broader shallow sea to the south and southwest. Subsequently, during the course of the major Early
Cretaceous (Aptian-Albian) transgression, this feature formed a seaway connecting two previously
separate basins (the North Sea Basin to the north and the Weald-Wessex-Channel Basin to the south) .
The Woburn Sands record this overall transgressive history i n the form of six main facies bodies, which
occur in five erosionally based units (from bottom to top) : (i) Orange and Heterolithic Sands; (ii) Silver
Sands; (iii) Silty Beds; (iv) Red Sands; and (vi) Transition Series.
The lowermost deposits (the Orange and Heterolithic Sands; equivalent to the Lower W oburn
Sands) display convincing evidence of tidal current deposition (e.g. bimodal-bipolar palaeocurrent
patterns, herringbone cross-bedding, clay drapes and wavy- flaser bedding) . They are further character­
ized by large-scale, subhorizontal and low-angle erosion surfaces, which are interpreted as tidal channel
bases and tidal shoal accretion surfaces, respectively. The overlying sand deposits (the Silver and Red
Sands; equivalent to the Upper Woburn Sands) display similar evidence of tidal current activity but are
distinguished by overall coarser grain sizes, better sorting, lack of clay layers and the abundance of
large-scale cross-bedding. The large-scale structures in the Silver and Red Sands dip mainly towards the
south (inferred ebb direction), whereas similar structures in the exposed Orange Sands dip mainly to the
northwest (inferred flood direction) .
The overall sequence is interpreted in terms of a transgressive tide-dominated estuary or embayment
model. The Lower Woburn Sands (Orange and Heterolithic Sands) were deposited in m utually evasive
ebb and flood tidal channels and intervening tidal shoals, probably in an inner estuarine environment . In
contrast, the higher energy Silver and Red Sands were deposited in the outer reaches of an estuary or
embayment where greater water depths allowed the build-up of large-scale bedforms. The southward
increase in both cross-bed set size and sand-body thickness in the Red Sands probably reflects general
southward deepening. The final element in this facies succession is the draping of the sand complex by
slowly deposited fossiliferous marine beds (Transition Series and Basal Beds of the Gault), which are
overstepped to the north by the shallow-marine muds of the Gault.
The description, interpretation and depositional model outlined here may assist in the recognition
and prediction of similar shallow marine sand bodies. This study demonstrates, for example, that thick,
high reservoir-quality sands with favourable geometries for stratigraphical traps can accumulate in
transgressive estuaries and embayments. The resulting sand complex could be expected to comprise
several erosively bounded, lenticular units displaying rapid lateral thickness and facies variations and an
upward increase in reservoir quality and sand continuity. This would contrast with the more tabular
geometry and more gradual lateral thickness and facies variations of similar tide-dominated deposits
developed in offshore/shelf environments.

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
17
18 H. D. Johnson and B.K. Levell

INTRODUCTION

Shallow-m arine sands (i.e those deposited in w ater both before and since completion of our work.
depths of 10-200 m and rangi n g from inshore/sub­ Secondly, the 1 980 Abstract is cle arly an i n adequate
tidal to offshore/neritic environmen ts) h ave received reference document, but is nevertheless quoted
less attention th an most other cl astic deposits and , by workers studying these exposures. Thirdly,
despite recent adv ances, current facies models these exposures comprise sand quarries, which are
remain rel atively generalized. constantly changing as a result of con tinuing s and
Recent studies demonstrate that sh allow-marine extraction. Hence, new observations are frequently
sands occur in a v ariety of settings and owe their m ade and so the d at a acquired during the course of
variability to several factors, p articularly the complex this study, despite the time l apse , are still worthy
relationships between fluctu ations in w ater depth , of fuller documentation. Publishin g the results of
subsidence rates, morphology of the coastal zone, this study will also allow our interpretation to
sediment supply and the hydraulic regime of the be more critically eval u ated and will form a more
b asin, includin g the shoreline (e . g. Swift & Thorne, l astin g contribution to the analysis of this important
1991) . Most studies of shallow-marine sand bodies geological unit.
distin guish those resulting from tide-domin ated The m ain aim of this p aper, therefore , is to provide
processes and those formed m ainly by w ave- and a sedimentological description and to argue a
storm-domin ated processes (Johnson & B aldwin , deposi tional model for the Lower Cretaceous
1986; Dalrymple, 1992; W alker & Plint, 1992). In (Apti an- Albi an) Woburn S ands at Lei ghton
this context the Lower Cretaceous Woburn S ands Buzzard, southern E n gl and (Fi gs 1 & 2). It is not our
of southern E n gl and h as been quoted repe atedly intention to comprehensively evalu ate our findin gs
as a prime example of an ancient tide-domin ated in the context of more recent rese arch undertaken
shallow-marine sand complex, with the spectacul ar on the Woburn S ands, p artly because much of this
l arge-scale cross-beddin g interpreted as the deposits remains unpublished, p articularly a detailed an alysis
of tid al sand w aves (e . g. Ben tley , 1970; de R aaf of sedimen tary structures of Buck ( 1987) , a litho­
& Boersma, 1 97 1; Walker, 1984; Buck , 1985 ; strati graphical study by Eyers ( 1992 a) and a recent
Dalrymple, 1992). However, there h as been less sequence strati graphical analysis by Won h am ( 1993).
agreement on the specific type of tid al environment
preserved in the Woburn S ands and on its overall
genetic evolution. Because the resultin g product is a GEOLOGICAL FRAMEWORK
thick sand of high reservoir quality, i t w as felt th at a
better understanding of the Woburn S ands would
Stratigraphic framework
assist in the development of strati graphical models
of tid al sand bodies. This could aid their recognition The Woburn S ands h ave been shown throu gh field
in the subsurface , allow comparison with the geo­ m appin g and borehole evidence to comprise a lens­
metry and intern al ch ar acteristics of other sh allow­ shaped sand complex up to c. 100 m thick (Fig. 3 ) ,
m arine s and bodies (e. g. Exum & H arms, 1968; which in fills a rel atively n arrow (25 -3 0 km wide)
McCubbin, 1969; Campbell, 1971; Spearin g, 1 975) N E - SW trendin g trough (Bristow, 1963; Wy att
and contribute to the development of depositional et a/., 1986). The trough, which m ay be p artly tec­
models for hydrocarbon exploration and production. tonic in ori gin (Eyers, 1 991) , came into existence in
This w as the b ackground to a field study conducted the L ate J ur assic, but w as infilled only in the E arly
in 197 9 when both authors were employed by the Cre taceous. At Lei ghton Buzzard, which is on the
Koninklijke/Shell Exploratie en Produktie L ab­ western m argin of the trough, the upper Aptian to
oratorium (KSEPL) in Rijswijk , The Netherlands. lower Alb i an Woburn S ands unconformably overlie
Subsequently, the results were presented at the Upper Jurassic cl ays and are overstepped northwards
1 980 Annu al Conference of the AAPG (Johnson & by the Albian G ault (Fig. 4).
Levell, 1980) and documented in an internal report Ammonites and brachiopods found in the b asal
in 1 982. This l atter report forms the b asis for this phosphatic gr avels of the Woburn S ands are assi gned
p aper, which is presented here for several re asons. to the upper Apti an ( nutfieldiensis zone; Casey,
First, the model presented here is different in 196 1) . Althou gh the Woburn S ands currently
several respects to those interpretations published exposed contain an abundan t and diverse ichno-
 Transgressive estuarine sand complex 19

2' 0' - Aptian-Albian Lower Greensand

outcrop distribution
0 50 100km

LHTON
BUZZARD

Fig. l. Location map showing the

distribution of Lower Cretaceous
outcrops.

fauna they do not contain a shelly fauna, possibly Barremian) the London- Brabant land mass formed
because of leaching of calcium carbonate. The over­ an intermi ttent land barrier between a shallow
l ying beds be long to the Transition Series, which marine southern North Sea Basin (the Boreal Sea)
comprises a thin ( 1 -2 m) , complex and relatively to the north, and a freshwater Wealden Basin to the
poorly exposed succession of variable lithologies, south (Fi g. 5A) . This southern basin, to ge ther with
including the Shenle y Limestone and various iron­ the Channel, Southwestern Approaches, Celtic Sea
cemented beds referred to locall y as 'Carstone ' and Bristol Channel Basins ( Ziegler, 1988, 1990) ,
( Fi g. 3). The Shenle y Limestone contains lower formed a series of mainly separate and active, fault­
Albian (tardefurcata to mammillatum zone) fauna bounded basins in the southern British Isles , which
and represents a depositional hiatus with a complex underwent a c. 40-million-year period of alluvial
depositional and diagenetic history (Eyers, 1992b). sedimentation and the deposition of 'Weald en ' facies
The condensed lower Albian Gau lt (dentatus zone) ( P. Allen, 198 1 ) .
represents a northward overstep, which can be This period o f non-marine sedimentation was
related to the ' 108 Ma' maximum floodin g surface of terminated by the major Aptian-Albian marine
Haq et a!. ( 1987). The paucity of datable fauna trans gression and the deposition of the Lower
within the Woburn Sands essentially precludes Greensand Group. This was mainly a consequence
further correlation wi th events elsewhere in the of continued sea-floor spreading and northward
Lower Greensand basin (Ruffell & Wach, 199 1 , extension of the Atlantic Ocean and the Early
i n press). Cretaceous eustatic sea-level rise. This resulted
in progressive, northeastward marine inundation
of the Southwestern Approaches, Channel and
Palaeogeographical setting and depositional
Weald Basins. Simultaneously, there was also
environments
northwestward-directed marine transgression
Duri n g the Early Cretaceous ( Ryazanian to through the Paris Basin into sou thern England and
20 H. D. Johnson and B.K. Levell

Three main periods of marine transgression are

@
f!7' Jane's Pit recorded in the Lower Greensand of southern
England, each associated with depos it ion of exten­
s ively cross-bedded t idal deposits (Bridges, 1982).
Heath and Reach
The upper Apt ian to lower Alb ian Woburn Sands
A are associated w ith the transgressive breaching of
the London-Brabant land ma�s (Figs SB & SC) .
However, although the Woburn Sands clearly
represent, in broad terms, a transgressive shallow­
marine sand deposit , more precise environmental
Old { @ interpretations have remained uncertain. The large­
Llnslade
Road
scale cross-bedding, extensive bioturbation and evi­
dence of reversin g currents has led most authors
to su ggest a t idal environment (Lamplugh, 1922;
Schwarzacher, 1953; Bentley, 1970; de Raaf &
Boersma, 197 1 ) , but these authors d isagree , or are
non-committal, about specific t idal subenviron­
ments, with su ggestions ranging from open shelf to
tidal fiat for different facies in the complex. From
these possibilities, two main deposit ional models
emerge for all or part of the Woburn Sands:
1 t idal shelf or seaway, such as the present-day
Straits of Dover/En gl ish Channel (e. g. Brid ges,
1982 ) ;
2 tidal estuary o r embayment (e. g. Johnson &
Levell, 1980) .
These alternatives w ill be considered here in the
light of our observations. Similarly, the uncerta inty
as to whether the breakthrou gh across the London­
Brabant land mass (the 'Bedforsh ire Strait' of
K irkaldy ( 1939)) occurred as a result of southward
or northward extension of a coastal embayment will
also be considered.
0 1km

Lithostratigraphical subdivision and relationships

Fig. 2. Location of the sand pits used in this study and the
line of the cross-section illustrated in Fig. 4.
cThe Woburn Sands comprise up to six facies bodies,
which are readily d istinguished on the basis of their
l ithofac ies characteristics (mainly grain size, com­
southerly expansion throughout the southern part of posit ion , colour, clay content, sed imentary struc­
the North Sea Basin and adjacent areas (e . g. West tures and bioturbation) . Most of these facies bodies
Netherlands, Broad Fourteens and Lower Saxony are separated by major subhorizontal to low-an gle
Basins; Ze igler, 1990) . The precise t iming of full erosion surfaces, and five discontinuity-bounded
marine connection between these basins is uncertain, units have been defined as follows (Table 1 &
but most recent reconstructions show this to have Fi g. 3 ) :
occurred by the late Aptian or early Alb ian, at 1 Transition Series (you n gest) ;
around the time of deposit ion of the Woburn Sands 2 Red Sands;
and the overlyin g Gault. Hence , on a regional scale,
the depositional h istory of the Woburn Sands would
appear to be related to marine transgression that
resulted in the connection of two main intracratonic Fig. 3. (Opposite.) Composite vertical section through the
basins. Woburn Sands.
 LITHOSTRATIGRAPHIC NOMENCLATURE

-�--'·-· -�------,
ENVIRONMENTAL SUMMARY
0- ·-:__-_-
---
--
---
BURROW TYPES

..
c
,.,
"'
:0
:;;: 0 :;
"'
:;
"'
����� Blanket shelf muds

..
:;; (.') (.')
t:=== =
:; - . "Dentatus "--
-'
3:
0
Cl
-:::::-=-
---
-
10 -- - _____/Slow deposition/reworked, transgressive
Carstone and Shanley e:i3>
Limestone Transition Series deposits
-

u "' "���� =

<1>
0:<1>
u
c --- -- · "'"''' =
\���-
<l>u "' Red Sands High energy, ebb dominated channel!
��
rocn
(f)
u
''- -- ...Y.C
=M shoal complex (estuary mouth or open

�\��
<1>
0
a: marine, sea strait environment)
u

''\�
20-
(:==.:r>

�
=

"'
�"'
en�
�u
:=C Silly Beds ==M /Low energy, estuary abandonment!
.
en"'
(f)
''''" ,,,,, transgressive deposits
''-'-'-'->--."."'-
'-':-
c
5 "'
u � J
.0"'
-�
c
"' 30-
�-g (f) Silver
Sands
��
�"'
:;; High energy, ebb dominated, estuary
Q)(f)
a. .2 mouth channel/shoal complex
a. Ui (= ebb-tidal delta environment)
=>
c
"'
"li
"' �-
..:
:;;
a.
u
c
"'
(/)
��
------�-
E
a.
40 - =
-s==:
=> ::l
=
.0
0
3: Heterolithic -
Sands �L
�-
= Low-moderate energy, estuary shoal
� = deposits

--
----�
.. ...
---� !"- .,_. =

"' -- -- ---- �
u �--.1!....,
c
"'
�
"'
(f)
u
c 50 - �
c "'
..}.\.._
5
��
(f) �
.0 c
0 3:
3:
:;;
e
en - ��
-'
3:
0
�-
·-··--�..... �
�
.__ Moderate-high energy, flood dominated,
Orange Sands
-:::'{J! -
channel-fill sands intercalated with tidal

�
shoal deposits
60-
-
� �
�

/// !))})
�--��---""!.

?
70 - Basal transgressive deposits with
== = phosphate nodules and reworked faunas

��[�
UPPER
JURASSIC

J:?/-::-1 Large-scale cross-bedding

EJ . Wavy bedding � Strongly bioturbated

B. Trough cross-bedding
c=J Flaser bedding 6:11 Shells and shell debris

Q Ripple cross-lamination
1:-�;1 Low-angle erosion surfaces Q Plant debris

D Intraclasts
E;J ' Concretions/nodules -& Occasional burrows
22 H. D. Johnson and B.K. Levell

Table 1. Summary of the l ithofacies and reservoir characteristics of the main units within the Woburn Sands

Interval Lithology Physical sedimentary structures Dep. environment

Gault Clay Grey fossiliferous claystones

Muddy shelf.

Transition Iron-cemented pebbly sands (basal beds); glauconitic & phosphatic fine-coarse, partly argillaceous Transgressive lag
Series sands. In-situ lenses of richly fossiliferous limestone (Shenley Lmst.). Reworked clasts of deposits.
iron-cemented sst. (Carstone) & Shenley Lmst. Rapid lateral lithological variations.
Three main types of cross-bedding:

Red Sands
I
0.-":
"//!/1/1/1 l Giant cross-bedding with
avalanche foresets &

Med.-v.coarse sand
Mod.-poorly sorted.
��T infilling, large scours ca 3m
deep and 1OOm wide.
High-energy,
Ferruginous with up
to 20% bv detrital
iron oxide (red
11 �1 Wedge-shaped sets (-2-3m
thick) superimposed on
ebb-dominated
complex
(estuary/embayment
colouration) up to low-angle (4-8°), S-dipping mouth or open marine
2.5% bv heavy
%
;; J surfaces. sea strait).
minerals 100%
sand.
III �:JtYti,;;tff -�-am0.1-4m thick tabular &
J trough cross-bedding.

Silty Beds Grey-green

glauconitic & lignitic Coarse sands have flat bases, large rippled or flat surfaces & internally
(/) clays, silts & f. cross-bedded or horizontally laminated.
Q sands. Minor crs., Transgressive or local
z well sorted sand abandonment
<(
layers & lenses. Fine sands occasionally show low-angle to horizontal lamination but mainly deposit.
(/)
bioturbated.
z
::;:)
a:

Ill
-20-40% sand.

��/-71"lr
Variety of large-scale cross-bedding:
0 Silver Med.-v.crs. sand. Northern area:
;: Sands 1 (J.t,t,l,ffZt:;Z,' Large-scale, low-angle
Well-sorted. Quartz
(2-4°) surfaces separated
arenites. Minor m High-energy
- by o.5-2m thick, tabular
carbonaceous ebb-dominated
cross-bed sets.

��: +m
debris. Locally 'L1 1 n-: embayment mouth
Fa-cemented e.g. 1-3m avalanche-type channel-shoal
clay clasts on Southern area: cross-bedding, partly filling complex (cf ebb-tidal
erosion surfaces). ���urs. Complex low-angle deltas)
100% sand. s

Heterolithic Fine to v.fine mod. Moderate-energy,

sorted sands. tidal shoal deposits
Sands Main structures (in order of decreasing importance) current ripple
Numerous thin clay cross-lamination, trough cross-bedding, scour & fill structures & low-angle within an inner
layers. Intraclasts of cross lamination. Herringbone cross-bed patterns. Abundant clay drapes estuarine/inner
clay & carbonac. embayment
produce wavy & !laser bedding. Large-scale low-angle surfaces (dipping
debris. -90% sand. -40) ?margmal to tidal
channel complex
(=Oranae Sands).
Orange Fine to crs. mod. Large-scale subhorizontal (1) & low-angle (2) erosion surfaces.
Sands sorted sands. High-energy
Scattered quartz
granules, clay
1
D �� 2
flood-dominated tidal
channel-fill sands with
pebbles & clay
drapes. Iron oxide /////////////////////I I � intercalated
moderate-to
cement in liesegang high-energy tidal
rings & around clay Large avalanche foresets fill deep scours (1-5m thick). Low angle surfaces shoal deposits.
deposits. -95-100% separated by cross-laminated, cross-bedded & bioturbated sands. Flaser &
sand. wavy bedding/clay drapes.

3 S ilty Beds; term inology of prev ious workers (Table 1 & Fig. 3 ) ,
4 Silver S ands; inclu d in g the schemes o f W yatt e t al. ( 1 986) and the
5 Oran ge S ands and the Heterolithic S ands more recent form al l ithostratigraphy of Sheph ard­
( Lower Woburn S ands or Brown S ands).
= Thorn et a/. ( 1986) .
This inform al scheme generally follows that of The vertical and l ater al rel ationsh ips of these units
Ben tley ( 1970) and is read il y correl ated w ith the are summar ized in Figs 3 & 4, respectively. The b ase
 Transgressive estuarine sand complex 23

Table 1. (Continued. )

Fauna and biogenic sedimentary structures Pal �'it�fe':.�ent Reservoir Thickness Geometry
characteristics

-?Om
Ammonites, belemnites, bivalves, brachiopods. SEAL max. Sheet-like

g
Abundant ammonites, bivalves, belemnites, astropods POOR to V-POOR
Sheet-like/tabular
& oysters. Partly reworked & phosphatized. henley -partly sealing due to 1-2m
Lmst.= brachipods, echinoids & crustacea. cementing & argill. content

No preserved fauna (?leached) 2 main types of Unimodal to

bioturbation: E-W lenticular
the S-SSE 0->15m
geometry in N
I Intense, small-scale colour mottling (5mm diam.)
-pale core & darker rim. Caused by horizontal burrows GOOD
& resulting in negligible destratificatation. Very locally reduced by
minor Fe cementation.
No shale layers.
(?up Large-scale
II Funnel to v-shaped burrows due to vertical animal Minor reversals
1o10's m southward
escape or sediment collapse/inti II. Large burrows (1O's & herringbone max.) thickening wedge
mm wide, up to -1OOmm high) caused by large bivalve patterns
in S.
or crustacaean.

POOR/SEAL Lenticular
No fauna observed.
Strongly bioturbated throughout- fine sediments g
-drapes irre .
effectively destratified. Isolated structures -thin permeable sands surface of ilver
1-2m Sand.
show dips to S probably laterally
No distinct burrow types. extensive (=Storm
layers) -dissected cut-out
by erosional base
of the Red Beds.

Bimodal-bipolar
Unfossiliferous. VERY GOOD
Negligible bioturbation-rare single clay-lined burrows Tabular within
vertically & laterally 2-15m
(Ophiomorpha) towards the top in some places (e.g. S-SW modes are uniform study area.
New Trees). dominant
(directions of all
major structures)

Unfossiliferous (?leached). Bimodal-bipolar Uncertain-restricted

Extensively bioturbated (ca. 10-50% of primary MODERATE to to E part of study
structures destroyed). Horizontal, slightly sinuous, POOR area.
clay-lined burrows are the most common type & occur WSW mode <25m Possibly interfinger
dominant & ENE -discontinuous shale to W with Orange
mainly in cross-laminated sands. Occasional vertical to
mode slightly layers Sands.
oblique burrows.
subordinate

Unfossiliferous (?leached). Bimodal-bipolar MODERATE to

Moderately to strongly bioturbated (30-50% of primary GOOD Uncertain,
structures destroyed) & variety of burrow types: up to greater N-S
(i) narrow vertical tubes, (ii) sinuous subhorizontal NW mode continuity ct.
dominant with -distinct higher -50m
burrows producing colour mottling. (iii) iron-cemented permeability zones E-W.
vertical to steeply inclined burrows, (iv) subhorizontal minor S-SSW
mode within N-S trending
branching burrows Thalassinoides. (v) v-shaped channels.
burrows.

of the lowest unit, the Orange and Heterolithic either lateral interfingering or erosional contact both
Sands, was not seen but is thought to directly overlie possible. However, facies similarities (discussed
the phosphati c gravels and sands of the fossiliferous later) suggest that the Orange and Heterolithic Sands
basal beds re corded in abandoned pits ( Lamplugh, are probably lateral equivalents.
1922). The relationship between these two lower­ The boundary between the top of the Woburn
most sands has also not been observed directly, with Sands and the overlying Gault is generally poorly
24 H. D. Johnson and B. K. Levell

D Gault Cia'{
D Transition Series
[(g);j Red Sands (RS)
�j:j:j:j Silty Beds
- Silver Sands (SS)
� Heterolithic Sands (HS)
0 2km
EO(,<�J Orange Sands (OS)
w

D Jurassic clays North

..,.
-- Approx. depth of
present-day
exposures

10 E:
"
0 20 )§
·t
30 <.)
10 Q)
<f)
"'
40.0
20 ::
0
50 a;
.0
30 <f)

60 �
Q)

40 70::2:

80
50

60

Fig. 4. Cross-section through the Woburn Sands illustrating the vertical and lateral relationships between the main
lithostratigraphical units (sec Fig. 2 for location) . The locations at which some of the main lithostratigraphical boundaries
can be seen arc indicated by single vertical lines. Note the dashed line indicating the approximate depth of present-day
exposures. Datum is the base of the cristatum subzone.

exposed and has not been studied here in any detail . SEDIMENTOLOGICAL
However, this impo rtant and richly fossilife ro us CHARACTERISTICS
inte rval has been studied e xtensively in the past by
palaeontologists and biostratigraphe rs, who have This section outlines in detail the sedimentological
meas ured many detailed vertical profiles (e.g. cha racteristics of the six main facies types. The
Lamplugh, 1 922; Wright & W right, 19 47; Casey, key points of description and interp retation are
19 61 ; Owen, 1 9 72). These data have been incor­ summarized in Table l .
porated into Table 1 and Figs 3 & 4.
 Transgressive estuarine sand complex 25

A B

BOREAL SEA
BOREAL SEA

:r:

,'

� Main palaeocurrent
directions

50 100km
�---

Fig. 5. Three schematic palaeogeographical maps

illustrating the transgressive history of the Lower
Cretaceous in the southern North Sea- English Channel
area, and the evolution of the 'Bedfordshire Strait' which
ultimately connected the Boreal Sea and the Wealden
Basin. Aptian-Albian outcrop shown in black
(a) Ryazanian-Valanginian; (b) Aptian: Woburn sands,
Folkcstonc sands, Hythe and Sandgate beds; (c) Albian, '----'"'---..J100km
Gault clay. (Based on Ziegler, 1988, 1990.)
26 H. D . Johnson and B. K. Levell

I Subhorizontal erosion surfaces a re essentially flat

Orange Sands
and extend up to 200 m. Lo cally they cut down in
con cave-upward scours 4- 6 m deep ( Fig. 6) . The
D_escription
e rosion surfaces are spa ced 5 - 10 m apart verti cally
Based on the available exposures at the time of our and a re normally overlain by coa rse lags of granules
field-work in 1 9 79 (B ryant's Lane, Stone Lane and and mud flakes.
Sheepcott quarries; Fig. 2), these sands a re moder­ 2 Low-angle erosion surfaces o ccur within the units
ately sorted, fine- to coa rse-grained and contain bounded by e rosion surfaces of type I and pass
some quartz granules, clay pebbles and wood frag­ laterally and down-dip into the subhorizontal e rosion
ments. The o range colour is due to widespread iron surfa ces ( Fig. 6) . They are spaced at intervals of a
oxide, whi ch o ccurs as a cement, in Liesegang rings few decimetres to 1 m and separate inte rvals with a
and in rims around clay (e.g. clay pebbles, drapes variable array of cross-bedding, cross-lamination and
and burrow linings). burrows ( Fig. 7) .
Large-scale erosion surfaces within the unit have The more deeply e rosive parts of the subhorizontal
been divided a rbitra rily into two types: erosion surfaces a re overlain by 1 -5 m thick tabular

NE SW
f<---- Flood-dominated tidal channel Tidal bar -----+1
metres
0

. Transport toN Photograph location

+----- 150metres--------•

Fig. 6. An example of large-scale facies relationships in a flood tidal-channel complex in the Orange Sands. The base of the
channel is a horizontal erosion surface lined with intraformational clay clasts. A series of low-angle (4-8°) erosion surfaces
(right side of the photograph) mark the flanks of a tidal bar, which is characterized by small-scale cross-bedding and
moderate to strong bioturbation . The inclinations of the low-angle erosion surfaces (inferred bar flank surfaces) increase
laterally (to the left) and eventually pass into high-energy, channel-fill deposits displaying tabular avalanche foresets up to
4 m high. A second flood-tidal channel sequence is also exposed in the lower part of the section. The simplified field sketch
(sec Fig. 3 for legend) shows the broader relations between the bioturbated tidal bar sands and the avalanche foresets of
flood-dominated, tidal channel-fill sands (from Bryant's Lane pit) .
 Transgressive estuarine sand complex 27

Fig. 7 . Physical and biogenic sedimentary structures i n the Orange Sands. (A) Tidal bar deposits comprising small- to
moderate-scale cross-bedding (10-SOcm thick) separated by horizontal and low-angle erosion surfaces with thin
intraformational mud-flake conglomerates, and occasional clay layers. The numerous low-angle reactivation surfaces give a
characteristic wedge-shaped appearance to the cross-bed sets (see also ( D ) ) . ( B ) Close-up of the central part of
(A) illustrating some details of the bioturbation. Note in particular the simple vertical burrows and a large V-shapcd
burrow (lower centre of photo) . (C) Large-scale tabular cross-bedding infilling a flood tidal channel (upper half of photo) .
Vertical burrows increase in density in the deeper part o f the channel and arc inclined perpendicular t o the forcscts. The
underlying deposits display oppositely-dipping cross-bedding, horizontal and inclined erosion surfaces and moderate
bioturbation. (D) Wedge-shaped cross-bedding with numerous reactivation surfaces, which arc occasionally overlain by
clay drapes or oppositely-dipping cross-lamination.

or wedge-shaped c ross-bedding with NE-dipping Burrowing has destroyed , on average, some 30-
avalanche foresets separated by large-scale , low­ 50% of the p rimary structures and has been sub­
angle erosion surfaces (Fig. 6) . These low-angle divided into five types:
e rosion surfaces may flatten up-dip into more closely 1 Na rrow ( c. 2 mm) vertical tubes that form a
spaced, subhorizontal e rosion surfaces. This is ac­ branching network with sections 10 -20 mm long.
companied by a change in sedimentary structures These tubes have no clay lining and a re extremely
from 1-5-m-thick avalanche cross-bedding to 0.2- fragile, being visible only on wind-sculpted faces.
0. 7-m-thick sets of t rough cross-bedding, cu rrent They resemble burrows p roduced by polycheate
ripple cross-lamination and flaser and wavy bedding. worms in modern sands of estuaries and tidal flats
Palaeocurrent directions from all these deposits a re (Schafer, 1 9 72 ) .
variable, but there is abundant evidence of reve rsals , 2 Sinuous, subho rizontal burrows p roducing c.
especially in the smaller-scale structures. The larger 5-mm-diameter colour mottling. These burrows
structures show mainly northwest-flowing palaeo­ occur mainly in the ripple-laminated sands.
cu rrents but with clear, subordinate reve rsals 3 Simple , ve rtical or steeply inclined tubes ( c. lO mm
(Fig. 8). diameter and 50-200 mm long) with clay linings,
28 H. D. Johnson and B.K. Levell

N Interpretation
I The interbedding of deposits wi th opposed palaeo­
N current modes, evidence of rapid lateral variation in
I flow regime (as shown by the intercalation of large­
and small-scale structures), the range of burrow
types , and the clay drapes, are all common character­
istics of high-energy, shallow-water tidal deposits
(de Raaf & Boersma , 197 1 ; N io & Yang, 1 99 1 ) . The
subhorizontal erosion surfaces, therefore, probably
define the bases of tidal channels and, at least in
the areas of deepest scour, these channels carried
Silver Sands northwestward flowing water. The low-angle erosion
Red Sands
n=157 surfaces are interpre ted as the accretionary flanks of
n=89
in-channel bars on which curren t dominance was
less pronounced and low-energy structures were
preserved (e.g. Yang & N io, 1989 ) . There is no
evidence that the preserved portions of these
bars were e ither emergent or suffered severe wave
activity . The Orange Sands thus represent a h igh­
energy, tidal channel complex w ith mutually evasive
ebb and flood tidal currents.

Heterolithic Sands

Description
This unit comprises moderately sorted, fine to
Heterolithic Sands very fine grained sands w ith numerous clay layers,
Orange Sands (scale x2) scattered clay flakes and woody detritus. The sands
n=203 n=52
are largely grey, while the clay drapes and surround­
ing sands are some times rust-coloured due to iron
Fig. 8. Palaeocurrent distributions based mainly on l arge­
scale cross-bedding. oxides.
The sands contain low-angle ( c. 4 °) erosion sur­
faces tens of metres long that closely resemble those
wh ich are frequently the sites of iron-oxide precipi­ of the Orange Sands, with the exception that they do
tation (Fig. 7A & B) . Th is type projects normal to not pass downwards into channel-fill facies and are
the bedding even when this is inclined (Fig. 7 C) , and often overlain by relatively continuous clay drapes.
is w idespread throughout the Orange Sands. The main sedimentary s tructures are, in order
4 Complex, subhorizontal to inclined, branching of decreasing importance: current ripple cross­
burrow networks c. 10-40 mm in diameter and lamination (Fig. 9A) , trough cross-bedding, scour­
w ith enlarged, bulbous junctions (Fig. 7D). Iron and-fill s tructures and low-angle cross-lamination.
cementation preserves these in three dimensions. Clay drapes occur on set boundaries and foresets
This type most closely resembles the crustacean w ith in all types of cross-stratification and sometimes
burrow system Ophiomorpha. produce wavy and flaser bedding (Fig. 9A & B).
5 Nested cone-shaped burrows (V-shaped in two Although not measured in detail , clay drape d istr i­
dimensions, Fig. 7 B) . These appear to represent the bution is suggestive of tidal bundles, possibly w ith
collapse of sedimentary lamination i n to 20-30-mm­ neap-spring tide cycles (e.g. Visser, 1980). Evidence
w ide horizontal tubes, but may also occur as iron­ of bidirectional currents is ubiquitous in all these
cemented , V-shaped laminae. They are especially s tructures (e.g. Fig. 9A) . The larger foresets
common in sands just above and below major erosion commonly have superimposed smaller sets w ith
surfaces. reversed dips . Cross-stratification type varies over
 Transgressive estuarine sand complex 29

Fig. 9. Physical sedimentary

structures in the Heterolithic Sands.
(A) Flaser bedding associated with
small-scale, current ripple cross­
lamination. Herringbone patterns
are occasionally developed but
normally the southwest ebb­
direction is dominant. (B) Clay­
draped foresets separated by thick
clay layers, which are internally
disrupted by bioturbation.

short distances, both laterally and vertically, and

Interpretation
there are no progressive vertical changes in either
set thickness or grain size. Palaeocurren ts are The bimodal-bipolar palaeocurrent pattern , and the
bimodal-bipolar, with a dominant southwest directed assemblage and variability of sedimentary structures
mode (Fig. 8). in this unit suggest a shallow-marine tidal origin .
Burrowing, which has destroyed around 10-50% The thickness o f this facies (up t o 25 m was recorded
of the primary sedimentary fabric, is dominated by Bentley ( 1970)) and the lack of features rep­
by sinuous, horizontal, clay-lined forms (Fig. 10). resenting emergence (rootlets, wave-reworked sur­
Burrows are most common in the flaser and wavy faces, desiccation cracks, etc . ) suggest deposition in
bedded subfacies and may be virtually absent in the a subtidal environment of moderate water depth
larger decimetre-scale cross-bed sets. The pre­ and fluctuating flow conditions. The lack of distinct
dominance of horizon tal burrows results in little channel-fill facies and the ubiqui tous presence of
disturbance of the sedimentary structures. Less low-angle inclined erosion surfaces wi th a constan t
common types are vertical to oblique and rare , south-eastward dip suggests accretion on a broad
spiral clay-lined burrows (Fig. lOD). subtidal shoal. Relatively low-energy curren ts are
30 H.D. Johnson and B.K. Levell

Fig. 10. Biogenic sedimentary structures i n the Heterolithic Sands. (A) Strong bioturbation i n current ripple cross­
laminated and partly ftaser bedded sands. The dominant biogenic structures are horizontal, clay-lined burrows.
(B) Moderate bioturbation mainly by horizontal, clay-lined burrows with a single inclined burrow (southeast of lens cap) .
Bidirectional, cross-lamination with occasional clay-ftasers is still well-preserved. (C) Plan view of the dominant burrow
type in this facies comprising horizontal, sinuous, clay-lined burrows with back-fill laminae. (D) Isolated example of a
vertical, clay-lined burrow resembling Ophiomorpha. Background facies is ripple laminated, ftaser bedded sand.

suggested by the small scale of cross-bedding and In the northwest of the area (around Heath and
the predominance of r ipple cross-lamination , Reach, Fig. 2) several pits expose up to 20 m of
whereas the extensive clay drapes suggest relatively S ilver Sands, and major low-angle (2-4° apparent
long periods of quiet water conditions. dips) erosion surfaces w ith a constant southwestward
dip can be p icked out throughout the u n it (Fig. llA ) .
Between these planar to slightly undulose erosion
Silver Sands
surfaces tabular cross-bed sets from 0.5 to 2 m thick
occur (Fig. 1 1B ) . The erosion surfaces terminate
Description
abruptly down-dip, resulting in thickening of some
The Silver Sands consist of well-sorted , medium- to cross-bed sets, and formation of hanging set bound­
very coarse-grained quartz arenites (previously used aries. No s ingle surface could be traced from the
as glass sands ) . Sooty and woody carbonaceous top of the 1 5 -m -thick unit to the base. There is
matter is locally abundant, but clay drapes are only occasional evidence of reversing palaeo­
absent. These sands truncate all earlier deposits w ith currents, such as at the base of the unit in Munday's
a major planar to regionally concave -upwards H ill quarry, resulting in an overwhelmingly domi­
erosion surface which is lined with granules and nant southwestward dip to all scales of forese t
clay flakes. The lag deposit overlying this erosion (Figs 8 & 1 1) .
surface is well cemented by iron oxides. I n the extreme north o f the area o f the Silver
 Transgressive estuarine sand complex 31

Fig. 11. Physical sedimentary structures i n the Silver Sands from the northern part of the study area. (A) A complete
vertical section through the Silver Sands ( c. 15 m thick) illustrating a sequence of moderate- to l arge-scale cross-bedding
(set thickness c. 0.5-2.0 m ) separated by undulatory horizontal to low-angle erosion surfaces (dipping to right ) . Cross­
bedding is mainly undirectional to the southwest (ebb-sets) . A prominent erosion surface lined with intraformational clay
clasts and quartz granules forms a ledge j ust below the contact between the dark and light coloured sands. The dark sands
represent ebb-dominated channel-fill sands within the upper part of the Orange Sands. (B) Closer view of the southwest
directed, ebb-dominated Silver Sands (quarry face c. 10 m high ) . Note the undulatory nature of the erosion surfaces and
corresponding irregularities in cross-bed set thickness (c. 0.5-2 m ) . There are no positive signs of reversing currents in
these ebb-dominated sands. The ledges produced along the various erosion surfaces are not lined with clay; these sections
are in c. 100% quartz sand.

Sands are somewhat thinner and more iron-stained. re-exposed recently ( J . Eyers, 1994 pers. comm . ) .
Cross-bedding is predominantly of the avalanche Internally these sets show complex lateral transitions
type in se ts 1 - 2 m thick . Three-dimensional ex­ from single avalanche se ts, which are a lways directed
posures show these sets to be the infills of spoon­ to the south or southwest (Fig. 8) , and which pass
shaped scours with the appearance of micro-de ltas, into thinner tabular or wedge-shaped sets of up- and
rather than the deposits of migrating large-scale down-slope dipping foresets, separated by low-angle
megaripp les (or sand waves) with flat bases. or subhorizontal erosion surfaces (Fig. 13). These
In the Shenley Hill area further to the east sets have only been seen in sections paralle l to the
(Fig. 2), tabular sets 2-4 m thick comprise the bulk transport direction , and their strike configuration is
of the Silver Sands (Fig. 1 2 ) . These were relative ly unknown. Bentley ( 1970) claimed to be able to map
poorly exposed during our fie ld -work but were the E -W trend of individua l sand-wave crests in this
photographed by Bentley ( 1970) and have been area.
 V.>
N

::t:
\:::)

0'
�
a
Cl
;:
.,
;:
"'­
t:l;)
;>;:
!:""'
"'
"'
"'
:::::

Fig. 12. Large-scale, sand-wave-type cross-bedding in the Silver Sands (Munday ' s Hill Quarry) . Individual sets are up to 4 m thick and are internally complex,
ranging from avalanche foresets to low-angle surfaces (6-8°) separated by upslope- and downslope-dipping cross-bedding (see Fig. 13 for details) . Note also the
sharp, planar contact between the Silver Sands and Silty Beds and the erosional contact between the Silver and Red Sands. The dark area along this contact
represents a ridge of strong iron cementation ('Carstone rib ' ) . Finally the sequence is capped by the marine Gault Clay.
 Transgressive estuarine sand complex 33

Fig. 13. Internal characteristics of the sand-wave-type cross-bedding in the Silver Sands in the Munday 's Hill- Double
Arches area (large-scale structures illustrated in Fig. 12). (A) Tabular cross-bedding with simple, angle-of-repose
foresets. Note small oppositely-dipping set (by lens cap) separating the main , thick sets and a prominent reactivation
surface in the upper thick set (Double Arches Quarry) . (B) Relatively small tabular cross-beds, with each displaying an
upward decrease in grain size (Munday's Hill Quarry) . (C) The low-angle surfaces dipping to the left represent an ebb­
directed (to southwest) sand-wave lee face . Internally these coarse to granule grade sands display upslope dipping cross­
bedding which represents flood-directed megaripples (Munday ' s Hill Quarry ) . (D) Herringbone patterns developed in
relatively small sets of trough cross-bedding (Munday's Hill Quarry ) .

In the extreme west of the area, west of Heath the lack or scarcity of clay drapes and burrows
and Reach (Sheepcott quarry; Fig. 2) , a heterolithic suggest deposition in a higher energy environment
facies of decimetre-scale cross-bedded sands with than both the Heterolithic and Orange Sands. None
persistent clay drapes overlies the Orange Sands. the less, the relatively infrequent evidence of revers­
These beds, formally assigned to the Lower Woburn ing palaeocurrents still suggests tidal deposition.
Sands, could be a lateral facies variant of the Silver The extensive subhorizontal erosion surface at the
Sands. base of the unit and the evidence of a bedform
In general, the Silver Sands are weakly bio­ complex up to 1 5 m thick, at least in the Heath and
turbated, apart from the upper few metres in Reach area, suggest deposition in a high-energy,
some pits (e.g. New Trees) , where clay-lined current-dominated bar system after a period of wide­
Ophiomorpha-type burrows occur. spread erosion. There is no clear channel-fill facies
and it is possible that deposition occurred in an
environment in which interbar depressions were
Interpretation
created by the accretion of shoals rather than the
The textural and mineralogical maturity of the Silver active cutting of channels.
Sands, the uniformly large size of cross-bedding and In more detail, the cross-bedding indicates large ,
34 H. D. Johnson and B.K. Levell

possibly sand-wave-type bedforms with super­

imposed megaripples over the majority of the out­
crop area, and microdelta-like forms infilling scour
hollows in the north. Taken together with a similar
change in cross-bedding style in the younger Red
Sands (see below ) , this change could be related to a
southward increase in water depth across the study
area (Fig. 2 ) , which allowed the construction of
larger bedforms in the south of the area.
In comparison with the shoal complex of the
Heterolithic Sands, current energy was apparently
much higher and there was less time for the settling
of suspended clay. The textural and mineralogical
maturity of the sands also indicates a more intensive
phase of reworking prior to final deposition.

Silty Beds

Description
The Silty Beds (Lamplugh & Walker, 1903;
Lamplugh , 1922) comprises a thin interval (mainly
1 - 2 m ; maximum 3 .6 m) of grey-green, glauconite­
and carbonaceous-rich clays, silts and argillaceous
fine-grained sands with rare thin beds of coarse- to
very coarse-grained, well-sorted sands. The fine­
grained sediments are strongly (90- 100 % ) bio­
turbated (Fig. 14) and have in places suffered
soft-sediment deformation. The coarse beds are
Fig. 14. Silty Beds, overlying the Silver Sands with a
50- 1 00 mm thick , have sharp bases and flat or
sharp, planar contact (next to trowel ) . The mottled texture
rippled tops, and are internally cross-laminated or in the Silty Beds is the .result of extreme bioturbation. A
plane-laminated. The unit sharply overlies a slightly low-angle laminated sand bed with an erosional base
undulating surface on top of the Silver Sands (Fig. (storm layer) is present in the upper part of the sequence
12). This undulose surface was considered by (Bryant's Lane Quarry) .
Lamplugh ( 1922) and Bentley ( 1970) to represent
the preserved morphology of shoal or sand-wave
crests in the underlying Silver Sands. The Silty Beds subtidal sand shoal and channel complexes described
are sharply overlain by either the Transition Series from the underlying deposits.
or the Red Sands (Fig. 4). The precise depositional environment of this unit
is debatable. The lack of marine fauna and abundant
carbonaceous material could indicate restricted
Interpretation
marine conditions, possibly within a deeper, pro­
This interval represents a period of slow sedimen­ tected part of an estuary ( Lamplugh, 1922). The
tation in a marine environment in which both current facies could also be interpreted as tidal flat deposits
and wave energy were apparently too weak to cause (Eyers, 1992b) , although the absence of tidal
significant erosion of the unconsolidated sands on channels and the characteristic fining upward
the sea-floor. The coarser grained sand layers are sequence capped by in situ marsh deposits (e.g.
interpreted as lags, resulting from reworking and Evans, 1965) fails to support this. Alternatively, the
winnowing of the Silver Sands (cf. Levell, 1980). abrupt cessation of high-energy conditions marked
The Silty Beds are interpreted to represent a phase by the sharp base of this unit, and the presence of
of low energy and low sediment influx (an estuarine reworked and glauconite-rich horizons within it, is
abandonment phase) relative to the high-energy, more characteristic of a sudden deepening, rather
 Transgressive estuarine sand complex 35

than shallowing in water depth. Hence , a more these tabular or wedge-shaped bundles is in turn
offshore shallow marine environmen t is a further subdivided by some times i n tersecting minor dis­
possibility . continuity planes that define the individual foresets.
The Silty Beds probably represen t a relatively The major reactivation surfaces are convex-up,
long time interval, which marks the end of the flattening toward the top of the set, where they are
underlying Orange- Heteroli thic and Silver Sands separated by cosets or sol i tary sets of decimetre­
depositional succession. This i n terpreta tion implies scale cross-bedding. At the top of the sets, where
that the overlying Red Sands form a genetically the reactivation surfaces flatten out, there are some
separate sand body. up-dip thinning sets 0.05 - 0.2 m thick of oppositely
dipping cross-bedding. This geometry closely
resembles, on a large scale, the ebb and flood shields
Red Sands .
described from several modern tidal environments
(e.g. Klein , 1970) .
Description
The unit comprises moderately to poorly sorted, Type ll is characterized by wedge-shaped cross-sets
medium- to very coarse-grained sands (occasionally (0.3- 2 m thick) with sou th- or southwest-dipping
gravelly), which are distinguished from all older foresets, which are superimposed on 2-4-m-thick
units of the Woburn Sands by their high conte n t (up sets of low-angle southward-dipping erosion surfaces
to 20% bulk volume) of detri tal iron oxide in the (dips 4-8°) . Individual cross-sets taper either up or
form of rust-coloured ooliths of goeth i te and angular down the inclined surfaces. Small sets (less than
ir9nstone chips. It is also rich in heavy minerals 0.5 m thick) occasionally thicken down-dip into
( Bentley, ( 1970) reports up to 2 . 5 % bulk volume) . single large-scale ( c. 2-3 m thick) avalanche sets.
I n addition, diagene tic iron oxide occurs a s a local Type II cross-bedding passes laterally i n to type I in
cement around mud clasts and on certain foresets. both down- and up-palaeocurren t direction .
The Red Sands erosionally truncate all earlier
deposits. In the Shenley Hill area they were shown Type III comprises 0. 1 -0.4-m-thick sets o f trough
by Bentley ( 1970) to occupy E- W trending shallow and tabular cross-bedding bounded by subhorizontal
scours up to 6 m deep and 100 m wide. South of erosion surfaces, and forms the bulk of the Red
this area the Silty Beds and the Silver Sands are Sands in the channelized facies of the northern area.
progressively cu t out by erosion beneath the sou th­
ward thickening Red Sands until, south of Leighton Additional information on the complex and varied
Buzzard, the sand pits expose only this last unit geometry of these sedimen tary structures has been
(Fig. 4), which is presumed to overlie Orange and/ obtained recently from ground-penetrating radar
or Heterolithic Sands. The erosion surface at the studies ( Bristow, in press) .
base of the Red Sands is i n terpreted as a sequence The Red Sands are extensively bioturbated (Fig.
boundary by Ruffell and Wach (in press). 16). The most widespread burrow type is a c. 5-mm­
Three main types of cross-bedding characterize diameter colour mottling in which pale-coloured
the Red Sands. burrow fills contrast with the darker iron-rich sands
surrounding them (Fig. 16A) . This structure results
Type I consists of thick (up to 5 . 5 m) sets of from horizontal burrowing and has caused only
avalanche cross-bedding that overlie prominent minimal disruption of the primary stratification. Such
erosional scours up to 3 m deep and more than 100 m bioturbation is apparently similar to that termed
wide . The overall set geome try thus resembles giant­ 'cryptobioturbation' by Howard & Frey ( 1975) . This
scale trough cross-bedding. The uppermost se t of structure was found by them in the G eorgia (USA)
the unit is overlain by a coset of successively smaller estuaries and a ttributed by them to amphipod
se ts with a particularly i n triguing geometry (Fig. crustaceans. The colour mottling could be due either
15). The thick set of avalanche foresets at the base to a slight grain-size fractionation ( the iron pellets
dip at 25 -30° and abut sharply against a basal fall mostly in the fine- to very fine-grained sand
erosion surface locally lined with granules and classes ( Bentley, 1970) or to some oxidation/
pebbles. Major reactivation surfaces periodically reduction difference caused by organic slime.
separate bundles of foreset laminae with an average The second and most distinctive burrow type
thickness of 130 mm (range 50-220 mm) . Each of is the nested inverted cone or funnel-shaped struc-
36 H. D. Johnson and B. K. Levell

Fig. 15. Large-scale (c. 3-4 m thick ) , ebb-directed (to south) , avalanche foresets and associated top set deposits.,.(A)
Foreset packets are separated by major reactivation surfaces. The latter are ascribed to stronger than normal flood-currents
because they are overlain by packets of oppositely dipping, wedge-shaped cross-beds. The large-scale structure is
interpreted as an ebb-dominated bedform which filled an erosional hollow. (B) Close-up of central part of (a) showing
oppositely-dipping (to northeast), flood-directed sets, possibly analogous to flood-shield bedforms. Locality: Pratts' Lane
Quarry.

ture (Fig. 1 6 B) . The V-shaped laminae (in two A third burrow type consists of horizontal 40-mm­
dimensions) of these structures could have formed diameter subhorizontal tubes with miniscus-shaped
either by sediment collapse i n to the underlying spreiten. The size, shape and back-filling spreiten
horizontal tubes, which are sometimes associated resemble the 'press structures' produced by £chino­
with these structures, or could have formed as an cardium cordatum ( Reineck & Singh , 1973). If these
escape structure by upward movement of an animal burrows are echinoid burrows, they would indicate
(e.g. large pelecypod or crustacean). The escape fully marine conditions.
s tructure interpretation may be supported by the
association of this sort of burrow with large erosion
Interpretation
surfaces overlain by migrating bedforms. Perhaps
the animals preferred the relatively slow deposition The Red Sands accumulated in a high-energy marine
areas of channel floors but were none the less able to or marginal marine environment dominated by
escape by efficient vertical burrowing when buried southward flowing, but sometimes reversing, cur­
beneath advancing bedforms. rents (Fig. 8) . There is no evidence of distinct
 Transgressive estuarine sand complex 37

Fig. 16. Biogenic sedimentary

structures in the Red Sands.
(A) Typical mottled appearance of
the Red Sands resulting from
horizontal burrows. Note the
alternation of non-bioturbated
laminae and the retention of
structure within the bioturbated
layers. Locality: Pratt's Lane.
(B) Two V-shaped burrows
occurring within a large-scale cross­
bedded sand body at the contact
between two different sediment
types (coarser, massive sands
above) . The lower cross-bedded
sands also display the mottled
texture detailed in (a) . Locality:
Grovebury South.

channel-fill or shoal deposits, and the whole uni t

Transition Series
seems to represent rather a blanke t-like field of
megaripples and megarippled sand waves (cf. J . R.L.
Description
Allen, 1980) , which increased systematically in
height southwards. The sedimentary s tructures The Transition Series comprises a d istinctive thin
closely resemble tidal sand waves found in modern (c. 1 -2 m thick) , and laterally extensive deposit at
subtidal environments (e.g. Berne et a!. , 199 1 ) . the junction between the previously described units
There i s no evidence of emergence o r s ignificant and the overlying Gault. I t has been studied inten­
wave activity , but the b ioturbation suggests similar sively by palaeontologists because of its rich fauna
marine or marginal marine cond itions to the under­ (for more extensive descriptions see Lamplugh
lying units. The lack of clay drapes indicates a more ( 1922 ) , Wright & Wright ( 1947 ) , Hancock ( 1958) ,
or less constantly active water column . Casey ( 1961 ) and Owen ( 1972)) .
Lithologically the Transition Series is laterally
variable , consisting mainly of partly cemented pebbly
38 H. D. Johnson and B.K. Levell

sands (at the base) and argillaceous, glauconitic and or because these flows reach maximum velocities at
phosphatic fine- to coarse-grained sands and sandy times of lower water, and are thus preferentially
clays. A distinctive and unusual limestone, the preserved in the topographically lower parts of the
Shenley Limestone, occurs as discontinuous in situ deposit. The predominance of southwestward­
bands, up to c. 0.6 m thick , and as reworked frag­ directed palaeocurrents leads to the proposal of
ments in the lower part of this unit. This limestone, open sea in the southwest.
which sharply overlies the underlying deposits, 3 The increase in set size southwards in the Red
contains a unique fauna, including rich and dis­ Sands (and possibly in the Silver Sands) , coupled
tinctive brachiopods and a unique assemblage of with a southward increase in the thickness of the
echinoids and Crustacea. The whole Transition Red Sands, could be explained by greater water
Series is highly fossiliferous, with abundant ammon­ depth in the south. Water depth is, of course , of
ites, bivalves, belemnites, gastropods and oysters. especial importance in the preservation of sets up to
There is abundant evidence of reworking and dis­ 5 . 5 m thick. In modern tidal basins (e.g. estuaries
continuous sedimentation in the form of reworked and embayments) the largest bedforms frequently
faunas and several condensed horizons. The Tran­ occur in the deeper areas (see Fig. 17).
sition Series forms a laterally extensive drape over 4 The geometry of the deposit is one of southward­
the underlying Woburn Sands and extends laterally dipping facies units (Fig. 4) . This is most readily
over Jurassic sediments (Owen, 1972) . explained by a northward transgression .
Two factors lead us to suggest that the Red Sands
should be considered separately from the older sub­
Interpretation
divisions of the Woburn Sands.
The Transition Series represents a composite trans­ 1 They rest erosively on the Silty Beds, which
gressive , shallow-marine lag, which forms a type of appears to be a slowly deposited unit representing
abandonment deposit on top of the underlying, the cessation of sand supply to the underlying Silver
mainly high-energy, subtidal deposits. The rework­ Sands shoal system.
ing of limestone beds, concretions and iron­ 2 Their distinctive petrography, rich in detrital iron­
cemented sandstones ( ' Carstone'), mixing of faunal oxide fragments with goethite ooliths, suggests
zones and condensed intervals indicate slow depo­ a change in the source material. The most likely
sition and strong current activity. The abundant possibility is the erosion of presently unexposed
and highly varied fauna supports a shallow marine, iron-rich intervals derived from further updip to the
probably offshore, depositional environment, with north , which may include the iron-stained Orange
the coarser grained layers representing winnowed Sands, the carstones of the Silver Sands or some
lag deposits. other, older or laterally equivalent formation(s)
(e.g. the early Cretaceous Carstone of the Norfolk
area, Casey & Gallois, 1 973). The only reported
PALAEOGEOGRAPHICAL EVOLUTION occurrence of similar goethite ooliths is from an
early Cretaceous oolith-rich sand unit further north
Reconstruction of the palaeogeographical evolution (Deepdale Pit), which may be equivalent to the
of the five erosionally bounded units depends not Silver Sands (Eyers, 1992a) .
only on the process interpretations given above but The deposition of the Woburn Sands began with
also on the overall palaeogeographical setting and the partly erosional and partly tectonic creation of a
the stratigraphical relationships between the units. trough trending NE-SW through the Bedfordshire
The first point to establish is the direction of the area. The basal deposits are gravels and phosphates
transgression . Several factors lead us to propose a with rolled, derived faunas and abundant wood
northward transgression . material, suggesting transgressive coastal erosion
1 The affinities of the fauna described by Casey (whether open-marine or on the margins of a
( 196 1 ) from the lower part of the complex are with brackish embayment is unknown).
faunas from the southern Wealden Basin, rather The overlying Orange and Heterolithic Sands
than from the northern Southern North Sea Basin. represent subtidal channel and shoal deposits,
2 In tidal embayments it is common for ebb-directed respectively. The relationship between these two
sets to be preferentially preserved, either because units could be either lateral interfingering or a
ebb flows are often enhanced by freshwater run-off sharp erosional contact (Fig. 17). l n the first case
 Transgressive estuarine sand complex 39

�� Tidal channel fills

} Orange

I . I High-energy tidal shoals

Sands

C:=J Moderate-energy tidal shoals } Heterolithic

Sands

b) Model 1
CHANNEL �
HIGH ENERGY CHANNE HIGH ENERGY MODERATE ENERGY
SHO L "'- L SHOAL SHOAL
A :, ::-:=--. ·:-::::_�
::::
__
-:-::=· ::- :: -_-.-_. -.::-:.: ::���-::-.�
.•.•.. •
.
.
-::-::-.·-. .�- �=._ � :s_ �:::. _.:.:.:::= � �;=-; ;�;�
::::::::::-::=-
·- - -. . .
. '\ - =-== =
. .� : - .:0 - _
;- _
�;:.;_
---
--=-
=_==
-==;::;===;;:::_::::
:: :;: B

5m
c) Model 2
A B

Fig.17. (a) Distribution of the Orange and Heterolithic Sands (Lower Woburn Sands or Brown Sands) below the Silver
Sands erosion surface, and (b) two possible models depicting possible lateral relationships (model 1 preferred) .
40 H. D. Johnson and B.K. Levell

(Fig. 17b) , the Heterolithic Sands would be situated sea-level rise leading to the area becoming an open­
towards the centre of the estuary or embayment, marine shelf rather than an estuary mouth. However,
with the Orange Sands being a marginal , flood­ alternative interpretations consider this interval to
dominated , channel complex. Such a palaeo­ represent tidal flat deposits and hence a relative
geography with marginal channels hugging the edges shallowing of sea-level of the facies succession has
of an embayment or estuary would be consistent been proposed (Eyers, 1992b) .
with the generally rotary circulation models for In gross facies terms (e.g. the style of cross­
estuaries, where tidal currents are deflected by bedding, grain size and lack of clay drapes) the Red
Coriolis force to the margins (e.g. Schubel , 197 1 ) . Sands closely resemble the Silver Sands. However,
Nevertheless, a problem remains with this inter­ there is a major difference in their petrography. The
pretation in that it is difficult to understand why the Red Sands could, therefore , represent a similar outer
Heterolithic Sands shoal complex accreted steadily estuarine shoal complex but formed at a time when
to a thickness of 25 m , without being dissected by a the up-dip tidal channels were eroding different
migrating channel system. However, the similarity material. They would therefore represent the shift
between the Heterolithic and Orange Sands leads of tidal channels to debouch again in the Leighton
us to infer a laterally interfingering relationship Buzzard Area. An alternative and more dramatic
(Fig. 17b). interpretation would relate the Red Sands to the
Two features place these two units in the relatively final breaching of the London-Brabant land mass
landward and protected parts of the embayment or and thus to continued coastal erosion at the margin
estuary, rather than on the open shelf. of the transgressing embayment. As discussed above,
1 The clear presence of distinct channel and shoal the erosion surface at the base of the Red Sands has
complexes is characteristic of the estuary proper, been interpreted as evidence for a sea-level low­
rather than the seaward extensions. Furthermore stand (sequence boundary) punctuating the overall
the flood dominance of the Orange Sands channels late Aptian transgression. This explanation would
indicates mutually evasive ebb and flood channels. presumably equate the depositional settings of the
2 The well-developed clay drapes indicate periods Silver Sands and Red Sands, with the Silty Beds
of slack water in a generally current-dominated representing either the most distal (offshore) or
regime. Whether these are diurnal drapes or not, most proximal (tidal flat) depositional settings.
periods of tidal slack water are more likely to occur On the grounds of the facies differences we have
in systems with rectilinear tides. Such tides are more described , and of simplicity, we prefer to view the
common in the inner and more restricted parts of succession as one of continued overall transgression
estuaries, whereas in estuary mouths and on the and attribute the baSal erosion surface of the Red
open shelf, tides tend to be rotary (Terwindt, 1973). Sands to changes in tidal dynamics, possibly associ­
The major erosion surface at the base of the Silver ated with the breaching of the Lori don - Brabant
Sands indicates a shift in depositional setting, and is land mass.
explained most readily by the lateral shift of a major, The Transition Series marks the final abandon­
wide, estuary mouth channel system. The overlying ment of the high-energy, tide-dominated sanely shoal
Silver Sands, with the lack of a distinct channel-fill complexes. Sedimentation was slow, discontinuous
facies and the large-scale bedform complexes up to and interspersed with periods of physical reworking.
15 m high , suggest, in the context of a continuing Subsequent basin deepening resulted in the wide­
northward transgression, deposition in estuary spread deposition of shallow-marine/shelf muds
mouth subtidal shoals. Modern analogues might be (the Gault) .
the outer reaches of Chesapeake Bay ( Ludwick 1970,
1974; Colman eta/. , 1988), or Delaware Bay ( Knebel
et al . , 1988) , or the seaward portions of ebb-tidal DEPOSITIONAL MODEL
deltas (Greer, 1975 ; Hayes, 1975 ) . FOR A TRANSGRESSI V E
A s mentioned above , the Silty Beds represent the ESTUARINE - EMBAYMENT SYSTEM
cessation of sand supply to the Silver Sands shoal,
followed by a period of slow deposition. Reasons for A generalized depositional model can be constructed
the abandonment of this shoal could be autocyclic for a transgressive estuarine -embayment system
(e.g. a shift in the shoal pattern related to the based on observations from both the Woburn Sands
movement of main tidal channels) or allocyclic (e.g. and from known facies and bedform distributions
 Transgressive estuarine sand complex 41

within modern meso- and macrotidal estuaries and 4 Reservoir quality and heterogeneity will be
embayments (Fig. 18). This model also can be strongly influenced by the contrast between channel
compared with estuarine facies models developed and shoal deposits. In general, the channel fills may
mainly from a synthesis of modern environments be expected to form discrete zones of higher per­
and from conceptual considerations (Dalrymple meability (Fig. 18b), whereas the tidal shoals will
et al., 1 99 1 ) . comprise variable but generally lower quality and
The essential elements o f the inner estuarine more heterogeneous reservoirs ( c.f. Orange and
environment are as follows. Heterolithic Sands).
1 Discrete , mutually evasive, ebb and flood tidal The essential features of the outer estuarine environ­
channels and interchannel tidal shoals. ment are as follows.
2 Tidal channel sands will be relatively coarse-grained 1 Minor lithological differentiation between tidal
and characterized by large-scale cross-bedding. channel and shoal deposits.
3 Tidal shoals will consist of more heterolithic sands, 2 Well-sorted sands with fewer clay intercalations
generally with smaller scale structures, stronger bio­ and less bioturbation compared with the inner
turbation and more frequent and laterally extensive estuarine sands.
clay layers than the higher energy channel-fill 3 Large-scale cross-bedding, partly reflecting large
deposits. estuary mouth bars and sand waves, and also

b) Tidal shoals

'

�
3

a) l
Basic Characteristics Woburn S. Equiv.
Fossiliferous clays Gault Clay

Reworked, fossiliferous,
TRANSGRESSIVE LAG Transition Series
partly cemented sands

Well-sorted sands.
Large-scale, complex
OUTER ESTUARINE/
cross-bedding. Silver Sands and
EMBAYMENT TIDAL
Ebb-dominated Red Sands
SHOALS
palaeocurrent directions.
Mainly minor bioturbuation.

Pre-transgressive deposits Upper Jurassic Clays

Fig. 18. A model for a transgressive estuarine- embayment depositional system (see Fig. 3 for legend). (a) Idealized
vertical section through a transgressive estuarine-embayment complex based on the Woburn Sands. (b) Block diagram
illustrating some of the essential differences in sand body characteristics between the inner estuarine/embaymer.t and outer
estuarine/embayment environments. (c) Idealized section through a transgressive estuarine- embayment indicating the
potential stratigraphic trap geometry.
42 H. D. Johnson and B.K. Levell

the infilling 0f deep ebb-dominated channels and allows consideration of their sequence stratigraphical
interbar troughs. implications (e.g. van Wagoner et at. , 1990) .
4 Possibly an overall seaward i ncrease in the size The major third-order lowstand and maximum
of cross-bedding in response to deeper-water flooding events would be recorded respectively by
conditions. the basal lag deposit of the Woburn Sands and by
5 Higher quality and more homogeneous reservoir the condensed horizons in the Lower Gault Clay.
characteristics compared with the inner estuarine We have seen no direct (i.e. preserved) evidence of
sands as a result of both stronger tidal currents and fluvial incision at the base of the Woburn Sands,
increased wave activity. Shale layers are likely to be which comprises a typical shallow-marine lag
very thin and discontinuous, or absent. deposit. The non-preservation of predicted lowstand
The vertical sequence through this type of trans­ fluvial deposits among shallow marine sandstones
gressive sand complex would tend to have lower successions is commonly argued to be the result of
reservoir quality sands in the lower part and higher either (i) minor lowstand fluvial deposition, due to a
reservoir quality sands towards the top (Fig. 1 8a) . predominance of fluvial bypassing during incision,
The transgressive nature of the sequence would lead and/or (ii) extensive shallow marine reworking
ultimately to the blanketing of the overall lenticular during subsequent relative sea-level rise. In many
sand-body complex by shelf muds and could thereby, cases this combination results in the apparent
form a stratigraphical trap (Fig. 1 8c) . merging of lowstand and initial marine flooding sur­
faces, and removal of all lowstand fluvial deposits
(e.g. Bergman & Walker, 1987; Plint & Walker,
DISCU S S ION 1987; Plint & Norris, 1 99 1 ; Thorne & Swift, 199 1 ;
Walker & Bergman, 1992). The greatest amount of
The difficulty of postulating a generalized facies such reworking is likely to occur in tide-dominated
model for this type of succession is that these deposits settings, particularly in confined areas such as
are influenced by a particularly wide range of estuaries or embayments, due to high-energy tidal
physical processes, both autocyclic and allocyclic; currents and their ability to incise deeply into the
they may receive sediment supplied from different substrate. Holocene incised valleys on the inner
sources (landward and/or seaward; Schubel, 197 1 ) continental shelf of southeastern USA , for example,
and their preservation potential i s linked closely to show a hierarchy of erosional events that reflect not
the rate and sense of relative sea-level fluctuations. only earlier fluvial incision but later deep tidal scour
These variables form the basis of Dalrymple et al. 's (c. 15 -30 m ) , particularly at the confluence of tidal
( 199 1 ) classification of estuarine facies, in which streams, in tidal channels associated with tidal flats,
they argue that these deposits are diagnostic of in tidal inlets and in channels associated with ebb
transgressive conditions. In this scheme the Woburn tidal delta shoals (Oertel et a/. , 199 1 ) .
Sands would be classified as a tide-dominated estuary Preservation potential o f transgressive estuarine
in which fluvial processes were negligible . Compar­ successions is relatively high due to their relatively
able modern environments in terms of physical protected position within palaeovalleys (Swift et a/. ,
processes, facies and sedimentary structures would 1980; Demarest & Kraft , 1987). The degree of
include the southwest Netherlands area (e.g. preservation is related most closely to the depth of
Haringvliet and Oosterschelde Estuaries) (Oomkens incision of the basal erosion surface , or sequence­
& Terwindt, 1960; Terwindt, 1973; van den Berg bounding fluvial incision, plus any additional down··
et al. , 1980; Yang & Nio, 1989 ) , Severn Estuary/ cutting by tidal scour associated with the initial
Bristol Channel area, (Hamilton, 1979; Harris flooding surface and , to a lesser extent, the degree
& Collins, 1985 ; J . R . L. Allen, 1 990) , Ossabaw of erosion associated with the ravinement surface
Sound (Greer, 1975 ) , Jade Estuary/German B ight (Dalrymple, 1992). The latter is equivalent to the
( Reineck, 1963; Reineck & Singh, 1973) and Broad ravinement/shoreface erosion surface of wave­
Sound, Australia (Cook & Mayo, 1977). dominated shorelines , but in tide-dominated
The stratigraphical relationships recorded in the systems it is characterized by tidal current erosion
Woburn Sands and summarized in the depositonal (e.g. the bedload parting zones of tidal seas; Stride,
model allows comparison with other stratigraphical 1982 ) . Complete preservation of an idealized tide­
models associated with tide-dominated estuaries, dominated transgressive stratigraphy, which seems
incised valleys and transgressive systems tracts , and to be extremely rare, would predict tidal shelf sand
 Transgressive estuarine sand complex 43

deposits (e.g. tidal sand ridges and tidal sheet sands REFERENCES
with sand waves and associated deposits; Stride,
1982; Nummedal & Swift, 1987; Dalrymple, 1992) ALLEN , J . R . L . ( 1 980) Sand waves: a model of origin and
overlying the ravinement surface . All the main sand internal structures. Sediment. Geol. , 26, 281 - 328.
ALLEN, J . R . L . ( 1 982) Mud drapes in sand wave deposits:
bodies in the Woburn Sands appear to occur below
physical model with application to the Folkestone Beds
the transgressive marine erosion surface and hence (Early Cretaceous, southeast England ) . Proc. R. Soc.
are assigned to the estuarine part of the sequence. London, Ser. A, 306, 291 -345 .
Shelf sand wave deposits may occur further to the ALLEN, J . R. L . ( 1990) The Severn Estuary in southwest
south and southeast within the Folkestone Beds Britain: its retreat under marine transgression, and fine
sediment regime. Sediment. Geol. , 66, 13-28.
(Narayan, 1 97 1 ; J . R. L . Allen, 1 982; Bridges, 1982). ALLEN, P . ( 1981 ) Wealden research - ways ahead. Proc.
The main difference between tidal shelf sands and Geol. Assoc. , 100, 529-564.
tidal estuarine sands is that the latter will be domi­ BANERJEE, I. ( 1991 ) Tidal sand sheets of the Late Albian
nated by complex interfingering of tidal-channel, Joli Fou-Kiowa-Skull Creek marine transgression,
Western Interior Seaway of North America. In: Clastic
tidal-shoal and, in places, tidal-flat deposits (cf.
Tidal Sedimentology (Eds Smith, D .G . , Reinson, G . E . ,
Maguregui & Tyler, 1 99 1 ) . Thus in the Woburn Zaitlin, B .A . & Rahmani, R . A . ) . Can. Soc. petrol.
Sands, concave-upward erosion surfaces and rela­ Geol. , Calgary, Memoir 16, 335-347.
tively rapid lateral thickness changes are common. BENTLEY, A . ( 1 970) Sedimentation studies in the Lower

In more basinward settings individual sand bodies Greensand; a report on the sedimentology of the Aptian
and lower Albian strata near Leighton Buzzard,
are likely to have tabular geometries, with individual Bedfordshire. Unpublished BSc thesis, University of
sand and shale layers having much higher continuity Oxford, 137 pp.
(cf. Surlyk and Noe-Nygaard, 1 99 1 ) . Distinguishing BERGMAN, K.M. & WA LKER , R.G. ( 1987) The importance
between these different types of shallow-marine sand of sea-level fluctuations in the formation of linear con­
glomerate bodies; Carrot Creek Member of Cardium
body is particularly important in evaluating the
Formation, Cretaceous Western Interior Seaway,
nature of such deposits in the subsurface because of Alberta, Canada. J. sediment. Petrol. , 57, 651 - 665 .
their different geometries and lateral extent (e.g. BERNE , S . , DuRAND, J. & W EB ER , 0. ( 1991) Architecture
Banerjee, 199 1 ; Brownridge & Moslow, 199 1 ; Jenette of modern subtidal dunes (sand waves ) , Bay of
et a!. , 1992). Bourgneuf, France. I n : The Three-Dimensional Facies
Architecture of Terrigenous Clastic Sediments and its
Implication for Hydrocarbon Discovery and Recovery
(Eds Miall, A . D . & Tyler, N . ) . Soc. econ. Paleontol.
ACKNOW L E DGEMENTS Mineral. , Concepts Sediment. Paleont . , Tulsa, 3 ,
245 -260.
BRIDGES, P . H . ( 1982) Ancient offshore tidal deposits.
We would like to extend our sincerest thanks to
I n : Offshore Tidal Sands: Processes and Deposits (Ed.
Harold Reading for giving us our initial introduction Stride, A . H . ) , pp. 172- 192. Chapman & Hall , London.
to the Woburn Sands and for making us aware of its BRJSTOW, C . R . ( 1 963) The stratigraphy and structure of the
sedimentological significance. The basis for this Upper Jurassic and Lower Cretaceous Rocks in the area
work, its methodology and evaluation stems directly between Aylesbury (Bucks) and Leighton Buzzard
(Beds) , Unpublished PhD thesis, University of London.
from the lessons learnt by the authors under Harold's BRISTOW, C . S . ( 1995) Internal geometry of ancient tidal
previous supervision. We are extremely grateful for bedforms revealed using ground-penetrating radar. In:
this and for his ongoing interest and influence in our Tidal Signatures in Modern and Ancient Sediments. (Eds
work. Fleiruning, B .W. & Bartholoma). Spec. Pubis. int. Assoc.
Sediment. , No. 24. Blackwell Science, Oxford (in press) .
The authors would also like to thank the Konin­
BROWNRIDGE, S . & MOSLOW, T.F. ( 1991 ) Tidal estuary
klijke/Shell Exploratie and Produktie Laboratorium, and marine facies of the Glauconitic Member, Drayton
Riswijk, The Netherlands, for supporting both the Valley, central Alberta. I n : Clastic Tidal Sedimentology
project and its publication. We also acknowledge (Eds Smith, D . G . , Reinson , G . E . , Zaitlin, B .A . &
Shell Research B . V. and Shell Internationale Pet­ Rahmani, R . A . ) . Can. Soc. petrol. Geol . , Calgary,
Memoir 16, 107- 122.
roleum Maatschappij B . V . for their permission to BucK, S. ( 1985) Sand-flow cross strata in tidal sands of the
publish this paper. We are also indebted to Guy Lower Greensand (early Cretaceous ) , southern England.
Plint for his patience and encouragement during J. sediment. Petrol. , 55, 895-906.
the preparation of this paper. The assistance and BucK, S . ( 1987) Facies and sedimentary structures in the
Folkestone Beds (Lower Greensand, Early Cretaceous)
openness of the two referees, Jill Eyers and Alistair
and equi valent strata in southern England. Unpublished
Ruffell , is also gratefully acknowledged. Finally, Ian PhD thesis, University of Reading.
Glenister is thanked for redrafting the figures. CAMPBELL, C.V. ( 1971) Depositional model - Upper
44 H. D. Johnson and B. K. Levell

Cretaceous Gallup beach shoreline, Ship Rock area, ology of fluctuating sea-levels since the Triassic
northwestern New Mexico. J. sediment. Petrol. , 41, (250 million years ago to the present) . Science, 235,
395-409. 1 156- 1 167.
CASEY, R. ( 1961 ) The stratigraphical palaeontology of the HARRIS, P.T. & COLLINS, M . B . ( 1985) Bedform distribution
Lower Greensand. Palaeontology, 3, 486-62 1 . and sediment transport paths in the Bristol Channel and
CASEY, R . & GALLOIS, R.W. ( 1973) The Sandringham Severn Estuary. Mar. Geol. , 62, 153- 166.
Sands of Norfolk. Proc. Yorks. geol. Soc. , 40, 1 -22. HAYES, M .O. ( 1 975) Forms of sand accumulation in
COLMAN , S . M . , BERQUIST, C . R . , J R . & HOBBS, C.H. , I I I estuaries. I n : Estuarine Research , Vol . II, Geology &
( 1988) Structure, age and origin o f the bay-mouth shoal Engineering (eel . Cronin, L . E . ) , pp. 3 -22. Academic
deposits, Chesapeake Bay , Virginia. Mar. Geol. , 83, Press, London.
95 - 1 13 . HowARD, J . D . & FREY, R.W. ( 1 975) Regional animal -
CooK, P . J . & MAYO, W . ( 1977) Sedimentology and sediment characteristics of Georgia estuaries.
Holocene history of a tropical estuary (Broad Sound, Senckenbergiana Mar. , 7, 33- 103.
Queensland) . Aust. Bur. Miner. Resow·. , Geol. JENElTE, D .C. , JONES, C.R. , VAN WAGONER, J .C. &
Geophys. Bull. , 170, 206 pp. LARSEN , J . E . ( 1992) High resolution sequence strati­
DALRYMPLE, R.W. ( 1992) Tidal depositional systems. In: graphy of the Upper Cretaceous Tocito Sandstone: the
Facies Models: Response to Sea Level Change (Eds relationship between incised valleys and hydrocarbon
Walker, R.G. & James, N . P . ) , pp. 195 - 218. Geological accumulation, San Juan Basin, New Mexico. I n :
Association of Canada, St Johns. Sequence Stratigraphy Applications to Shelf Sandstone
DALRYMPLE, R.W. , ZAITLIN , B . A . & B OYD , R . ( 1991 ) Reservoirs: Outcrop to Subswface Examples (Eels van
Estuarine facies models: conceptual basis and strati­ Wagoner, J.C. , Jones, C.R . , Taylor, D . R . , Nummedal,
graphic implications. J. sediment. Petrol. , 62, 1 130- 1 146. D . , Jenette, D .C. & Riley, G.W. ) . American Associ­
DEMAREST, J .M. , II & KRAFT, J . C. ( 1 987) Stratigraphic ation of Petroleum Geologists Field Conference,
record of Quaternary sea levels: implications for September 199 1 .
more ancient strata. In: Sea-Level Fluctuation and JOHNSON , H . D . & BALDWIN , C.T. ( 1986) Shallow silici­
Coastal Evolution (Eds Nummedal, D . , Pilkey, O . H . & clastic seas. In: Sedimentary Environments and Facies
Howard , J . D . ) . Spec. Pub!. Soc. econ. Paleont. Miner . , (Ed. Reading, H . G . ) , pp. 229-282. Blackwell Scientific
Tulsa, N o . 4 1 , 223-229. Publications, Oxford .
DE RAAF, J . F . M . & BoERSMA, J . R . ( 197 1 ) Tidal deposits JoHNSON, H . D . & LEVELL, B . K . ( 1 980) Sedimentology of
and their sedimentary structures (seven examples from a Lower Cretaceous subtidal sand complex, Woburn
Western Europe) . Geol. Mijnbouw, 50, 479 - 50 1 . Sands, southern England. Bull. Am. Assoc. petrol. Geol. ,
EvANS, G. ( 1 965) Intertidal flat sediments and their 64, 728-729.
environments of deposition in the Wash. Q. J. geol. Soc. KIRKALDY, J .F. ( 1939) The history of the Lower Cretaceous
London, 121, 209-245. period in England. Proc. Geol. Assoc. , 50, 379-413.
Exu M , F. A. & HARMS, J .C. ( 1 968) Comparison of marine­ KLEI N , G . DE VRIES ( 1970) Depositional and dispersal
bar with valley-fill stratigraphic traps, western Nebraska . dynamics of intertidal sand bodies. J. sediment. Petrol. ,
Bull. Am. Assoc. Petrol. Geol. , 52, 1 85 1 - 1868. 40, 1 095 - 1 127 .
EYERS, J. ( 199 1 ) The influence of tectonics on early KNEBEL, H . J . , FLETCHER, C . H. , III & KRAFr, J . C . ( 1 988)
Cretaceous sedimentation in Beclfordshire, England. Late Wisconsinan - Holocene paleogeography of Dela­
J. geol. Soc. London , 148, 405 -414. ware Bay: a large coastal plain estuary. Mar. Geol. ,
EYERS, J . ( 1 992a) Lithostratigraphy of the Lower Greensand 83, 1 15- 133.
and Gault Clay (Lower Cretaceous) of the Bedfordshire LAMPLUGH, G.W. ( 1922) On the j unction of Gault and
Province, England. Unpublished PhD thesis, Open Lower Greensand near Leighton B uzzard (Beclfordshire) .
University. Q. J. geol. Soc. London, 78, 1 -80.
EYERS, J. ( 1992b) Sedimentology and palaeoenvironment LAMPLUG H , G.W. & WALKER, J.F. ( 1903) On a fossiliferous
of the Shenley Limestone (Albian, Lower Cretaceous): band at the top of the Lower Greensand near Leighton
an unusual shallow-water carbonate. Proc. Geol. Assoc. , Buzzard. Q. J. geol. Soc. London , 59, 234-265 .
103, 293-302. LEVELL, B . K. ( 1 980) Evidence for currents associated
GREER, S . A . ( 1975) Estuaries of the Georgia coast USA: with waves in Late Precambrian shelf deposits
III. Sandbody geometry and sedimentary facies at the from Finnmark, North Norway. Sedimentology, 27,
estuary/marine transition zone, Ossabaw Sound Georgia; 153- 166.
a stratigraphic model. Senckenbergiana Mar. , 7, LuDWICK, J .C. ( 1 970) Sand waves and tidal channels in
105 - 135 . the entrance to Chesapeake Bay, Virginia. Va. J. Sci. ,
HAMILTON, D. ( 1979) The high-energy, sand and mud 21 ' 178- 184.
regime of the severn estuary, S . W . Britain. In: Tidal LUDWICK, J .C. ( 1 974) Tidal currents and zig-zag sand
Power and Estuary Management (Eds Severn, R .T. , shoals in a wide estuary entrance. Geo/. Soc. Am. Bull. ,
Dineley, D . L . & Hawker, L . E . ) , pp. 162- 1 72. Trans­ 85, 717-726.
atlantic Arts Incorporated, Albuquerque, Colston MAGUEREGUI, J . & TYLER, N. ( 1991 ) Evolution of Middle
Paper 30. Eocene tide-dominated deltaic sandstones, Lagunillas
HANCOCK, J . M . ( 1 958) The Lower Cretaceous near field, Maracaibo Basin, Western Venezuala. In: The
Leighton Buzzard. I n : The London Region Geologists Three-Dimensional Facies Architecture of Terrigenous
Association Guide, No. 30, pp . 36-40. Clastic Sediments and Its Implication for Hydrocarbon
HAQ, B . - U . , HARDENBOL, J. & VAIL, P . R. ( 1987) Chron- Discovery and Recovery (Eels Miall, A . D . & Tyler, N . ) .
 Transgressive estuarine sand complex 45

Soc. econ. Paleontol Mineral . , Concepts Sediment. Aktua-palaeontologie nach Studien in der Nordsee,
Paleont., Tulsa, 3 , 233-244. p. 666. W. Kramer, Frankfurt. ) .
McCuBBIN, D . G . ( 1 969) Cretaceous strike-valley sandstone ScHUBEL, J . R . ( 1971 ) Sources o f sediments to estuaries.
reservoirs, Northwestern New Mexico. Bull. Am. Assoc. In: The Estuarine Environment, Estuaries and Estuarine
petrol. Ceo!. , 53, 2 1 14-2140. Sedimentation (Eel. Schubel, J . R . ) , pp . V, 1 - 19.
NARAYA N , J. ( 197 1 ) Sedimentary structures in the Lower American Geological Institute, Washington.
Greensand of the Weald, England, and Bas-Boulonnais, SCHWARZACHER, W. ( 1953) Cross-bedding and grain size in
France . Sediment. Geol. , 6, 73 - 109. the Lower Cretaceous sands of E. Anglia. Geol. Mag. ,
N10, S.-D. & YANG, C .-S. ( 1991 ) Diagnostic attributes 90, 322-330.
of clastic tidal deposits: a review. I n : Clastic Tidal SHEPHARD-THORN , E . R . , HARRIS, P . M . , HIGHLEY, D . E . &
Sedimentology (Eds Smith, D . G . , Reinson, G . E . , THORNTON, M . H . ( 1 986) An Outline Study of the Lower
Zaitlin, B . A . and Rahmani , R . A . ) . Can . Soc. petro l . Greensand of Parts of South-east England. Report for
Geo l . , Calgary, Memoir 16, 3-28. the Department of the Environment, British Geological
NuMMEDAL, D . & SwiFr, D .J.P. ( 1987) Transgressive Survey, Keyworth.
stratigraphy at sequence-bounding unconformities: some SPEARING, D.R. ( 1975) Shannon Sandstone, Wyoming. In:
principles derived from Holocene and Creataceous Depositonal Environments as Interpreted from Primary
examples. In: Sea-Level Fluctuation and Coastal Evol­ Sedimentary Structures and Stratification Sequences.
ution (Eds Nummedal, D . , Pilkey, O . H . & Howard, Society of Economic Paleontologists and Mineralogists,
J . D . ) . Spec. Pub!. Soc. econ. Paleont. Mineral, Tulsa, Short Course No. 2, pp. 104- 1 14.
4 1 , 241 - 260. STRIDE, A . H . ( 1982) Offshore Tidal Sands: Process and
OERTEL, G . F . , HENRY, V . J . & FOYLE, A . M . ( 1 99 1 ) Impli­ Deposits. Chapman & Hall, London, 213 pp.
cations of tide-dominated lagoonal processes on the SuRLYK, F. & NoE-NYGAARD, N . ( 1991 ) Sand bank and
preservation of buried channels on a sediment-starved clune facies architecture of a wide intracratonic seaway:
continental shelf. In: Shelf Sand and Sandstone Bodies: Late Jurassic- Early Cretaceous Raukelv Formation,
Geometry, Facies and Sequence Stratigraphy (Eds Swift, Jameson Land, East Greenland. I n : The Three­
D . J . P . , Oertel, G.F. , Tillman, R . W . & Thorne, J . A . ) . Dimensional Facies Architecture of Terrigenous Clastic
Spec. Pubis i n t . Assoc. Sediment., No. 1 4 , 379-393. Sediments and Its Implication for Hydrocarbon Discovery
Blackwell Scientific Publications, Oxford . and Recove1y (Eels Miall, A . D . & Tyler, N . ) . Soc. econ.
OoMKENS, E . & TERWINDT, J . H . J . ( 1960) Inshore estuarine Paleontol. Mineral . , Concepts Sediment. Paleont . ,
sediments in the Haringvliet, Netherlands. Ceo/. Tulsa, 3 , 261 - 276.
Mijnbouw, 39, 701 -710. SwiFT, D . J . P . & THOR N E , J .A . ( 1 99 1 ) Sedimentation
OwEN , H.G. ( 1 972) The Gault and its junction with on continental margins, 1: a general model for shelf
the Woburn sands in the Leighton Buzzard Area, sedimentation. I n : Shelf Sand and Sandstone Bodies:
Bedfordshire and Buckinghamshire. ?roc. Geol. Assoc. , Geometry, Facies and Sequence Stratigraphy (Eels Swift,
83, 287-312. D . J . P . , Oertel, G.F. , Tillman, R.W. & Thorne, J . A . ) .
PuNT, A.G. & WALKER, R . G . ( 1 987) Morphology and Spec. Pubis i n t . Assoc. Sediment . , No . 1 4 , p p . 3 - 3 1 .
origin of an erosion surface cut into the Bad Heart Blackwell Scientific Publications, Oxford.
Formation during major sea-level change, Santonian of SWIFT, D . J . P . , MOIR, R . & FREELAN D , G . L. ( 1980)
West-Central Alberta, Canada. J. sedim. Petrol. , 57, Quaternary rivers on the New Jersey shelf: relation of
639-650. seafloor to buried valleys. Geology, 8, 276-280.
PuNT, A . G . & NORRIS, B . ( 1 99 1 ) Anatomy of a ramp TERWINDT, J . H . J . ( 1 973) Sand movement in the in- and
margin sequence: facies successions, palaeogeography, offshore tidal area of the southwestern part of The
and sediment dispersal patterns in the Muskiki and Netherlands. Geol. Mag. , 52, 69-77.
Marshybank formations, Alberta Foreland Basin. Bull. THORNE, J . A . & Swwr, D . J . P. ( 1991) Sedimentation on
Can. Pet. Geol. , 39, 18-42. continental margins, VI: a regime model for depositional
REINECK, H . E . ( 1963) Seclimentgefuge in Bereich cler sequences, their component systems tracts, and bounding
Sucllichen Norclsee. Abh. Senckenb. Natwforsch. Ges, surfaces. I n : ShelfSand and Sandstone Bodies: Geometry,
505, 1 - 138. Facies and Sequence Stratigraphy (Eds Swift, D . J . P . ,
REINECK, H . E . & SING H , I . B . ( 1973) Depositional Sedi­ Oertel , G . F . , Tillman, R . W . & Thorne, J . A . ) . Spec
mentary Environments with Reference to Terrigenous Pubis int. Assoc. Seclimento l . , No. 14, pp. 189-255.
Clastics. Springer-Verlag, New York, 431 pp. Blackwell Scientific Publications, Oxford.
RuFFELL, A . H . & WACH, G . D . ( 199 1 ) Sequence strati­ VAN DEN BERG, J . M . , GoESTEN , M . J . B . G . & S M U LDERS, F.
graphic analysis of the Aptian- Albian Lower Greensand ( 1980) Dynamics and sequential analysis of a mesoticlal
in southern England. Mar. Petrol. Ceo/. , 8, 341 -353 . shoal and intershoal channel complex in the eastern
RuFFELL, A . H . & WACH, G . D . (in press) Comparison Scheidt (southwestern Netherlands). Sedimem. Geol. ,
of depositional sequences in the arenaceous beds of 26, 263 -279.
the Aptian-Albian boundary (Cretaceous) in southern VAN WAGONER, J .C . , MITC H U M , R . M . , CAMPIO N , K . M . &
and eastern England. In: The Mesozoic and Cenozoic RAHMA N IAN , V . D. ( 1 990) Siliciclastic Sequence Stra­
Sequence Stratigraphy of European Basins (Eels tigraphy in Well Logs, Cores, and Outcrops: Concepts
Graciansky, P.C. & Jacquin, T . ) . Soc. econ . Paleont. for High-resolution Correlation of Time and Facies.
Mineral. Tulsa. American Association of Petroleum Geologists, Tulsa,
ScHAFER, W. ( 1972) Ecology and Palaeoecology of Marine Methods in Exploration, 7, 55 pp.
Environments , p. 568. Oliver & Boyd, Edinburgh, (Also: ViSSER, M . J . ( 1 980) Neap-spring cycles reflected in
46 H. D. Johnson and B. K. Levell

Holocene subtidal large-scale bedform deposits: a WRJGHT, C.W. & WRIGHT, E . ( 1947) The stratigraphy of
preliminary note. Geology, 8, 543-546. the Albian Beds at Leighton Buzzard. Geol. Mag. , 84,
WALKER, R.G. ( 1 984) Shelf and shallow marine sands. I n : 1 6 1 - 168.
Facies Models ( E d . Walker, R . G . ) . Geosci. Can . , WvArr, R.J . , LAKE, R . D . , MoORLOCK, B . S . P . & SHEPHARD­
Reprint Ser. 1 , 1 4 1 - 170. THOR N , E . R . ( 1986) The Leighton Buzzard Area: Geo­
WALKER, R.G. & BERGMAN , K .M. ( 1992) Late Cretaceous logical Report on 1: I 0 000 Sheets SP82NE, Southeast;
clastics: cores and outcrops. In: Field Trip Guidebook: SP92NW, Southwest; SP93NW, Northeast, Southwest,
Trip No. 6. Canadian Society of Petroleum Geologists/ Southeast. British Geological Survey, Keyworth .
American Association of Petroleum Geologists, Calgary, YANG, C.-S. & N i o , S.-D. ( 1989) An ebb-tide delta depo­
77 pp. sitional model - a comparison between the modern
WALKER, R . G . & P uNT, A . G . ( 1992) Wave- and storm­ Eastern Schelde tidal basin (southwest Netherlands) and
dominated shallow marine systems. In: Facies Models: the Lower Eocene Roda Sandstone in the southern
Response to Sea Level Change (Ed. Walker, R.G. & Pyrenees (Spain ) . Sediment. Geol. , 64, 175- 196.
James, N . P . ) , pp. 2 19-238. Geological Association of ZIEGLER, P . A . ( 1 988) Evolution of the Arctic- North
Canada, St Johns. Atlantic and the Western Tethys. Mem . Am. Assoc.
WoNHAM, J . P . ( 1993) Sedimentology and sequence stra­ petrol. Geo l . , 43, 198 pp.
tigraphy of tidal sandstone bodies: implications for ZIEGLER, P.A. ( 1990) Geological Atlas of Western
reservoir characterisation. Unpublished PhD thesis, and Central Europe. Shell Internationale Petroleum
University of Liverpool. Maatschaappij B . V . , The Hague, 239 pp.
Spec. Pubis int. Ass. Sediment. (1995) 22, 47-74

An incised valley in the Cardium Formation at Ricinus, Alberta:

reinterpretation as an estuary fill

R O G ER G . W A L K E R
Department of Geology, McMaster University,
Hamilton, Ontario L8S 4Ml, Canada

ABSTRACT

The Ricinus incised valley (Turonian-Coniacian Cardium Formation) is at least 50 km long, 10 km wide
and up to about 30 m deep. It trends NW-SE, parallel to tectonic trends and shoreline trends in the
Cardium Formation. I t cuts into shallow-marine mudstones and sandstones (Raven River Allomember
of the Cardium), and is truncated by a transgressive surface of erosion (E5). The valley fill forms a small
oil reservoir in the subsurface, but the field is disrupted by thrust faults, and the western margin is poorly
constrained because of thrusting and Jack of wells.
The valley fill consists of interbedded sandstones and mudstones, with both brackish and fully marine
trace fossil assemblages. Some of the sandstones are structureless, with individual beds apparently over
4 m in thickness; others are parallel-laminated with beds up to 3 m thick. Cross-bedding in sets thicker
than 5 em is extremely rare. The mudstones contain some thin silty and sandy laminations, but are
commonly extensively bioturbated with a fully marine trace fauna (robust Helminthopsis, Zoophycos,
Planolites, Anconichnus, Skolithos, Ophiomorpha, Thalassinoides, Teichichnus, Asterosoma, Terebellina
and Rosselia). The mudstones can be traced as a sheet about 4 m thick along the entire length of the
valley; they appear to pinch out against the eastern wall of the valley. Chert-pebble conglomerates are
scattered throughout some wells, but tend to be concentrated near the northern end of the valley. Here,
they arc associated with sandstones and with fully marine mudstones.
The valley was formerly interpreted to have been cut and filled by turbidity currents, but this
interpretation is difficult to reconcile with the depositional environments of the rest of the Cardium
Alloformation. It is suggested here that the valley was cut by a river during a falling stage of relative sea­
level, and was filled transgrcssivcly as an estuary. There is absolutely no evidence for tidal currents, and
little indication of waves. The very thick and structureless sandstones suggest pulses of sediment swept
into the estuary from the bay head, possibly as density underflows. There is little or no evidence for the
reworking of these deposits within the estuary. There arc no central basin mudstones (no turbidity
maximum), although this could be due to incomplete preservation of the entire length of the estuary.
The geometry of the estuary and the facies distribution of the fill suggest that the open sea Jay to the
north. Chert pebbles moving in the longshore drift system were introduced into the estuary from the
north by storms; they were not introduced from the fluvial (southern ) end of the estuary. The
orientation of the incised valley, parallel to depositional strike for over 50 km, remains a maj or problem.

INTRODUCTION

The Ricinus sand body in the Cardium Formation of storm-dominated shallow-marine sandstones (Walker
Alberta (Figs 1 & 2) is over 50 km long, at least & Eyles, 1 988) , and transgressive incised shoreface
10 km wide, and up to about 30 m thick (Walker, deposits ( Bergman & Walker, 1987 , 1988; Pattison
1985 ) . I originally interpreted it as a channel fill, & Walker, 1 992). The main purpose of this paper
with channel cutting and filling by turbidity currents. is to re-examine the allostratigraphical (NACSN ,
This interpretation has always been difficult to 1983) position of the sand body (Figs 1 & 2 ) , and to
reconcile with the sedimentology of the other facies reinterpret the incised valley fill as estuari ne.
in the Cardium Formation, which consist largely of Ricinus is a small oil field, discovered in 1969.

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
47
48 R.G. Walker

10 9 8 7

37
RICINUS
0 Wells
.
Cores

OFF-FIELD

o Wells
36 " Cores

60'

ALBERTA 35

054' 0

Fig. I. Map showing locations of

3 4 Cardium fields associated with the

/�K�\�9'
E4 (fields outlined) and E5 (black)
\o so bounding discontinuities. Note

1 5:
6
location of Ricinus close to and

I
partly straddling the edge of the
� EDMONTON deformed belt (thrust symbol, inset
33
lower left). The main map shows
w--.:_ E5 fields locations of wells and cores. All of
the wells immediately east of
�
�
Ricinus are shown, but only a few
Caroline wells have been used. Note
RICIN
�\,
U '\.c::
E4 fields
also the locations of cross-section
32 A-A' through to G-G'. Number
� CALGARY 6 miles at right indicates Townships, and

116' 114' 5
. - 1' -10 KM number along top indicates Ranges
west of the 5th Meridian (114°).

sw c.20 km NE
Fig. 2. Stratigraphy of the Cardium
Alloformation in the Ricinus area,
drawn roughly to scale vertically.
Bounding discontinuities E2, E3,
E6 and E6.5 cannot be recognized
here. Sandier upward successions
A-D are assinged to the Hornbeck
Allomembcr (note that succession
D is commonly truncated by E4).
E5 RSE (regressive surface of
erosion) defines the base of the
Ricinus Allomember, and E5 TSE
(transgressive surface of erosion)
defines the top. There is some
evidence that E5 TSE cuts
downward progressively
southwestward. IT and RT refer to
initial and resumed transgressions,
as discussed in the text.
 Cardium Formation incised valley 49

Initial oil-in-place was estimated as 2.88 x 106m3

Structural problems at Ricinus
(18. 1 million bbl) , with initial recoverable oil esti­
mated at 2.27 x 106m3 ( 14.3 million bbl) (Geological The exact morphologies of the Ricinus valley and
Survey of Canada, 198 1 a , b). However, the signifi­ sand body are difficult to reconstruct because Ricinus
cance of the field goes well beyond its minor econ­ lies at the eastern limit of the Foothills deformed
omic potential. It is the only incised valley known in belt (Figs 1 , 10 & 1 1 ) . The sand body is broken by
the Cardium, which is surprising in a formation several SW-NE trending faults, interpreted to have
dominated by eight regionally extensive erosion formed in response to different distances of thrusting
surfaces, each controlled by major fluctuations of toward the northeast. The fault positions have been
relative sea-level. The incised valley interpretation located mainly by offsets in the northeastern margin
is complicated by the fact that the valley is straight, of the field, which can be fairly well constrained by
at least 50 km long, and is parallel to regional strike the close juxtaposition of wells with 'Ricinus' well­
(Fig. 1 ) . log responses versus 'off-field' log responses (Figs 1
The new interpretations are based upon all o f the & 3-7). Within Ricinus, the sand body is repeated
data available as of April 1993 , as illustrated in by thrusting in many wells (see Fig. 5 ) . In a few wells
seven new cross-sections presented here (see Figs the dips exceed 45° and are locally vertical.
3-9 and Appendix 1 ) . There are 172 on-field wells,
of which 71 are cored (Fig. 1 ) . All measurements are
The former turbidite interpretation
given in metric units, except where specific reference
is made to wells where the well logs and cored The interpretation of Ricinus as a channel cut and
intervals were originally recorded in feet. filled by turbidity currents (Walker, 1985) was based
on the presence of very thick structure less sandstones
and the presence of many beds characterized by a
INTRODUCTION TO THE Tabc and Toe internal structure (Bouma, 1962). In
CARDIUM ALLOFORMA TION the early 1980s, the possibility of turbidity currents
in the Western Interior Seaway was suggested by the
The Cardium Alloformation (NACSN , 1983) is need to explain how a very large number of sharp­
Upper Cretaceous (Turonian - Coniacian), and based HCS beds had been emplaced well below fair­
occurs in the Alberta foreland basin. This basin is weather wave base. Thinking was also influenced by
part of the Western Interior Seaway, which was the occurrence of well-defined turbidites overlain by
open from the Boreal Ocean to the Gulf of Mexico HCS beds in the Jurassic Passage Beds in the Banff
during the Turonian . Sediment was supplied from area. Here , both the turbidites and HCS beds have
the active cordillera in the west. identical palaeocurrent orientations (Hamblin &
The Cardium is about 100 m thick, and is Walker, 1979), suggesting emplacement of the HCS
dominated by open-marine mudstones and storm­ beds by turbidity currents, but above storm wave
influenced sandstones with hummocky cross­ base.
stratification (HCS). There is one prograding Turbidite ideas were first applied to the Cardium
shoreface succession (Kakwa Allomember; Plint before the presence and significance of sea-level
& Walker, 1987) overlain by non-marine deposits changes in the Cardium had been documented (Piint
(Musreau Allomember), and several examples et a/., 1986), and before much of the important
of incised transgressive shoreface deposits. The recent work on estuarine deposits had been pub­
relationships of these various deposits have been lished (Dalrymple et a/. , 1992). The main purpose of
documented by mapping eight bounding discon­ this paper is to suggest an alternative to turbidity
tinuities, at least five of which can be traced basin­ currents for cutting and filling the incised valley at
wide (Piint et a/. , 1986; Wadsworth & Walker, 199 1 ; Ricinus.
Walker & Eyles, 199 1). These were originally termed
erosion surfaces and lettered El- E6, E6. 5 and E7.
Allostratigraphy
The regional trend of the various shoreface sand
bodies is consistently NW-SE (Fig. 1 ) , parallel to The most useful subdivision of the Cardium is
the active cordillera to the west of the foreland allostratigraphical (NACSN , 1983) rather than
basin. lithostratigraphical (Fig. 2). An allostratigraphical
unit 'is a mappable stratiform body of sedimentary
50 R.G. Walker

rock that is defined and identified on the basis of its regionally extensive sheet-like sandier upward
bounding discontinuities ... [one may thus define] as successions of the Hornbeck Allomember (Fig. 2).
single units discontinuity-bounded deposits charac­ These successions are labelled A through to D
terized by lithic heterogeneity' (NACSN , 1983 , in Figs 2-8 and collectively form the datum in
p. 865). I regard the Cardium Alloformation as these cross-sections. Succession C is cored in wells
a formally defined stratigraphical unit, and have 7-9-36-8W5 and 10-17-34-7W5 (Fig. 4; all wells lie
appended some thoughts on allostratigraphy in west of the 5th Meridian , and the W5 will be omitted
Appendix 2. The Cardium Alloformation has been from all subsequent well numbers cited). Succession
subdivided into 14 allomembers, with eight well­ C consists of bioturbated sandy mudstones that
defined bounding discontinuities (E1 through to E7; gradually become sandier upward; the top consists
Plint et al. , 1986) . Throughout this paper, informal of a 1-cm-thick layer of chert grains about 0.5 mm
use of the term Cardium (without a modifier) implies (7-9-36-8) or 3 mm (10-17-34-7) in diameter. This
the Cardium Alloformation. The term Cardium sandier upward succession, capped by a gritty layer,
Formation will be used only to designate the litho­ is similar to those described by Walker & Eyles
stratigraphical unit. ( 1 988) from the Raven River Allomember; suc­
cessions A, B and D have not been cored but are
probably similar to succession C.
OUTLINE OF CARDIUM STRATIGRAPHY The sand bodies that form reservoirs at Garrington,
AND SEDIMENTOLOGY Crossfield and Caroline (E4 fields in Fig. 1; Pattison
& Walker, 1992) are assigned to the Burnstick
The base of the Cardium Alloformation is taken at Allomember. They rest in NW-SE trending steps
bounding discontinuity El (Figs 2 -6). In the Ricinus on the E4 erosion surface (Fig. 2) , and are inter­
area E1 has never been cored, and it is probably a preted as transgressive shorefaces. Each incised
correlative conformity (see Appendix 2) rather than shoreface is underlain by a surface of initial trans­
a surface with demonstrable erosion . Its log identi­ gression (IT in Fig. 2) , and is truncated by a surface
fication is compatible with the E 11og identifier used of resumed transgression (RT in Fig. 2). An RT
by Plint et al. ( 1 986) and Wadsworth & Walker surface can pass southwestward into an IT surface if
( 199 1 ) . Erosion surfaces E2 and E3 cannot be ident­ and when the transgression pauses and another still­
ified in this area. Surface E4 has demonstrable stand shoreface becomes incised. An example can
erosion (Figs 3, 5 & 6), and cuts into a series of be seen in well 7-33-36- 10, section A (Fig. 3 ) , where

7-33-36-10 7-26-36-10 7-36-36-10 1-31-36-9 8-32-36-9 11-33-36-9 10-34-36-9 11-36-36-9 7-32-36-8

® @
-- - ---'
?-
--

F
� 0

� ���
�
�
�

� �
�Es� �TSE� � � �
�

�--

_t_ -EI-

Fig. 3.Gamma-ray and resistivity well-log cross-section A-A', hung collectively on a series of more -or-less parallel
markers above and below the Cardium. Depths are marked at 50-m intervals and cored intervals are shown as black bars.
Wells 7-33-36-10, 7-26-36-10, 7-36-36-10 and 7-32-36-8 have 'off-field' well-log signatures.
 Cardium Formation incised valley 51

7-36-36-10 11-18-36-9 5-10-36-9 4-18-36-8 10-18-36-8 7-19-36-8 7-9-36-8

15-23-36-10

® ®
- -

�� �

--- �/
�
z - ----
0
-TSE �co�
::;;�
0::
?- ' /
0 -
LL __ f\_SE___ ___________ ,
'3 BLACK
�E4�
fc
<i.� �E4�
::;;­
::;) _,
0
0::
"'
u

j \
-E. I-

Fig. 4. Cross-section B-B', hung communally on the sandier upward successions A-C, and the top of the black blanket
(dashed line). See caption of Fig. 3 for other conventions, and note draping of upper markers over the Ricinus sand body.
The shoulder on the resistivity and gamma-ray logs typical of E7 is not present in 7-9-36-8, nor is the narrower waist in the
well-log profile immediately below E7. The waist appears to be cut out by E7 toward 7-9-36-8, and E7 is probably only
about 2 m above E5 (see Figs 12-14). Wells 7-36-36-10, 15-23-36-10, 11-18-36-9, 7-19-36-8 and 7-9-36-8 have 'off-field' well­
log signatures.

1-5-34-8 6-9-34-8 14-10-34-8 10-10-34-8 6-14-34-8 6-24-34-8 11-20-34-7 10-21-34-7

© ©
F?

--

---- §
---- a>
---- g
"'
"'

-- EI - -
g----
"'

Fig. 5.Cross-section C-C', hung on sandier upward successions A-C. Other conventions explained in the caption of Fig.
3. Wells 1-5-34-8 and 10-21-34-7 have 'off-field' well-log signatures. F indicates a fault repeat of the top of the Ricinus
Allomember in 10-10-34-8, and F? indicates possible faulting. RRl is the first sandier upward succession in the Raven River
Allomember.
52 R.G. Walker

1 0-8-33-7 4-9-33-7 3-10-33-7 7-10-33-7 10-12-33-7 6-24-33-7 11-25-33-7

® ®

g
�­
z
0
"
N
�
-- ? ?
0
�:::!: 0
N

0:::
0
LJ...

E4---
0 BLANKET
...J

rc
...J
<[
�N --

� TB 0
15 +A g
N
0:::
<[
u
t - EI- -

Fig. 6.Cross-section D-D', hung on sandier upward successions A-C. Other conventions explained in the caption of Fig.
3. Note the erosional truncation of markers between ES TSE and E7 by the E7 bounding discontinuity between wells
7-10-33-7, 10-12-33-7 and 6-24-33-7. Wells 10-12-33-7, 6-24-33-7 and 11-25-33-7 have 'off-field' signatures.

11-18-33- 7 3-19-33-7 8-19-33-7 6-20-33-7 4-21-33-7 6-23-33-7

© Fig. 7. Cross-section E-E', hung

on the B sandier upward succession.
Note truncation of the top of the
'black blanket' and top of RR1 by
bounding discontinuity ES RSE,
and the apparent pinchout of facies
D between wells 3-19-33-7 and
8-19-33-7 against the wall of the
incision. Facies D and the
sandstones above (3-19-33-7)
----- ----
appear to be eroded by ES TSE
toward well 11-18-33-7.

E4 formed during an initial transgression , and sand blanket is taken at the first deflections of the gamma­
was deposited on the surface during a stillstand in ray and resistivity logs (the dashed line in Figs 4-8).
transgression . This sand was truncated as the E4 Here, the black blanket grades into bioturbated
transgression continued, leaving a 4-m-thick sand siltstones and sandstones, and HCS sandstones ,
body. However, over most of the study area, E4 is organized into sandier upward successions. The first
probably overlain only by a thin lag. of these successions is identified as RR-1 (Figs 5 &
The Raven River Allomember extends from the 7). The successions are eroded to various depths
top of this lag to the ES bounding discontinuity (Fig. by ES.
2). The lower part consists of black mudstones (the Erosion surface E5 has received much attention
'black blanket' of Walker, 1983) characterized by (Fig. 2; Walker & Eyles, 199 1 ) , partly because of
parallel and featureless gamma-ray and resistivity its economic importance in locating the shoreface
logs (Figs 2 & 4-8). In logs, the top of the black gravels in the Carrot Creek area (Bergman &
 Cardium Formation incised valley 53

7-16-34-8 16-16-34-8 10-22-34-8 6-24-34-8

® !15 ®

Fig. 8. Cross-section F-F', hung

comunally on sandier upward
successions A-C. Note thinning
and disappearance of facies C and/
or D from well 16-16-34-8 eastward
�,L
�ISE"""'
�

toward well 10-22-34-8, and possible BLANKET

erosion of these facies by the thick
sandstone in well 7-16-34-8 (possible
erosion surface shown by question
marks). Note also the truncation of
markers between ES TSE and E7 by
the E7 bounding discontinuity.

'-"'"""E7"'"""-
3-19-33-7 9-16-34-8 16-30-34-8 1 2 -19-35-8 ll-30-35-8 16-30-35-8 3-6-36-8
�

3-6-36-8 7-18-36-8 10-18-36-8 16-13-36-9 6-24-36-9 13-26-36-9 1 2-34-36-9 2-4-37-9

)
-�
!15
�0 M

Fig. 9. South (G) to north (G') cross-section showing the continuity of facies C and/or D along the length of the field.
These muddy facies can be recognized by their distinctive indentation in the gamma-ray and resistivity well-logs. The
mudstones occur in the uppermost part of the core in 12-34-36-9 (Fig. 26), and also appear to occupy the same position in
wells 7-18-36-8, 10-18-36-8 and 13-26-36-9. In these wells, the Ricinus Allomember is unusually thin, suggesting that
sandstone originally above facies C and/or D has been eroded at the ES TSE bounding discontinuity.
54 R.G. Walker

Walker, 1987, 1988) , and also for determining the body that rests on E5 RSE and is truncated by E5
remnant erosional topo·graphy on top of major TSE. The extent of this sand body, and the type
Cardium reservoirs at Pembina, Willesden Green wells are the same as for the Ricinus Member as
and Ferrier (shown black and identified as E5 fields defined by Walker ( 1985, p. 1 98 1 ) .
in Fig. 1 ; Walker & Eyles, 1988, 199 1 ) . In this Discontinuity E 6 i s a n erosion surface with little
paper, the base of the Ricinus sand body is correlated topographical relief and it cannot be identified
with E5 - specifically with a falling stage of relative reliably in this area. Although E6.5 has more relief
sea-level (E5 RSE- regressive surface of erosion) . than E6 it also is difficult to identify in the study
Over most of the Cardium basin, the subaerial E5 area. Discontinuity E7 consistently forms a small
surface was modified during transgression , and is shoulder on the gamma-ray and resistivity logs, and
designated E5 TSE (transgressive surface of erosion) can be identified easily. The topography of the E7
in Figs 3-9. Erosion beneath E5 RSE is illustrated surface has been mapped by Wadsworth & Walker
by the truncation of log markers, particularly in ( 199 1 ) . They showed a deep 'valley' or 'low' on the
Figs 5 - 8 . erosion surface j ust east of Ricinus, and part of this
The extent o f erosion a t the E 5 R S E surface can can be seen in Figs 4 & 6. In Fig. 4, E7 is not present
also be shown by isopaching the E4 to E5 RSE in the 'expected' position in 7-9-36-8 (Fig. 12) . The
interval. If it is assumed that E4 is relatively planar lag associated with E5 is prominent in core (Fig. 13),
in the Ricinus area, the shape of the Ricinus incision and is overlain by transgressive mudstones. A second
can be seen in Fig. 10. The new term Ricinus lag, with chert pebbles up to 1 . 1 em diameter, occurs
Allomember (Fig. 2) is introduced here for the sand about 2 m above E5 (Fig. 14) ; it shows on the

(?) in Fig. 4. A similar relationship is shown in Figs

gamma-ray and resistivity logs and is identified as E7

10 9 8 7 6-8, where log markers in the E5 TSE to E7 interval

are truncated by E7. Log marker E7, which is nor­
37 mally well developed and easy to identify, is absent

10 9 8 7
36

37

35

36

34

6
35

33

34
Isopach,
E4 to E5 RSE 6
(base of Ricinus 32
incised valley),
metres 33
lsop1ch,
E5 RSE to E5 TSE,
Fig. 10. Isopach map of the interval between E4 and E5. metres
On field, the map shows the E4-E5 RSE interval, and off
field, the E4-E5 TSE interval. Isopachs show height 32
above E4; thus the 4 m isopach shows the deepest part of
the valley. Note that the edge of the valley is approximated
by the 28 m isopach. The line F shows one of the faults at
Ricinus (Fig. 1); north of this fault, all of the data points Fig. 11. Isopach map of the E5 RSE to E5 TSE interval
have been shifted southwestward to line up the axis of the (the thickness of the Ricinus sand body). See Fig. 10 and
field as defined by the well density - see Appendix C. Appendix C for explanation of line F.
 Cardium Formation incised valley 55

Fig. 12. Core from well 7-9-36-8, which lies just east of the
margin of Ricinus. B =bottom, T =top. Note siderite
horizon (below label) and chert pebbles (above label)
associated with ES, and the interval of laminated
mudstones (the 'laminated blanket' of Walker, 1983)
between ES and E7. Note siderite horizon (under label)
and chert pebbles (above label) associated with E7, and
the much blacker mudstones without silty laminations
above E7. The separation of ES and E7 over Ricinus is
40-48 m, but is about 2m in this core. Thus there is about
40 m of erosional relief on E7 immediately to the east of
Ricinus (Wadsworth & Walker, 1991).

in the easternmost wells in these cross-sections,

suggesting that E7 and E5 are co-planar (or that E7
has actually cut into and removed E5). The top of Fig. 13. Detail of siderite horizon, chert pebbles, and
eroded sandstone clast (large triangular shape) at the ES
the Cardium Alloformation is taken at the top of the
TSE horizon, well 7-9-36-8, 8634 ft. See Fig. 12 for
lag that rests on E7. stratigraphical context of this photograph, but note that
core has been rotated in Fig. 13 to show different features.
Core is 3 in. (7.6 cm) wide.
GEOMETRY OF THE
RICINUS INCISED VALLEY mined that bounding discontinuities E7 , E5 TSE , E5
AND SAND BODY RSE , E4 and El could be traced and identified
throughout the study area, maps of various strati­
The geometry of the Ricinus sand body was estab­ graphical intervals were made using SURFER
lished initially from many cross-sections - seven are 4.0 (Golden Software Inc. , Golden, Colorado)
presented here (Figs 3 - 9). After it had been deter- (see Appendix 3). ·
56 R.G. Walker

which is the base of the valley. The eastern edge of

the valley is approximated by the 28 m isopach. The
valley trends roughly NW-SE, with a straight and
well-defined eastern margin that has at least 24 m of
relief in T34 R8-9 (Fig. 10). This eastern valley wall
has a gradient of about 0.007° westward into the
valley. The eastern margin remains a straight and
well-defined feature from T35 R8-9 to the southern­
most fault in T32 R7, but northward, the valley
seems to broaden and become less well defined in
T36 R9. Erosion at the eastern valley wall is evident
in the cross-sections; Figs 3 - 8 all show truncation of
log markers by E5 RSE (particularly in Figs 5, 7 and
8) , or convergence of E5 RSE toward the top of the
'black blanket' (Fig. 6) , or convergence toward E4
(Fig. 5). The eastern margin is characterized by a
rapid change from on-field to off-field log signatures,
allowing a boundary to be drawn to within 600 m
(e.g. in land section 2 1 , T35 R8 and land sections
17-20, T36 R8; Fig. 1 ) .
The western margin i s very poorly defined because
there is little well control; many of the well logs
suggest extensive thrusting and are impossible to use
in the isopach maps. The maximum depth of the
valley is in the centre of the field (T34 R8 -9), and
the valley becomes shallower both northwestward
and southeastward.
The northern and southern ends of the valley are
faulted. Wells on line with the Ricinus sand body
but outside the faulted ends have 'off-field' well-log
signatures.

Sand body geometry

The sand body geometry was defined by isopaching

the E5 TSE to E5 RSE interval (Fig. 1 1) , with the
assumption that E5 TSE (the upper bounding dis­
continuity) approximates a planar surface. The sand
Fig. 14. Detail of siderite, bioturbation and chert pebbles
body isopach map closely resembles the map of
at the E7 horizon, well 7-9-36-8, 8639 ft. See Fig. 12 for
stratigraphical context of this photograph. Core is 3 in. the E4 to E5 RSE interval (the morphology of the
(7. 6 cm) wide. base of the valley) , with the sand body thinning
northwestward and southeastward from a maximum
thickness greater than 24 m in T34 R8. In the
northwest, the isopach of the sandstone seems to
Incised valley geometry
define a valley more clearly than the E4 to E5 RSE
An isopach map of the E4 to E5 interval is shown in map. The prominent eastward deflections of the
Fig. 10. On-field, the map shows the E4 to E5 RSE 8 and 4 m isopachs in T34 R7 -8 are due to the
interval, and off-field, it shows E4 to E5 TSE. 1 1-m-thick sandstone in well 1 1-20-34-7; this sand­
Erosion surface E4 is not exactly planar (Figs 3 -8) , stone has apparently been thrust unusually far to the
but regionally it may be taken to approximate a northeast (Figs 2 & 5). With this exception, the
plane. If so, the E4 to E5 isopach can be interpreted eastern edge of the sand body is remarkably straight.
as showing the on-field topography of E5 RSE, The sand body is characterized by a very sharp
 Cardium Formation incised valley 57

base, and blocky to blocky and serrated gamma-ray

Facies A: structureless sandstones
and resistivity well-log signatures (Figs 3 - 8) ; these
will be termed 'on-field' signatures. They contrast This facies consists of sandstones that lack any
with off-field wells (e .g. Fig. 3, wells 7-33-36- 10, physical or biological structures (Figs 1 5 - 17), or
7-26-36- 10 and 7-36-36- 10) , which are characterized that are dominantly structureless but contain parallel
by a gradual sandier upward signature between E4 lamination and/or ripple cross-lamination in the
and E5 and a capping sand body that is only a metre uppermost part of the bed . Grain size is mostly in
or two thick. the fine-grained sand range, with a few beds of very

FACIES OF THE
RICINUS ALLOMEMBER

The Ricinus sand body contains a suite of facies

that are quite unlike those found in any other part of
the Cardium. Five facies were recognized originally
(Walker, 1985 ) , and this scheme is modified and
simplified here.

Fig. 16. Structureless sandstone. The absence of structure

Fig. 15.Structureless sandstones with no apparent appears to be primary - there is no evidence of the
bedding surfaces and no mudstone partings. B =bottom, destruction of former structure by dewatering or
T =top. Core 3 in. (7 .6 em) wide, each core sleeve 2.5 ft bioturbation. Core 3 in. (7. 6 cm) wide. Well 8-10-34-8,
(about 75cm) long. Well 8 -10-34-8, 8109-8124 ft. 8115 ft.
58 R.G. Walker

Structureless sandstones:
percent of core
... .
100 ••• •• . .

6 u
Structureless sandstones:
average bed thickness 1m)
5

• .
.
.
.
• ...
.. . .
. . .. .
. .
• .

0
33 34 34 35 36 36 37

Southeast Northwest

Fig. 17.
Distribution of facies A structureless sandstones
from T33 northward to T37.

fine-grained sand and even fewer of medium-grained

sand. X-radiographs have not been made of these
sandstones, but it is believed that the absence of
sedimentary structures is real. Structures are readily
observed in other sandstone facies of similar grain
size (Figs 18- 20) . It seems reasonable to infer that
where the core shows absolutely no indication of
structure (Figs 15 - in the four core sleeves on the
right - & 16) , there is probably nothing there. Fig. 18. Sandstone showing crude parallel stratification at
The percentage of facies A beds in core varies the base, well-developed parallel lamination in the middle,
and ripple cross-lamination at the top. Core 3 in. wide
from zero to 100, but there is no trend from south to
(scale in em). Well 11-18-33-7, 8541 ft.
north (Fig. 17). Because of the intense fracturing,
particularly in the southern part of Ricinus, it is
difficult to estimate individual bed thicknesses.
However, several cores contain sandstones that are broken or cut. Nevertheless, there appear to be
unfractured, contain no internal grain-size breaks, at least four single beds thicker than 4 m (in wells
have no mudstone partings, and have no layers of 1 1-1 8-33-7 , 7-36-33-8, 2-1-34-8, 14-27-36-9), with two
eroded mud clasts (Fig. 15 - note that there is one more reaching thicknesses of 6.3 (7-28-34-8) and
mudstone parting near the top of the fourth sleeve 6.7 m ( 16- 10-34-8). There is an overall tendency for
from the left). Possibly some of these criteria for the thicker beds to be concentrated in the southern
multiple beds have been lost in places where core is part of Ricinus; the thinner ( < 1 m) beds occur
 Cardium Formation incised valley 59

Fig. 19. Almost every piece of core in these boxes shows

parallel lamination. The individual pieces of core are cut,
obscuring possible bed contacts. There are no angular

B =bottom, T =top. Core is 3 in. ( 7. 6 cm) wide, and each

divergences, intersections or truncations of lamination.

core sleeve is 2. 5 ft ( about 75em) long. Well 14-2-34-8,

7726-7740 ft.
Fig. 20. Parallel lamination, commonly emphasized by
finely comminuted organic material. Well 14-2-34-8,
7708ft.
throughout the sand body but are more common
in the north, particularly in Townships 36 and 37
(Fig. 17). body; the average thickness is about 50 em (range
These thinner beds have sharp bases marked by a 10- 135 cm , n= 153 beds).
jumble of eroded mud clasts and/or a concentration Trace fossils are very rare in this facies, and are
of coarse grains and granules of chert. Grading restricted to Ophiomorpha.
occurs in about half of the beds, either because
of an upward disappearance of the coarse grains
Facies B: parallel-laminated sandstone
and granules (coarse tail grading) , or because of a
progressive upward fining throughout the bed (dis­ These sandstones are characterized by parallel
tribution grading) . Internally, the beds are domi­ lamination (Figs 19 & 20); the stratification is
nantly structureless, but some show the development characterized by subtle changes in grain size and by
of crude parallel lamination toward the top. About very thin carbonaceous laminae (Fig. 20) . I ndividual
one-third of the beds have a single layer of ripple beds appear to be as thick as 2 m (maximum 6 m) in
cross-lamination at the top. The most common the southern part of the sand body, and thin north­
association of structures is Tbe (Fig. 18), with a few ward to an average of about 60 em . The percentage
beds showing the succession Tal>e· These beds are of parallel-laminated sandstones in the cores shows
much more common in the northern part of the sand a very weak tendency to decrease northward (Fig.
60 R.G. Walker

21).Grain sizes are consistently in the fine-grained 100

sand range. Trace fossils are rare , and are restricted
to Ophiomorpha.
80 Parallel-laminated sandstones
percent of core
60 •

Facies C: interbedded bioturbated sandstones 40

and mudstones
20 . .
..

This facies consists of alternations of sandstones . .

.
·- .
(thickness typically·1 -3 em) and mudstones (thick­ 0 I
·-···· I I
··-····-·-·--· ·-·.· --
nesses a few millimetres to about 1 em) (Figs 22 & •
23). I t occurs throughout the sand body (Fig. 9), but 6
Parallel-laminated sandstones
is thickest (10-15 m) in the central area (T34 R8)
average bed thickness (m)
(Fig. 11). Contacts with the other sandstone facies 5
(structureless, parallel-laminated) tend to be gra­
dational , and granules and pebbles of chert are rare 4
(compare with facies D, below) .
The sandstones are characterized by small-scale 3
ripple cross-lamination. Where ripple forms can be •
seen, they are commonly asymmetrical, suggesting 2 ..
current ripples rather than wave ripples. The facies •

is not intensely bioturbated; the commonest trace

fossil is Planolites (Fig. 23), with some examples of
•
•

·. ·.
.
. ;, ..
••

Ophiomorpha, Arenicolites , Conichnus and Rosselia 0 �---.----.--,,--r-

33 34 34 35 36 36 37
(Fig. 22 - the Rosselia in sleeve 12, counting from
Southeast Northwest
the right, was figured by McEachern & Pemberton ,
1992, p. 63, A). Fig. 21.Distribution of facies B parallel -laminated
sandstones from T33 northward to T37.

Facies D : bioturbated mudstones

clasts rarely occur in distinct beds - they are
This facies is less sandy and more thoroughly bio­ more commonly bioturbated into the sandy mud­
turbated than facies C (Figs 24 - 26). It contains a stone. The clasts occur throughout the field, from
larger suite of trace fossils; the commonest traces the southernmost (11-21-34-8) to northernmost
are simple, horizontal sand-filled tubes about 1 em ( 12-34-36-9) occurrences of the facies.
diameter (Planolites), with Helminthopsis (Fig. 27). Where the lower contact of this facies is seen in
Zoophycos (Fig. 27), Anconichnus, Skolithos, core, it is always sharp, and marked by the incoming
Ophiomorpha, Thalassinoides , Teichichnus, Aste­ of chert granules and pebbles (Figs 24, 25 & 28),
rosoma, Terebellina and Rosselia. together with rip-up clasts of mudstone and siderite.
Scattered chert granules and pebbles are com­ Good examples occur in wells 11-21-34-8 (1 mm
monly associated with the bioturbated sandy mud­ chert grains; Fig. 24), 16-30-34-8 (3 mm grains) ,
stones (Fig. 26), and they occur throughout the 2-34-36-9 (with chert pebbles up to 1 . 4cm),
thickness of the facies in many wells. The coarser 12-34-36-9 (1.3 em pebbles; Fig. 25) and 2-4-37-9

Fig. 22. ( Opposite. ) Well 12-19-35-8 boxed with the base of the well at the lower right; B =bottom, T =top. Note the
sharp base of the Ricinus sandstone (BASE R), overlain by parallel-laminated sandstone that grade into facies C, the
interbedded bioturbated sandstones and mudstones (sleeves 11-14 from right). Toward the top of this facies, it becomes
interbedded with parallel-laminated sandstones (sleeves 14 and 15), and grades up into structureless sandstones (top five
sleeves). Core is 3 in. (7.6 cm) wide, and each core sleeve is 2.5 ft (about 75cm) long. Well 12-9 -35-8, 8990-9048 ft. Note
Rosselia (ROSS) in sleeve 12 - this example was figured by McEachern & Pemberton, 1992. During core photography,
handwritten labels with way-up arrows, well locations and the words 'Ricinus' and 'bottom' were pinned on to the core
boxes. Thus 'Ricinus bottom' refers to each photograph of three or four core boxes, not to the bottom of the Ricinus
Allomember. The base of the Ricinus Allomember is marked by the printed letters BASE R.
62 R.G. Walker

and pebble lags mark the very sharp contact with

underlying facies. These lags are normally only
one to a few centimetres thick, with chert granules
and pebbles up to 4 cm diameter, and abundant
sideritized mud clasts. These basal lags occur from
the southern to northern parts of the sand body, but
there is no systematic distribution of clast sizes.
Conglomerates also mark the top of the Ricinus
Allomember, where they occur as lags on the E5
TSE surface (Fig. 25) . This surface truncates the
Ricinus Allomember, and the conglomerates
actually belong to the Carrot Creek Allomember.
These conglomerates have been cored only in three
wells (7- 19-35-8, 12- 18-36-8 and 12-34-36-9) ; they
are 1 .2-2.5 m thick and have chert pebbles up to
4 cm diameter (in 12- 18-36-8).
In T36 R9 (particularly in land sections 24 -27)
conglomerates are abundant from bottom to top of
the Ricinus sand body (Figs 26 & 29). The coarser
clasts occur at the bases of graded sandstone beds,
or form conglomerate beds up to about 1 m thick,
with chert pebbles 1 -2 em diameter. A few pebbles
reach 3 - 4 cm, with a maximum of 7.5 cm diameter
(in 3-27-36-9). These bedded conglomerates are
structureless to crudely stratified, and are inter­
bedded with structureless or parallel-laminated sand­
stones, and with thin mudstone partings (Fig. 29).
The scattered granules and pebbles associated
with the bioturbated mudstone facies have been
described above. They are typically mixed into the
mudstones by bioturbation, or may fill distinct bur­
rows (Fig. 30) .

LATERAL AND VERTICAL

FACIES DISTRIBUTIONS

Sandstones
Fig. 23. Interbedded bioturbated sandstones and
mudstones, showing abundant Planolites (P), and a The Ricinus Allomember trends NW-SE, with a
possible example of Zoophycos (Z) at the bottom of the maximum thickness of about 25 m in T34 R8 (Fig.
photograph. Note ripple cross -lamination within the 1 1 ) . It thins against the eastern wall of the valley
sandstones. Well 12-19-35-8, 9016 ft.
(Fig. 1 1) but the western margin is poorly constrained
by data and disrupted by extensive thrusting. Within
( 1 .7 em pebbles; Fig. 28). Thus the coarser clasts at the main sand body, the sandstone facies (structure··
this sharp contact become generally fi n er grained less, and parallel-laminated) show no preferred
southward. vertical trends. Some wells are almost exclusively
structureless, others almost exclusively parallel··
laminated. Most wells show interbedding of these
Facies E: conglomerates
two facies. There is a weak tendency for the average
Conglomerates and scattered pebbles occur through­ bed thickness in both facies to decrease northwest­
out the Ricinus Allomember. At the base, granule ward (Figs 17 & 2 1 ) , but there is almost no change
 Cardium Formation incised valley 63

Fig. 24. This core is boxed top-to-left; B =bottom, T =top. The parallel-laminated sandstone facies is overlain sharply by
the bioturbated mudstone facies. The sharp contact is marked by the incoming of chert granules (CH). The bioturbated
mudstones contain Zoophycos (Z; it first appears near the base of the facies), Helminthopsis (H), Rosselia (R),
Thalassinoides (TH) and Teichichnus (TE). Chert granules and pebbles are dispersed throughout the bioturbated
mudstone facies, but become abundant and coarser toward the top (in the sandier core sleeve on the left). Core is 3 in.
(7.6 cm) wide, and each sleeve is 2.5 ft (about 75 em) long. Well 11-21-34-8, 7022-7052 ft. See comment about handwritten
label 'Ricinus bottom' in caption of Fig. 22.

along the field in the percentage of each facies in relationship can be interpreted to imply partial valley
core (Figs 17 & 2 1 ). filling (more sand in the centre, less toward the
margin) before the muddier facies C and D were
deposited as a sheet that pinched out toward the
Mudstones
valley wall.
Because mudstones form an important part of many The wells that contain facies C or D appear to
estuary fills, the position of Facies C and D within have no preferred locations within the field. This
Ricinus was investigated in detail. These facies may be due to thrusting, particularly on the western
occur throughout the field (Fig. 9) , with a maximum side. Alternatively, there is some evidence that the
thicknesses in T34 R8-9. An isopach map (not distribution of facies C and D is masked by erosion
included here) of the interval between the base of of E5 TSE into the Ricinus sand body. Erosion of
the incised valley and the base of facies C or D facies C and D is strongly suggested in Fig. 7 ,
shows that in the valley axis these facies occur about between wells 3-19-33-7 and 1 1-18-33-7, and in Fig.
12 m above the base. However, toward the north­ 9 (wells 7-1 8-36-8 and 10- 18-36-8, and at the northern
eastern margin , the height above the base decreases end in wells 13-26-36-9, 12-34-36-9 and 2-4-37-9).
progressively to about 4 m (Figs 7 & 8). The isopachs The patchy distribution of facies C and D may also
are parallel to the trend of the valley wall. This be due to original depositional pinchout against the
64 R.G. Walker
 Cardium Formation incised valley 65

eastern valley wall, as suggested in Fig. 7 (between The facies overlying the valley-floor scour at
wells 3-19-33-7 and 9-19-33-7) and Fig. 8 (between Ricinus are completely different from those over­
wells 16- 16-34-8 and 10-22-34-8) . Also , it is possible lying E4 and ES elsewhere. At Ricinus, the typical
that some deposits of facies C and D were eroded facies are thick, structureless and parallel-laminated
during subsequent filling of the Ricinus valley, and sandstones. Elsewhere, the E4 and ES incised scours
before the development of ES TSE . This is suggested are overlain by up to 20 m of conglomerate (Carrot
in Fig. 8, where facies C and/or D in well 9-16-34-8 Creek Allomember; Bergman & Walker, 1987,
might possibly be scoured by the sandstone with a 1988 ) , or coarse and pebbly sandstones with sets of
blocky well-log response in 7-16-34-8. cross-bedding 1 0 cm or more thick (Burnstick Allo­
Although relatively few well-logs display the member; Pattison & Walker, 1992). The coarse
deeply indented pattern shown in Fig. 9 (indicative facies in the Carrot Creek and Burnstick Allo­
of facies C and/or D ) , and despite the limited core members are interpreted as transgressive shoreface
control , it seems likely that these facies originally deposits. The structureless and parallel-laminated
extended from the northern to southern parts of the sandstones at Ricinus are found nowhere else in the
field as a sheet. This sheet averages about 4 m in Cardium.
thickness, and seems to pinch out against the eastern Typical open-marine Cardium deposits consist
wall of the valley. The western margin is unknown. of mudstones, and wave-rippled and HCS sand­
The mudstones of facies D, with their fully marine stones interbedded with bioturbated mudstones.
trace fauna, are apparently limited to the northern There are no definite HCS sandstones at Ricinus.
and central parts of the field (2-4-37-9 southward to Common and abundant ichnospecies in the open­
1 1-21 -34-8) . They have very sharp, and probably marine mudstones include Chondrites, Gyrochorte,
erosive bases in wells 1 1-21-34-8, 16-30-34-8, Palaeophycus, Planolites, Rhizocorallium , Thalas­
2-34-36-9 and 12-34-36-9. sinoides and Zoophycos (Pemberton & Frey, 1984,
p. 282) . This is a different suite from that found in
the mudstones at Ricinus, where Rosselia is much
COMPARISON WITH more abundant, Palaeophycus is rare and Chondrites
OTHER CARDIUM ALLOMEMBERS was not observed.
Finally, the one prograding shoreface in the
The major differences between the Ricinus ABo­ Cardium , the Kakwa Allomember, is dominated by
member and other Cardium allomembers are swaley cross-stratification (Plint & Walker, 1987 ) .
reviewed here. Because all of the facies at Ricinus Swaley cross-stratification has not been identified at
are apparently marine, I will not make any com­ Ricinus.
parisons with the non-marine Musreau Allomember.
There are many asymmetrical scoured surfaces
in the Cardium. The scours are linear, oriented INTERPRETATION OF RICINUS
NW-SE, and have gentle dips to the southwest and
steeper dips to the northeast. Surfaces E4 and ES
Could Ricinus be an incised shoreface?
are typical (Walker & Eyles, 199 1 ; Pattison &
Walker, 1992) . Ricinus is also a linear scour oriented The orientation of the Ricinus sand body suggests
NW-SE, but incomplete data for the southwestern the possibility that it is an incised shoreface , similar
side preclude any analysis of its symmetry. to those at Garrington , Crossfield and Caroline

Fig. 25. ( Opposite. ) Sharp base of Ricinus sand body (BASE R) on underlying mudstones of the Raven River
Allomember. Note lag j ust above base (L). The sandstones are parallel-laminated and structureless, and are very abruptly
overlain by the bioturbated mudstones facies in the upper half of the figure, at CH. Note sudden incoming of chert pebbles
(CH) up to 1 em in diameter. The top of the Ricinus sand body is marked by erosion surface E5 TSE (E5 in photograph),
which is characterized by eroded sandstone clasts at the base (j ust above E5 label in photograph). The lag on E5 is about
15em thick, and is overlain by transgressive laminated mudstones (the 'laminated blanket' of Walker, 1983; LB in
photograph). B =bottom, T =top. Core is 4 in. (10. 2em) in diameter, and each core sleeve is 2ft (about 60cm) long. Well
12-34-36-9, 9254-9307 ft. See comment about handwritten label 'Ricinus bottom' in caption of Fig. 22. Core is boxed top­
to-right and is continuous from upper to lower panels in the figure.
66 R.G. Walker

Fig. 26. Bioturbated mudstone facies with abundant scattered chert granules and pebbles (small white specks throughout
photograph), and a prominent conglomerate bed (CGL). B =bottom, T = top. Core is 3 in. (7.6 cm) wide and each core
sleeve is 2.5 ft (75 em) long. Well 3-32-34-8, 9040-9066 ft. See comment about handwritten label 'Ricinus bottom' in
caption of Fig. 22.

(associated with E4; Pattison & Walker, 1992), and Ricinus, the mudstones appear to pinch out against
at Carrot Creek (associated with E5; Bergman & the bounding discontinuity on the eastern side. If
Walker, 1987, 1988) . However, the abundance of this discontinuity were interpreted as the initial trans­
thick beds of structureless sandstone at Ricinus is gressive surface underlying a Ricinus shoreface, the
most atypical of a shoreface, where the sand is mudstones would downlap on to this surface . This
normally being constantly reworked by longshore would be an almost impossible situation for a minor
currents, rip currents and waves. The Carrot Creek transgression of a shoreface ; the mudstones ought to
shorefaces are characterized by conglomerates, and pinch out against the western wall and continue
the Burnstick shorefaces by cross-stratified coarse­ seaward parallel to the eastern bounding discon­
and granule-sandstones, and conglomerates. The tinuity. All of these arguments, taken together,
prograding Kakwa shoreface is dominated by swaley suggest that Ricinus is not any form of incised shore­
cross-stratification. None of these features is present face, despite the orientation of the sand body.
at Ricinus.
Mudstones 4-5 m thick are not characteristic of
Origin of the valley
shorefaces, although it could be argued that the
mudstones at Ricinus formed during a minor trans­ The comparisons of Ricinus with other parts of the
gression of the postulated shoreface . However, their Cardium, along with discussions of a possible shore­
geometry is incompatable with this hypothesis. At face origin , suggest that Ricinus is (i) not an incised
 Cardium Formation incised valley 67

Fig. 27. Zoophycos (Z) and robust Helminthopsis

(H; flattened black marks) in well l l -21-34-8, 7035 ft. Core
is 3 in. (7.6 cm) wide.

transgressive shoreface deposit, (ii) not a prograding Fig. 28. Sharp and erosive basal contact of bioturbated
shoreface , and (iii) not an open-marine HCS sand mudstone facies D on underlying sandstones in well 2-4-37-
body like those in the Raven River Allomember 9, 9326 ft. Chert pebbles are up to 1. 7 em diameter, and
(Walker & Eyles, 1988) . The basal incision (E5 siderite intraclasts are larger than 2 em. Note possible
Zoophycos (Z) only a few centimetres above base of the
RSE) appears to define a NW-SE trending valley at bioturbated mudstone facies. Core 3 in. (7 .6 em) diameter.
least 24 m deep, with a marine fill.
I now reject my earlier suggestion that the valley
was cut by turbidity currents (Walker, 1985). The
Cardium contains mainly shallow-marine to shore­
line deposits, influenced by about eight major fluc­ pebbles and ripped-up mudstone clasts. There is no
tuations of relative sea-level . In this overall setting, preserved angle-of-repose cross-bedding in sets
the most likely origin for the Ricinus valley is fluvial , thicker than about 5 em, nor any other facies indica­
with incision during a falling stage of sea-level. The tive of fluvial deposition. It therefore appears that
valley was subsequently filled in an estuarine setting. any fluvial sediment in the valley was completely
The basal deposits of the Ricinus sand body consist reworked by waves during initial transgression of
of structureless and parallel-laminated sandstones, the valley. At the same time , the valley walls were
commonly with a lag of chert pebbles, siderite probably wave scoured and the valley widened. The
68 R.G. Walker

Fig. 29. Interbedded parallel-laminated sandstones (facies B, centre of figure), bioturbated sandstones and mudstones
(facies C), and conglomerates (facies E, in 2nd sleeve) in well 3-27-36-9, 8885-8914 ft. Chert pebbles are up to about 2 em
diameter. B =bottom, T =top. Core is 3 in. (7 .6 em) wide, and each core sleeve is 2.5 ft (75 em) long. See comment about
handwritten label 'Ricinus bottom' in caption of Fig. 22.

interpretation of the base of the valley as a regressive the structureless and parallel-laminated sandstones
surface of erosion (ES RSE) applies to the main decrease northward (Figs 17 & 2 1 ) . Third, conglom­
phase of valley erosion; if the floor and walls of the erates are best developed at the northern end of the
valley have indeed been modified during trans­ valley (T36 R9, land sections 24-27) , but they are
gression , the surface seen today is strictly a ravine­ not cross-bedded fluvial conglomerates. They are
ment surface, or surface of initial transgression interbedded with mudstones and sandstones with a
(IT) . marine trace-fossil fauna, suggesting that the pebbles
were being moved by waves in the longshore drift
system, and introduced into the marine end of the
P alaeogeography
estuary by storms. There are far fewer granules and
There are no unquestionable data that indicate flow pebbles associated with the thick structureless sand­
direction in the valley, and the problem of palaeo­ stones at the southern end of the valley. Finally, the
geography is compounded by the incomplete preser­ mudstones of facies D have a sharp base with a
vation of the valley; both ends are faulted. The conglomeratic lag. In the four wells with data avail­
following indicators, taken together, suggest that able , the clast size in the lag decreases progressively
the sea lay to the north. First, the valley appears to southward , from 1 .7 em to about 1 mm. This also
widen northward (Fig. 10), and the eastern wall suggests marine transgression from north to south,
becomes less steep. Second, the bed thicknesses of and hence an open sea to the north.
 Cardium Formation incised valley 69

are both dominated by dune bedforms, and presum­

ably would be dominated by angle-of-repose cross­
bedding if these estuaries were preserved in the
geological record. Such cross-bedding is absent at
Ricinus.
Wave-dominated estuaries are characterized by
a tripartite facies distribution consisting of (i) an
upstream river-dominated bayhead delta, (ii) a
central basin (turbidity maximum) where mudstones
are deposited, and (iii) a downstream, marine­
dominated area where sand is introduced into the
estuary from the sea. This tripartite subdivision is
not seen at Ricinus, although this might be due to
incomplete preservation of the entire length of the
estuary. There is no suggestion of a bayhead delta
prograding northward, or marine sands prograding
southward; in both cases, sandier upward facies
successions would be expected (but are not present)
as the sands built on to central basin mudstones.
Wave-formed or wave-influenced sedimentary
structures (symmetrical ripples, hummocky cross­
stratification) are rare at Ricinus. As in many modern
estuaries, this is probably due to partial closing of
the estuary at the seaward end, and limited pen­
etration of large waves from the sea. Wave base in
the Ricinus estuary may therefore have been only a
few metres below sea level.
The sandstones at Ricinus commonly contain
evidence of rapid deposition from strong currents,
with no subsequent reworking. Rapid deposition is
suggested by the thick, structureless sandstones;
some of the flows may have deposited beds thicker
than 4 m, mostly at the southern end of the valley
Fig. 30.Chert granules and pebbles scattered within the (T33 and T34). Other flows continued along the
bioturbated mudstone facies. Note granules filling distinct length of the valley to deposit structureless sand­
burrows (centre), and Zoophycos (Z) in upper part of stones 50- 150 em thick in T36 and T37 (Fig. 17).
photograph. Well 3-32-3411, 9050ft. Core is 3 in. (7.6 cm)
wide. The common occurrence of parallel lamination
suggests that flows were rapid. The presence of Tahc
and Tbe beds (Fig. 18) indicates that some of the
rapid flows waned during deposition. Flow distance
from the southernmost fault in T32 and T33 to the
Deposition in the Ricinus estuary
northern end of the valley (T37) is more than 45 km.
Deposition in the Ricinus estuary is not closely The beds that contain only parallel lamination also
comparable with that of any modern estuaries, imply rapid flows, with single depositional events
nor does it conform to any existing estuary models forming beds up to 3 m thick in the south, and about
( Dalrymple et al. , 1992; Reinson , 1992). The absence 1 m thick in the north. The preservation of structure­
of angle-of-repose cross-bedding in sets thicker than less, parallel-laminated, and Tabc and Tbc beds indi­
about 5 em indicates the absence of two- and three­ cates that there was no wave reworking after the
dimensional dunes in the estuary. Tide-dominated original deposition of the sand. However, the estua1y
estuaries (e.g. the Gironde (Allen , 199 1 ) and the was not necessarily deep because these facies have
Bay of Fundy ( Dalrymple et al . , 1990)) and river­ been preserved; their preservation may simply reflect
dominated estuaries (e.g. the Mgeni (Cooper, 1993)) the proposed shallow depth to wave base.
70 R.G. Walker

Because there is no evidence for fluvial or tidal well 16-1 6-34-8 (Fig. 8) have been cut out in well
facies in the estuary, and no evidence for wave 7-16-34-8 (Fig. 8). Examination of sandstone cores
reworking, I suggest that the structureless, parallel adjacent to preserved mudstones showed no lag
laminated, Tabc and Tbc beds were emplaced by horizons or facies changes that might indicate
density underflows generated at a bayhead delta scouring.
(not observed) in the southern end of the estuary. Estuarine deposition was terminated by a final
Similar underflows appear to have been generated in rise of relative sea-level , and the estuary deposits
the estuary of the Congo (Zaire) River (Heezen were truncated by ES TSE. There appears to be
et al. , 1964) , and finer grained underflows originated some topographical relief on this surface, with
from the delta of the Colorado River in Lake Mead deeper scouring down to the west (Figs 3 - 5 & 7).
and flowed the entire length of the lake (about The ES TSE surface is typically overlain by a thin lag
lOO km; Grover & Howard, 1938). of chert pebbles (Figs 13, 25 & 29) , and then by
The mudstones of facies C and D appear to occur transgressive mudstones with thin silty laminations ­
as a sheet about 4-5 m thick that spreads from end the 'laminated blanket' of Walker, 1983.
to end of the valley. In many places, the structureless
and parallel-laminated sandstones grade vertically
into the interbedded bioturbated sandstones and VALLEY ORIENTATION
mudstones of facies C (Fig. 22) . However, the mud­
stones of facies D are characterized by a very sharp The Ricinus valley trends parallel to the incised
base, and a basal layer of chert pebbles and mudstone shorefaces in the Burnstick (Pattison & Walker,
clasts (Figs 24, 25 & 29) . The trace-fossil fauna of 1992) and Carrot Creek (Bergman & Walker, 1987,
these mudstones is fully marine. I suggest that after 1988) Allomembers (Fig. 1 ) ; an estuary would be
deposition of the lower sandstones, the estuary was expected to be perpendicular. The valley is incised
rapidly flooded, with wave scouring of the substrate, into black mudstones and open marine HCS sand­
cut-off of sediment input from the fluvial (southern) stones and bioturbated mudstones of the Raven
end of the estuary, introduction of pebbles by storms River Allomember, and it is the only known incised
from the open (northern) end of the estuary, valley in the Cardium (both in subsurface and out­
and quiet deposition of mud. Wave scouring in crop). The orientation of the Ricinus valley seems to
the estuary could have increased during flooding require that the depositional strike in the basin
( deepening ) if the barrier at the seaward end was changed from NW -SE (incised transgressive shore­
removed. Deposition of mud was interrupted faces associated with E4) , to N E - SW (perpendicular
periodically by storms, introducing sand and pebbles to the Ricinus valley), and back to NW-SE (incised
at various stratigraphical horizons within the transgressive shorefaces associated with ES).
mudstones. One possibility that has been considered (and
The relationship between the mudstones of facies rejected) is that the Ricinus valley is a strike-parallel
C and D is not clear. The log response suggests that tributary of a main valley that trends SW- NE. The
both facies tend to occur at the same stratigraphical marine end of Ricinus appears to be at the northern
horizon (Fig. 9); however, facies C appears to have end, so the postulated main valley would most likely
gradational contacts with facies A and B, whereas be north of Ricinus. No such valley is known. It also
facies D has a distinctly erosional relationship with seems unlikely that the gravel that occurs throughout
underlying facies. None of the cores shows facies C the section at the northern end of Ricinus could
and D occurring together, and there is not enough have been emplaced into the main valley, and thence
data to determine their possible lateral relationships. into the Ricinus valley via a right angle turn. The
After deposition of the sheet of mudstones, sand orientation of Ricinus remains enigmatic!
was again introduced into the estuary, and the same
facies occur both above and below the mudstones.
Because of the patchy distribution of the mudstones, CONCLUSIONS
there is a strong possibility that a falling stage of
relative sea-level caused reintroduction of sand into 1 The sand body at Ricinus is defined by bounding
the estuary and local scouring of the mudstones. discontinuities below and above. The lower discon­
This is tentatively suggested in Fig. 8. The depth of tinuity defines an incised valley at least 50 km long,
scouring would be at least 15 m if facies C and D in 24 m deep and 10 km wide. The valley trends
 Cardium Formation incised valley 71

N W- S E , parallel t o regional tectonic trends and of the better ideas now incorporated into allostra­
parallel to regional shoreline trends in the Cardium. tigraphy and sequence stratigraphy.
2 The valley is filled with structureless and parallel­
laminated sandstones. There are also mudstones
that form a layer 4-5 m thick that appears to extend AP PENDIX 1
as a sheet along the entire length of the valley. The
sheet probably pinches out against the eastern wall
Cross-sections
of the valley. The mudstones contain a fully marine
trace-fossil fauna. The choice of a datum for the cross-sections is
3 The geometry of the valley and distribution of difficult. Possible datums above Ricinus tend to
facies suggest that the open sea lay to the north. drape over the sand body. The 'Cardium zone
4 There is no evidence of fluvial or tidal processes , marker' (E7 in the cross-sections) commonly used
and wave reworking of the sandstones is rare. At by industry and some other workers is a very bad
least part of the fill (the structure less sandstones choice because it is a mappable erosion surface with
and the Tahc and Tbe beds) appears to have been more than 40 m of relief (Wadsworth & Walker,
emplaced by density currents that maintained 1 99 1 ; Figs 4 & 6-8). In this paper I have chosen a
sand in suspension, and then deposited it very group of resistivity signatures A - D in Figs 3-8 that
rapidly. The sand was not reworked after its initial appear to define four sandier upward successions
deposition. This suggests that the valley was domi­ immediately below E4. The top of the 'black blanket'
nated by input of large amounts of sand from the (Figs 4-8) is almost parallel to the sandier upward
fluvial end. This sand appears to have slumped, successions. Together, horizons A - D can be taken
perhaps on a bayhead delta, and then travelled to approximate an original , almost fiat, sea-floor,
along the estuary as a density underflow. into which the Ricinus valley is incised.
5 A phase of marine transgression terminated sand All 71 cores have been measured and photo­
input and led to the deposition of a valley-wide layer graphed. All of the off-field well-logs west of Ricinus
of marine mudstone. A subsequent drop of relative have been used where the Cardium is present and
sea-level led to renewed sand emplacement, prob­ recognizable , but some cannot be used because the
ably with local scouring of the mudstones. rocks west of Ricinus are too deformed. Enough
6 The valley is interpreted as a lowstand incision wells north and south of the valley (Fig. 1) have
formed by a river. The fill consists of a variety of been used to define the ends of the valley, and most
facies, but broadly can be termed estuarine. of the wells immediately east of the valley (between
7 The orientation of a major valley 50 km long, Ricinus and Caroline) have been used. Only a small
parallel to depositional strike, remains enigmatic. proportion of Caroline wells are shown on the map
(Fig. 1 ) .
The cross-sections show gamma-ray and resistivity
ACKNOWLEDGEMENTS well-logs, with cored intervals shown as black bars.
Depths are mostly in metres, with depth markers
I began work at Ricinus in 1982 whilst on sabbatical shown at 50-m intervals, but a few older wells are
leave at Amoco Canada. I thank the company for shown in feet (numbers greater than 6000, with
their support. Work on this paper commenced in depth markers at 100-ft intervals) . In the northern
1993 , and I thank David James and Wascana Energy and central parts of the field , a few cores show
I nc. for access to well logs. The research has been dipping beds. However, the generally parallel nature
financially supported by the Natural Sciences and of E 1 , successions A - D , and the top of the black
Engineering Research Council of Canada, in the blanket, shows that thicknesses in most wells are not
form of Operating and Strategic Grants. I particu­ seriously distorted by variations in dip.
larly thank James McEachern for his help in ident­ The southern part of the field appears to be more
ifying and explaining the trace fossils. And finally, extensively thrusted, with more fracturing of the
none of the work would have been possible without sandstones in core. Steeply dipping beds in section
the help of Harold Reading; it was Harold who first D can be identified in cores, and in those parts of the
emphasized to me in 1961 the difference between well-logs where thicknesses appear to be expanded.
gradational and abrupt facies contacts, following the I have assumed that in section D (Fig. 6) the eastern­
thinking of Johannes Walther and anticipating many most well ( 1 1-25-33-7) is undeformed because the
72 R.G. Walker

thicknesses of various stratigraphical intervals are resolution sequence stratigraphy'. This subdiscipline
very similar to those in the other cross-sections. is converging on allostratigraphy, but unfortunately
Other wells in section D have been drilled into retains the concept that units are bounded by uncon­
dipping rocks, and I have therefore reduced the well formities, and that the strata bounded by the uncon­
logs so that the thicknesses of units A - D below E4 formities are 'genetically related' .
are roughly constant across the section. I n the Cardium (Fig. 2) , the only type 1 sequence
boundary, with valley incision, is E5 RSE at Ricinus.
If I ignore Ricinus for a moment, it is possible to
AP P ENDIX 2 define a 'sequence' between E4 and E5, by arguing
that both are unconformities, and that both
developed initially during subaerial exposure (E4,
Allostratigraphy
see Pattison & Walker, 1992; E5, see Walker &
Allostratigraphy first appeared m the North Eyles, 199 1 ) . The implication is that the rocks
American Stratigraphic Code in 1983 (NACSN , between E4 and E5 are 'genetically related' as a
1983 ) . It does not appear to have been used in the sequence. This sequence consists of coarse sand­
Western Interior Seaway of North America until stones and conglomerates deposited in a trans­
1987 (Plint et a!. , 1987, p. 366) , with reference to the gressing shoreface (Burnstick member, Fig. 2),
unconformity-bounded members of the Cardium overlain by black mudstones (the 'black blanket',
Formation proposed by Plint et al. ( 1 986) . An Fig. 2) and very fine- to fine-grained hummocky
allostratigraphical unit 'is a mappable stratiform body cross-stratified sandstones in the upper part of the
of sedimentary rock that is defined and identified on Raven River Allomember. There is presumably a
the basis of its bounding discontinuities' (NACSN, maximum flooding surface immediately below or
1983 , p . 865). I suggest (following the definitions j ust within the black blanket. I suggest that on
of sequence stratigraphy) that this definition be sedimentological grounds, there is absolutely no
extended to read ' . . . bounding discontinuities and 'genetic relationship' between the transgressive con­
their correlative conformities' . The bounding dis­ glomerates and the overlying (highstand) black
continuities were not defined in the stratigraphical mudstones and HCS sandstones. Water depths were
code, but clearly include erosion surfaces (regressive very different, as were the wave climates, sediment
and transgressive surfaces of erosion) , condensed transport processes and grain sizes of available sedi­
horizons of reduced or zero deposition (commonly ment. The salinities may also have been different.
maximum flooding surfaces), and firmgrounds and Thus sequence stratigraphy is not the most appro­
hardgrounds. Allostratigraphy is descriptive, and priate technique for descriptive subdivision of a unit
does not involve any of the interpretive aspects of like the Cardium. Allostratigraphy does not involve
sequence stratigraphy. It can commonly be applied problems of defining terms such as unconformity ,
on a smaller scale than sequence stratigraphy (where genetically related, and evidence of subaerial­
a sequence is a 'relatively conformable succession of erosional truncation, and allows the Cardium to be
genetically related strata bounded by unconformities subdivided using bounding discontinuities (however
or their correlative conformities'; van Wagoner et they formed). The most useful bounding discon­
al. , 1990, p. 22) . An unconformity was defined as 'a tinuities are the transgressive surfaces of erosion
surface separating younger from older strata along (E4, E5 TSE, etc. ) , and the flooding surfaces that
which there is evidence of subaerial-erosional trunc­ j uxtapose relatively offshore quietly deposited mud­
ation, or subaerial exposure, with a significant stones on top of nearshore conglomerates (Fig. 25).
hiatus indicated' (van Wagoner et al. , 1990, p. 22) . The allostratigraphical units so defined correspond
In the Cardium , there i s absolutely n o 'evidence of to transgressive systems tracts (the Burnstick
subaerial-erosional truncation' on surfaces E5, E6, Allomember and its correlative conglomerate lag on
E6.5 and E7, although the surfaces are interpreted to E4) and highstand systems tracts (the Raven River
have initially formed subaerially. The 'evidence' Allomember) ; these systems tracts do contain
was entirely removed during erosive shoreface 'genetically related' strata.
transgression . The Ricinus sand body is also defined by two
Sequence stratigraphers are now recognizing the bounding discontinuities; E5 RSE and E5 TSE. It
need to subdivide many stratigraphical units on a is therefore defined in this paper as the Ricinus
finer scale, thus giving rise to discussions of 'high- Allomember of the Cardium Alloformation. It is
 Cardium Formation incised valley 73

genetically distinct both from the highstand deposits Alloformations and allomembers will commonly
of the Raven River, and the transgressive deposits be heterogeneous, encompassing both horizontal
of the Carrot Creek Allomembers. and vertical facies changes. Thus the concept
The scale of allostratigraphical subdivision, like of formally defining a 'type section' needs to be
that of lithostratigraphical subdivision , is left to the re-evaluated. Also, any formal definition of an allo­
discretion of the individual worker. In the Cardium , stratigraphical unit must include descriptions of the
the transgressive surfaces of erosion are complicated bounding discontinuities, so that they can be recog­
by the fact that in many places there appear to be nized by other workers.
stillstands of sea-level during the overall trans­
gression. Incised shorefaces are cut during these
stillstands. The shoreface is floored by the initial AP PENDIX 3
surface of transgression (IT), and is truncated by a
surface of resumed transgression ( RT) (Fig. 2,
P reparation of computer maps
Burnstick Allomember ) . However, the RT surface
will pass laterally into another IT surface if there is It has not been possible to restore all of the thrusts
another stillstand during the transgression (Fig. 2). that displace sandstones within the field. Conse­
Thus the two Burnstick sand bodies shown in Fig. 2 quently, some of the wells penetrate sandstones that
are defined by different bounding discontinuities, have been thrust northeastward, and some thickness
and hence would qualify as different allomembers. may be slightly exaggerated if the well penetrates
Such fine subdivision seems unnecessary at present; dipping strata. The effect of exaggerated thickness is
the erosional envelope beneath the sand bodies can reduced in the mapping because SU RFER calculates
be regarded as one bounding discontinuity, as can each grid-point using data from the 10 closest wells
the contact between the RT surfaces and the over­ using a kriging method. Thus the thickness distor­
lying transgressive mudstones. Thus the Burnstick tions are probably not great, but the maps should be
Allomember (and the Carrot Creek Allomember) stretched southwestward in the reader's imagination
consist of a series of separate sand bodies at slightly to allow for the foreshortening due to thrusting.
different stratigraphical horizons, but all developed It can be seen in Fig. 1 that the main sand body as
during one overall phase of transgression. defined by drilling density does not seem to be
My final comment on allostratigraphy concerns displaced laterally by the faults that define the
names. The stratigraphical code does not address southernmost three compartments. In my data base,
this problem - it simply states that 'the principles the locations of the wells in the northern compart­
and procedures for naming allostratigraphical units ment have been shifted about 4 km southwestward
are the same as those for naming of lithostratigraphi­ to bring them on line with the rest of the field; off­
cal units' (NACSN, 1983 , p. 867). The Cardium field wells to the southwest were also shifted, but
Formation is a formally defined lithostratigraphical off-field wells to the northeast were not relocated.
unit. Some workers have suggested that if an allo­
stratigraphical scheme is to be used, the term
Cardium Alloformation is forbidden because REFERENCES

Cardium has already been used. It is clear that if

existing names are banned, the number of names ALLEN, G.P. (1991) Sedimentary processes and facies in
the Gironde estuary: a recent model for macrotidal
will double if existing units are subdivided allo­ estuarine systems. In: Clastic Tidal Sedimentology (Eds
stratigraphically. This is ridiculous. I do not believe Smith, D.G . , Reinson. G.E. , Zaitlin, B . A . & Rahmani,
that the term Cardium Alloformation is in any way R.A.). Can. Soc. petrol. Geol . , Calgary, Memoir 16,
confusing. It is defined as those rocks between E l 29- 40.
BERGMAN, K . M. & WA L KER , R.G. (1987) The importance
and the top of the transgressive lag o n E7, and
of sea level fluctuations in the formation of linear con­
encompasses roughly the same rocks as the Cardium glomerate bodies: Carrot Creek Member, Cretaceous
Formation. It has the advantage of emphasizing a Western Interior Seaway, Alberta, Canada. J. sediment.
natural break in sedimentation, whereas the base of Petrol. , 57, 65 1 - 665.
the Cardium Formation is defined at the incoming of BERG MA N , K . M. & WALKER, R.G. (1988) Formation of
Cardium erosion surface E5, and associated deposition
sandstones at some arbitrary point within the first
of conglomerate; Carrot Creek field, Cretaceous Western
sandier upward succession. There is room for both a Interior Seaway, Alberta. In: Sequences, Stratigraphy,
Cardium Formation and a Cardium Alloformation. Sedimentology: Surface and Subswface (Eds James, D.P.
74 R. G. Walker

& Leckie, D.A.). Can. Soc. petrol. Geol . , Calgary, mation (Upper Cretaceous) at Seebe, Alberta. In: The
Memoir 15, 15-24. Mesozoic of Middle North America (Eds Stott, D.F. &
B o u M A , A.H. ( 1 982) Sedimentology of some Flysch Glass, D.J.) Can. Soc. petrol. Geol . , Calgary, Memoir
Deposits. Elsevier, Amsterdam, 168 pp. 9, 281 -304.
CooPER, J .A.G. ( 1993) Sedimentation in a river dominated PuNT, A.G. & WALKER, R. G. ( 1987) Cardium Formation
estuary. Sedimentology, 40, 979- 1 0 1 7 . 8. Facies and environments of the Cardium shoreline
DALRYMPLE, R.W., KNIGHT, R.J . , ZAITLIN , B.A. & and coastal plain in the Kakwa field and adjacent areas,
MIDDLETON , G. V . ( 1 990) Dynamics and facies model of northwestern Alberta. Bull. Can. petrol. Geol. , 35,
a macrotidal sand-bar complex, Cobequid Bay - Salmon 48-64.
River · estuary, Bay of Fundy, Canada. Sedimentology, PUNT, A. G., WALKER, R.G. & BERGMAN, K.M. ( 1986)
37, 577- 6 1 2 . Cardium Formation 6. Stratigraphic framework of the
DALRYMPLE, R.W. , ZAITLIN , B.A. & BoYD, R. ( 1992) Cardium in subsurface. Bull. Can. petrol. Geol. , 33,
Estuarine facies models: conceptual basis and strati­ 2 1 3-225 .
graphic implications. J. sediment. Petrol. , 62, 1 1 30- 1 146. PUNT, A. G., WALKER, R.G. & BERGMA N , K.M. ( 1 987)
GEOLOGICAL S U RVEY OF CANADA ( 198 1a) Gas Pools Cardium Formation 6. Stratigraphic framework of the
of Western Canada. Geol. Surv. Canada, Map 1558A, Cardium in subsurface: Reply. Bull. Can. petrol. Geol. ,
Ottawa. 35, 365-374.
GEOLOGICAL S u RVEY OF CANADA ( 1981b) Oil Pools REINSON, G.E. ( 1 992) Transgressive barrier island and
of Western Canada. Geol. Surv. Canada, Map 1558B, estuarine systems. In: Facies Models - Response to Sea
Ottawa. Level Change (Eds Walker, R.G. & J ames, N.P. ) .
GROVER, N.C. & HowARD, C.S. ( 1938) The passage of pp. 179- 194. Geological Association o f Canada.
turbid water through Lake Mead. Trans Am. Soc. civ. VAN WAGONER, J.C., MITCHUM, R. M. , CAMPION, K.M. &
Eng. , 103, 720- 790. RAHMANIAN , V.D. ( 1990) Siliciclastic Sequence Stra­
HAMBLIN, A.P. & WALKER, R.G. ( 1979) Storm-dominated tigraphy in Well Logs, Cores, and Outcrops. American
shallow marine deposits: the Fernie-Kootenay Association of Petroleum Geologists, Tulsa, Methods
(Jurassic) transition, southern Rocky Mountains. Can. in Exploration Series, 7 , 55 pp.
J. Earth Sci. , 16, 1673- 1690. WADSWORTH, J.A. & WALKER, R.G. ( 1 99 1 ) Morphology
HEEZEN, B.C., MENZIES, R.J . , SCHNEIDER, E. D . , EWING, and origin of erosion surfaces in the Cardium Formation
W.M. & GRANELLI, N.C.L. ( 1964) Congo submarine (Upper Cretaceous, Western Interior Seaway, Alberta)
fan. Bull. Am. Assoc. petrol. Geol. , 48, 1 126- 1 149. and their implications for rapid sea level fluctuations.
McEACHER N , J.A. & PEMBERTO N , S. G. ( 1992) lchnological Can. J. Earth Sci. , 28 , 1507- 1520.
aspects of Cretaceous shoreface successions and shore­ WALKER, R. G. ( 1983) Sedimentology and stratigraphy in
face variability in the Western I nterior Seaway of North the Caroline-Garrington area, Bull. Can. petrol. Geol. ,
America. In: Applications of Ichnology to Petroleum 3 1 , 2 1 3 -230.
Exploration; a Core Workshop (Ed. Pemberton, S. G. ) WALKER, R.G. ( 1 985) Cardium Formation at Ricinus Field,
Society of Economic Paleontologists and Mineralogists, Alberta: a channel cut and filled by turbidity currents in

NACSN ( N ORTH AMERICAN COMMISSION ON STRATIGRAPHIC

Tulsa, Core Workshop 1 7 , 57-84. the Cretaceous Western Interior Seaway. Bull. Am.

NoMENCLATURE ) ( 1 983) North American stratigraphic

Assoc. petrol. Geol. , 69, 1963- 198 1 .
WALKER, R.G. & EYLES, C.H. ( 1988) Geometry and facies
code. Bull. Am. Assoc. petrol. Geol. , 67, 84 1 -875. of stacked shallow marine sandier upward sequences
PATTISO N , S. A.J. & WALKER, R. G. ( 1992) D eposition and dissected by erosion surface: Cardium Formation,
interpretation of long, narrow sand bodies underlain Willesden Green, Alberta. Bull. Am. Assoc. petrol.
by a basinwide erosion surface: Cardium Formation, Geol. , 72, 1469- 1494.
Cretaceous Western Interior Seaway, Alberta, Canada. WALKER, R. G. & EYLES, C.H. ( 199 1 ) Topography and
J. sediment. Petrol. , 62, 292-309. significance of a basinwide sequence-bounding erosion
PEMBERTO N , S. G. & FREY, R.W. ( 1984) I chnology of storm­ surface in the Cretaceous Cardium Formation, Alberta,
influenced shallow marine sequence: Cardium For- Canada. J. sediment. Petrol. , 61, 473-496.
Spec. Pubis in!. Ass. Sediment. (1995) 22, 75-99

Gravelly shoreface and beachface deposits

B R U C E S . H A R T* L and A . G UY P Ll N T"!
*Geological Survey of Canada, Box 6000, Sidney,
British Columbia VBL 4B2, Canada; and
"I Department of Earth Sciences, University of Western Ontario,
London, Ontario N6A 5B7, Canada

ABSTRACT

Modern gravel-dominated shoreface deposits are poorly documented, especially from progradational
settings. This makes the interpretation of possible ancient analogues very difficult. In the absence of
unique diagnostic features, ancient conglomeratic shoreface deposits may be identified on the basis of:
(i) the inferred relationship between sedimentary process and deposit, based on sedimentary structures,
fabric and texture of the deposit, palaeocurrent data, etc . ; (ii) vertical and lateral successions of facies;
(iii) gross morphology of a deposit; and (iv) stratigraphical setting. Much more is known about gravelly
beachface deposits because modern analogues are easily accessible.
Several distinct 'end-member' types of gravelly shorcface/beachface deposits can be recognized:
(i) river mouth bar deposits, which show evidence of rapid deposition, scouring, reworking by waves and
(possibly) post-depositional failure; (ii) mixed sand-gravel systems in which sedimentary structures
within the sandstones allow determination of depositional environment; (iii) swash-aligned gravelly
systems dominated by ncar-perpendicular wave approaches, which show shore-normal facies zonation;
(iv) driji-aligned gravelly systems , which show evidence of pronounced longshore transport generated by
oblique wave incidence; and (v) transgressive lag deposits. Observations and interpretations of ancient
conglomeratic shorefore deposits raise process-related questions that have yet to be addressed by studies
of modern equivalents.

INTRODUCTION

It is commonly accepted that one of the most fruitful depositional environments must rely on interpret­
approaches to facies analysis involves comparison of ation of the processes which gave rise to the observed
modern sedimentary environments and potential lithofacies , and relating these interpreted processes
ancient analogues. This approach, based on the to stratigraphical context, partially analogous
uniformitarian principle that the present is the key modern environments, and the three-dimensional
to the past, forms the basis of most texts dealing morphology of the deposit.
with facies models and its roots can be traced back Progradational gravel-dominated shoreface and
for at least a century to Johannes Walther. Walther beachface systems are not common, and are rarely
advocated an 'actualistic' approach to the examin­ documented. Three 'end-member' settings for grav­
ation of ancient sedimentary deposits, but knew elly shorefaces may be identified. The first involves
that ancient facies might exist which have no direct areas of active erosion of coastal bedrock outcrops
modern analogues (Middleton, 1973) . In such cases, (e.g. Zenkovitch , 1967; Ogren & Waag, 1986). The
as explained by Reading (1986), reconstruction of second setting is in high-latitude areas that were
subject to Pleistocene glaciation (e.g. Forbes &
1 Present address: Department of Geosciences, Penn­
Taylor, 1987 ) . There, gravel can be derived from
sylvania State University, U niversity Park, Pennsylvania tills, outwash deposits, or other coarse-grained
168 02, USA. sub- to proglacial deposits. Thirdly, progradational

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0

75
76 B.S. Hart and A . G. Plint

gravelly coastlines may develop where the shoreline of the data presented was drawn from studies of
is adjacent to uplands yielding abundant coarse­ conglomerates in the Upper Cretaceous Cardium
grained sediment. Most modern examples of such Formation of western Canada (Piint & Hart, 1988;
coastlines are associated with fan-deltas. In fan­ Hart & Plint, 1989, 1 99 1 ; Hart, 1990, 199 1 ) . In the
delta settings, gravelly shorefaces develop in a 'tran­ Cardium Formation, sedimentological studies are
sition zone' between alluvial plain and offshore complemented by extensive stratigraphical studies
settings (e .g. Wescott & Ethridge, 1980) . Holocene (e.g. Plint et a!., 1986; Bergman & Walker, 1987;
sea-level rise has inhibited widespread develop­ Hart, 1990; Hart & Plint , 1993a ) . This work has
ment of progradational systems, and therefore most demonstrated that the chert-pebble conglomerates
studies of modern gravelly coasts have involved in the Cardium were deposited many tens to hun­
transgressive gravelly shorelines along bedrock dreds of kilometres basinward of any potential
coasts and in glaciated areas. source -they are therefore not interpretable as
Given the relative scarcity of gravel dominated alluvial fan or fan-delta deposits.
coastal deposits, and the logistical difficulties
involved in their investigation, it is no surprise that
modern gravel-dominated shorefaces represent a SEDIMENT TRANSPORT PROCESSES
type of environment that has received very little
attention from sedimentologists. Modern gravel There is considerable disagreement between authors
beachfaces have been the object of considerable and between disciplines (e.g. geologists and coastal
study (Bluck , 1967; Carr, 1969; Dobkins & Fol k , geomorphologists and engineers) about the termin­
1970; Carter & Orford , 1984; Forbes & Taylor, ology used to describe coastal zonation. The shore­
1987 ) , but there have been very few studies of the face is defined here as the broadly concave-upward
movement of pebble-sized material on the shoreface zone where sediment is transported by wave pro­
(Neate, 1967; Mathews, 1980; Gillie, 1983) and the cesses on a regular basis (i.e. between low-tide level
nature of sedimentary structures present on modern and 'fair-weather wave base'; Fig. 1). The upper
gravelly shorefaces is poorly documented. Studies of portion of the shoreface is the region of breaking
gravel on modern inner shelves (e.g. Gillie, 1983; waves (surf ) or 'nearshore zone' of coastal geomor­
Forbes & Boyd, 1987) are helpful in this regard , and phologists (see Komar, 1976; Dubois, 1992 ) . The
gravels have also been reported from modern sand­ lower shoreface extends seaward from the breaker
dominated shorefaces (e . g . Hunter et a/., 1979; zone to the depth at which shoaling surface gravity
Shipp, 1984; Ross & Long, 1989 ) . waves no longer entrain sediment (sometimes des­
Ancient conglomeratic shoreface/inner shelf cribed as the 'transition zone' between the nearshore
deposits have been recognized in outcrop (e.g. and offshore; Komar, 1976; Dubois, 1992). In the
Clifton, 1973, 198 1 ; Leckie & Walker, 1982; Leithold stratigraphical record , the vertical transition from
& Bourgeois, 1984; Leckie, 1988; Massari & Parea, interbedded sandstones and mudstones to thick sand­
1988; Cheel & Leckie, 1992) but in most of these stones (and/or conglomerates) in ancient prograding
cases, pebble-size material forms a small pro­ shoreline successions is typically defined as the tran­
portion of a sand-dominated sedimentary deposit. sition from inner shelf to shoreface deposition (e.g.
Kleinspehn et a!. ( 1984), Massari & Parea ( 1988) Plint & Walker, 1987 ) . The width and maximum
and Massari et a!. ( 1 986) described progradational water depth of the shoreface and nearshore are a
shoreface and beachface gravels associated with fan­ complex function of sea-floor slope, wave climate
delta settings. and tides, all of which vary between sites and over
Conglomerates have been described using subsur­ the course of time at any given site (Wright & Short,
face data (Bergman & Walker, 1987; Arnott, 199 1 ) 1984) . The beachface comprises that region between
and interpreted a s shoreface deposits o n the basis of low-tide level and the landward limit of wave action
stratigraphical position and the geometry of the (Fig. 1 ) . There exists abundant literature concerning
conglomerate bodies as a whole. Ancient con­ shoreface and beachface processes (e.g. Zenkovitch,
glomeratic beachface deposits are clearly recogniz­ 1967; Komar, 1976; Niedoroda et a/., 1985; Carter,
able, both in outcrop and core (e.g. Nemec & Steel 1988) , and an extended discussion will not be
1984; Plint & Walker 1987; Postma & Nemec, 1990) . presented here.
This paper will review the sedimentology of gravel­ Numerical analysis of the relationship between
dominated shoreface and beachface deposits. M uch flow parameters and the movement of gravel in
 Gravelly shorefaces 77

SHOREFACE

Swash
�....:::--
-- ------------------- limit

����-----�--�- Mean
low tide
Depth at Fairweather
breaker zone wave base

Fig. l. Schematic diagram showing

coastal zone terminology used in
this paper.

shallow marine settings has not advanced to the motion under the trough, plus gravity) . For a given
point where predictive estimates of sediment trans­ grain size, slope and wave conditions, there will exist
port can be made confidently. This shortcoming can a point at which the upslope and downslope forces
be related both to the logistical problems of working are balanced, and forward motion will equal the
in the high-energy shoreface zone and the problems backward motion . Seaward of this point, sediment
inherent in understanding the movement of pebble­ will migrate offshore, whereas further landward,
size materials in general (see below). Bulk textural sediment will migrate onshore. This effect is grain­
properties of a deposit, clast form and fabric all size dependent: for the same wave conditions and
appear to play important roles, which are not (yet) bottom slope, fine-grained sediments may move off­
fully quantifiable. Williams et al. ( 1989 ) , for exam­ shore whereas coarse-grained sediments move
ple, used sophisticated instrument packages to study onshore (Zenkovitch, 1967) . When a mix of grain
·
the movement of marine gravels, yet felt compelled sizes is available on the sea-floor, a graded profile can
to state that 'caution is required if attempting to be developed with large particles moving onshore
deduce bedload transport rates for marine gravels and smaller particles offshore (Zenkovitch, 1967 ,
with significantly different grain size distribution or p. 1 15 ) .
in flow regimes radically different from the present' . Shoreface profiles are a function o f several factors,
Given these uncertainties, only a qualitative over­ including wave height, period and sediment grain
view of the subject will be presented in this section. size. Although gravelly systems appear to respond
less quickly to changes in wave regime than their
sandy counterparts (Carter, 1988), Hart & Plint
Shoreface processes
(1989) concluded that gravel on the shoreface at
In a general way, there are three main forces respon­ Chesil Beach moved offshore during storms then
sible for sediment transport on the shoreface. onshore during fair-weather conditions. Gillie ( 1983)
These are: (i) wave motion; (ii) currents induced by described both onshore and offshore movement of
shoaling waves; and (iii) other currents. gravel on the inner shelf in response to differing wave
During wave shoaling, progressive distortion of conditions. Kidson & Carr ( 1959) used radioactively
sinusoidal deep-water waves results in the forward 'tagged' pebbles and found that at some critical (but
motion under the wave crest becoming shorter but undefined) distance offshore, pebble movement was
faster, and the backward motion under the trough very limited (even during storms) and exchange of
becoming longer and slower. This time-velocity gravel between the shoreface and shelf did not occur.
asymmetry has important consequences for sedi­ I ntuitively, this 'critical' distance must be related to
ment. As described by Zenkovitch ( 1967) and Bruun the depth at which oscillatory wave motion can
( 1988) , the shoreface profile is a function of sediment initiate movement, and is therefore a function of
grain size and wave type (height, wavelength, period, sediment size, submarine slope and wave climate.
form ) . For any particle resting on the shoreface Longshore currents (sensu stricto) are generated
there are forces acting to move the particle upslope over the upper shoreface by the dissipation of the
(stronger but shorter landward motion under the longshore component of energy in breaking waves.
crest) and downslope (slower but lon2:er seaward The upper shoreface, owing to the high energy levels
78 B.S. Hart and A. G. Plint

and potential for unidirectional longshore currents, Finally, in some settings, seaweed can enhance
is considered to be a major sediment transport both landward and seaward transport (Woodborne
zone for coarse-grained sediments (e.g. Carter & et a!., 1989 , and references therein ) , although the
Orford, 199 1 ) . geological significance of this process is hard to
Rip currents are another type of current induced assess. Pebbles and shell material attached to hold­
by breaking waves. Channels can be excavated fasts are commonly observed along beaches adjacent
by these seaward flows (Davidson-Arnott & to seaweed beds (Fig. 2 ) . Seaweed communities are
Greenwood, 1976; Gruszcyriski et a!., 1993) and best developed along rocky or gravelly coasts in cool
pebbles and even boulders can be transported off­ waters of high latitudes (Carter, 1988). Once trans­
shore (Zenkovitch, 1967; Gruszczynski et a!., 1993 ) . ported out of its initial habitat, the plant dies and
In sandy settings, coarser material accumulates as decays, leaving no lasting trace of the holdfast on
lags in the base of the rip channels (Davidson-Arnott the clast. Woodborne et a!. ( 1 989) reported kelp
& Greenwood, 1976; Hunter et a!., 1979). Rip cur­ rafting of a 30-kg boulder, and kelp rafted pebbles in
rents are not typically well developed on steep, 350 m water, 120 km offshore, and also reported that
'reflective' shorelines such as those associated with about 15% of the kelp along the South African
most modern gravelly shorelines (Carter, 1988) . coast is torn from the bottom during storms each
However, it must be remembered that these are year. If only 1% of the algae carried clasts, pebble
generally transgressive systems and progradational transport by this process could be 'significant' over
shorelines may have different profiles. geological time (Woodborne et a!., 1989 ) .
Tidal, wind-induced and fluvial currents may also
transport sediment on the shoreface . In wave­
Beachface processes
dominated settings, perhaps the most important role
of tides is to change water depth and hence the Beachface processes along gravelly coasts are better
geographical limits of the surf and swash zones documented than those below low-tide level. The
(Ross & Long, 1 989) . Currents generated by roughness of gravel beachfaces can reduce the velo­
coastal set-up (with onshore winds) flow offshore city and distance travelled by the swash, relative to
along the sea-floor in the shoreface zone, and thus sandy systems (Carter, 1988). The high porosity and
enhance offshore sediment transport during storms permeability and low water table of gravelly beaches
(Niedoroda et at., 1985). Close to river mouths, also reduces the run-up distance of swash, and
interaction of fluvial and shoreface processes is to be reduces or eliminates backwash by allowing water to
expected. Gravity-driven mass transport processes freely percolate into _the sediment. When the water
may occur in 'mouth bar' settings (Kleinspehn et a!., table is high, for example during falling tides or
1984) as rapid deposition leads to oversteepening storms, surface clasts on the beachface are more
and resultant instability. susceptible to movement by wave swash and back-

Fig. 2. Clasts attached to seaweed

holdfasts on a transgressive
shoreline, Island View Beach,
British Columbia. Lens cap for
scale.
 Gravelly shorefaces 79

wash . Forbes & Taylor ( 1 987) reported that beach across sandy beaches (Clifton et al., 1973; Fig. 3A) .
morphology is sensitive to the amount of sand in the The permeability of gravel-dominated systems tends
system. Oblique swash motion on the beachface , to limit the development of such small channels,
generated by oblique wave approach , transports although they may nevertheless develop (Fig. 3 B ) .
clasts alongshore, with grains returning directly sea­ Bluck ( 1 967) suggested that particles can b e trans­
ward during the backwash. This alongshore 'zig-zig' ported through the sediment body in certain portions
motion should not be confused with longshore cur­ of coarse-grained beachfaces. Bluck suggested
rents in the surf zone, even though both processes that the backwash percolating through the gravel
transport clasts along the shoreline. Overwash may appeared to 'flush' finer particles seaward. How­
be important during storms on some gravel barrier ever, unless the beach consisted of an open, clast­
systems (Carter & Orford , 198 1 , 1984) . supported framework of pebbles and cobbles, the
Pebble-filled streams and drainage rills (developed sieving process is unlikely to operate.
by groundwater flow during falling tide) may cut

Fig. 3. Streams cutting across

beachface. (A) Pebble-lined stream
cutting across sandy beachface,
China Beach, British Columbia.
Note steep, locally overhanging
channels walls. View toward back­
beach area. Note 'pulse' of gravel
in background being transported
seaward through the channel.
(B ) Large drainage rills incised
into Chesil Beach, England.
80 B.S. Hart and A . G. Plint

GRAIN SIZE AND SHAPE axis length) and D10 from two measured sections at
the Bay Tree locality are shown in Fig. 5. Regression
As with sandy deposits, the textural characteristics lines from D10 in section BT2, and L and I from
of gravels reflect the processes responsible for their both sections have positive slopes, indicative of an
deposition. Unfortunately, gravels do not readily up-section increase in grain size, and high correlation
lend themselves to the same types of routine size coefficients. These results suggest that upward­
analyses as sands, and therefore much less is known coarsening trends (not necessarily 'recognized' by
about their grain-size distribution. On the other D10) may develop on some gravelly shorefaces,
hand , the size of gravels makes them much more although the lack of a distinguishable trend for the S
amenable to analysis of clast shape, usually expressed axis suggests that this dimension may not have been
as form indices derived from measurement of long hydrodynamically significant at Bay Tree.
( L), intermediate (!) and short ( S) axes. As noted Dupre et al. ( 1980) , in a study of a Pleistocene,
in the preceding section, accurate predictions of mixed sandy and gravelly shoreface system, reported
entrainment velocity are very difficult to quantify as that the coarsest clasts were to be found within
a result of sorting and clast shape effects. deposits of the upper surf zone. A comparable up­
section increase in the thickness, abundance and
lateral continuity of conglomerate beds in sandstone­
Textural trends
dominated shoreface deposits can be seen in some
The threshold conditions for movement of a given sections of the Cardium Formation (Fig. 6).
clast depend not only on flow parameters but also on Some studies of beachface deposits indicate that
clast shape, fabric of the deposit and the size of the the coarsest sediments are found on the uppermost
grain relative to that of the host deposit (Komar & (landward) part of the beachface (Biuck, 1967; Kirk,
Li , 1986) . Although the largest clasts are sometimes 1980; Massari & Parea, 1988) , suggesting that
used to evaluate flow competence , such grains on upward-coarsening trends should develop in pro­
the surface of a deposit project higher into the flow gradational systems. Maejima ( 1982 ) , however,
and are thus subjected to higher shear stress than reported that the coarsest sediment was present in a
the bulk of the sediment. Komar ( 1 987) suggested zone of berm accretion on the beachface. In mixed
that flow competence is therefore best evaluated by sand-gravel systems, the plunge point (at the toe of
relating the bed shear stress and the ratio of the the beach during fair-weather conditions) may also
diameter of the largest clast to the median diameter be a position where coarse sediment accumulates
of the deposit as a whole . Maejima ( 1 982) found, (Dupre et al., 1980; Kirk, 1980; Short, 1984). Well­
however, that mean grain size and D10 show very sorted and stratified very coarse sandstones and
similar trends across gravelly beaches, lending some granule conglomerates are present as beachface
justification to the use of D10 (D10 =intermediate deposits in the Cardium Formation (Piint & Walker,
diameter of the 10 largest clasts ) . Williams & 1987; Arnott, 199 1 ) .
Caldwell ( 1 988) suggested that the short axis of Gravelly shorefaces and beaches may reveal along­
clasts was the most 'hydraulically sensitive' length shore trends in grain size generally taken to represent
on gravel beaches. Further investigation appears downdrift fining. Two examples from outcrops of
warranted. the Cardium Formation show that such trends may
In the previous section it was noted that graded not always be clear (Fig. 7 ) . Pattison & Walker
shoreface profiles are expected to develop under ( 1992) used along-strike grain-size variations to infer
equilibrium conditions, with larger clasts moving sediment dispersal direction in pebbly sandstone
onshore. A vertical section through a progradational shoreface deposits of the Cardium Formation in
shoreface deposit might therefore display an upward­ the subsurface. This type of trend was inferred to
coarsening trend . To date, the only attempt to quan­ develop in response to preferential longshore trans­
tify such a trend appears to be the work presented port of finer grades, with smaller clasts 'out running'
by Hart & Plint ( 199 1 ) on the Bay Tree outcrop of coarser sizes. Although such processes are docu­
the Cardium Formation. Representative measured mented from modern settings, Carter ( 1988) has
sections through two, well-exposed examples of grav­ emphasized that downdrift coarsening may also
elly shoreface deposits in the Cardium Formation develop under some circumstances.
are presented in Fig. 4. Vertical trends in L (long Gravels and conglomerates typically consist of at
axis length) , I (intermediate axis length) , S (short least two grain-size classes of material, gravel clasts
 Gravelly shorefaces 81

BT7

BT6

BT2

Ravinement Surface ---r-t:l;;;;::--..:.o

(Hart & Plint 1993a)
A

MNA·IO MNA·9
, __

0
.-�Channelized base
/
B

Fig. 4. Measured sections of Cardium Formation. (A) Shoreface conglomerates at Bay Tree (Alberta), represented by
seven measured sections along cliff face oriented nearly parallel to depositional strike. Total distance between BT7 and
BT4 about 300m. Nearly constant thickness along strike suggests tabular geometry. Beach face deposits are preserved at
top of BT7. (B) Shoreface/mouth-bar conglomerates at Elephant Ridge (British Columbia) . Sections measured along diff
face at slight angle to depositional strike. Total distance along strike about 200m. Note suggestion of channel form, with
conglomerates scouring into u nderlying swaley cross-stratified shoreface sandstones. Sections hung on coarse-grained wave
rippled surface inferred to be a nearly planar marine flooding surface (Hart, 1990) . Facies codes: A, clast-supported pebble
conglomerate; B, imbricated clast-supported conglomerate; C, graded conglomerates; D, decimetre-thick cross-bedded
conglomerates; E, matrix-supported conglomerate; F, interstratified conglomerate and sandstone (beachface deposits);
G, amalgamated conglomerates; H, matrix-supported pebbly granule conglomerate; I, poorly sorted pebbly sandstones
with localized shallow scour-fill conglomerates; 16, swaley cross-stratified sandstones; 17, cross-bedded sandstones, locally
pebbly.
82 B.S. Hart and A. G. Plint

�• • • •
100
(A) BT2

•• • ••
(B) BT7

I
0� 80 -----
60 � • •

•
0 40
�
�
o 2
I I I

.g � 100

iJ !IL�I :
"'
c
0
·o; � 100
I . ..

jJ�� • r: .. �
. _,
� 80

�•••100
• . .

I = r:
. .
X 80 Fig. 5. Trends in non-dimensional
E 60 - ..
L, S, I, and 010 versus non­
� 40
dimensional height above base from
20
I I ----=L " I __JI two sections at Bay Tree , Alberta.
0 0 L_ ..1. _..1.
o 25 50 75 100 O _25- 50_ 7 5 _ 100 See text for explanation. Relative
Non-dimensional height above base section locations shown in Fig. 4A.

and sand and/or mud matrix. The proportion of of the body. The matrix-supported conglomerates
each defines whether the deposit is open framework were interpreted to have been deposited in the
(lacking matrix completely) , clast supported (clasts vicinity of fluvial distributary mouths.
in contact with spaces infilled by matrix) or matrix Although these results and others suggest that
supported (clasts 'floating' in matrix). These differ­ texture (in particular sorting) is a valuable tool for
ences can have significant economic importance, as discriminating between fluvial and littoral conglom­
they directly affect reservoir properties such as erates, texture alone cannot be used to distinguish
porosity and permeability (e.g. Clarke, 1979; Cant between fluvial and shoreface conglomerates
& Ethier, 1984; Arnott, 1991). (Nemec & Steel, 1984). Matrix- or clast-support of
There is a consensus that shoreface and beachface conglomerates should not be considered diagnostic
conglomerates tend to be better sorted than fluvial of, respectively, fluvial and littoral deposits. For
conglomerates, and that sand and gravel tend to example, Howell & Link (1979) found that, in the
be better segregated in deposits of shallow marine Eocene of California, non-marine conglomerates
environments (e.g. Emery, 1955; Clifton, 1973). It tended to be clast-supported (in places open frame-·
has been suggested that rapid deposition of sand and work), whereas wave-influenced deposits include
gravel in fluvial deposits may produce pebbly sands, both clast- and matrix-supported conglomerates.
whereas prolonged reworking of a mixture of sand Leithold & Bourgeois (1984), N ielsen et al. (1988)
and gravel by waves will tend to segregate the two and Massari & Parea (1988) have also reported
constituents (Clifton, 1973). Well-sorted, clast­ matrix-supported conglomerates interpreted to have
supported conglomerates are common in shoreface been deposited in a shoreface setting. Studies of
conglomerates at the Bay Tree locality and elsewhere fluvial conglomerates suggest that they too are typi··
in the Cardium Formation (Figs 4 & SA & B ) . cally clast supported (Harms et al., 1982). Low··
Arnott (1991) examined a linear conglomerate wave-energy shorelines may lack sufficient energy to
body of the Cardium in the subsurface. Although effect wave segregation of sand- and pebble-size
dominantly clast-supported, matrix-supported con­ material. Finally, the nature of the sediment supplied
glomerates were also present but tended to be con­ to the shoreline (ratio of sand to gravel, fluctuations
centrated preferentially along the landward margin of this ratio through time) and sediment transport
 CJ
;:;
'<::
"'
�
"'
;::..
0

�
.,

Fig. 6. Up-section increase in thickness of shallow, pebbly scour fills in swaley cross-stratified shoreface sandstones from a single outcrop of Cardium Formation at
Cutbank Lake, Alberta. Photographs (A)-(D) represent a vertical interval of about 6 m . Conglomerates are interpreted as rip-current and longshore trough
deposits. Pocket knife (arrowed) for scale. 00
w
84 B.S. Hart and A. G. Plint

infiltrate? Comparison of the size of clasts inferred

to have been rolled on the bed with the size of the
N
sand in the matrix is sometimes used as a starting
point (Harms et al., 1982) . The application of this
approach to shoreface conglomerates is difficult
Vl
c
0
because: (i) shallow marine settings are typified by
·.::;
"' combined flows (especially during significant sedi­
>
Q; ment transport events such as storms) and criteria
Vl
.c for the threshold of movement under such conditions
0
.....
0 are not easily established (especially for gravels),
Q; and (ii) this approach is probably not valid for high
.c
E sediment concentrations or transport rates such as
::J
z
might typify the surf zone.
Textural evidence can sometimes provide clues as
BT4
to the origin of the matrix. For example, in the
9 I q C? I 9 s Cardium Formation at Bay Tree, a sandy matrix is
A 20 30 40 50 60 70 common in clast-supported shoreface conglomerates
010 (mm) but is typically absent from adjacent finer (granule
or fine pebble) conglomerates (Fig. SA) . This may
MNA-10 N suggest that the sand was able to infiltrate the larger
MNA-9
pore throats of the coarser deposits but not through
the more restricted openings between smaller clasts.
· MNA-8

MNA-6 Clast shape

MNA-4 Many studies have suggested that form indices can

MNA-3 be useful for distinguishing between 'beach' and
'fluvial' pebbles (Cailleux, 1945; Dobkins &
MNA-2 s
Folk , 1970). The study by Dobkins & Folk ( 1970)
suggested that the 'maximum projection sphericity'
10 20 30 40 50 and 'oblate- prolate index' were useful discrimi-­
B 010 (mm)
nators for distinguishing between beach and river
clasts and several studies have since investigated the
Fig. 7. Longshore trends in D10 in shoreface
conglomerates of Cardium Formation at: (A) Bay Tree usefulness of these criteria (partial summary in Hart
and (B) Elephant Ridge (Mount Niles), showing a weak, 199 1 ) . Hart (199 1 ) , using pebble shape data from
southeasterly fining trend, which, based on regional Bay Tree and Chesil Beach , concluded that no form
palaeocurrent data (Hart, 1990; Hart et a/., 1990) is the parameter derived to date can be used to unambigu­
dominant longshore transport direction. Each
'observation' at Bay Tree corresponds to a D10 value ously distinguish shoreface or even beachface gravels
calculated from individual beds in each section. The D10 and conglomerates from fluvial deposits.
values at Elephant Ridge represent averaged values for Thus, although pebble-shape segregation does
each measured section. The location of sampling points is develop in fluvial and beachface environments, and
shown in Fig. 4A & B .
pebble zonation does develop on individual beaches
(Orford, 1975), shape criteria alone cannot be con­
sidered diagnostic of depositional environment .
Although Howard ( 1 992) was able to show statistical
mode (see next section) must also be considered. differences between shape indices of ancient fluvial
Carter & Orford (1991) emphasized that episodic and beach gravels, considerable overlap was present
supply can induce textural changes that are inde­ between fields and none of the 'cri tical values' derived
pendent of process regime. from studies of modern gravels provided a successful
The origin of a sandy matrix in a conglomerate means of differentiating beachface conglomerates.
can be problematic. Was the sand transported and
deposited with the gravel , or is the sand a later
 Gravelly shorefaces 85

Fig. 8. Clast-supported pebbly

shoreface conglomerate. (A) Bay
Tree outcrop; note open framework
portions. (B ) Core ( 10-18-63-5W6)
note crude stratification and
possible imbrication to right.

B EDFORMS AND be nearly shore-parallel due to wave refraction (Fig.

SEDIMENTARY STRUCTURES lOA}, and both onshore and offshore migration
(suggested by asymmetric profiles) have been
In most ancient deposits, sedimentary structures reported. Gravel wave ripple form sets may be
provide the key information for palaeoenvironmen­ preserved , especially when the bedform is draped
tal reconstruction. Knowledge of the types of bed­ by finer-grained sediment. Cross-bedding may be
forms and morphological elements in analogous produced by migration of the bedform under the
modern environments is therefore critical . Unfortu­ influence of asymmetric wave motion or wave motion
nately little is known about the variety of bedforms with weak superimposed unidirectional currents.
that may form on gravelly shorefaces. Bi- and polymodal fabrics with modes oriented in
onshore and offshore directions may indicate the
presence of shore-parallel symmetrical gravel wave
Shoreface
ripples (ct. Leckie & Walker, 1982) even if form sets
Gravel wave ripples (Fig. 9A & B ) , formed by are not observed (Hart & Plint, 1989 ) . Alternatively,
oscillatory water motion are excellent indicators of migrating gravel wave ripples may produce fabrics
subaqueous deposition at depths above storm wave with dominant dips in the direction of ripple
base but can be formed at considerable depths, well migration (Hart & Plint, 1989; Fig. lOB) . In the
beyond the limit of the shoreface. Leckie (1988) absence of preserved form sets, it may be difficult
reviewed existing literature on the generation and to distinguish landward dips generated by onshore
environmental significance of 'coarse-grained' rip­ migration of gravel wave ripples from those produced
ples. Although most examples appear to be from in gravels by strong unidirectional offshore flows (cf.
shelf settings, gravel wave ripples have been reported Cheel & Leckie, 1992) .
from modern (Shipp , 1984; Hart & Plint, 1 989} and Nearshore bars of various types are also formed
ancient (Leithold & Bourgeois, 1984; Massari & by wave processes. These bars develop on gently
Parea, 1988} gravelly shorefaces. Crestlines tend to dipping shorelines which tend to be sand-dominated
86 B.S. Hart and A. G. Plint

Fig. 9. Coarse-grained wave ripples

developed in poorly sorted coarse
sandy conglomerate; Cardium
Formation on Elephant Ridge. (A)
Symmetrical crestlines and (B)
interference ripples. Scale bar in
centimetres.

(Carter, 1988). Large, broadly convex upward capped by planar stratified sandstones or pebbly
'macroforms' apparently representing nearshore sandstones (Fig. llA & B ) .
bars have been reported from shoreface successions Troughs associated with longshore bars have been
comprising sandstones, pebbly sandstones and con­ recognized (Leithold & Bourgeois, 1984; Massari &
glomerates by Leithold & Bourgeois (1984), Massari Parea, 1988) as broadly lenticular conglomerates
& Parea (1988) and Massari eta!. (1986). Such forms with concave upward bases cut in sandstones. Such
are found in the upper shoreface (Davidson-Arnott troughs should be elongate and nearly shore-parallel.
& Greenwood , 1976) and may or may not migrate. Ross & Long (1989) cored a modern nearshore bar
Migratory bars centimetres to decimetres in height system in a mixed gravel and sand setting and found
may move onshore and (particularly in tidal settings) that gravel was not restricted to the troughs, but
weld themselves to the beachface, typically produc­ could also be found on the bar crest. The presence
ing onshore oriented tabular to sigmoidal cross-beds of longshore bars and troughs from ancient gravelly
 Gravelly shorefaces 87

A 8

Mean x-bed
orientation

Mean crestline
Crestline orientation
orientation

Fig. 10. Summary of palaeocurrent information from Elephant Ridge section of Cardium Formation (Fig. 4B) .
(A) Coarse-grained wave ripple crestlines (a) and decimetre-scale cross-bedded conglomerates (b) . Wave ripples
indicate NW-SE trending shoreline, in agreement with other data from the Cardium Formation in ths area (Hart,
1990) . (B) Pebble fabric from a near-symmetrical coarse-grained wave ripple at Elephant Ridge suggesting onshore
migration of the wave ripple form.

shoreface deposits suggests that steep, 'reflective' wavelengths typically about 5 m and crests oriented
profiles (typical of modern transgressive gravelly at high angles to the shoreline from the lower beach­
shorefaces) were not developed at the time of depo­ face (intertidal) . These bedforms appeared to have
sition (Wright & Short, 1984). been produced originally by longshore currents at
Broad, pebble-lined scours that are oriented per­ high tide.
pendicular, or nearly so, to the palaeoshoreline At Bay Tree (Fig. 4A) over 50% of the conglomer­
in shoreface sandstones may be interpreted as ate is organized into crudely bedded, amalgamated
rip-current channels (e.g. Fig. 6C; Leithold units, 1-3 m thick. Centimetre to decimetre thick,
& Bourgeois, 1984; Massari & Parea, 1988; massive clast-supported pebble conglomerate beds
Gruszcyzynski et a/., 1993 ) . Shell debris may form a are the primary constituent of the amalgamated
significant component of such lags, facilitating the units, but open framework , imbricated, matrix­
recognition of depositional environment. supported conglomerates and thin sandstone beds
Cross-bedded sandstones and conglomerates are (generally < lO cm thick) are also present. Shallow
common in shoreface deposits, and may be oriented scours, depositional thinning and lateral facies
onshore, offshore or alongshore (Dupre et al., 1980; changes limit the lateral extent of individual beds to
Leithold & Bourgeois, 1984; Massari & Parea, 1988). a few metres. Despite the internal scouring, the
These structures may be the product of asymmetrical overall appearance of the conglomerate, revealed
wave motion, and rip, longshore, tidal and (close to especially by the sandstone beds, is of a pronounced
river mouths) fluvial currents . At the Bay Tree horizontal stratification in the strike sections at Bay
section of the Cardium Formation, cross-bedded Tree (Fig. 14) . Presumably, 'clinostratification', such
conglomerates several decimeters thick (with cross­ as that illustrated by Massari & Parea ( 1 988), would
bedding defined by size or textural variations) are be observable in shore-normal sections.
present. This large-scale cross-bedding (Fig. 12A) Most of these shoreface conglomerates can be
was apparently produced by the migration of dune­ interpreted as the products of gravel bedload sheets
like forms at least 60 em high. Palaeocurrent evi­ such as described from gravelly fluvial systems (Hein
dence indicates that these bedforms and similar & Walker, 1977; Whiting et a/., 1988). The lateral
cross-sets in other outcrops of the Cardium migrated dimensions of these 'bedforms' were probably of
in a shore-parallel direction (Figs lOA, 12B & 13A ) . the order of several metres, explaining the lateral
Such bedforms have not been documented from impersistence of individual conglomerate beds in the
modern gravelly shorefaces, although B ujalesky & amalgamated deposits. The possible existence of
Gonzalez-Bonorino (199 1 ) reported asymmetric this type of sediment transport has been suggested
'washed-out (gravel) dunes' up to 10 em high with by Mathews ( 1980), who found that tracers on a
88 B.S. Hart and A. G. ?lint

Fig.1 1. Migratory bars of upper

shoreface and beachface. (A)
Poorly sorted pebbly sand , Island
View Beach, British Columbia ( B )
Sigmoidal, landward-directed
sandstone cross-set in upper portion
of Cardium Formation shoreface
conglomerates at Bay Tree,
Alberta.
 Gravelly shorefaces 89

Fig. 12. Decimetre-scale cross­

bedding in Cardium Formation
shoreface conglomerates. (A) Bay
Tree (scale= 15 em) and (B)
Elephant Ridge (pocket knife for
scale) . In both cases, migration
direction was alongshore, to the
southeast.

modern gravelly foreshore moved alongshore on the of quantitative techniques such as plots of maximum
lower beachface (i.e. in the surf zone at high tide) as clast size versus bed thickness or Markov chain
discrete 'slugs' of sediment representing volumes of analyses become impractical.
1-10m\ rather than as a uniform gravel sheet. Pebble fabrics from the shoreface portion of the
These 'slugs' moved 10- SOm alongshore during conglomerate at Bay Tree are typically polymodal
periods of high wave energy, but were reworked (Fig. 13B ) . In most cases, the principal mode is
during fair-weather conditions. One may speculate towards the south or southeast (alongshore) . Other
that, during strong longshore flows, the bedload modes are observed at about 90° and 180° to the
sheets may have grown in height and developed principal mode in nearly all cases. All of these
slipfaces, such as has been observed in modern samples represent crudely stratified to cross-bedded
gravelly streams (Whiting et at., 1988) . Because of clast-supported pebble conglomerates generally
the amalgamated (and hence indistinguishable) con­ from the middle to lower portions of the section.
tacts between individual beds ('flow units') , the use The dominant dip direction is approximately that of
 (A�
90 B.S. Hart and A. G. Plint

(B) (C) between bedload sheets and dunes. Massari & Parea

�
(1988) also reported the presence of polymodal
fabrics from shoreface conglomerates but did not

�
elaborate on the potential significance of their obser­
vations. They also indicated that rose diagrams of
clast long axes from upper shoreface deposits could
be isotropic, polymodal, bimodal or unimodal.
Clifton (1973) suggested that pebble segregation
(E) (F)
and the lenticularity of conglomerates, although not

� �
necessarily diagnostic, were useful criteria for
distinguishing between wave-worked and fluvial
gravels. Wave-worked gravels tend to be more 'lat­
erally regular' than lenticular, owing to the uniform­
ity with which waves act over large areas. He also
suggested that winnowing of a deposit by waves
(G) (Hl (I)

� �
could segregate sand- and gravel-sized material . Sub­

~
sequent studies (Leithold & Bourgeois 1984; Massari
& Parea 1988; Hart & Plint 1989) have tended to
confirm Clifton's textural and stratigraphical obser­
vations. It should be reiterated, however, that rip­
Fig. 13. Palaeocurrent data from Bay Tree section. (A)
current deposits and longshore trough conglomerates
Cross-bedded sandstones and conglomerates. (B- I )
Pebble fabrics from conglomerates. Pebble fabrics are for may both have lenticular geometries, providing that
sites labelled in Fig. 4A: (B) BTI-11 (beachface they were buried below the influence of wave rework­
conglomerates); (C) BTI- 15 (plunge-step ing. Lenticular, channelized conglomerates may also
conglomerates); (D) BT4-2 (?wave rippled be present in river mouth settings (Kleinspehn et al.,
conglomerate) ; (E) BTI-5; (F) BT3-7; (G) BT3- 1 2; (H)
BT5-9; ( I ) BT4- 13. See text for further description. 1984; Fig. 4B).
Planar cross-bedded conglomerates several deci­
metres thick at the top of some sections at Bay Tree
(Fig. 15) probably record the progradation of a
the dominant cross-bed orientation in the sand­ gravel plunge step at the toe of the beach (low­
stones. This suggests a downcurrent dip of clasts tide level) -a feature common on coarse-grained
lying on the lee side of gravel bedforms transitional beaches (Forbes & Taylor, 1987; Massari & Parea,

Fig. 14. Pronou nced horizontal

stratification in shoreface
conglomerates at Bay Tree.
Stratification emphasized by
sandstone interbeds.
 Gravelly shorefaces 91

Fig. 15. Planar tabular

conglomerate cross-set at Bay Tree,
i nferred to have been produced by
the offshore progradation of a
plunge step at the base of the
beachface. From top of section BT2
( see Fig. 4A) .

1988; Hart & Plint, 1989) . Somewhat larger steps

are reported from some gravelly shorefaces (Kirk,
1980; Carter, 1988; Massari & Parea, 1988; Postma
& Nemec, 1990) . A pebble fabric from the plunge­
step deposits at Bay Tree shows offshore dipping
pebbles (Fig. 13B) . B ased on (unpublished) obser­
vations at Chesil Beach, progradation of the plunge
step seems most likely to occur either following high
tide or after a storm surge when the water table in
the beach is still high. At such times, backwash
dislodges pebbles from the beachface and transports
them seaward. These clasts then avalanche over the
plunge step causing it to prograde.
Body and trace fossils provide the clearest
evidence of marine deposition (Clifton, 1973;
Kleinspehn et a/., 1984; Leithold & Bourgeois,
1984) . Unfortunately, neither are common in high­
energy, conglomerate-dominated shoreface deposits
(Leithold & Bourgeois, 1984; Massari & Parea,
1988; Hart & Plint, 1989 ) . Trace fossils are rare in
conglomeratic shoreface deposits of the Cardium
Formation, although traces are present locally in
thin interbedded sandstones (Fig. 16) . Although
gravels are generally inhospitable to infaunal macro­
benthos, we have observed (using SCUBA) crab
burrows in gravels on the Chesil Beach shoreface.
However, the preservation potential of these bur­
rows seems very low owing to the lack of a contrast­ Fig. 16. Burrowing ( Diplocraterion) in sandstone within
ing sediment fill. In general , the abundance of trace shoreface conglomerates, Cardium Formation (core from
fossils may be related directly to the percentage of 16-19-5 1 - 1 0W6, scale bar= 5 em).
92 B.S. Hart and A. G. Plint

sand, partly because this forms a more hospitable considered diagnostic of gravelly beaches (Fig. 17 A
substrate, and partly because sand is better able to & B ) . In examples from the Cardium Formation,
preserve traces of burrowing. individual gravel ·laminae/beds can be normally or
inversely graded.
Sand tends to be well-segregated from gravel, at
Beachface
least in high-energy systems , and commonly forms
The structure of beachface gravels and conglomer­ 'sand runs' in the intertidal zone (Fig. 18A) . Beach­
ates is well established (Maejima, 1982; Nemec & face dips are generally inversely related to the per­
Steel, 1984; Forbes & Taylor, 1987; Massari & Parea, centage of sand in the system and directly related to
1988; Hart & Plint, 1 99 1 ) . Well-stratified, laterally grain size. Shape sorting and seaward-dipping pebble
continuous gravels and interbedded sands that dip imbrication (Fig. 18B) may be well developed.
gently (typically S0-20°) in a seaward direction are Landward-dipping stratification and imbrication may

Fig. 17. Beachface deposits.

(A) Gently dipping interstratified
beachface conglomerates and
sandstones, Bay Tree. Note low­
angle truncation surface. (B) Gently
dipping stratification in beachface
conglomerates, Cardium
Formation, Cutpick Hill, Alberta.
 Gravelly shorefaces 93

Fig.18. Modern beachface gravels

at Chesil Beach, England.
(A) Good segregation of sand- and
pebble-sized constituents on
beachface. (B) Offshore (to right)
imbrication of beachface gravels.
Shape sorting is not well developed
at Chesil Beach.

be present in cases, and are apparently the result of outcrops of the Cardium Formation (Fig. 19). The
berm accretion (Maejima, 1982; Forbes & Taylor, scours may have steep walls (in places stepped or
1987 ) . Maejima ( 1 982) noted that broad , concave­ undercut) that scour into hummocky/swaley cross­
upward scour surfaces may represent the effects of stratified and planar laminated sandstone. Multiple
beach cusp formation. Massari & Parea ( 1988) found phases of fill can be inferred for some units, and
that beachface conglomerates were internally trunc­ sandstone 'intraclasts' , woody debris and shell frag­
ated in places by low-angle erosion surfaces inter­ ments are observed in places.
preted to represent storm-wave planation of the Leithold & Bourgeois ( 1 984) suggested very rapid
beachface (Fig. 17A). cut-and-fill (nearly instantaneous) at shallow depths
Enigmatic 'high-angle scours' were described just seaward of the breaker zone. An alternative
by Leithold & Bourgeois ( 1 984) and Bourgeois & interpretation is that the scours are the product of
Leithold ( 1 984) from gravelly sandstone 'shoreface' small streams incised into the beachface. Intuitively,
deposits; similar structures have been observed in it seems unlikely that high-angle scours can be
94 B.S. Hart and A. G. Plint

Fig. 19 Conglomerate filling �teep­

sided scour cut into planar­
laminated, fine-grained sandstone
?beachface deposits of Cardium
Formation along Murray River,
British Columbia. The scour may
have been cut subaerially by a small
stream crossing the beach, or may
be the product of subtidal scour by
rip-currents (Chiocci & Clifton,
1991) . See text for discussion.

maintained in non-cohesive sandy substrates in a and beachfaces can be very sensitive to textural
subaqueous setting, especially in the presence of factors such as the amount of sand in the· system.
intense wave activity. However, Gruszcyzynski The effects can be non-linear: process affects mor­
(pers. comm . , 1993) reported seeing (using SCUBA) phology which in turn affects process. In this respect ,
vertical-sided, rip-current channels cut into sandy it may be overly simplistic to attribute facies vari­
sediment, and speculated that bacterial or algal ations at any given site to short-term changes in
binding may have rendered the sediment cohesive. energy level (e.g. storm -non-storm) , sea-level , etc.
Alternatively, high-angle walls (locally overhanging) Modern gravelly beachfaces show much intersite
can be cut in loose sand, and maintained in subaerial variation and , in places, much temporal variation at
settings by capillary forces (Fig. 3A) . Cohesion may any given site. Facies models based on 'snapshots' of
be promoted by salt cementation (in a spray zone any particular beachface or shoreface are therefore
j ust above the limit of the swash) and by organic clearly of limited applicability. It is equally clear that
films. The example shown in Fig. 19 is from planar sedimentary facies and textural or pebble shape
stratified sandstone, interpreted as beachface , which criteria derived from the beachface must not be
directly overlies swaley cross-stratified (shoreface) applied when attempting to identify shoreface
sandstone, the stratigraphical position of which, and conglomerates.
relation to relative sea-level changes, is well con­ Several previous authors have presented sum­
strained (Hart & Plint, 1993a) . maries of criteria for distinguishing between fluvial
Facies descriptions of wash over deposits have been and shoreface/beachface gravels (Clifton, 1973;
provided by Carter & Orford ( 198 1 ) and Forbes Leckie & Walker, 1982; Ethridge & Wescott, 1984;
& Taylor (1987) . Massari & Parea ( 1988) have Nemec & Steel, 1984; Reddering & Illenberger,
described washover deposits from Messinian and 1988). Based on these studies, and our own studies
Pleistocene gravelly. shorelines. of shoreface and beachface gravels and conglomer­
ates, an attempt is made in Table 1 to outline the
expected character of preservable gravelly shoreface
FACIES SUCCESSIONS and beachface deposits as a function of several vari­
ables. The factors that appear to be of greatest
Stochastic variations in storminess, sediment supply, significance are: (i) change in relative sea-level
relative sea-level and other factors probably account (direction and magnitude) , (ii) textural character-­
for much more of the variation in sedimentary struc­ istics of the sediment, including proportion of sandl
tures and sediment textures than is commonly in the system , (iii) proximity to sediment source
appreciated . The morphology of gravelly shorefaces (river mouth) ; and (iv) orientation of the shoreline
 Gravelly shorefaces 95

Table 1. Summary of expected preservable characteristics of gravelly shoreface and beachface deposits

Geometry and
stratigraphical
position Processes Facies

Transgressive Thin lags continuous with Barrier rollover Thin lags

flooding surfaces Rafting Wave ripples and storm
Topographically 'trapped' beds
shoreface deposits

Regressive
River mouth Lobate (plan) Rivers/waves/tides Lenticular beds
Lenticular (cross-section) Slope failures Mixed palaeocurrents
Correlate with other Mix of fluvial and marine
progradational deposits facies
Mixed sand and gravel Tabular (cross-section) Waves/tides - shoreface Longshore bars/troughs
?Shore-parallel Swash/backwash - beachface Rip-current scours
elongation Palaeocurrents reflect
Correlate with other processes
progradational deposits Segregation of sand and
gravel increases with
wave energy
Gravel - drift aligned Tabular (cross-section) Longshore currents dominant Weak shore-normal
Shore-parallel elongation zonation
Correlate with other Bedload sheets/dunes
progradational deposits Palaeocurrents shore
parallel
Gravel - swash aligned Tabular (cross-section) Onshore/offshore motions Pronounced shore­
?Beach ridges normal zonation
Shore-parallel elongation Wave ripples
Correlate with other Onshore/offshore
progradational deposits palaeocurrents

with respect to dominant wave-approach direction. both in outcrop and in core and the relative strati­
It is expected that in most instances transgressive graphical position can be used to help infer a shore­
gravelly shorelines will leave little permanent record face origin for the underlying deposits. Non-marine
of their existence. Thin pebbly lags representing facies may or may not be present above the beach
abandonment of isolated clasts during shoreface facies.
retreat, and possibly organic rafting seaward of the Roughly tabular geometries are expected for indi­
shoreface (see Leithold, 1989; Woodborne et a!., vidual progradational packages, except close to river
1989; Liu & Gastaldo, 1992), are the expected litho­ mouths where lenticular geometries may occur (Fig.
logical expression . Thicker successions might be 4B) . These river-mouth deposits will reflect the
preserved where the shoreface has 'backed up' interaction of fluvial, tidal and wave-induced cur­
against pre-existing relief ('bevels', cliffs, etc . ; see rents. Rapid deposition in the transition zone will
Bergman & Walker, 1987, 1988; Hart & Plint, 1993b) . generate poorly sorted deposits and , potentially,
Stratigraphically, these deposits will correlate with instability related gravity flows (Kleinspehn et a!.,
marine flooding surfaces (Plint et a!. , 1986) . 1984 ) . Channelized and burrowed deposits may be
Progradational packages are likely to represent found in close proximity.
regional shoreline regression, although they may also Away from river mouths, the influence of marine
develop in areas of high sediment supply during processes will become more apparent. In 'mixed'
relative sea level rise (e.g. modern fan-deltas) . settings with appreciable amounts of sand, segre­
Ideally the shoreface conglomerates will overlie shelf gation of sand and gravel constituents should be
deposits and , in turn, be overlain by beachface proportional to the wave energy level of the environ­
conglomerates. The latter tend to be quite distinctive ment. As a general rule , the potential for finding
96 B.S. Hart and A. G. Plint

marine trace fossils increases with the amount of the development of innovative sampling and moni­
sand in the deposit. Longshore bar systems and rip­ toring techniques. Difficulties encountered with
current deposits may be distinguishable (Leithold establishing the threshold of grain movement and
& Bourgeois, 1984; Massari & Parea, 1988; bedload transport rates under combined flows typical
Gruszcynski et al., 1993). of sandy shoreface settings are doubly present on
Studies of modern gravelly barriers have led to a gravelly shorefaces, where clast morphology, bulk
distinction between swash-aligned and drift-aligned textural parameters and fabric play important roles.
barrier systems, each with distinctive process­ The establishment of phase diagrams for bed con­
morphology characteristics (Carter & Orford, 199 1 ) . figurations remains problematic even in sandy shore­
Swash-aligned barriers are oriented nearly perpen­ face settings, and extremely little documentation is
dicular to the direction of wave approach and are available for gravelly shoreface settings. Empirical
dominated by onshore-offshore motions. Drift­ investigations of ancient deposits may provide, for
aligned barriers are those oriented at an angle to the some time to come, the best insight into the processes
direction of wave approach, and are dominated by operative on modern gravelly shorefaces.
longshore transport. Shore-parallel facies zonation
is best developed in swash-aligned beaches and are
'smeared' in drift-aligned systems (Carter & Orford, ACKNOWLEDGEMENTS
199 1 ) . Gravelly shoreface deposits tend to appear
massive, with crudely developed stratification and The work on shoreface conglomerates of the Car­
sedimentary structures revealed by subtle textural dium Formation which forms the nucleus of this
variations (Bergman & Walker, 1987 , 1988; Hart & paper was completed during the senior author's
Plint, 1989, 199 1 ; Plint & Hart, 1988). By careful doctoral studies at the University of Western
study of palaeocurrent information, both in the Ontario. Funding for that study was made possible
conglomerates and in interbedded sandstones, it through research grants (to A . G . P . ) from the Natu­
should be possible to recognize the dominant direc­ ral Sciences and Engineering Research Council, the
tion of wave approach (Hart & Plint, 1989 ) . Department of Energy, Mines and Resources and
the University of Western Ontario. Additional logis­
tical support was provided by Canadian Hunter Ltd . ,
SUMMARY Esso Resources Ltd . , Home Oil Ltd. and Unocal
Canada Ltd. To all these agencies and companies,
At present, sufficient criteria exist to permit the we are very grateful . Helpful comments by reviewers
identification of gravelly shoreface deposits, at least E. Leithold and F. Massari are greatly appreciated.
in outcrop. However, few truly 'diagnostic' criteria We thank the Canadian Society of Petroleum Geol­
exist, with faunal or trace faunal evidence being ogists for permission to reproduce Figs 4A, SA, 9A
possible exceptions. The inadequacy of previously and 12A, originally published in Plint & Hart ( 1988).
proposed diagnostic criteria (such as shape para­ This paper was written during the tenure of a Visiting
meters) reflects the range of processes that can Fellowship by the senior author at the Geological
affect the character of shoreface conglomerates. Survey of Canada's Pacific Geoscience Centre.
In outcrop, the stratification style, sedimentary
structures and palaeocurrent information can be
powerful indicators of depositional process. In the REFERENCES
subsurface , conglomerate body morphology, orien­
tation and stratigraphical relation to other units may ARNOTT, R.W.C. ( 1991) The Carrot Creek 'K' Pool, Car­
aid interpretation. In both outcrop and subsurface , dium Formation, Alberta - a conglomeratic reservoir
detailed facies analysis should make it possible to related to a wave-reworked distributary mouth bar com­
plex. Bull. Can. petrol. Geol. , 39, 43-53.
distinguish subtypes of gravelly shoreface deposits BERGMAN, K . M . & WA L KER , R . G . ( 1987) The importance
by clarifying the nature and relative importance of of sea level fluctuations in the formation of linear con­
specific depositional processes (e.g. wave-induced glomerate bodies; Carrot Creek Member of Cardium
and gravity driven currents, etc . ) . Formation , Cretaceous Western Interior Seaway,
Alberta, Canada. J. sediment. Petrol. , 57, 651 - 665 .
More studies are needed t o document the range of
BERGMAN, K . M . & WALKER, R.G. ( 1 988) Formation of
sedimentary processes and structures present on Cardium erosion surface E5 , and associated deposition
modern gravelly shorefaces. To do so will require of conglomerate: Carrot Creek Field, Cretaceous West-
 Gravelly shorefaces 97

ern Interior Seaway, Alberta. I n : Sequences, Strati­ small coastal streams, southwestern Oregon . I n : Coastal
graphy, Sedimentology: Surface and Subswface (Eds Geomorphology (Ed. Coates, D . R . ) , pp. 115- 140. State
James, D . P . & Leckie, D . A . ) . Can . Soc. petrol. Geol . , U niversity of New York at Binghamton .
Calgary, Memoir 1 5 , 15-24. DAVIDSON-ARNOlT, R . G . D . & GREENWOOD, B . ( 1976)
BLUCK, B . J . ( 1967) Sedimentation of beach gravels: Facies relationships on a barred coast, Kouchibouguac
examples from South Wales. J. sediment. Petrol. , 37, Bay, New Brunswick, Canada. I n : Beach and Nearshore
128- 156. Sedimentation (Eds Davis, R.A. & Ethington , R . L . ) .
BOURGEOIS, J. & LEITH OLD, E.L. ( 1984) Wave-worked Spec. Pub ! . Soc. econ. Paleont. Mineral, Tulsa, 24,
conglomerates - depositional processes and criteria for 149- 168.
recognition. I n : Sedimentology of Gravels and Conglom­ DoBK.JNS, J . E . & FOLK, R . L . ( 1970) Shape development on
erates (Eds Koster E . H . & Steel, R . J . ) . Can . Soc. Tahiti-Nui. J. sediment Petrol., 40, 1167- 1203.
petrol. Geol . , Calgary, Memoir 10, 331 - 343. D uBOIS, R . N . ( 1992) A re-evaluation of Bruun's Rule and
B RU U N , P . ( 1988) The Bruun Rule of erosion by sea­ supporting evidence. J. coast. Res. , 8, 618 - 628.
level rise: a discussion on large-scale two- and three­ DUPRE, W . R . , CLIFTON, H . E . & HUNTER, R.A. ( 1980)
dimensional usages. J. coast. Res. , 4, 627-648. Modern sedimentary facies of the open Pacific Coast and
B UJALESKY, G . C . & GONZALEZ-BONORINO, G. ( 1991) Pleistocene analogs from Monterey Bay, California. I n :
Gravel spit stabilized by unusual (?) high-energy wave Quaternary Depositional Environments of the Pacific
climate in bay side, Tierra del Fuego. In: Coastal Sedi­ Coast (Eds Field, M . E . et a!.), pp. 105 - 120. Soc. econ.
ments '91, American Society of Civil Engineers, Seattle, Paleont. Mineral. Pacific Section, Tulsa.
pp. 960-974. EMERY , K . O . ( 1955) Grain size of marine beach gravels.
CAILLEUX, A. ( 1945) Distinction des galets marins et flu­ J. Ceo!. , 63, 39-49.
viatiles. Bull. Soc. geol. France, 5, 325-404. ETHRIDGE , F.G. & WESCOTT, W.A. ( 1984) Tectonic setting
CANT, D . J . & ETHIER, V . G . ( 1984) Lithology-dependent and recognition of hydrocarbon reservoir potential of
diagenetic control of reservoir properties of conglomer­ fan-delta deposits. I n : Sedimentology of Gravels and
ates, Falher Member, Elmworth Field, Alberta. Bull. Conglomerates (Eds Koster, E . H . & Steel, R.J . ) . Can.
Am. Assoc. petrol. Ceo!. , 68, 1044- 1 054. Soc. petrol. Geol . , Calgary, Memoir 10, 2 17 -235 .
CARR, A.P. ( 1969) Size grading along a pebble beach: FORBES, D . L . & BOYD, R . ( 1989) Gravel ripples o n the
Chesil Beach, England. J. sediment. Petrol. , 39, inner Scotian Shelf. J. sediment. Petrol., 57, 46-54.
297-311. FoRBES, D . L. & TAYLOR, R . B . ( 1987) Coarse-grained beach
CARTER, R.W.G. ( 1988) Coastal Environments. Academic sedimentation under paraglacial conditions, Canadian
Press, New York, 617 pp. Atlantic coast. I n : Glaciated Coasts ( Eds Fitzgerald , D .
CARTER, R . W . G . & ORFORD, J . D . ( 1981) Overwash pro­ & Rosen P . ) , pp. 5 1-86. Academic Press, New York.
cesses along a gravel beach in southeast Ireland. Earth GILLIE, R . D . ( 1983) Field measurements of gravel move­
Sutf Process. Landf , 6, 413-426. ment by wave action on the inner continentia! shelf. I n :
CARTER, R . W . G . & ORFORD, J .D . ( 1984) Coarse clastic Proceedings of the Canadian Coastal Conference (Ed.
barrier beaches: a discussion of the distinctive dynamic Holden, B . J . ) , pp. 73-89. National Research Council of
and morphosedimentary characteristics. Mar. Ceo!. , 60, Canada, Ottawa.
377- 389. GRUSZCZYNSKI, M . , RUDOWSKI, S., SEMIL, J . , SLOMINSKI,
CARTER, R.W.G. & ORFORD, J . D . ( 1991) The sedimentary J. & ZROBEK, J. ( 1993) Rip currents as a geological tool.
organization and behaviour of drift-aligned gravel bar­ Sedimentology, 49, 2 17 -236.
riers. l n : Coastal Sediments '91, American Society of HARMS, J . C . , SOUTHARD, J . B . & WALKER, R . G . ( 1982)
Civil Engineers, Seattle, pp. 934- 948. Structures and Sequences in Clastic Rocks. Society of
CHEEL, R.J . & LECKIE, D .A. ( 1992) Coarse-grained storm Economic Palaeontologists and Mineralogists, Tulsa,
beds of the Upper Cretaceous Chungo Member (Wapiabi Short Course 9, 200 pp.
Formation), southern Alberta, Canada. J. sediment. HART, B .S. ( 1990) The sedimentology and stratigraphy of
Petrol., 62, 933-945. the Upper Cretaceous Cardium Formation in northwestern
CHIOCCI, F.L. & Cuno N , H . E . ( 1991) Gravel-filled gutter Alberta and adjacent British Columbia. Unpublished
casts in nearshore facies - indicators of ancient shore­ PhD thesis, University of Western Ontario, London, ·

line trend. In: From Shoreline to A byss: Contributions in Ontario, 505 pp.
Marine Geology in Honor of Francis Parker Shepard. HART, B.S. ( 1991) A study of pebble shape from gravelly
(Ed. Osborne, R . H . ) Soc. econ. Paleont. Miner. Spec. shoreface deposits. Sediment Ceo!. , 73, 185 - 189.
Pub!. No. 46, pp. 67-76. HART, B . S . & Pu NT , A . G . ( 1989) Gravelly shoreface
CLARKE, R . H. ( 1979) Reservoir properties of conglomer­ deposits: a comparison of modern and ancient facies
ates and conglomeratic sandstones. Bull. Am. Assoc. sequences. Sedimentology, 36, 55 1-557 .
petrol. Ceo!. , 63, 799-803. HART, B . S . & Pu NT , A . G . ( 1991) Conglomeratic shore­
CuFrON , H . E . ( 1973) Pebble segregation and bed lenticu­ face deposits from the Cretaceous Cardium Formation,
larity in wave-worked versus alluvial gravel. Sedimen­ Alberta, Canada. l n : Coastal Sediments '91, A merican
tology, 20, 173- 187. Society of Civil Engineers, Seattle, pp. 949-959.
CunoN, H . E . ( 1981) Progradational sequences in Miocene HART, B.S. & Pu N T , A . G . ( 1993a) Origin of an erosion
shoreline deposits, southeastern Caliente Range, surface in shoreface sandstones of the Kakwa Member
California. J. sediment. Petrol. , 5 1 , 165 - 184. (Upper Cretaceous Cardium Formation, Canada ) :
CLIFTON , H . E . , PHILLIPS, R.L. & HUNTER, R . E . ( 1973) importance for reconstruction o f stratal geometry and
Depositional structures and processes in the mouths of depositional history. I n : Sequence Stratigraphy and Facies
98 B.S. Hart and A. G. Plint

Associations (Eds Posamentier, H.W. , Summerhayes, dominated environments - examples from the Miocene
C . P . , Haq, B . U . & Allen, G . P . ) , Spec. Pubis int. Ass. of south-west Oregon. Sedimentology, 3 1 , 749-775.
Sediment No . 18, 45 1 - 467. Blackwell Scientific Publi­ L1u , Y. & GASTALDO, R . A . ( 1992) Characteristics and
cations, Oxford. provenance of log-transported gravels in a Carboniferous
HART, B.S. & Pu NT , A . G . ( 1 993b) Tectonic influence on channel deposit. J. sediment. Petrol. , 62, 1072 - 1083.
deposition and erosion in a ramp setting: Upper MAEJIMA, W. ( 1982) Texture and stratification of gravelly
Cretaceous Cardium Formation, Alberta foreland basin. beach sediments, Enju Beach, Kii Peninsula, Japan . J.
Bull. Am. Assoc. petrol. Ceo/. , 77, 2092-2 107. Geosci. Osaka City Univ. , 25, 35-5 1 .
HART, B . S . , VANTFOORT, R . M . & Pu NT , A.G. ( 1 990) Dis­ MASSARI, F. & PAREA, G.C. ( 1988) Progradational gravel
cussion on 'Is there evidence for geostrophic currents beach sequences in a moderate- to high-energy, micro­
preserved in the sedimentary record of inner to middle­ tidal marine environment. Sedimentology, 35 , 88 1 - 9 1 3 .
shelf deposits?' J. sediment. Petrol. , 60, 633 -635. MASSARI, F . , PAREA, G . C . , RAINONE, M . L . , VEDOVATO,
HEI N , F.J. & WALKER, R . G . ( 1 977) Bar evolution and L. & YIVALDA, P. ( 1 986) Elementi di sedimentologia
development of stratification in the gravelly, braided, delle paleospiagge Pleistoceniche marchigiane. I n : Alti
Kicking Horse River, British Columbia. Can. J. Earth Riunione Gruppo Sedimentologia CNR, Ancona, 5 - 7
Sci. , 14, 562-570. giugno 1986, p p . 8 1 - 103.
HowARD, J . L. ( 1 992) An evaluation of shape indices as MATHEWS, E . R. ( 1 980) Observations of beach gravel trans­
palaeoenvironmental indicators using quartzite and port, Wellington Harbour entrance, New Zealand. N.Z.
metavolcanic clasts in Upper Cretaceous to Palaeogene J. Geol. Geophys. , 23 , 209-222.
beach, river and submarine fan deposits. Sedimentology, MIDDLETON , G . V . ( 1 973) Johannes Walther's law of the
39, 47 1 -486. correlation of facies. Geol. Soc. Am. Bull. , 84, 979-988.
HowELL, D . G . & LINK, M . H . ( 1 979) Eocene conglomerate NEATE, D.J . M . ( 1 967) Underwater pebble grading of Chesil
sedimentology and basin analysis, San Diego and the Bank . Proc. Ceo/. Assoc. , 78, 419-426.
southern California borderland. J. sediment. Petrol. , 49, NEMEC, W . & STEEL, R.J. ( 1 984) Alluvial and coastal
5 1 7 -540. conglomerates: their significant features and some com­
HUNTER, R . E . , CLIFTON , H . E . & PHILLIPS, R . L . ( 1 979) ments on gravelly mass-flow deposits. In: Sedimentology
Depositional processes, sedimentary structures and of Gravels and Conglomerates (Eds Koster, E . H . &
predicted vertical sequences in barred nearshore sys­ Steel, R.J . ) . Can. Soc. petrol. Geol . , Calgary, Memoir
tems, southern Oregon coast. J. sediment. Petrol. , 49, 10, 1 - 3 1 .
7 1 1 -726. NIEDORODA, A . W . , Swwr, O . J .P. & HOPKINS, T.S. ( 1 985)
KmsoN, C. & CARR, A . P . ( 1 959) The movement of shingle The shoreface. I n : Coastal Sedimentary Environments
over the seabed close inshore. Geogr. J. , 125 , 380-389. (Ed. Davis, R . A . , Jr. ) , pp. 533 - 624, Springer-Verlag,
KIRK, R . M . ( 1 980) Mixed sand and gravel beaches: mor­ New York.
phology, processes and sediments. Prog. phys. Geogr. , NIELSEN, L . H . , JOHANNESSEN, P.N. & SURLYK, F. ( 1 988) A
4, 189 - 2 10. Late Pleistocene coarse-grained spit-platform sequence
KLEINSPEH N , K.L. , STEEL, R .J . , JOHANNESSEN , E . & N ET­ in northern Jylland, Denmark. Sedimentology, 35 ,
LAND, A. ( 1 984) Conglomeratic fan-delta sequences, 9 1 5 - 937.
Late Carboniferous-Early Permian, Western Spits­ OGREN , D . E . & WAAG, C.J. ( 1 986) Orientation of cobble
bergen. I n : Sedimentology of Gravels and Conglomerates and boulder beach clasts. Sediment. Geol. , 47, 69-76.
(Eds Koster, E . H . & Steel, R.J . ) . Can . Soc. petrol. ORFORD, J . 0 . ( 1975) Discrimination of particle zonation
Geo l . , Calgary, Memoir 1 0 , 279- 294. on a pebble beach. Sedimentology, 22, 441 -463.
KOMAR, P . O . ( 1 976) Beach Processes and Sedimentation . PArnsoN , S . A . J . & WALKER, R . G . ( 1 992) Deposition and
Prentice-Hall, Englewood Cliffs, N J , 429 pp. interpretation of long, narrow sandbodies underlain by a
KOMAR, P . O . ( 1 987) Selective gravel entrainment and the basinwide erosion surface: Cardium Formation, Cre­
empirical evaluation of flow competence . Sedimentology, taceous Western Interior Seaway, Alberta, Canada. J.
34, 1 165- 1 176. sediment. Petrol. , 42, 292-309.
KOMAR, P.O. & L1, Z. ( 1 986) Pivoting analysis of the PuNT, A . G . & HART, B .S . ( 1 988) Field Guide to the Upper
selective entrainment of sediments by shape and size Cretaceous Dunvegan (Cenomanian) and Cardium
with application to gravel threshold . Sedimentology, 33, (Turonian) Formations in the Dawson Creek - Fort St.
425-436. John area, British Columbia. Canadian Society of Pet-·
LECKIE, D . A . ( 1 988) Wave-formed, coarse-grained ripples roleum Geologists' Field Guide to 'Sequences, Stratigra-·
and their relationship to hummocky cross-stratification. phy, Sedimentology: Surface and Subsurface Technical
J. sediment. Petrol. , 58 , 607 :..._ 6 22. Meeting', 14- 1 6 September, Calgary, Alberta, 5 1 pp.
LECKIE, D . A . & WALKER, R . G . ( 1 982) Storm- and tide­ PuNT, A.G. & WALKER, R . G . ( 1 987) Cardium Formation
dominated shorelines in Cretaceous Moosebar-Lower 8. Facies and environments of the Cardium shoreline
Gates interval: outcrop equivalents of Deep Basin gas and coastal plain in the Kakwa Field and adjacent areas,
trap in Western Canada. Bull. Am. Assoc. petrol. Geol. , northwestern Alberta. Bull. Can. petrol. Geol. , 35,
66, 138- 157. 48-64.
LEITHOLD, E.L. ( 1989) Depositional processes on an PUNT, A . G . , WALKER, R . G . & BERGMAN, K . M . ( 1 986)
ancient and modern muddy shelf, northern California. Cardium Formation 6. Stratigraphic framework of the
Sedimentology, 36, 179-202. Cardium in subsurface. Bull. Can. petrol. Geol. , 34,
LEITH OLD, E . L . & BouRGEOIS, J. ( 1 984) Characteristics of 2 1 3 -225.
coarse-grained sequences deposited in nearshore, wave- PoSTMA , G. & NEMEC, W . ( 1 990) Regressive and trans-·
 Gravelly shorefaces 99

gressive sequences in a raised Holocene gravelly beach, southeast Jamaica. Bull. Am. Assoc. petrol. Ceo!. , 64,
southwestern Crete. Sedimentology, 37, 907-920. 374-399.
READING, H . G . ( 1986) I ntroduction . I n : Sedimentary En­ WHITING , P .J . , DIETRICH, W . E . , LEOPOLD, L . B . , DRAKE,
vironments and Facies (Ed. Reading, H .G ) , pp. 1 -3 .
. T . G . & SHREVE, R . L . (1988) Bedload sheets in hetero­
Blackwell Scientific Publications, Oxford. geneous sediment. Geology, 16, 105- 108.
REDDERING, J . S . V . & ILLENBERGER, W . K . ( 1988) Dis­ WILLIAMS, A .T. & CALDWELL, N . E . ( 1988) Particle size
cussion on palaeoenvironmental significance of clast and shape in pebble-beach sedimentation. Mar. Ceo!. ,
shape in the Nardouw Formation, Cape Supergroup. S. 82, 199-215.
Afi·. J. Ceo!. , 9 1 , 554-555 . WILLIAMS, J .J . , THORN E , P.O. & H EATHERSHAW, A.D.
Ross, N. & LONG, B . F . ( 1989) Evolution morpho­ ( 1 989) Comparisons between acoustic measurements and
sedimentaire de Ia barre de deferlement: un example predictions of bedload transport of marine gravels.
dans le golfe du Saint-Laurent, Quebec. Geogr. phys. Sedimentology, 36, 973-979.
Quat. , 43, 377-388. WooDBORNE, M . W . , RoGERS, J. & JARMA N , N. ( 1989) The
SHIPP, R.C. ( 1984) Bedforms and depositional sedimentary geological significance of kelp-rafted rock along the west
structures of a barred nearshore system, eastern Long coast of South Africa. Ceo-Mar. Lett. , 9, 109 - 1 18 .
Island, New York. Mar. Ceo!. , 60, 235-259. WRJGHT, L . D . & SHORT, A . D . (1984) Morphodynamic
SHORT, A . D . ( 1984) Beach and nearshore facies: southeast variability of surf zones and beaches: a synthesis. Mar.
Australia. Mar. Ceo!. , 60, 261 -282. Ceo!. , 56, 93- 1 18.
WESCOTr, W.A. & ETHRJDGE, F.G. ( 1 980) Fan-delta sedi­ ZEN KOVITCH , V.P. ( 1967) Processes of Coastal Develop­
mentology and tectonic setting - Yallahs fan delta, ment. Oliver and Boyd, London, 738 pp.
Spec. Pubis int. Ass. Sediment. (1995) 22, 101-135

The return of 'The Fan That Never Was':

Westphalian turbidite systems in the Variscan Culm Basin:
Bude Formation (southwest England)

R OB E R T V . B U R N E
Australian Geological Survey Organisation, P O B ox 378, Canberra,
ACT 2601, Australia

ABSTRACT

The Westphalian Bude Formation of southwest England was one of the first successions to be
interpreted as a subsea fan deposit. This interpretation has been questioned recently, and the alternative
depositional environment of a wave-affected shelf has been proposed. The tectonic setting, palaeo­
geographical relationships and facies succession of the formation have been reassessed in an attempt to
resolve this controversy. The formation was deposited in the synorogenic Culm Basin. It contains thick
sandstone units, but is in other respects similar to the underlying turbidite-bearing Crackington
Formation (Namurian-Westphalian) . There is no depositional continuity with the contemporaneous
paralic and deltaic deposits of the Bideford Group that occur in a tectonically distinct succession near
Westward Ho! Four facies are identified in the Bude Formation; the black shale facies, the muddy
siltstone facies, the interbedded sandstone-shale facies and the sandstone-dominated facies. Both
thickening upward and thinning upward facies successions occur at some levels, but generally the facies
succession shows less predictable, although not random, alternations between the four facies. The
suggestion that the Bude Formation represents the deposits of a wave-influenced shelf does not stand up
to scrutiny. There is no substantial evidence for shallow water environments in the Formation . Evidence
for storm-generated hummocky cross-stratification and combined-flow wave ripples is equivocal. The
ichnofacies have no depth connotation and there is no evidence for sedimentological continuity with
undoubted deltaic successions. By contrast there is overwhelming evidence for deposition of turbidite
systems in an isolated fresh- or brackish-water basin. The black shale facies represent the deposits of
fine-grained turbidity currents in an anaerobic or dysaerobic basin. The muddy siltstone facies represent
the deposits of aerobic environments characterized by gentle currents, possibly fan-levee environments.
The interbedded sandstone-shale facies consist of a variety of turbidites and were deposited in either
levee, lobe or interchannel environments. The sandstone-dominated facies were deposited in fan-valley
environments, and include channel-fill successions. It is concluded that the Bude Formation was
deposited as subsea fans on the northern, inactive margin of a land-locked, foreland basin.

INTRODUCTION

The Westphalian Bude Formation (Owen, 1934; the coast near Westward Ho! in north Devon (Figs 2
King, 1966, 1967, 1971; Freshney et al., 1979) was & 3). Reading (1963) drew attention to the funda­
deposited in the synorogenic Culm Basin of south­ mental difference of opinion about the deposition
west England (Sedgewick & Murchison, 1840; of these successions. Owen (1950) and Prentice
Ussher, 1892; Thomas, 1988) (Figs 1 & 2). The (1960a, b) considered that they were examples of
formation crops out along the coast of north paralic 'Coal Measures', but Ashwin ( 1957, 1958)
Cornwall and west Devon (Fig. 2) and it had become interpreted them as turbidites. Reading ( 1963,
the custom to equate it with a contemporaneous p. 69) contrasted the evidence in the rocks around
succession, the Bideford Group, which crops out on Westward Ho! for 'successive advances into a basin
Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
101
102 R. V. Burne

52 '

----

5 1'

ST

50'

6' 3'

D Permian- Tertiary • Lizard and Start complexes ----&....- Thrust fault

[] I+++I
-- Normal fault
Upper Carboniferous Granite
� Strike- slip fault

� Devonian - Lower Carboniferous

BCFZ Bristol Channel fracture zone

SLFZ Sticklepath - Lustleigh fracture zone

Fig. 1. Location of the Culm Basin. (A) Location of southwest England with respect to the British Isles. (B) Location of
the Culm Basin with respect to the other principal tectonic elements of southwest England (after Hartley & Warr, 1990).
Suggested Silesian position of southwest England with respect to France produced by pre-deformation restoration along the
Bristol Channel-Bray Fault (after Holder & Leveridge, 1986), all other structures shown in present-day positions.

of a low coastal plain, possibly fronted by deltas, Subsequent workers confirmed the existence of
alternating with periods of reduced supply when paralic cycles in the Bideford Group, (comprising
basin conditions were re-established' with the fact the Northam and Abbotsham Formations), around
that 'No cyclic pattern of sedimentation has been Westward Ho! (Walker, 1964a, b; de Raaf et al. ,
recognized, nor has any unequivocal evidence of 1965; Money; 1966; Elliott, 1976). They have also
shallow water or coastal plain deposition been confirmed Reading's statement that (1963, p. 69)
observed' in the Bude Formation. This implied that 'The precise depositional environment of the Bude
'whilst coastal plains reached into north Devon in Sandstones is not easy to establish'. Sedimentologists
Lower Westphalian times, there is no evidence that had generally recognized that the Bude Formation
they extended into Cornwall'. was, in part, a turbidite succession, but did not
 Bude Formation subsea fans 103

'40 '20 '30 4"20'

Bideford Bay

'30
Hartland
Point

51"00'

'20

Bude Bay

'10

50'50' BUDE

· . p;.
.

'00

]
• Post- Silesian strata --Fault

SLFZ Sticklepath - Lustleigh fracture

WESTWARD HO!­ zone
BIDEFORD FAULT BLOCK
m Greencliff Beds
CULM BASIN

D Bude Formation
� Town

SILESIAN
t-=-� Bideford group & Westward
Ho! Formation
[·.:.·: ·_.:J Crackington Formation
D Village

1 30 National grid number tor Great

Britain

� Pre- Silesian strata

Fig, 2, Geological map of north Cornwall and west Devon. (Modified after Burne & Moore, 1971; Thomas, 1988.)

appear to be a 'true flysch' (Goldring & Seilacher, basin was generally filled with fresh or brackish
1971). It contained some unusual characteristics, water, although rare marine incursions occurred,
such as massive sandstones, which de Raaf (pers. probably the result of glacio-eustatic oscillations.
comm., 1966) had informally termed 'Budes' and The mature, fine-grained sandstones were derived
interpreted as some form of unusually fine-grained from a land mass of older sedimentary rocks to the
'fluxoturbidite' (Unrug, 1963, 1965). north and were deposited as turbidites from bottom­
A sedimentological facies analysis of the middle hugging underflows. The turbidites included con­
Bude Formation (Fig. 4) was undertaken to identify ventional Bouma sequences, traction carpet deposits
features definitive, rather than indicative, of depo­ with current-reworked tops, irregularly laminated
sitional processes and environments (Burne, sandstones, and 'slurried' beds produced by the
1969a, b; 1970, 1971, 1973, 1976; Burne & Moore, impact of dense turbidity currents on a soft substrate.
1971). It was concluded that the Bude Formation The major 'slumped' beds had the same origin as the
had been deposited in a land-locked basin, akin to 'slurried' beds, but some sandstone dykes and
the present Black Sea or Caspian Sea, that had been volcanoes were formed by load-induced post­
isolated by Hercynian continental collision. The depositional water-escape. The depositional
104 R. V. Burne

R
HARTLAND
TO
BOSCASTLE

a
WESTWARD
HOI

z
"'
:J
"'
I
0..
f­
Amaliae Marine Band rJ)
UJ
s:

Usteri Marine Band

Marine

• Greencliff Beds

Q Bideford Group

D Westward Ho! Formation

� Bude Formation

� Crackington Formation

INDEX SHALES

ccs Clovelly Court Shale

DPS Deer Park Shale

ESS Embury Shale

GAS Gull Rock Shale

HQS Hartland Quay Shale Fig. 3. Stratigraphical correlation

SMS Sandy Mouth Shale between the Culm Basin and
SPS Saturday's Pit Shale Westward Ho! Silesian sequences.
TCS Toms Cove Shale
Brigantian
(Modified from King, 1967 ;
WGS Warren Gutter Shale •oom Thomas, 1988.)

environment was that of a submarine fan, with open­ Melvin (1976, 1986) reached similar conclusions
fan, fan-levee, active fan-channel, and inactive fan­ from a study of higher levels in the Bude Formation
channel facies being distinguished. Despite the (Fig. 4). He suggested that deposition was from
apparently random nature of much of the succession, turbidity currents in a relatively shallow basin and
it was recognized that predictable associations noted that sediment coarser than fine-grained sand
occurred in parts (Burne, 1969b, Fig. 20) (Fig. 5), was absent, and the organization of the bundles of
including thickening upward successions, interpreted sandstones was complex. He suggested that sediment
as being due to the advance of a fan-channel/levee was supplied from a delta to the north by resedimen··
complex over an open fan, and thinning upward tation of delta-front sands, and by slumping of delta··
successions, interpreted as being the result of fan­ front muds and silts to form a lower prodelta turbidite
channel fill and abandonment. fan. Melvin could not identify well-developed
 Bude Formation subsea fans 105

B Nodular shale

UJ
§- Shale

0
::;)
"' � Siltstone

D Mudstone

G Slumped bed

Thick-bedded and massive

[]] sandstone characteristics of

BudeFormation

� Shales with thin sandstone

�.
Medium to thinly bedded
sandstones with subordinate
shales and siltstones

INDEX SHALES
600
WGS Warren Gutter Shale
'Anthracoceras' aegiranum
horizon

SMS Sandy Mouth Shale

SPS Saturday's Pit Shale (key

shale W)

TCS Tom's Cove Shale (key shale

Q)

LS Longpeak Shale

HQS Hartland Quay Shale

Gastrioceras amaliae horizon

GRS Gull Rock Shale

Gastrioceras /isteri horizon

ES Embury Shale
Gastrioceras subcrenatum
400 200 horizon

Fig. 4. Representative lithological section of the Crackington Formation and Bude Formation based on exposures between
Duckpool (SS 201 1 15) and Embury Beach (SS 214 195) (modified after Freshney et a/., 1979). King (1967) originally
defined the base of the Bude Formation at a level 265 m below the base of the Tom's Cove Shale and the top of the
formation at a level 150 m above the same datum, a section here referred to as the middle Bude Formation.

thickening upwards and thinning upward successions structures occur throughout the succession, spaced
as described by Burne (1969b) from lower in the no more than a few metres apart.'
Bude Formation. However, he presented a facies This paper attempts to resolve the controversy
model of various fan, fan-levee and fan-channel that has resulted from these contrasting inter­
relationships that accounted for the facies suc­ pretations by reconsidering the tectonic setting,
cessions observed in measured sections. Apart from palaeogeographic relationships and facies analysis of
different degrees of organization in the succession, the Bude Formation.
there is general correspondence between the fan
interpretation suggested by Burne (1969b) and that
proposed by Melvin (1986). THE TECTONIC SETTING
Higgs ( 1987) strongly criticized Melvin's inter­ OF THE CULM BASIN
pretation in a vigorous discussion provocatively
titled 'The Fan That Never Was?'. Higgs (1983, The convergence and collision of the Laurussian
1984, 1986a, b; 1991) tentatively identified wave­ and Gondwanan continental masses between the
influenced structures in the Bude Formation, and Devonian and the Early Permian (Burne, 1969b,
pursued the significance of this interpretation to its 1973; Ziegler, 1990) generated a deformation front
logical conclusion, that the 'Bude Formation was that migrated northwards with time (Besly, 1988) to
deposited largely, if not entirely, above storm wave form the Variscan orogenic belt. During the Middle
base, as indicated by the fact that wave-influenced Devonian and Early Carboniferous, volcanicity and
106 R. V. Burne

COMPOSITE SEQUENCE Possible interpretation

Mudstone
Open fan or inter
dominated
-channel sediment
sequence

'Thinning upward'
sequence
1-:: .
/
--- --- -�-
::--......_
�-=--

�
::-�-- :� ----------- ----:-:-,
l Sediment in confines
of partly active
channel

-----==- � -==-

Sandstone
Sediment in confines
dominated ----�-�-
of active large
sequence
channels

.. ·.:::.:.:_"· Sediment of channel

'Thickening upward'
levee
sequence

Mudstone
dominated
sequence
Open fan or inter­
channel sediment

I Bed scale
Approx. 1m
(bu1 variable)

16/04117
 Bude Formation subsea fans 107

subsidence accompanied a period of extension 197 1) . The occurrence of correlatable marine bands
in southwest England. Regional inversion of this throughout these basins indicates that they were
Devonian to Early Carboniferous basin occurred in separated by very low topographical relief, with
the early Namurian. Loading of its northern margin the marine bands resulting from short-lived marine
by the resulting thrust nappe development (Hecht, transgressions that entered the basins from the
1992) formed the Culm Basin that lay north of Moscow Depression (Ziegler, 1990). The trans­
the developing Variscan Orogen and south of the gressions probably reflect high sea-levels in a series
cratonic Wales-Brabant Massif (Hartley & Warr, of short-term glacio-eustatic fluctuations, the last of
1990) (Fig. 1). Evidence for a thin skinned thrust­ which occurred in Westphalian D time.
and-nappe regime implies that the orogen did not It was only in the Culm Basin that deep-water
attain 'Alpine' proportions (Isaac et a/. , 1982) . conditions persisted into the Westphalian (Ziegler,
Hartley & Warr (1990) suggest that the Culm Basin 1990). Deposition took place on the northern, non­
maintained its present geographical relationship with orogenic flank of the basin, as shown by the lack of
South Wales throughout the Variscan deformation. palaeocurrents from the south (Fig. 6) (Ashwin,
However, others suggest that the northern boundary 1957; Burne, 1969b; Freshney eta/. , 1979; Melvin,
of the basin corresponded to an area west of present­ 1986; Higgs, 1991) and by the compositionally
day Le Havre, France (Fig. 1) (Holder & Leveridge, mature, fine-grained nature of the sediments (Fig. 7)
1986), and subsequent transcurrent faulting along (Burne, 1969b; Freshney eta/. ; 1979; Melvin, 1986;
the Bristol Channel and Bray Fault has displaced the Haslam & Scrivener, 199 1). Namurian processes
basin to its present position, south of the Bristol deposited thin-bedded turbidites from currents flow­
Channel. ing eastward along the axis of the basin, forming the
In the early Namurian the Culm Basin was prob­ Crackington Formation (Freshney et a/. , 1979) .
ably part of a continuous deep-water flysch basin, Deposition of the Bude Formation began in the
the Rhenohercynian Basin, that extended from Early Westphalian. Although 50% of the beds in the
Cornwall to Upper Silesia (Ziegler, 1990) . Sedimen­ Bude Formation resemble those of the underlying
tation began to exceed subsidence in parts of this Crackington Formation, it is distinguished by its
basin during the Namurian, and the Culm Basin content of massive sandstones (Fig. 4). The currents
became one of a number of basins isolated during that deposited the Bude Formation generally flowed
the last stage of continental collision (Franke & towards the south and southwest, although easterly
Engel, 1988). In the Culm Basin evidence for flowing currents occurred at some horizons (Fig. 6)
this change may be seen in the turbidites of the (Burne, 1969b, Freshney et a/. , 1979, Melvin, 1986).
Crackington Formation. Marine goniatites, which No post-Bude-Formation sediments are preserved
had been scattered throughout the shales lower in in the Culm Basin.
the succession, became restricted in the uppermost
Namurian to marine bands deposited by marine
incursions into an otherwise brackish-water basin TECTONICALLY JUXTAPOSED
(Freshney et a/., 1979). By the early Westphalian SUCCESSIONS IN THE
most Rhenohercynian basins had shallowed and CULM BASIN
established paralic conditions providing the classic
localities for coal-measure cyclothems (Reading, Freshney & Taylor ( 1972) proposed a simple stra-

Fig. 5. (Opposite.) Composite 'sequence' (i.e. succession in present usage) for the Bude Formation reproduced from
Burne ( 1969b, Fig. 20) showing the original interpretation, including the first recognition of 'thickening' and 'thinning'
upward 'sequences' in subsea fan deposits. The composite succession includes only one representative of the various bed
types· found in each part of the succession , and the scale shown relates to the thickness of these individual beds. Complete
successions are actually 10-60 m thick. In the usage of the present paper the mudstone-dominated sequence corresponds to
the black shale facies, now interpreted as fine-grained turbidites in an anaerobic or dysaerobic basin; the thickening upward
sequence corresponds to part of the interbedded sandstone- shale facies, now interpreted as turbidites deposited in either
levee , lobe or interchannel environments; the sandstone-dominated sequence corresponds to the sandstone-dominated
facies, now interpreted as the deposits of fan-valley environments, including channel-fill successions; and the thinning
upward sequence corresponds to part of the interbedded sandstone-shale facies now interpreted as the deposits of aerobic
environments characterized by gentle currents, possibly fan levees.
108 R. V. Burne

A Tool Marks Westphalian deposits which 'while they still contain

sequences of turbidites, have within them substantial
numbers of massive sandstones and other associated
facies characteristic of the rocks described by King
(1966)' (Freshney & Taylor, 1972, p. 467). An ar­
bitrary junction between the two formations was set
at the top of the Hartland Quay Shale (Freshney
et al., 1979), a marker bed containing goniatites of
the Gastrioceras amaliae marine band. The Bude
Directions Formation includes the youngest Westphalian rocks
exposed in southwest England, the Warren Gutter
Senses Only
Shale, which contain a fauna correlated with the
'Anthracoceras' aegiranum marine band (Freshney
8 Current Scours et al., 1979).
The sedimentological contrast between the
Crackington and Bude Formations and the con­
temporaneous paralic sediments exposed near
Westward Ho! led Reading (1965) to suggest that
the Culm Basin contained a number of sedimen­
tologically distinct though synchronous strati­
graphical successions separated by low-angle thrust
faults. Burne & Moore (1971) concluded that
Directions two distinct stratigraphical successions could be
Senses Only
recognized; a northern, essentially deltaic succession

�
(the Westward Ho! Formation, Northam Formation,
C Cross Laminations
Abbotsham Formation and the Greencliff Beds),
N and a contrasting southern succession, consisting
entirely of basin facies, comprising the Crackington
I Formation and the Bude Formation of Freshney
et al. (1979) (Figs 2 & 3). The two successions are
10 15 20
%
separated by a major E- W trending normal fault,
with downthrow several hundred metres to the north
Directions N = Number of readings
(Freshney & Taylor, 1972). This fault is exposed on
Fig. 6. Structurally corrected palaeocurrents recorded the coast (Burne & Moore, 1971, pp. 293-294) and
from sole marks and cross laminations in the middle Bude can be traced inland eastward to the Mole Valley
Formation (data of Burne, 1969b). (National Grid Reference SS 675 260), 3 km east of
South Molton (Fig. 1).
tigraphy for the Culm Basin (Figs 3 & 4) in which all The existence of tectonically juxtaposed suc··
the dominantly turbidite sandstone-mudstone cessions in the Culm Trough is also indicated by
successions of Namurian and basal Westphalian contrasting levels of organic maturation shown by
rocks were classified as Crackington Formation, vitrinite reflectance measurements from the two suc­
whereas the Bude Formation contained the younger cessions (Cornford et al., 1987). Samples from the

Fig. 7. (Opposite.) Petrology and grain-size distributions of the various bed types in the Bude Formation. (A) Comparison
of sandstone composition of the Bude Formation (data of Burne, 1969b; Freshney et al., 1979; Melvin, 1986); the
Crackington Formation (data of Melvin, 1986); and the present-day Mississippi subsea fan (data of Roberts & Thayer,
1985). (B) Matrix composition of various Bude Formation bed types. Note high matrix content of 'slumped' and 'slurried''
beds, low matrix content of laminated and structureless sandstones, and intermediate matrix content of graded silty
sandstones (data of Burne, 1969b) . (C) Plot of mean grain size (<!>) against sorting ( oG ) of Bude Formation sandstone grain··
size distributions, with and without matrix included (data of Burne, 1969b). Fields recognized by Melvin (1986) for the
Bude Formation and Crackington Formation sandstones are outlined (note that Melvin measured o1 ) .
 Bude Formation subsea fans 109

A QUARTZ

�
Bude Formation
sandstones

D
Crackington Formation
sandstones

E2J] Mississippi Fan sands

FELDSPAR ROCK FRAGMENTS

8 QUARTZ

m; Cross laminated beds

Structureless & irregularly

LIJ laminated sandstones

� Graded silty sandstones

E:;l
"Slurried" &
D "slumped" beds

FELDSPAR MATRIX
&
ROCK FRAGMENTS

c
3.0

2.0

MELVIN (1986)
0G Bude Formation data

}
MELVIN (1986)
Crackington Formation data
1.0
0 Matrix Excluded
BURNE (1969b)
6. Matrix Included Bude Formation data
110 R. V. Burne

section between Westward Ho! and Abbotsham form the succession in the area of Westward Ho! and
a statistically different group compared with those Bideford is tectonically juxtaposed to the Culm
from stratigraphically equivalent rocks from the area Basin succession (Figs 1 & 2).
between Hartland Quay (National Grid Reference
SS 223 248) and Crackington Haven (SX 143 968).
This suggests that the two areas have suffered differ­ FACIES ANALYSIS
ent thermal histories, i.e. different burial histories. OF THE BUDE FORMATION
'These two areas must either now be in juxtaposition
due to post burial tectonic events (e.g. thrusting) or Four facies are recognized in the Bude Formation,
they are not stratigraphic equivalents, the Bideford although these are, to a large extent, intergra­
Formation (i.e. Bideford Group) being younger dational. The term facies is used here to denote the
and hence less deeply buried than the Bude and assemblage of beds deposited within a depositional
Crackington Formations' (Cornford et al., 1987, system. A depositional system comprises a sedimen­
p. 463). The stratigraphical equivalence of the sec­ tary environment together with the processes that
tions was established by Edmonds et at. (1975, 1979) may operate within it. Beds are formed by individual
who identified the Gastrioceras amaliae marine band depositional processes.
550 m above the base of the Bideford Group. The
Hartland Quay Shale, although of a contrasting
Facies descriptions
facies, also contains this horizon and has been
defined by Freshney et al. (1979) as underlying the
Black shale facies
base of the Bude Formation (Fig. 3). Edmonds et at.
(1979) suggested that depositional continuity existed Black shales (Fig. 8) are the most persistent facies
between these successions and concluded that the in both the Bude Formation and the underlying
Northam and Abbotsham Formations (for which Crackington Formation and form valuable lithofacies
they proposed the name Bideford Formation, but for stratigraphical correlation (King, 1967; Freshney
for which the term Bideford Group is here retained)
lie conformably between the Crackington Formation
and the Bude Formation. Freshney et at. (1979) and
Melvin (1986) both based their interpretation of the
palaeogeography and facies of the Bude Formation
on implied continuity between the deltas of the
Bideford group and the depositional environments
of the Bude Formation. According to Freshney
(pers. comm., 1994) 'there is no physical evidence of
thrusts in the Bideford area, but it is possible that
strike-slip, east-west trending faults may define areas
of contrasting thermal and burial history'. Although
Freshney (pers. comm., 1994) points out that
'wedge-bedded sandstones characteristic of the
Bideford Group can be traced eastward beyond
Umberleigh (National Grid Reference SS 610 237)
and appear to pass further east into a number of
massive Bude-type sandstones', it has yet to be
established whether this transition represents depo­
sitional continuity or tectonic juxtaposition. How­
ever, Cornford et at. (1987) concluded that the
fluvial-deltaic Bideford Group is allochthonous
and therefore the palaeogeography of the basin
should not be constrained by the apparent associ­
Fig. 8. Black shale facies showing regular millimetre-scale
ation between the Bideford Group and the deeper
graded mudstone laminae with interbedded grey, muddy
water Bude Formation. This conclusion supports siltstone laminations, some with thin ripple cross­
the interpretation of Burne & Moore (1971) that laminations. Scale in centimetres (SS 1985 1313).
 Bude Formation subsea fans 111

& Taylor, 1972). The facies may attain a thickness of and Skolithos (Higgs, 1991). Freshney & Taylor
20 m (Fig. 4), and is dominated by uniform, dark (1972) found that in the lower parts of the Crack­
mudstone composed almost entirely of sharp-based, ington Formation a marine fauna of goniatites occurs
graded laminae 5-10 mm thick. Thin sharp-based throughout the shales, whereas from the Namurian
sets of ripple cross-laminated silt occur at the lower G1 zone the goniatites occur only in isolated bands
boundary of some of these laminations. Sole marks within some black shales. The Goniatites, which
on these silts comprise prod marks or, rarely, are truly marine, occur in bands between layers of
obstacle scours around burrow entrances. Isolated nodules that contain fossils of fish that were tolerant
thicker beds can occur in this facies. of low-salinity environments. Goniatites have been
The clay minerals of the shales consist dominantly found only at two horizons in the Bude Formation
of illite (Burne, 1969b). Undeformed shales also (Fig. 4), the Sandy Mouth Shale and the Warren
contain kaolinite, whereas in deformed shales this is Gutter Shale. In the Warren Gutter Shale, goniatites
replaced by chlorite, and grey shales contain more are present only as spat (Freshney et at. , 1979).
kaolinite (Freshney et al., 1979). Siderite bands and Goniatites have not been found in the middle
ankerite nodules are found in some shales. The Bude Formation, although fish remains and
grading of the laminae is due to an upward increase coprolites occur at two horizons (King, 1967).
in the content of organic carbon. Finely comminuted Cornuboniscus budensis and Elonichthys aitkeni,
plant debris is common on parting surfaces. Freshney both palaeoniscid bony fishes, the acanthodian
et al. (1979) record jarosite-natrojarosite efflor­ Acanthodes wardi and an eocarid crustacean
escences in sulphurous shales. Higgs (1991) found Crangopsis huxleyi occur in the Saturday's Pit Shale,
that carbon/sulphur (CIS) ratios (Berner & Raiswell, whereas the Coelocanth Rhabdoderma elegans
1983, 1984) in four out of five black-shale samples occurs in the Tom's Cove Shale (Fig. 4).
were less than 10.
The shales contain little evidence of biological
Muddy siltstone facies
activity apart from the common occurrence of
Planolites (Fig. 9) (King, 1967). Tops of interbedded This facies consists of grey to dark grey laminated
sandstones may contain Diplocraterion parallelum, muddy siltstones and thin, sharp-based silty sand­
an'd, more rarely Arenicolites, Teichichnus, Phyeodes stones that show parallel laminations and ripple drift
cross-lamination. In places the facies is disrupted by
intrastratal folding and small-scale synsedimentary
faults. Scouring occurs locally (Fig. 10). The muddy
siltstones are composed of carbonaceous material,
clay minerals, and silt-sized quartz particles. The
grading of the 1-10-mm-scale laminae reflects a
decrease in the size of the quartz grains and an
increase in the proportion of clay (Fig. 11). The
silstones are composed of quartz, illite, kaolinite,
chlorite, siderite and minor feldspar (Merriman,
quoted in Higgs, 1991). Higgs (1991) found that
organic carbon varied from 1.8% to 2.9%, plant
fragments were conspicuous, burrows are absent,
and C/S ratios (Berner & Raiswell, 1983, 1984)
range between 22.8 and 39.0.
King (1967) described Kouphnichnium (King,
1965; Goldring & Seilacher, 1971), i.e. xiphosurid
feeding trails from this facies. The animals moved
over silt-covered surfaces, prodding down into lower
silty laminae. Current sole marks were formed before
the trackways were emplaced, and the tracks are
Fig. 9. Replacements of Planolites burrows preserved in a
confined to one level within the graded laminated
siderite vein in the black shale facies. A 4-cm-long live units, indicating that little sediment was deposited
limpet for scale (SS 1983 0313). during the time in which the track was being made
112 R. V. Burne

Fig. 11. Photomicrograph of graded muddy siltstone

lamination in the muddy siltstone facies. Carbonaceous
material appears dark (SS 1 995 0400).

Fine-grained graded beds range from 1 to 12 em in

thickness. Beds have sharp bases, in places bearing
Fig. 10. Muddy siltstone facies showing alternations of sole marks, and comprise laterally continuous,
laminated muddy siltstone and ripple cross-laminated normally graded layers of siltstone and muddy silt­
units. Note irregular scouring and deposition of cross­
laminated beds at the level of the hammer (circled). stone (Fig. 13). The sediment is poorly sorted with
Section youngs to the right (SS 201 077). median grain size 3.5 <P and contains up to 58% clay
minerals and other matrix (Fig. 7).
Laminated beds are sharp-based, fine-grained
sandstones, ranging in thickness from 3 mm to 13 em.
(Goldring & Seilacher, 1971). Other tracks initially Bounce and prod moulds are the most common sole
regarded by King (1965) as xiphosurid mating-traces marks, but groove, current-ridge and, rarely, flute
(Fig. 12), were reinterpreted by Higgs (1988) as moulds also occur. The beds are composed of uni­
Undichna or trails left by fish dragging their fins form plane laminations, and/or current ripple cross­
along the soft bottom during feeding. laminated fine-grained sandstones (Figs 13-15). The
cross-lamination occurs either as single sets or as
climbing ripples, in some cases without stoss-side
Interbedded sandstone-shale facies
erosion (Walker, 1963)". Ripple crests are either
This facies comprises interbedded 'event' beds straight or cuspate (Figs 16 & 17). The cross­
(Einsele & Seilacher, 1982) and black shales. The laminated sandstones pass up into alternating dark
relative abundance of various types of event beds in and light coloured, parallel-laminated silts. Beds are
this facies shows no consistency of grouping. At generally laterally extensive, but can show variations
some horizons beds of one of the following types in thickness due to both scouring and localized
dominate the succession. In other cases a predictable concentration of sedimentation. They are relatively
succession of event beds occurs, whereas elsewhere well sorted (excluding matrix aG 0.3-0.5) with
the facies consists of a less predictable succession of median grain sizes varying between 3.5 and 4.5 <P
beds. The following types of event beds can be and matrix contents of 15% (Fig. 7).
distinguished. Graded silty sandstones form sharp-based,
 Bude Formation subsea fans 113

parallel-sided beds 5-25 cm thick. They are gen­

erally laterally continuous, but may be scoured or
may themselves expand into pre-existing scours.
The sole marks are predominantly linear current­
ridge moulds, flute or groove moulds. The lower
part of each bed is structureless and grading is
obvious only in the topmost part, where aggregates
of plant material and fragments of black mudstone
may be concentrated (Fig. 18). Median grain size at
the base of the unit is around 2.5 ¢, decreasing
to about 3.0¢ at the top. The long axes of grains
(Fig. 19) in the structureless unit show slight pre­
ferred orientation in the same direction as that of
current flow indicated by sole marks on the base of
the bed, and, although the bed as a whole is poorly
sorted, the grain-sorting at any one level is moderate
to good (excluding matrix oG 0.6-0.75), although
matrix content may be up to 50% (Fig. 7). The
structureless silty sandstone unit is overlain by a
5-20-mm-thick unit of uniform parallel-laminated
silty sandstone of slightly finer grain size and then by
a similar thickness of current ripple cross-laminated
silty sandstone with either single sets or ripple drift.
Water-escape structures or convolution are generally
not found in these beds. The cross-laminated unit is
overlain by a unit of alternating light and dark
parallel-laminated silts, which is, in turn overlain by
Fig. 12. Traces in the muddy siltstone facies originally structureless grey silt.
interpreted by King (1965) as traces of xiphosurid mating
Structureless sandstones are sharp-based and
behaviour, but reinterpreted by Higgs (1 988) as trails left
by the fins of bottom-feeding fish (Undichna). Structures between 10 cm and 1 m thick (Figs 20 & 21). The
have a wave length of 9 em (SS 2017 0751). most common sole marks are flute moulds and linear

Fig. 13. Grey muddy siltstone

laminations (base of section)
overlain by fine-grained graded
beds, then passing up into laminated
beds showing 'type 3' ripple drift of
Walker (1 963) (SS 201 074).
114 R. V. Burne

Fig. 14. Sharp based, laminated

bed showing irregular ripple-drift
with stoss side erosion and
convolutions due to water-escape.
Note graded top of the 12-cm-thick
bed (SS 1999 0493) .

Fig. 15. View to the right of Fig. 14

showing a sharp erosive top to the
same bed, and an absence of
convolutions (SS 1999 0493) .

current-ridge moulds. Most of the bed is ungraded sharp tops or may grade up, through an abrupt and
and most beds are completely structureless, although irregular contact, into either irregular lamination
some contain faint traces of lamination, often dis­ comparable to hummocky cross-stratification,
rupted into dish structures (Stauffer, 1967). They current ripple cross-lamination (Fig. 22) or, very
are composed of moderately well-sorted fine-grained rarely, dune-shaped cross-sets. The beds are gen­
sand (excluding matrix aG 0.85-0.9; median grain erally laterally continuous, but are prone to amal­
size 2-2 4 <jl; matrix content 14-25%) (Fig. 7),
. gamation. They also can be scoured or can thicken
with grains orientated either parallel to the current in channels. They may also be confined to channels.
direction indicated by sole marks, or bimodally 'Slurried' beds (Wood & Smith, 1959) consist of
about this direction (Fig. 19). The beds may have sharp-based units of a poorly sorted slurry of sand,
 Bude Formation subsea fans 115

Fig. 16. Sandstone surface bearing straight-crested ripples Fig. 17. Sandstone surface bearing cuspate-crested ripples
(SS 2014 0779). (SS 2014 0774).

silt and mud (excluding matrix oa 0.9; including

matrix oa 2.5; median grain size 3.5 ¢, matrix
content 56-70%) (Fig. 7). They are generally
between 2 and 20 em thick (Figs 23 & 24). The sharp
base of the bed may bear sole marks, most commonly
groove or prod and bounce moulds. The lower parts
of the beds show grain orientation in the same sense
indicated by sole marks (Fig. 19). The beds may be
ungraded or graded only in the topmost part, but
they commonly show a two-part division, with a
basal massive, indistinctly laminated or, rarely,
cross-laminat.ed silty sandstone, and an upper, more
poorly sorted 'slurried' unit often containing large
contorted fragments of mudstone. In places beds
have 'frozen' in the act of peeling up fragments of
mudstone from the substratum (Fig. 25). Sandstone Fig. 18. Interbedded sandstone-shale facies showing
graded silty sandstones (A) separated by a 'slurried' bed
or siltstone pseudo-nodules (Fig. 18), interpreted as (B). Note load balls falling from laminated unit on top of
load balls (Kuenen, 1965), and water-escape tubes the 'slurried' bed, and laminated top of the underlying
may be found in the upper part of the 'slurried' unit. graded silty sandstone (SS 2018 0740).
116 R. V. Burne

A Irregularly Laminated Sandstones C Graded Silty Sandstones

N 41 0 N 31 5 N 30 2
Spec. 48 Spec. 14 Spec. 49 Spec. 2A

B Structureless Sandstones

N 31 8
Spec. 2C
N 295
Spec. 2B

N 3 83 N 40 6
60

1
Spec. 73 Spec. D "Slumped" Bed E "Slurried" Bed

I
N 27 5
N 2 87
Spec. 4
1
N

I
Spec. 30 N 274
Spec. 52 b
N 3 02
Spec. 31

10 15 20
%

Fig. 19. The orientations of long axes of elongate grains

within representative specimens of various types of
N 2 69 sandstone bed (data from Burne, 1969b). Measurements
Spec. 58
were made from orientated thin-sections sawn parallel to
stratification. Palaeocurrent directions from sole marks or
N = Number of grains
cross-laminations associated with the beds are indicated by
Spec. = Specimen number 10 15 20
(BURNE 1969b) % an arrow on the rose diagram.

This unit may be overlain by a thin unit of laminated true slump deposits. They are of variable extent.
sediment consisting of either current ripple cross­ Some appear to have a sheet-like form correlatable
lamination or the cross-laminated deposits of small between sections (King, 1967), whereas others are
sand-volcanoes. known from only one section (Edmonds eta!. , 1979;
'Slumped' beds are 4-20-m-thick beds of disrupted Freshney et a!., 1979) and may pass both laterally
sediment (Figs 26-29), similar in character to the and downward into undeformed material. The beds
thinner 'slurried' beds (Figs 7 & 19). The name is are composed of sediment with three separate
well established (e.g. Freshney et al., 1979), but origins; exotic muddy sandstone (excluding matrix
is misleading because the beds are probably not o0 1.2; including matrix o0 2.5; median grain size
 Bude Formation subsea fans 117

Fig. 20. Structureless sandstones within the interbedded sandstone- shale facies showing traces of irregular lamination,
possibly due to the amalgamation of individual units. Length of extended tape is 135 em (SS 1998 061 1 ) .

3.5 <P; matrix content 56-70%) (Fig. 7); sandstone than load moulds are not common, although moulds
and mudstone fragments (Figs 26 & 28); and sand­ of transverse mud ripples have been observed. They
stone balls (Fig. 29). The surfaces of some beds bear may be laterally continuous, or they may wedge
sand volcanoes (Fig. 27). out laterally. The beds are characterized by various
irregular and often interfering non-parallel lami­
nations, although the internal structures are some­
Sandstone-dominated facies
times difficult to distinguish due to lack of grain size
Thick sandstones with few visible structures are contrast. In places, uniform cross-stratification in
diagnostic of the Bude Formation (Fig. 30). Some scalloped based sets up to 20 em deep and 30 em
sandstones are truly structureless, but in others across occur throughout the bed (Fig. 31). The beds
structure seems to have been obscured by the lack of are composed of well-sorted medium sand, and
contrast in grain-size in the beds. Irregular bedding, represent the coarsest sands observed in the Bude
cross-bedding, parallel lamination, convolute lami­ Formation. The grain fabric shows a preferred
nation and parting lineation can all be discerned in orientation at right angles to the dip of the lamination
places. Some sandstones consist of an amalgamated (Fig. 19).
series of laterally continuous beds of structureless Small 1-3-m-deep scours occur near the top of
sandstones similar to those described above. some sandstones (Fig. 32). These scours are often
Also found in this facies are irregularly laminated filled by overlying beds draping into them. The fill is
sandstones, which are relatively coarse-grained similar to the sediment into which the scour is cut.
(median grain size of 2-2.9 <P), moderately sorted Deeper (> 4 m), fiat-bottomed channels also occur
(excluding matrix oG 0.6-0. 65; matrix content of in this facies. The fill is confined by the channel
12-15%) (Fig. 7) and generally form beds 1-4 m margins and comprises only a few units of either
thick. They are sharp based, but sole marks other massive sandstone or irregularly laminated sand-
118 R. V. Burne

Fig. 21. Structureless sandstones showing irregular

structures possibly due to amalgamation. Length of
extended tape is 70 cm (SS 2014 0775).

Fig. 22. Structureless sandstone with upper part reworked

into irregular laminations (SS 199 038) .

Fig. 23. 'Slurried' bed showing three part division of graded base, centre with homogenized shale fragments, and thin
laminated top with load deformation (SS 201 1 0764) .
 Bude Formation subsea fans 119

Fig. 24. 'Slurried' bed with contorted fresh shale fragments in upper part of bed (SS 2014 0775 ) .

Fig. 25. Lower part o f 'slurried' bed showing incorporation o f underlying sediments into the bed as How 'froze'
(SS 201 077) .

stone. The basal unit of the channel fill can contain An example of a channel complex occurs in the
large mudstone clasts, suggesting that the processes massive sandstone facies between King's (1967)
that deposited these units were directly connected key shales P and Q, in repeated sections between
with the erosion of the channel. In places there Northcott Mouth (National Grid Reference SS 202
is evidence that the deposition of sediment was 085) and Crooklets Beach (National Grid Reference
originally localized at one side of the channel, SS 202 085) (Figs 33 & 34). Mapeo & Andrews
possibly as some form of point bar. Successive (1991) and Tanner (1992) have provided structural
channel cuts can be observed within one complex. interpretations for these features, however, the
120 R. V. Burne

Fig. 26. 'Slumped' bed about 80 m

above the Sandy Mouth Shale with
sand volcano (arrowed) on surface.
Hammer for scale (SS 2015 1026).

Fig. 27. Sand volcano shown

arrowed in Fig. 26. Note central
crater and Hank lineations
(SS 2015 1026).

evidence for sedimentologically distinct channel fills the black shale facies indicates deposition under
and channel erosion indicate that the discordances tranquil conditions in a basin where sediment was
are true channel cuts rather than being the result of introduced by discrete and gentle currents, at least
early thrusting, although these bed boundaries would some of which were bottom-hugging. The rare
have provided planes for preferential movement occurrences of pelagic goniatites and bivalves can be
during deformation. correlated with marine bands in other Westphalian
Basins and probably represent brief connections
with open marine conditions during eustatic high ..
Facies interpretation
stands. There is no other indication of marine
conditions in the succession. The preservation of
Black shale facies
nektonic fish remains along with their faecal material
The preservation of well-laminated mudstones of in well-laminated strata lacking remains of macro··
 Bude Formation subsea fans 121

Fig. 28. The Black Rock 'slumped' bed exposed north of Upton Cross. Note load balls foundering from the top of the bed
and 1 . 8 m long sandstone clast (arrowed) (SS 200 053).

Fig. 29. Sandstone clast shown

arrowed in Fig. 28. Note that the
deformation is reminiscent of that
found in load balls, and when the
sandstone clast is restored to its
former geometry an original
lenticular cross-section is revealed,
which , together with the irregular
internal laminations, suggest that
the feature may originally have
formed as a sand volcano on the top
of the 'slumped' bed (SS 2000 0529).

benthos indicates an anaerobic environment (Savrda brought in by the fine-grained turbidity currents in
& Bottjer, 1991), although the restricted ichnofacies the form of finely comminuted terrestrial plant
dominated by Planolites, suggests dysaerobic con­ matter. Evidence for anaerobic and dysaerobic
ditions (Sageman et a!. , 1991). The oxygen depletion conditions is supported by the CIS ratios for the
may have been caused by the increased oxygen shales (Berner & Raiswell, 1983, 1984) reported
demand of 'surplus' organic matter (Wetzel, 1991) by Higgs (1991). The bed-top burrows reflect
122 R. V. Burne

Fig. 3 1 . Vertical section through climbing cross-stratified

sets in irregularly laminated sandstones. Arrow shows
younging direction (SS 2003 0818).

Muddy siltstones facies

The regular lamination of the muddy siltstone facies,
absence of body fossils, presence of bottom-feeding
traces, and high CIS ratios indicate that this facies
was probably formed below storm wave base in an
oxygenated and possibly freshwater environment.
Higgs ( 1991) concluded that xiphosurid tracks are
Fig. 30. Thick, irregularly bedded sandstones of the known only from rocks that appear, on other evi­
sandstone-dominated facies. Figure for scale, section dence, to have been deposited in waters no deeper
youngs to the right (SS 281 080) . than a few tens of metres. However, Goldring &
Sielacher ( 1971) point out that arthropods inhabit
opportunistic colonization. They do not extend practically all marine environments, and that deeper
above the top of the sand bed, indicating that the parts of sedimentary basins where silt and clay
animals responsible died when mud-deposition laminae are deposited from low-velocity currents or
resumed (Higgs , 1991) , a characteristic of turbidite distal turbidity flows provide the highest fossilization
successions (Seilacher, 1982) . The laminations either potential for xiphosurid trackways. It is concluded
result from the settling of dispersed suspended that this facies was deposited by a series of relatively
material that was received in discrete pulses, possibly gentle but persistent bottom-hugging currents
due to seasonal flooding of distant rivers (Burne, (Burne, 1969b; Goldring & Seilacher, 197 1; Melvin,
1969b; Melvin, 1976; Higgs, 199 1) , or as high­ 1986) , the deposits of which Melvin (1986) has
concentration sediment clouds (fine-grained tur­ compared to T0 and T023 fine-grained turbidites o:f
bidity currents) (Burne, 1969b; Melvin, 1976; Glenn Stow & Shanmugam (1980).
& Kelts, 1991). Melvin ( 1986) has pointed out that
these laminations could be described as T3_7 and
Interbedded sandstone-shale facies
T6_7 mud turbidites using the terminology of Stow
& Shanmugam (1980). All the bed types encountered in the interbedded
 Bude Formation subsea fans 123

Fig. 33. 'Thickening upward' succession of black shale

Fig. 32. Small scour in sandstone-dominated facies facies, interbedded sandstone-shale facies, and sandstone­
(outlined) with fill of structureless sandstone beds dominated facies, the latter forming a channel fill.
(SS 2014 0771 ) . Structureless sandstone beds have accreted against the
channel margin (accretion surfaces outlined) . Cliffs are
60 m high, section youngs to the right (SS 2014 0630) .

sandstone- shale facies originated from short-lived

bottom-hugging density currents with a higher bidites in which the structureless graded beds result
content of suspended matter than the surrounding when rapid deposition of sediment outpaces the
water, i.e. turbidity currents in the sense of Sundborg ability of a decelerating current to make stable bed­
(1956, p. 60). This conclusion is supported by the forms (Burne, 1969b; cf. Melvin, 1986). The result is
high matrix content of all the beds (Fig. 7). The beds a graded unit in which grains do not have an oppor­
are all composed of fine-grained sediment (virtually tunity to be preferentially orientated and are depo­
no sediment particles larger than 1 ¢) that could sited in a reasonably stable packing position.
be carried in turbulent suspension by turbidity Laminated beds are interpreted as laminated tur­
currents of reasonable velocity (Lowe, 1982) and bidites, equivalent to the Bouma (1962) Toed, Ted or
would favour the development of autosuspension Tct types. The rate of deceleration of the depositing
(Bagnold, 1962 ) . current was low enough for the deposited sediment
The following detailed interpretations may be to construct stable bedforms beneath the waning
made for each of the bed types in this facies. flow (Burne, 1969b; Melvin, 1986).
Fine-grained graded beds are interpreted as tur- Graded silty sandstones are interpreted as
 A >-'

ss 20000817 �

::>;:)
:<::::
B O:l
!::
ss 20020774 3
D
Structureless sandstones "'
SS 201 60n2

[:. :: : ·.]
.. Graded silty-sandstones

� �
�==j
Cross-bedded sandstones

Grey muddy-siltstone & laminated beds

r:t//1 'Slurried' beds & fine-grained graded beds

• Black shales

1 m[ � Duplex

Sm
Fig_ 34_ Cross-sections of a channel complex in sandstone-dominated facies between key shale P and key shale Q (Tom's Cove Shale) (King, 1967). The complex
consists of an earlier channel filled with beds of structureless sandstone, and a later channel filled with cross-stratified irregularly laminated sandstone. (A) Exposure
in section 39 of King ( 1967), North of Earthquake Cliff. Note shale clasts at base of channel fill, apparent lateral accretion of structureless sandstones, and final cross­
bedded channel-fill. (B) The same complex 500 m south in section 36 of King ( 1 967) .
 Bude Formation subsea fans 125

turbidites showing the almost complete Bouma se­ continue infinitely but for the fact that the gradual
quence Tabcde (Burne, 1969b; Melvin, 1986) , although loss of water from the current reduces the current's
the 'd' division is rarely well developed (Bouma, volume to a point where the concentration of the
1962). In the rapid deposition of the 'a' division, suspended sediment suppresses fluid turbulence,
grains do not have an opportunity to be preferentially and the current would essentially evolve into a
orientated, although they are deposited in a reason­ cohesive debris flow, or mudflow, without passing
ably stable packing position. through an intervening tractional stage (Postma,
Deposition of the massive units of the structureless 1986; Postma et al. , 1988). This dense current flowing
sandstones may have taken place from the bedload over a soft muddy bottom would tend to peel up the
of a current flowing during the strongly fluctuating substrate and incorporate it into the flow in the
hydrodynamic conditions and vigorous burst-sweep manner described by Dzulynski & Radomski (1966) .
cycles of large turbidity currents (Hiscott, 1994). The mud would be contorted by the flow, and would
Deposition would be either as successive sheets tend to disaggregate. However, it would also increase
of sediment (under conditions of plane bed with the concentration of transported solids in the flow,
movement) or as more irregular lenses of sediment thereby encouraging, 'freezing' of the flow. Postma
(under conditions of antidune flow). Subsequently et at. (1988) concluded that a rapid transformation
the escape of pore-water, trapped in the sediment of flow behaviour is necessary for 'soft' intraclasts to
during deposition, and the compaction of the unit be preserved. When the critical concentration of
may have destroyed this primary structure. Alterna­ solids is reached the flow would stop abruptly and
tively the structureless parts of these beds may have leave a quick bed. The coarser sediment. of the unit
developed if the tangential stress applied by the would not be disturbed by this and would preserve
depositing current was high enough to maintain the any original preferred orientation (Fig. 19). Any
dispersive stress of a bedload, the volume of which sediment deposited on the surface of the unit at this
was being greatly increased by sediment falling out stage would tend to founder into it because of the
of suspension. This process of transport causes pro­ lack of strength of the bed (Figs 18 & 29) . Excess
nounced orientation in the grain fabric and is the water would be expelled during the compaction of
equivalent of the traction carpet of Dzulynski & the bed, forming water-escape structures, pipes of
Sanders ( 1962), but is different from the process of matrix-free sand, and constructing sand volcanoes
grain flow (Sanders, 1965 ; Stauffer, 1967, p. 502) in on the surface of the bed (Fig. 27) . This sequence
which the tangential stress is applied by gravity to of post depositional events has been reproduced
sediment resting on a slope. Traction carpets are experimentally by Nichols et al. (1994). The surface
essentially current-impelled grain flows (Middleton of the bed may continue to be sheared by a residual
& Hampton, 1976) or cohesionless debris flows current, with the result that a thin unit of current
(Postma, 1986). As the current waned, the tangential ripple cross-lamination forms on the upper surface
stress applied to the bedload would decrease to a of the bed. Sand volcanoes on the surface of 'slurried'
point where the dispersive stress in the traction beds may show an asymmetry in the same sense as
carpet could no longer be maintained. At this point that indicated on the sole mark of th.e bed as a result
the traction carpet would abruptly cease to move. of eruption into a current (Burne, 1970) .
The reworked nature of the tops of these beds
results from the tail of the turbidity current con­
Sandstone-dominated facies
tinuing to shear the surface of the sediment deposited
by the body of the current, and reflects the probable The comparatively coarse, moderately sorted
irregularities of that surface. , nature of the beds of the sandstone-dominated facies
The grain size distributions of the 'slurried' beds suggests that they have been deposited from the
and 'slumped' beds are very similar to that of modern bedload of a transporting current. The structureless
mudflows (Baker, 1965) except that there are no sandstones of this facies have a similar origin to
very coarse grains. All the features of the units those of the interbedded sandstone shale facies, but
considered here can be explained by deposition from are commonly amalgamated to form thick sand­
turbidity currents carrying a very high concentration stones. The irregularly laminated sandstones are of
of fine-grained sediment in suspension. Bagnold two types. In the majority of beds the irregular
( 1968) has demonstrated that such currents expend lamination cannot be related to any bedform in
little energy and sediment transport might therefore the lower flow regime, but strongly resemble those
126 R. V. Burne

produced by Middleton (1965, Fig. 3) in an exper­ much of the Bude Formation does not conform to
imental study of deposition from a current forming well-developed thickening-up or thinning-up suc­
antidunes. ln other sandstones, better organized cessions (Melvin, 1986). The organization of these
cross-stratification can be discerned, which can be parts of the Bude Formation is less predictable,
related to migrating dunes. In both these sandstone although it is not random (Fig. 4).
types the structures are persistent throughout the
bed and there is little evidence of grading, or waning
Interpretation
flow conditions. Burne (1969b) considered that they
were deposited by single events characterized by
The case against a storm-influenced shelf
discrete, short-lived, but relatively powerful currents
of almost constant velocity. Higgs (1991) concluded that The Bude Formation
was deposited largely, if not entirely, above storm
wave base, as indicated by the fact that wave­
FACIES SUCCESSION influenced structures occur throughout the suc­
cession, spaced no more than a few metres apart.'
Melvin (1986) thought that the presence of Planolites,
Description
xiphosurid trackways and possible hummocky cross­
The long-term evolution of a sedimentary succession stratification in the succession meant that the
at any given place involves temporal transitions Bude Formation had formed as a subsea fan 'within
between environments, each of which is character­ relatively shallow (shelf as opposed to abyssal) water
ized by different depositional systems producing depths' (p. 26). Higgs (1987) queried how a fan
distinct facies. The result is a facies succession. could form at such shallow depths 'where the normal
Reading ( 1971) noted the lack of predictability of configuration is a wave-cut shelf ' (p. 378). Melvin
Bude Formation facies successions when compared (1987, p. 382) replied that it is 'hard to imagine how
with other cyclic Silesian successions, such as the 1300 m could accumulate by the passive aggradation
synchronous paralic or deltaic facies succession of of storm deposits on a lacustrine shelf, wherein
the Bideford Group (de Raaf et al. , 1965). The facies changes are determined by subtle salinity
facies succession of the Bude Formation (Figs 35 & variations associated with a yo-yo effect of eustatic
36) shows alternation between end-member con­ sea level changes'. This refers to Higgs's (1986b,
ditions of tranquil mudstone sedimentation and the 1991) interpretation that each alternation of 1-10-
rapid emplacement of channelled massive sandstones m-thick sandstone with shales of similar thickness
with a complete gradation of intermediate environ­ within the Bude Formation resulted from fluctuations
ments. Burne ( 1969b) identified a composite suc­ in depth and salinity brought about by intermittent
cession in the Bude Formation consisting of a over-topping of the sill of a silled basin. According
succession of black shale facies, a thickening upward to Higgs (1991), the sandstones were deposited from
succession of interbedded sandstone-shale facies, a river-generated storm flows, diverted and modified
channelled sandstone dominated facies, and then by wave activity. The river-fed turbidity current and
a thinning upward succession of interbedded sand­ the wave-induced oscillatory current never varied
stone-shale facies and silty mudstone facies that to the point where the wave processes dominated
passes abruptly up into another black shale facies distally, as would be expected in the case of a
(Fig. 5). This succession was derived largely from localized river underflow operating through a
relationships observed in the section between the tempest (Brenchley, 1985). Higgs (1991) interpreted
key shale 0 and the top of the Tom's Cove Shale the ripple cross-lamination developed in the upper
Member of King (1967) (Figs 35 & 36). However, parts of these sandstone beds as 'quasi-symmetrical

Fig. 35. (Opposite.) Lateral variation in facies successions between key shale 0 and the top of key shale Q (Torn's Cove
Shale). Key shales and section numbers of King ( 1967). The individual beds of the interbedded sandstone-shale facies are
shown in a generalized way to demonstrate the variability of this facie. Note the tendency for thickening upward sequences
to follow the black shale facies, and for the muddy siltstone facies to occur beneath the black shale facies. Note the position
of the channel complex of Fig. 34 in sections 39 and 36. (Locations: Earthquake Cliff, SS 201 079; Upton Cross, SS 200 049;
Lower Longbeak, SS 198 032; Widemouth Sand, SS 199 023).
 Bude Formation subsea fans 127

Section 5
(Widemouth Sand)

D..
ly laminated or
Irregular
structure less sandstones Section 9
(South of Lower
Long beak)
B
"Slurried beds" or fine
grained graded beds

§ Cross laminated beds

0
�m�
Graded silty sandstones

� Muddy siltstones I
t:::::=::j
Section 1 8 1
D Black shales (Upton Cross) 1

Section 36 I
(South of earthquake) I
· I
>: : · . ·:. : - : ·
.
I
-: : : · : - ·
I

[ TOM'S
(Key
COVE
shale
SHALE
Q)

Section 39
(North of earthquake)

.. .· ·. · · . : .

· · ::'
.. ·
. . .

shale P)

0)
128 R. V. Burne

Fig. 36. A 60-m-high cliff exposure of the facies succession of section 36 (SS 201 077) as logged in Fig. 35 . Section youngs to
the right. Arrows mark the three 'thickening upward' successions from key shale 0 (left), key shale P (centre) and Tom's
Cove Shale or key shale Q (right) . Highlighted in the foreground is the channel-cut shown in the upper part of Fig. 34B.
This is filled with cross-stratified, irregularly laminated sandstones comparable to those shown in Fig. 31 from the laterally
equivalent channel fill shown in the upper part of Fig. 34A.

ripples' intermediate between current ripples and body of the same turbidity current (Middleton &
wave ripples. However, the statistics quoted by Hampton, 1976). Larue & Provine (1988) illustrate
Higgs ( 1991) in support of a wave-generated origin similar irregular cross-laminated sets from turbidites
for these ripples are not definitive (Reineck & on Barbados, which they attribute to deposition as
Wunderlich, 1968) and they could equally represent vacillatory turbidites emplaced by multiple surging
non-equilibrium bedforms developed under waning flows.
flow conditions (Baas, 1994). Unequivocal oscil­ Higgs ( 1991) cited the occurrence of multi­
lation ripples have not been reported from the Bude directional tool marks and mud-draped scours
Formation (Goldring & Seilacher, 1971), neither as further evidence of wave action in the Bude
have 'clean' sandstones characteristic of wave sorting Formation. Diametrically opposed sole marks on
(Fig. 7; Seilacher, 1982). The 'quasi-symmetrical the same surface are a characteristic of some tem­
ripples', described by Higgs (1983, 1984, 1987, 1991) pestites (Seilacher, 1982), but do not occur in the
as storm-wave influenced structures are found in Bude Formation. The multidirectional sole marks
beds that are quite different from the wave-generated that do occur are similar to those that Middleton &
lithotypes described by de Raaf et al. (1977), and I Hampton ( 1976) have attributed to divergence of
consider that these structures have no depth-related flow directions within a strongly lobate turbidity
connotation and are more logically interpreted as current head. Mud-draped scours are not definitive
the product of reworking by the tail of a turbidity of wave action because Normark & Piper ( 1991)
current of sediment previously deposited by the have recorded them from submarine fan-channels.
 Bude Formation subsea fans 129

Rare occurrences of hummocky cross-stratification The lack of predictable vertical associations in the
in the Bude Formation were cited by Higgs (1991) as Bude Formation is of itself evidence against a shallow
evidence of storm-wave activity reworking bottom water or paralic origin for the succession, because a
sediments. Einsele & Seilacher (1991) claim that 'large proportion of turbidite successions cannot be
hummocky cross stratification is not found in tur­ assigned to any form of regular cyclicity' (Piper &
bidites, but Prave & Duke ( 1990) have concluded Stow, 1991, p. 371). Hiscott (1981) suggested that
that small-scale hummocky cross-stratification does cyclical turbidite successions, identified solely on
not indicate a particular flow condition or depo­ visual criteria, could be explained by chance occur­
sitional environment. They interpret examples rences within unordered successions of turbidite
similar to those described by Higgs (1983, 1984, beds. However, Burne ( 1969b) recognized thicken­
1991) to have been formed by antidunes generated ing upward and thinning upward successions, which
by standing waves at the interface between the he related to subsea fan environments. Similar
depositing body and tail of a turbidity current and an successions have since been described from both
overlying low-density layer. Similar structures in ancient and modern turbidite fan environments. For
turbidites also have been interpreted as antidune example, Mutti & Ricci-Lucchi ( 1972) and Ricci­
structures by Walker ( 1967), Skipper (1971), Hand Lucchi ( 1975) described turbidite successions from
et al. ( 1972) and Yagishita ( 1994) . Alternatively, the northern Appenines of Italy that contain sub­
Larue & Provine (1988) have interpreted hummocky marine fan successions or 'second-order cycles'.
cross-stratification found in turbidites as being due These either followed one another directly or were
to deposition of laminated sand on a surface modified separated by 2-100 m of monotonous succession
by fluidization and current scour - a situation interpreted as the deposits of the open fan or basin
comparable to that inferred for some of the Bude plain. They contained thickening upward cycles that
Formation examples. Mutti (1977, fig. 16) figures represented prograding lobes (60% of these cycles
thin parallel laminae in a basin-plain-facies turbidite were non-channelled), and thinning upward cycles
that closely resemble the structures interpreted that represented a channel-fill succession (80% of
by Higgs ( 1983, 1984, 1991) as hummocky cross­ these were channelled).
stratification. In the absence of associated unequi­ It is concluded that the Bude Formation can also
vocal evidence for shelf sedimentation, such as that be interpreted as the deposits of a turbidite system
cited by Monaco (1992), the rare hummocky cross­ or subsea fan (defined without reference to shape).
stratification recorded from the Bude Formation is Thus the sandstone- dominated facies may represent
interpreted to represent the result of either upper the deposits of an accretionary fan-valley system
flow-regime structures or traction across a quick bed (O'Connell et al. , 1991). The fan-valley is on a scale
during deposition from a turbidity current. much greater than that of the outcrop, but within
Brenchley et at. (1993) have presented an in­ this overall valley 'thalweg' channels occur with a
terpretation for the Lower Ordovician Bell Island scale observable in outcrop. The facies has the
Group of Newfoundland that is remarkably similar complex organization typical of either the fan­
to that presented by Higgs (1991) for the Bude channel facies (Mutti, 1977), characterized both by
Formation, but the facies succession is quite differ­ thick-bedded channelized units and thinner bedded
ent. In contrast to the Bude Formation, the Bell units, or of the channel-mouth facies of Mutti (1977),
Island Group shows more predictable facies suc­ in which beds show lensing and wedging, and massive
cessions, contains no sole marks, but wave ripples, units are overlain abruptly by dune or ripple cross­
hummocky cross-stratification, herring-bone cross­ stratification.
stratification, gutter casts and bioturbation are all In the silty mudstone facies the evidence of
common. frequent though gentle turbidity currents suggests
deposition either in an elevated position within a
fan-valley complex or on a channel levee. This
The case for turbidite systems
interpretation is supported by the evidence of
The substantial thickness of the Bude Formation, apparent flow instability (Piper & Stow, 1991).
the relative continuity of single beds, and the absence Normark & Piper (1991) described levee deposits as
of a shallow-water benthic fauna all support a containing either discontinuous sands with wavy
turbidite-basin interpretation rather than that of a bedding and climbing ripples or abundant thin sand­
storm-affected shelf (Einsele & Seilacher, 1991). silt laminae. Further support for this interpretation
no R . V. Burne

is provided by Mutti ( 1977), who found that thin, known to be possible in turbidite systems (Skipper
rippled sandstone units that diverged or expanded in & Middleton, 1975 ; Middleton & Hampton, 1976;
thickness were characteristic of a channel-margin Postma, 1986; Postma et al., 1988; Normark & Piper,
facies. 1991; Ghibaudo, 1992) .
There are three possible environments in which Reading (1987 , p . 8 ) has reminded us that: 'So
the interbedded sandstone- shale facies could form often a new fashion or model is developed by
(Mutti, 1977; Normark & Piper, 1991) . In channel the subordination of alternative hypotheses or
levee and interchannel environments, turbidite possibilities'. The model developed by Higgs (1991)
deposits will reflect the degree to which sequential for a storm-influenced shelf is based on the equi­
channelled flows overtop the levee. They will show vocal evidence of hummocky cross-stratification
variability in both vertical succession and thick­ and 'quasi-symmetrical' ripples (Higgs, 1983, 1984,
ness. Deposition on a lobe at the end of a leveed 1987, 1988, 1991). It is confirmed neither by facies
fan-valley system will show less vertical or lateral associations nor by stratification characteristics of
variability, but preserve evidence of continuity of associated sediments, and, although Melvin ( 1 987)
aggradation. Deposits on an unchannelized open fan considered that the presence of wave ripples in the
will tend to be composed of laterally continuous Bude Formation is not in dispute, no convincing
beds of rather uniform character, although reflecting illustration of them has been published to date. The
currents of different sizes. The character of the problems regarding the depositional environment
interbedded sandstone- shale facies is generally not of the Bude Formation all disappear when it is
uniform and it is concluded that it represents either realized that there is no other substantial evi­
the deposits of levee and interchannel environments dence for shallow-water environments in the Bude
or of lobes that accumulated at the end of leveed Formation. The ichnofacies have no depth conno­
fan-valleys. tation (Goldring & Seilacher, 197 1 ; Sageman et al.,
The black shale facies is composed of fine-grained 1991; Goldring, 1993). Whalley & Lloyd ( 1986)
rhythmic couplets typical of profundal lake deposits note that the thickness of the Bude Formation is
(Glenn & Kelts, 199 1 ) and interpreted as the deposits exceptional for the depositional environment pro­
of open fan or basin sedimentation. posed by Higgs (1983, 1984), and Higgs ( 199 1 )
himself found it 'curious that such a thick 1300 m
shallow water succession is totally devoid of near­
CONCLUSIONS shore or emergent features'.
AU the beds of the Bude Formation have features
It is concluded that the Bude Formation was depo­ consistent with having been emplaced by turbidity
sited in a land-locked synorogenic foreland basin. currents and there is no evidence that sediment
Deposition occurred on the northern, inactive flowed into the basin by any other mechanism. What
margin of this basin. Despite the generally unpredict­ evidence is there for the form of this turbidite
able nature of much of the succession, periods of system? Although some have argued for a prodelta
black shale sedimentation alternated with periods setting (Melvin, 1986) and others for a base of
of turbidite deposition in which subsea fans were slope ramp (Hartley, 1991) , the deposits are more
constructed. The conclusion of Melvin ( 1986) that compatible with a basinal rather than a slope
parts of these fans did, on occasion, rise above storm environment. The only constraint on the depth of
wave base cannot be absolutely excluded, but it this basin is the requirement that the basin should
seems more likely that the environment was a pro­ have a marginal slope sufficient to ensure either the
fundal lake or inland-sea basin (Glenn & Kelts, maintenance of hyperpycnal flow or the 'ignition'
1991) in some ways comparable with the East African (Parker, 1982) of turbidity currents (Normark &
Lakes (Cohen, 1990; Scholz & Rosendahl, 1990; Piper, 199 1 ). Although the black shalefacies contains
Baltzer, 1991; Scott et al., 1991) or Lake Baikal evidence of changes in basin salinity, marine
(Hutchinson et al. , 1992). incursion, and degree of bottom oxygenation, none
Earlier alternative interpretations of the Bude of these appear to have had any direct influence on
Formation (Owen, 1950; Prentice, 1962; King, 1965, either the supply of sand to the basin or the gen­
1967; Edmonds et al. , 1968; Freshney et al., 1972, eration of turbidity currents. The supply of sediment
1979) seem to be based, in part, on a lack of appreci­ to the basin was probably greatest during periods of
ation of the extensive variation of structures now relatively low basin water level and consequently
 Bude Formation subsea fans 131

very few powerful turbidity currents were initiated the Bude Formation was deposited on the inactive
during periods of high water level. margin of the basin (Burne, 1969b; Haslam &
The compositions of the various turbidites indicate Scrivener, 1991). The sandstones are actually richer
that they all have the same provenance, but there in quartz (Burne, 1969b; Freshney et al. , 1979;
are marked differences in the amount of matrix in Melvin, 1986) than are the sands of a modern passive­
the various beds (Fig. 7). A possible explanation for margin fan, the Mississippi fan (Roberts & Thayer,
these different depositional characteristics from 1985) (Fig. 7). Shanmugam & Moiola (1988) con­
similarly sourced turbidity currents is to view them cluded that interpretations derived for active-margin
in terms of high- and low-density flows. Low-density fans should be applied with caution to passive­
flows would exhibit a proximal to distal progression margin fans because of differences in spatial distri­
from full Bouma sequences Tabcde through to Tde bution of turbidite facies and their associations (cf.
sequences by sedimentation from fully turbulent fig. 8 of Normark & Piper, 1991). Shanmugam &
flows. High-density flows would deposit bedload, Moiola ( 1988) found that characteristics of mature
perhaps from traction carpets, in more proximal passive-margin taos included: low sand-to-mud ratios
situations, whereas distally concentrated auto­ in their sediment supply; hyperpycnal flow is import­
suspensions would flow until grain concentration ant in fan evolution; flows are uniform and of low
increased to a limiting point for turbulence and the velocity; and very large muddy flows may be triggered
flow would be rapidly arrested by transformation by sea-level changes. The result is a fan characterized
into a debris flow. It is possible that the spilling of by steady growth of narrow levees, progradation of
part of the turbidity current over channel levees channel-levee systems, and uniform aggradation of
would suddenly dissipate energy in the channelled lobes - all features consistent with the facies suc­
flow, causing bedload deposition, possibly as a result cession of the Bude Formation.
of traction carpet collapse, whereas the overbank
part of the flow would carry suspended sediment
only and would also dissipate energy until it increased
its concentration to that of a high-density flow, ACKNOWLEDGEMENTS
eventually 'freezing' to form a 'slurried' or 'slumped'
bed. These beds of contorted sediment have con­ Harold Reading introduced me to the challenges of
ventionally been regarded as having originated either the Bude Formation, and supervised my doctoral
as slumps or debris flows (Freshney et al., 1972; research. A.F. King generously made correlations
Melvin, 1986; Hartley, 1991) and their interpretation of the Bude Formation available to me prior to their
as mudflows or cohesive debris flows , which are publication. E.C. Freshney, W. D. Gill, L.R. Moore,
the distal deposits of high-concentration turbidity J.F.M. de Raaf, E.R. Oxburgh, D. Tappin, E. K.
currents, is a novel concept with considerable Walton, G. Whitnall, A. Wood and G. Young are
palaeogeographical significance. Thjck and extensive thanked for discussions of various aspects of the
'slumped' beds interbedded with black shales may study. Norma Burne made invaluable contributions
represent large muddy flows triggered by changes in to this research. The Chadd family of Bude provided
water level in the basin. The thicker, more restricted hospitality, assistance and friendship. R.J. Korsch,
'slumped' beds probably represent the distal deposits J.F. Lindsay, R. Norris and T.S. Loutit commented
of channelled flows that deposited massive sands in on an early draft of this paper. R. G. Walker provided
more proximal environments. These units may be encouragement over a long period, and, together
compared with the chaotic silt beds recorded by with E.C. Freshney and A.G. Plint, vigorously
Nelson et al. (1992) from the outer lobes of the reviewed the manuscript. I gratefully acknowledge
Mississippi fan. the support and assistance of the Australian Geo­
The determination of whether a fan model is logical Survey in many aspects of the preparation
appropriate for the Bude Formation depends on the of this manuscript. Gail Hill of the Cartographic
recognition and distinction of fan-lobe, fan-channel/ Services Unit (AGSO) drafted the diagrams.
mouth-bar and interchannel deposits. Some parallels
can be drawn with the fan successions described
by Mutti & Ricci-Lucchi (1972) and Ricci-Lucchi REFERENCES
(1975), though these were formed on active tectonic
margins. Although the Culm Basin was synorogenic, AsHWIN, D . P . ( 1957) The structure and sedimentation of
132 R. V. Burne

the Culm sediments between Boscastle and Bideford, BURNE, R.V. & MooRE, L.J. ( 1971 ) The Upper Carbon-­
north Devon. PhD thesis, University of London. iferous rocks of Devon and North Cornwall. Proc. Ussher
AsHWtN , D . P . ( 1 958) The coastal outcrop of the Culm Soc. , 2, 288-298.
Measures in south-west England. In: Abstracts of the COHEN, A.S. (1990) Tectono-stratigraphic model for sedi-­
Proceedings of the Conference of Geologists and Ceo­ mentation in Lake Tanganyika, Africa. In: Lacustrine
morphologists in the southwest of England, Exeter, 1 958 Basin Exploration: Case Studies and Modern Analogs
(Eds Simpson, S. & Robson, J . ) , pp. 2 - 3 . Royal Geo­ (Ed. Katz, B . J . ) , Mem. Am . Assoc. petrol. Geol. , Tulsa,
logical Society of Cornwall, Penzance. 50, 137- 150.
BAAS, J . H . ( 1 994) A flume study on the development and CORNFORD, C . , YARNELL, L. & M URCHISON , D.G. ( 1987)
equilibrium morphology of current ripples in very fine Initial vitrinite reflectance results from the Carboniferous
sand. Sedimentology, 41, 185-209. of north Devon and north Cornwall. Proc. Ussher Soc. ,
BAGNOLD, R.A. ( 1962) Auto-suspension of transported 6, 461 -467.
sediment; turbidity currents. Proc. R. Soc. London, Ser. DE RAAF, J.F.M . , BOERSMA, J.R. & VAN GELDER, A .
A , 265, 315-319. ( 1977) Wave-generated structures and sequences from
BAGNOLD, R.A. ( 1 968) Deposition in the process of a shallow marine succession, Lower Carboniferou s ,
hydraulic transport. Sedimentology , 1 0 , 45 -56. County Cork , Ireland. Sedimentology, 24, 45 1 -483.
BAKER, P.E. ( 1965) The geology of the Mt Misery Volcano, DE RAAF, J .F.M . , READING, H. G. & WAL KER , R.G. ( 1 965)
St Kitts. PhD thesis, University of Oxford. Cyclic sedimentation in the lower Westphalian of north
BALTZER, F. ( 1991) Late Pleistocene and Recent detrital Devon, England. Sedimentology , 4, 1 - 52 .
sedimentation in the deep parts of northern Lake DZULYNSKI, S. & RADOM SKI , A. ( 1 966) Experiments on
Tanganyika (East African Rift) . In: Lacustrine Facies bedding disturbances produced by the impact of heavy
Analysis (Eds Anad6n, P . , Cabrera, L l . & Kelts, K . ) , suspensions upon horizontal sedimentary layers. BulL
Spec. Pubis int. Ass. Sediment . , No . 13, p p . 147 - 173 . Acad. Pol. Sc. Ser. Sci. Geol Geogr., 14, 227-230.
Blackwell Scientific Publications, Oxford. DZULYNSKI, S. & SANDERS , J . E . ( 1962) Current marks on
BERNER, R.A. & RA ISWELL , R. ( 1983) Burial of organic firm mud bottoms. Trans. Conn. A cad. Arts Sci., 42,
carbon and pyrite sulfur in sediments over Phanerozoic 57-96.
time: a new theory. Geochim. Cosmochim. Acta, 47, EDMONDS, E . A . , McKEow N , M . C . & WILLIAMS , M. ( 1 975)
855-862. British Regional Geology, South-West England. HMSO,
BERNER, R.A. & RAISWELL, R. ( 1 984) CIS method London.
for distinguishing freshwater from marine sedimentary EDMONDS, E . A . , WRIGHT, J . E . , BEER, K . E . , HAwKES ,
rocks. Geology, 12, 365-368. J . R . , WILLIAMS , M . , FRESHNEY, E . C . , et al. ( 1968)
BESLY, B . M . (1988) Late Carboniferous sedimentation in Geology of the country around Okehampton. Memoir of
northwest Europe: an introduction. In: Sedimentation in the Geological Survey of Great Britain for the One-inch
a Synorogenic Basin Complex: the Upper Carboniferous Geological Sheet 324, New Series. Institute of Geological
of Northwest Europe (Eds Besly, B.M. & Kelling, G . ) , Sciences, HMSO, London, 256 pp.
pp. 1 - 7 . Blackie, Glasgow. EDMONDS, E . A . , WtLLlAMS, B . J . & TAY LOR , R.T. ( 1979)
BouMA, A.H. ( 1 962) Sedimentology of some Flysch Geology of Bideford and Lundy Island. Memoir of
Deposits. A Graphic Approach to Facies Interpretation. the Geological Survey of Great Britain for 1 : 50 000
Elsevier, Amsterdam, 168 pp. Geological Sheet 292. New Series, with sheets 275, 2 76,
BRENCHLEY, P.J. (1985) Storm influenced sandstone beds. 291, and part of sheet 308. Institute of Geological
Modern Geol. , 9, 369-396. Sciences, HMSO, London, 143 pp.
BRENCHLEY, P.J . , PICKERILL, R.K. & STROMBERG , S.G. EtNSELE, G. & SEtLACHER, A. ( 1982) Cyclic and Event
( 1 993) The role of wave reworking on the architecture Stratification. Springer-Verlag, Berlin , 536 pp.
of storm sandstone facies, Bell Island Group (Lower EtNSELE, G. & SEtLACHER, A. ( 1991) Distinction of
Ordovician), eastern Newfoundland. Sedimentology, 40, tempestites and turbidites. In: Cycles and Events in
359-382. Stratigraphy (Eds Einsele, G . , Rieken, W. & Seilacher,
BuRNE, R.V. ( 1969a) Observations on the deposition of A . ) , pp. 377-382. Springer-Verlag, Berlin.
the Bude Sandstones. Proc. Ussher Soc. , 2, 109 - 1 10. ELLIOlT, T. (1976) Upper Carboniferous sedimentary
BuRNE, R.V. ( 1969b) Sedimentological studies of the Bude cycles produced by river-dominated elongate deltas .
Formation. PhD thesis, University of Oxford. J. geol. Soc. London, 132, 199-208.
BURNE, R.V. ( 1 970) The origin and significance of FRANKE, W. & E NGEL , W. ( 1 988) Tectonic settings
sand volcanoes in the Bude Formation (Cornwall). of synorogenic sedimentation in the Variscan Belt of
Sedimentology, 15, 2 1 1 - 228. Europe. In: Sedimentation in a Synorogenic Basin
BuRNE, R.V. ( 1971 ) Sedimentological unit definition , Complex: the Upper Carboniferous of Northwest Europe
hypothesis testing, and multi-step dependence analysis (Ed. Besly, B.M. & Kelling, G . ) , pp. 8-23. Blackie .
for lithological sequences. Program with Abstracts, VIII FRESHNEY, E.C. & TAY LOR , R.T. ( 1972) The Upper
International Association of Sedimentologists, Congress, Carboniferous stratigraphy of north Cornwall and west
August 31-September 3 Heidelberg, p. 16. Devon. Proc. Ussher Soc., 2, 464-47 1 .
BuRNE, R.V. ( 1973) Palaeogeography of south west FRESHNEY, E.C. , McKEOWN , M . C . & W ILLIAMS , M. ( 1 972)
England and Hercynian continental collision. Nature Geology of the Coast between Tintagel and Bude.
(Phys. Sci.) , 241, 129 - 13 1 . Memoirs of the Geological Survey of Great Britain -­
BuRNE, R.V. (1976) Submarine fans and their hydrocarbon Explanation of part of One-inch Geological Sheet 322,
potential. Abstracts, 25th International Geological New Series. Institute of Geological Sciences, HMSO,
Congress, August 1 7-24, Sydney, Vol. 2, pp. 340-34 1 . London, 92 pp.
 Bude Formation subsea fans 133

FRESHNEY, E . C . , EDMONDS, E .A . , TAYLOR, R.T. & turbidites - fact or fiction? 1. sediment. Res. Sect. A:
WILLIAMS, B . J . (1979) Geology of the country around sediment. Petrol. Process. , A64, 204-208.
Bude and Bradworthy. Memoir of the Geological Survey HOLDER, M .T. & LEVERIDGE, B . E . (1986) Correlation of
of Great Britain for I : 50 000 Geological Sheets 307 and the Rhenohercynian Variscides. 1. geol. Soc. London,
308. Institute of Geological Sciences, HMSO, London, 143, 1 4 1 - 147.
62 pp. HUTCHINSON, D . R . , GOLMSHTOK, A . J . , ZONENSHAIN, L.P. ,
GHtBAUDO, G . ( 1 992) Subaqueous sediment gravity flow MooRE, T.C. , ScHoLz, C.A. & KuTGORD, K . D . ( 1 992)
deposits: practical criteria for their field description and Depositional and tectonic framework of the rift basins of
classification. Sedimentology, 39, 423-454. Lake Baikal from multichannel seismic data. Geology,
GLEN N , C.R. & KELTS, K. ( 1 991) Sedimentary rhythms in 20, 589-592.
lake deposits. In: Cycles and Events in Stratigraphy (Eds IsAAc, K . P . , TuRNER, P.J. & STEWART, l.J. ( 1982) The
Einsele , G . , Rieken, W. & Seilacher, A . ) , pp. 188-22 1 . evolution of the Hercynides of central SW England.
Springer-Verlag, Berlin. 1. geol. Soc. London, 139, 521 - 53 1 .
GoLDRING, R. (1993) Ichnofacies and facies interpretation. KlNG, A.F. ( 1 965) Xiphosurid trails from the Upper
Palaios, 8, 403-405. Carboniferous Bude Sandstones, north Cornwall. Proc.
GOLDRING , R. & SEtLACHER, A. ( 197 1) Limulid undertracks Geol. Assoc., 1626, 162- 165 .
and their sedimentological implications. Neues 1ahrb. KlNG, A.F. ( 1 966) Structure and stratigraphy o f the Upper
Geol. Palaeontol. Abh., 137, 422-442. Carboniferous Bude Sandstones, north Cornwall. Proc.
HAND, B . M . , MIDDLETON , G .V. & SKIPPER, K. ( 1972) Ussher Soc., 1 , 229-232.
Antidune cross-stratification in a turbidite sequence, KlNG , A.F. (1967) Structure and stratigraphy of the Upper
Cloridorme Formation, Gaspe, Quebec: Discussion. Carboniferous Bude Formation, north Cornwall. PhD
Sedimentology, 18, 135- 138. thesis, University of Reading.
HARTLEY, A. ( 1 991) Debris flow and slump deposits KlN G , A.F. ( 1971 ) Correlation of the Upper Carboniferous
from the Upper Carboniferous Bude Formation of SW Bude Formation, north Cornwall. Proc. Ussher Soc., 2,
England: implications for Bude Formation facies models. 285-288.
Proc. Ussher Soc. , 7, 424-426. KuENEN, P.H. ( 1 965) Value of experiments in geology.
HARTLEY, A.J. & WARR, L.N. (1990) Upper Carboniferous Geol. Mijnbouw, 44, 22-36.
foreland basin evolution in SW Britain. Proc. Ussher LARUE, D.K. & PROVINE, K.G. ( 1 988) Vacillatory tur­
Soc., 7, 2 12-216. bidites, Barbados. Sedimentary Geol., 57, 21 1 -219.
HASLAM, H.W. & SCHRIVENER, R.C. ( 1991) A geochemical LowE, D . R. ( 1982) Sediment gravity flows II: depositional
study of turbidites in the Bude and Crackington forma­ models with special reference to the deposits of high­
tions north of Exeter. Proc. Ussher Soc. , 7, 421 -423. density turbidity currents. 1. sediment. Petrol. , 52,
HECHT, C.A. (1992) The Variscan evolution of the Culm 279-298.
Basin, south-west England. Proc. Ussher Soc. , 8, MAPEO, R . B . M . & ANDREWS, J.R. ( 1991) Pre-folding
33-38. tectonic contraction and extension of the Bude Forma­
HtGGS, R . (1983) The possible influence of storms in the tion, North Cornwall. Proc. Ussher Soc. , 7, 350-355.
deposition of the Bude Formation (Westphalian), north MELVIN, J . ( 1 976) Sedimentological studies in the Upper
Cornwall and north Devon. Proc. Ussher Soc. , 5 , Palaeozoic Sandstones near Bude Cornwall and Walls,
477-478. Shetland. PhD. thesis, University of Edinburgh.
HIGGS, R. (1984) Possible wave-influenced sedimentary MELVIN, J. ( 1 986) Upper Carboniferous fine-grained tur­
structures in the Bude Formation (Lower Westphalian, biditic sandstones from southwest England: a model for
south-west England), and their environmental impli­ growth in an ancient, delta-fed subsea fan. 1. sediment.
cations. Proc. Ussher Soc., 6, 88-94. Petrol. , 56, 19-34.
HtGGS, R . ( 1 986a) 'Lake Bude' (early Westphalian , SW MELVIN, J . ( 1987) Upper Carboniferous fine-grained
England): storm-dominated sedimentation in an equa­ turbiditic sandstones from southwest England: a model
torial lake. Proc. Usshser Soc. , 6, 416-417. for growth in an ancient, delta-fed subsea fan: reply.
HtGGS, R . (1986b) A facies analysis of the Bude Formation 1. sediment. Petrol. , 57, 380-382.
(Lower Westphalian), SW England. PhD thesis, Uni­ MIDDLETON , G.V. (1965) Antidune cross-bedding in a
versity of Oxford. large flume. 1. sediment. Petrol., 35 , 922-927.
HtGGS, R. ( 1987) The fan that never was? - Discussion of MIDDLETON , G.V. & HAMPTON , M . A . (1976) Subaqueous
'Upper Carboniferous fine-grained turbiditic sandstones sediment transport and deposition by sediment gravity
from southwest England: a model for growth in an flows. In: Marine Sediment Transport and Environmental
ancient, delta-fed subsea fan'. 1. sediment. Petrol., 57, Management (Eds Stanley, D . J . & Swift, D . J . P . ) ,
378-379. p p . 197-218. Wiley, New York.
HIGGS, R. (1988) Fish trails in the Upper Carboniferous of MONACO, P. ( 1992) Hummocky cross-stratified deposits
south-west England. Palaeontology, 31, 255-272. and turbidites in some sequences of the Umbria-Marche
HIGGS, R . ( 1991) The Bude Formation (Lower West­ area (central Italy) during the Toarcian. Sedimentary
phalian), SW England: siliciclastic shelf sedimentation in Geol. , 77, 132- 142.
a large equatorial lake. Sedimentology, 38, 445-469. MoNEY, N.J. ( 1 966) Carboniferous rocks of north Devon:
Htscorr, R.N. ( 1981) Deep-sea fan deposits of the the Appledore Formation. Proc. Ussher Soc. , 1 ,
Macigno Formation (middle- upper Oligocene) of the 235-238.
Gordana Valley, northern Appenines, Italy: Discussion. MuTT!, E. (1977) Distinctive thin-bedded turbidite facies
1. sediment. Petrol. , 5 1 , 1015- 1033. and related depositional environments in the Eocene
HiscoTT, R.N. ( 1994) Traction-carpet stratification in Hecho Group (South-central Pyrenees, Spain).
134 R. V. Burne

Sedimentology, 24 , 107 - 13 1 . Nature, 208, 745 -747 .

Murr1, E . & RICCI-LUCCHI, F. (1972) L e torbediti READING , H . G . (1971) Sedimentation sequences in the
dell'Appenino settentrionale: introduzione all'analisi di Upper Carboniferous of Northwest Europe. Comptes
facies. Mem. Soc. Geol. !tal., 1 1 , 161 - 199. rendues 6e Congress international de Stratigraphie et
NELSON , C.H. , TwiCHELL, D . C . , SCHWAB, W . C . , LEE, Geologie Carbonifere, Sheffield, September 11 -14, 1967,
H . J . & KENYON, N. (1992) Upper Pleistocene turbidite Vol. 4 , pp. 140 1 - 1412.
sand beds and chaotic silt beds in the channelized, distal, READING , H.G. (1987) Fashions and models in sedi­
outer-fan lobes of the Mississippi fan . Geology, 20, mentology: a personal perspective. Sedimentology, 34,
693-696. 3-9.
NICHOLS, R .J . , SPARKS, R.S.J. & WILSON , C.J.N. ( 1994) REINECK, H .-E. & WUNDERLICH, F. ( 1968) Zur unter­
Experimental studies of the fluidization of layered sedi­ scheidung von asymmetrischen Oszillationsrippeln und
ments and the formation of fluid escape structures. Striimungsrippeln. Senckenbergina Lethaea, 49, 321 -·
Sedimentology , 41, 233-253. 354.
NoRMARK, W.R. & PIPER, D .J.W. ( 1991) Initiation Ricci-LUCCHJ, F. (1975) Depositional cycles in two turbidite
processes and flow evolution of turbidity currents: formations of northern Appenines (Italy) . J. sediment.
implications for the depositional record. In: From Shore­ Petrol. , 45, 3-43 .
line to Abyss: Contributions in Marine Geology in RoBERTS , H . H . & THAYER, P . A . (1985) Petrology of
Honour of Francis Parker Shepard (Ed. Osborne, R . H . ) , Mississippi Fan depositional environments. In: Sub­
Spec. Pub!. Soc. econ. Paleontol. Mineral. , Tulsa, 46, marine Fans and Related Turbidite Systems (Eds Bouma,
209-229. A . H . , Normark, W.R. & Barnes, N . E . ) , pp. 331 -339 .
·

O'CONNELL, S . , NORMARK, W.R. & KENYON, N . H . ( 1991) Springer-Verlag, Berlin.

An entrenched thalweg channel on the Rhone Fan: SAGEMAN , B . B . , WIGNALL, P . B . & KAUFFMAN , E . G . (1991)
interpretation from a SEABEAM and SEAMARC 1 Biofacies models for oxygen-deficient facies in epi­
survey. In: From Shoreline to Abyss: Contributions in continental seas: a tool for paleoenvironmental analysis .
Marine Geology in Honour of Francis Parker Shepard In: Cycles and Events in Stratigraphy (Eds Einsele, G . ,
(Ed. Osborne, R . H . ) , Spec. Pub!. Soc. econ. Paleontol. Rieken, W. & Seilacher, A . ) , pp. 542-564. Springer­
Mineral . , Tulsa, 46, 259-270. Verlag, Berlin.
OwEN, D . E . (1934) The Carboniferous rocks of the north SANDERS, J . E . (1965) Primary sedimentary structures
Cornish coast and their structures. Proc. Geol. Assoc. , formed by turbidity currents and related resedimentation
45, 451 -471 . mechanisms. In: Primary Sedimentary Structures and
OwEN , D . E . (1950) Carboniferous deposits i n Cornubia. their Hydrodynamic Interpretation (Ed. Middleton,
Trans. R. geol. Soc. Cornwall, 18 , 65 - 104. G.V.) Spec. Pub!. Soc. econ. Paleontol. Mineral. , Tulsa,
PARKER, G. (1982) Conditions for the ignition of cata­ 12, 192-219.
strophically erosive turbidity currents. Mar. Geol. , 46, SAVRDA, C.E. & BoiTJER, D . J . (1991) Oxygen-related bio­
307-327. facies in marine strata: an overview and update. In :
PIPER, D.J.W. & STOw, D.A.V. (1991) Fine-grained tur­ Modern and Ancient Continental Shelf Anoxia (Eds
bidites. In: Cycles and Events in Stratigraphy (Eds Tyson, R.V. & Pearson, T . H . ) , Spec. Pub!. geol. Soc.
Einsele, G . , Rieken, W. & Seilacher, A . ) , pp. 360-376. London, 58, 201 -219. Geological Society of London,
Springer-Verlag, Berlin. Bath.
PoSTMA, G. (1986) Classification for sediment gravity-flow ScHOLZ, C.A. & RoSENDAHL, B .R . (1990) Coarse-clastic
deposits based on flow conditions during sedimentation. facies and stratigraphic sequence models from Lakes
Geology, 14, 291-294. Malawi and Tanganyika, East Africa. In: Lacustrine
POSTMA, G. , NEMEC, W. & KLEINSPEHN , K.L. (1988) Large Basin Exploration: Case Studies and Modern Analogs
floating clasts in turbidites: a mechanism for their (Ed. Katz, B . J . ) , Mem. Am. Assoc. petrol. Geol. , Tulsa,
emplacement. Sediment. Geol. , 58, 47-61. 50, 151- 168.
PRAVE, A.R. & Du KE , W.L. (1990) Small-scale hummocky Scorr, D . L . , NG'ANG'A, P . , JoHNSON, T.C. & RosENDAHL,
cross-stratification in turbidites: a form of antidune B . R . (1991) High-resolution acoustic character of Lake
stratification . Sedimentology, 37, 531-539. Malawi (Nyasa) , East Africa and its relationship to
PRENTICE, J.E. (1960a) Dinantian, Namurian and sedimentary processes. In: Lacustrine Facies Analysis
Westphalian rocks of the district south-west of (Eds Anad6n, P . , Cabrera, Ll. & Kelts, K.), Spec. Publ.
Barnstaple, north Devon. Q. J. geol. Soc. London, 1 15 , int. Assoc. Sediment. , No. 13, 129-145. Blackwell
261-289. Scientific Publications, Oxford.
PRENTICE, J .E. (1960b) The stratigraphy of the Upper SEDGEWICK, A. & MURCHISON , R.I. ( 1840) On the physical
Carboniferous rocks of the Bideford region, north structure of Devonshire, and on the subdivisions and
Devon. Q. J. geol. Soc. London, 116, 397-408. geological relations of its older stratified deposits. Trans.
PRENTICE, J . E . (1962) The sedimentary history of geol. Soc. London, 5, 633-705.
the Carboniferous in Devon. In: Some Aspects of SEILACHER, A. ( 1982) Distinctive features of sandy
the Variscan Fold Belt (Ed. Coe, K.), pp. 93- 108. tempestites. In: Cyclic and Event Stratification (Eds
Manchester University Press, Manchester. Einsele, G. & Seilacher, A . ) , pp. 333-349. Springer­
READING , H . G . (1963) A sedimentological comparison of Verlag, Berlin.
the Bude Sandstones with the Northam and Abbotsham SHANMUGAM, G. & MoiOLA, R.J. ( 1988) Submarine fans:
Beds of Westward Ho! Proc. Ussher Soc. , 1, 68-70. characteristics, models, classification , and reservoir
READING , H . G . (1965) Recent finds in the Upper Carbon­ potential. Earth Sci. Rev. , 24, 383-428.
iferous of south-west England and their significance. SKIPPER, K. (1971) Antidune cross-stratification in a tur-
 Bude Formation subsea fans 135

bidite sequence, Cloridorme Formation, Gaspe, Quebec. Somerset archaeol. nat. Hist. Soc. , 38, 1 1 1 -2 19.
Sedimentology, 17, 5 1 -68. WALKER, R.G. ( 1963) Distinctive types of ripple-drift cross­
SKIPPER, K. & MIDDLETON , G.V. ( 1975) The sedimentary lamination. Sedimentology, 2, 173- 188.
structures and depositional mechanics of certain WALKER, R . G . ( 1 964a) Some aspects of the sedimentology
Ordovician turbidites, Cloridorme Formation, Gaspe of the Shale Grit and Grindslow Shales (Namurian R 1c ,
Peninsula, Quebec. Can. J. Earth Sci. , 12, 1934- 1952. Derbyshire) and the Westward Ho ! and Northam
STAUFFER, P.H. ( 1967) Grain-flow deposits and their Formations (Westphalian, north Devon) . PhD thesis,
implications, Santa Ynez Mountains, California. University of Oxford .
J. sediment. Petrol. , 37, 487-508. WALKER, R.G. ( 1964b) Some aspects of the sedimentology
STOw, D . & SHANMUGAM, G. ( 1 980) Sequence of structures of the Westward Ho! and Northam Formations. Proc.
in fine-grained turbidites: comparison of recent deep sea Ussher Soc. , 1 , 85-87.
and ancient flysch sediments. Sediment. Geol. , 25, WALKER, R.G. ( 1967) Upper flow regime bed forms in tur­
23-42. bidites of the Hatch Formation, Devonian of New York
S uNDBORG, A. ( 1956) The river Klaralven, a study of State. J. sediment. Petrol. , 4, 1052- 1058.
fluvial processes. Geogr. Annaler, 38, 127-316. WETZEL, A. ( 1991) Stratification in black shales: depo­
TANNER, P.W.G. ( 1 992) The duplex model: implications sitional models and timing - an overview. In: Cycles
from a study of flexural-slip duplexes. In: Thrust and Events in Stratigraphy (Eds Einsele, G . , Rieken, W.
Tectonics (Ed. McClay, K . R . ) , pp. 201 - 208. Chapman & Seilacher, A . ) , pp. 508-523. Springer-Verlag, Berlin.
& Hall, London. WHALLEY, J . S . & LLOYD, G . E . ( 1986) Tectonics of the
THOMAS, J . M . (1988) Basin history of the Culm Trough of Bude Formation , north Cornwall - the recognition of
southwest England. In: Sedimentation in a synorogenic northerly directed decollement. J. Geol. Soc. London,
basin complex; the Upper Carboniferous of Northwest 143, 83-88.
Europe (Eds Besly, B .M. & Kelling, G . ) , pp. 24-37. WooD, A. & SMITH , A.J. ( 1959) The sedimentation and
Blackie, Glasgow. sedimentary history of the Aberystwyth Grits (Upper
UNRUG, R. ( 1963) Istebna Beds - a fluxoturbidity forma­ Llandoverian) . Q. J. Geol. Soc. London, 114, 163- 195 .
tion in the Carpathian Flysch. Ann. Soc. geol. Pol. , 33, YAGISHITA, K. (1994) Antidunes and traction-carpet
49-92. deposits in deep-water channel sandstones, Cretaceous,
UNRUG, R. (1965) Le facies de flysch a fluxoturbidites dans British Columbia, Canada. J. sediment. Res. Sect. A:
les Carpathes Polonaises. Carpatho-Balkan Geological sediment. Petrol. Process. , A64, 34-4 1 .
Association, VII Congress, Sofia, September, Reports, ZIEGLER, P.A. ( 1990) Geological Atlas of Western and
Part II, pp. 301 -303. Central Europe. Shell Internationale Petroleum
UsSHER, W.A.E. ( 1 892) The British Culm Measures. Proc. Maatschappij B . V . and the Geological Society, London.
Spec. Pubis int. Ass. Sediment. (1995) 22, 137-156

Depositional controls on
iron formation associations in Canada

P HI LI P F R A L I C K* and T I M O T HY J. B A R RE T Tt
*Department of Geology, Lakehead University,
Thunder Bay, Ontario P7B 5El, Canada; and
tMineral Deposit Research Unit, Department of Geological Sciences, University of British Columbia,
Vancouver, British Columbia V6T 1Z4, Canada

ABSTRACT

Understanding Precambrian iron formation depositional processes has been hindered by the lack of
precise modern analogues. However, by combining a regional basin analysis of sedimentary and
volcanic rocks surrounding an iron formation with detailed examination of sedimentary structures and
lithic associations within an iron formation, the depositional setting and physical processes of sedimen­
tation can be inferred. Six iron formations present in the Canadian Shield were examined using this
approach. The Palaeoproterozoic Gunflint Formation consists of strand-proximal stromatolites and
oolitic shoals, transitional distally to grainstones and parallel-laminated chemical muds. This succession
was laid down on a wave- and tide-dominated inner shelf. In the northern Labrador Trough, the Middle
Member of the Palaeoproterozoic Baby Formation represents an outer shelf to slope environment. Here
massive, graded and ripple-laminated siltstone-shale couplets dominate the succession, with iron-rich
chemical sediment forming fine-grained tops to some couplets; thick assemblages of parallel laminated
chemical sediments also occur. In the Beardmore-Geraldton area, a submarine environment is
represented by oxide facies Archaean iron formations. The chemical units accumulated where clastic
mud would normally be found in coarsening and thickening upward levee and ramp assemblages. In the
Terrace Bay area, Archaean abyssal plain deposits consist of black, graphitic slate with pyrite layers and
zones of chert; these sediments lie on volcanics and are overlain by a submarine-ramp clastic assemblage.
Archaean volcanic-associated iron formation is represented by thin interflow sediment packages in a
submarine lava-plain south of Beardmore, and as a sulphide lens in a volcanic edifice south of Schreiber.
In the first five examples iron formation occupies the niche usually dominated by clastic mud. The iron
formation is able to form because of a reduced clastic supply reflecting some combination of the
following factors: a peneplained source region; a relative rise in sea-level creating sediment storage
capacity in subaerial and shallow-water regions; development of sediment bypass systems; and cessation
of active volcanism near the depositional area. The facies of iron formation in a given area is dependent
on factors such as water depth (both Eh and energy), degree of basin isolation from clastic sedimentation,
and amount of hydrothermal input or upwelling. Proximal to submarine hydrothermal vents, iron
formation accumulation can dominate without a reduction in clastic input; the mineralogy and layering
are controlled directly by the temperature, Eh and drift direction of hydrothermal discharge.

INTRODUCTION

Sediments accumulating in modern environments depositional systems, but in the Precambrian the
commonly provide a basis for comparison with present is sometimes not the key to the past. In the
ancient successions, in order to reconstruct the Archaean and Palaeoproterozoic, grossly different
palaeogeography and depositional processes that atmospheric and water chemistry, and possibly tem­
operated in a region. This approach is probably the perature, led to the deposition of chemical sediments
most useful technique for reconstructing ancient that have inadequate modern analogues . Iron
Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
137
138 P. Fralick and T.J. Barrett

formations, which are chemical sedimentary rocks bearing minerals, in either subaerial (Garrels et at.,
containing greater than 15% Fe (James, 1966), are 1973; Drever, 1974) or subaqueous (van Hise &
volumetrically the most important member of this Leith, 19 1 1; Huber, 1959) environments, and hydro­
group. Modern subaqueous hot-spring activity thermal ion exchange ( Goodwin, 1956; Gross, 1965;
provides clues to understanding the precipitation of Fralick, 1987). Simonson ( 1985) has shown that
Fe-rich sediments, but the mineralogy and areal lithological data from the Animikie Group and
extent of modern deposits is not comparable to Labrador Trough do not support a cratonic source
many iron formations deposited during the Earth's for the iron. Chemical data on trace elements,
early history. Theoretical chemical modelling has, to rare earths and stable isotopes do not differentiate
date, been unable to explain the interlayering of between low-temperature weathering of mafic vol­
iron-rich sediment and chert. canic rocks and high-temperature hydrothermal
A broader approach is necessary when conaucting alteration as possible sources for the iron. Iron
field studies of iron formation. For example, sedi­ released by both processes will be in solution in any
mentary and volcanic rocks surrounding iron for­ ocean, and although hydrothermal input is obviously
mations provide data on depositional settings; dominant in some situations, low-temperature
sedimentary structures and lithic associations within leaching may have, at times, played an important
iron formations give information on physical process role. In order for small quantities of Fe2+ to go into,
of deposition; and regional basin analysis provides and stay in solution, p02 must have been very low
a framework within which to view iron formation (Holland, 1973).
genesis. Combining these techniques makes palaeo­ A hydrothermal source has been proposed for the
geographical reconstruction and depositional model­ silica (Gross, 1965; Gross & Zajac, 1983; Fralick,
ling feasible. 1987), although Cloud ( 1973) believed the silica in
Iron formations are often classified into two main Palaeoproterozoic iron formations may have orig­
groups, those of Superior type, which were deposited inated from subaerial weathering. In either case, the
as thick, laterally extensive units in tectonically stable probable lack of silica-secreting organisms at this
areas, with sedimentary structures and textures time would have resulted in an ocean saturated with
indicative of shallow-water environments, and those H4Si04 (Siever, 1957; Cloud, 1973; Holland, 1973).
of Algoma type, which accumulated in tectonically The movement of large amounts of water into an
active regions as well-layered but discontinuous environment with a different Eh, pH, or temperature
units, mainly associated with volcanic rocks or deep­ is needed to precipitate iron and silica in the quan­
water sediments (Gross, 1965, 1983). This rather tities necessary to form an iron formation. This can
arbitrary subdivision is similar to the geosynclinal be achieved through the hydrothermal venting of
theory, initially useful, but not suited to our present hot, reduced, acidic solutions (Fralick, 1987; Barrett
data base. It limits iron formation to two types, et a!. , 1988), or upwelling of deep ocean waters with
when in reality there are a multitude of types. Iron a low p02 on to a more oxygenated shelf (Cloud,
formation characteristics are controlled by depo­ 1973, 1983; Holland, 1973; Drever, 1974). However,
sitional settings and processes that are complex these scenarios do not explain deep marine Archaean
multivariate systems and have little likelihood of iron formation deposited distally to hydrothermal
being repeated exactly in time or space. Thus, iron activity. In this situation, and with the oceanic
formations may be similar, but will not be identical, bottom waters saturated with iron and silica, hydro­
and each is capable of providing further information thermal venting may produce a zone of oversatu­
on how this unusual rock type forms. Classification ration and precipitation which extends away from
systems are of course useful in our nomenclature, the vent area for tens to possibly hundreds of kilo­
but they must not become an end in themselves. A metres (Fralick, 1987). In the Palaeoprotero­
fuller understanding of process and depositional zoic, and possibly the Archaean, microorganisms
setting should be the primary goal of research on influenced precipitation of iron and silica as they
iron formations. This requires more detailed work added 02 to the environment (Cloud, 1973), and
on individual units. possibly directly precipitated iron and silica as
Factors involved in the deposition of iron for­ coatings and tests (LaBerge et al., 1987), although
mation include the source of iron and silica, transport the latter process has been questioned by Oehler
mechanisms, and controls on precipitation. Possible ( 1976). Alternatively, near-surface photo-oxidation
sources for the iron include weathering of iron- of Fe2+ could have formed the iron-rich layers
 Depositional controls on iron formation 139

(Cairns-Smith, 1978; Braterman & Cairns-Smith, Formation (Fig. 1). This sedimentary package was
1987), and pressure reduction during upwelling could deposited on a peneplained Archaean surface
have caused silica to precipitate ( Holland, 1973). forming the southwestern margin of Superior
Studies of specific iron formation basins should Province (the edge of the North American craton at
attempt to integrate a number of the processes that time). Minor mafic flows and extensive tuf­
described above (cf. Morris, 1993), faceous horizons within the formation indicate that
In this paper we provide six examples from the the region was volcanically active at the time of
Canadian Shield which highlight the control that basin formation and subsidence. The Gunflint For­
depositional setting had on iron formation character­ mation is underlain locally by a basal conglomerate;
istics. The depositional environments described con­ the Gunflint has a gradational upper contact with
stitute a succession from shallow shelf through slope, black shales and siltstones of the Rove Formation.
rise and abyssal plain to areas of active volcanism. It The Gunflint Formation is divisible into two mem­
is apparent from these examples that iron formation bers and averages 120 m in thickness.
type is, like clastic deposits, controlled by depo­ The lower member of the Gunflint Formation
sitional setting. Of course we are not the first commonly has basal algal and oolitic cherts (Fig. 2).
to emphasize this fact. Gross ( 1980), and many The microfossil-bearing algal structures often occur
subsequent researchers (e.g. Ojakangas, 1983; on boulders of the basal conglomerate, which pro­
Simonson, 1985), have stated or implied similar vided a stable substrate. The remainder of the lower
ideas. What we strive to emphasize here is the member commonly consists of grainstone beds sep­
benefit of conducting iron formation research in a arated by fine-grained sediment layers (Fig. 3a).
manner similar to studies of clastic depositional Locally, this facies is termed cherty iron formation.
systems. Individual fine-grained layers are homogeneous and
iron-rich, but mineralogy varies from layer to layer.
Iron-oxide-chert and iron-silicate-chert mixtures
CASE STUDIES are the most abundant. Sand-sized, chemical mud
intraclasts forming the grainstones are commonly
similar in mineralogy to interbedded fine-grained
Objectives
layers. Larger rip-ups of material similar to the
The rationale for providing this series of case studies underlying substrate are present in many beds (Fig.
is to emphasize the control exerted by depositional 3c). Internal laminations are often difficult to discern
processes on iron formation attributes. The physical in the grainstone beds, due to the effects of dia­
features that iron formations exhibit are largely genesis, but trough cross-stratification (Fig. 3d) is
the result of their depositional environment. The not uncommon and some sections exhibit hummocky
headings in this section are not meant to be a cross-stratification (HCS). Small outcrop size and
classification system, but are purely descriptive. We normal faulting in the area, make analysis of
believe it is more productive to look upon each iron up-section trends within the Gunflint Formation
formation as a distinct entity reflecting the processes difficult, and limited amounts of drill-core must be
operative in the setting in which it formed. used. The section depicted in Fig. 2 exhibits an
upward decrease, then increase in the thickness of
grainstone beds through the lower member of the
Shallow shelf: Gunflint Formation
formation. Interstratified with the grainstones in
Iron formation, associated with other types of some sections are units metres to tens of metres
chemical and clastic sediment, deposited in shallow thick consisting of laminated, fine-grained chemical
marine settings is most common in the Palaeo­ sediment (Fig. 3b), locally called slaty iron for­
proterozoic, although well-documented examples mation, and similar in all aspects to the fine-grained
from the Archaean also exist. The Canadian Shield material in the grainstone assemblages. The lower
contains two extensive regions dominated by iron member is capped sporadically by breccia. In Fig. 2
formation associated with shallow marine sediments: this breccia is underlain by a massive layer of silt­
the Animikie Group and portions of the Labrador sized material.
Trough. The upper member of the Gunflint Formation is
Iron formation within the Canadian portion of the similar in stratigraphy to the lower member. Basal
Animikie Group is contained in the 2000 Ma Gunflint algal cherts (Fig. 3e), often developed directly on
140 P. Fralick and T.J. Barrett

c
Ungava Bay

Lake Superior 0 100km

Fig. I. (A) Location of the six iron

� Proterozoic
� Archean intrusive
L...±_j metamorphic terrains
formations discussed. (B) Regional
geology of the area north of Lake
rv:v.l Archean volcanic­ Superior. (C) Regional geology of
� sedimentary terrains the northern Labrador Trough.

top of fragments of the breccia, are succeeded by a 1988). Stromatolites and accretionary grains com-·
sequence of thinning upward grainstone beds separ­ monty appear to be limited to near-strand positions
ated by chemical sediment. Regionally extensive (Morey et at., 199 1). This spatial restriction is
tuffaceous shales, with accretionary lapilli present at not due to browsing metazoans, as in the case of
some localities, are interbedded with the grainstone Phanerozoic stromatolites, but probably due to
succession. A limestone unit caps the upper member frequent current activity causing large areas of
and separates the Gunflint Formation from the shales the offshore granular substrate to be unstable.
of the basal Rove Formation. Ojakangas ( 1983) has shown that water depth was
Lougheed ( 1983) attributed the depositional the major control on the development of 'cherty'
environments represented by the Animikie Group (grainstone dominated) versus 'slaty' (chemical mud
iron formations to supratidal, intertidal and subtidal dominated) iron formation. The grainstones were
zones of a broad shelf. The importance of subtidal deposited at shallower depths by tide-and-wave/
and possibly lower intertidal sand shoals as depo­ storm driven currents (Fig. 4). The chemical muds
sitional sites for the grainstone was emphasized by accumulated during fair-weather times in shallow-·
Simonson ( 1985). Bidirectional palaeocurrent indi­ water regions, and at all times in the deeper offshore.
cators at some sites indicate that tidal currents were Alternating successions of grainstone-dominated and
responsible for the accumulation of some grain­ mud-dominated iron formation packages suggest
stone units (Ojakangas, 1983). Hummocky cross­ fluctuations in relative sea-level (White, 1954;
stratification present in other sequences indicates Morey, 1983) and can be used to create sea-level
the importance of storm activity in creating and curves. Upward-thinning and upward-thickening
moving sand-sized intraclasts seaward (Fralick, successions of grainstone beds also suggest, respect-
 Depositional controls on iron formation 141

I LEGEND
?
Inter layered
....
Q)
Grainstone
.0
and Mudstone
E
Q)

50
�
Mudstone
....
( Volcaniclastir.)

lJ
Q)
c..
c..
:::J

Chert

J
Stromatolites

40 4 Quartz

\
Barite
Vein

Archean
Granitic

l30
Basement

Contorted
.�
1\
....
Q)
.0 <1>3: Bedding
-'<Q.
E .�:::>
Q) .c:
� 1--

� Breccia

·r
....

v
Q)
3: "E
0
0
_J
�g
"'3:
20

I �
'
Rip- up
Clasts

10 j
IJ
Cross-
Stratification

Hummocky

�
+ + +
Cross-

0
+ +
.S? "0 �
Fig. 2. Stratigraphical section of the + + + �
c 0
<1>0-
�
"'
Stratification
o�- c.
-
Gunflint Formation, Animikie + + .c 0 "'
� (j) .c "'
Group. The upper 20% of the + + + (j) Ul 0
formation was not present in this
drill hole. For detailed location see >.-.,
c == c
-UJ o
Kronberg & Fralick (1992). u
(j)

ively, relative rise and fall of sea-level (Fig. 2). The shift from chemical sedimentation to a volcaniclastic­
bed thickness trends shown in Fig. 2 suggest that dominated system.
sea-level fluctuations controlled the cyclic nature of The source of the iron- and silica-rich precipitates
the lower and upper members. The positioning of that form the majority of the Gunflint Formation is
the breccia and possible vadose silt unit at the lowest still controversial. Simonson ( 1985) favoured a direct
sea-level stand, inferred from bed thickness, suggests hydrothermal source in the offshore with coprecipi­
that these units formed through subaerial dissolution tation of silica and iron-rich phases in the nearshore
of carbonates. Increasing volcanic activity in the area due to ambient high levels of silica and iron
upper Gunflint Formation may have resulted in a in the ocean. Dissolved hydrothermal iron and
142 P. Fralick and T.J. Barrett

Fig. 3. Photographs of inner-shelf-associated iron formation in the Animikie Group. (a) Grainstone (intraclast)-dominated
succession. The grainstone beds are arranged in lenses which may cross-cut one another. Fine-grained chemical muds (m)
are interlayered with sand lenses. (b) Slaty iron formation. This chemical mud-dominated succession also contains
grainstone lenses but they are less abundant and smaller than in the cherty iron formation. (c) Thick-bedded cherty iron
formation with abundant rip-up clasts (r). (d) Inclined rip-up clasts in a grainstone lens lying on the backset slope of a dune.
(e) Rare calcium carbonate stromatolites; most other mounds are silicified. Lens cap is 5.5 em in diameter.
 Depositional controls on iron formation 143

Grainstone
(reworked
chemical
sediment)

Fine-grained chemical sediment

Fig_ 4_ Block diagram of the depositional environments in which the Gunflint chemical sediments accumulated. Deeper
areas not affected by currents induced by storms and tides are dominated by laminae of silica, iron silicate, iron carbonate
and iron oxide. Shallower areas affected by sporadic current activity consist of interlayered grainstone and beds of fine­
grained chemical sediment similar to that accumulating in deeper areas. Grainstone production occurred in the shallows
due to erosion and abrasion of chemical muds, with the material transported offshore during times of increased current
activity. Ooid shoals developed in strand-proximal zones of maximum turbulence, and stromatolites formed along the
strand where the substrate had been diagenetically hardened. Karstification occurred in subaerially exposed areas.

silica may have been carried up to shelf depths by departs from this trend with the upward appearance
upwelling along cratonic margins. and then domination of the unit by iron formation
(Fig. 5). This interval may be correlated with iron
formation present in the Knob Group to the south­
Outer shelf and slope: Middle Member, west of the Baby Formation.
Baby Formation Clastic layers within the Baby Formation are
commonly 0.5- 10 mm thick and non-graded with
Examples of iron formations deposited in outer shelf sharp lower and upper contacts (Fig. 6a). They are
and slope settings are rare. Similarly, descriptions of composed of siltstone, silty shale and occasionally
shallow-water and deep-sea clastic units are common very fine-grained sandstone. Silty shale layers often
but outer shelf and slope secessions are not well contain coarser silt streaks only a few grains thick.
represented. This may be due to their more restricted Graded units similar in other respects to the non­
distribution and lower preservation potential. graded layers also are common. The grading may
The Palaeoproterozoic Baby Formation of the consist of either an upward grain-size decrease with
northern Labrador Trough (Fig. 1) provides an no internal laminations, or alternating siltstone
example of iron formation deposition in an outer and silty shale laminae that thin and fine upwards.
shelf to slope setting. The trough separates the Current-ripple-laminated layers of fine sandstone
Superior Province from the Rae Province and under­ are interbedded with the fine-grained clastic suc­
went a major orogenic episode during closure cession. Beds are 1-4 em thick, wavy bedded and
between these two land masses. Sediments of the non-graded. Current ripples in the thicker units are
Baby Formation were deposited prior to the major larger and are commonly stacked. The undulating
phase of orogenic activity, probably during an earlier upper surfaces of the rippled units are sharply over­
extensional phase (Wares & Goutier, 1990; Shulski lain by silty shale, which fills troughs and covers
et al., 1993). They are underlain by dolomite and crests. Thicker, compound fine-grained sandstone
overlain by tholeiitic basalt. beds are also present in the succession. These are
The kilometre-thick Baby Formation is dominated up to 20 cm thick and are composed of vertically
by a monotonous, thin-bedded succession of fine­ stacked, 1-5-cm-thick, massive to parallel lami­
grained clastic sediment. The Middle Member nated, non-graded sandstone. The layers are some-
144 P. Fralick and T.J. Barrett

M Middle Member of the Baby Formation (Figs 5 &

6). Chemical sediment may form discrete millimetre­
scale layers either between graded silt-shale
60 Fe-C, couplets, or interlaminated with the shaley top of
Fe-0,
Fe-S
the couplet. Iron-rich minerals may be iron silicates,
oxides, carbonates or sulphides. One succession of
rhythmites containing thicker (average thickness=
50 4 em) carbonate grainstone layers is also present.
This succession is separated from a chemical sedi­
ment assemblage by a massive dolomite, 1.5 m thick.
ss
Carbonate layers in the grainstone succession contain
small- to medium-scale trough cross-stratification.
40
Chemical-dominated successions are composed of
SS, thinly to thickly laminated iron-rich, fine-grained,
Fe-C
sediment locally interstratified with chert or jasper
layers (or more rarely lenses). Carbonate, silicate,

f,
30 oxide and sulphide facies iron formation are all
Fe-S
present within the Baby Formation. Interlayering
of the facies is common, although all facies rarely
Fe-C
occur in the same interval. The iron-rich layers are
_
Overburden centimetre- to decimetre-scale in thickness, with
20
internal millimetre- and submillimetre-scale lami­
nations caused by differences in crystal size (Fig.
ss
6c). Microscopic examination reveals that the
laminae reflect changes in the amount of clastic
Fe-0,
10 Fe-S, Si material and chert present. The cherts interbedded
SS, Fe-C with the iron-rich units are commonly 1-3 em thick
SS, Fe-C, Si and exhibit no internal laminations except for rare
SS, Fe-P submillimetre-scale iron silicate layers.
0 SS, Fe-0, Fe-S, Si The Baby Formation forms part of an assemblage
that has been interpreted as a passive margin suc­
Fig. 5. Stratigraphical section of a portion of the Middle cession, thickening towards the east (Dimroth, 1981;
Member, Baby Formation. All units are internally layered
LeGallais & Lavoie, 1982; Wardle & Bailey, 1981).
on the millimetre- to centimetre-scale. Chemical sediment
occurs in the shale-rich tops of some clastic beds. SS, The abundant fine-grained sediments, the lack of
siltstone and shale; Si , chert; Fe-S, iron silicate; Fe-C, iron wave or tidal deposits and the rare presence of
carbonate; Fe-0, iron oxide; Fe-P, iron sulphide. carbonate grainstones transported by bottom cur­
rents suggest that deposition took place on the outer
shelf or upper slope (Fig. 7). The rhythmites were
deposited from low-density turbidity currents similar
times separated by millimetre-scale shale drapes. to those attributed to comparable units in the off­
Rare, thicker beds in the Middle Member consist of shore of northeastern North America (Chough
two types. (i) Massive, metre-scale medium-grained & Hesse, 1980; Hesse & Chough, 1980; Stow &
sandstone beds with coarse sand grains scattered Shanmugam, 1980). These sediment clouds may
throughout. Reverse grading may be developed near have been raised by sporadic slumping on the slope,
their tops. (ii) Organized beds, with a thin, basal, or as overflows from channelized bypass systems.
current-rippled, fine-grained sandstone overlain by The latter mechanism is preferred because slump
millimetre-thick, alternating coarse siltstone and silty scars are rare. Thin current-rippled units represent
shale laminae that thin and fine upwards. The upper intermittent bottom currents. Rippled units of this
halves of the units are composed of normally graded type are rarely described from slope deposits. The
silty shale. thick sandstone units provide better evidence for a
Iron formation occurs interbedded with clastics slope. Their massive nature combined with reverse
and as dominantly chemical successions within the grading indicates that downslope grainflow processes
 Depositional controls on iron formation 145

Fig. 6. Photographs of the iron formation associated with outer-shelf-slope facies in the Baby Formation. (a) Close-up of
thin-bedded sediments showing lenses and continuous layers of silt (light-coloured layers), interbedded with shales.
(b) Magnetite laminae assemblages (m), interlayered with lighter siltstone and shale laminae assemblages (s).
(c) Laminated magnetite. Grain-size differences produce the layering. (d) Interbedded siltstone (light, s) and magnetite
(dark, m); grey units (g) represent mixtures of magnetite and clastics. (e) Reflected-light photomicrograph of pyrite bands
(p) interlayered with clastic material (s). Scale bar is 0.5 mm.
146 P. Fralick and T.J. Barrett

Grain-flow

Channelized
turbidity
flows

Fig. 7. Block diagram of the

depositional environments in which
the Baby Formation chemical
sediments accumulated. Thinly
laminated iron-rich precipitates and
chert form discrete units or are
interlayered with Bouma d-e
turbidites in an outer shelf and
slope setting. Clastic material was
delivered to this environment
through channel overflow, slump­
generated low-density turbidity
currents, and grainflows originating
at times when the shelf was
dominated by quartz sand. Rare
iron carbonate sands were brought
into this environment when the shelf
Rainout of chemical sediment was starved of clastic material.

were at least in part responsible for their emplace­ resulting in the varied mineralogy of the iron for­
ment. The trough cross-stratified carbonate grain­ mation. The dominance of rainout processes and the
stone was obviously moved into the area by fairly variability of sediment influx and water chemistry
strong bottom currents originating on the inner shelf, produced an iron formation that is both thinly
where grain production would have taken place. laminated and laterally persistent.
This places the Middle Member of the Baby For­
mation on the outer shelf to slope break, past the
Submarine rise:
mud line (Stanley & Wear, 1978) but close enough
Beardmore-Geraldton Clastic Associated Iron
to areas affected by tide-, or storm-produced currents
Formation
to receive shelf sediments during unusually high­
velocity flow events. Archaean iron-formation-bearing successions that
The presence of iron formation in the Middle were probably deposited on submarine rises are
Member requires the clastic input to be greatly present in the Rainy Lake and Spirit Lake areas
diminished or chemical precipitation rates to be (Wood, 1980), Manitou Straits district (Teal &
greatly accelerated. Not enough data are present Walker, 1977), Lake St Joseph region (Meyn &
to choose between these two alternatives. Clastic Palonen, 1980), the Abitibi greenstone belt (Hyde,
supply may be controlled by: (i) the rate at which 1980), and the Beardmore-Geraldton terrain
clastic materials are supplied to the shelf; and (ii) (Barrett & Fralick, 1985, 1989). The latter will be
the storage capacity of the shelf combined with the used as the case example (Fig. 1B).
efficiency of sediment bypass systems in the outer The Beardmore-Geraldton terrane consists of
shelf to slope area. The rate of chemical precipitation three metasedimentary belts, each resting on a thick
was probably controlled by upwelling rates from the volcanic assemblage. A metavolcanic terrain lies to
deep ocean. As these variables fluctuated, the system the north and the gneisses and granites of the Quetico
oscillated between clastic and chemical dominance. Subprovince lie to the south. The northern meta­
The Baby Formation has a great variety of iron­ sedimentary belt is composed of fluvial conglom­
rich mineral phases, which probably reflects the erates and sandstones (Devaney, 1987), the central
physiographic setting on the outer shelf. There, belt consists of a prograding shoreline to offshore
upwelling currents would first encounter shallower turbidites (Devaney & Fralick, 1985), and the
waters. This may have led to large changes in Eh, southern metasedimentary belt contains turbidites
pH and dissolved concentrations through time, and iron formation (Barrett & Fralick, 1985, 1989).
 Depositional controls on iron formation 147

The belts are kilometres wide and tens of kilometres tiguous or separated by millimetre-thick laminations
in length, with subvertical bedding younging to the of magnetite-rich sediment (Fig. 9a, b, c & e). IFLA
north. The assemblage represents a fore-arc basin c: sand-rich composite units up to about 1 m thick,
that has been stacked tectonically (Barrett & Fralick, generally consisting of thin, stacked, ungraded,
1989; Devaney & Williams, 1989). Oxide facies iron laminated sand beds. These units consist of medium­
formation occurs interbedded with clastics in the to coarse-grained sand. The composite units are
turbiditic portion of the assemblage (Fig. 8). separated by intervals of magnetite or magnetite and
The turbidites are grouped into thick-, medium­ siltstone up to 15 em thick (Fig. 9a & b). IFLA d:
and thin-bedded turbidite-dominated associations, framework-supported polymictic conglomerate beds
and a thin-bedded iron-formation-clastic-sediment up to a few metres thick, interbedded with sandstone
association (Barrett & Fralick, 1989). The bulk of and minor iron formation, or fairly thick iron for­
the succession is composed of the first three associ­ mation and thin-bedded sands (Barrett & Fralick,
ations, which forms a submarine ramp, with attri­ 1985, 1989). The first three lithofacies associations
butes suggesting fan development at some locations. are commonly, although not exclusively, organized
The thin bedded, iron-formation-clastic-sediment into coarsening upward successions 1-50 m thich
association forms packages metres to tens of metres (Fig. 8) (Fralick, 1987). Sandstone assemblages are
thick, interstratified with the ramp/fan turbidites, sharply overlain by IFLA a, which is gradational
and is itself divisible into four iron formation upwards through IFLA b to IFLA c. This package is
lithofacies associations (IFLA) (Figs 8 & 9). then overlain by ramp/fan turbidites (Barrett &
Their characteristics are as follows. IFLA a: domi­ Fralick, 1989).
nantly magnetite-rich sediment with millimetre- to The ramp-fan system represented by the
centimetre-scale, graded or ungraded silt interbeds Beardmore-Geraldton turbidites provided few
(Fig. 9a, b, d, f & g). IFLA b: centimetre-scale, locations where clastic influx was low enough to
graded to sharply bounded silt beds, either con- allow iron formation accumulation (Fig. 10). To the

A B
IFLAa
M M
• interlaminated
magnetite,
50 6
siltstone,
and slate
IFLAa
II interlaminated
magnetite,
40 4 jasper,
siltstone,
and slate

D 0
Sandstone

30 2
lliJ IF LAc
interbedded
sandstone,
siltstone,
slate and
magnetite
0
IFLAb
20
� - interbedded
siltstone,
Fig. 8. Stratigraphical sections of slate and
magnetite
iron-formation-bearing units in the
Beardmore-Geraldton area.
10
(A) Outcrop section measured at
the Leitch Mine (detailed location
in Barrett and Fralick, 1985).
(B) Outcrop section measured at
Solomon's Pillars (detailed location
in Fralick, 1987). 0
148 P. Fralick and T.J. Barrett

Fig. 9. Photographs of iron formation associated with an inner submarine ramp-fan complex in the Beardmore­
Geraldton region. (a) Magnetite layers (dark) interstratified with Bouma d-e turbidites. Clastic and chemical dominated
intervals are visible (IFLA a=a; IFLA b=b; IFLA c=c). (b) Interlayered magnetite and d-e turbidites forming two thin
coarsening and thickening upward successions (arrows). (c and d) Alternating magnetite-rich (dark) and clastic-rich (light)
laminations (prints of thin-sections). Scale bar is 5 mm. (e) Photomicrograph showing gradation from clastic-dominated
bottom to magnetite-dominated top of two thin d-e turbidites. Scale bar is 1 mm. (f) Photomicrograph of interlaminated
magnetite (dark) and silt (light). Scale bar is 0.5 mm. (g) Photomicrograph of a magnetite-dominated sequence. The lighter
layers are mixtures of magnetite and chert. Scale bar is 0.1 mm.

south in Quetico Subprovince, where prograding activity sporadically shifted in the arc to the north,
channel-fed lobes merge, iron formation is rare. In sediment influx rates also shifted laterally along
the Beardmore-Geraldton area, interchannel ramp the multiple sediment entry points to the basin.
development limits iron formation sedimentation. This, and probably autocydic processes of channel
Chemical sediment could form only in interchannel switching, led to the transformation of clastic­
areas where ramp progradation was limited by an dominated interchannel areas into iron-oxide­
insufficient sediment supply. As areas of volcanic dominated environments. Either levee outbuilding,
 Depositional controls on iron formation 149

Fine-grained
chemical sediment
Debris-flow

Fig. 10. Block diagram of the

depositional environment in which
the Beardmore-Geraldton
chemical sediments accumulated.
Iron oxide successions interbedded
with d-e turbidites developed in
clastic-sediment-starved
interchannel areas not receiving
ramp turbidites. These successions
commonly coarsen and thicken
upwards, probably due either to
levee or to ramp progradation into
the interchannel sites.

caused by channel re-establishment in the area, or

ramp progradation, due to renewed high rates of
sediment supply, overwhelmed these chemical
systems, causing a gradual return to a clastic­
dominated bottom. Conglomerates interbedded with
iron formation attest to an upper ramp/fan position
for the chemical sediments. Conglomerates are not
present in the more distal sediments to the south
(Fralick et al., 1992).

Abyssal plain:
Terrace Bay Clastic Associated Iron Formation

Areas described as graphitic shear zones are common

throughout the Canadian Shield. They are usually 3
black, graphitic, pyritiferous slates, metres to tens of
metres in thickness, which have been sheared, due � Felsic porphyry
intrusive
to their incompetence. Depositional environments
for these units vary, but the majority probably lnterlaminated

represent abyssal oceanic muds which accumulated �- pyritic-carbonaceous

slate and pyrite
in settings distal to active sediment sources. The
area to the east of Terrace Bay (Fig. 1B) contains a • Massive pyrite

number of graphitic slate assemblages. They appear

to represent fault and fold repetition of the transition � Chert

from sea-floor volcanics to overlying clastics (Fralick

& Barrett, 1991; Eriksson etal., 1994).
The Terrace Bay sediments occupy a basin
bordered to the northeast and southwest by arc­
related volcanics. The sediments also overlie a Fig. 11. Representative stratigraphical section of a portion
volcanic pile kilometres in thickness. Thin, pyri­ of the Kingdom iron formation in the Terrace Bay area.
tiferous black slates occur between some of the flows The 40-m-thick iron formation is underlain by altered,
near the top of the pile (Fig. 12a) and a fine-grained intermediate, pillowed volcanics and overlain by
turbidites. Finely interlaminated carbonaceous slate,
clastic-chemical sedimentary unit, up to 40 m in
pyrite and chert form most of the iron formation, with
thickness, directly overlies the volcanics. This unit occasional pyrite layers up to 20 cm in thickness (detailed
consists of black, graphitic slate with intricately location in Schnieders, 1987).
150 P. Fralick and T.J. Barrett

Fig. 12. Photographs of iron formation associated with an abyssal plain assemblage in the Terrace Bay area. (a) Pyrite and
chert form a zone (p) between two pillows (v). The pillow selvages have been highly silicified (si). These chemical
sediments and volcanics lie directly below the iron formation (Kingdom Occurrence in Schnieders, 1987). (b) Interlayered
slate (s), pyrite (p) and chert (c). Deformation has caused some disruption but the fine layering is still visible.
(c) Interlayered pyrite (p) and black slate (s). (d) Turbidite succession above graphitic slate-chemical sediment zone.

laminated pyrite layers and zones of chert (Fig. 1 1; or coarse-grained sandstone to siltstone or slate and
Schnieders, 1987; Barrett et a!., 1988). The pyrite exhibit features typical of turbidites (Fig. 12d). The
occurs as massive layers (Fig. 12c) up to 25 em thick; beds vary in thickness from several centimetres to
as thin, millimetre-scale laminations in the mudrock several metres. Successions of turbidite beds range
and chert (Fig. 12b & c); as centimetre-scale spheres from a-dominated to d-e-dominated; and thinning­
floating in the mudrock and chert (Fig. 12a & b); fining and thickening-coarsening upward success­
and as disseminated pyrite cubes. The cherts form ions are present, although uncommon.
zones, up to tens of centimetres thick, of either pure Deposition in the area began with subaqueous
silica or mixtures of silica, clay and/or pyrite (Fig. volcanism building a thick extrusive succession.
12b). The black slates are composed of quartz, Waning volcanic activity allowed sediment to
sericite, carbonate, chlorite, pyrite and graphite, accumulate between successive flows. With total
with scattered, floating, angular quartz and feldspar cessation of volcanic activity, a blanket of fine­
grains of possible tuffaceous origin, and rare volcanic grained sediment was deposited (Fig. 13). Venting
shards. Bouma d-e turbidites are occasionally inter­ hydrothermal fluids produced alteration zones in the
bedded with the metalliferous succession. underlying volcanics and added Si, Fe and trace
A metasandstone-slate assemblage (kilometres amounts of other metals to the fine-grained mud
in thickness) overlies the succession of metalliferous raining down on the bottom. Outbuilding of a sand­
sediments. Beds are commonly graded from medium- rich submarine fan-ramp complex ended deposition
 Depositional controls on iron formation 15 1

Sand lobes

Fig. 13. Block diagram of the

depositional setting in which the
Terrace Bay chemical sediments are
interpreted to have accumulated.
Intermediate to mafic volcanic sea­
sediment
floor was covered by carbonaceous
and clay
muds interlayered with iron
sulphides and chert. Prograding
turbiditic fan/ramps built out over
the chemical-clastic mixture. Intermediate and mafic volcanics

of the metalliferous muds. The metal-rich deposits Province. Examples from the Abitibi Subprovince
are interbedded with the outer-fan/ramp turbidites comprise major portions of cycles I and II of Dimroth
of this complex (Fralick & Schnieders, 1986; et at. ( 1982), the basal komatiitic and tholeiitic
Schnieders, 1987). portions of supergroups described by Jensen ( 1985)
and, in particular, the Kinojevis Group (Jensen,
1978a, b, 198 1). Sedimentary rocks associated with
Submarine lava plain:
the lava-plain successions are not abundant and
Beardmore Volcanic Associated Iron Formation
comprise thin, interflow units, Fralick ( 1987)
Thick lava-plain successions are common in Superior described a series of interflow sedimentary units

A B

Layered siltstone
and shale

Massive
siltstone

Layered pyrite
E2J
1-
Mafic
intrusive

Pyrite and
layered chert
� Pillowed and
massive flows

Pyrrhotite and
\A/\ Volcanic ash

disrupted chert
� Chert

II
f+:+l Felsic dike lnterlaminated
� magnetite
and chert
Intermediate
volcanics

Fig. 14. Stratigraphical sections representative of volcanic-associated iron formation in (A) the Schreiber area (location in
Fralick et al.,1989), and (B) the Beardmore-Geraldton area (location in Fralick, 1987).
152 P. Fralick and T.J. Barrett

present in a volcanic pile underlying a turbiditic with proximity of the hydrothermal vent sources
succession in the Beardmore area (Fig. 1B). Oxide (Fralick, 1987; Fig. 16). The thickness of the inter­
and carbonate iron formation constitute the domi­ flow sediment packages is controlled by four factors.
nant lithologies present in the interflow sediment Time duration between flows is obviously important,
(Fig. 14B). Layers of magnetite (Fig. 15c & d) or but the rate of hydrothermal emission, proximity to
·
side1:ite (Fig. 15a & b) range from submillimetre, to the active vents and bottom-current patterns are
centimetres in thickness, with interbedded chert also major controls on unit thickness (Fralick, 1987).
(Fig. 15a, b, c & d) sometimes attaining thicknesses
of metres. Laminations within beds are not common.
Submarine extrusive edifice: Morley occurrence
Chlorite-rich units (with plagioclase and quartz) are
often associated with the iron-rich laminae, forming Submarine volcanoes produce relief of the sea­
bundles of alternating iron-rich and chlorite-rich floor. The Canadian Shield contains many such suc­
layers. cessions, usually dominated by intermediate and
Clastic supply to the area was limited to volcani­ felsic volcanics. Associated iron formations may be
clastic ash, which forms the chloritic layers and similar to those described in the section on lava
isolated grains in the iron formation. The chemical plains, although they have a tendency to be inter·­
sediment layers tend to be purer and thicker than stratified with thicker successions of pyroclastic rocks
the iron formations previously described. This is due to the more explosive nature of the volcanism.
probably the result of limited clastic supply together Massive sulphide build-ups are also common in this

Fig. 15.Photographs of volcanic-associated iron formation in the Beardmore area. (a) lnterlayered chert, c, and iron
carbonate, a, at the Empire Mine, south of Beardmore. (b) Close-up of (a). The dark layers are siderite. (c) Interflow
sediment consisting of chert, c, and magnetite, m. The thickest magnetite layer is loaded into the underlying chert.
(d) Photomicrograph of interlayered magnetite-rich, m, and chert-rich, c, laminae. Scale bar is 0.5 mm.
 Depositional controls on iron formation 153

Off-axis
volcanism

proximal
chemical
sediment

sediment
and clay

Fig. 16. Block diagram representing the depositional setting of iron formation types similar to those near Schreiber and
Beardmore-Geraldton. Thick successions of pyrite and chert formed in locations close to vents, probably associated with
axial valleys, off-axis volcanism or rifted arcs. Iron oxides, silicates, carbonates and chert were associated with either
less i ntense and/or lower temperature portions of discharging hydrothermal zones, or were precipitated distal to major
hydrothermal point sources.

setting. Iron-rich massive sulphide deposits are found floor (Fralick et at., 1989; Fralick & Barrett, 199 1).
in clastic- and volcanic-dominated abyssal plain Clastic supply to this area was limited (Fig. 16),
assemblages, but tend to be more prolific in felsic­ although some fine-grained material, of probable
dominated piles. The Morley occurrence will be volcaniclastic origin, was being delivered. Each
used as an example of this type of iron formation. It lamina in the pyrite beds probably reflects a short­
is located south of Schreiber in a volcanic pile term hydrothermal injection into a stagnant bottom
immediately west of the area discussed for the layer of water. Upward-thinning bundles of laminae
abyssal plain association (Fig. 1B). may correspond to medium-duration hydrothermal
The volcanic succession containing the iron for­ events. The domal structures, and high carbon con­
mation was deposited about 2. 7 Ga ago, and rep­ tent of the sediment provide evidence for relatively
resents a large, arc-type edifice (Schnieders, 1987; deep-water organic mats during chemical precipi­
Fralick & Barrett, 199 1). The Morley deposit is a tation (Fralick et at., 1989).
lenticular chemical sedimentary unit up to 7 m thick
which is underlain by intermediate flows and pyro­
clastic rocks, and overlain by thin turbidites and DISCUSSION
structurally emplaced mafic flows (Fralick et at.,
1989). The lower portion of the unit (Fig. 14A, The case studies illustrate that iron formations could
6 -; 10 m) contains pyrite with interbanded light and form in a wide range of marine environments, given
dark chert: the upper half consists of bedded to the right conditions. Shallow shelves appear to
laminated pyrite (Fig. 14A, 10- 1 1.5 m). Within the be 'all or nothing ' environments. Here sediments
pyrite-rich upper part of the succession, a variety of are distributed on the shelf commonly by non­
bedding structures are developed. The pyrite layers channelized flows. Clastics delivered to the near­
contain delicate internal laminations of pyrite and shore will be spread over large areas in the offshore,
carbonaceous chert from 0.02 to 1 mm thick. Near limiting iron formation development. To form iron
the tops of individual pyrite laminae the proportion formations, clastic delivery must be minimal, such
of chert and disseminated clastic debris is greater. that chemical sedimentation has the opportunity to
Colloform pyrite domes up to 3 em across and 2 em dominate. This leads to the development of Superior
thick are also present. Millimetre-scale mudstone Type iron formation with great lateral extent, and
laminae thin over the small pyrite domes and thicken sedimentary structures indicative of shallow water.
in flanking depressions. Discordant pyrite growth­ Limited clastic delivery can be achieved if sediment
structures on domes, together with inclined pyrite storage sites are available near the strand, thereby
crusts and microslumps, indicate that pyrite accumu­ preserving a chemical-dominated offshore. An
lation produced an irregular microrelief on the sea- example of this occurs in Gunflint correlatives in
154 P. Fralick and T.J. Barrett

Michigan, where clastic tidal flats with magnetite Portions of oceanic abyssal plains removed from
layers served as depositional sites for sands, silts and major clastic supply provided the most st:�ble sites
clays, leaving subtidal areas dominated by iron for­ for iron formation development. Here Precambrian
mation and chert (LaBerge et al . , 1992). sedimentary successions are commonly dominated
Other shelves dominated by chemical sediment by iron formation and fine-grained clastics. Factors
display features quite different from the Gunflint controlling chemical sediment deposition in other
succession. Carbonates and cherts of the Archaean areas were of little consequence on the abyssal plains.
Steep Rock Group exhibit prolific stromatolite A mixture of iron formation and fine-grained clastic
development (Wilks & Nisbet, 1988) and lack grain­ sediment could continue to be deposited until vol­
stones. This may be due to quiet-water conditions or canism or outbuilding of a clastic pile buried this
other, as yet unexplained, factors. Waves, storm lithofacies.
surges and tidal currents operating on shelves, com­ Depositional environments in volcanic terrains
bined with relative sea-level changes, variation in vary as much as in their clastic counterparts.
clastic influx , climate and mixing rates with offshore Although only two examples of volcanic-associated
water, all govern the physical attributes of the iron formation are discussed here, they do, however,
chemical sediments deposited in this setting. illustrate some general characteristics of chemical
Physical controls on chemical sedimentation on sediment deposited in this setting. Layering of
outer shelves and slopes are less well understood. chemical sediments in volcanic terrains reflects the
Clastic supply is obviously a major control. Sea-level dominant effect of discrete hydrothermal venting
rise may be a key to limiting clastic supply as it events. Layer thickness and type are inferred to be
provides more storage capacity on the shelf. It there­ controlled by variations of temperature and com­
fore becomes important to study the linkage in position of venting fluids (Fralick, 1987 ; Fralick
depositional response between coeval inner shelf, et al., 1989). Interbedded ash layers may reflect
outer shelf and slope systems during periods of sea­ magma recharge events with related increases
level rise. The development of sediment bypass zones in hydrothermal activity. These types of iron for­
can also produce interchannel areas dominated by mation should provide the best source of data on the
chemical sediment. Upper slope interchannel areas relationship between iron formation bedding attri­
are poorly described in the literature. This hinders butes and hydrothermal activity. Layered poly­
evaluation of sediment bypass as a control on iron metallic massive sulphide deposits are likewise an
formation development. excellent source of this type of information.
In submarine fan and ramp environments, devel­ The above discussion highlights our relative lack
opment of sediment bypass systems has been of knowledge of physical controls of iron formation
documented as a major control on formation of accumulation. Further work emphasizing the linkage
areas in which chemical sediment can accumulate. between various scales of depositional process and
The development of a channel does not always iron formation attributes is needed. More work
prevent interchannel ramp progradation from is also needed on chemical controls. Studies of
flooding off-channel areas with clastic material major and trace element chemistry have been of
(Barrett & Fralick, 1989). Changes in location of limited value, especially where whole-rock samples
volcanic activity alter sediment supply rates along rather than individual layers have been analysed.
the ramp, switching depositional systems from Stable isotope and REE studies have proven more
chemical to clastic and visa versa. Bottom conditions interesting, although more data are needed on mono­
become more unsuitable for iron formation accumu­ mineralic samples. Experimental modelling of the
lation further downslope on the ramp or fan. In this chemical systems responsible for Fe and Si precipi­
area, flows become non-channelized, spreading tation appears to be the missing link at present.
out over the bottom. Sediment-starved areas are In particular, we need to study the way in which
uncommon and likewise so is iron formation. To progressive saturation of stable bottom-water layers
deposit iron formation here, major portions of the by hydrothermal injections controls the nature and
clastic supply system must be shut off. This may be sequence of precipitation of iron-rich minerals. Only
accomplished through sea-level rise creating more by integrating sedimentological, geochemical and
sediment storage capacity in upslope environments experimental studies of iron formation can we further
or climatic change in the hinterland causing sediment unravel the processes that formed these intriguing
transport systems to dry up. sediments.
 Depositional controls on iron formation 155

ACKNOW LEDGEMENTS DEVANEY, J .R. & WILLIAMS, H.R. (1989) Evolution of an

Archean subprovince boundary: a sedimentological
We are particularly grateful to the Thunder Bay and structural study of part of the Wabigoon -Quetico
boundary in n orthern Ontario. Can. J. Earth Sci. , 26,
staff of the Ministry of Northern Development and 1013-1026.
Mines (Mines and Mineral Division) for helpful DIMROTH , E. (1981) Labrador Geosyncline: type example of
discussions and information on occurrences of early Proterozoic cratonic reactivation. In : Precambrian
iron formation in this area. Field-work in northern Plate Tectonics (Ed. Kroner, A.), Developments in
Precambrian Geology 4, pp. 331-352. Elsevier,
Quebec was assisted by Bob Wares. Useful com­
Amsterdam.
ments and suggestions on an earlier version of the DIMROTH , E., IMREH, L., ROCHELEAU, M. & GOULET, N .
manuscript were provided by Richard Hyde, Guy (1982) Evolution of the south-central part of the Archean
Plint and an anonymous reviewer. Figures were Abitibi Belt, Quebec. Part I: stratigraphy and paleo­
drafted by Sam Spivak and the word processing was geographic model. Can. J. Earth Sci. , 19, 1729- 1758.
DREYER, J . l . , (1974) Geochemical model for the origins of
conducted by Wendy Bourke. This research was Precambrian banded iron formations. Ceo/. Soc. Am.
supported by the Natural Sciences and Engineering Bull. , 85 , 1099-1106.
Research Council of Canada. ERIKSSON , K. A., KRAPEZ, B. & FRALICK, P.W. (1994)
Sedimentology of Archaean greenstone belts: signatures
of tectonic evolution. Earth Sci. Rev. , 37, 1-88.
FRALICK, P.W. (1987) Depositional environment of
REFERENCES Archean iron-formation: inferences from layering in
sediment and volcanic hosted end members. In : Pre­
BARRETT, T.J. & FRALICK, P.W. (1985) Sediment redepo­ cambrian Iron- Formation (Eds Appel, P. & LaBerge,
sition in Archean iron formation: examples from the G.), pp. 251-266. Theophrastus Publications, S.A.
Beardmore-Geraldton greenstone belt, Ontario. J. Athens.
Sediment. Petrol. , 55, 205-212. FRALICK, P.W. (1988) Microbial bioherms, Lower Proter­
BARRETT, T.J. & FRALICK, P.W. (1989) Turbidites and iron ozoic Gunflint Formation, Thunder Bay, Ontario. In:
formation, Beardmore-Geraldton, Ontario: application Reefs. Canada and A djacent A reas (Eds. Geldsetzer,
of a combined ramp/fan model to Archaean clastic and H.H.J., James, N . P. & Tebbutt, G.E.), Mem. Can. Soc.
chemical sedimentation. Sedimentology , 36, 221-234. petrol. Geol., Calgary, 13, 24-29.
BARRET!', T.J . , FRALICK, P.W. & JARVIS, I. (1988) Rare­ FRALICK, P.W. & BARRETT, T.J. (1991) Precambrian
earth-element geochemistry of some Archean iron for­ depositional systems along the southwestern edge of the
mations north of Lake Superior, Ontario. Can. J. Earth Superior Craton. Geological Association of Canada,
Sci. , 25, 570-580. Annual Meeting Fieldtrip A3 Guidebook, 54 pp.
BRATERMAN , P.S. & CAIRNS-SMITH , A.G. (1987) Iron photo­ FRALICK, P.W. & SCHNIEDERS, B.R. (1986) Depositional
precipitation and the genesis of the banded iron­ environments and tectonic implications of an Archean
formation s. In : Precambrian Iron-formations (Eds volcanic-metalliferous slate-turbiditic succession, north­
Appel, P.W.U. & LaBerge, G.L.), pp. 215-245. ern Ontario, Canada, 12th International Sedimentological
Theophrastus Publications, S.A., Athens. Congress, Programme with Abstracts, Canberra, p. 110.
CHOUGH, S.K. & HESSE, R. (1980) The northwest Atlantic FRALICK, P.W., BARRETT, T.J., JARVIS, K.E., JARVIS, I.,
Mid-Ocean Channel of the Labrador Sea: III. Head spill ScHNIEDERS, B.R. & VANDE KEMP, R. (1989) Sulfide­
deposits from turbidity currents on natural levees. facies iron-formation at the Archean Morley occurrence,
J. sediment Petrol. , 50, 227-234. northwestern Ontario: contrasts with oceanic hydro­
CLOU D , P. ( 1973) Paleoecological significance of the banded thermal deposits. Can. Mineral. , 27, 601-616.
iron-formation . Econ. Ceo!. , 68, 1135-1143. FRALICK , P.W., Wu, J. & WILLIAMS, H.R. (1992) Trench
CLOUD, P. (1983) Banded iron-formation - a gradualist's and slope basin deposits in an Archean metasedimentary
dilemma. In: Iron-formation Facts and Problems (Eds belt, Superior Province, Canadian Shield. Can. J. Earth
Trendell, A.F. & Morris R.C.), pp. 401-416. Elsevier, Sci. , 29, 2551-2557.
New York. GARRELS, R.M., PERRY, E.A. J R . & MACKENZIE, F.T.
CAIRNS-SMITH , A. G. (1978) Precambrian solution geo­ (1973) Genesis of Precambrian iron formations and the
chemistry, inverse segregation, and banded iron for­ development of atmospheric oxygen. Econ. Ceo/. , 68,
mations. Nature, 148, 27-35. 1173-1179.
DEVANEY, J . R. (1987) Sedimentology and stratigraphy of GooDWIN , A.M. (1956) Facies relationships in the Gunflint
the northern and central metasedimentary belts in the iron-formation. Econ. Ceo!. , 5 1 , 565-595.
Beardmore- Geraldton area of northern Ontario. GRoss, G.A. (1965) Geology of iron deposits in Canada,
Unpublished MSc thesis, Lakehead University, Thunder Vol. 1, general geology and evaluation of iron deposits.
Bay, Ontario, 227 pp. Ceo/. Surv. Can. econ. Ceo/. Rep . , 22, 181 pp.
DEVANEY, J.R. & FRALICK, P.W. (1985) Regional sedimen­ GRoss, G.A. (1980) A classification of iron formations
tology of the Namewamenikan Group, northern Ontario: based on depositional environments. Can. Mineral. , 18,
Archean fluvial fans, braided rivers, deltas and an 215-222.
aquabasin. In : Current Research, Part B, Paper 85- J B , GRoss, G.A. (1983) Tectonic systems and the deposition
pp. 125-132. Geological Survey of Canada, Ottawa. of iron-formation. Precam. Res. , 20, 17 1-187.
156 P. Fralick and T.J. Barrett

GRoss, G.A. & ZAJAC , I.S. (1983) Iron formations in MoREY, G.B., SouTHWICK, D . L. & ScHOlT LE R , S.P. (1991)
fold belts marginal to the Ungava Craton. In: Iron­ Manganiferous zones in Early Proterozoic iron­
formation: Facts and Problems (Eds Trendell, A.F. & formation in the Emily district, Cuyuna Range, east­
Morris, R.C.), pp. 253-294. Elsevier, New Y ork. central Minnesota. Minn. geol. Surv. Rep. Invest. , 391,
HESSE, R. & CHOUGH, S.K. (1980) The northwest Atlantic 42 pp.
Mid-Ocean Channel of the Labrador Sea: Il. D eposition MoRRIS, R.C. (1993) Genetic modelling for banded iron­
of parallel laminated levee-muds from the viscous sub­ formation of the Hamersley Group, Pilbara Craton,
layer of low density turbidity currents. Sedimentology , Western Australia. Precam. Res. , 60, 243-286.
27, 697-711. OEHLER, J .H. (1976) Hydrothermal crystallization of silica
HoLLAND, H.D. (1973) The oceans: a possible source of gel. Geol. Soc. Am. Bull. , 87, 1143-1152.
iron in iron-formations. Econ. Geol. , 68, 1169-1172. 0JAKANGAS, R.W. (1983) Tidal deposits in the early
HuBER, N . K. (1959) Some aspects of the origin of the Proterozic basins of the Lake Superior region. The Palus
Ironwood iron formations of Michigan and Wisconsin. and the Pokegama Formations; evidence for subtidal­
Econ. Geol. , 54, 82-118. shelf deposition of Superior-type banded iron-formation .
HYDE, R.S. (1980) Sedimentary facies in the Archean In : Early Proterozoic Geology of the Lake Superior
Timiskaming Group and their tectonic implications, Region (Ed. Medaris, L. G.) , Geol. Soc. Am. Mem.,
Abitibi Greenstone Belt, N ortheastern Ontario, Canada. Boulder, 160, 49-66.
Precam. Res. , 12, 161-195. ScHNIEDERS, B. R. (1987) The geology of sulfide-facies iron
JAMES, H.L. (1966) Chemistry of the iron-rich sedimentary formations and associated rocks in the Lower Steel River-­
rocks. U.S. geol. Surv. Prof. Pap . , 440 J, p. 61. Little Steel Lake area, Terrace Bay, Ontario.Unpublished
JENSEN , L.S. (1 978a) Geology of Thackeray, Elliott, MSc thesis, Lakehead University, Thunder Bay,
Tannahill and D okis Townships, District of Cochrane, Ontario, 196 pp.
Ontario. Ont. geol. Surv. Rep . , 165, 71 pp. SHULSKI, T. , WARES, R.P. & SMITH, A. D. (1993) Early
JENSEN, L.S. (1978b) Geology of Stoughton and Marriott Proterozoic (1.88-1.87 Ga) tholeiitic magmatism in the
Townships District of Cochrane. Ont. geol. Surv. Rep . , New Quebec orogen. Can. I. Earth Sci. , 30, 1505-1520.
173, 73 pp. SIEVER, R. (1957) Silica budget in the sedimentary cycle.
JENSEN , L.S. (1981) A petrogenic model for the A rchean Am. Mineral. , 42, 821 - 841.
Abitibi Belt in the Kirkland Lake area, Ontario . SIMONSON , B.M. (1985) Sedimentological constraints on
PhD thesis, University of Saskatchewan, Saskatoon, the origins of Precambrian iron formations. Geol. Soc.
Saskatchewan, 520 pp. Am. Bull. , 96, 244-252.
JENSEN, L.S. (1985) Stratigraphy and petrogenesis of STAN LEY, D.J. and WEAR, C.M. (1978) The 'mud-line': an
Archean metavolcanic sequences, southwestern Abitibi erosion-deposition boundary on the upper continental
subprovince, Ontario. In: Evolution of A rchean Supra­ slope. Mar. Ceo!. , 28, 19-29.
crustal Sequences (Eds Ayres, L. D., Thurston, P.C., STOw, D.A.V. & SHANMUGAM, G. (1980) Sequences and
Card, K. D. & Weber, W.), Geol. Assoc. Can. Spec. structures in fine-grained turbidites; comparison of recent
Pap. , 28, 65-87. deep-sea and ancient flysch sediments. Sediment Geol. ,
KRONBERG, B.l. & FRALICK, P.W. (1992) Geochemical 25, 23-42.
alteration of felsic Archean rocks by Gunflint Formation­ TEAL, P.R. & WALKER, R.G. (1977) Stratigraphy and
derived fluids, Quetico-Superior region, northwest sedimentology of the Archean Manitou Group, north­
Ontario. Can. I. Earth Sci. , 29, 2610-2616. western Ontario. Geol. Surv. Can. Pap . , 77-1A ,
LABERGE, G.L., ROBBINS, E.l. & HAN , T.-M. (1987) A 181-184.
model for the biological precipitation of Precambrian VAN H 1 s E , C.R. & LEITH, C.K. (1911) The geology of the
iron-fonnations - a: geological evidence. In: Precambrian Lake Superior region. U.S. geol. Surv. Monogr. , 52,
Iron-formations (Eds Appel, P.W . U . & LaBerge, G.L.), 641 pp.
pp. 69-96. Theophrastus Publications, S.A., Athens. WHITE, D.A. (1954) Stratigraphy and structure of the
LABERGE, G.L., 0JAKANGAS , R.W. & KICHT, K. (1992) Mesabi range, Minnesota. Minn. geol. Surv. Bull. , 38,
Early Proterozoic and Archean rocks of the Gogebic 92.
Range. Institute of Lake Superior Geology, 38th Annual WARDLE, R . J . & BAILEY, D . G. (1981) Early Proterozoic
Meeting, Part 2, Fieldtrip Guidebook. Hurley, Wisconsin. sequences in Labrador. In: Proterozoic Basins in Canada
LAGALLAJS, C.J. & LAVOIE , S. (1982) Basin evolution of (Ed. Campbell, F.H.A.), Geol. Surv. Can. Pap.,
the Lower Proterozoic Kaniapiskau Supergroup, central Ottawa, 81-10, 33 1-358.
Labrador miogeocline (trough), Quebec. Bull. Can. WARES, R.P. & GouTIER, J . (1990) Deformational style in
petrol. Geol. , 30, 150-166. the foreland of the northern New Quebec Orogen.
LouGHEED, M.S. (1983) Origin of Precambrian iron Geosci. Can. , 17, 244-249.
formations in the Lake Superior region. Geol. Soc. Am. WILKS, M.E. & NISBET, E.G. (1988) Stratigraphy of the
Bull. , 94, 325-340. Steep Rock Group, northwest Ontario: a maj or Archean
MEYN, H.D. & PALONEN , P.A. (1980) Stratigraphy of an unconformity and Archean stromatolites. Can. J. Earth
Archean submarine fan. Precam. Res. , 12, 257-285. Sci. , 25, 370-391.
MOREY, G.B. (1983) Animikie Basin, Lake Superior region, WooD, J. (1980) Epiclastic sedimentation and stratigraphy
U.S.A. In : Iron-formation: Facts and Problems (Eds in the N orth Spirit Lake and Rainy Lake areas: a
Trendell, A.F. & Morris, R.C.) pp. 13-68. Elsevier, comparison. Precam. Res. , 12, 227-255.
New Y ork.
Spec. Pubis int. Ass. Sediment. ( 1995) 22, 157-193

Facies models in volcanic terrains:

time's arrow versus time's cycle

G EO F F R E Y J . O RTON
Department of Geology, McMaster University,
Hamilton, Ontario L8S 4Ml, Canada

ABSTRACT

Vertical and lateral facies variations in volcanic terrains are abrupt owing to the sudden input of
additional sediment from one or more volcanic vents. Further complications arise from the influence of
volcanic activity on the subsidence history of the basin. A detailed study of one small depositional
system from the Ordovician Llewelyn Volcanic Group near Tryfan Fach, North Wales is used to
illustrate the problems this causes for developing facies models.
The Tryfan Fach Member is made up of four facies associations characterized by their geometries,
composition, and sedimentary structures. Each is assigned to a particular depositional setting: marine
shelf, braided stream, floodbasin and alluvial fan. The basal portion comprises a comparatively thick ,
mudstone-dominated succession deposited in a quiet-water marine setting largely below fair-weather
wave base. In contrast, coarse-grained, rhyolite-bearing sandstones were deposited in shallow, flood­
prone, southward flowing bedload-dominated braided streams and as unconfined sheet floods. These
sandstones amalgamate to form a laterally extensive sheet c. 15m thick and at least 2 km wide, which lies
with sharp contact on subjacent marine mudstones. Sandstones pass gradationally upwards into
interbedded coarse and fine sandstones, laminated vitric siltstones with accretionary lapilli, and
mudstones, deposited on a low-gradient coastal plain. Conglomeratic alluvial fan deposits, derived from
older deposits of granite, tuffaceous sediment and rhyolite, occur along the northeast margin of the
basin.
Although several features can be explained by envisaging the whole succession as the product of one
linked depositional system , the differences in sediment composition and palaeocurrent trends raise
problems. These cast doubt on the strict application of Walther's Law to the total succession, and
demand at least three genetically unrelated depositional systems. The rhyolitic braidplain to floodbasin
succession is attributed to subaerial aggradation of primary and reworked pyroclastics, and had a
different source from subjacent marine mudstones and the adjacent alluvial fan. The sharp basal
contact of the braidplain sandstones is interpreted as an erosional unconformity, and is inferred to have
resulted partly from volcano-tectonic uplift in advance of the volcanic eruption. The contemporaneous
progradation of an epiclastic alluvial fan from the opposite side of the basin is related to accelerated but
more differential basin subsidence during volcanism. Within the syneruptive deposits it could not be
established whether facies change reflected sorting processes on the alluvial plain or progressive
eruption of finer grained ash. As a consequence, vertical sequences are difficult to interpret, palaeo­
geographical maps are speculative, and the timing of relative sea-level changes could not be assessed
fully.

INTRODUCTION

Vertical-sequence analysis, essentially an extension environments through time from two-dimensional

of Walther's Law, is probably the single most sequences (de Raaf et al. , 1965 ; Visher, 1965). As
important advance in sedimentology in recent Walther stressed, the law can be applied only to
decades (Dott, 1983) . It provides a powerful tool successions with gradational , non-erosive contacts
with which to interpret changes in sedimentary between facies and/or environments (see Middleton,

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0

157
158 G.J. Orton

1973). Also implicit in its application is the assump­ concepts, labelled by Gould ( 1987) as time's arrow
tion that the control on changes in grain size and/or (a linear succession of unique events) and time's
facies largely lies within the area of deposition. cycle (recurrent patterns in a world that remains
External parameters (e.g. sediment supply, relative essentially unchanged), embody the major issues
sea-level) change gradually rather than abruptly; within analysis of every volcaniclastic succession .
continuity rather than discontinuity is assumed. A small-scale but well-exposed sequence from the
Although modern facies models do allow for 'cata­ Ordovician Llewelyn Volcanic Group near Tryfan
strophic' events (e.g. storms) , these catastrophes Fach, North Wales is used to illustrate the differences
are assumed to recur with the same frequency in and interaction between these two concepts. At this
the past, present and future . And the thickness of locality, a 15-m-thick sheet of coarse-grained rhyo­
sedimentary deposits resulting from catastrophes litic volcaniclastics is bounded by marine mudstones
is normally not sufficient to alter radically the or fine-grained sandstones, and lies a few kilometres
pre-existing geography and patterns of sediment away from conglomeratic alluvial fan deposits. The
dispersal. main problem addressed in this paper is whether
Things are not so simple in volcanic terrains. B y the rhyolitic sandstones are reworked pyroclastics
their very nature, explosive volcanic eruptions are and indicate penecontemporaneous volcanism, and
catastrophic and rapidly change the supply of sedi­ whether the other stratigraphical units had the same
ment available within a catchment area. The grain source. Through analysis of sediment composition ,
size and type (crystals, lithics, glass shards) of sedi­ it can be shown that there is no genetic relationship
ment supplied can change with equal rapidity during between vertically and laterally adjacent strati­
the course of a volcanic eruption owing to changes in graphical units. Each represents a separate depo­
the rate of magma extrusion, volatile content of sitional system. The implications for facies modelling
magma erupted or the amount of interaction with and reconstructing environments in volcanic settings
external water in the vent (see Heiken & Wohletz, are discussed.
199 1 ; White, 199 1 ) . During eruptions, patterns of
sediment dispersal are not always controlled by pre­
existing topography; ash-fall deposits in particular TERMINOLOGY
are controlled by wind patterns and are distributed
over broad areas independent of topography and , to Sands rich in volcanic debris are of two dominant
some extent, gravity. Depositional basins around types according to the method of clast formation .
volcanoes are often infilled from multiple sources. Pyroclastic material (shards, pumice, crystals) is
The growth of intrabasinal volcanoes can form generated by explosive volcanism , and is often
physiographic barriers that isolate depositional contributed directly to the sedimentary record as
systems and limit the fetch (and hence size) of waves primary pyroclastic deposits. In contrast, epiclastic
reaching coastlines. Relative sea-level can change volcaniclastic materials result from erosion and
quickly owing to uplift or subsidence resulting from weathering of older volcanic rocks, including lithified
a volcanic eruption. tuffs, and do not usually reflect contemporaneous
As a consequence, ancient volcaniclastic suc­ volcanicity. A 'grey area' within this classification
cessions are notoriously difficult to interpret. Two concerns reworking of unconsolidated pyroclastic
concepts jockey for recognition in every ancient debris. Many authors (particularly Fisher, 196 1 ;
succession: (i) each bed in the succession represents Fisher & Schmincke, 1984) regard recycled pyro­
a unique historical event (i.e. a new eruption, a new clastic material, redeposited by wind or water as
type of eruption) with no genetic relation to the bed secondary pyroclastics; others (e.g. Cas & Wright,
before; (ii) the succession represents a 'group of 1987) suggest that pyroclastic particles reworked by
facies genetically related to one another and which water or other agents should be called epiclastic.
have some environmental significance' (Collinson, In older and more deformed successions, dis­
1969). The first perspective relates facies change tinction becomes even fuzzier, and one is often
to changing source parameters. Sedimentology lucky if one can demonstrate that some particles
essentially loses its predictive capability and ability were produced by explosive volcanism. A further
to define environments. The second is an ahistorical complication is that many volcaniclastic deposits
perspective in which laws, processes and the pace consist of mixtures of both syneruptive pyroclasts
of change remains predictable . I think these two and epiclastic material (sensu Fisher & Schmincke ,
 Facies models in volcanic terrains 159

1984) . The recognition of pyroclastic debris where mined configuration of sub-basins, rates of fault­
such material is diluted by and mingled with sands of block subsidence, the distribution of volcanic and
other origins is not easy. Resolving this problem , sedimentary units, and depositional settings. The
however, is far from an exercise in semantics but is Snowdon graben can be divided into at least two
crucial in determining the nature of the volcanic structural basins, 10- 15 km wide, about Llanberis
contribution to sedimentation. Calling all pyroclastic Pass (Fig. 1) based on the distinctive petrography
material that has suffered additional transport of their contained sediment (Orton, 1990) and the
'epiclastic' hides the fact that volcanicity may have distribution of basalt intrusions (Campbell et a!. ,
been active at that time, and breaks the genetic 1988) . The basin northeast of Llanberis Pass has been
lineage between deposits and their source. referred to as the Tryfan depocentre (Orton, 1990) .
In this study the term pyroclastic is used where the Volcanism, and the deposition of coarse clastic
composition, texture, variability, and geometry of material, varied in time and space. It can be divided
the deposits all suggest that clasts were produced by into two eruptive cycles that are reflected in two
a contemporaneous volcanic eruption. Pyroclastic volcanic groups (Howells et at. , 199 1 ) : a lower
material that was retransported by 'normal' hydro­ Llewelyn Volcanic Group in northeast Snowdonia
logical processes shortly after its initial deposition is (Fig. 1) and an upper Snowdown Volcanic Group in
referred to as reworked pyroclastics in order to southwest Snowdonia. Although this activity was
emphasize that sedimentation was still occurring in confined to just two chronostratigraphical stages
response to volcanism. (Soudleyan and Longvillian) , about 5% of total
(72 Ma) Ordovician time , its products comprise
about half the total thickness of the Ordovician
GEOLOGICAL FRAMEWORK sequence. The average rate of sediment accumu­
lation during the Soudleyan and Longvillian
During the latter part of the Ordovician, North approached 1 m lOOO yr-1 (Orton, 199 1 ) .
Wales was part of an extensional or transtensional The earliest volcanic activity in the Llewelyn
marginal basin (Campbell et al. , 1988; Kokelaar, Volcanic Group developed from at least four, partly
1988) sited on continental crust comprising accreted contemporaneous , centres. Rhyolite lavas and silicic
volcanic arcs (Thorpe , 1979). The Precambrian crust ash flow tuffs (Conwy Rhyolite, Braich tu Du
formed part of a small microcontinent, Eastern Formation) , trachyandesite lavas and tuffs (Foel
Avalonia (Soper & Hutton, 1984) , derived from Fras Formation) and basaltic-andesite lavas (Foel
continental Gondwanaland, and separated from the Grach Basalts) were all erupted (Fig. 1). Extrusive
North American continent (Laurentia) and Baltica rocks are associated mainly with marine mudstones
by the Iapetus ocean and the Tornquist's sea. Palaeo­ and show little evidence of reworking; they were
magnetic and faunal reconstructions record the probably ponded in subsiding, fault-bounded areas
northward movement of Avalonia throughout Early of the sea-floor (Howells et al. , 199 1 ) . However, the
Ordovician to Early Devonian times (Cocks & Fortey, latest activity in the Llewelyn Volcanic Group, the
1982; van der Voo, 1983, 1988) , with temperate Capel Curig Volcanic Formation, was dominated by
southerly latitudes (c. 30-35°S) indicated for mid­ larger scale subaerial eruptions of silicic magma,
to late Ordovician (Caradoc) times (Torsvik & giving rise to widespread ash flow tuff deposits.
Trench, 1991). Alluvial fan facies suggest a humid Volcaniclastic successions underlying the Capel
climate (Orton, 199 1 ) . Curig Volcanic Formation, herein referred to as the
Ordovician sedimentation i n North Wales was Tryfan Formation (Figs 1 & 2), reflect a compara­
dominated by muddy facies. Volcanism and depo­ tively reduced amount of volcanic activity. The
sition of associated coarse-grained clastic material Tryfan Formation can be divided into at least eight
was confined largely to the Caradoc series, and members (Fig. 2), based on differences in sedi­
confined to a NE-SW oriented rift termed the mentary facies and/or environments, and the
Snowdon graben. An orthogonal array of deep­ amount, petrography, chemical composition and
seated fractures cutting the ensialic basement location of any volcanicity associated with sedimen­
(Campbell et al . , 1988; Kokelaar, 1988) controlled tation. Each member represents a distinct depo­
the position of eruptive centres within the graben. sitional episode with no genetic relation to the
Reactivation of these and shallower fractures as episode before. The Gwern GofTuff, and underlying
faults, often in conjunction with volcanism , deter- fluvial- deltaic deposits were derived from a source
160 G.J. Orton

� CapeiCt.rig
�"J Volcanic Formation
(;:::;::::::1 Tryfan Formation
r:Mi:!;(!Jil Foel Fras Volcaric CC>I"\l)lex
- Foel Grach Basalt Fm.
� &aichTuDu
� Volcanic Formation
� Coowy Rhyoite Formation
�
Stbvolcaric ntrusions
o- -skm
Important faults

Fig. 1. Llewelyn Volcanic Group: distribution of formations and related subvolcanic intrusions. Box locates Fig. 2.
Intrusions in the south identified as follows: MP, Mynydd Perfedd; BC, Bwlch y Cywion; T, Talgau; CL, Carnedd Llewelyn
(modified from Howells et al. , 1991). On inset map, shaded area denotes approximate position of Snowdon graben, MS
refers to the Menai Straits.

to the east-southeast (Orton, 1988) . Progradation Fach Member (Figs 2 & 3 ) . It varies in thickness
and/or transgression of coarse-grained fluvio-deltaic from 30 to 120 m , and has been recognized only
systems was the dominant mode of basin infill above within the Tryfan anticline (Figs 1 & 2) , over an
the Gwern Gof Tuff. Each of these prograded toward extremely small area ( < 20 km2) . Exact information
the south and southeast, indicating a 'flip' in the on the time-scale spanned by the Tryfan Fach
polarity of infill of the Tryfan depocentre. The depo­ Member cannot be provided. However, it was of the
sitional package of concern here lies directly above order of a few hundred thousand years and certainly
the Gwern Gof Tuff and is referred to as the Tryfan not millions of years.
 Facies models in volcanic terrains 161

68

I
I

* fossillocalities

500 m 66

Fig. 2. Simplified geology of the Tryfan anticline (modified after British Geological Survey, 1985, unpublished 1:10000
sheets; Orton, 1990). Numbers on border refer to UK Ordinance Survey National Grid (grid square SH) and are 1 km
apart. Circled letters on map refer to location of the main sedimentary sections, whereas numbers on stratigraphical log
(and associated textures) denote the three basinal facies associations: 1, marine shelf mudstones; 2, Rhyolitic braided
stream sandstones; 3, floodbasin sandstones .

FACIES ASSOCIATIONS stone. One bed of coarse-grained lithic sandstone

containing rounded rhyolite pebbles (Fig. 4, 49 m)
The stratigraphical record of the Tryfan Fach also occurs. Rare brachiopods (Rostricellula) indicate
Member can be divided into four facies associations, a marine setting (Tunnicliff & Rushton, 1980).
each assigned to a particular depositional setting. Thicker horizons (to 2 m thick) of rhyolitic fallout
As cleavage obscures sedimentary facies in fine­ tuff (Fig. 3) consist of thin, 8 -1 0 cm thick, massive
grained lithologies (unless rhyolitic) , environmental to graded beds of vitric siltstone.
interpretations are based largely on data collected The succession generally becomes sandier upwards
from sedimentary rocks of medium silt grade and with bioturbated sandy siltstone giving way to inter­
coarser. stratified very fine sandstone (beds 10- 100 em) and
mudstone (beds to 20 cm) . Most sandstone beds are
sharp based, massive, sometimes normally graded,
Marine shelf mudstones
and rarely infill isolated-scours 30 em deep. Beds
The basal portion of the Tryfan Fach Member in the near the top of the succession , however, contain
south consists of poorly exposed mudstone with ripple cross-stratification, small-scale (to 15 em)
sharp-based massive or massive to horizontally trough cross-stratification , planar horizontal lami­
laminated beds (to 1 0 cm) of coarse siltstone and nation , and more rarely undulating lamination.
fine-grained sandstone. Upper surfaces of sandstone Vv'here trough cross-stratification occurs, sets are
beds are either sharp or grade into overlying mud- often separated by discontinuous siltstone lenses.
162 G.J. Orton

Facies associations

0">
en
�
;
100 Thinly bedded floodbasin
sandstones

u::

' " Rhyolitic braided stream

��
sandstones

a:
(!)

....
:2
I 60
u
<!
u..
z en
<! u::
u..
>-
a:
f-

40

Marine shelf m udstones

20

GWERN GO F TUFF

Fig. 3. Composite log through the Tryfan Fach Member (from the vicinity of log C, Fig. 2) showing stratigraphical
relationships between facies associations. Ruled vertical bars denote position of more detailed sedimentary logs. The
Gwern Gof Tuff (base of section) is about 40 m thick and only its top portion is depicted. The top of the braided stream
association forms the popular rock-climbing slab at Tryfan Fach.
 Facies models in volcanic terrains 163

Facies interpretation

65

Small scale trough and

ripple
? Shoreface
5cm
Cross stratification

Low-angle Wave swash on

planar lamination sand bar?

------ -- Fairweather wave base-??

Bioturbated siltstone

55

Alternation of
rapid deposition by
storms and
Ripple waning alluvial
cross­ floods and
stratification slow suspension
sedimentation

Pebbly sandstone

Graded
sandstone beds

45

1m

Fig. 4. Section from mid- to upper portion of the marine shelf mudstone association. Scale is in metres above Gwern Gof
Tuff. From section B, east of Tryfan Fach, SH 67205995. Log is continued in Fig. 5. See Fig. 3 for stratigraphical position
of log.
164 G.J. Orton

The top few metres of the association are again assigned to an 'outer shelf setting' (Fritz & Howells ,
fine-grained (Fig. 5 ) . Mudstones are slightly tuf­ 199 1 ) . Care must be taken, however. As noted
faceous and a laterally continuous comparatively by Pickerill & Hurst ( 1983), if suspension clouds
pure unit of vitric siltstone (50- 100 em thick) produced by storms or river floods contain , in
lies directly beneath the coarse-grained rhyolitic addition to sand and silt-grade material, sufficient
volcaniclastic material of the braided stream associ­ mud to provide cohesiveness, modification by later
ation (Figs 5 & 6) . Mudstones contain thin ( < 30 cm) currents could be precluded even though deposition
tabular sandstone beds, most of which display bio­ occurred above normal, even 'fair-weather' , wave
turbation. Although the sandstone beds do not base.
pinch and swell they could not be traced between The upper sandier part of the succession contains
sections only a few hundred metres apart. Indistinct unequivocal evidence for deposition in shallower
parallel lamination and more rarely trough cross­ water. Planar lamination (Fig. 4, c. 61 m) may reflect
stratification were the only sedimentary structures wave swash processes. However, a similar facies ,
observed. termed quasi-planar lamination, described from
Overall, facies and fossils indicate deposition in Lower Cretaceous lower shoreface to shelf deposits
a quiet-water muddy environment. Environmental of Montana has been attributed to deposition under
interpretation of Caradoc fauna is facilitated by single-event, high-energy combined-flow conditions
comparison within coeval strata in the nearby (Arnott, 1993 ) . Outcrop quality is not adequate to
(< 100 km) Berwyn Hills (Pickerill & Brenchley, discern whether lamination remains perfectly planar
1979). Here Rostricellula is associated with the or gently undulates, although the position of the
Macrocoelia subcommunity of a Dinorthis com­ lamination at the base of an upward-shoaling unit
munity and is considered to have colonized high­ lends some credence to a combined-flow origin .
energy non-turbid, well-oxygenated environments Undulatory lamination, considered to form under
on silty substrates under moderately high rates of intense oscillatory flow (Allen, 198 1 ) or combined
sedimentation. A maximum water depth of 25 m was flow (Myrow & Southard, 1991) occurs nearby at
assigned to the Macrocoelia subcommunity based about the same stratigraphical level and provides
on sedimentary structures, and the presence of the additional evidence for wave processes. The twice
fossil-boring Vermiforichnus (Pickerill , 1976) . observed, sandstone-filled scours are similar in size
The bottom part of the Tryfan Fach Member was and orientation (but not abundance!) to gutter casts
dominated by deposition of mud from suspension. described by Myrow ( 1992) from shallow subtidal
Coarser grained sediment was introduced period­ deposits in Newfoundland. In his model the subtidal
ically in decelerating flows. These could represent zone , dominated by fine-grained sediment, is largely
suspension clouds deposited during the waning stages a zone of sediment bypass in which high-velocity
of a storm (Nelson, 1982) or underflows (hyper­ sediment-laden flows erode shore-normal scours
pycnal effluent) from river-derived flood events. preserved as gutter casts.
Although underflows are documented most com­ The trough cross-stratification of the sandiest
monly where river waters enter freshwater lakes, deposits (Fig. 4, 63-66 m) is thought to be produced
they also occur when bedload-dominated streams by waves and/or river mouth floods. The absence
enter marine basins. Wright et al. (1986) confirmed of facies such as lenticular and flaser bedding,
that hyperpycnal underflows carrying silt occur off reactivation surfaces, rhythmic bedding and mud
the mouth of the Huanghe (Yellow) River. Under­ drapes suggest that tidal processes were absent or
flows carrying sand develop when floodflows from unimportant. The small size of the cross-stratification,
the San Lorenzo River (California) scour the river inter-bedding of mudstone, and absence of swaley
mouth and pass offshore as a plane jet (Hicks & or hummocky cross-stratification indicates low near­
Inman, 1987) and are suspected also to be a common shore wave power. This requires further comment.
phenomena on fan-deltas along fjords (Prior et al. , Wave power at a coastline depends on the maximum
1987 ) . deep-water wave energy and its shallow-water fric­
I f storm o r river-introduced silts and sands d o not tional attenuation, which is a function of the sub­
exhibit evidence of reworking by waves, it is usually aqueous slope. Younger shorefaces of the Tryfan
inferred that beds reflect deposition below effective Formation commonly contain abundant hummocky
wave base. A comparable facies described from and swaley cross-stratification, undulating lami­
other Ordovician successions in North Wales was nation, and large wave ripples (e.g. Orton, 1988;
 .
c B
E
u:;:;:;;;;:::s.;;;- , 0
"'i>=
o --..
0

o-
, •
FLOODBASIN

� ri o
D •

1-"',
8-------
500 m o-- ---- 400 m
�

l2
0

400 m
o

"'
118

ri®�
- o--
------1

�
�)�
:·�
.. '
-� --8- --- ----
------ sil
- -ts
---t...
one
... . �-;·
�
marker -..!!!!..�
.
--...i'.J-·
.. .... ..
"'J"=
("')
�·
;:s

n·v
�
_.,._.
0
� 0
0 Channel fill BRAIDED �
�· Channel fill
STREAMS 1:.:;-
,.._
� �� n=ll
5'
"'
120 0
0 128
0
2
;:,
c:;·
0 <;;
KEY
I
.----
�
5'
Bioturbation
Indistinct stratification "'

MARINE MUDSTONES
I�
Graded sandstone
Ripple cross-strat
Horizontal lamination
I
I �.
v\
" v v
,
--- -- .......
. . Trough cross-strat
. ..,
Ashfall tuff
Bedded sandstone

�>
Mudstone
0 3-10
0 11-20
0 21-30 SET SIZE
0 >30 INCM

Fig. 5. Detail of facies variation in upper part of Tryfan Fach Member on west limb of Tryfan anticline. See Fig. 2 for location of sections. Section B is a continuation
(without a stratigraphical gap) of section B depicted in Fig. 4.

......
0\
Vl
166 G.J. Orton

Fig. 6. Contact between (A) fine­

grained bioturbated sandstones of
the marine shelf association
(beneath hammer) and (B) trough
cross-stratified coarse-grained
volcaniclastic sandstones of the
braided stream association (above
hammer), which often contain a thin
(30-50cm) bed of siliceous
mudstone. General fluvial
palaeocurrent is to the left and
towards the reader. Hammer is
30cm long.

Fritz, 199 1 ) . As it seems unlikely that deep-water sandstone. Fine-grained intervals are laterally
wave energy changed dramatically over a few continuous for distances of about 200 m, with the
hundred thousand years, the low wave power exception of a tuffaceous, green-coloured siltstone
documented throughout the Tryfan Fach Member is that extends for at least 900 m (Fig. 5). Sandstone
related to a low subaqueous slope. At the time sheets are characterized by extremely abrupt vertical
of emplacement of the Gwern Gof Tuf (Fig. 3) , changes in grain size and texture . Vertical-sided ,
the offshore slope must have dipped to the west­ laterally stepped erosional scours up to 30 em deep
northwest. In contrast, the rhyolitic braidplain sometimes occur along their base. In thicker sheets,
deposits of the Tryfan Fach Member indicate a slope shallow trough cross-stratification is the dominant
to the east-southeast. Sometime during the inter­ sedimentary structure , with rare climbing ripple
vening time period, that is when the marine shelf cross-lamination and horizontal lamination. Palaeo­
association was being deposited, the slope must current distribution is unimodal towards the
have rotated, either tectonically or by deposition, southeast. Many thick sheets fine upwards , from
through the horizontal. Low-gradient muddy shelves conglomeratic coarse sandstone to siltstone (Fig. 8).
are highly efficient at attenuating wave energy When this occurs the set size of cross-stratification
(cf. Wells & Coleman, 198 1 ). On the Surinam coast­ decreases upwards ( c. 20 cm to Scm) in conjunction
line the total wave-energy loss through slow shoaling with grain-size changes. Thinner sandstone beds
across the inner shelf ranges from 9 3 to 9 6% , and display a wider range of sedimentary structures.
most waves do not reach the shoreline nor break. Coarse-grained beds are commonly normally graded ,
fining up from conglomeratic sandstone to siltstone
over short (< 50 em) vertical distances. Finer grained
Rhyolitic braided-stream sandstones
beds are massive, have massive bases with cross­
A multistorey, 12- 15 m thick , laterally extensive stratified tops, horizontal lamination, or undulating
(about 2 km) horizon of poorly sorted granule lamination.
conglomerate to fine sandstone directly overlies The geometry of the above sandstone packages ,
the muddy coastal sediments (Figs 3, 5, 6 & 7). their internal structures, coarse texture and uni­
Tuffaceous siltstone rip-up clasts (to 5 em) and sub­ modal palaeocurrents indicate deposition by uni­
angular rhyolitic pebbles (to 2 cm) are common . The directional tractional currents under a broad range
sandstones were deposited as broad tabular sheets of flow strengths. A subaerial setting is assumed
20-250 cm thick separated by thin ( < 20 cm) beds because: (i) there is no evidence (e.g. wave ripples)
of vitric siltstone or interbedded siltstone and fine within fine-grained beds for marine processes ,
 Facies models in volcanic terrains 167

Facies interpretation

Scm
Stacked
'channel' fills
5cm Small scale
trough

Cross-stratification

Aggradation in
shallow
Low-angle unconfined
....
�:..:...:_ ,_ �,�·
braided stream
" indistinct
A /

� � ··<-'"·�-
/ stratification
..,
io
-.--- , 10cm

Sea-green siltstone
z horizontal lamination
0
and rippling
t=
<t
u lnfill of
0
(/) deeper
(/)
<t fluvial
channel
�
<t 20cm Shallow trough
U.J
a: cross-stratification
f- ;...;,.:...��·
(/) 1 5 cm lnfill of
0 minor
U.J 30cm
0 'channels'
<( 10cm
a:
a:l
'Overbank'
floods
Graded
sandstone beds

Undulating
lamination

Upper flow regime

Horizontal
sheetfloods
lamination
with small
scours

Fig. 7. Complete section through

braided stream facies association.
Rhyolitic siltstone Ashfall tuff
From Section D, east slope of
Tryfan, SH 6683 5945. See Figs 3 &
5 for stratigraphical position of log. Bedded sandstone

Numbers to right of log denote

thickness of cross-strata. In thin
beds that fine upwards the set size
indicated refers only to sets at the
base of the beds. M Si Sa G

(ii) there is no bioturbation as occurs within other The thicker sandstones are best interpreted as
facies associations, (iii) sandstones are extremely infill of fluvial , probably braided, channels (Miall,
poorly sorted, (iv) tidal indicators (e.g. mudstone 1977). In relation to modern braided rivers, these
drapes) are absent. sandstones resemble to some extent the South
168 G.J. Orton

(Campbell, 1976; Godin, 199 1 ) , or Pennsylvanian

South Bar (Rust & Gibling, 1990a) and Boss Point
(Browne & Plint, 1994) Formations are similarly
dominated by trough cross-stratification. The
absence of planar cross-stratification has been
explained in at least two ways. Some suggest that
lateral migration of deeper channels across a braided
sand-flat would lead to the preferrential preservation
of three-dimensional dunes at the expense of cross­
channel bars characterizing shallower channel por­
tions (Campbell, 1976). Alternatively, channel
avulsion, causing rapid abandonment of the pre­
viously active tract, combined with flood-stage
erosion of top-stratum would prevent development
of fining upward waning channel fill sequences
and result in a sheet-like sandbody dominated by
trough cross-strata (Rust & Gibling, 1990a; Browne
& Plint, 1994) .
Several important observations suggest a differ­
ent interpretation for the trough cross-stratified
'channel' fill units of the Tryfan Fach Member.
1 There is little evidence, such as channel forms,
deep scours or channel base lags, that bedforms
were confined within deep channels. Instead, fluvial
cycles rest on virtually planar erosion surfaces
or wedge out at their base against broad, shallow
channels.
2 Although sediments are poorly sorted, cosets often
fine upwards with a corresponding decrease in set­
Fig. 8. Channel-fill sequence from a 2-m-deep pebbly size (Figs 7 & 8) . In contrast , no clear grain size
braided stream (braided stream association). Base of
sequence is just beneath the field of view. Height of cross­
trends in the facies cycles are present within the
stratification decreases upwards, from 20cm to 4cm. Flow Boss Point Formation and co-sets are inferred
of fluvial current was towards the reader. to represent partial-channel fills (Browne & Plint,
1994). Although trough cross-sets in the Boss Point
range from 50 em to 120 em in thickness, vertical
variation in set-size is not reported.
Saskatchewan River (Cant & Walker, 1978) in being 3 The set size is small enough that sandstone sheets
dominated by trough cross-stratification. The South could represent complete 'channel' fills. A broad
Saskatchewan maintains average channel depths of relationship between bedform height (H) and flow
3m (maximum 5 m) , in which trough cross-stratified depth (h) in rivers has been established (e.g. Allen,
sands accumulate by migration of sinuous crested 1984, fig. 8-20). Preserved set thickness will be
dunes. Fining, upward sequences are generated as considerably less than the height of dunes originally
the channel shallows due to lateral migration. As present at flood stage because of relaxation during
this occurs, the trough sets are overlain and/or falling stage (Allen, 1 974) and truncation by sub­
interbedded with stacked sets of planar cross­ sequently migrating scours. Consequently, Rust &
stratified sand formed by cross-channel migration of Gibling (1990a) suggest adding 50% as a conservative
shallow bar complexes. estimate of the amount lost. This approach gives a
Unlike the South Saskatchewan, the Tryfan Fach flood-stage dune height of about 30 em for the largest
Member contains no planar cross-stratification. dunes at the base of cosets and a few centimetres
Although puzzling this is not unique and numerous for bedforms at their tops. The mean relationship
other ancient braided stream successions, such (H!h 0. 15) of Allen ( 1984) yields a flood-stage
=

as the Jurassic Westwater Canyon Formation flow depth of about 2 m for the large basal dunes, a
 Facies models in volcanic terrains 169

depth comparable to the thickness of the cosets. deposits could occur on the floodplain or within
4 The broad, extremely low-angle, and shallow channel belts at times of falling stage when water
nature of the trough cross-stratification (Fig. 8) depths were locally very shallow.
indicates either extremely rapid flows and/or shallow Facies of modern , shallow, sand-bed braided rivers
flow depths (Jopling, 1965 ) . There are similarities differ considerably according to discharge and dis­
with proglacial sandur deposits (e.g. Ruegg, 197 1 } . charge variability, depth of flood waters, and grain­
For all o f the above reasons, there seems iittle size of sediment transported . Large perennial ,
doubt that most of the fining upward trough cross­ comparatively sluggish rivers with a finer sediment
stratified sandstone sheets record virtually uncon­ load are usually characterized by migration of large
fined flood events. The thickness and type of lobate linguoid bars , which deposit superimposed
sequence resulting from such flood events would suites of planar-tabular cross-bedding. Examples
depend on the depth and duration of the flood, and include the Platte River of Nebraska (Crowley,
the amount and grain size of sediment available. 198 3) or outwash fans in Iceland (Boothroyd
Three possibilities are considered . The thickest & Nummedal, 1978) . In contrast, with smaller,
sheets require that a range of grain sizes are available ephemeral, flood-prone streams with considerable
and that flow-depths change in a comparatively short-term variability in discharge there is either
gradual fashion so that bedforms have time to insufficient time or insufficient depths of water for
respond to the decreasing flow strengths. These large linguoid bars to develop. This can be again
cosets represent the most prolonged (and most illustrated by outwash fans in Iceland. Catastrophic
widespread} discharges, and may have involved floods caused by subglacial volcanic eruptions
some channel incision or migration. Thinner sand­ (jokulhlaups) have discharges many orders of
stone sheets where trough cross-stratification grades magnitude larger than the normal summer discharges
rapidly into ripple cross-stratification would develop (Maizels, 1989) . Sandy deposits that result are
where floods waned extremely rapidly, whereas thin structureless, horizontally laminated, or trough
incomplete (top missing) successions could result if cross-stratified and lack stacked sets of planar cross­
flood waters simply ran out of sediment to deposit. stratified sand formed during normal discharges.
Although longitudinal sections are rarely pre­ Sinuous-crested megaripples (St) were also the most
sented, some of the cross-strata (termed low-angle common bedform in channelized stream flood in
indistinct stratification on logs) appears similar central Australia (Williams, 197 1 ) .
to scour-fill bedding described from deposition of I n summary, I conclude that Tryfan Fach Member
pebbly sand sheets under high-discharge , shallow 'rivers' varied from a few centimetres to perhaps 2 m
flow conditions (cf. Smith , 1986, 1987, 1988) . Sand­ in depth, and were virtually unconfined. Sedi­
stone sheets with this type of cross-strata are mentation proceeded through vertical aggradation
extremely poorly sorted, rarely fine upwards, and beneath shallow ephemeral sheet-floods rather than
set size is typically less than 10 em (cf. Fig. 7, top of by incision, lateral migration, and avulsion of
log). These features suggest that an abundant sedi­ more perennial channels. The term 'sheet braide d'
ment supply 'forced' rapid aggradation of low relief (Cotter, 1978) is an apt description of the type of
sinuous-crested dunes on the river bed. 'river' envisaged for the Tryfan Fach Member. This
The thin sandstone beds without cross-stratifi­ fluvial style is not common at the present day,
cation contain sedimentary structures indicating except in arid areas where ephemeral runoff forms
high-velocity shallow-water sheetflood conditions. a network of shallow, interlacing poorly defined
The massive to stratified facies, for instance, channels. Ancient examples that are dominated by
resembles hyperconcentrated flood flow deposits trough cross-stratification include the Devonian Peel
described from Holocene and ancient alluvial Sound Formation (Miall & Gibling, 1978) and
sequences of both volcanic (Smith, 1986; Maizels, perhaps distal successions of the Miocene Ellensburg
1989) and non-volcanic (Simpson & Eriksson, 1989} Formation (Smith, 1988).
settings. Beds containing horizontal lamination
would represent upper flow regime plane-bed con­
Thinly bedded floodbasin sandstones
ditions. Following Rust & Gibling (199Gb) the beds
with undulating lamination can be interpreted as A finer grained interval of interbedded sandstones
vertically aggrading antidunes that may have formed and mudstones gradationally overlies the cross­
in 10 -15 em of water flowing at 1 m s- 1 . These stratified rhyolitic sandstones sheet (Fig. 9). Beds
170 G.J. Orton

Facies interpretation
Massive-sandstone
bedded Debris flow
��Rhizocorallium siwilnlttteshrtloaccranemineandattioenarydsandstlapionel i Reworashfalkedl
. �: -..---1
�I Quisuspensi
et-watoenr
" v

""I
tsiLamiulfftsatceous
nated
one sedimentation
" ...,...
I
::·:'/}';;-.':'·:."?:·'�?1i 1r
!;:.�:.:.:.t.r:-:.�"..':.Y-�·;);;.� V
Bioturvbeatsandst
massi ed ones
-------

V-
-··- ·· :z-x7) 15cm Small scale Minorchannel
fluviasl
tstrough
ratificcratoioss­n
-
lHoramiiznontatioanl
�

��'�··
Skolithos
Horlamiiznontatioanl Sheetshalfllooodsw watinteor
I and bioturbation
--'lJ0==-1 lr
....:. �'<yl

---=-J
· 10cm
Fig. 9. Representative section from
Fluvial channel floodbasin facies association. From
1mI- Amalsheetgamatfloodsed section C, just south of wire fence
_-
___

-
wisiBedded
tltshtonesandst one
lenticles
on east slopes of Tryfan, SH 6696
5960. See Fig. 3 for stratigraphical
position of log and Fig. 2 for
M Si Sa 1 G location of section.

are sheet-like on outcrop scale with even thin ( 10 em) beds. The intervening siliceous siltstones display
beds traceable for 100 m. Sandstone units vary sub­ horizontal or undulating lamination and small wave
stantially in composition and thickness. Very coarse ripples (Fig. lOA). Lamination in fine-grained beds
sandstones (beds 20-50 cm thick) are either massive often contains water-escape structures or is con­
(Fig. lOA) or contain small-scale (lO cm) trough voluted. Subhorizontal burrows in the top of
cross-stratification, which yields palaeocurrents the succession (Fig. lOB ) (Cruziana ichnofacies:
toward the southeast. Coarse-grained lithologies Rhizocorallium) indicate a low-energy depositional
sometimes load into underlying mudstones (Fig. environment, but cannot be used to infer water
lOA). Climbing ripple cross-lamination and horizon­ depths or salinity because Rhizocorallium has been
tal lamination are the domjnant lithofacies within described from coastal lagoon and freshwater set­
the finer grained and thinner ( 1 0- 15 em) sandstone tings (Fiirsich & Mayr, 1 98 1 ) .
 Facies models in volcanic terrains 171

Fig. 10. Facies o f floodbasin facies

association. Photograph is located
11m above base of Fig. 9.
(A) Climbing? wave ripples,
undulose and horizontal lainination
within siliceous siltstones. Deposits
lie about 85 em stratigraphically
above position where photograph B
was taken. Massive beds (M) may
be bioturbated. Coarse-grained
crystal-lithic sandstones at the top
of the photograph represent the
base of a thin (50em) debris flow
deposit. Lens cap is 6 em in
diameter. (B) Horizontally
laminated tine-grained vitroclastic
(i.e. predominantly devitritied glass
shards) tine-grained sandstones and
siltstones. Notice lack of grading or
organization to sedimentary units.
Letters denote: R, subhorizontal
Rhizocorallium burrows;
W, vertically-aggrading wave
ripples. Coin is 3.5 em in diameter.

Facies indicate a quiet subaqueous environment. ically interrupted by input of coarser sediment.
They are termed floodbasin deposits because of Fine-grained horizontally laminated or rippled sand­
their fine grain-size , because there is little evidence stones, and coarse-grained massive sandstones
(e.g. desiccation cracks, palaeosols) for prolonged probably represent subaqueous high concentration
subaerial exposure, and because they appear to have sheet flows (cf. Dam & Andreasan, 1990). The thin
accumulated simultaneously across a basin several coarse-grained sandstone beds with trough cross­
kilometres wide. Rare occurrences of accretionary stratification are not unlike beds from the previously
lapilli within siltstones provide evidence for some described braided-stream association, which were
penecontemporaneous rhyolitic volcanicity. The attributed to waning alluvial floods and are inferred
evenly laminated siltstones indicate episodic sedi­ to represent a similar phenomena. Whether the
mentation out of suspension. Bioturbation and wave floodwaters developed on an alluvial plain cr
rippling (Fig. lOB) within siltstone units indicate that whether beds represent the subaqueous extension of
they represent an amalgamation of several smaller the flood events is a moot point.
events. These quiet-water conditions were period- Plausible settings for the floodbasin association
172 G.J. Orton

include: (i) an interdistributary bay, (ii) lagoons flows figured prominently on a nearby alluvial plain,
along the delta front behind barrier islands, the alluvial input was denser than basinal waters
(iii) lakes on a floodplain adjacent to an active (hyperpycnal flow) allowing flood waters to pass
alluvial tract and (iv) a transgressed (or flooded) unchecked into the ?marine realm. Mixing between
low-gradient alluvial plain or braidplain. floodwaters and ambient waters appears to have
The first environmental setting is dscounted been extremely limited (cf. Dam & Andreasan, 1990)
because there is no evidence (deeply incised and the subaqueous flood deposits are internally
channels, levees) for the distributary channels. If the similar to subaerial sand and gravel sheets deposited
fine-grained succession represented lagoon deposits earlier (braided stream association) on an alluvial
onlapping a braidplain during a marine transgression plain. The association has many features in common
then one would expect: (i) some evidence for with distal floodbasin/lacustrine deposits described
landward-direct currents (e.g. from storm wash­ from a rapidly subsiding Devonian strike-slip basin
overs), and (ii) the lagoon deposits to be overlain by (Hornelen Basin, Norway) (Steel & Aasheim, 1978) .
sandstones representing the protective barrier islands/ Similarities include the thickness of depositional
bars. Yet all palaeocurrents are seaward directed units, occurrence of horizontally laminated, rippled
(Fig. 5) and the floodbasin association is overlain by and trough cross-stratified sandstones inter­
a black laterally extensive mudstone interpreted as stratified with mudstones, convoluted lamination,
offshore marine (Orton, 1990) . and absence of obvious intermediate (2-20 m) or
A lacustrine origin, the third alternative, is more larger scale cyclicity.
difficult to fully disprove. Lakes can develop on
poorly drained areas of an alluvial floodplain and •
receive sediment from nearby river tracts. Quiet­
Conglomeratic alluvial fan sandstones
water suspension deposition of mud and silt occurs
during normal river discharges , with sheets of coarser Conglomeratic alluvial fan deposits characterize the
sediment introduced during large river floods, when northeast margin of the Tryfan depocentre and are
crevasse splays form. The grain size of sediment well exposed between Craig yr Ysfa and Carnedd
within crevasse splays reflects that carried by the Llewellyn (Fig. 1 ) . The alluvial succession that is
parental river tract, and can include coarse sands correlated with the Tryfan Fach Member (Fig. 1 1 )
(e.g. Pollard et al. , 1982). Palaeocurrent directions overlies bioturbated siltstones t o very fine-grained
from crevasse-splay deposits are usually at a high rippled and horizontally laminated sandstones
angle to the associated river deposits, and crevasse­ thought to reflect deposition in a shallow quiet­
splay sandstones often display coarsening or water environment (Howells & Leveridge , 1980).
thickening upward cycles according to position with Alluvial-fan facies rest erosively on these bioturbated
respect to the active river tract (e.g. Galloway, siltstones, and a scour at (, :ast 50 m wide and more
198 1 ) . If the floodbasin association is considered in than 1 .5 m deep has been identified (Fig. 1 1 , 2 1 .5 m).
relation to the underlying rhyolitic fluvial succession , The lowermost alluvial sediments (Fig. 1 1 , 2 1 . 5 -
the above lacustrine alternative seems improbable. 3 6 m ) consist o f either (i) poorly sorted medium­
The most important observations are: (i) the contact grained to conglomeratic sandstone with trough
between the fluvial and floodbasin succession cross-stratification or (ii) scours infilled by normally
is gradational (over about 2 - 3 m), (ii) the lateral graded , conglomeratic sandstone to fine sandstone.
extent (perpendicular to palaeoslope) for the braided Palaeocurrent directions are all to the west in keeping
stream and floodbasin associations is the same, with those obtained from younger fan successions at
(iii) palaeocurrent directions from the floodbasin Craig yr Ysfa (Orton, 199 1 , fig. 4) . The trough
association are consistent with those obtained from cross-stratified sandstones suggest deposition by
the underlying fluvial association, and (iv) little sys­ braided streams. Compared with the cross-stratified
tematic vertical change in facies or grain size could sandstones of the braided stream facies association
be discerned within the floodbasin sections. these sandstones (i) have larger foresets (sets to
I therefore conclude that the floodbasin associ­ 70 cm) ; (ii) less commonly display upwards changes
ation represents a submerged low-gradient alluvial in facies, grain size, or set size; and (iii) contain
plain. During most of the time, quiet-water sus­ planar cross-stratification (Sp ). These observations
pension sedimentation of silt and mud occurred. imply that rivers or floodwaters were comparatively
However, when flash floods and sediment gravity deep, that channel avulsion and/or flood-stage erosion
 Facies models in volcanic terrains 173

60 ·+---.....,
Wave reworking
)-
Wave swash Fan
abandonment
Small braided streams
--- -
. •.
·J
----=r
_J ------
Thin sheetfloods
50 <>.o•,•- •J'ng",\ G ms
.·

:-: -:-.--7.-:�:: :·- :

- :;_���" :�
) Sb Disedilutme ent gravity flows
:_ � Sm
�--�- - :.�
-
¢�
-�--·-�:· __G_m_G______
. ' :_:_:_:_:,_f_:_::_:_�_t_:_:_____ Proxi
·. _

al l u vi malal
es fi loewsd fan
-
��-
•

----;:..-' <:_... J
_.A...-.
�
vi s cous
Stream flow
debri
Poorlsy flchannel
debri ows ized
Ext r e me var i a ti o n i n Mid-fan
::o·,.::-.<;:-:<:1
-�� -----------------
St 10cm flow regime
_:�1) llonw-fil sofinuosimobityle
�:�4
�--� j
brai--d--ed channel s-----­
30 � �
-
� �
-�- � -)-
� -- -- ---- -- --
_) Sg Channel
sedi m izedgradiviltuyteflows
ent
�_:, and shal l o w ( 1 m deep!
--:� d: O o • C
UI ___,)
st
--------·
'
r e amfl o w alfanDilsutalvial
� ..
_./
...A: _ ·�
- < __ /�/
Stem30 -.... llonwfil siofnuosi3-4 tmy channel
/
deep
� · --
__/'-...�
. /

BROAD
�::;:=
��70cm �::::r:-;'St - ------
CHANNEL

20 §�,
Flinutovilaolwsheetwavefloenerods gy Shoreface
/

8HYOLITE
VOLCANICITY
,1P Sh lake
------

rs
Bioturbation ' Of shore'
M Sl Sand G
Fig. 11. Alluvial fan facies association . From section A, between Carnedd Llewellyn and Craig yr Ysfa, SH 6885 6450.
Distances are metres above the base of the Tryfan Formation . Note occurrence of rhyolitic ash-fall tuffs in fine-grained
deposits beneath the coarse-grained alluvium .
174 G.J. Orton

of topstratum prevented the formation of fining FACIES ORGANIZATION

upward cycles , and that in-channel bars (hence Sp) AND DISTRIBUTION
occasionally developed (see earlier discussions).
The normally graded conglomeratic sandstones
Facies relations in the Tryfan anticline
resemble hyperconcentrated flow deposits (e.g.
Smith, 1986; Maizels, 1989). Deposits in the Tryfan The first three facies associations are best displayed
Fach Member are either internally structureless or on the northwest limb of the Tryfan anticline, where
contain poorly developed trough cross-stratification four sections, spaced about 400 m apart have been
within the finer grained upper portions. The latter measured (Figs 2, 5 & 12) . These can be correlated
relationship indicates an initial rapid fallout of with a section (F) on the opposite limb of the anti­
suspended load and deposition by direct suspension cline, about 3 km to the southeast (after unfolding) .
sedimentation, followed transitionally by traction Along the northwest limb of the anticline , the
sedimentation within the more dilute low-density marine shelf association thickens northeastward
upper portions (see also Lowe, 1982). The inter­ from 30 to 70 m. Although vitric siltstone beds are
calation of such debris flows within finer grained thicker and have less admixed epiclastic material in
stream flow deposits indicates rapid variation in the southwest , sandstones are coarsest and most
fluvial discharge conditions on the alluvial fan. prevalent in central sections. On the southeast limb
The overlying alluvial fan sediments (Fig. 1 1 , 36- (section F) , the total succession is muddier. How­
51 m) are dominated by unstratified conglomeratic ever, a 1 m thick, medium-coarse-grained sandstone
sandstone occurring in tabular beds 20 -50 cm thick. containing moderate-scale trough cross-stratification
They commonly contain subvertical rhyolite pebbles (25-cm sets) lies 1 8 m beneath the top of the marine
(4 em size) or rip-up clasts of siliceous siltstone (30 em mudstone. Although reliable palaeocurrents could
long). The sharp non-erosive bases, occurrence of not be obtained, a fluvial origin seems unlikely given
well-defined units, lack of sorting or grading and its stratigraphical position. As there is no evidence
nature of fabric are all in accord with deposition by for tidal processes during deposition of the Tryfan
cohesive matrix-supported debris flows. Formation , this lends further credence to the idea
The overall succession can be interpreted in terms that the marine shelf association represent a low­
of increasing proximity to sediment supply. Although gradient platform on which wave energy decreased
deposits could represent a valley-fill, with axial shoreward.
braided streams and laterally derived debris flows, The relative thicknesses of braided stream and
there is no evidence for this physiography in the floodbasin components (Fig. 9) indicate that the
sedimentary record. Instead, a simpler model is most persistent axis of coarse fluvial deposition was
favoured involving progradation of a small alluvial along the southwest side of the braidplain. Fluvial
fan westward into the Tryfan depocentre . The main­ discharge and attendant sediment/water ratios also
tenance of deep channels in a distal fan setting is not appear to have been more variable along this side.
common because channel bifurcation usually results For instance, the thickness of cross-stratified sand­
in flow expansion, progressive shallowing, and stone sheets implies that most 'channels' were quite
eventually unchannelized flow. It implies both an shallow; thin debris flows and sheetfloods were an
abundant supply of water and confinement of flow important part of aggradation on the alluvial plain.
by channel banks. In contrast, to the northeast, successions indicate
The uppermost alluvial sediments of the Tryfan infill of several deeper (3 -4 m) 'channels' and sheet­
Fach Member (Fig. 1 1 , 5 1 -60 m) are better sorted flood deposits are absent. Despite good exposure
and finer grained. Trough cross-stratified sandstones (> 70% outcrop) more detailed stratigraphical cor­
yield palaeocurrents in a variety of directions. relation was not possible.
Thin beds of fine-grained sandstone display planar To the southeast (section F) , the cross-stratified
horizontal to low-angle cross-cutting lamination braidplain sandstone sheet can be correlated with an
and contain concentrations of magnetitic heavy erosively based , 2.5 m thick, fining upwards suc­
minerals. Together these facies suggest reworking of cession of trough cross-stratified medium- to fine­
the fan surface , perhaps by waves, following its grained rhyolitic sandstone (Figs 12 & 13). Fine
abandonment. siltstone occurs as thin discontinuous lenses between
individual sets, indicating numerous slack-water
periods during aggradation. The cross-stratified
 . . ... - · ·····
;:;;;-

�
"'

�-
�
0
50 m >:>..
"'
<:;-
;:; ·
�
0
"'
.,
;::
'0 ;:; ·
;;:-

�
FACIES �
;:; ·
ASSOCIATION "'
'V
vv

�
·v
FLOODBASIN v v
·v
,;
:
vv l KM
·vv c
v
BRAIDED STREAM �'Vv V
vv 60 � 6
vv
.···
... .
Ill
'VV
MARINE SHELF 'vv . o
,, ::) · ' r:::;J CAPEL CURIG
:'7 E (,../ F ., . c::J VOLCANIC FM.
Eg�·jj ALLUVIAL FAN ,
._.• ,_.,,,./ ,..,.., GWERN GOF TUFF
; ; : _.....�.
. .
70

Fig. 12. Distribution of facies associations in Tryfan anticline. Inset shows location of sections described in this paper. Horizontally ruled vertical bars denote
......
location of detailed sections shown in Figs 4, 7 , 9 & 13. -..)
Ul
176 G.J. Orton

20

Interbeddedandbiositurltsbtatoneed
sandstone Transgrn teossiofvfsehore
floodbasi

15

Trmedioughumcrtooss-finse-trgartaifiineedd Fluvial flood currents

sandstone
mudstone iwinterthbeds
Thinly bedded
sandstone ? Shoreface

Fossiliferous sandstone
Mudstone Offshore
Fig. 13. Section through upper
portion of Tryfan Fach Member,
southeast limb of Tryfan anticline.
M Si Sand See Fig. 12 for location of section .

sandstones overlie a 2-m-thick sharp-based suc­ against the background of these sections . Two
cession of massive (?bioturbated) and horizontally interpretations warrant further comment. The
bedded sandstones, and are overlain by interbedded cross-stratified rhyolitic sandstones are similarly
fine -coarse-grained sandstones, siliceous siltstones, interpreted as representing distal fluvial currents .
and interlaminated sandstones and siltstones. If the cross-sets were wave produced , their size
Overall the succession mimics the facies suc­ suggests a high nearshore wave energy, yet wave
cessions seen on the northwest side of the anticline power was low enough that mudstone could be
and interpretation of section F (Fig. 13) is made deposited during interludes between input of coarser
 Facies models in volcanic terrains 177

sediment. This does not negate the possibility, facies successions either missing or thinner than
however, that some or all of the cross-stratified expected. Even if the braided streams prograded
succession may have been deposited below sea-level. into a low-wave-energy basin one would expect
The second important difference is that the cross­ more widespread shoreface or delta-front deposits
stratified sandstones lie above fine-grained sand­ between the alluvial fan/braidplain and interfan/
stones, and rest erosively on them. offshore mudstones (e.g. Vos & Eriksson, 1977;
On both sides of the anticline, the floodbasin Fernandez et al. , 1988; Bardaji et al. , 1990).
association gradationally overlies the fluvial deposits, Consequently, the abrupt juxtaposition of 'fluvial'
indicating a reduction in the amount of coarse sedi­ and 'marine' along the northwest limb of the Tryfan
ment deposited on the braidplain. Successive sand­ anticline, as well as the scoured base to the more
stone beds, however, do not become finer grained distal ?fluvial sandstones suggests a relative sea-level
ur-section . They merely become thinner or less fall during sedimentation, with possible removal
common (i.e. more widely separated in vertical of upper shoreface/nearshore deposits by erosion.
section). 'Uplift' appears to be greatest in the northwest,
where braidplain deposits rest with sharp contact on
marine mudstones. The succession to the southeast
Bounding surface between the braided stream
(Fig. 13) suggests more continuous aggradation,
and marine shelf associations
although the condensed nature and sharp base to
One fundamental question concerns the position of probable shoreface sandstones indicates that relative
the marine shelf association with respect to a shore­ sea-level was still falling ( cf. Plint, 1988).
line. At least two interpretations are possible, largely A similar abrupt alternation between braid­
because of uncertainty concerning the amount and plain and shallow water environments has been
source of the sand along the coastline. Neither described from other strike-slip basins, including the
model is entirely satisfactory. Pennslyvanian Boss Point Formation (Browne &
The first alternative treats the sandy portion of Plint, 1994) and English Westphalian B strata
the marine shelf association as part of a shoreface. (Haszeldine & Anderton, 1980) . In the Boss Point,
The mudstone-dominated interval that caps the as­ braidplain packages rest erosively on, and locally cut
sociation may represent the shoreline itself. Walker deep ( 15 m or more) scours into underlying lacustrine
and Harms ( 1975) and Hamblin ( 1992) have de­ mudstones. A reason for the sharp contact was not
scribed a tideless muddy coastal deposit with some given , although uplift due to episodic tectonic source
similarities. However, in the marine shelf association area rejuvenation seems likely.
no indicators (e.g. change in colour or reddening
of mudstones, desiccation cracks, palaeosols, beach/
Extent of the deposit
foreshore deposits) for subaerial or near subaerial
conditions were observed, and even the uppermost It is pertinent here to dispel the notion that the
deposits appear to be fully marine. The second deposits described represent a small or distant
alternative considers the sandy succession to be a component of a much larger volcaniclastic apron.
small and offshore component of an even muddier In the first place, deposition of the Tryfan Fach
coastline. On such coastlines bed thicknesses and Member occurred only to the northeast of Llanberis
grain sizes first increase than decrease away from the Pass (Fig. 1 ) . The Snowdon region was uplifted
shoreline. This pattern arises because of in situ and subaerial throughout deposition of the Tryfan
resuspension of ambient muddy sediment with fines Formation (Orton , 1991 ) and contemporaneous
transported landward combined with offshore marine mudstones further to the southwest belong
transport of coarser sediment during storms (cf. Hill to a separate depocentre. Although contempor­
& Nadeau, 1989; Myrow, 1992) . In this scenario , aneous mudstone-dominated successions occur
the uppermost mudstone-dominated interval of the 10 km to the southeast, distal 'feather-edge' sand­
marine shelf association would correspond to both stones from the Tryfan Fach depositional system
muddy shoreline and nearshore subtidal deposits. have not been recognized. In light of the rate at
Yet it contains no evidence (flaser and wavy bedding, which the braided stream sandstones pinch out across
current ripples, gutter casts) for a nearshore setting. the Tryfan anticline, correlation is not suspected.
In both alternatives the transition between the Less than 5 km to the northwest , adjacent to the
two associations appears too abrupt, with expected Nant Peris fault (Fig. 1 ) , all Ordovician successions
178 G.J. Orton

are stratigraphically condensed (Trythall et a!. , 1987 ; cal instability and reactiVIty of their framework
Orton, 1990). The entire Tryfan Formation is about grains. Glass suffers particularly, with chalcedony,
50 m thick, compared with more than 700 m in the opal, zeolite and clay minerals being alteration
Tryfan anticline. It consists of mudstone or fine­ by-products (e.g. Davies et al. , 1979; Surdam &:
grained sandstone with no sandstones being similar Boles, 1979) .
in composition to the rhyolitic sandstones of the The end result of sorting and alteration processes
Tryfan Fach Member. These must correlate with a in distal exposures is usually a crystal- lithic sand-­
zone of non-deposition and/or an unconformity near stone, with framework grains set in a fine-grained,
Nant Peris. variably altered and/or recrystallized, non-descript
Ordovician successions also occur an additional matrix (e.g. Runkel, 1990; Cather & Folk , 1991) .
30 km to the northwest, on Anglesey (Fig. 1 , inset) . Thus many features, including the geometry of the
Although substantial Precambrian strike-slip move­ deposit, its sedimentary facies, the variability of
ment occurred along the narrow stretch of water its composition, shape and sorting parameters ,
(Menai Straits) that separates Anglesey from the and the extent that sediment composition differs
Welsh mainland (Gibbons, 1983) , lithostratigra­ from the 'norm' for sandstones within the basin,
phical correlation of Cambrian rocks across the are needed to establish penecontemporaneous pyro··
Menai Straits indicates little or no movement of clastic volcanism.
Anglesey relative to the Welsh mainland since the
Early Cambrian (Reedman et a!. , 1984). Ordovician
Petrography
successions on Anglesey consist of deep-marine
muds and/or mass flows deposited in smaLl fault­
Alluvial fan deposits
bounded basins (Bates, 1972). These include
volcanogenic massive sulphide deposits hosted in Alluvial fan deposits contain significant mono­
rhyolitic volcanics and pillow basalts of probable crystalline quartz (24% ) , little feldspar (8%) and
Caradocian age (Pointon & lxer, 1980; Westhead, lithic fragments dominated by aphanitic or sparsely
1991) . Thus, although there might be some Cardocian porphyritic quartz-bearing rhyolite (20%) , siltstone
volcanism on Anglesey, there was certainly no (26% ) , or granitic (3% ) lithic fragments (Table 1) .
upstanding subaerial volcanic arc that could have Composition indicates derivation from a mixed
sourced the Tryfan Fach Member. granite , tuffaceous sediment, and rhyolite volcanic
For the above reasons, it is concluded that the source. Although volcanics still comprise about half
depositional system responsible for the Tryfan Fach of the lithic population, lithics are thought to
Member affected areas within only a few kilometres be derived from older rhyolites of the Llewellyn
of Tryfan Fach itself, and that the deposits of the Volcanic Group.
Tryfan anticline represents a substantial portion of
the area traversed by the system.
Braided stream and floodbasin deposits
Sandstones contain large (to 4 mm) euhedral sani··
EV IDENCE FOR VOLCANISM dine (25%) and smaller plagioclase crystals (17% ) .
The percentage of monocrystalline quartz (5% )
Characteristic petrographic features indicative of is low compared with the 'norm' for the Tryfan
explosive volcanism include glass shards, pumice Formation (Table 1) . Lithic fragments (to 1 em) are
and scoria, bipyramidal �-quartz, and euhedral extremely angular and dominated by rhyolitic
zoned feldspars. To those not familiar with volcani­ volcanics. These are mostly porphyritic two-feldspar
clastic deposits, it may seem an easy task to establish felsic volcanics (32%) with phenocrysts of albite­
when volcanism occurred. However, volcanic clasts oligoclase and simply twinned alkali feldspar set in a
differ widely in size, shape, density and durability, recrystallized quartzo-feldpathic matrix that some··
with mean settling velocities varying by several times displays a trachytic texture . Quartz pheno··
orders of magnitude. During reworking, hydraulic crysts are rare and deeply embayed when they do
processes can quickly remove most of the less dense occur. Other lithic clasts include aplite, ?rhyolitic
or smaller ash and pumice fraction. Volcaniclastic tuff, or perlitically fractured glass (8% ) . Clasts are
deposits are also more susceptible to diagenesis set in a quartzo-feldspathic matrix ( c. 20% of totall
than any other type of sand because of the chemi- rock) . The texture of the matrix and lithic fragments
 Facies models in volcanic terrains 179

Table 1. Composition of sandstones

Tryfan Fach Member

Tryfan Fach Craig yr Ysfa Tryfan Formation

braidplain (n = 7) alluvial fan ( n = 5) (n = 101)

Mean SD Mean SD Mean SD

Monocrystalline quartz 4.8 2.9 23.4 9.8 32.6 15.6

Polycrystalline quartz 2.6 3.9 10.6 5.2 6. 1 5.4
Chert 1.6 1.7 11.2 6.4 7.8 7.2
K-feldspar 25.3 15.2 5.3 3.2 14.0 1 1.4
Plagioclase 17.5 9.8 2.3 1.2 6.5 6.5
Rhyolite 31.9 14.5 10.7 6.0 7.2 12.3
Rhyolitic tuff and perlite 8.5 8.2 Absent 1.8 5.8
Aplite 1.1 1.3 Absent 1.2 1.5
Mafic volcanics 2.4 3.7 3.8 2.5 1.7 3.9
Siltstone 1.9 2.5 26.0 12.5 14.8 14.8
Granite 0.2 0.4 2.7 3.6 0. 9 1.6
Opaques 1.3 2.0 0.5 0.6 2.9 5.8
Other heavy minerals 0.4 1.0 0.2 0.3 0.4 0. 9
PF ratio 0.43 0. 19 0.33 0.12 0.36 0.25
LvL ratio 0. 94 0. 09 0.51 0. 13 0.58 0.30
QF ratio 0. 14 0. 1 1 3.3 0.7 2.4 1.9
QFR 9 : 43 : 47 46 : 8 : 46 41 : 21 : 38
QmFLt 5 : 43 : 52 24: 8 : 68 34: 21 : 45
QpLvLs 5 : 89 : 8 17 : 44: 39 16 : 51 : 33
LvbLvrLsp 6 : 83 : 1 1 6 : 36 : 58 4: 46 : 50

PF, plagioclase/total feldspar ratio ; LvL, proportion of volcanic lithics within lithic population; QF ratio, quartz/total
feldspar ratio. Q, total quartzose grains; F, total feldspar grains; R, unstable lithic fragments; Qm, monocrystalline quartz;
Lt, total lithics R + polycrystalline quartz; Qp, polycrystalline quartz; Lv, volcanic lithic fragments; Ls, sedimentary
=

lithic fragments; Lvb, mafic volcanics; Lvr, rhyolitic volcanics; Lsp, sedimentary and plutonic lithics.

is similar and it is often difficult to tell the two apart. rarely exceeds 50% (Ewart, 1979), the crystal
Although geochemical data are not available, the content of sandstones must have been enriched,
crystal population (both free and within lithics) either by fractionation processes accompanying
suggests derivation from a rhyodacite to rhyolite the eruption and deposition of the tephra and/or
magma. hydraulic sorting during fluvial reworking. The
The abrupt appearance of a monomict sheet erupted volume of magma would therefore be much
consisting of feldspar crystals and angular volcanic larger than the volume of deposits in the Tryfan
lithics within an environment previously receiving Fach region, perhaps double. Although still small,
mudstone suggests a penecontemporaneous volcanic the volume is still well in excess of the amount
eruption. Unequivocal evidence for explosive pyro­ of pyroclastics (c. 0. 15 km 3) extruded in the 1980
clastic volcanicity is provided by the vitric siltstones eruption of Mount St Helens (Rowley et al. , 198 1 ) .
containing accretionary lapilli. Although the nature
of the eruption is hard to pin down, the high pro­
Pre-eruptive marine mudstone association
portion of non-vesicular lithics and crystals, low
abundance of vitric-material, absence of pumice, The fine-grained sandstones intercalated with the
and presence of accretionary lapilli suggest small­ marine mudstones are moderately sorted quartzo­
scale phreatomagmatic eruption(s) (cf. Sheridan & feldspathic and lithic arenites. No samples were
Updike, 1975; Heiken & Wohletz, 199 1 ) . point-counted owing to the fine grain size and dif­
The rhyolitic deposits have a combined volume of ficulty of identifying components, but composition
at least 0.2 km3. As crystal content in volcanic rocks appears similar to the 'norm' for the Tryfan For-
180 G.J. Orton

mation (Table 1 ) . The anhedral nature of framework This remains to be located . There are several
grains, the high percentages of chloritic matrix subvolcanic intrusions in the Llewelyn Volcanic
and the dominance of sedimentary lithics suggests Group for which comagmatic extrusive correlatives
derivation from a source of tuffaceous sedimentary have not been recognized (Fig. 1) and which invite
rocks and mudstone. Although rhyolitic ash-fall comparison. These lie about 5 km or more (after
deposits occur near the base of the succession unfolding strata) to the west of Tryfan Fach along
(Fig. 3) there is no evidence elsewhere for pene­ the northeast trending Nant Peris fault. The largest
contemporaneous volcanicity, with the exception of intrusion (Bwlch y Cwyion) consists of microgranite
the vitric siltstones at the very top of the succession. (rhyodacite/dacite) with later marginal intrusions
of rhyolite (Howells et a!. , 1991) whereas smaller
intrusions are rhyodacites , granodiorites and feld­
Possible position of the volcanic source
spar porphyries.
In many cases volcaniclastic successions can be
interpreted in terms of proximity to a central volcano
(e.g. Smith, 1988) by comparison with well-studied REASSEMBLING THE
modern examples. In the Tryfan Fach Member, the STRATIGRAPH ICAL P IECES
absence of primary volcanic deposits (larvas, pyro­
clastic flow deposits) , coarse breccias and debris In the Tryfan Fach Member, each facies association
flows is consistent with deposition in a distal setting. is in some way different, in composition and/or
Similar monolithological syneruptive sands occur on facies, to adjoining facies associations. It will be
the most distal ring plains that flank major central shown that the meaning one attaches to facies
volcanoes at distances of 30- 100 km from volcanic changes between and within associations depends on
vents. whether the rhyodacitic sandstones are attributed to
I begin by discounting these distances. Within the penecontemporaneous explosive volcanism.
Llewelyn and Snowdon Volcanic Group, the absence
of systematic facies variation across the entire out­
One epiclastic system: time's cycle
crop belt, as well as the common identification
of intrabasinal vents, argues against derivation of The simplest solution, and the one that would be
volcaniclastic material from a line of volcanoes adopted if the petrography of sandstones and
beyond the reach of the present outcrop. Instead the composition of rhyolite clasts was unknown, is to
stratigraphy reflects a rather haphazard assemblage integrate the four facies associations into one linked
of small volcanoes, none of which had appreciable depositional system or systems tract (Fig. 14) . As
subaerial relief, erupting into and within subsiding noted by Hutton ( 1788) it is with 'pleasure that
sedimentary lowlands (Howells et al. , 1 99 1 ) . [we] observe order and regularity in the works of
Although the Welsh basin i s typically portrayed a s a Nature, instead of being disgusted with disorder and
back-arc basin, this is not essential, and such low­ confusion'.
lying volcanic fields can also occur due to intra-arc In this interpretation , a tributary alluvial fan at
extension (e.g. White & Robinson, 1992). In such Craig Yr Ysfa would feed sediment on to an axial
settings, 'distal' or more accurately 'low-gradient' braidplain confined to the southwest by additional
deposits (e.g. Tryfan Fach Member) are often locally uplift along Llanberis Pass. Younger alluvial systems
interstratified with more 'proximal' facies (e.g. lavas, in the Tryfan Formation exhibit this drainage pattern
pyroclastics). In North Wales, most large Ordovician (Orton, 199 1 ) . The sandier upwards marine sheU
volcanic eruptions had vents within a few tens of association would represent offshore to coastal plain
kilometres, or less, of the sites of deposition of environments reflecting deposition in front of,
their primary products (excluding fallout tuffs) adjacent to , or between alluvial fan salients that
(Kokelaar, 1988, 1992; Howells et al. , 199 1 ) . It is were deprived of coarser sediment. Tectonic uplift
not necessary to suppose that smaller eruptions had along the basin margin resulted in progradation of
more distant volcanic sources. The absence of the alluvial fan and a derivative braidplain into the
a correlative coarse rhyolitic unit amongst more marine basin. Lesser amounts of uplift within the
proximal successions about 5 km away (see above basin resulted in a relative sea-level fall and braid­
'Extent of the deposit' section) also lends support plain sandstones rest with sharp contact on a
for a local source for the Tryfan Fach Member. condensed nearshore succession . Following tectonic
 SEDIMENT
FACIES ENVIRONMENT PALAEOGEOGRAPHY RELATIVE SEA-LEVEL
COMPOSITION

LACUSTRINE

BRAIDPLAIN

�
"'

�·
;:!
Q
�
1::;-
;: ·
-.:
Q
;::;-
.,
;:s
r; ·
�

Gwern Gof Tuff

�
�-

OUTER 51-ELF

COM'OSITION

l [I]
'•'• •'•
••••• • • •
vvvv
vvvv ..-·.·.·
vvvv

pyroclastic reworked epiclastic +

pyroclastic Low + _ High
V+ + r· I

Fig. 14. Epiclastic interpretation of vertical successions. Block diagram (looking northeast) shows palaeogeography if morphological elements are genetically
related. An unchanging marine transgression is assumed throughout. Place names are: TF, Tryfan Fach; CY, Craig yr Ysfa. Facies associations as in Fig. 12. ......
00
......
182 G.J. Orton

uplift, a more gradual shift in the position of the occasional high-velocity floods to introduce coarse
aLluvial fan lobe aLlowed floodbasin waters to material on to the shelf; (iii) the thickness and
gradually transgress on to the fan/braidplain surface. 'purity' trends in reworked ash-fall tuffs at the top of
Floodbasin siltstones therefore increase in pro­ the marine shelf association , as well as grain-size
portion upwards at the expense of coarse sandstone trends across the braidplain (Fig. 5) suggest a source
beds. to the southwest rather than to the northeast.
There are many pieces that do not fit into this Together these features cast doubt on the strict
puzzle: (i) the composition of braided stream application of Walther's Law to the total sequence
deposits differs from sandstones of the alluvial fan at and seems to demand three weakly or unlinked
Craig Yr Ysfa, implying derivation from a differe;1t depositional systems that developed in advance of,
aLluvial fan or from a more distant axial source; and in conjunction with penecontemporaneous
(ii) coarse rhyolitic sand is never present in the volcanism (Fig. 15).
marine shelf association - one would expect

· ·· ·
:-. . · · · · :·.. · �

SEDIMENT
COMPOSITION
FACIES ENVIRONMENT • . :;, , J:�· ·[?i;��t��)· �t:�l;-
ALLUVIAL PLAIN

BRAIDED SAND·Sf£ET

NNER SHELF

OUTER SH::LF

Fig. 15. Pyroclastic (or syneruptive) interpretation of the vertical succession. Block diagrams (also looking northeast)
depict palaeogeography if Walther's Law cannot be applied to the full succession and morphological elements need to be
considered separately. Two types of 'reworked pyroclastic materials' are portrayed on 'sediment composition' column in
order to illustrate uncertainty about the number of volcanic events. Darker tone indicates pyroclastic materials that have
been reworked more extensively (i.e. they were deposited a greater length of time after the volcanic eruption). Several
(A, B and C) possible interpretations of relative sea-level change are presented for the syneruptive period. See text for
discussion. See Fig. 14 for key to sediment composition.
 Facies models in volcanic terrains 183

(3 - 15 km) magma chambers, with rates of uplift

Three unrelated depositional systems: time's arrow
ranging from 0.002 to almost 1 m per year (cf.
McKee et al. , 1984; Hill et al. , 1985; Luongo &
Progradation of a muddy shelf and postulated
Scandone, 199 1 ) . These rates obviously are extremes
volcano-tectonic uplift
or we would not notice them. The unique longer
An initial supply of fine-grained silts and mudstones term (2000 yr) historic record from Campi Flegrei
(largely non-volcanic) was delivered to a low-wave­ caldera (Italy) shows that episodes of unrest are
energy basin, probably by small channels or sheet­ often separated by decades of quiescence (Lyell,
floods (cf. Wells, 1983). Sand distribution suggests 1853, Caputo, 1979; Dvorak & Gasparini, 199 1 ) .
that the source lay to the northwest to west (Fig. 1 5 , Here, 10 centuries o f gradual subsidence ( 1 1 m) was
bottom block diagram). Reworked rhyolitic ash-fall followed by more gradual uplift beginning in the
tuffs were derived from a similar direction and may eleventh century. In 1538, a mere 0.03 km 3 of ash
represent small eruptions precursor to the main and scoria erupted from Monte Nuovo. This was
volcanic event. The lateral variation in the amount preceded, however, by 40 years of episodic uplift
of sand suggests that sediment input occurred (and land formation) culminating in 7 m of uplift
from localized point sources with minimum lateral in the two days preceding the eruption at a site
reworking at the shoreline. Widespread tectonic (Pozzuoli) several kilometres away.
uplift preceding and associated with volcanicity As the Tryfan Fach Member was not deposited
resulted in a relative sea-level fall. Hence coarse­ within or nearby such a caldera, it is pertinent to
grained sandstones derived from the volcanism rest query whether the above data is relevant. Extrusive
with a sharp, probably erosional contact on a rocks of the lowermost Llewelyn Volcanic Group
condensed nearshore succession . Five to 15 m of were ponded in subsiding, probably fault bound,
uplift would be needed to cause erosion of upper sectors of the sea-floor (Howells et al. , 199 1 ) , the
shoreface/nearshore deposits. The timing of uplift thickest accumulations (both about 1000 m) being
with respect to volcanism in the Tryfan Fach Member rhyobtic lavas and ash-flow deposits of the Conwy
is not thought to be fortuitous and some of the uplift Rhyolite and trachyandesite lavas and tuffs of the
is related to buoyant forces associated with the Foe! Fras Volcanic complex (Fig. 1 ) . As has been
intrusion and/or migration of magma. All large noted recently (Lipman, 1984; G.P.L. Walker, 1984) ,
rhyolitic eruptions in North Wales show evidence it is a semantic issue whether the rectangular volcano­
for tens of metres of tectonic uplift in the vicinity of tectonic depressions that contain these deposits
volcanic centres in advance of eruptions (Howells should be termed caldera. The final volcanism of the
et al. , 199 1 ; Orton, 1991) and there is no reason why Llewelyn Volcanic Group, the Capel Curig Volcanic
smaller amounts of uplift might not precede smaller Formation involved eruption of two rhyolitic ash­
rhyolitic eruptions. A volcano-tectonic control on flow tuffs (each > 20 km3) from a postulated vent in
uplift and/or development of regressive cycles has the north near Conwy, as well as smaller eruptions
also been suspected within other ancient sedimen­ (c. 1 km3) from volcanic centres just north of
tary, particularly extensional, basins (cf. Leeder, Carnedd Llewelyn and just south of Tryfan (Fig. 1 ) .
1983 ; Henry & Price, 1984; Watkins, 1986; Leckie & The continuity o f compositional variation within
Cheel, 1989; Sloan & Williams, 199 1 ) . the Llewelyn Volcanic Group suggests that various
Our understanding of the influence o f volcanism intrusive and extrusive products were all originally
on basin subsidence is limited by the short time comagmatic and related by closed-system fractional
span < 100 yr) over which detailed observations crystallization (Howells et al. , 199 1 ) . This necessi­
from modern volcanoes have been recorded and the tates a parental magmatic system that stretched
difficulty in ancient successions to separate volcano­ 25 km from the Llanberis Pass fault toward Conwy,
tectonic from tectonic effects. Most modern data and extended about 10 km in a transverse direction.
is from volcanic fields where large silicic caldera­ This magma 'chamber' may have been an elongate
forming eruptions previously occurred (Newhall & mid- to subcrustal beam or dyke controlled by the
Dzurin, 1988). Vertical ground movements are dominant structural grain , a shape documented by
best recorded at calderas located by the sea, where geophysics from modern continental rifts (e.g.
elevation changes, forming land , are very con­ Cordell et al. , 1985 ) . Petrochemical and exper­
spicuous. Ground deformation in calderas occurs imental data (Ball & Merriman , 1989) show that
readily because of the existence of large shallow rhyolites derived from this magmatic system evolved
184 G.J. Orton

at relatively shallow crustal levels (PH,o < 500 bars) . where (more proximally, or laterally) with the
Thus many of the conditions needed for volcano­ coarser grained braided stream deposits. If one was
tectonic 'unrest', such as earlier caldera-forming trying to determine relative sea-level changes, the
eruptions, existence of large , shallow magma top of the braided sandstone sheet might be picked
chambers , volcanism in an extensional graben-type as a marine flooding event and used to infer a rise of
structure, existed within the Snowdon graben at relative sea-level (sea-level curve B , Fig. 15).
Tryfan Fach time. 2 A contrasting scenario results if one allows
volcanism to continue and the type of volcanism to
change with time. Time lines would essentially
Aggradation of synerup tive pyroclastics
parallel bedding contacts rather than cut across
Sediment composition within braidplain and flood­ them as in the first scenario . In this case, it is assumed
basin settings all indicate deposition in conjunction that all vitric siltstones are primary, or only slightly
with a period of explosive rhyodacitic volcanicity reworked, pyroclastic materials. Rather than pro­
(Fig. 15, top block diagram). The sudden increase in ducing coarser grained pyroclastic material, the
sediment grade and abundance resulted in a brief volcano produced fine-grained ash. The occurrence
period of aggradation within shallow ephemeral of accretionary lapilli suggests that this change may
channels. This braidplain prograded rapidly as have been triggered by access of water to the vent.
a sheet across the pre-existing shoreline and/or Fluvial channels did not become more distant (i.e.
exhumed marine mudflat into a low-wave-energy distal) from the source up-section, they just became
basin, with facies suggesting a more direct input of overwhelmed by aggradation of fine-grained primary
volcanically produced sediment loads on to the and reworked fallout ash. Hence, coarse sandstone
southwest portion of the braidplain. progressively formed a smaller proportion of the
It is not certain whether sandstones represent sediment deposited on the alluvial plain. Laminated
pyroclastic material reworked from a series of small ashfall siltstones may have been deposited in ponded
eruptions rather than a single eruption (left- and depressions on the alluvial plain; a marine trans­
right-hand sediment composition columns of Fig. gression may have occurred only at the very top
15). Consequently, the thin siliceous mudstone of the succession, after all volcanism had ceased
beds within braidplain deposits could reflect: (curve C, Fig. 15).
(i) periods of quiet-water sedimentation following Two observations favour the second interpret­
lateral migration of the fluvial channels (the most ation. Firstly, sandstones at the top of the floodbasin
'traditional' interpretation) , or (ii) fallout of fine­ association (Fig. lOA) are as coarse as those lower
grained ash following episodes of rapid syneruptive down. Secondly, lamination in many of the siltstones/
aggradation of coarse clastic material. Fine ash may fine-grained sandstone beds is often . diffuse (Fig.
have been erupted at the same time as the underlying lOB) indicating that fine sediment from one event
coarse, clastic material but initially dispersed within was still being deposited when coarsest grains from
an eruption cloud. the next event arrived . Near-source ash-fall deposits
There are also several interpretations that can be commonly display such weak bedding (e.g. Branney,
presented for the braidplain to floodbasin transition. 1991) owing to fluctuations in the energy of the
1 The upwards decrease in the proportion of coarse­ eruption, direction and strength of winds or over­
grained sandstones may reflect gradual waning of lapping of several eruptive pulses. In contrast,
the coarse sediment supply, perhaps due to erosive sedimentary laminae in lakes are commonly sharp­
lowering of the volcanic edifice and/or near-vent based (e.g. Hamblin, 1992) because they reflect
subsidence following the volcanic eruption. In this infrequent episodes (e.g. seasonal varves) of clastic
model, the maximum basinward advance of the input.
syneruptive braidplain is 'picked' as the top of the
coarse-grained rhyolitic sandstones. Fine-grained
A symbiotic alluvial fan ?
floodbasin siltstones would largely represent matrix
of the pyroclastic flows, liberated by fluvial rework­ The relationship of the alluvial fan deposits to the
ing, and transported to the site of deposition by low­ other two genetic units is most perplexing. One
gradient streams. In this model, the floodbasin might at first expect that the two stratigraphical units
successions are distal and should interfinger else- were not contemporaneous. However, as four
 Facies models in volcanic terrains 185

coarsening upward alluvial fan sequences can be active volcanism and concluded that 'volcanic activity
recognized in the Tryfan Formation at Craig yr Ysaf was related in a close but subtle way to the subsidence
(Orton, 1988, Fig. 4) , and four braidplain -deltaic history of the basin, and thus to the sedimentology
successions have been recognized in the Tryfan anti­ of its fill'.
cline it seems reasonable to make a one-to-one
correlation. None of the other alluvial fan suc­
cessions contain feldspar-phyric rhyolitic volcanics THE STRA T IGRA P H ICAL RECORD
that could suggest otherwise . Rhyolitic ash-fall OF VOLCANISM
deposits within fine-grained 'shoreface' deposits that
underlie the coarse-grained alluvial fan deposits lend Sedimentary sequences associated with active
further support to this correlation. volcanoes are often considered to consist of two
The evidence suggests that the alluvial fan pro­ elements (G. Smith , 199 1 ; White, 199 1 ) : (i) syn­
graded at about the same time as volcanism was eruptive sequences, which result from primary
active, or shortly thereafter, but from the opposite volcanic deposition and immediate post-eruptive
side of the basin. Tectonic changes associated with reworking, (ii) inter-eruptive sequences, which
the volcanism must have produced enough relief record deposition without significant influence of
between Craig yr Ysfa and the Tryfan depocentre to volcanic activity where normal sediment delivery
allow fan progradation. Migration of magma may processes are dominant (Fig. 16) . The syneruptive
have been the mechanism driving this subsidence condition is a perturbation brought about by the
pattern. Magma migration should have two effects: volcanic eruption and includes not only the time­
(i) it will produce relative uplift in the immediate scale of the eruption but also the period following
vicinity of the volcanic vent; (ii) fault blocks that cessation of volcanism while hydrological and hill­
lose some of their supporting 'cushion' of magma slope conditions remain disturbed. In humid cli­
could subside relative to surrounding fault blocks, mates, and without vegetation, the syneruptive
even those more distant from the volcanic centre. In condition may continue for a long time, at least one
some ways this process is analogous to the movement or two decades, until most loose debris from the
of salt, laterally and upwards, toward the position of eruption is removed completely. In the Tryfan Fach
least overburden pressure. Member it includes the braided stream and flood­
Magma migration is documented most commonly basin facies associations. Syneruption sedimentation
in extensional volcanic settings. It has been invoked is in large part aggradational and the resultant stra­
in the Rio Grande rift to explain progressive increase tigraphy is an event stratigraphy with time planes
in the gradient ( 1 m km- 1 to 7 . 1 m km- 1) of rift essentially parallel to bedding surfaces.
drainage systems from late Pliocene to Holocene
times (Bachman & Mehnert, 1978) and present-day
Around stratovolcanoes
subsidence that is similar in size but opposite in sign
to uplift in the Socorro area of New Mexico further Most conceptual models for subaerial volcaniclastic
along the rift axis (Reilinger & York, 1979). A more alluvium are based largely on sedimentation within
convincing illustration is provided by the 1912 volcaniclastic aprons deposited adjacent to relatively
eruption of Novarupta (Alaska) where a depression high-standing volcanic arcs (e.g. van Houten 1976;
600 m deep (volume 6 km3) formed at Mount Katmai Vessell & Davies, 198 1 ; Smith, 1987, 1988). During
10 km away from the eruption site, implying a volcanic eruptions the instantaneous production of
complex magmatic plumbing system (Hildreth, large volumes of sediment, combined with enhanced
1987). In the basaltic Bedded Pyroclastic Formation and more variable runoff, result in sedimentation
(North Wales), episodes of subsidence preceded principally by aggradation of high-sediment-load
periods of copious magmatism and rapid sediment flood and debris-flow processes. Inter-eruption
accumulation (Kokelaar, 1992); this subsidence periods are often characterized by erosion and
pattern could be explained by lateral migration of channel incision; incision (Vessell & Davies, 1981)
magma away from deposition sites and toward following syneruptive episodes was an integral aspect
eruptive vents. Within a Variscan pull-apart basin in of the predictive facies model presented by G . A .
Spain, Besley & Collinson, ( 1991) also documented Smith ( 1991) for continental volcaniclastic sediments.
accelerated but more differential subsidence during Inter-eruption streams draining the high-standing
186 G.J. Orton

volcanic highlands are commonly of gravel-bedload 1 Slope was insufficient. Rhyolitic volcanic fields
character. The overall geometry of the succession typically have a subdued topography. In North
that results is controlled by the relative importance Wales, the stratigraphical architecture and the dis­
of syneruptive versus inter-eruptive conditions, tribution of products from volcanic cones suggest
which in turn are controlled by eruptive frequency that maximum relief was a few hundred metres
and rates of tectonic subsidence (G . A . Smith, 199 1 ; (Howells et al. , 199 1 ; Orton, 199 1 ) . Walton ( 1986)
Fig. 16a). also argued that the paucity of debris-flow deposits
in the Tertiary rhyolitic volcaniclastic sediments of
West Texas resulted from the low-gradient of source
Within a humid low-relief continental rift
areas. At the Taupo volcanic centre , although debris
The Tryfan Fach Member succession differs from flows occur in post-eruptive reworked pyroclastic
the above scenario in three ways: (i) debris flow deposits (R. Smith, 199 1 ) , their development
deposits are largely absent from the syneruptive depended on the moderate slopes ( 10°-20°) of the
sedimentary record; (although debris flows occur in inward dipping pre-existing topography surrounding
a penecontemporaneous basin-margin alluvial fan, Lake Taupo. This topography is a tectonically
this occurrence was unrelated to high sediment loads uplifted ridge of Upper Palaeozoic- Mesozoic grey­
due to volcanism); (ii) incision, if anywhere, occurred wackes that has nothing to do with the volcano itself
at the base of syneruption deposits; and (iii) fine­ (Cole , 1984).
grained muds, rather than gravels, were deposited 2 Depositional site was too distal. Cross-stratified
after or between volcanic events. These differences sands, not unlike those of the braided stream associ­
are thought to arise largely because the Tryfan Fach ation, occur within most distal volcaniclastic suc­
sedimentary system developed within a humid­ cessions. As volcanoes range from a few tens of
region, actively subsiding, continental rift. metres (e.g. tuff rings) to several kilometres (e.g.
some stratovolcanoes) in height, the term 'distal'
has no singular meaning in terms of distance from
Syneruptive facies and debris flows
source , with 'distal' facies developing as close as
Three reasons are given for the paucity of debris a few kilometres to more than 100 km from the
flows. volcano.

A: STRATOVOLCANOES 8: LOW RELIEF HUMID RIFT

·
f:t�: .��:�i�?:�\:�-?�_�.� �??�� ·:�_ :·:::.:s:�-:;-�-: :·: :: :};����.-: _-.=�-:: �--�_-��·:;-::-��
,. -. .. . ..�-

��� -� .� �::.:.C.. ::.�.l.�:�;�<��-:., ;,:;.).�:� i..�:�.-:��:<;

- -��- -0�_-.· .·..:;
� ;.:� · - ·-�·:.-:·:· ::: . .·o . •:a. ,..,uo "-o.-��-o.an.: .•:.
· ·.".:
.-c

-·
••

primary pyroclastics
SYN­ debris flow
ERUPTIVE hyperconcentrated flows

braided streams

Fig. 16. The stratigraphical record of volcanicity: (A) around stratovolcanoes. Pre-eruptive and post-eruptive successions
are not separated (from G.A. Smith, 1991). (B) Fluvial-deltaic successions of the Tryfan Formation, North Wales. Lower
example would represent Tryfan Fach Member.
 Facies models in volcanic terrains 187

3 Pyroclastic materials were reworked from a series basins, the details of the relative timing of sub­
of small eruptions rather than a single eruption. sidence and volcanism is a 'chicken-and-egg' type of
Although most small monogenetic volcanoes are question. Did increased transtension and associated
short-lived , multiple soil horizons described from subsidence cause the volcanic eruption or did the
the 'flanks' of some volcanoes (Wells et al. , 1990) , volcanic eruption occur and allow or drive renewed
direct observation of volcanic activity (Foshag & subsidence?
Gonzalez, 1956) and ancient examples (e.g.
Kokelaar, 1992) indicate repeated eruptions over a
Inter-eruptive sedimentation
long time period (decades to hundreds of thousands
of years) with reuse of the same magma plumbing Inter-eruption or 'background' facies and facies
system. Due to the small size of individual eruptions, geometries depend on the topography, relief, type
and time between eruptions, primary products of bedrock , climate, subsidence history, etc. of the
(pyroclastic flows, debris flows) could be completely drainage basin. Almost any facies could conceivably
reworked into cross-stratified sandstones. occur. In low-relief extensional basins fine-grained
mud commonly is deposited after volcanic events
on alluvial plains or within lacustrine to playa
Bounding surface between syn- and post-eruptive
environments (e.g. Besley & Collinson, 199 1 ;
deposits
Buesch, 199 1 ; Turbeville, 1991 ; White & Robinson,
All sedimentary basins are the result of isostatic 1992; this study) . These environments arise because
adjustment of the lithosphere to a tectonic driving of the low topography of the basin, because syn­
force , enhanced by subsequent sediment loading eruptive deposition flattens and infills topography,
(Watts, 1989), and in this case modulated by the because low-gradient drainage systems are easily
presence of an underpinning, buoyant magma body. blocked or re-routed by lava flows or debris flows, or
The amount and rate of surface deformation is because volcanism drives basin subsidence.
controlled ultimately by the strength of the litho­
sphere (Karner & Dewey, 1986) and gravity controls
on the upwards migration and accumulation of FACIES MODE L L ING :
magma (G.P.L. Walker, 1989). If the lithosphere is T IME ' S ARROWS V ERSUS
highly fractured, or magma is of low density and/or T IME ' S CYCLE
viscosity, deformation will be compensated locally.
Unconformities can develop within any sedimen­ The stratigraphical record allows us to define 'bite­
tary basin, their occurrence depending on rates sized' pieces or facies (R. G. Walker, 1990). In facies
of subsidence, rates of sediment production, and modelling we assume 'there is a system and order in
eustatic sea-level change. The only base-level Nature' (R.G. Walker, 1992) and attempt to re­
scenario that can prevent syneruption deposits of assemble these 'bites' into a group of facies that are
the Tryfan Fach Member from being incised after genetically related to each other and have some
aggradation is to invoke a rapid rise in relative sea­ environmental or palaeogeographical significance.
level in conj unction with the explosive volcanism. The two end-member conditions, time's arrow (a
Given the short time available, a eustatic control unilinear succession of unique events) and time's
seems unlikely, and the sea-level rise is best related cycle (recurrent patterns in a world that remains
to accelerated subsidence of the underlying fault essentially unchanged) are categories of our invention,
block. Although this subsidence could be purely erected here to illustrate the problems of interpreting
tectonic and have no relation to the volcanism, the volcaniclastic successions. Although both approaches,
coincidence of the two events seems hard to explain. by themselves, are incorrect (as noted by Gould,
Several other authors have also noted that vol­ 1987), one may become dominant.
canism was associated with accelerated and/or dif­ Two variables unite to determine the degree to
ferential basin subsidence (Kokelaar, 1988; Besley which Walther's Law is applicable; these are the
& Collinson, 199 1 ; see above). This genetic linkage completeness of the sedimentary record (the num­
between rates of basin subsidence (e.g. relative sea­ ber of time gaps), and the extent to which the sedi­
level rise) and sediment load (e.g. an eruption) ment source changed, in terms of the amount and
is uncommon in normal sedimentary successions. grain size of supply, throughout deposition (Fig.
To some extent, however, particularly in strike-slip 17) . The two end-member conditions are: continu-
188 G.J. Orton

SEDIMENT STRATIGRAPHIC COMPLETENESS WALTHER'S

COMPOSITION "BITES" OF RECORD MARITAL STATE

t
(/)
(/)
w
z
�
u
�
None A lot
AMOUNT SOURCE CHANGED PURE BLISS DIVORCED
BETWEEN SUCCESSIVE BEDS

Fig. 17. Application of Walther's Law to syneruptive and inter-eruptive volcaniclastic successions (using ideas from Dott,
1983; Gould, 1987; Walker, 1990, 1 992). Walther's marital state is discussed further in the text.

ous sedimentation and/or an unchanging source transformed by introduction of coarse-grained

(Walther's Law fully applicable) ; episodic sedi­ rhyodacitic material associated with the explosive
mentation and/or a constantly changing source volcanic eruption.
(Walther's Law inapplicable) . As all environments, The time's cycle metaphor is most applicable
including non-volcanic ones, are to an extent within the pre-eruptive marine mudstones and the
influenced by episodic events (Dott, 1983 ) , the epiclastic alluvial fan succession. Because source
metaphor that becomes dominant depends on the was comparatively constant and unchanging (or
time frame over which observations are made. changing gradually) a facies model can be provided
The time's arrow metaphor embodies the entire for the whole succession, vertical sequences can be
Tryfan Fach Member. The coarse-grained cross­ interpreted, and palaeogeographical maps can be
stratified and thinly bedded rhyodacitic sandstones presented . Although unique events did occur (e.g.
are unrelated to and lie with probable erosional the ash-fall beds, flood events from the alluvial
unconformity on subjacent marine mudstones. The plain) , their magnitude was insufficient to radically
up-sequence progression of facies on this time-scale alter pre-existing sediment dispersal systems and/or
is essentially unpredictable, as pre-existing sediment environments. Episodic hydrological and oceano­
dispersal systems were overwhelmed and probably graphic events (storms, river mouth floods) become
 Facies models in volcanic terrains 1 89

'normal' over the time period (tens of thousands a nearby, but unrelated, source. However, within
of years) involved and Walther's Law is fully syneruptive pyroclastics it could not be determined
applicable. whether coarse sandstones were reworked from a
In non-volcanic terrains , and in the epiclastic single or several eruptions, and whether rhyolitic
successions noted above, it is implicit (de Raaf et a/. , siltstones on the alluvial plain represent fine-grained
1965) or explicit (Busch , 1971) that any progressive material extracted from older pyroclastic flows
facies association with gradational facies contacts during reworking or influx of ash from new volcanic
is the result of one related set of depositional eruptions.
conditions. However, within syneruptive reworked The key to unlocking the story hidden within such
pyroclastic deposits, gradational grain-size and facies volcaniclastic sequences is to recognize when
changes are virtually meaningless (or at least very volcanism was active. The basic question that is
difficult to interpret! ) in two-dimensional sequences addressed is whether facies changes in a vertical
unless the eruptive history of the volcano is known. sequence reflect changing energy levels within the
This is rare, even in historic eruptions. Many volcanic depositional setting (Walther's Law is applicable) or
eruptions consist of a series of closely spaced but changing source parameters because of or during an
smaller events, with the amount, grain size, density, eruption (Walther's Law does not work ) . In this
shape and/or 'distribution of eruptive products regard, the study of volcaniclastic materials is still in
gradually changing owing to progressive changes in its infancy, and is the area where much new research
the amount of magma vesiculation, the eruptive flux is needed. The above understanding is possible only
rate, vent parameters , and the supply rate of water. if volcanism-induced eruptive and emplacement
Although each bed is unique, the succession of syn­ processes can be recognized and separated from
eruptive beds may not be random but linked by 'background' conditions of sediment supply (see
gradual (?and predictable) changes in source region also McPhie & Allen, 1992). Even when the eruption
parameters. The problems this causes in sequence is observed (e.g. Mount St Helens) , it may be dif­
analysis are well illustrated in the Tryfan Fach ficult to distinguish between primary and reworked
Member by the braidplain to floodbasin transition, pyroclastic material, and to recognize individual
where it could not be determined whether facies eruptive episodes in the depositional record. These
change reflected sorting processes on the alluvial distinctions becoming increasingly difficult to make
plain or progressive eruption of finer grained volcanic when studying the ancient record , where syneruptive
ash. Environments were difficult to reconstruct and sheets may be difficult to separate (G. Smith, 1991)
palaeogeographical maps could not be produced. As and material from smaller eruptions becomes
a consequence the timing of other controls on the admixed with 'background' inter-eruption detritus
stratigraphy (e.g. sea-level change) could not be (Hackett & Houghton, 1989). In these successions,
fully assessed . Thus although one cannot totally it may not be possible to determine whether facies
divorce Walther's Law from these volcaniclastic change reflects a change in type and/or energy of the
sequences it certainly must be applied with caution, environment, or a change in the type of sediment
with the relationship best described as an unhappy produced by the volcano, or both. We should be
marriage (Fig. 17) . prepared to admit this and realize the implications
for the accuracy of environmental and/or sequence
interpretations .
CONCLUSIONS

Coarse-grained sandstones of the Tryfan Fach ACKNOWLEDGEMENTS

Member represent aggradation and progradation of
reworked pyroclastic materials deposited in a braid­ Research formed part of a doctoral thesis ( 1990)
plain setting in conjunction with penecontempor­ completed under the helpful guidance of Harold
aneous ?phreatomagmatic rhyodacitic volcanicity. It Reading (Oxford University) and Roger Suthren
can be shown with reasonable certainty that these (Oxford Brookes University) . Later visits to North
deposits have no genetic relation to either muddy Wales were financed by the Monash University
shelf deposits , upon which they lie with probable Development Fund and the Australian Research
erosional unconformity, and/or adjacent alluvial fan Council. Over the years stimulating field discussions
deposits, which prograded at the same time from in North Wales with Dairmid Campbell, Ray Cas,
190 G.J. Orton

Bill Fritz, Malcolm Howells, Adrian McArthur and Petrol. , B64, 40-59.
Tony Reedman have contributed to my understand­ BuESCH, D . C. ( 1991) Changes in depositional environ­
ments resulting from emplacement of a large-volume
ing. Ideas gathered in North Wales have benefited
ignimbrite. In: Sedimentation in Volcanic Settings (Eds
by additional research experiences in Australia Fisher, R.V . & Smith, G.A.), Spec . Publ. Soc. econ.
(Exhibition of 1851 Fellowship) and Japan (STA & Paleontol. Mineral. , Tulsa, 45, 139- 154.
AIST Fellowships) where similar environmental BuscH, D . A. ( 1971 ) Genetic units in delta prospecting.
Bull. A m . Ass. petrol. Geol. , 55, 1 137- 1 154.
problems were encountered. Ideas presented in this
CAMPBELL, C .V. ( 1976) Reservoir geometry of a fluvial
paper were greatly sharpened by thorough and sheet sandstone . Bull. Am. Assoc. petrol. Geol. , 70,
constructive reviews by Gary Smith, James White, 1009 - 1020.
and the editor, Guy Plint. CAMPBELL, S.D.G. , HowELLS, M.F. , SMITH , M . &
REEDMAN, A.J. ( 1 988) A Caradoc failed rift within the
Ordovician marginal basin of Wales. Geol. Mag. , 125,
257-266.
REFERENCES
CANT, D.J. & WALKER, R.G. ( 1 978) Fluvial processes
and facies sequences in the sandy braided South
ALLEN, J.R.L. ( 1974) Reaction, relaxation and lag in Saskatchewan River, Canada. Sedimentology , 25, 625-·
natural sedimentary systems . Earth Sci. Rev. , 10, 263- 648.
342. CAPUTO, M. ( 1 979) Two thousand years of geodetic and
ALLEN, J. R.L. ( 1 984) Sedimentary Structures: their geophysical observation in the Phlegran Fields near
Character and Physical Basis.Elsevier, Amsterdam, Naples. Geophys. J. R. astron. Soc. , 56, 3 19-328.
663pp . CAS, R.A . F. & WRIGHT, J .V. ( 1987) Volcanic Successions:
ALLEN, P.A. ( 1981) Wave-generated structures in the Modern and Ancient. Allen & Unwin, London, 528 pp .
Devonian lacustrine sediments of south-east Shetland CATHER, S.M. & FoLK, R.L . ( 199 1) Pre-diagenetic
and ancient wave conditions . Sedimentology, 28, sedimentary fractionation of andesitic detritus in a semi­
369-379. arid climate: an example from the Eocene Datil Group,
ARNOTT, R.W .C . ( 1 993) Quasi-planar-laminated sandstone New Mexico. In: Sedimentation in Volcanic Settings (Eds
beds of the Lower Cretaceous Bootlegger Member, Fisher, R .V. & Smith, G.A.), Spec. Publ. Soc. econ.
North-Cental Montana: evidence of combined-flow Paleont. Mineral. , Tulsa, 45, 2 1 1 -226.
sedimentation. J. sediment. Petrol. , 63, 488-494. CoCKS, L. R .M . & FORTEY, R.A. ( 1 982) Faunal evidence
BACHMAN, G.O . & MEHNERT, H.H. ( 1 978) New K - Ar for oceanic separations in the Palaeozoic of Britain.
dates and the late Pliocene to Holocene history of the J. geol. Soc. London, 139, 465 -478.
central Rio Grande region, New Mexico. Geol. Soc. CoLE, J .W . ( 1 984) Taupo-Totorua depression: an ensialic
Am. Bull. , 89, 283-292. marginal basin of North Island, New Zealand. In:
BALL, T.K. & MERRIMA N, R.J. ( 1989) The petrology and Marginal Basin Geology: Volcanic and Associated
geochemistry of the Ordovician Llewelyn Volcanic Sedimentary and Tectonic Processes in Modern and
Group, Snowdonia, North Wales. British Geological A ncient Marginal Basins (Eds Kokelaar, B.P. & Howells,
Survey Research Report, No. 89/ 1 . HMSO, London . M .F.), Spec . Publ . geol. Soc. London, No. 16, pp. 109-
BARDAJI , T. , DABRJO, C.J., GoY, J.L., SoMOZA, L. & 120. Geological Society of London, Bath.
ZAZO, C . ( 1 990) Pleistocene fan deltas in southeastern CoLLINSON, J . D. ( 1 969) The sedimentology of the
Iberian peninsula: sedimentary controls and sea-level Grindslow Shales and the K.inderscout Grit: a deltaic
changes. In: Coarse-grained Deltas (Eds Colella, A. & complex in the Namurian of northern England. J.
Prior, D.B.), Spec. Pubis int . Ass . Sediment., No. 1 0 , Sediment. Petrol. , 39, 194- 221 .
pp. 129- 1 5 1 . Blackwell Scientific Publications, Oxford. CORDELL, L. , LoNG, C.L. & JONES, D.W. ( 1985) Geo··
BATES, D . E .B . ( 1 972) The stratigraphy of the Ordovician physical expression of the batholith beneath Questa
rocks of Anglesey . Geol. J. , 8, 29-60. caldera, New Mexico. J. geophys. Res . , 90, 263-269.
BESLEY, B .M . & COLLIN SON , J.D. ( 199 1 ) Volcanic and COTTER, E. ( 1 978) The evolution of fluvial style, with
tectonic controls of lacustrine and alluvial sedimentation special reference to the central Appalachian Palaeozoic.
in the Stephanian coal-bearing sequence of the Malpas­ In: Fluvial Sedimentology (Ed. MiaU, A.D.), Mem . Can.
Sort Basin, Catalonian Pyrenees . Sedimentology , 38, Soc. petrol . Geol. , Calgary, 5, 361 -384.
3 - 26. CROWLEY, K.D. ( 1 983) Large scale bed configurations
BooTHROYD, J.C. & Nu MME DA L , D. ( 1978) Proglacial (macroforms), Platte River Basin, Colorado and
braided outwash: a model for humid alluvial fan deposits. Nebraska: primary structures and formative processes.
In: Fluvial Sedimentology (Ed. Miall, A.D.), Mem. Can. Geol. Soc. Am. Bull. , 94, 1 17 - 133.
Soc . petrol. Geol., Calgary, 5, 641 - 668. DAM, G. & AN DREASAN , F. ( 1990) High energy ephemeral
BRANNEY, M.J . ( 1 99 1 ) Eruption and depositional facies of stream deltas: an example from the Upper Silurian
the Whorneyside Tuff Formation, English Lake District: Holmestrand Formation of the Oslo region, Norway.
an exceptionally large magnitude phreatoplinian erup­ Sediment. Geol. , 66, 197-226.
tion. Geol. Soc. Am. Bull. , 103, 886-897. DAVIES, O . K. , ALMON, W. R. , BoNIS, S. B . & HuNTER,.
BROWNE, q.H. & PuNT, A . G . ( 1 994) Alternating braid­ B . E . ( 1 979) Deposition and diagenesis of Tertiary­
plain and lacustrine deposition in a strike-slip set­ Holocene volcaniclastics, Guatemala. In: Aspects of
ting: the Pennsylvanian Boss Point Formation of the Diagenesis (Eds Scholle, P.A. & Schluger, P.R.), Spec.
Cumberland Basin, Maritime Canada. J. sediment. Publ. Soc. econ. Paleont. Mineral. , Tulsa, 26, 281 -306.
 Facies models in volcanic terrains 191

DE RAAF, J.F.M. , READING, H. G. & WALKER, R.G. (1965) HAMBLIN , A. P. ( 1992) Half-graben lacustrine sedimentary
Cyclic sedimentation in the Lower Westphalian of North rocks of the lower Carboniferous Strathlorne Formation,
Devon, England. Sedimentology , 4, 1 -52. Horton Group , Cape Breton Island , Nova Scotia,
Dorr, R.H. (1983) 1982 SEPM presidential address: Canada. Sedimentology , 39, 263 -284.
Episodic sedimentation - how normal is average? how HASZELDI N E , R.S. & AN DERTO N , R. (1980) A braidplain
rare is rare? does it matter? I. sediment. Petrol. , 53, facies model for the Westphalian B Coal Measures of
5-23. north-east England. Nature, 284, 5 1 - 53 .
DvoRAK, J.J. & GASPARINI, P. (1991) History of earth­ HEIKEN , G. & WoHLETZ, K. (1991) Fragmentation pro­
quakes and vertical ground movements in Campi Flegrei cesses in explosive volcanic eruptions. In: Sedimentation
caldera, southern Italy: comparison of precursory events in Volcanic Settings (Eds Fisher, R.V. & Smith, G.A.),
to the AD 1538 eruption of Monte Nuovo and of activity Spec. Pub!. Soc. ccon. Paleont. Mineral. , Tulsa, 45,
since 1968. I. volcano!. geotherm. Res. , 48, 199-222. 19-26.
EwART, A. (1979) A review of the mineralogy and chemistry HENRY, C.D. & PRICE, J.G. (1984) Variations in caldera
of Tertiary- Recent dacitic , latitic, rhyolitic and related development in the Tertiary volcanic field of Trans­
sialic volcanic rocks. In: Trondhjemites, Dacites and Pecos Texas. I. geophys. Res. , 89, 8765-8786.
Related Rocks (Ed. Baker, F.), pp. 113 - 121. Elsevier, HICKS, D.M. & INMAN , D.L. (1987) Sand dispersion from
Amsterdam. an ephemeral river delta on the central California coast.
FERNANDEZ, L . P . , AGUEDA, J.A . , COLMENERO, J . R . , Mar. Geol. , 77, 305-3 18.
SALVADOR, C.!. & BARBA, P . (1988) A coal-bearing fan­ HILDRETH , W. (1987) New perspectives on the eruption of
delta complex in the Westphalian D of the Central Coal 1912 in the Valley of Ten Thousand Smokes, Katamai
Basin, Cantabrian Mountains, northwestern Spain: National Park, Alaska. Bull. Volcano!. , 49, 680-693.
implications for the recognition of humid-type fan deltas. HILL, D.P. , BAILEY, R.A. & RYALL, A.S. (1985) Active
In: Fan Deltas: Sedimentology and Tectonic Setting (Eds tectonic and magmatic processes beneath Long Valley
Nemec, W. & Steel, R.J.), pp. 286-302. Blackie, Caldera, Eastern California: an overview. J. geophys.
Glasgow. Res. , 90, 1 1 1 1 1 - 1 1 120.
FISHER, R.V. (1961) Proposed classification of volca­ HILL, P.R. & NADEA U , O . C. (1989) Storm dominated
niclastic sediments and rocks. Geol. Soc. Am. Bull. , 72, sedimentation on the inner shelf of the Canadian
1395- 1408. Beaufort Sea. I. sediment. Petrol. , 59, 455-468.
FISHER, R. V. & SCHMINCKE, H.U. (1984) Pyroclastic Rocks. HowELLS, M.F. & LEVERIDGE, B.E. (1980) The Capel
Springer-Verlag, Berlin, 472 pp. Curig Volcanic Formation. Institute of Geological
FosHAG , W.F. & GoNZALEZ-R, J. ( 1956) Birth and devel­ Sciences Report, No. 80/6, pp. 1 -23. HMSO, London.
opment of Paricutin volcano, Mexico. U.S. geol. Surv. HoWELLS, M . F. , REEDMAN, A.J. & CAMPBELL, S.D.G.
Bull. , 965-D, 355-489. (1991) Ordovician (Caradoc) Marginal Basin Volcanism
FRITZ, W.J. (1991) Theoretical modelling of large wave in Snowdonia (North- west Wales) . British Geological
ripples in volcaniclastic sediments, Ordovician Llewelyn Survey, HMSO, London, 191 pp.
Volcanic Group , North Wales. Sediment. Geol. , 74, HunoN, J. (1788) Theory of the Earth. Trans. R. Soc.
241 -250. Edinburgh , l , 209-305.
FRITZ, W.J. & HOWELLS, M.F. (1991) A shallow marine JOPLING, A.V. (1965) Hydraulic factors controlling
volcaniclastic facies model: an example from sedimentary the shape of laminae in laboratory deltas. I. sediment.
rocks bounding the subaqueously welded Ordovician Petrol. , 35, 777-791.
Garth Tuff, North Wales, U . K. Sediment. Geol. , 74, KARNER, G.D. & DEWEY, J .F. (1986) Rifting: lithospheric
217-240. versus crustal extension as applied to the Ridge Basin of
FORSICH, F.T. & MAYR , H. (1981) Non-marine Rhizo­ southern California. In: Future Petroleum Provinces of
corallium (trace fossil) from the Upper Freshwater the World (Ed. Halbouty, M.T.), Mem. Am. Assoc.
Molasse (Upper Miocene) of southern Germany. Neues petrol. Geol., Tulsa, 40, 317-337.
Jahrb. Geol. Palaeontol. Monatsh. , 198 1 , 321 -333. KOKELAAR, B.P. (1988) Tectonic controls of Ordovician
GALLOWAY, W.E. (1981) Depositional architecture of arc and marginal basin volcanism in Wales. I. geol. Soc.
Cenozoic Gulf Coast plain fluvial systems. In: Recent London, 145, 759-775.
and Ancient Non-marine Environments: Models for KoKELAAR, B.P. (1992) Ordovician marine volcanic and
Exploration (Eds Ethridge, F. G. & Flores, R.M.), Spec. sedimentary record of rifting and volcanotectonism:
Pub!. Soc. econ. Paleont. Mineral., Tulsa, 3 1 , 127- 155. Snowdon, Wales, UK. Geol. Soc. Am. Bull. , 104,
GIBBONS, W. (1983) Stratigraphy, subduction and strike­ 1443 - 1455.
slip faulting in the Mona Complex of North Wales - a LECKIE , D.A. & CHEEL, R.J. ( 1989) The Cyprus Hills
review. Proc. Geol. Assoc. , 94, 147 - 164. Formation (upper Eocene to Miocene): a semi-arid
GoDIN, P.D . (1991) Fining-upward cycles in the sandy braidplain resulting from intrusive uplift. Can. I. Earth
braided-river deposits of the Westwater Canyon Member Sci. , 26, 1918-1931.
(Upper Jurassic), Morrison Formation, New Mexico. LEEDER, M. R. ( 1983) Lithospheric stretching and North
Sediment. Geol. , 70, 6 1 - 82. Sea clastic sourcelands. Nature, 305, 510-513.
GouLD, S.J. (1987) Time's Arrow, Time's Cycle: Myth and LIPMAN , P.W. (1984) Roots of ash-flow calderas in western
Metaphor in the Discovery of Geological Time. Penguin North America: windows into the tops of granitic batho­
Books, New York, 222 pp. liths. I. geophys. Res. , 89, 8801 -8841.
HACKEIT, W.R. & HouGHTON , B.F. (1989) A facies model LowE, D.R. ( 1982) Sediment gravity flows: II. Depo­
for a Quaternary andesitic composite volcano: Ruapehu, sitional models with special reference to the deposits of
New Zealand. Bull. Volcano!. , 5 1 , 5 1 - 68. high-density turbidity currents. J. sedime'ru. Petrol. , 52,
192 G.J. Orton

279-297. 'offshore bars' in the Cardium Formation of Alberta :

LuONGO, G. & SCAN DONE, R. ( 1991) Campi Flegrei. their relationship to relative changes in sea level. I n :
J. Volcano!. Geotherm. Res. , 48, 227 pp. Sea-level Changes: a n Integrated Approach (Eds Wilgus ,
LYELL, C . ( 1853) Principles of Geology 9th edn. John C . K . , Hastings, H . W . , Kendall, C . , Posamentier, H . W . ,
Murray, London, 835 pp. Ross, C . A . & Van Wagoner, J . C . ) , Spec. Pub!. Soc.
MA.tzELS, J. ( 1989) Sedimentology, palaeoflow dynamics, econ. Paleont. Mineral . , Tulsa, 42, 357-370.
and flood history of jokulhlaup deposits: palaeo­ POINTON , C . R . & IXER, R . A . ( 1 980) Parys Mountain
hydrology of Holocene sediment sequences in southern mineral deposit, Anglesey, Wales: geology and ore
Iceland sandur deposits. J. sediment. Petrol. , 59, mineralogy. Trans. Inst. Min. Metal/. , 89, 143 - 155 .
204-223. POLLARD, J . E . , STEEL, R . J . & UNDERSUD, E. ( 1 982) Facies
McKEE, C . P . , LOWENSTEI N , P . L . , DESAINTOURS, P . , TALA!, sequence and trace fossils in lacustrine fan delta deposits ,
B . , htKARAI, I. & Motu, J . L . (1984) Seismic and ground Hornelen basin, Middle Devonian, Western Norway.
deformation crises at the Rabaul caldera: prelude to an Sediment. Geol. , 32, 63-87.
eruption? Bull. volcano!. , 47, 397 -41 1 . PRIOR, D . B . , BORNHOLD, B . D . , WiSEMA N , W . J . & LOWE,
McPHIE, J . & ALLEN, R . L . ( 1992) Facies architecture of D . R . ( 1987) Turbidity current activity in a British
mineralized submarine volcanic sequences: Cambrian Columbian fjord. Science, 237, 1330- 1333.
Mount Read Volcanics, Western Tasmania. Econ. Geol. , REEDMAN, A .J . , LEVERIDGE, B . E . & EVANS, R . B . ( 1 984)
87, 587-596. The Arfon Group ('Arvonian') of North Wales. Proc.
MlALL, A . D . (1977) A review of the braided river depo­ Geol. Assoc. , 95, 3 1 3 - 32 1 .
sitional environment . Earth Sci. Rev. , 13, 1 - 16. REILINGER, R . E . & YORK, J . E . ( 1 979) Relative crustal
MlALL, A . D . & GlBLI N G , M . R . ( 1978) The Siluro-Devonian subsidence from levelling data in a seismically active part
clastic wedge of Somerset Island, Arctic Canada, and of the Rio Grande rift, New Mexico . Geology , 7 ,
some regional implications. Sediment. Geol. , 2 1 , 85 - 139 - 143.
127 . ROWLEY , P . D . , KUNTZ, M . A .
MACLEO D , N . S . (1 981)
&
MIDDLETON , G . V . (1973) Johannes Walther's Law of the Pyroclastic-flow deposits. The 1 980 Eruptions of
In:
correlation of facies. Geol. Soc. Am. Bull. , 84, 979-988. Mount St. Helens (Eds Lipman, P . W . & Mullineaux,
MYROW, P . M . ( 1992) Bypass-zone tempestite facies models D . R. ) . U . S . geol. Surv. Prof. Pap . , 1250, 489 - 5 12 .
and proximality trends for an ancient muddy shoreline RUEGG, G.H. ( 1971) Features of middle Pleistocene
and shelf. J. sediment. Petrol. , 62, 99 - 1 15. sandur deposits in the Netherlands. Geol. Mijnbouw, 56 ,
MYROW, P.M. & SouTHARD, J.B. ( 1991) Combined-flow 5 - 24.
model for vertical stratification sequences in shallow RuNKEL, A . C . ( 1990) Lateral and temporal changes in
marine storm-deposited beds. J. sediment. Petrol. , 6 1 , volcanogenic sedimentation; analysis of two Eocene
202-210. sedimentary aprons, Big Bend Region , Texas. J.
NELSON, C.H. ( 1 982) Modern shallow-water graded sand sediment. Petrol. , 60, 747-760.
layers from storm surges, Bering shelf: a mimic of RusT, B . R . & GtBLI N G , M . R . ( 1990a) Braidplain evolution
Bouma sequences and turbidite systems . .1. sediment. in the Pennsylvanian South Bar Formation near Sydney,
Petrol. , 52, 537-545. Nova Scotia. J. sediment. Petrol. , 60, 59-72.
NEWHALL, C.G. & DzuRJSIN, D. ( 1988) Historical unrest RusT, B.R. & GtBLIN G , M.R. (1990b) Three-dimensional
at large calderas of the world . U . S . geol . Surv. Bull . , antidunes as HCS mimics in a fluvial sandstone: the
1855, v . 1 , 598 pp. Pennsylvanian South Bar Formation near Sydney, Nova
ORTO N , G . J . ( 1988) A spectrum of mid-Ordovician fan and Scotia. J. sediment. Petrol. , 60, 540-548.
braidplain deltaic sequences, North Wales: a conse­ SHERIDAN , M.F. & UPDIKE, R.G. ( 1975) Sugarloaf
quence of varying fluvial clastic input. I n : Fan Deltas: Mountain tephra - a Pleistocene rhyolitic deposit of
Sedimentology and Tectonic Setting (Eds Nemec, W. & base-surge origin in Northern Arizona. Geol. Soc. Am.
Steel, R . J . ) , pp. 23-49. Blackie, Glasgow. Bull. , 86, 5 7 1 - 5 8 1 .
ORTON , G.J. ( 1 990) Volcaniclastic sedimentation in a SIMPSON , E . L . & ERIKSSON , K . A . ( 1 989) Sedimentology of
Caradocian marginal basin, North Wales . Unpublished the U nicoi Formation in southern and central Virginia:
PhD thesis, Oxford U niversity. evidence for late Proterozoic to Early Cambrian rift-to··
ORTO N , G .J. ( 1991) Emergence of subaqueous depositional passive margin transition. Geol. Soc. Am. Bull. , 101 ,
environments in advance of a major ignimbrite eruption, 42-54.
Capel Curig Volcanic Formation, North Wales: an SLOAN, R.J. & WILLIAMS, B .P . J . ( 199 1) Volcano-tectonic
example of regional volcanotectonic uplift? Sediment. control of offshore to tidal flat regressive cycles from the
Geol. , 74, 25 1 - 288. Dunquin Group (Silurian) of southwest Ireland. In:
PtCKERJLL, R . K . ( 1976) Vermiforichnus borings from the Sedimentation, Tectonics and Eustacy: Sea-level Changes
Ordovician of central Wales. Geol. Mag. , 1 13, 159- 164. at Active Plate Margins (Ed . Macdonald , D . I . M. ) , Spec.
PICKERJLL, R . K . & BRENCHLEY, P.J. ( 1979) Caradoc marine Pub!. int. Ass. Sediment . , No. 12, pp. 105 - 1 19 .
benthic communities of the south Berwyn Hills, North Blackwell Scientific Publications, Oxford .
Wales. Palaeontology , 22, 229-264. SMITH, G . A . ( 1986) Coarse-grained nonmarine volcani··
PtCKERJLL, R . K . & H uRST, J . M . ( 1983) Sedimentary facies, clastic sediment: terminology and depositional process.
depositional environments, and faunal associations Geol. Soc. Am. Bull. , 97, 1 - 10.
of the lower Llandovery (Silurian) Beechill Cove SMITH, G.A. ( 1987) The influence of explosive volcanism
Formation, Arisaig, Nova Scotia. Can. J. Earth Sci. , 20, on fluvial sedimentation: the Deschutes Formation
1761- 1779. (Neogene) in Central Oregon. J. sediment. Petrol. , 57 ,
PuNT, A . G . ( 1988) Sharp-based shoreface sequences and 6 1 3 - 629.
 Facies models in volcanic terrains 193

SMITH , G.A. (1988) Sedimentology of proximal to distal South Africa. Sediment. Geol. , 18, 161 - 174.
volcaniclastics dispersed across an active foldbelt: Ellens­ WALKER, G . P . L . (1984) Down-sag calderas, ring faults,
berg Formation (late Miocene) , central Washingto n . caldera sizes, and incremental caldera growth. J. geophys.
Sedimentology, 3 5 , 953-977. Res. , 89, 8407-8416.
SMITH, G.A. ( 1991) Facies sequences and geometries i n WALKER, G . P . L . ( 1989) Gravitational (density) controls on
continental volcaniclastic sequences. I n : Sedimentation volcanism , magma chambers and intrusions. Aust. J.
in Volcanic Settings (Eds Fisher, R.V. & Smith, G . A . ) , Earth Sci. , 36, 149-165.
Spec. Pub!. Soc. eco n . Paleontol. Mineral . , Tulsa, 45, WALKER, R.G. ( 1 990) Facies modelling and sequence stra­
109 - 122. tigraphy. J. sediment. Petrol. , 60, 777-786.
SMITH , R . C . M . ( 1991) Landscape response to a major WALKER, R . G . ( 1 992) Facies, facies models and modern
ignimbrite eruption , Taupo volcanic center, New stratigraphic concepts. I n : Facies Models: Response to
Zealand . In: Sedimentation in Volcanic Settings (Eds Sea Level Change (Eds Walker, R . G . & James, N . P . ) ,
Fisher, R . V . & Smith, G . A . ) , Spec. Pub!. Soc. econ . p p . 1 - 14 . Geological Association o f Canada, St John's.
Paleontol. Mineral . , Tulsa, 45, 123 - 138. WALKER, R . G . & HARMS, J . C . ( 1975) Shorelines of weak
SoPER, N.J. & HurroN , D .H . W . ( 1 984) Late Caledonian tidal activity: Upper Devonian Catskill Formation,
sinistral displacements in Britain: implications for a three Central Pennsylvania. In: Tidal Deposits: a Casebook of
plate collision model. Tectonics, 3, 791 - 794. Recent Examples and Fossil Counterparts (Ed. Ginsburg,
STEEL, R.J. & AASHEIM, S . M . ( 1978) Alluvial sand depo­ R . N . ) , pp. 103 - 108. Springer-Verlag, New York.
sition in a rapidly subsiding basin, Devonia n , Norway. WALTON, A . W . ( 1 986) Effect of Oligocene volcanism on
In: Fluvial Sedimentology (Ed. Miall, A . D . ) , Mem. Can. sedimentation i n the Trans-Pecos volcanic field of Texas.
Soc. petrol. Geol . , Calgary, 5, 386-412. Geol. Soc. Am. Bull. , 97, 1 192- 1207.
SuRDAM, R.C. & BoLES, J . R . ( 1979) Diagenesis of volcanic WATKINS, R.T. ( 1986) Volcano-tectonic control on
sandstones. I n : Aspects of Diagenesis (Eds Scholle, P . A . sedimentation i n the Koobi Fora sedimentary basin. In:
& Schluger, P . R . ) , Spec. Pub!. Soc. econ. Paleontol. Sedimentation in the African Rifts (Eds Frostick, L . E . ,
Mineral . , Tulsa, 26, 227-242. Renault, R . W . , Reid, I . & Tiercelin, J . - J . ) , Spec. Pub!.
THORPE, R . S . ( 1979) Late Precambrian igneous activity in geol. Soc. London, No. 25 , 85-98. Geological Society
Southern Britain . I n : The Caledonides of the British of London, Bath.
Isles - Reviewed (Eds Harris, A . L . , Holland , C . H . & WAITS , A . B . ( 1989) Lithospheric flexure due to prograding
Leake, B . E . ) , Spec. Pub!. geol. Soc. London, No. 8 , sediment loads: implications for the origin of offlap/
p p . 579-584. Geological Society of London, Bath. on lap patterns in sedimentary basins. Basin Res. , 2 ,
ToRSVJK, T.H. & TRENCH, A . ( 1991) The Ordovician history 133 - 144.
of the Iapetus Ocean in Britai n : new palaeomagnetic WELLS, J.T. & COLEMA N , J . M . ( 1981) Physical dynamics
constraints. J. geol. Soc. London , 148, 423 -425. and fine-grained sediment dynamics, coast of Surinam,
TRYTHALL, R.J.B . , EccLES , C., MOLYNEUX, S.G. & South America. J. sediment. Petrol. , 51, 1053 - 1063.
TAYLOR, W . E . G . ( 1987) Age and controls of ironstone WELLS , N.A. ( 1983) Transient streams i n sand-poor
deposition (Ordovician) North Wales. Geol. J. , 22, redbeds. In: Modern and Ancient Fluvial Systems (Eds
3 1 - 43 . Colliso n , J . D . & Lewin , J . ) , Spec. Pubis int. Assoc.
TUNNICLIFF, S . P . & RuSHTO N , A . W . A . ( 1980) Faunas from Sediment . , No . 6, pp. 393-403. Blackwell Scientific
sediments associated with the Capel Curig Tuffs: One­ Publications, Oxford .
inch sheet 106 (Bangor). British Geological Survey WELLS, S . G . , McFADDEN , L . D . , RENAULT, C . E . & CROWE,
Open File Report, PD 80/12. B . M . (1990) Geomorphic assessment of late Q uaternary
TuRBEVILLE, B.N. (1991) The influence of ephemeral volcanism in the Yucca Mountain area, southern
processes on pyroclastic sedimentation in a rift-basin, Nevada: implications for the proposed high-level radio­
volcaniclastic- alluvial sequence, Espanola basin, New active waste disposal . Geology, 18, 549-553.
Mexico. Sediment. Geol. , 74, 139 - 155 . WESTHEAD, S .J . (1991) Prospects at Parys Mountains.
VAN DER Voo, R. (1988) Paleozoic paleogeography of Geol. Today , 7 , 130 - 135.
North America, Gonwana, and intervening displaced WHITE, J . D . L. (1991) The depositional record of small,
terranes: comparisons of paleomagnetism with paleo­ monogenetic volcanoes within terrestrial basins. I n :
climatology and biogeographic patterns. Geol. Soc. Am. Sedimentation in Volcanic Settings ( E d s Fisher, R . V . &
Bull. , 100, 3 1 1 -324. Smith, G . A . ) , Spec. Pub!. Soc. econ. Paleontol.
VAN HouTE N , F . B . (1976) Late Cenozoic volcaniclastic Mineral . , Tulsa, 45, 155- 174.
deposits, Andean foredeep, Columbia. Geol. Soc. Am. WHITE, J.D.L. & ROBINSON, P.T. ( 1992) Intra-arc
Bull. , 87, 481-495. sedimentation in a low-lying marginal arc, Eocene Clarno
VESSELL, R.K. & DAVIES, D.K. ( 1981) Non-marine Formation, central Oregon. Sediment. Geol. , 80,
sedimentation i n an active fore-arc basin. In: Recent 89 - 1 14 .
and Ancient Non-marine Environments: Models for WiLLIAMS, G.E. ( 1971) Flood deposits of sand bed
Exploration (Eds Ethridge, F . G . & Flores, R . M . ) , Spec. ephemeral streams of central Australia. Sedimentology ,
Pub!. Soc. eco n . Paleontol. Mineral . , Tulsa, 3 1 , 3 1 - 45 . 17, 1 -40.
V!SHER, G . S . (1965) Use o f the vertical profile i n environ­ WRIGHT, L .D . , YANG, Z . S . , BORNHOLD, B . D . , KELLER,
mental reconstruction. Bull. Am. Assoc. petrol. Geol. , G . H . , PRIOR, D . B . & WiSEMAN , W . J . ( 1986) Hyper­
49, 4 1 - 6 1 . pycnal plumes and plume fronts over the Huanghe
Vos, R . G . & ERIKSSON , K . A . ( 1 977) An embayment (Yellow River) delta front . Geomarine Letters, 6, 97-
model for tidal and swash deposits occurring within 106.
a fluvially dominated middle Proterozoic sequence in
 Tectonics and Sedimentation

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
Spec. Pubis int. Ass. Sediment. (1995) 22, 197-217

Coarse-grained lacustrine fan-delta deposits (Pororari Group)

of the northwestern South Island, New Zealand:
evidence for Mid-Cretaceous rifting

M A L C O L M G. LAIRD
Institute of Geological and Nuclear Sciences, c/o Department of Geology,
University of Canterbury, Christchurch, New Zealand

A BS T R A C T

Mainly coarse-grained non-marine sediments of the Pororari Group, dominated b y breccia and
conglomerate, and of mid- to late Albian age, are widespread west of the Alpine Fault in the South
Island of New Zealand. They represent an unconformity-bounded sequence, resting with marked
angular discordance on Early Palaeozoic sediments of the Greenland Group or non-conformably ori
mid-Palaeozoic and Early Cretaceous granites, and overlain with regional discordance by coal measures
of Maastrichtian age. In the southern Paparoa Range, where the Pororari Group is in excess of 2000 m
thick, it can be divided into three major facies assemblages: (1) matrix- and clast-supported breccia; (2)
pebbly sandstone and massive or graded sandstone; and (3) laminated dark mudstone with thin graded
sandstones. Facies assemblage 1 is inferred to represent the deposits of debris flows and sheet flows on
alluvial fans; facies assemblage 2, the deposits of debris flows and turbidites in a dominantly lacustrine
environment; and facies assemblage 3 was deposited in a distal lacustrine environment. The three facies
assemblages intertongue, and are inferred to represent elements of a fan-delta system debouching into a
lake.
The palaeocurrent pattern deduced from clast imbrication in breccia units, and from sedimentary
structures in the lacustrine deposits, indicates a dominantly NNE-SSW direction of sediment transport.
The consistent and rapid thinning of lithostratigraphical units from north to south is compatible with the
palaeocurrent results. Palaeocurrent plots and clast lithologies suggests that several point sources
supplied coarse sediment during· the history of deposition of the Pororari Group, each giving rise to
distinct but overlapping fans.
The great thickness of coarse deposits (a minimum of 1200 m for the breccia assemblage alone) and
the angularity of clasts suggests that these sediments were generated by contemporaneous tectonic
activity. Fluctuations in lake level also may have been controlled by active faulting at the basin margin,
although climatic changes may have played a part. The fan-delta system is likely to have occupied an
active subsiding half-graben. The consistently south-southwest directed sediment gravity flows suggest
that the active fault was likely to have been oriented in a WNW-ESE direction, downthrown to the
south, and with an implied north-northeast direction of tectonic extension. A similar extension direction
for the Paparoa Range area during early. Albian times has been deduced from independent structural
research on a closely associated metamorphic core complex, and is also compatible with the WNW­
E SE orientation of normal faults bounding adjacent offshore undrilled half-grabens infilled with
deposits of prdbably mid-Cre��ceous age. The coarse deposits of the Pororari Group are likely to have
been associated with an early period of rifting in the Albian, pre-dating by perhaps 25 Ma the separation
of New Zealand from Gondwana in the late Campanian or early Maastrichtian.

INTRODUCTION

Coarse-grained non-marine sediments, o f mainly conglomerate, are widespread throughout the

mid- to late Albian age, dominated by breccia and southern and western South Island of New Zealand.

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0

197
198 M.G. Laird

Most are poorly exposed, and detailed descriptions spread compressional deformation, metamorphism,
are sparse. One exception is the Kyeburn Formation, uplift, erosion and calc-alkaline plutonism.
which lies in the southeast of the South Island Over much of New Zealand the rocks of the
(Bishop & Laird, 1976; Fig. 1), and which represents Rangitata Orogen are separated by a major uncon­
talus to braided river and lacustrine deposits of mid­ formity from younger, less-deformed strata. This
Albian to at least late Cenomanian age, emplaced unconformity marks the change in tectonic regime
in a locally developed, triangular, fault-bounded from convergent margin to extension, attributed by
depression. Another well-described exception is the Bradshaw (1989) to collision with the New Zealand
Puysegur Formation of the southwestern South margin of the spreading ridge between the Phoenix
Island (Lindqvist, 1990; Fig. 1), which is inferred to and Pacific Plates. Mid Cretaceous extensional tec­
have been deposited in terrestrial, lake margin, and tonism is best exemplified by the post-unconformity
lacustrine environments in a fault-controlled basin. development, in the mid Cretaceous, of half grabens
All of these mid-Cretaceous sediments rest uncon­ throughout much of the New Zealand region, both
formably on older rocks, and clearly post-date the onshore and offshore. Subsidence continued along
Early Cretaceous Rangitata Orogeny, which marked the bounding faults in the western and southern
a significant change of tectonic pattern in the New South Island, at least in some instances, from mid­
Zealand region. From the Permian to the Early Albian until Santonian or early Campanian times
Cretaceous, most rocks formed under the influence (Laird, 1993, 1994). Both the Kyeburn and the
of convergent-margin tectonics and comprise incom­ Puysegur Formations are inferred to occupy basins
plete remnants of magmatic arcs, fore-arc basins, formed east of the Alpine Fault in mid-Albian times
trench-slope basins, .;:tnd accretionary complexes under the extensional tectonic regime.
(Bradshaw, 1989). The later stages of this regime, in On the west coast of the South Island, west of the
the Early Cretaceous, were accompanied by wide- Alpine Fault, outcrops of breccia (collectively known

Kyeburn
•

Fig. 1. Locality map of the South

Island of New Zealand showing ( in
black) the main outcrop areas of
100 km mid-Cretaceous non-marine coarse
Puysegur clastic deposits. B, Beebys
Point
Conglomerate.
 Cretaceous rift deposits of New Zealand 199

as the Hawks Crag Breccia, forming part of the G E OL O G IC A L S E T TI N G

Pororari Group) are scattered but widespread over a
NE-SW extent of approximately 320 km (Fig. 1). The Pororari Group comprises an unconformity­
The breccia is matrix- or clast-supported with a bounded sequence resting on Early Palaeozoic
sparse sandy matrix, and is often reddish-brown in turbidites of the Greenland Group, or on mid­
colour. Structure is generally simple, most suc­ Palaeozoic and Early Cretaceous granitoids.
cessions dipping uniformly at angles up to 50°, or It is overlain by the Paparoa Coal Measures of
folded into simple, open synclines. Thickness of the Maastrichtian age. In the southern Paparoa Range,
Pororari Group varies considerably, but a maximum the Pororari Group crops out in two extensive areas
is c. 4 500m. on the eastern and western flanks of the range, and
The age of the group has been determined solely in small exposures in Fox River to the north (Fig. 2).
from pollen. Palynological studies (Raine, 1984) In the western outcrops, four lithostratigraphic units
indicate that the bulk of the Pororari Group lies can be recognized (Laird, 1988; Figs 2 & 3): (1) a
within the Lycopodiacides bullerensis assemblage, locally occurring basal unit, dominated by inter­
with the youngest rocks reaching into the lower part bedded matrix- and clast-supported breccia, massive
of the overlying Trichotomosulcites subgranulatus or normally graded sandstone, and laminated mud­
assemblage. Absence of angiosperms suggests that stone (Watson Formation), followed gradationally
the Pororari Group lies within the New Zealand by (2), thick-bedded matrix- and clast-supported
Motuan Stage (mid- to late Albian). The original breccia (Hawks Crag Breccia), followed by (3), inter­
upward time range of the Group in the region is bedded thin-bedded sandstone and matrix-supported
unknown, being constrained only by the local occur­ breccia and conglomerate beds (Bovis Formation),
rence of unconformably overlying coal measures of which are in turn overlain by (4), laminated carbon­
Maastrichtian age. However, a possible correlative aceous mudstone (Bullock Formation). The eastern­
of the Pororari Group, lying c. lOOkm northeast of outcrops lack the basal unit, and Hawks Crag Breccia
the Paparoa Range (Beebys Conglomerate; Fig. 1) rests directly on basement rocks.
has yielded a poorly constrained microfloral age The landscape is extremely rugged, and covered
in the range late Cenomanian to early Santonian with dense temperate rainforest (Fig. 4). This limits
(Johnston, 1990), indicating that Pororari Group the accessible and usable exposure to the gorges and
deposition may have extended locally well into the tributaries of the main streams, which have provided
Late Cretaceous. the measured stratigraphical sections (Fig. 3).
The origin and depositional setting of the Hawks
Crag Breccia has been the object of considerable
speculation in the past: explanations have included D E S C RIPT I O N A N D I N T E R PR E T A TI O N
glacial or glaciofluvial action (McKay, 1883), talus OF FACIES A S S E MBLA G E S
or alluvial fan formation (Morgan & Bartrum, 1915),
or the deposits of air-lubricated landslides (Korsch Three facies assemblages are recognized within the
& Wellman, 1988). In most occurrences, these southern Paparoa Range outcrops: (1) matrix- and
breccia units make up the entire exposure, and the clast-supported breccia; (2) pebbly sandstone, and
mechanism of emplacement and geometry of the massive or graded sandstone; and (3) laminated
bodies has been difficult to interpret because of their dark mudstone with thin graded sandstones.
apparently uniform nature over large areas. How­
ever, in the Paparoa Range in the northwest of the
Assemblage 1: matrix- and clast-supported breccia
South Island, where the most extensive areas of
outcrop occur, other lithofacies are also present,
Facies descriptions
and these throw light on sedimentary processes and
environments of deposition of both the Hawks Crag This facies assemblage makes up the entire Hawks
Breccia and of the Pororari Group as a whole. This Crag Breccia, and also occurs within the Watson
study examines the sedimentology of the Pororari Formation, both as the basal 165m exposed in Tin­
Group in the southern Paparoa Range, where the dale Creek and as a 100-m-thick unit near the base
breccia lithofacies is associated with fine-grained of the formation in Pororari River (Fig. 3). Breccia
deposits (Fig. 2). units increase in frequency near the top of the
Watson Formation, which passes gradationally
200 M.G. Laird

LJ Younger strata N

; Unconformity
Pororari Group

Bullock Formation
Bovis Formation
Hawks Crag Breccia
t

Watson Formation

Unconformity
0 Basement
_.- Fault

0 2 3 4 5 km

Fig. 2. Map of part of the southern Paparoa Range showing Pororari Group outcrop and palaeocurrent plots. See Fig. 12
for key to the palaeocurrent plots.

upwards into the Hawks Crag Breccia. Maximum The assemblage consists dominantly of poorly
thickness of the facies assemblage is hard to estimate, sorted matrix-supported breccia consisting of sub­
but in the Bullock Creek area in the north, where it angular to subrounded clasts (often outsize) of base­
is thickest, it is likely to reach 1500 m. ment material in a coarse sandy matrix (Fig. 5).
 Cretaceous rift deposits of New Zealand 201

PAPAROA
N s
COAL Bullock Slaty
MEASURES
M Creek M ·Creek

BULLOCK
FORMATION

1 ''LACUSTRINE
BOVIS
(distal) Pororari Facies
FORMATION

M River Tindale assemblage 3

120D-fi:I:i'Iii ''LACUSTRIN E '
Creek 1000
_
'
(proximal) 2000 M

1800 I-
ALLUVIAL
-�<(
-' -

�0
a._ LL
<(cr:
l9cr: (/)(.9:'!
<(t:lw --- -w

<(I�cr:
FAN

Ugj �
I

<(cr:cr: /
Facies
assemblage 1
0cr: /
0a._
Obscurred by
Quaternary
' deposits
800
r---- I
I
I
LACUSTRINE I
I KEY
600 Clast-supported breccia
z I
zo I Pebby sandstone
ALLUVIAL
I-<(
5;i=
<(� FAN I
Massive or graded sandstone
Mudstone
s:�LL I
I

LACUSTRI NE I
I

GREENLAND ° Fault
GROUP/
GRAr-.'ITE
+------ 3·2 km-------1-8 km--+

Fig. 3. Stratigraphical columns showing relationships between the major formations and facies assemblages of the Pororari
Group.

Fig. 4. Hogback and cliffed

topography developed on Hawks
Crag Breccia (Pororari Group),
with steep east-facing dip slopes.
The prominently bedded unit in the
upper middle of the photograph is
20-30m thick. The valley of the
upper Pororari River is in the
foreground: crest of the Paparoa
Range is in the background.
202 M.G. Laird

Fig. 5. Thick matrix-supported

(upper) and clast-supported (lower)
breccia beds, Fox River mouth.
Beds young towards the left.
Hammer for scale . Photograph
courtesy of L. Homer.

Fig. 6. Succession of thin clast­

supported breccia beds, Fox River
mouth. Note normal grading in the
bed below the geological hammer.
Some of these beds may represent
the deposits of highly sediment­
charged sheet flows. Beds young
upwards. Hammer for scale.

Clast-supported breccia makes up a minority of the less than 1m thick but reaching a maximum of 3 m in
beds (Fig. 5). Breccia beds most commonly lie in the thickness, and separated by thin (up to 10 em) beds
range 1 -6 m thick, but some breccia units (probably of poorly sorted coarse to pebbly sandstone (Fig. 6).
composite) may reach 20-30 m in thickness and Poorly developed symmetrical grading, reverse
extend with little thickness change for at least several grading, or normal grading occurs locally (Figs 6 &
hundred metres (Fig. 4). Most lack internal stratifi­ 7), although this is uncommon. Individual clasts
cation and bedding is in many cases indistinct. Mud­ reach a maximum size of 3m, and in some instances
stone is absent, and thin sandstone beds up to 20 em show imbrication and preferred orientation of long
thick occur only rarely. In rare instances clast­ axes.
supported breccia forms well-defined beds, usually Carbonaceous material is common, particularly at
 Cretaceous rift deposits of New Zealand 203

Fig. 7. Thick, matrix-supported

breccia bed, with thin sandstone bed
at base, Pororari River. Note lack of
internal structure, but the presence
of symmetrical grading. Beds young
to left. Hammer for scale.

the interfaces between beds, and small coalified logs Cenozoic alluvial fans (cf. Ballance, 1984; Blair,
occur rarely. At the mouth of Fox River several 1987), and attributed to the deposits of sheet flows
thick beds of poorly sorted breccia show crude emplaced during catastrophic floods. The breccia
fining-upward grading into thin graded sandstone beds at Fox River mouth, which pass upwards into
(in some instances showing synsedimentary defor­ graded sandstone and then into coaly material, may
mation structures) and then into coaly material. No represent the results of flash floods and the delayed
root horizons have been recognized, and all plant settling out of fines, with transported vegetation
material appears to have been transported from its being the last to settle.
source. Shallow scours or channels, up to 1m deep, Assemblage 1 breccias are thus inferred to have
occur locally. Other sedimentary structures, apart been deposited on a series of alluvial fans, dominated
from rare synsedimentary deformation features, are by debris-flow depositional processes, closely similar
notably absent. to the humid-temperate climate, late Quaternary
and modern alluvial fans that are characteristic of
the mountainous regions of New Zealand (cf. Pierson,
Interpretation
1980, 1981; McArthur, 1987) and elsewhere (e.g.
The preponderance of poorly sorted, poorly strati­ Kochel & Johnson, 1984). The abundance of debris
fied matrix-supported breccia with common outsize flows, and the poor sorting and angularity of many
clasts, a subordinate proportion of clast-supported of the clasts suggests that source slopes were steep
breccia, the rarity of sedimentary structures within and lay in close proximity. The common occurrence
breccia beds, and the absence of cross-stratification of carbonaceous material, including rare coalified
and only minor presence of thin well-bedded sand­ logs, indicates that the source slopes were vegetated.
stones, suggests that mass wasting events played a No fine-grained subfacies potentially representa­
predominant role in the formation of assemblage 1. tive of interdistributary deposits (cf. Lewis &
The major facies of matrix-supported breccia closely Ekdale, 1991) has been recognized within the assem­
resembles the deposits of debris flows, which domi­ blage, and fans were therefore likely to have been
nate sedimentary processes on many modern alluvial coalescing.
fans, whereas the subordinate clast-supported
breccia is characteristic of associated debris flow
A ssemblage 2: pebbly sandstone and
levees (cf. Pierson, 1980; Blair & McPherson, 1992).
massive or graded sandstone
Thin, moderately well-sorted sandstone beds and
thin-bedded breccia and sandstone couplets, forming
Facies descriptions
a minor proportion of assemblage 1, closely resemble
facies prominently represented in some modern and This facies assemblage makes up most of the Watson
204 M.G. Laird

and the Bovis Formations. It consists dominantly of tures, and most are ungraded. Clasts commonly
matrix-supported conglomerate and pebbly sand-. occupy specific horizons within each bed, but random
stone, graded or massive very fine-grained to coarse­ distribution also occurs. Clasts are of basement gran­
grained sandstone, and alternating sandstone and ite or Palaeozoic sandstone, and range up to 0.5 m in
mudstone successions. Thickness is variable, but an size. Imbrication is relatively common (Fig. 8). Inter­
aggregate thickness of 600 m is reached in the Watson leaved very fine- to coarse-grained, moderately well­
Formation exposed in Pororari River, and a sorted sandstones vary in thickness up to 1 m, and
maximum thickness for the Bovis Formation is include both graded and ungraded beds (Fig. 9).
approximately 400 m (in Slaty Creek}. Clasts in the Some have scoured bases, and show sole marks
conglomerate and pebbly sandstone of assemblage 2 consisting mainly of grooves and less common prod
are notably better rounded than in the breccia of marks or flutes. Primary current lineations and cur­
assemblage 1, and all beds are matrix-supported. rent ripples are also present on bedding planes and
Conglomerate and pebbly sandstone beds, which parallel-lamination is common in the finer sandstone
are up to 2 m thick, lack internal sedimentary struc- beds. Isolated horizons of current-rippled sandstone

Fig. 8. Imbricated pebbles of

granite and of Greenland Group
sediments supported by a matrix of
very coarse to gritty sandstone,
Bovis Formation, Slaty Creek
(at approximately 500m, Slaty
Creek column, Fig. 3).
Palaeocurrent direction from right
to left (N-S). Beds young towards
top right. Note the thin interbeds of
carbonaceous mudstone. Hammer
for scale.

Fig. 9. Amalgamated beds of fine

sandstone showing no internal
stratification or grading, in the
lower portion of the Watson
Formation, Pororari River (at 120m
in column, Fig. 12). Beds young
towards top. Pocket knife is 9cm
long.
 Cretaceous rift deposits of New Zealand 205

are common in the intervening mudstones (Fig. 10). formation is particularly well-exposed in Pororari
Load casts and flame structures are common at the River and south-flowing tributaries, although its
interface between sandstone and mudstone beds basal contact is faulted. In the adjacent Tindale
(Fig. 11), suggesting rapid deposition, and some thin Creek, the basal unit consists of clast-supported
sandstone layers are intersected by small low-angle breccia consisting entirely of Palaeozoic sandstone
synsedimentary thrust faults. The thrusting is in the clasts resting uncomformably on Palaeozoic strata
same sense as the overturning of flame structures (Fig. 3). This basal unit, and a petrographically
and of other palaeocurrent indicators. Widely sep­ similar 100-m-thick unit of clast-supported-breccia
arated burrowed horizons also occur. occurring about 300 m above the basal fault in
Some variation within the facies assemblage is Pororari River, are considered to represent an inter­
evident between the Watson and Bovis Formations. calation of Assemblage 1. The remainder of the
The Watson Formation is characterized by inter­ formation is dominated by alternating sandstone
leaved components of facies assemblages 1 and 2, and mudstone, including common slumped and
although assemblage 2 is dominant (Fig. 12). The hydroplastically deformed beds (Fig. 13). These beds

Fig. 10. Mudstones of the Watson

Formation containing thin massive
or laminated beds of fine-grained
sandstone, and horizons of isolated
ripples (same locality as Fig. 9).
Beds young upwards. Pocket knife
is 9cm long.

Fig. 11. Load casts and flame

structures in very fine-grained
sandstone beds, lower portion of the
Watson Formation, Pororari River
(at 110m in column, Fig. 12).
Palaeocurrents flowed from left to
right (N-S). Beds young towards
top. Pocket knife is 9cm long.
206 M.G. Laird

KEY

v Sole mark fj, Fining-upward

• Sole mark W Parting lineation
Palaeocurrent Plots:
Slump
17
units
Coarsening-upward
• Slumping direction [] Cross lamination
u Channel
V units --+ Imbrication direction

Sequence Palaeocurrents Lithology and Facies Assemblage Palaeoenvironment

M
0
<{<{
a:­
u U
(flu Clast-supported breccia, ALLUVIAL FAN
dominated by granitic clasts (A 1)
�w

��
<{
J:

�----- N --------1- -----------------------+------------�

�
Alternating sandstone and mudstone,
with minor clast-and-matrix­ BRAID PLAIN
supported breccia and conglomerate. AND
Rare cross-bedding. Channelling LACUSTRINE
· ·
': and small fining-upward sequences FAN
n = 11 common in the upper part (A2)

Alternating thin-bedded
sandstone and mudstone, LACUSTRINE
organized into coarsening­ FAN
upward sequences (A2)
z
0
=
� n 21
:::i:
a:
0
lL
z u
0
(/)

� Clast-supported breccia,
� \
clasts consisting of ALLUVIAL FAN
Palaeozoic sandstone(A1)

LACUSTRINE
FAN
Alternating thin-bedded
sandstone and mudstone (A2)

Fig. 12. Detailed stratigraphical column of the Watson Formation, Pororari River, showing main sedimentary
characteristics, palaeocurrent plots, and inferred palaeoenvironments. Al, A2, facies assemblages 1 and 2.
 Cretaceous rift deposits of New Zealand 207

Fig. 13. Slumped strata near the

base of the Watson Formation,
Pororari River (at approximately
20m in column, Fig. 12). Beds
young towards top of photograph.
Pocket knife is 9em long.

commonly alternate with packets of thick-bedded, thick (Fig. 14). Small scours or channels up to 0. 5 m
graded or massive sandstone. Some 460 m above the deep, infilled with fine-grained sandstone, in some
faulted base of the section, 6 m of pebbly mudstone instances slumped, occur at rare intervals in the
occurs, containing indurated sandstone pebbles up succession (Fig. 15).
to 20 em, and large rafts of thinly bedded sandstone A change in lithofacies occurs at about 800 m
up to 140 cm long, some showing evidence of syn­ above the base of the Pororari River section. At this
sedimentary deformation. This unit is interpreted to horizon, 5 m of cross-bedded sandstone occurs, the
be a mass flow deposit. Above this prominent slump cross-bedded sets averaging 10-20 em in thickness.
horizon, a series of thickening and coarsening Some sets have suffered hydroplastic deformation.
upward units occurs, varying in thickness from 15 to At about the same level, thin beds of matrix- or
110m. The base of each unit consists of laminated clast-supported graded breccia, the clasts consisting
mudstone, passing upwards into alternating mud­ of granite fragments, start to appear, commonly
stone and thin sandstone beds, and finally up into with erosive bases (Fig. 16). Units of alternating
thick, often amalgamated sandstone beds up to 1 m sandstone and mudstone, although still common,

Fig. 14. Thick, massive fine-grained

sandstone beds (upper one with a
scoured base) in a fining upwards
succession, middle portion of the
Watson Formation, Pororari River
(at 650m in column, Fig. 12). Note
pseudonodule horizon below the
lower sandstone bed. Beds young
towards top of photograph). Pocket
knife is 9em long.
208 M.G. Laird

Fig. 15. Channel cut in mudstone

and infilled with slumped sandstone,
Watson Formation, Pororari River
(at 440m in column, Fig. 12). Beds
young towards top of photograph.
Hammer for scale.

Fig. 16. Thin graded breccia beds in

the upper part of the Watson
Formation (at approximately 820m
in column, Fig. 12). Note slumped
laminated mudstones between the
breccia beds. Beds young to the left
of the photograph.

become less frequent upwards. The top 10-20 m of massive matrix-supported breccia beds compnsmg
the Watson Formation is dominated by matrix­ the basal beds of the Hawks Crag Breccia.
supported conglomerate, carbonaceous sandstone, In the Bovis Formation, assemblage 2 forms part
and thin coaly layers in a series of fining upward of one systematically fining upwards succession, with
units 1. 5-3 m thick. Each unit comprises a basal only minor intercalations of other facies associations
conglomerate consisting of moderately well-rounded compared with the Watson Formation. There is an
pebbles of basement rocks, predominantly granite, upward gradation from matrix-supported breccia of
up to 15 em in diameter, commonly resting on a assemblage 1 into metre-bedded pebbly coarse sand­
scoured surface, passing upwards into carbonaceous stone of assemblage 2, containing scattered sub­
parallel- or cross-bedded very coarse sandstone, and angular and subrounded clasts of basement rocks up
then into thin beds of carbonaceous siltstone or coal to 1m (Figs 8 & 9), with minor interbeds of decimet:re
seams up to 20 em thick (Fig. 17). The fining upwards to metre-bedded medium- to coarse-grained massive
units pass with rapid gradation upwards into thick sandstone. The proportion of conglomerate to sand-
 Cretaceous rift deposits of New Zealand 209

Poorly-stratified matrix-supported
conglomerate, scoured at base

Cross-bedded carbonaceous sandstone

Poorly-stratified matrix-supported
conglomerate, scoured at base

Carbonaceous siltstone(lenses)

Poorly-stratified matrix-supported
conglomerate, scoured at base,
and fining upwards

Carbonaceous sandstone
Coal
Carbonaceous sandstone

Poorly-stratified matrix-supported
conglomerate, scoured at base
Fig. 17. Sketch of conglomeratic
fining upwards units at the top of the
Watson Formation, Pororari River Poorly-stratified sandstone
(940m in column, Fig. 12).

stone beds decreases upwards, and in its upper

Interpretation
portions the succession consists almost entirely of
massive or normally graded sandstone beds separ­ The high proportion of mudstone throughout assem­
ated by thin layers of carbonaceous mudstone (Fig. blage 2 in the Watson Formation and in the upper
18). Some of the massive sandstone beds contain portions of the Bovis Formation, preponderance of
pebble trains or scattered clasts, the clast size sediment gravity flow deposits and scarcity of traction
decreasing upwards in the succession. The top few current deposits, and the occurrence of rare bur­
tens of metres of the Bovis Formation consists of rowed horizons suggests sedimentation in a standing
centimetre to decimetre graded very fine to fine­ water body. The highly carbonaceous nature of the
grained sandstone alternating with mudstone. deposits and the absence of shelly fossils suggests
Initially thin mudstone interbeds between sand­ that this is likely to have been a lake. The very poor
stones thicken upwards until near the top of the sorting of the matrix-supported breccia beds, the
unit, the predominant lithofacies is alternating mud­ presence of megaclasts in some beds, and absence of
stone and fine-grained graded or massive sandstone. sedimentary structures suggests deposition of these
Flame-structures are commonly developed at the beds by subaqueous debris flows. Common loads
base of sandstone beds, and show a uniform direction and flame structures at the interface between sand­
of overturning. Sole structures, mainly grooves, but stone and mudstone beds suggest rapid deposition of
a few flutes, also occur. Burrows occur locally. the coarse sediments, and the increasingly common
210 M.G. Laird

Fig. 18. Thin, massive to poorly

graded fine- to medium-grained
sandstone beds with thin partings of
carbonaceous mudstone, near the
top of the Bovis Formation, Slaty
Creek. Beds young towards top of
photograph. Hammer for scale.

presence of normally graded bedded sandstones Slaty Creek. Near the base, assemblage 3 contains
upwards in the Bovis Formation suggests deposition mudstone interbedded with decimetre-bedded fine­
by turbidity currents. The common occurrence of to very fine-grained, parallel laminated sandstone.
slump deposits in the Watson Formation suggests Sandstone beds are commonly normally graded, and
the presence of a significant slope throughout much have sharp bases and tops. This facies passes rapidly
of its depositional history. up into finely laminated, dark, carbonaceous
The thin fining upwards successions near the top mudstone (Fig. 19). Sole marks, consisting mainly of
of the formation in Watson Creek are inferred to grooves, but also with rare prod and flute casts,
have resulted from a high-.energy braided stream occur on the base of some sandstone beds. Isolated
system developed at the toe of a Hawks Crag alluvial vertical burrows occur at some horizons. Rare
fan, which prograded over a vegetated coastal flat slumped units up to 1. 5 m thick also are present near
between the advancing alluvial fan and the standing the base of the assemblage. In Slaty Creek, a horizon
water body. Miall's (1978) 'Scott-type' braided of coal 0. 5 m thick, bounded above and below by
stream model, considered typical of proximal dark mudstone with prominent vertical burrows,
braided stream deposits on alluvial fans, is prob­ occurs a few metres above the gradational contact
ably a close analogue. The Scott River was also between assemblages 2 and 3. The coal could not be
represented as typically containing minor interbeds traced more than 5 m laterally and is almost certainly
of cross-bedded sand and small fining upwards cycles. lenticular. No root horizons are present, and
A similar environment has been inferred by Hamblin the coal is inferred to represent allochthonous
(1992) for parts of the Lower Carboniferous Horton material.
Group of Nova Scotia.
Interpretation
Assemblage 3: laminated mudstone
The ubiquitous presence of highly carbonaceous
mudstone, absence of traction current deposits, and
Facies descriptions
absence of marine shelly fossils but presence of
This facies assemblage, which makes up the bulk of burrowed horizons, suggests deposition in a standing
the Bullock Formation, consists dominantly of dark body of water, inferred to be a lake. The rare
brown to grey, highly carbonaceous laminated mud­ occurrence of sediment gravity flow deposits and
stone. It passes upwards with rapid gradation from slumps suggests that deposition occurred distal to
alternating thin sandstones and mudstones of assem­ the sediment distributary system, and probably well
blage 2, attaining a thickness in excess of 500 m in out on the floor of the lake basin.
 Cretaceous rift deposits of New Zealand 211

Fig. 19. Laminated carbonaceous

mudstone, Bullock Formation,
Slaty Creek. Beds young towards
top of photograph. Hammer for
scale.

D EPO SIT IO N A L SYS T E M inundation by a relative rise in lake level and depo­
O F THE PO R O R A R I G R O U P sition of proximal lacustrine sediments is inferred to
account for association 2, with a brief re-establish­
The three facies assemblages intertongue, and are ment of an alluvial fan represented by association 1
obviously closely related portions of a common deposits from 32 5 to 425 m in the Pororari River
depositional system. The Watson, Hawks Crag section (Fig. 12). The lacustrine deposits of the
Breccia, and Bovis Formations of the western Watson Formation were increasingly encroached on
Paparoa Range, and their associated facies assem­ and finally overwhelmed by the alluvial fans making
blages also show rapid systematic thinning from up the Hawks Crag Breccia, which represented the
north to south (Fig. 3). This, together with the major alluvial fan development of the Pororari
nature of the interrelationships between the facies Group. The end of Hawks Crag Breccia times shows
assemblages supports the interpretation of the increasing incursion and swamping of the fan by first
Pororari succession as a lacustrine fan-delta system. proximal and then distal lacustrine associations in all
The alluvial fan deposits of assemblage 1 are inferred areas of outcrop.
to have fed either directly , or via a narrow braid­ The thickening and coarsening upward successions
plain, into a lake, where deposition continued on a in the middle part of the Watson Formation may
fan consisting of assemblage 2 proximal deposits have been autocyclic, i.e. represented periodic
of the Watson and Bovis Formations. Lacustrine, diversion of the main feeder channels of a lacustrine
mainly quiet-water conditions, are represented by delta and the build-up of a new pro-delta lobe, with
the assemblage 3 mudstones of the Bullock Forma­ accommodation provided by continued subsidence
tion. A simplified reconstruction showing the of the basin. However, the repetitive nature of the
inferred relationship of assemblages to each other, successions, and the evidence for major changes in
and the evolution of the fan delta system, is given in base level outlined above, suggest that a more likely
Fig. 20. cause was fluctuations in lake level. The mudstone
The most informative profile is that given by the at the base of each cycle is inferred to represent a
Pororari River and its tributaries (Fig. 12). Here flooding event caused by a rise in lake level, the
there is clear evidence of a fluctuating lake shoreline, subsequent development of the thickening upward
with irregular progradation of the subaerial portion succession caused by lobes of sediment prograding
of an alluvial fan into the lake, and subsequent into the lake during relative high-stand. The fluctu­
periods of transgression. The presence of 165m of ations in lake level could have been brought about
basal breccia in Tindale Creek suggests that depo­ by a variety of mechanisms, including tectonic
sition started with an alluvial fan. Subsequent activity on a fault or faults controlling the depo-
212 M.G. Laird

N s

Lake surface
a

Cl Cl Cl
Cl Cl Cl
Cl Cl Cl
Cl
Cl l:.
l:. Alluvial fan l:.
l:.
Cl l:.
l:. l:. CRAG
Cl Cl l:. Cl
l:.
l:. BRECCIA
Assemblage 1 l:.
l:.
l:. l:.
Cl
l:. l:. Cl l:.

Cl l:. Cl l:.
Cl
l:. Cl l:. Cl
l:. l:.
l:. Cl l:. l:.
l:.
l:. Cl

Fig. 20. Palaeoenvironmental/tectonic sketch (not to scale), showing the inferred relationships of the facies assemblages
making up the Pororari Group. a-a' shows the approximate location of the Pororari River measured section (Fig. 3).

sitional basin, or by climatic cycles. Cyclic suc­ of transport. Only a small number of directional
cessions present in the dominantly lacustrine early features were recorded in the Bullock Formation in
Mesozoic Newark Supergroup of eastern North the eastern Paparoa Range.
America have been attributed to climate-controlled The palaeocurrent and palaeoslope trend in the
high-frequency fluctuations in the depth of rift lakes Watson Formation is generally from north to south.
(e.g. Olsen, 1986, 1991). However, in the thick, well-exposed succession
in the Pororari River and its tributaries, plotting
palaeocurrents in segments according to facies
P A L A E O C U R R E N T PAT T E R N groups clearly shows that significant variations occur
within the succession (Fig. 12). The lower, poorly
Palaeocurrents (summarized in Fig. 2 ) were deter­ exposed 300 m, consisting of alternating sandstones
mined from a wide variety of sedimentary features, and mudstones of association 2, has palaeocurrents
including grooves, prods, flutes, primary current directed towards the south or south-southwest. By
lineation, channel margins, and cross-lamination, contrast, the similar facies higher in the column
and palaeoslopes from slump fold axes in the lacus­ between 430 and 800 m in the section has palaeocur­
trine facies associations 2 and 3. Palaeocurrent rents directed towards the south-southeast. The sole
indicators are relatively common in the Watson imbrication measurement recorded in the inter­
Formation, and in the Bovis and Bullock Formations vening 100-m unit of assemblage 1 breccia (between
of the western Paparoa Range. Although no suitable 330 and 430 m, section Fig. 12), although statistically
bedding surfaces were exposed for measurement of insignificant, is also directed to the south-southeast,
palaeocurrent directions in the Bovis Formation of suggesting a source in common with the overlying
the eastern Paparoa Range, numerous overturned sediments, but differing from the source providing
flame structures, and local occurrence of imbrication the earlier sediments. The 100-m thick breccia unit
in pebbly beds indicate a consistent general direction is thus inferred to represent an abrupt incursion of
 Cretaceous rift deposits of New Zealand 213

an alluvial fan into the lake. The monolithological cation were relatively uncommon however, and
nature of clasts, which consist of indurated basement only a few measurements were obtained. Although
sandstone, and their high degree of angularity, there is some variation, a dominant NNE-SSW, or
suggests a very proximal source. The same fai'!_. prob­ NE-SW trend is apparent.
s
ably continued to feed clastic detritus in 'a 0uth­ In the Bovis Formation, palaeocurrents are widely
southeasterly direction into the lake up until:a�:'reast variable (Fig. 2), but still show a generally southerly
the 800 m mark (Fig. 12), an inference supportetl by trend. The information from sole marks and other
the fact that the clasts in the 6-m bed of slumped features is strongly reinforced by the direction of
pebbly mudstone 20m above the breccia (at 460 m, overturning of flame structures, which occur at the
Fig. 12) also consist largely of indurated basement base of many sandstone beds. Sole marks from the
sandstone. Bullock Formation show a closely constrained
The upper part of the Watson Formation, between and consistent trend from north-northeast to south­
800 and 950 m in Fig. 12, although dominated by southwest (Fig. 2).
assemblage 2 facies, includes thin beds of breccia The consistent thinning of the Watson Formation,
and conglomerate containing granitic clasts, indi­ Hawks Crag Breccia, and Bovis Formations from
cating a different provenance. This mixture of north to south in the western Paparoa Range also
sources is reflected by the wide variation in current supports the inference of a transport direction in a
directions recorded, ranging through southeast to generally N-S direction.
southwest. Southwest and west-southwest directions
are indicated by clast imbrication on two surfaces in
the overlying Hawks Crag Breccia (Fig. 12), and a T E C T O NIC 1M PLIC A T I O N S
south-southwest direction of transport is recorded
overall for the Hawks Crag Breccia in the study area I t has been widely recognized that fan-deltas are
(Fig. 2). Palaeocurrent data were not available from preferentially located in areas of active tectonism,
the Watson Formation in the adjacent Tindale and are commonly controlled by an active fault or
Creek, the only other area of major outcrop, thus faults (e.g. Ethridge & Wescott, 1984). The great
these changes of palaeocurrent trend could not be thickness of the clastic wedge (a minimum of 1200 m
confirmed as a regional phenomenon. However, the for the Hawks Crag Breccia alone), and the relative
inference is that at least three distinct sources and angularity of clasts, suggests that they were gener­
fans influenced sedimentation: the earliest from the ated by an active fault system in close proximity.
north or north-northeast; the next from a Palaeozoic The evidence suggests that the fan-delta system occu­
area to the north-northwest, and the upper part of pied a half-graben, downthrown to the south (Fig.
the Watson Formation is increasingly influenced by 20). The facies relationships, with thick coarse sub­
debris derived from a granitic source to the north­ aerial deposits concentrated in the north, and rapid
east. Derivation of indurated sandstone clasts in the thinning and fining to the south, i.e. downcurrent,
second fan from the north-northwest is consistent also closely fits the model proposed by Surlyk (1978)
with the presence of Early Palaeozoic turbidites of for fan-delta sedimentation in a half-graben. Move­
the Greenland Group cropping out to the west of ment on the inferred bounding fault to the north
the Pororari Group. However, basement rocks cur­ continued during much of the deposition of the
rently lying to the northeast of the Hawks Crag Pororari Group, and was responsible not only for
breccia and extending for many kilometres consist generating the coarse facies of the Hawks Crag
entirely of foliated gneiss of the Charleston Meta­ Breccia and the associated lacustrine sediment grav­
morphic Group (Nathan, 1978; Laird, 1988), a lith­ ity flows, but also for the necessary subsidence to
ology almost absent from the Hawks Crag Breccia, provide accommodation for the great thickness of
which is dominated by granitic clasts. This problem the deposits. Contemporaneous fault movement may
of derivation is discussed below under Tectonic also have been a factor in causing the fluctuations in
Implications. lake level noted earlier, although climate also may
Clast imbrication is the only usable palaeocurrent have played a role. Judging from the consistently
indicator in the alluvial fan assemblage represented s-ssw directed sediment gravity flows, both sub­
by the Hawks Crag Breccia, which is dominated by aerial and subaqueous, the active fault scarp was
granitic clasts. Suitable surfaces for recording imbri- likely to have been oriented in a westerly or west-
214 M.G. Laird

South Island
New Zealand

N
�
"i

t
0 10
-
km

KEY

- Late Cenozoic fault

c;:::? 9 Mid Cretaceous fault(?)

;> Palaeocurrent azimuth

Pororari group

Greymouth [I] Upper lake beds

Fig. 21. Map showing generalized
m Hawks Crag Breccia palaeocurrent trends and inferred
alignment of the mid-Cretaceous
j...: - : l
: Lower lake beds fault bounding the half-graben
containing the Pororari Group.

northwest direction with an implied N-NNE direc­ lacustrine strata in the outcrop area on the eastern
tion of tectonic extension (Fig. 21). flank of the Paparoa Range, which lies south­
As noted previously, the similarity of Lithofacies, southeast of that on the western flank, i.e. downcur­
stratigraphical position, and palaeocurrent pattern rent, do not conform with the southward thinning
of the Hawks Crag Breccia and, in particular, the pattern of units clearly shown by the western out­
overlying lake beds in the now widely separated crops. The thickness of the Hawks Crag Breccia of
outcrops on the eastern and western flanks of the the eastern outcrop is similar to the maximum thick­
Paparoa Range suggests that they were formed in ness in the north of the western unit, but the over­
the same depositional basin. However, the thick­ lying strata of association 2 are more than twice the
nesses of the Hawks Crag Breccia and the overlying thickness of the western maximum. This is hard to
 Cretaceous rift deposits of New Zealand 215

explain by invoking a change in sedimentation pat­ infilled by the Beebys Conglomerate (Johnston ,
tern, as the palaeocurrent pattern is essentially indis­ 1990; Fig. 1) is also likely to have been formed
tinguishable between the two outcrop areas. A more during the same period of north-northeast extension.
likely explanation is that transcurrent movement These data fit well with the north-northeast exten­
along a north-northeast trending fault (Hawera sion direction for the Paparoa Range area during
Fault?) subsequent to deposition has offset the late Aptian to early Albian times deduced from
eastern part of the half graben to the south by at structural research on the basement noted above.
least 10-15 km. Tulloch & Kimbrough (1989) provided evidence
It was noted previously that the clasts making up that stretching lineations associated with low-angle
the Hawks Crag Breccia consist almost entirely of normal detachment faults separating the upper from
granite , inferred to have been derived from a proxi­ the lower plate defined a regionally consistent north­
mal source to the northeast. The basement rocks to northeast extensional trend, which coincides with
the northeast and north, however, consist for many that inferred from the half-graben.
kilometres of foliated gneiss of the Charleston Meta­ It has been suggested (Laird, 1993, 1994) that
morphic Group, a lithology that occurs only rarely half-graben formation and synchronous infilling were
in the Hawks Crag Breccia. An added apparent associated with a New-Zealand-wide early period of
complexity is the evidence provided by Tulloch & extension beginning in the mid-Albian, pre-dating
Palmer (1990) that granite samples typical of the by approximately 2 5 Ma the separation of New
bulk of the Hawks Crag Breccia at the Fox River Zealand from Gondwana and sea-floor spreading in
mouth and from several large exposures in Bullock the Tasman Sea.
Creek are similar to the Buckland Pluton, the closest
outcrop of which lies 20 km to the northeast. The
clasts are distinct from the Meybille Granite , which CONCLUSIONS
crops out to the immediate south of Fox River
mouth. The inference drawn by Tulloch & Palmer 1 The Pororari Group of the southern Paparoa
(1990) is that, at the time of their deposition, either Range represents a fault-controlled lacustrine fan­
the Fox River and Bullock Creek deposits were delta complex. The Hawks Crag Breccia and equiv­
situated 20-2 5 km to the northeast, or the eroded alent matrix- and clast-supported breccias represent
upper parts of the Buckland Pluton had a consider­ deposits of humid-temperate climate alluvial fans,
ably greater horizontal extent. Tulloch & Kimbrough while intertonguing and overlying finer grained sedi­
(1989) inferred that the Greenland and Pororari ments were deposited predominantly in a lacustrine
Groups, together with the granite from which the environment.
latter had b'een derived, were part of an upper plate 2 Mass-flow processes dominated both alluvial fan
that detached from a metamorphic core complex and lacustrine sedimentation, suggesting that tec­
(Charleston Metamorphic Group) which comprised tonic activity continued throughout most of the
a lower plate. On this hypothesis, dislocation of period of deposition of the Pororari Group, dying
the Hawks Crag Breccia from its original source is away only towards the end. The abundance of debris­
possible. flow deposits in the alluvial fan association, and the
A possible analogue of the southern Paparoa poor-sorting and angularity of many of the clasts
Range half-graben has been recognized in seismic suggests that source slopes were steep and lay in
profiles offshore from Greymouth, 30 km to the close proximity. The ubiquitous occurrence of
southwest (Bishop, 1992; Fig. 21). This half-graben debris-flow deposits, turbidites and slumps in the
is also aligned WNW-ESE, and is inferred to be proximal lacustrine succession indicates the presence
infilled largely with Pororari Group sediments. It is of a significant slope on the subaqueous fan, which
clearly bounded to the north by a southward-dipping may account for the limited presence of deposits
normal fault, and to the south by a pinchout. It is transitional between the alluvial and sublacustrine
divided by major north-northeast striking transfer fans.
faults into several segments. The orientation of the 3 Fluctuations in lake level are indicated by episodic
half-graben and bounding normal fault clearly sup­ transgressions and regressions of the lake with res­
ports the direction of tectonic extension suggested pect to the adjacent alluvial-fan system, and by the
for the southern Paparoa Range half-graben. The presence of repetitive thickening and coarsening
inferred E-W to SE-NW oriented half-graben upwards successions in the proximal lacustrine
216 M .G. Laird

deposits of the Watson Formation. These fluctuations ETHRIDGE, F.G. & WESCOTT, .W.A. (1984) Tectonic setting,
may have been caused by active faulting at the basin recognition and hydrocarbon reservoir potential of fan­
delta deposits. In: Sedimentology of Gravels and Con··
margin, by climatic changes, or by a combination of
glomerates (Eds Koster, E.H. & Steel, R. J . ), Mem.
both. Can. Soc. petrol. Geol., Calgary, 10, 217-235.
4 The fan-delta system is inferred to have occupied HAMBLIN, A.P. (1992) Half-graben lacustrine sedimentary
an actively subsiding half-graben. The consistently rocks of the Lower Carboniferous Strathlorne Formation,
south-southwest directed sediment gravity flows and Horton Group, Cape Breton Island, Nova Scotia,
Canada. Sedimentology, 39, 263-284.
rapid southward thinning of the sedimentary wedge JOHNSTON , M.R. (1990) Geology of the St Arnaud District,
suggests that the active basin margin fault was likely Southeast Nelson (Sheet N29). New Zealand Geological
to have been oriented in a WNW-ESE direction. Survey Bulletin 99. New Zealand Geological Survey,
The implied NNE-SSW direction of tectonic exten­ Lower Hutt, New Zealand, 119 pp.
KocHEL, R.C. & JOHNSON, R.A. (1984) Geomorphology
sion compares well with a similar extension direction
and sedimentology of humid-temperate alluvial fans,
determined for this area from structural research on central Virginia. In: Sedimentology of Gravels and Con­
basement rocks, and with that of a probable mid­ glomerates (Eds Koster, E . H. & Steel, R.J.). Mem.
Cretaceous half-graben recognized on offshore seis­ Can. Soc. petrol. Geol., Calgary, 10, 109-122.
KoRSCH , R.J. & WELLMAN, H.W. (1988) The geological!
mic profiles. The orientation of the half-grabens is
evolution of New Zealand and the New Zealand region.
essentially parallel to the future spreading axis associ­ In: The Ocean Basins and Margins, Vol. 7B (Eds Nairn,
ated with the opening of the adjacent Tasman Sea. A.E.M., Stehli, F.G. & Uyeda, S. ), pp. 411-482.
This opening occurred at about 80 Ma, suggesting Plenum, New York.
that initial rifting and half-graben formation occurred LAIRD, M.G. (1988) Sheet S37 Punakaiki. Geological
map of New Zealand 1 : 63 360. New Zealand Geologi­
20-25 Ma before sea-floor spreading began.
cal Survey, Department of Scientific and Industrial
Research, Wellington.
LAIRD, M.G. (1993 ) Cretaceous continental rifts. New
ACKNOWLEDGEMENTS Zealand Region. In: Sedimentary Basins of the World.
South Pacific Sedimentary Basins (Ed. Ballance, P.F. ),
pp. 37-49. Elsevier, Amsterdam.
I am grateful to D . Lewis, J . Lindqvist and A. LAIRD, M.G. (1994) Geological aspects of the opening
Tulloch for stimulating discussions, both in the field of the Tasman Sea. In: The Evolution of the Tasman Sea
and in the office, which helped to clarify my ideas on Basin (Eds van der Lingen, G.J., Swanson, K. & Muir,
processes of emplacement and geological setting of R.J.), pp. 1-17. Balkema, Rotterdam.
LEWIS, D.W. & EKDALE, A.A. (1991 ) Lithofacies relation­
the Pororari Group. Thorough reviews by T. Astin,
ships in a late Quaternary gravel and loess fan delta
S. Flint, D. Lewis, J. Lindqvist and G. Plint greatly complex, New Zealand. Palaeogeogr. , Palaeoclimatol. ,
improved the manuscript and are appreciated. Palaeoeco/. , 81, 229-251.
LIN DQVIST, J.K. (1990) Puysegur Group: a mid Cretaceous
lacustrine fan-delta complex, Balleny Basin, southwest
Fiordland. Geol. Soc. N. Z. Misc. Pub/. , SOA, 83.
REFERENCES McARTHUR, J.L. (1987) The characteristics classification,
and origin of Late Pleistocene fan deposits in the Cass
BALLANCE, P.F. (1984) Sheet-flow-dominated gravel fans Basin, Canterbury, New Zealand. Sedimentology, 34,
of the non-marine middle Cenozoic Simmler Formation, 459-471.
central California. Sediment. Geol. , 38, 337-359. McKAY, A. (1883) On the geology of the Reefton District,
BISHOP, D.J. (1992) Extensional tectonism and magmatism lnangahua County. Geol. Survey Reports during 1882 ,
during the middle Cretaceous to Paleocene, North Vol. 15, pp. 142-144. Wellington.
Westland, New Zealand. N.Z. J. Geo/. Geophys. 35, MLALL, A.D. (1978) Lithofacies types and vertical profile
81-91. models in braided river deposits: a summary. In: Fluvial
BISHOP, D.G. & LAIRD, M.G. (1976) Stratigraphy and Sedimentology (Ed. Miall, A.D.), Mem. Can. Soc.
depositional environment of the Kyeburn Formation petrol. Geol., Calgary, 5, 597-604.
(Cretaceous), a wedge of coarse terrestrial sediments in MORGAN , P.G. & BARTRUM, J.A. (1915) The geological
Central Otago. J. R. Soc. N.Z. , 6, 55-7 1. and mineral resources of Buller-Mokihinui Subdivision,
BLAIR, T.C. (1987) Sedimentary processes, vertical stratifi­ Westport Division. Geological Survey Bulletin No. 1 7
cation sequences, and geomorphology of the Roaring (new series) . New Zealand Department of Mines, Geo­
River alluvial fan, Rocky Mountain National Park, logical Survey Branch, Wellington.
Colorado. J. sediment. Petrol. 57, 1 - 18. NATHAN , S. (1978) Sheet S31 & Part S32 Buller- Lyell.
BLAIR, T.C. & McPHERSON, J.G. (1992) The Trollheim Geological Map of New Zealand 1 : 63 360. New Zealand
alluvial fan and facies model revisited. Geol. Soc. Am. Geological Survey, Department of Scientific and Indus­
Bull. , 104, 762-769. trial Research, Wellington.
BRADSHAW, J.D. (1989) Cretaceous geotectonic patterns in OLSEN, P.E. (1986) A 40-million-year lake record of early
the New Zealand region. Tectonics, 8, 803-820. Mesozoic orbital climatic forcing. Science, 234, 842-848.
 Cretaceous rift deposits of New Zealand 217

OLSEN, P.E. ( 1991) Tectonic, climatic, and biotic modu­ Geological Survey Report 109 Department of Scientific
lation of lacustrine ecosystems - examples from Newark and Industrial Research, New Zealand.
Supergroup of eastern North America. In: Lacustrine SuRLYK, F. ( 1978) Submarine fan sedimentation along fault
Basin Exploration. Case Studies and Modern Analogs scarps on tilted faultblocks (Jurassic-Cretaceous bound­
(Ed. Katz, B.J.) , Mem. Am. Assoc. petrol. Geol. , Tulsa, ary, East Greenland). Gr¢n. geol. Unders. Bull. , 128,
50, 209-224. 108 pp.
PIERSO N , T.C. ( 1 980) Erosion and deposition by debris TULLOCH , A.J. & KIMBROUGH, D.L. ( 1989) The Paparoa
flows at Mt Thomas, north Canterbury, New Zealand. Metamorphic Core Complex, New Zealand: Cretaceous
Earth Surf Process. , 5, 227-247. extension associated with fragmentation of the Pacific
PIERSON, T.C. ( 1981) Debris flows. An important process margin of Gondwana. Tectonics, 8, 1217- 1234.
in high country gully erosion. 1. Tussock Grasslands TuLLOCH, A.J. & PALMER, K. (1990) Tectonic implications
Mountain Lands Inst. , Rev. , 39, 3- 14. of granite cobbles from the mid-Cretaceous Pororari
RAINE, J.l. ( 1984) Outline of a palynological zonation of Group, southwest Nelson, New Zealand. N.Z. 1. Geol.
Cretaceous to Paleogene terrestrial sediments in West Geophys. , 33, 205 -217.
Coast region, South Island, New Zealand. New Zealand
Spec. Pubis int. Ass. Sediment. (1995) 22, 219-236

Sedimentation and tectonics of a synrift succession:

Upper Jurassic alluvial fans and palaeokarst
at the late Cimmerian unconformity,
western Cameros Basin, northern Spain

N IG E L H . P L A T T
Geco-Prakla Schlumberger, Schlumberger House, Buckingham Gate,
Gatwick Airport, West Sussex RH6 ONZ, UK

ABSTRACT

The Upper Jurassic (late Kimmeridgian to Berriasian ?) Senora de Brezales Formation of the western
Cameros Basin, northern Spain, comprises a laterally variable succession of continental conglomerates,
sandstones and pedogenetic carbonates deposited in wadi-type channels, alluvial fans, sandflats and
palaeosols in a semi-arid environment.
This succession rests on a complex unconformity surface that developed in response to relative falls
in sea-level and strong Late Jurassic extensional faulting. Footwall uplift resulted in truncation of
underlying strata at the crestal axes of fault blocks. The clastic rocks of the Senora de Brezales
Formation represent the erosion products of the Jurassic anq older strata. Lateral facies variations
reflect the changing lithology beneath the unconformity surface, whereas strong lateral thickness
variations record fault control on sedimentation.
In areas of limited erosion, karst surfaces developed on subaerially exposed lower Kimmeridgian
limestones. Where these were remo ved, erosion of upper Oxfordian marginal marine sandstones led to
the deposition of red continental sandstones. In other areas, erosion led to the incision of channels into a
pediment of Middle Jurassic carbonates. The channels were filled with conglomerates largely deri ved
from erosion of the Jurassic marine limestones. The location of channels was strongly influenced by
NE-SW faults. Laminar and nodular calcretes formed directly on the unconformity surface, in
interchannel areas, and within the o verlying clastic succession.
The Senora de Brezales Formation is an outcrop analogue for Upper Jurassic clastic successions
present in the subsurface of basins to the west of Britain on the northern Biscay margin . Field studies
highlight the complex lateral facies variations within this synrift succession and underline the control of
seismic-scale and subseismic-scale faults on erosion and sedimentation patterns at the late Cimmerian
unconformity.

INTRODUCTION AND
GEOLOGICAL SETTING

Many of the Mesozoic basins bordering the North et a!.,1985; Platt, 1989a, 1990; Platt & Pujalte,
Atlantic display complex unconformities and thick 1994), Ireland (Petrie et a!., 1989; Shannon, 1991)
continental successions recording major rifting and southern England (Kamerling, 1979; Chadwick,
events during the Late Jurassic to Early Cretaceous 1985; Evans, 1990; Ruffell & Coward, 1992).
(Tankard & Balkwill, 1989; Hiscott et a!., 1990). Upper Jurassic clastic rocks are reported from the
Improved understanding of the stratigraphy and offshore basins to the southwest of the UK and
tectonic evolution of rifts bordering the Bay of Biscay Ireland on the northern Biscay margin (Millson,
has highlighted the similarities between the offshore 1987; Evans, 1990; Shannon, 1991; Moore, 1992),
and onshore basins of Spain (Garcfa-Mondejar but there has been limited sedimentological study of
Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
219
220 N.H. Platt

these subsurface successions to date. In northern the modern Florida Everglades or to the marshlands
Spain, on the southern Biscay margin, Upper of southern Iraq (Platt, 1989b, c; Platt & Wright,
Jurassic-Lower Cretaceous continental clastic de­ 1991, 1992). The Rupelo Formation is overlain by
posits are well-exposed, permitting detailed sedi­ oncoidal limestones, sandstones and mudstones
mentological study and mapping of lateral facies and of the ?Valanginian-Barremian Hortigtiela For··
thickness variations within the Upper Jurassic synrift mation (Fig. 2), and a thick ?Barremian-Aptian
succession (Platt, 1986; Clemente & Perez-Arlucea, series of alluvial clastic rocks comprising the sandy-·
1993; Gomez Fernandez & Melendez, 1994). muddy Piedrahita de Mufi6 Formation and the
This paper presents the results of a sedimen­ conglomeratic-sandy Salas Group. This succession
tological study of the Upper Jurassic Senora de is capped by alluvial sandstones and conglomerates
Brezales Formation of the western Cameros Basin of the upper Albian to lower Cenomanian Utrillas
(Fig. 1), outlining the tectonic controls on alluvial Formation.
fan deposition and palaeokarst development at the
late Cimmerian unconformity.
LITHOFACIES

STRATIGRAPHY The Senora de Brezales Formation shows strong

lateral variability (see discussion below), but in all
The stratigraphy of the western Cameros Basin, cases comprises three main facies: sandstones, con­
described by Platt (1986, 1989a, b) and Platt & glomerates and carbonate rocks. Figure 3 presents a
Pujalte (1994), is outlined in Fig. 2. A thick suc­ representative logged sedimentological section from
cession of Early-Middle Jurassic marine carbonates the Senora de Brezales Formation.
is truncated by a complex series of unconformities
(Mensink & Schudack, 1982; Schudack, 1987;
Red sandstones
Platt et al., 1991). In the south, Bathonian­
Callovian limestones are discordantly overlain by Red sandstones reach a maximum of 40 m in thick­
(?Callovian-) Oxfordian sandstones and cal­ ness and are present:
carenites of the San Leonardo Formation (Wilde, 1 forming the lower part of the Senora de Brezales
1988) and Oxfordian to basal Kimmeridgian car­ Formation to the southwest of the San Leonardo
bonates of the Talveila Formation (Dfaz et al., 1983; Fault (Fig. 1);
Platt, 1986; Thalmann, 1989). Sedimentary cycles 2 above basal Senora de Brezales Formation lime­
within this complex and heterogeneous succession of stone conglomerates in the northeast of the study
marginal marine carbonates and littoral clastics (Platt, area.
1986) may be compared with those present in the The sandstones are typically dark brick-red in
Corallian of southern Britain (Talbot, 1973; Sun, colour and are of medium grain size. They have
1989). subrounded grains, are generally quartzose, and
The middle Oxfordian to Kimmeridgian shallow­ contain scattered quartz pebbles 1-4 mm in dia­
to marginal-marine rocks are absent in many areas, meter. The sandstones are generally structureless
reflecting erosion prior to deposition of the Senora (Fig. 4A), although trough cross-bedding, thin
de Brezales Formation, which is a laterally variable pebbly lags and plane lamination are locally discern­
succession of continental deposits present through­ ible (Fig. 4B). White-red mottling and/or 1-2 cm
out the western part of the basin (Platt, 1989a). diameter burrows are developed at a few localities.
The Senora de Brezales Formation comprises poly­ Vertical cylindrical structures 5-10 em in diameter
genetic conglomerates, red sandstones and calcareous are also present locally, as are 2-cm-diameter tubu­
palaeosols, and is probably of Kimmeridgian­ lar structures parallel to bedding.
Berriasian age (Platt, 1986; Wright et al., 1988). The In the west, at Hortezuelos and Mamolar (Fig. 1),
Senora de Brezales Formation reaches a maximum the sandstones are cut by vertical and horizontal
thickness of 75 m to the north of Espej6n (Figs 1 & cracks. Brecciated, nodular carbonate horizons
3), and is overlain by lacustrine-palustrine lime­ 0.5-1 m thick are also present. These are commonly
stones of the Rupelo Formation (Berriasian). The mottled and display floating quartz grains, angular,
Rupelo Formation is interpreted as a succession of branching spar-filled cracks and grey carbonate
freshwater carbonates deposited in a low-gradient, stringers 1 mm in width. These carbonate facies are
low-energy system of lakes and swamps similar to described in detail below.
 Alluvial fans, palaeokarst and tectonics 221

Fig. 1. (A) Palinspastic

reconstruction of the North A tlantic �
and Bay of Biscay region for Late QCF R .'......._ '-...
Jurassic times showing the location
� ... '-...
"'-..
of major Mesozoic basins. The
Cameros Basin is in central Iberia, ... "'
a · ML
• :'
He :' J """
•
...
to the south of the Biscay rift. •

/JCF PM •
(B) Map of the western Cameros .·

Basin showing major structural Q__J9km

elements and localities mentioned in
the tex t: C, Castrovido;
H, Hortigiiela/Valparaiso;
Ho, Hortezuelos; HR, Huerta del
Rey; J, Jaramillo Quemado; LG , La
Mo
Gallega; M, Mamolar; •
Mo, Moncalvillo; ML, Mambrillas �
de Lara; PB, Pinilla de los
Barruecos; PM, Pinilla de los
Moros; Q, Quintanilla de las Vinas;
R, Rupelo; SI, Salas de los Infantes;
SL, San Leonardo de Yagiie; T,
-- basin margin
Talveila; QCF, Quintanilla­ --'--- normal fault

thrust
Castrovido Fault; JCF, Jaramillo­
___.___
Covarrubias Fault; SLF, San
Leonardo Fault. (C) Inset showing · · · · · · · · transfer fault
area covered by map B .

places, but the rare presence of lags and mesoscale

Interpretation
cross-bedding suggests that the sandstones were
The homogeneous character of the sandstones is waterlain. Mud is absent; flows were prol::ably short­
consistent with strong bioturbation. Primary sedi­ lived with no deposition from suspension. Thin sec­
mentary structures are preserved in only a few tion observation shows that the red colour is the
 W CAMEROS BASIN- STRATIGRAPHY

(/)
::::l

w w
a::O
a..U Albian
::::>1-
a..<{
w
a:
u

Aptian

Barremian

(/)
::::l

ww
a::O !'-'"""
:s:;u
o<�: Hauterivian
....Jtu
a:
u

Valanginian

Berriasian

Tithonian
lfiii±�;ijj£/�,��;P""�
�--------��:0�::;i;�:�i:�:,�r:: �:��i.;�:��:f}:+:?:>:xj;fj}/j;;����:;;.;
Kimmeridgian ,;;.;..;. �
· ·

S: Talveila��

Oxfordian

-,
Callovian

KEY

Evaporite Marl

Oncoidal limestone Sandstone

Organic-rich limestone Conglomerate

Fig.2. Upper Jurassic to Lower

Limestone Cretaceous stratigraphy of the
western Cameros Basin.
 Alluvial fans, palaeokarst and tectonics 223

z
0
�
�
a: Ladera
f2 Member
0
...J
LU
a..
� Las Vinas
Member

KEY

(/)
(ill!illJ�illn Quartz conglomerate

LU Charophyte-gastropod
...J
<t: limestone and marl

w
Nz
0:::o
­
Brecciated limestone
COI­ Polymict conglomerate
w< Reticulate calcrete
0�
<t:o:::
o:::O
Laminar calcrete
ou.. Burrows and roots
Cross-bedded red sandstone
•Z
LU
(/) Massive red sandstone
0 j���� Marine limestone
Fig. 3. Logged sedimentological
section from the Senora de Brezales
Formation, section near chapel of Marine
Nuestra Senora de Brezales, 2 km Jurassic
north of Espej6n.

product of iron oxide grain coatings, typical of con­ prevented the preservation of organic material so
tinental 'red bed' deposits. This interpretation is that the roots are preserved only as moulds.
consistent with the sedimentological evidence of
subaerial exposure. This includes cracks, interpreted
Conglomerates
as sheet cracks developed during desiccation, mottles
and tubular structures of varying morphology and Conglomerates are common in the Senora de
size, interpreted as 1-2-cm-diameter burrows and Brezales Formation. These are poorly sorted, with
2-10-cm-scale roots. The brecciated, nodular and subangular to subrounded clasts, mostly of carbonate
mottled fabrics of the carbonates are typical of pedo­ and generally varying in size from 2 to lOcm.
genetic horizons, favouring an interpretation as cal­ Locally, the clasts reach 30-100 em, as at Jaramillo
cretes. The presence of floating quartz grains indicates Quemado (Fig. 4C). The matrix consists of medium­
the in situ growth of replacive calcite. grained red sandstone with 1-2-cm rounded quartz
The red sandstones are thought to represent pebbles, which are more abundant in the southwest.
ephemeral stream deposits. The climate was prob­ Conglomerates commonly occur in channels. The
ably semi-arid, with sporadic rainfall permitting only channels are 10-50 m across, have erosive bases and
short-lived stream flow; in the intervening periods are 2-3 m deep (Fig. 5). Coarse conglomerates at
the sandstones were homogenized by burrowing and the channel bases pass up into red, trough cross­
root action. The occasional rainfall permitted the bedded sandstones. Many channels are cut directly
growth of hardy scrub vegetation and periodically into marine Jurassic carbonates, as in the northeast
allowed the establishment of high or perched water of the study area. At Castrovido (Fig. 1), channelized
tables, favouring the development of palaeosol pro­ conglomerates rest on lower Callovian limestones,
files and pedogenetic carbonates (see also below). passing up into interbedded conglomerates and
Desiccation and strongly oxidizing conditions were red sandstones. Channelized conglomerates also
responsible for the red colour of the sediments and rest unconformably on lower Callovian marine
224 N.H. Platt
 �
f2
""

�
'i:i'
·"'
;:;

�
�
�

�
C)
,.,.
�

�
�
;:;
.,__

�
0
;:;

�·

Fig. 4. Lithofacies and sedimentary structures within the Senora de Brezales Formation. (A) sandstones- massive. Huerta del Rey. Person for scale.
(B) sandstones- cross-bedded. Huerta del Rey. Hammer for scale. (C) conglomerates- very coarse conglomerates containing locally derived clasts of lower
Callovian limestone. Jaramillo Quemado. Lens cap for scale. (D) conglomerates- stacked, fining up sheet-flood conglomerates. Arrows mark bases of sheet-flood
units. La Gallega (Navas road) . Hammer for scale. (E) carbonates- cross-cutting laminar calcrete present within conglomerate, Hortezuelos. Lens cap for scale.
(F) carbonates- stacked, horizontally bedded laminar calcretes in sandstone matrix, Mamolar. Lens cap for scale. (G) karst surface- karstic cavity developed in
Lower Kimmeridgian limestones, filled by limestone breccia and red internal sediment, Talveila. Hammer for scale. (H) karst surface- red sandstone, marked by 1-'3
V1
arrow, piped down into lower Callovian limestone surface, Pinilla de los Moros. Hammer for scale.
 �
0\

:<:
�
�
:::::

Fig. 5. Channels of polymict conglomerate cut into red sandstones and marls, Espej6n. The arrows mark the bases of three distinct channels. These channels are not
fault-bounded. Bedding is vertical, younging away from camera: The photograph was taken standing on Jurassic marine limestones and looking southwards. The
prominent ridge running left to right across the photograph above the Senora de Brezales Formation channels is composed of non-marine limestones of the o verlying
Rupelo Formation (R). The village and cornfields beyond lie on Valanginian-Cenomanian clastic rocks (KL), whereas the mountain on the horizon consists of
Upper Cretaceous marine limestones (KU).
 Alluvial fans, palaeokarst and tectonics 227

carbonates at Quintanilla de las Vinas and Hortigiiela

Interpretation
in the northwest, an area where red sandstones are
absent. In the southwest, conglomerates occur inter­ The channels provide evidence of powerful, strongly
bedded with sandstones or with palaeosols some erosive flows which were probably intermittent
metres above the base of the Senora de Brezales events associated with sporadic heavy rainstorms.
Formation. The wadi-type channels were incised into the uncon­
Many channels incised into the basal unconformity formity surface on uplifted blocks, and locally, as at
are bounded on one or both margins by minor NE­ Espej6n, into finer grained deposits of the Senora de
SW trending faults. Lateral variations in conglomer­ Brezales Formation. Away from the fault scarps,
ate thickness also commonly occur across these flows spread out into sheet-floods, depositing tabular
faults. In contrast, channelized conglomerates pre­ units of conglomerate and sandstone.
sent at higher levels within the Senora de Brezales
Formation are not fault-bounded. At Espej6n {Fig.
Carbonate rocks
1), a number of individual channels are spectacularly
exposed. The channels are asymmetric, with strongly Pedogenetic carbonate horizons are interbedded
erosive bases. The coarsest conglomerates fill the with the red sandstone facies of the Senora de
base of the channel, passing upwards and towards Brezales Formation (see above). Carbonate horizons
the channel margin into red, trough cross-bedded also rest directly upon marine Jurassic limestones in
sandstones. interchannel areas.
Sheet-like conglomerates are also common; these Laminated limestones are present in the north of
show planar erosive bases and typically occur in the study area, lying either directly on the uncon­
2-5-m beds (Fig. 4D). Several pulses of conglomer­ formity surface (as at Castrovido and Mambrillas de
ate occur at Senora de Brezales; these are traceable Lara; Fig. 1), interbedded with or cross-cutting the
laterally for 20-50 m. sandstones and conglomerates (as at Hortezuelos
At Hortezuelos (Fig. 1), conglomerates near the and Mamolar; Fig. 4E & F) or developed as exten­
base of the succession are monomict, dominated by sive sheets resting on conglomerate bodies (as at
angular 2-4-cm clasts of grey micrite. Elsewhere, Quintanilla de las Vinas).
the conglomerates are polymict ('polygenetique' of The laminae are irregular and undulating, consist­
the French usage; Salomon, 1982). Clast lithologies ing of 1-mm alternations of light and darker grey
include: (rarely black) micrite. Many of the laminated hor­
1 fossiliferous yellow-orange packstones containing izons display a fabric of horizontal 1-mm pale beige,
0.5-mm quartz grains and brachiopod fragments- concentrically laminated carbonate tubules 0.5 mm
these are derived from the marine lower Callovian; in diameter filled with sparry calcite.
2 pink to grey micrites, commonly partially At Senora de Brezales and at Espej6n (Fig. 1), a
recrystallized- these may represent subaerially succession of marly, brecciated and heavily mottled
weathered clasts of marine carbonate from the carbonate horizons is interbedded with sheet-flood
Talveila Formation, Middle Jurassic, or Liassic; and channelized conglomerates. These sediments
3 pink oolitic grainstones- these are especially display a complex reticulate fabric produced by
common in the southwest of the study area, and horizontal 2-20-mm layers of laminar calcrete and
were probably derived from Middle Jurassic shallow vertical tubular structures 1-10 em in diameter and
water carbonates; up to 1 m in length. Some of the vertical structures
4 rarer, small ( < 1 em) clasts of red sandstone- are composed of homogeneous micrite, but others
these may be derived from red sandstones in the show an internal fabric similar to that of the laminar
Senora de Brezales Formation, or from Triassic red­ horizons. Between these structures, the mottled
beds. marly, or sandy, matrix commonly contains abun­
The conglomerates show few sedimentary struc­ dant fine carbonate strings and spar-filled angular,
tures, and are mostly clast-supported. Matrix­ branching cracks. At Senora de Brezales, stacked
supported conglomerates occur at a few localities, as laminar horizons form mound-like structures 5 m
at Senora de Brezales and La Gallega (Navas road), across with a relief of 0.5 m (fig. 9a in Wright et a/.,
but even in these cases there is little mud and the 1988).
matrix is comprised of sandstone.
228 N.H. Platt

unravelling the depositional history of the study

Interpretation
area. Effects of local faulting were certainly signifi­
The carbonate horizons within the Senora de cant in the localized preservation of strata beneath
Brezales Formation are interpreted as calcareous the unconformities.
palaeosols. The laminated limestones are similar to The Senora de Brezales Formation rests on the
laminar calcretes described from modern exposure most important intra-Mesozoic unconformity in the
surfaces in Florida by Multer & Hoffmeister (1968). study area. This unconformity records a period of
Although similar in some respects to laminar stro­ uplift, faulting and erosion that began some time
matolites, an absence of diagnostic algal structures, after the early Kimmeridgian. It marks a major
such as calcified filaments (Read, 1976), and their change in basin configuration, and a change from
association with pedogenetic features, suggests inter­ relatively uniform, gentle subsidence and generall.y
pretation as biogenic laminar calcretes or 'rhizolites' shallow marine conditions from the Rhaetian to
(Wright et a!., 1988). The fine spar-filled tubules are early Kimmeridgian, to the more rapid, strongly
interpreted as root structures (rhizoliths). The beige localized subsidence and continental sedimentation
micrite coating may either be a replacement of the that characterized the latest Jurassic and Early
root tissue itself (internal septae commonly form the Cretaceous.
characteristic 'alveolar texture' of Esteban, 1974), The combined effects of tectonic activity and ero­
or may be made up of layers of concretionary car­ sion during this period resulted in the formation of
bonate precipitated around the root. A more detailed a faulted surface with significant relief. Thus the
description of the laminated horizons is to be found Senora de Brezales Formation lies unconformably
in Wright et a!. (1988). on a variety of different strata, from Bathonian-­
The thickness and abundance of laminar calcrete Callovian shallow-marine limestones, to upper
horizons towards the top of the formation point to Oxfordian to lower Kimmeridgian marginal marine
long periods of root colonization, both on the surface sandstones and limestones of the San Leonardo and
and beneath it. These times were punctuated by Talveila Formations.
occasional input of fine-grained distal alluvial clastic Where the Senora de Brezales Formation rests on
material, producing an alternation of laminar cal­ Kimmeridgian and Callovian limestones, it reslts
cretes with mottled sandstone or marl. Desiccation variably upon karst surfaces on palaeohighs and on
and the action of larger roots were probably respon­ an erosive channelled base in the lows. Where sand­
sible for most of the observed brecciation of the stones of the San Leonardo Formation are truncated
carbonate horizons. by the unconformity, the basal contact is planar
Brecciated, mottled carbonate horizons are also without the development of palaeokarst or significant
common within the basal part of the overlying relief (Platt, 1986).
Rupelo Formation, and in places there may be an
upward transition between the Rupelo and Senora
Karst surface
de Brezales formations. Although Platt ( 1986) con­
sidered these to be two essentially separate units, The Senora de Brezales Formation lies on karstified
they may be partial lateral equivalents. lower Kimmeridgian limestones at Talveila (Mauthe,
1975; Fig. 40), Pinilla de los Barruecos and Horte­
zuelos (Fig. 1). At all these localities, red internal
GEOMETRY OF sediment fills cavities and joints within the underlying
LATE JURASSIC UNCONFORMITIES carbonate rocks. At Hortezuelos, the karstified sur­
face is cut by a small fault with a throw of 1 m. Here,
The study of Upper Jurassic facies distributions is deeper karstification and/or partial erosion of the
complicated by the presence of at least two import­ fault scarp led to the local deposition of a 1-2-m
ant unconformities, one beneath the San Leonardo lens of coarse chaotic conglomerates.
Formation (probably late Callovian to middle At Castrovido in the north of the study area
Oxfordian in age) and a later one beneath the (Fig. 1), a karst surface is developed on marine
Senora de Brezales Formation (probably intra­ Callovian limestones in interchannel areas. The
Kimmeridgian). Further discontinuities may exist marine Jurassic carbonate surface is locally dis­
within the San Leonardo Formation. The superpos­ coloured and brecciated and laminar calcretes rest
ition of these unconformities causes problems in on the unconformity. A similar situation occurs
 Alluvial fans, palaeokarst and tectonics 229

at Mambrillas de Lara (Fig. 1), where the lower gian marine carbonates were exposed. Erosion of
Callovian limestone surface is draped by laminar these highs acted to equalize the pediment relief.
calcretes penetrating downwards for 30 em along A schematic NW-SE stratigraphical section for
joints and cracks (see Wright et al., 1988). At Pinilla the Senora de Brezales Formation in the north of
de los Moros (Fig. 4H), a karstified surface of lower the study area (Fig. 6) demonstrates the tectonic
Callovian limestone is penetrated by vertical cracks control on facies that is evident after restoration to
1 em across filled with red sandstone. the base of the overlying Rupelo Formation. The
At these localities, subaerial exposure and erosion presence at Jaramillo Quemado of coarse basal con­
prior to deposition of the Senora de Brezales For­ glomerates with clasts of lower Callovian limestone
mation resulted in karstification of underlying lime­ up to 1 m in size (Fig. 4C) probably records erosion
stones. Red internal sediment was piped into joints of a faulted high to the west of the Jaramillo­
and karstic cavities; laminar calcretes formed at the Covarrubias Fault, a major NE-SW transfer struc­
surface and along joints and cracks beneath. ture (Platt, 1990). At this locality there may be a
gradation to karst conglomerates, where brecciation
was followed by only limited transport, so that the
Planar contact
locally derived clasts were deposited more-or-less in
To the north of Hortezuelos, and at Mamolar and situ, infilling an irregular karstic relief. Red sand­
La Gallega (Fig. 1), red sandstones of the Senora de stones are restricted to the area to the east of the
Brezales Formation rest on yellow sandstones of the fault, whereas immediately to the west, between
San Leonardo Formation and paleokarst is not devel­ Valparaiso and Mambrillas de Lara, karst surfaces
oped. A S-cm-thick, dark red, ferruginous-cemented are developed on the marine Jurassic, and the Senora
sandstone layer is locally present at the contact. de Brezales Formation is thin or absent. Moving
The poorly defined contact between the red sand­ further to the west, conglomerates are present firstly
stones and the underlying yellow sandstones suggests on isolated fault blocks and eventually in fault­
that the red sandstones may have been at least controlled channels.
partially derived by reworking of the San Leonardo Many of the conglomerate-filled channels are
Formation. bounded on one or both sides by NE-SW faults
(Fig. 6); good examples occur at Castrovido,
Quintanilla de las Vinas, and between San Leonardo
Channels
de Yagiie and Hontoria del Pinar (Fig. 1). These
Channels filled with polymict conglomerates occur minor faults are parallel to the Jaramillo-Covar­
at the base of the Senora de Brezales Formation in rubias Fault. At La Gallega, a NE-SW fault with a
the north of the study area, where the San Leonardo throw of 25 m on the western side of the Navas road
and Talveila Formations are absent. bounds a depression filled with Senora de Brezales
The channels record powerful stream erosion of Formation conglomerates and sandstones. Many
the limestones beneath. The siting of channels on more minor faults occur, typically at a 100-m spacing.
the downthrown sides of NW-SE trending faults These faults have throws of only 2-10 m, but still
(Fig. 6) suggests that they formed in structural lows controlled lateral variations in the subcrop to the
produced by the Late Jurassic faulting. unconformity and exerted strong local influence
on facies development and thicknesses within the
Senora de Brezales Formation.
TECTONIC CONTROLS Changes in facies and thickness also occur across
ON FACIES DEVELOPMENT NW-SE faults. Although the exposure is not con­
tinuous in the dip direction and the basin structure
The style of sedimentation within the Senora de only allows comparison between outcrops on parallel
Brezales Formation strongly reflected the pattern of anticlines several kilometres apart, general contrasts
fault blocks and the lithologies exposed. Where the are evident. Thick successions with red sandstones
sandstones of the San Leonardo Formation were up to 40 m thick occur near the southwestern basin
eroded, planation of the contact ensued and rework­ margin, whereas thinner successions are present to
ing led to the deposition of red sandstones. Karst the north. In the north of the study area, the Senora
surfaces developed on fault-bounded structural de Brezales Formation is thicker to the south of the
highs, where Middle Jurassic and lower Kimmerid- Quintanilla-Castrovido Fault than immediately to
230 N.H. Platt

NW m ottled I brecciated lim estone SE

MARINE JURASSIC PEDIMENT

JCF
Quintanilla Mambrillas Valparaiso Jaramillo Pinilla de Castrovido
de las Vinas de Lara Hortiguela Quemado los Moros

KEY ?Berriasian Rupelo Formation non-marine limestone & marl

(Las Vinas Member)
............................................................................................................... J.+.�+----············-·······-·-············-··············
red sandstone

?Kimmeridgian Senora de Brezales conglomerate

- Berriasian Formation

laminar calcrete
..................................................•....•............•....•••.................................• ee;;r;<;i;l·.................•....................................

Lower Callovian top of Marine Jurassic marine limestone

Fig. 6. Cartoon strike section between Quintanilla de las Vinas and Castrovido, illustrating tectonic control on lateral
changes in facies development within the Senora de Brezales Formation. Section restored to the base of the Rupelo
Formation to eliminate the effects of post-depositional faulting of Lower Cretaceous and Tertiary age (see Platt, 1990).
Horizontal distance between Quintanilla de las Vinas and Castrovido is 30 km; maximum vertical thickness of the Senora
de Brezales Formation is 50 m.

the north. Red sandstones are well-developed to the clast shapes and abundance of coarse carbonate
east of the Jaramillo-Covarrubias Fault between clasts suggests limited clast transport; clast derivation
Jaramillo Quemado and Castrovido but are very was chiefly from erosion of the marine Jurassic
thin or absent immediately to the northwest of the limestones of underlying or adjacent fault blocks.
fault. These lateral variations suggest that both The vertical succession may therefore record the
NW-SE and NE-SW fault sets were probably progressive erosional history ('unroofing') of the
active during deposition of the Senora de Brezales source areas.
Formation, defining a complex mosaic of fault blocks 1 Locally, as at Hortezuelos, a basal conglomerate
with differing subsidence histories and different is present, containing clasts only of grey micritic
detailed stratigraphies. limestone. These clasts appear to have been derived
from laminar calcretes deposited in interchannel
areas and from Talveila Formation lagoonal
EROSION, CLAST DERIVATION carbonates.
AND DEPOSITIONAL HISTORY 2 Red sandstones above consist mainly of quartz,
probably including a component reworked from the
The composition of the conglomerates within the sandstones of the San Leonardo Formation.
Senora de Brezales Formation strongly reflects the 3 Polymict conglomerates above these contain
subcrop to the unconformity (Fig. 7). The irregular abundant reddened oolitic limestone clasts, indi-
 Alluvial fans, palaeokarst and tectonics 231

DEPOSITIONAL HISTORY CLAST DERIVATION

4
Valanginian - quartz sandstones and conglomerates- Talveila Formation:

����!��� ���!�: �; ���!?.���

Cenomanian progressive unroofing and erosion of
----- - -- - _ _ - - - -----

• :1111������ �;;;���;,��=�:.
· ···············

San Leonardo Formation:

.....'!��-��!��!-�-��!�-�-�-���-�!��!:�----

:���
Middle Jurassic:
Formation limited clastic supply shallow-marine limestones

..................... --------------------- --·-············-·----------------·

Lower Jurassic:
polymict conglomerates -
marine limestones and marls
reworking of marine Jurassic,
especially Middle Jurassic
Senora de
Brezales red sandstones - reworking Triassic:
Formation of San Leonardo Fm continental, mostly sandstones

monomict basal conglomerate -

-------------------------------- ��:-y-���!�-�-?!.!.����i!� .F.�.: .................
.

Fig. 7. U pper Jurassic-lower cretaceous clast derivation . Clastic rocks- vertical succession illustrating likely derivation of
clasts from the successive 'unroofing' and erosion of underlying stratigraphical units. See text for explanation. The
schematic section shown is a composite based on outcrops on the northeastern flank of the Hortezuelos anticline.

eating erosion and reworking of Bajocian-Callovian this may reflect lengthier transport in more perma­
shallow marine carbonates. Their deposition may nent river systems developed under more humid
therefore date the 'unroofing' of the underlying conditions.
Middle Jurassic.
Similar polymict conglomerates containing abun­
dant coarse limestone clasts occur in a number of ENVIRONMENTAL SYNTHESIS
areas, in each case recording local clast derivation
from the marine Jurassic. Examples occur in the Figure 8 presents a schematic depositional model for
north of the study area, as described above (Fig. 6). the Senora de Brezales Formation, showing the
The San Leonardo Formation is absent in the north, main environments represented and the control of
but the local presence of red sandstones, as at the unconformity subcrop and of tectonics on facies
Castrovido, may record transport of eroded material development. Sedimentation took place under a
elsewhere from the San Leonardo Formation or semi-arid climate. Small-scale wadi-type channels
from basement lithologies. were incised directly into the underlying marine
In the southwest of the study area, the presence of limestone pediment, but much of the sandstone and
quartz pebbles may record erosion of quartzitic conglomerate was deposited from laterally extensive
material from the San Leonardo Formation or from sheet-floods spreading out from these channels.
basement rocks on the Duero block to the south. Pedogenetic carbonates developed on the pediment
However, widespread erosion of Triassic sandstones in interchannel areas, and also formed as surficial
and Palaeozoic quartzites probably did not com­ layers and nodular brecciated horizons on or within
mence until much later (Platt, 1989a). The low the exposed alluvial clastic deposits. Bioturbation
clastic content in the freshwater carbonates of the and root penetration acted to destroy primary
overlying Rupelo Formation suggests that this phase sedimentary structures, particularly within the
of peneplanation was complete by the Berriasian, sandstones.
although the catchment area geology was still appar­ The association of channels and sheet-floods is
ently dominated by carbonate lithologies. The first consistent with deposition in small-scale alluvial fans.
wide-spread erosion of the Triassic and Palaeozoic is Basin-bounding NW-SE faults and NE-SW trans­
probably dated by the appearance of sandstones fer faults both influenced sedimentation, which acted
within the Barremian-Aptian Hortigiiela and Pied­ to infill fault-bounded depressions; clastic supply
rahita de Mufi6 Formations (Figs 2 & 7). These came dominantly from the erosion of neigbouring
rocks do not contain Jurassic limestone clasts- uplifted blocks.
232 N.H. Platt

pedogenesis - laminated crusts

& paleosols on alluvial substrates

cross-graben wrench small fans sourced

or "transfer" faults at fault intersections

laminated crusts on
interchannel areas
of pediment
\
channels cut
into pediment -
located along
transfer faults

basin-bounding
extensional normal faults _.-

U
I
Oxfordian arenites

erosion of Mid. Jurassic sequence

- polymict conglomerates

2 localised erosion of Lr Kimmeridgian limestones

- monomict conglomerates

Fig.8. Schematic depositional model for the Senora de Brezales Formation, illustrating sedimentary environments
represented and the influence of tectonics and the underlying geology on facies development and clast derivation.

REGIONAL PERSPECTIVE Brezales Formation occur at the base of the laterally

equivalent Aguilar Formation (see Platt & Pujalte;
1994). Here, too, the subcrop to the Late Jurassic
Upper Jurassic clastic rocks
unconformity was dominated by marine Jurassic
The Kimmeridgian-Tithonian saw a major re­ limestones developed in facies similar to those of the
gression in Iberia and across much of northwest Cameros. Faulting and continental erosion then
Europe. In the Lusitanian Basin of Portugal (Wison permitted the deposition of locally derived, tec­
et al., 1989), tectonic movements led to an acceler­ tonically controlled conglomerates of similar
ation of subsidence; the lacustrine-lagoonal car­ composition.
bonates of the Oxfordian Caba'<os and Vale Verde Upper Jurassic continental clastic sediments are
Formations were superseded by thick Kimmeridgian also reported from several basins to the west of
clastics (Hill, 1989). In the Asturias Basin .. on the Britain on the northern Biscay margin, where they
·
north Spanish coast (Valenzuela et al., 19,86), con­ form potential hydrocarbon reservoir targets. Exam­
glomerates, red sandstones and palaeosols similar to ples occur in the Celtic Sea Basins (Millson, 1987;
those of the Senora de Brezales Formation occur in Petrie et at., 1989; Shannon, 1991) and the Brittany
the poorly dated La Nora Formation. These are Basin of the Southwestern Approaches (Evans,
overlain by marginal marine deposits of lower . 1990). in these areas, a major period of Late Jurassic
Kimmeridgian age, passing up into Upper Jurassic extension led to fragmentation of the basins into
fiuvio-deltaic deposits of the Lastres Formation. In fault blocks, which were tilted and eroded. These
neighbouring Cantabria, polymict conglomerates Late Jurassic fault block structures form a major
and sandstones similar to those of the Senora de exploration target in the Porcupine Basin of offshore
 Alluvial fans, palaeokarst and tectonics 233

Ireland (MacDonald et al., 1987). In the Porcupine, Timoleague Formation show similarities with the
the Kimmeridgian saw a change in tectonic regime, heterogeneous marginal marine succession devel­
as widespread broad subsidence gave way to fault­ oped in the San Leonardo and Talveila Formations.
controlled subsidence and sedimentation. This evol­ Like the western Cameros, the North Celtic Sea
ution is parallel to that recorded by deposition of the saw a change to continental conditions during the
Senora de Brezales Formation in the western Kimmeridgian. Here, the coarse- to fine-grained,
Cameros Basin. locally conglomeratic sandstones of the Clonakilty
Moore (1992) recognized an Upper Jurassic synrift Formation may correspond with the clastics of the
to post-rift transition succession up to 500 m thick Senora de Brezales Formation in the Cameros. This
within faulted sub-basins of the Porcupine Basin. interpretation is further supported by Millson's
These rocks are characterized on seismic data as (1987) descripton of thin, concretionary, silty micritic
mounded, fan-style deposits resting on a faulted carbonate horizons, possibly representing calcretes,
unconformity, passing laterally and upwards into within the Clonakilty Formation.
parallel, onlapping reflectors of a ?marginal marine
succession. Moore suggests that these fans may have
Late Jurassic unconformities
significant reservoir potential, particularly in the
sand-prone proximal-mid- and mid-fan regions. Many authors have recognized a Late Jurassic to
Proximal areas of the fan contain poor reservoir­ Early Cretaceous unconformity in northwest Europe
quality unsorted conglomerates (Shell well 34/19), (see e.g. Ziegler, 1990). This 'late Cimmerian'
whereas the distal part is expected to contain fine unconformity, which is variably dated as earliest
siltstone and shale. The clastic fans were bounded Cretaceous or Volgian-Tithonian, is known from
by faults on one side, passing basinwards into finer the North Sea (Rawson & Riley, 1982), Porcupine
grained distal fan deposits and onlapped by Lower Basin (Croker & Shannon, 1987; Tate, 1993), North
Cretaceous shales. Although this succession was Celtic Sea (Millson, 1987; Shannon, 1991; McCann
deposited in a marginal marine setting, the facies & Shannon, 1993), and Cantabria (Platt & Pujalte,
geometries developed are essentially identical to 1994). Accurate dating of this surface is commonly
those inferred for the alluvial clastics of the Senora hindered by problems of stratigraphical resolution in
de Brezales Formation (Fig. 9) and the overlying marginal/restricted Upper Jurassic facies; for exam­
low-gradient lacustrine deposits of the Rupelo ple, in the study area, the basal Senora de Brezales
Formation. Formation unconformity can be dated only as post­
In the North Celtic Sea (Millson, 1987), inter­ lower Kimmeridgian and pre- lower Berriasian.
bedded marginal marine carbonates and littoral and Superposition of several unconformities means
continental clastics of the mid- to upper Oxfordian that the time gap represented by the unconformity is

Clastic fans Onlapping infill Clastic fans

Basal unconformity

Fig.9. Sketch showing inferred stratigraphical relationships in the Upper Jurassic Senora de B rezales Formation of the
western Cameros Basin. Figure is modified after a cartoon by Moore ( 1992) depicting seismic relationships in the Upper
Jurassic synrift to post-rift transition of sub-basins within the Porcupine Basin (Moore, 1992). The two successions share a
common tectonic evolution and are thought to show e ssentially identical geometry.
234 N.H. Platt

of variable extent in different areas. In basinal areas were derived from localized reworking of lower
of the North Sea and the Wessex Basin, the 'base Kimmeridgian limestones, red sandstones from
Cretaceous unconformity' is represented by a con­ erosion of upper Oxfordian sandstones, and polymict
densed section of Volgian-Ryazanian (Tithonian­ conglomerates were derived from erosion of the
Berriasian) age (Rawson & Riley, 1982; Ruffell, entire Jurassic succession, in particular the Middle
1992). However, on structural highs and in areas Jurassic.
adjacent to the Biscay rift, Lower Cretaceous rocks The Senora de Brezales Formation offers a useful
may rest directly upon Liassic or Triassic rocks, as outcrop analogue for Upper Jurassic clastic suc­
occurs in western parts of the Celtic Sea basins cessions in basins bordering the northern Biscay
(Millson, 1987), the Western Approaches (Evans margin to the west of Britain. A depositional model
1990) and northern Cantabria (Garcia-Mondejar developed for the Senora de Brezales Formation
et a!., 1985; Hines, 1988). (Fig. 9) serves to illustrate the importance of seismic­
Study of the Cameros Basin indicates that the scale and subseismic-scale fault movements in con­
complex association of superimposed unconformities trolling sedimentation and facies distributions and
in the areas around Biscay is the product of vari­ highlights the strong lateral variability so character­
ations in relative sea-level (notably falls in the early istic of synrift deposition.
Oxfordian and ?mid-Kimmeridgian and rises in
the late Oxfordian and ?late Kimmeridgian to
Berriasian), coupled with widespread Late Jurassic ACKNOWLEDGEMENTS
extensional block faulting. The effect of these,
reflected in the basal contact, deposition and deri­ My sincere thanks go to Dr Harold Reading for
vation of the Senora de Brezales Formation, was to support and encouragement. I am grateful to the
produce truncation during times of lower base level, students from Oxford and Bern Universities who
particularly on the footwall crests of extensional helped with detailed mapping and logging. Helpful
fault blocks, and complex sedimentary onlaps during reviews and comments were provided by Guy Plint,
infilling of local accommodation space created by Alastair Ruffell, Bruce Sellwood and Pat Shannon.
faulting or rises in relative sea-level. My special thanks go to Adrian Forrer for discussion
and good humour in the field.

CONCLUSIONS

REFERENCES
The Senora de Brezales Formation is a laterally
variable Upper Jurassic succession of continental CHADWICK, R . A .( 1 985) End Jurassic-early Cretaceous
red-beds deposited under a semi-arid climate. Con­ sedimentation and subsidence (late Portlandian to
glomerates and sandstones were deposited in wadi­ Barremian), and the late Cimmerian u nconformity. I n :
Atlas of Onshore Sedimentary Basins i n England and
type channels and by sheet-floods in small-scale
Wales (Ed. Whittaker, A . ) , pp. 9-25. British Geological
alluvial fans. Bioturbation, desiccation and root Survey, Keyworth.
action led to homogenization of the waterlain sand­ CLEMENTE, P. & PEREZ-ARLUCEA, M. ( 1993) Depositional
stones during protracted subaerial exposure. Pedo­ architecture of the Cuerda del Pozo Formation, Lower
genetic carbonates represent palaeosols and laminar Cretaceous of the extensional Cameros Basin, north­
central Spain. J. sediment. Petrol. , 63, 437 -452.
calcretes formed within and at the base of the clastic
CROKER, P.F. & SHANNON, P . M. ( 1987) The evolution and
succession. hydrocarbon prospectivity of the Porcupine Basin, Off­
The Senora de Brezales Formation rests on the shore Ireland. In: Petroleum Geology of North West
late Cimmerian unconformity, which truncates a Europe (Eds Brooks, J. & Glennie, K . ) , pp. 633- 642.

thick succession of marine Jurassic carbonates. Graham & Trotman, London.

DfAz, M. , GoY, A . & YEBENES, A . ( 1983) Un complejo de
Lateral variations in thickness in the Senora de isla barrera-lagoon del Jurasico superior (Talveila, Soria).
Brezales Formation reflect Late Jurassic extensional X Congreso Nacional de Sedimentologfa, Menorca,
block faulting; many channels are fault-bounded Abstracts, 2.58-2.62 .
and the coarsest conglomerates occur adjacent to EsTEBAN, C . M. ( 1 974) Caliche textures and 'Microcodium'.
Boll. Soc. geol. !tal. , 92 (Suppl . ) , 1 05- 125.
major faults. The deposition of the Senora de
EvANS, C.D. R . ( 1 990) United Kingdom Offshore Regional
Brezales Formation recorded erosion of the under­ Report: the Geology of the Western English Channel and
lying marine Jurassic; monomict conglomerates its Western Approaches. British Geological Survey,
 Alluvial fans, palaeokarst and tectonics 235

HMSO, London. Western Cameros Basin , Province of Burgos, northern

GARclA-MONDEJAR, J . , HlNES, F.M . , PU J ALTE , v. & Spain. Unpublished PhD. thesis, University of Oxford,
READING, H.G. ( 1985) Sedimentation and tectonics in 250 pp.
the western Basque- Cantabrian area (northern Spain) PLATT, N.H. (1989a) Continental sedimentation in an
during Cretaceous and Tertiary times. Excursion 9. In: evolving rift basin: the Lower Cretaceous of the Western
Excursion Guidebook 6th European Regional Meeting, Cameros Basin (Northern Spain). Sediment. Geol. , 64,
International Association of Sedimentologists , Lleida, 9 1 - 109.
Spain (Eds Mila, M.D. & Rosell , J . ) , pp. 308-392. PLATT, N . H . ( 1989b) Lacustrine carbonates and pedo­
GOMEZ FERNANDEZ, J.C. & MELENDEZ, N. (1994) Estra­ genesis: sedimentology and origin of palustrine deposits
tigrafia de Ia 'Cuenca de Cameros' (Cordillera Iberica from the Early Cretaceous Rupelo Formation, Western
Noroccidental, N de Espana) durante el transito Jun1sico­ Cameros Basin, Northern Spain. Sedimentology , 36,
Cretacico . Revista de Ia Sociedad Geologia de Espana , 7, 665-684.
12 1 - 139. PLATT, N.H. ( 1989c) Climatic and tectonjc controls on
HILL, G. (1989) Distal alluvial fan sediments from the sedimentation of a Mesozoic lacustrine sequence : the
Upper Jurassic of Portugal: controls on their cyclicity Purbeck of the West Cameros Basin, Northern Spain.
and chanel formation. 1. geol. Soc. London, 146, Palaeogeogr. Palaeoclimatol. Palaeoecol. , 70, 187- 197 .
539-555. PLATT, N.H. ( 1 990) Basin evolution and fault reactivation
HINES, F.M. ( 1988) The sedimentation, tectonics and stra­ in the western Cameros Basin, Northern Spain. 1. geol.
tigraphy of the Cretaceous/ Tertiary sequence of north­ Soc. London, 147, 165- 175.
western Santander, northern Spain. Unpublished PhD PLATr, N.H. & PuJALTE V. ( 1994) Correlation of Upper
thesis, University of Oxford, 250 pp. Jurassic- Lower Cretaceous of continental sequences
HISCOlT, R . N . , WILSON, R.C.L . , GRADSTEIN, F. M . , from the southern Biscay margin northern Spain. 1. geol.
PUJALTE, V . , GARciA-MoNDEJAR, J . , BouDREAU , R . R . , Soc. London, 1 5 1 , 7 15 -726.
e t al. (1990) Comparative stratigraphy and subsidence PLATT, N . H . & WRIGHT, V.P. ( 1991 ) Lacustrine carbonates:
history of Mesozoic rift basins of North Atlantic. Bull. facies models, facies distributions and hydrocarbon
Am. Assoc. petrol. Geol. , 74, 60-76. aspects. In: Lacustrine Facies Analysis (Eds Anad6n, P.
KAMERLING, P. (1979) The geology and hydrocarbon habi­ & Cabrera, LI.), Spec. Pubis int. Assoc. Sediment . ,
tat of the Bristol Channel Basin. 1. petrol. Geol. , 2, N o . 13, pp. 57-74. Blackwell Scientific Publications,
75-93 . Oxford.
MACDONALD, H., ALLAN, P.M. & LOVELL, J.P.B. ( 1 987) PLATT, N.H. & W RIG HT , V.P. (1992) Palustrine carbon­
Geology of oil accumulation in Block 26/28, Porcupine ates, calcretes and the Florida Everglades: towards
Basin, Offshore Ireland. In: Petroleum Geology of an exposure index for the fresh-water environment?
North West Europe (Eds Brooks; J . & Glennie, K . ) , 1. sediment. Petrol. , 62, 1058- 107 1 .
p p . 643-65 1 . Graham & Trotman, London. PLATr, N . H . , BURNS, S . J . , PLATT, J . D . & WIGNALL , P.B.
MAUTHE, F. ( 1975) Palaokarst im Jura der Iberischen ( 1991 ) Organic-rich deposits from the Callovian of the
Ketten (Prov. Soria, Nordspanien). Neues 1ahrb. Geol. Sierra de Ia Demanda, Northern Spain. Eclogae geol.
Palaeontol. Abh. , 150 , 354-372. Helv. , 84(2), 483-508.
McCAN N , T. & SHANNON, P.M . ( 1993) Lower Cretaceous RAwsoN , P . F. & RILEY, L.A. ( 1 982) Latest Jurassic-Early
seismic stratigraphy and fault movement in the Celtic Cretaceous events and the 'late Cimmerian' unconform­
Sea Basin. First Break , 1 1 , 335-344. ity in the North Sea area. Bull. Am. Assoc. petrol.
MENSINK, H., & ScHUDACK, M. ( 1982) Caliche, Boden­ Geol. , 66, 2628-2648.
bildungen und die Palaogeographischen Entwicklung in READ, J . F. ( 1976) Calcretes and their distinction from
der westlichen Sierra de los Cameros. Neues 1ahrb. stromatolites. In: Stromatolites (Ed. Walter, M . R . ) ,
Geol. Palaeontol. Abh . , 163, 49-80. pp. 255 -27 1 . Developments i n Sedimentology, 20,
MILLSON, J . A . ( 1987) The Jurassic evolution of the Celtic Elsevier, Amsterdam.
Sea Basins. In: Petroleum Geology of North West Europe RuFFELL, A.H. & CowA RD , M . P. ( 1 992) Basement tecton­
(Ed. Brooks, J. & Glennie, K.), pp. 599...:.610. Graham ics and their relationship to Mesozoic megasequences in
& Trotman, London. the Celtic Seas and Bristol Channel area. In: Basins on
MooRE, J .G. ( 1 992) A syn-rift to post-rift transition the Atlantic Seaboard: Petroleum Geology, Sedimen­
sequence in the Main Porcupine Basin, offshore western tology and Basin Evolution (Ed. Parnell, J . ) , Spec. Pub!.
Ireland. In: Basins on the Atlantic Seaboard: Petroleum Geol. Soc. London, Vol. 62, pp. 385-394.
and Basin Evolution (Ed. Parnell, J . ) , Spec. Pub!. geol . RuFFELL, A . H . ( 1992) Early to mid-Cretaceous tectonics
Soc. London, No. 62, pp. 333-349. Geological Society and unconformities of the Wessex Basin (southern
of London, Bath. England). 1. geol. Soc. London, 149, 443-454.
M u LTER, H.G. & HoFFMEISTER, J . E . ( 1968) Subaerial SALOMON, J. ( 1 982) Les formations continentales du
laminated crusts of the Florida Keys. Geol. Soc. Am. Jurassique superieur-Cretace inferieur (Espagne du
Bull. , 79, 183- 192. Nord, Chaines Cantabrique et NW Iberique). Mem.
PETRIE, S . H . , BROWN, J . R . , GRANGER, P.J. & LOVELL, Geol. Univ. Dijon , 6, 228 pp.
J.P.B. ( 1989) Mesozoic history of the Celtic Sea basins. ScHUDACK, M. ( 1987) Charophytenftora und fazielle
In: Extensional Tectonics and Stratigraphy of the North Entwicklung der Grenzschichten mariner Jura/Wealden
Atlantic Margins (Eds Tankard, A.J. & Balkwill, H . R . ) , in den Nordwestlichen Tberischen Ketten (mit Verglei­
Mem . A m . Assoc. petrol. Geol . , Tulsa, 46, 433-444. chen zu Asturien und Kantabrien). Palaeontogr. Abt.
PLATT, N.H. ( 1986) Sedimentology and tectonics of the Ser. B, 204, 1 - 180.
236 N . H. Platt

SHANNON, P.M. ( 1991 ) The development of Irish sedimen­ J. & SUAREZ DE CENT!, C. ( 1 986) The Jurassic sedimen­
tary basins. J. geol. Soc. London , 148, 1 8 1 - 189. tation in Asturias. Trab. Geol. Univ. Oviedo, 16,
SuN S.Q. (1989) A new interpretation of the Corallian 1 2 1 - 132.
(Upper Jurassic) cycles of the Dorset coast, southern WILDE, S. (1988) The Bathonian and Callovian of the
England. Geol. Mag. , 24, 139- 158. northwest Iberian Range: stages of facial and palaeo­
TALBOT, M.R. ( 1973) Major sedimentary cycles in the geographical differentiation on an epicontinental plat­
Corallian Beds (Oxfordian) of Southern England. form. Cuad. Geol. Iberica, 14, 1 13- 142.
Palaeogeogr. Palaeoclimatol. Palaeoecol. , 14, 293-317. WILSON, R.C.L . , HISCOTT, R.N. , WILLIS, M . G . &
TANKARD, A.J. & BALKWILL, H . R. (eds) ( 1 989) Extensional GRADSTEIN , F.M . (1989) The Lusitanian Basin of west­
Tectonics and Stratigraphy of the North Atlantic Margins. central Portugal: Mesozoic and Tertiary tectonic, strati­
Mem. Am. Assoc. petrol. Geol . , Tulsa, 46, 639 pp. graphic and subsidence history. In: Extensional Tectonics
TATE, M.P. ( 1993) Structural framework and tectono­ and Stratigraphy of the North Atlantic Margins (Eds
stratigraphic evolution of the Porcupine Seabight Tankard, A.J. & Balkwill, H.R.), Mem. Am. Assoc.
Basin, offshore western Ireland. Mar. petrol. Geol. , 10, petrol. Geol . , Tulsa, 46, 341 -361.
95 - 123. WRIGHT, V . P . , PLATI, N . H . & WIMBLEDON, W . A . ( 1 988)
THALMA N N , C. ( 1989) Die Geologie entlang der San Biogenic laminar calcretes: evidence for calcified root
Leonardo- Oberschiebung zwischen San Leonardo de mat horizons in palaesols. Sedimentology, 35, 603-620.
Yagiie und Cubilla, NW-Keltiberische Ketten, Provinz ZIEGLER, P.A. ( 1 990) Geological Atlas of Western and
Soria, Spanien .
Unpublished Diplom-Arbeit, Universitiit Central Europe, 2nd (and completely revised) edn,
Bern, Teil B, pp. 126-270. 2 vols. Shell Internationale Petroleum Maatschappij BY,
VALENZUELA, M . , GARclA-RAMOS J . C . , GONZALEZ-LASTRA, The Hague. 239 pp.
Spec. Pubis int. Ass. Sediment. (1995) 22, 237 - 2 64

The use of geochemical data in determining the provenance

and tectonic setting of ancient sedimentary successions:
the Kalvag Melange, western Norwegian Caledonides

RODMAR RAVNAS and HAR ALD FURNES

Geological Institute, Univers ity of Bergen, Allegt. 41, N-5007 Bergen, Norway

ABSTRACT

The inferred Silurian Kalvag Melange is a thick (> 2500 m ) bedded olistostromal melange, which
formed due to gravitational resedimentation driven by tectonic activity along the steep submarine slope
of a basin margin and was deposited in the lower slope or base-of-slope environment. The melange
consists of olistoliths, more than 2.8 km in length and up to 0.5 km in thickness, of shallow-marine
deposits, deep-marine turbidites, bedded cherts, pebbly mudstones, calcareous rocks, lava-flow basalts
and andesites, rhyolites and rhyolitic ignimbrites embedded in a sheared olistostromal 'groundmass'.
The latter comprises chiefly debris-flow deposits. Interfingering channel-fill conglomerates and gully-fill
turbidites with erosional contacts to the olistostromal 'groundmass' are present. The channel-fill
conglomerates are dominated by debris of lithic-volcanic arenites and volcanic rocks. Olistoliths of
volcanics and the volcanic debris in the channel-fill conglomerates were derived from alkaline, calc­
alkaline , !AT, MORll and boninitic sources.
Geochemical and isotope data from the volcanic rock olistoliths reflect an evolution from magmatism
associated with initial rifting to incipient back-arc spreading, and finally to volcanic arc magmatism.
Geochemical analyses of conglomerate debris and sedimentary-rock olistoliths suggest derivation from
an evolved, mature island arc, and an ophiolitic source. The volcanic arc is thought to have formed by
rifting of accreted terranes along an active continental margin, and the MORJJ-type and boninitic debris
most likely represent derivation from an older, accreted ophiolitic terrane.
The sedimentological and geochemical data jointly suggest that the deposition of the Kalvag
Melange took place in a back-arc basin , formed by progressive cannibalization of the volcanic arc apron .
Based on the present knowledge of the development of marginal basins within the outboard terranes of
the Caledonian orogen, the basin-evolution history recorded by the Kalvag Melange can be correlated
with the history of other late Ordovician to early Silurian marginal basins.

INTRODUCTION

The primary control that tectonics exerts on the Dorsey, 1988; Smith et al., 1988; Haughton &
production, d ispersal, accumulation and lithification Halliday, 1991; Marsaglia et al., 1992). Intercalated
of sediments has long been recognized (Pettijohn volcaniclastic rocks and clasts in coarse conglom­
et al., 1987). Provenance regions, and especially erates can be characterized and sometimes dated,
those forming parts of the overriding plate at destruc­ providing supplementary geochemical and isotopic
tive plate margins, can u ndergo significant and rapid information about the hypothetical provenance area
uplift. Much information is lost due to subsequent and the original plate tectonic setting of the depo­
erosion, deformation and/or tectonic displacement, sitional basin.
and a complete in s itu record of the source rocks is The mineralogical and geochemical composition
rarely preserved. The sedimentary record of adjacent of terrigenous sediments and sedimentary rocks
sedimentary basins can provide some constraints depend chiefly upon provenance, relief and climate
toward redressing this preservational bias (e.g. of the source region(s), weathering, transport

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
237
238 R. Ravnas and H. Fumes

processes, basin geometry and morphology, depo­ cambrian migmatitic orthogneisses (Jostedalen
sitional e nvironment and diagenesis (e. g. Bhatia, Complex), paragneisses and associated augen­
1983; Basu, 1985; Helmold, 1985; McBride, 1985; gneisses, amphibolites, eclogites and meta­
Mutti, 1985; Normark, 1985; Ricci Lucchi, 1985; anorthosites (Fjordane Complex), and low- to
Dorsey, 1988; Johnsson & Stallard, 1989; McLennan medium-grade metamorphic sedimentary, volcanic
et a/., 1990). Plate-tectonic setting has been advo­ and plutonic rocks (Askvoll Group) of the Western
cated as the primary control on sediment com­ Gneiss Region. A quartz diorite from the Askvoll
position (Blatt et a/., 1980; Pettijohn et a/., 1987) Group has yielded a middle Proterozoic magmati.c
and the link between plate-tectonic setting and U-Pb age of 1640. 5 ± 2.3Ma (Skaret a/., 1994).
modal framework of modern sediments (mainly The middle tectonic unit is separated from the
sands and sandstones) appears empirically robust lower by the extensional Kvamshesten Detachment
(e. g. see Crook, 1974; Schwab, 1975; Dickinson & Zone. It consists of the Dalsfjorden Suite, composed
Suczek, 1979; Ingersoll & Suczek, 1979; Dickinson of various types of syenitic to charnockitic ortho­
& Valloni, 1980; Valloni & Mezzardi, 1984; gneisses, granites and gabbros (Kolderup, 1921),
Dickinson, 1985; Valloni, 1985). The nature of the and the unconformably overlying pre-Silurian
primary framework grains can, however, be modified and Silurian metasediments of the H�yvik and
significantly, obliterated or destroyed during post­ Herland Groups, respectively. The H�yvik Group
depositional diagenesis, burial, deformation and/or experienced polyphase deformation and associated
metamorphism. The geochemical composition, metamorphism prior to deposition of the u nconform­
element ratios and isotope composition of sediments ably overlying Herland Group (Brekke & Solberg,
seem less depe ndant upon these post-depositional 1987; Andersen et a/., 1990).
processes (Hiscott, 1984; Robertson & Henderson, The upper tectonic unit comprises the Solund­
1984; Argast & Donnelly, 1987; Larue & Sampayo, Stavfjord Ophiolite Complex and its cover of meta­
1990), and as such have proved powerful and more sediments and metavolcanics (Fumes eta/., 1990). It
reliable in provenance studies and discrimination is separated from the middle tectonic unit by a
of plate-tectonic setting of ancient sedimentary composite terrane, the Sunnfjord Melange, inter­
successions (Bhatia & Taylor, 1981; Bhatia, 1983; preted to have developed during obduction of the
Bhatia & Crook, 1986; Roser & Korsch, 1986; ophiolite complex upon an e arly to middle Silurian
McLe nnan et a/., 1990). continental margin represented by the Herland
The present paper discusses the provenance and Group (Berg, 1988; Andersen et a/., 1990; Alsaker
tectonic setting of an olistostromal melange, the & Fumes, 1994). The Solund-Stavfjord Ophiolite
inferred Silurian Kalvag Melange of the western Complex has yielded a U - Pb age of 443 ± 3 Ma
Norwegian Caledonides (Fig. 1), on the basis of data (Du nning & Pedersen, 1988). Single zircon U-Pb
obtained from the melange's ' groundmass', olisto­ datings from its cover sediments have yielded both
liths (of volcanic and sedimentary rocks) and cobbles Archaean, Proterozoic and early Ordovician ages
and boulders of interbedded polymictic conglom­ (Pedersen & Dunning, 1993).
erates. Secondly, an attempt is made to further The Kalvag Melange is separated from the
detail the reconstruction and probable evolution of gneisses, metasediments and metamorphic plutonic
the melange's provenance region and depositional and volcanic rocks of the lower, middle and upper
setting by integrating the present knowledge of the tectonic units to the north by a major thrust
island arc/ophiolite terranes and marginal basins of zone (Hartz et al., 1994). The metasediments on
the Scandinavian Caledonides. Bremanger have bee n correlated with the H�yvik
Group (Fumes et a/., 1990; Hartz et a/., 1994). The
metamorphic igneous rocks include a number of
GEOLOGICAL SETTING small serpentinite bodies, amphibolites, green­
stones, gabbros and schists (Bryhni et a/., 1981;
The area between Solund and Bremanger (Fig. 1) Hartz et a/., 1994), which may possibly correlate
forms part of the westernmost Norwegian Caledo­ with the Sunnfjord Melange and/or the Solund·­
nides. Three tectono-stratigraphical units are Stavfjord Ophiolite Complex and its cover. More­
recognized: the lower, middle and upper tectonic over, Hartz et a/. (1994) recognized major E-W
units (Brekke & Solberg, 1987; Andersen et a/., striking, southward dipping extensional shear
1990). The lower tectonic unit comprises Pre- zones along the northern and central parts of
 TECTONOSTRATIGRAPHY

Devonian
�

UPPER
TECTONIC
� Sunnfjord
Melange (SM)
UNIT

� Herland Gp. �
� MIDDLE Cl
"'
TECTONIC
Hoyvik Gp.
� 0
UNIT
Dalsfjord Suite
g.
Western Gneiss Region LOWER
TECTONIC �
c:;·
undifferentiated
UNIT

LEGEND
�

� !}
I:..!..:....J Devonian
s
LEGEND Q �����g��;f�� -diorite
...,
'0'
•
[ill
Younger igneous instrusions
Kalvc\g Melange
Granodiorite
D Sunnfjord Melange �
� Gabbronorite/diorite UPPER �
[ill] �e�����j��::;:g ���f � s
nic
TECTONIC
s·
The KalvAg Melange (SSOC & cover/S�1?)
UNIT
V7777i1
A
Melange �gmundmass"
�
Cover of metasediments
and lavas/intrusions �·
L_j �
Coarse conglomerates
Solund-Stavfjord OphiolitE �
�
..:
cs
Bedded olistostromal facies Complex (SSOC)

OHstoHhs
�
� Herland Group (HG) "'

Qm ;:s
Arenitic shallow-marine sediments
� :::>
.
(protolith blocks) Hoyvik Group

G:::::J
;:s
Chert
R
Q
Dalstjord Suite
Deep-marine sandy turbidites r=l
Q
Western Gneiss Region LOWER
L=...J undifferentiated TECTONIC
Ignimbrite/rhyolite UNIT
....L.....J..
O sasalt Detachments between
Lower. Middle and

.
0 1 km 0 5 10 15 20 25 km .....1-..1- Upper Plates
'
---Fault

Fig. 1. Simplified geological map and tectono-stratigraphy of the So lund- Bremanger area, compiled from Fumes eta/. ( 1990), Hartz eta/. (1994) and Osmundsen &
Andersen ( 1994) , and simplified geological map of Frs;lya, modified from Bryhni & Lyse ( 1985).

t;5
\0
240 R. Ravnas and H. Fumes

Bremangerlandet (Fig. 1) that separate major plates, terms of its two principal components: the bedded
analoguous to the tectono-stratigraphy of the olistostromal 'groundmass', and the olistoliths that
Sunnfjord area (Brekke & Solberg, 1987; Andersen are scattered in this groundmass. The olistoliths,
et al. , 1990; Andersen & Jamtveit, 1990; Osmundsen which may exceed 2. 8 km in length and up to 0. 5 km
& Andersen, 1994). in thickness, comprise shallow-marine (offshore to
The Kalvag Melange has been interpreted in terms shoreface) sandstones and mudstones, deep-marine
of a tectonized olistostrome formed in an island arc sandy turbidites, bedded cherts, pebbly mudstones,
or accretionary prism setting (Bryhni & Lyse, 1985; calcareous rocks, subaerial and subaqueous lava··
Fumes et al., 1990), and was tentatively included in flow basalts and andesites with interstitial calcareous
the upper tectonic unit by Fumes et al. (1990). The rocks and intercalated sediments, rhyolite and
melange succession is intruded by two plutons of rhyolitic ignimbrite (Fig. 2A-E). The olistoliths are
gabbronoritic/dioritic and granodioritic composition interpreted to have been emplaced as large slides
(Fig. 1). The older gabbronorite/diorite has yielded and slumps. The melange 'groundmass' comprises
a Sm-Nd age of 380 ± 26 Ma (Fumes et al., 1989), mainly matrix-rich sedimentary breccias and gravelly
thus providing a minimum age for the melange mudstones (Fig. 2F & G) that are inferred to
formation. represent chiefly cohesive debris-flow deposits.
The uppermost stratigraphical level of the region The interbedded conglomerates and sandstones
is represented by early to middle Devonian clastic (Fig. 2H), interpreted to have been deposited from
sediments, deposited in late to post-Caledonian gravelly non-cohesive debris-flows and high-density
collapse basins (Hossack, 1984; Norton, 1986; turbidity currents and sandy high- to low-density
Seranne & Seguret, 1987; Andersen & Jamtveit, turbidity currents respectively, commonly have
1990). In the Bremanger area these sediments rest sharp, erosive basal contacts, locally displaying large
unconformably upon the pre-Devonian rocks as well loadcasts. They form broad, lenticular bodies that
as on the granodioritic pluton. locally interfinger laterally with the olistostromal
facies of the melange's 'groundmass', and are inter­
preted to represent submarine slope channel- and
THE KALVAG MELANGE gully-fill deposits. Black shales and subordinate fine··
grained turbidites, the latter probably representing
The Kalvag Melange forms a more than 16-km-wide overbank deposits of the submarine channels, form
rock unit on the island s of Fqllya and Bremanger minor constituents of the melange succession.
(Fig. 1). On Fr¢ya, the Kalvag Melange makes up a Locally the 'groundmass' is intruded by fine- to
minimum 2500-m-thick succession consisting of medium-grained sandstone dykes and sills.
clasts and large detached blocks (olistoliths) of native The Kalvag Melange is interpreted to have been
and exotic rocks (sensu Raymond, 1984) d ispersed formed due to gravitational resedimentation driven
in a pelitic 'groundmass' , and interbedded conglom­ by tectonic activity along the steep submarine slope
erates and sandstones (Figs 1 & 2) (Ravnas, of a basin margin, and deposited in the lower slope
1991). Graptolites and tabulate corals of probable to base-of-slope environment (Ravnas, 1991).
Ordovician-Silurian age were reported by Reush Several points of evidence suggest that the entire
(1903) and Kolderup (1928), but have not been melange succession chiefly is of sedimentary origin.
confirmed by later studies. The melange succession On the island of Fr¢ya the melange d isplays a
shows no pervasive shear fracturing, but local zones pervasive, although subtle, bedding defined by
of pervasively sheared matrix occur. Strong tec­ textural and compositional variations. The general
tonic deformation, including folding, faulting and stratigraphical way-up direction is recognizable,
boudinage, is limited to discrete, isolated zones (fault and there seem to be some large-scale thickening/
zones and shear fracture zones). Primary structures thinning upwards trends. These trends are correla­
within olistoliths and the fragmental fabric of the tive throughout the island area, marked predomin­
melange 'groundmass' are generally well preserved antly by the stratigraphical thickness distribution of
within the contact aureoles of the two plutons, the m ass-flow facies (Ravnas, 1991). The slide and
allowing distinction between different types of olisto­ slump units (olistoliths) stand out as isolated, out­
liths and recognition of their primary depositional sized peaks on this background, and are correlative
environments. (Fig. 3) only when associated with the maxima of the
The melange as a whole can be considered in mass-flow facies thickening. The olistoliths that are
 Geochemical data for determining provenance 241

encased in the bedded 'olistostromal' matrix show national basalt standards and the recommended
primary depositional contacts, and are in some cases values of Govindaraju (1984) for calibration.
draped by black shales. In some places, gradations Possible post-depositional alteration of basaits, vol­
down the inferred palaeoslope from sedimentary­ caniclastic materials and sediments have been dis­
rock olistoliths of slide/slump origin into debris-flow cussed in a number of studies. The elements Ti,
deposits can be seen. The matrix of the melange is P, V, Zr, Y, Nb and Cr, with the addition of Si, Fe,
thought to have been derived from the fine-grained Mg, La, Th, Ce and Co in sedimentary rocks, are
interlayers of the shallow marine sediments in considered relatively stable during low-grade meta­
addition to slope mud (see below). Partly similar morphism (e.g. Cann, 1970; Coish, 1977; Shervais,
clast lithologies are present in the interfingering 1982; Bhatia, 1983; Staudigel & Hart, 1983; Hiscott,
channel-fill conglomerates and the olistostromal 1984; Robertson & Henderson, 1984; Bhatia
'groundmass' or debris-flow deposits. Locally the & Crook, 1986; Cas & Wright, 1988; Larue &
overbank and channel-fill deposits show evidence of Sampayo, 1990) and are used systematically through­
slumping prior to deposition from the subsequent out this study. Ca, Na and K in sedimentary rocks
flow, and the olistostromal 'groundmass' deposits show increased mobilization with increasing meta­
locally are found as olistoliths and conglomerate morphic grade (Larue & Sampayo, 1990) and thus
debris. The spatial distribution of facies associations are less useful as indicators of primary composition.
and olistoliths (Fig. 3) suggests that the melange The melange succession, forming mainly a biotite­
developed during a series of successive resedimen­ quartz hornfels (Bryhni & Lyse, 1985), shows
tation events (Ravnas, 1991). prograde contact metamorphism towards the two
Basal and internal penecontemporaneous, 'soft­ plutons, especially the gabbronoritic/dioritic pluton,
type' deformation (folding, faulting and boundinage) which has a border zone up to 150 m thick of migma­
of sedimentary-rock olistoliths is attributed to their tized and foliated host rock (Fumes et a!., 1989). To
emplacement by sliding and slumping while still in a minimize the effect of remobilization of elements
semi-consolidated state. The volcanic-rock olistoliths due to the increase in metamorphic grade, most
show little internal deformation, apparently due to samples were collected well away from the two
their inherently rigid character. Sandstone dykes intrusive bodies. To avoid possible systematic com­
show cross-cutting relationships with the sheared positional variations of sediments related to vari­
olistostromal 'groundmass' and locally are found to ations in grain size (Roser & Korsch, 1986; Argast &
intrude the deformed basal part of sedimentary-rock Donnelly, 1987), chiefly medium- to fine-grained
olistoliths. High-pressure-low-temperature miner­ arenites (greywackes) were selected for analyses.
alogy has not been observed, neither in olistoliths
nor in the sheared matrix.
Conglomerate debris of channel-fill conglomerates
The melange has been divided broadly into a
lower part comprising mainly sedimentary slides, Clasts of well-rounded igneous and well-rounded
slumps and cohesive debris-flow deposits, and an and angular sedimentary rocks are present, rep­
upper part comprising mainly cohesive debris-flow resenting gabbro, diorite, metabasalts, trachyande­
deposits and subordinate slumps, fine-grained turbi­ site/dacite, granite, rhyodacite/rhyolite, bedded
dites and resedimented coarse conglomerates. The chert, quartzite, subarkose, lithic arenite, lithic­
lower part contains few or no volcanic-rock olisto­ volcanic arenite and wacke, black shale and
liths, whereas the upper part shows an upward calcareous rocks. The chemical analyses of rep­
increase in the content of volcanic-rock olistoliths. resentative samples are presented in Table 1.
The metabasalts can be divided into three different
types based on their geochemical characteristics
GEOCHEMISTRY (Fig. 4, Table 2). These are MORB (mid-ocean
ridge basalts), tAT (island arc tholeiites) and rocks
Chemical analyses of samples from olistoliths, approaching boninitic composition. All types show
conglomerate debris and the olistostromal 'ground-· negative Nb anomalies and exhibit either a flat
mass' were carried out by X-ray fluorescence. The pattern or show slight or pronounced depletion of P,
glass bead technique of Norrish & Hutton (1969) Zr, Ti andY in the MORB normalized, multi-element
was used for major element oxides, and pressed variation diagrams (Fig. 4A), typical of back­
powder pellets for the trace elements, using inter- arc basin basalts, tAT and boninites respectively
242 R. Ravnas and H. Fumes

OLISTOLITH COMPOSITION

Alkaline potassic
meta basalts
Subalkaline
MORB-Iike metabasalts
Calcareous rocks

Bedded chert

Transitional type
metabasalts

Large olistoliths of Ignimbrite and

native and exotic IAT metabasalts
blocks

Deep-marine
sandy turbidites

'Background' sedimentary
mass of bedded olistostrome
(gravitational melange);
---contacts sedimentary,
non-ophiolitic matrix
E and rock clasts/blocks
8
"'
N

Shallow marine
siliciclastic sediments

Alkaline potassic metabasalts,

Ignimbrite and bedded chert

Shallow marine
siliciclastic sediments

Fig. 2. (Above. ) Composite profile through the Kalvag Melange, compiled from measured sections along the eastern and
western coast of Frll\ya. (Right. ) Locations of field photographs are indicated by letters. (A) Flow-folded, welded
ignimbrite. Lens cap is 6cm in diameter. (B) Massive andesitic and rubbly aa-type basaltic flows. Length of hammer is
40cm. (C) Shallow-marine (shoreface) sediments; interbedded inferred wave-formed ripple cross-laminated and parallel
stratified sandstones. Scale bar is 5 em. (D) Normally graded, sandy Bouma-type Tabc and Tab turbidites. Lens cap is 6em in
diameter. (E) Evenly bedded, turbiditic ribbon chert. Length of hammer is 40 cm. (F) Disintegration of olistolith of
shallow-marine deposits into (G) clast-rich, debris-flow deposits of the melange 'groundmass'. Scale bars of (F) and (G) are
20cm and Scm respectively. (H) Clast-supported conglomerate representing inertial, non-cohesive, high-density gravelly
debris-flow. Note polymictic clast population. Length of hammer is 40cm.
Geochemical data for determining provenance 243
 t
B
A c

----
- ? --
----
--
-
? -

2000
+---- 2.0km ___ _.·����:�Jg�6�c��IC RO��§ts !D D
• 2.5km
-o-- CHANNEL-/GULLEY
oo o o"';--
FILL DEPOSITS
1500
��-�:-<�:
. ..:.; · ---<<
. ...:-:-:�.::.:�:�
.
!J . ·. : .- :·- : : : ·(::-: :. ·� .. . ·. .· . ·. ·
: -. ·�:-:
- : · 1500: �
"�-·����OLI TH OF DEEP-MARINE
SANDY TURBIDITES 1500
..,
. · .
�
?. · . ·. · ·
. . .
..
--<-
.

�
.
/
'/
2.8km -------!�

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

Stratigraphic
;_.:.-·-
height :;_,
.--:--=--.� �.-.-,. ------- ___ _.,..-,""'":"-:- .-.-:
uncertain
. . E :;_,

--��-�-� �)1 "'

""'

/
;:s

: :- ;,·, !:;·
<:... . . ·
_.;./ · "'
;:s
Stratigraphic\..·...:....:...:_>­ : "'-
height
500 �
�
uncertain

·�· �
: �

:;i�::.:::.�::::-:-:::.-:-::.:-6
�:
.�
("

:
17
.:: & &

Fig. 3. Spatial distribution of olistoliths within the Kalvag Melange. Possible correlations are marked by dashed lines . Sections A and B are generalized logs
interpreted from the map, whereas sections C, D and E are simplified versions of measured sections described by Ravnas (1991).
Table 1. Representative geochemical analyses of conglomerate debris
Sample no. Location Debris type Si02 Ti02 Al203 Fe203" MnO MgO CaO Na20 1<20 P205 LOI Total v � � � � � � & y li • � � � � Th

49,60 1,65 14,43 12,60 0,29 6,92 9,01 2,86 1,00 0,13 0,79 99,29 232 203 54 45 20 109 41 157 35 103 31 19 705
R8-157 Mulen MORB
R8-163 Mulen MORB 51,37 1,18 16,58 9,34 0,23 5,79 10,62 2,84 0,32 0,10 0,30 98,66 175 308 39 62 3 88 8 -� 71 24 18 233

50,52 1,47 13,65 12,98 0,24 6,59 9,23 2,76 0,66 0,12 0,60 98,8� 237 257 56 78 6 72 20 1� 38 83 24 21 260
R8-164 Mulen MORB
R8-166 Mulen MORB .50,23 1,55 15,13 11,62 0,26 6,53 9,80 3,59 0,44 0,17 0,34 99,65 221 346 50 101 3 101 7 =n 88 23 17 147 7

MORB 49,00 1,91 14,34 13,05 0,34 6,77 9,93 2,48 0,25 0,10 0,88 99,05 269 314 54 51 23 117 2 195 39 114 34 19 260 4
R9-17 Mulen
R9-21 Mulen MORB 49,82 1,63 14,19 12,51 0,30 7,21 9,07 2,65 1,12 0,13 0,70 99,34 234 288 53 66 11 114 42 160 37 99 28 15 566 5

MORB 48,47 2,30 14,35 13,87 0,28 6,63 7,65 2,13 1,12 0,21 1,66 98,66 295 167 60 48 17 171 45 110 51 163 53 12 345 16
R9-15 Mulen
14,88 10,22 0,23 5,09 8,51 2,73 0,32 0,13 0,95 98,19 195 170 43 35 36 149 6 -47 150 39 25 81 8
R9-16 Mulen MORB 53,74 1,40

IAT 53,22 0,79 22,86 4,83 0,08 2,32 7,13 4,62 1,07 0,11 1,87 98,89 99 48 18 22 61 43 550 10 32 18 28 836
R8-155 Mulen
R8-156 Mulen IAT 51,87 0,55 17,06 14,48 0,26 3,91 4,80 3,81 1,41 0,03 1,20 99.40 250 68 63 27 11 95 57 161 13 32 36 1 480 C)
130 128 185 21 34 32 17 651
"'
6,91 4,87 3,71 2,70 'o.o5 1,49 98,47 175 178 49 61
R8-162 Mulen IAT 49,71 0,58 18,05 10,23 0,17 0
R8-165 Mulen IAT 53,64, 0,98 15,28 10,10 0,24 5,88 10,10 3,21 0,25 0,13 0,55 100,36 197 277 40 75 73 2 312 25 43 24 17 232 4 (")
�
199 1050 62 120 349 66 66 9 26 47 4 374 "'
Mulen boninitic 51,33 0,33 13,58 13,23 0,23 9,04 3,74 1,62 1,34 0,03 3,96 98,42
R9-18
52,58 0,28 17,71 10,51 0,19 6,23 4,13 3,94 1,57 0,03 1,91 98,57 177 57 52 27 158 78 154 11 28 27 13 328 �
R9-22 Mulen boninitic
R8-172 Mulen boninitic 53,61 0,26 16,93 9,95 0,20

0,16.
6,30

7,93
3,98

7,40
3,89

3,04
1,48

1,98
0,07

0,03
2,56

1,30
99,23

98,59
173

116
54

170
52

43
22

53
165

51
73

82
163

313
10

10
24

24
29

25
14

16
M8

1192
[
R8-153 Mulen gabbro 50,23 0,30 17,04 9,18
R8-154 Mulen diorite 54,89 0,25 14,45 9,37 0,17 '5,36 6,77 3,12 1,13 0,03 0,75 96,29 144 18 41 23 18 71 48 251 8 35 21 15 922 §­
!:)
12,41 2,76 0,03 0,61 2,88 5,81 0,32 0,03 0,58 98,30 38 57 22 14 201 10 87 10 39 220 7 18
R8-158 Mulen granite (co) 72.69 0,18
R8-159 Mulen granite (co) 77,45 0,13 11,24 1,51 0,01 0,39 1,85 4,65 0,64 0,02 0,90 98,80 33 27 27 14 145 21 95 24 42 506 13 8 'Cj'
....
11,32 3,55 0,04 0,97 2,99 4,57 0,38 0,04 0,70 100,12 42 7 10 46 32 12 186 20 306 24 38 161 9 10
R9-23 Mulen granite (co) 75,36 0,22
R9-24 Mulen granite (co) 72,48 0,29 12,56 3,59 0,06 1,09 2,89 5,20 0,63 0,08 1,15 100,02 51 15 10 20 40 23 215 25 96 20 38 216 15 3
f}
R8-161 Mulen granite (m) 76,13 0,16 12,00 2,21 0,03 0,58 2,30 5,51 0,44 0,04 0,55 99,86 40 12 67 21 12 141 11 91 38 47 213 19 29
1i)
�·
11,36 3,22 0,04 1,17 2,37 4,32 0,69 0,07 1,16 99,65 38 4 23 31 18 231 17 35 18 35 174 13 5
R9-30 Mulen granite (m) 75,03 0,22
71,06 0,27 12.55 5,52 0,10 1,44 2,34 5,08 0,80 0,08 0,59 99,82 50 17 5 26 63 24 67 20 71 24 43 310 10
R8-167 Mulen 1gn./rhyol.
R8-168 Mulen ign./rhyol. 76,09 0,22 11,37 2,66 0,04 0,39 2,88 3,98 0,40 0,03 1,44 99,50 33 14 5 29 26 8 202 18 79 19 35 105 9 s
R8-169 Mulen dac./tr.and. 69,24 0,29 14,37 2,56 0,06 0,94 4,55 5,27 0,40 0,14 0,75 98,56 47 42 15 17 62 13 231 12 150 11 78 47 82 33 27 �-
R9-38 Mulen dac.ltr.and. 67,05 0,43 16,52 3,77 0,07 1,74 4,50 4,18 0,42 0,23 1,02 99,93 69 45 12 23 15 70 16 255 18 218 14 144 69 118 56 44
'1::3
R8-170 Mulen chert 100,60 0,02 0,35 0,27 0,01 0,04 0,08 0,22 0,01 O,Q2 0,23 101,83 M � 14 10 3 28 39 12 \3
H U 16 11 3 143 39 ""
R8-171 Mulen chert 98,51 0,_03 0,33 0,51 0,02 0,04 0,53 0,53 0,02 0,53 101,05 "'
R9-27 Mulen chert 98,00 0,05 1,39 0,63 0,02 0,12 1,04 0,50 0,02 0,30 102,05 29 13 4 12 55 5 47 41 11 ;:s
0,01 14 5 13 6 9 4 75 50 14 !:>
0,59 0,34 0,01 0,08 0,04 0,51 0,04 0,13 101,44 25
R9-28 Muten chert 99,65 0,03 ;:s
0,26 1,24 0,07 0,02 0,30 99,76 � � 5 12 48 6 84 6 47 15 11
R9-36 Mulen chert ang. 92,54 0,07 3,67 0,61 0,01 0,96
�
subarkose? 92,06 0,07 4,26 0,32 0,02 0,16 1,52 2,03 0,18 0,90 101,53 30 16 13 13 66 59 23 54 57 17
R8-160 Mulen
R9-29 Mulen subarkose? 91,18 0,11 4,59 1.24 0,30 1,04 1,84 0,21 0,02 0,47 101,02 M � 17 14 65 146 10 39 111 10

p. mudst. 55,95 0,77 17,42 6,98 0,09 3,08 4,40 4,54 1,54 0,15 3,54 98,47 234 196 31 89 39 130 72 261 24 156 11 36 30 517 13 17
R8-151 Mulen

bl. mudsh. 73,74 0,47 10,46 3,83 0,04 1,53 2,90 2,69 1,00 0,11 2,49 99,24 171 151 14 53 62 69 37 241 15 89 4 19 26 432 17 10
R8-152 Mulen
Mulen bl. mudsh. 68,44 0,50 12,01 5,52 0,04 2,05 2,78 2,73 1,51 0,06 2,38 98,03 231 149 21 69 33 95 65 133 22 103 5 28 32 363 22 9
R9-33

x-lam arenite 76,70 1,62 8.81 4,84 0,07 1.62 1,94 2,52 1.22 0,09 0,75 100,18 89 54 16 23 84 51 93 24 1529 43 52 47 265 22 26
R8-175 Mulen

Mulen lit.-volc. aren. 58,92 0,97 16,04 8,17 0,08 3,38 2,76 4,45 1,63 0,10 1,75 98,24 140 180 38 45 29 105 69 262 34 255 14 90 40 727 29
R9-31
Mulen tit.-volc. aren. 63,31 0,94 13,96 7,86 0,10 3,12 3.02 3,31 1,34 0,09 1.18 98,22 129 187 33 36 26 97 55 220 29 242 11 72 45 627 30 10
R9-32

LOI, loss on ignition; Fe.o; , total iron as Fe.03 + 1 . 1 x FeO.

�
 A 8

COMENOITE
I
CD
PANTELLEAITE
100 I
\ PHONOLITE
'
'

............
RHYOLITE
10 '
0

0,1

Zr/Ti02 0

0,1
- - -� --- �
0,01
ANDESITE I BASALT /
. ., . /
®
100 -------- ALKALI­

.... .. . . -
····-......
BASALT

SUB-ALKALINE BASALT
�

10 ;;_�// 0,001 -l-------+--+---1

/ 0,01 0,1 10
-��2�----- Nb/Y

1000 ®
0,1

I
,,,,
I
0,01 I
I
I D
I
I
100
® ppm Cr 100
I
I
I

\ t,
\,.rJl t: WPB \
10 II I
ID
\ IAT I\ I
a: \ I\ I
0
:IE \ I \ I
[] I
><
\ \ I
0 \ \ :
+------+-'�-���--_L-�
0
a: 10
0,1
10 100
ppmY

0,01 10 ®

100
0 Trachyandesilic/dacilic debfis
'
'
\
I
I
10 \
WITHIN \
PLATE '
'
Ti02(%) 1 I
I

0,1

100
®
Granitic and rhyolite debris
0,1 +---..C.:..---1----_____,
10 10 100 1000
ppmZr

• MORB melabasatts

a Boninilic melabasalts and

gabtxoic rocks

0,1 • tAT metabasatts

o granitic and rhyolitic

debris

• trachyandesitesl dacites
Sr K Rb Ba Nb Ce p Zr Ti y Cr
 Geochemical data for determining provenance 247

Table 2. Classification of metabasaltic and gabbroic conglomerate debris according to the discriminant plots of Fig. 4B

A B c D

Discriminant plot Metabasaltic debris Metabasaltic debris Metabasaltic debris Gabbro/diorite debris

Nb/Y-Zr/Ti02 Subalkaline andesite/ Subalkaline andesite/ Subalkaline andesite/ Subalkaline andesite/

basalt basalt basalt basalt
Y-Cr MORB/WPB LAT (MORB) LAT !AT
Zr-Ti02 (%) MORB (LATfWPB) LAT Boninite Boninite
Total MORB LAT Bonini tic Boninitic

MoRB, mid ocean ridge basalts; WPB, within plate basalts; LAT, island arc tholeiites.

(Saunders & Tarney, 1979; Basaltic Volcanism Study close to that of the rhyolite and rhyolitic ignimbrite
Project, 1981; Hickey & Frey, 1982; Cameron et al., olistoliths, thereby suggesting a common protolith
1983; Thompson et al., 1984; Crawford, 1989; (see interpretation below).
Wilson, 1989). This tripartite subdivision is also Bedded chert occurs as large angular blocks up to
suggested by the discriminant plots of Fig. 4B. The 1. 7m across. Angular cobbles of 'quartzite' are,
debris of phaneritic, mafic volcanic rocks (gabbro based on mineralog.ical and geochemical criteria,
and diorite) show a similar chemical composition to interpreted to represent disrupted chert beds derived
the boninitic samples, as suggested by their nearly from the ribbon chert. The subarkosic debris are
identical MORB normalized, multi-element variation geochemically similar to chert, but differ with respect
diagram patterns (Fig. 4A) and the discriminant plot to their coarser grain size and higher proportion of
of Fig. 4B. The gabbro and diorite are interpreted plagioclase. However, detrital sodic plagioclase
accordingly to represent the plutonic equivalents of locally is present in some chert beds of the larger
the boninitic metavolcanics (see further discussion bedded chert olistoliths, suggesting that a common
below). parent cannot be ruled out. Well-rounded quartzite
Although the value of geochemical analyses of clasts are recrystallized completely, and their
acidic volcanic rocks has been questioned (e.g. see protolith (whether representing vein quartz,
discussion in Cas & Wright, 1988, and references quartzitic sandstone or chert, see below) cannot be
therein), the chemical composition of the trachy­ determined.
andesite/dacite cobbles testifies to their alkaline Well-rounded cobbles of lithic arenites differ
affinity (Fig. 4) and demonstrates their incompati­ mineralogically and geochemically from the angular
bility with the granitic and rhyolitic rock debris. The debris of lithic-volcanic arenites. The lithic arenites
rhyolitic and fine- to coarse-grained granitic clasts consist, in decreasing order, of quartz, lithic sedi­
display a similar mineralogy (particularly diagnostic mentary grains (quartzite/quartz schist), sodic
is the presence of blue quartz in both types) and plagioclase, alkali feldspars, ilmenite, zircon and
nearly identical chemical compositions, and are minor apatite and opaques. Modal ap.alyses indicate
interpreted to represent extrusives and their intrusive either a continental craton or recycled orogen/
equivalents. Their geochemical signature (e.g. pro­ collision orogen as provenance (Dickinson, 1985;
nounced negative Nb, P and Ti anomalies, see Fig. Valloni, 1985), as is also suggested by their chemical
4A) indicates derivation from a depleted tAT parent composition. Lithic-volcanic arenites, on the other
(Pearce et al., 1984; Holm, 1985), and suggests hand, are texturally, mineralogically and geochemi­
erosion of low-K granitic rocks, typical of island cally similar to the olistoliths of shallow-marine
arcs (Ewart, 1982). Moreover, the rhyolitic debris sandstones and deep-marine turbidites (see
exhibits a mineralogical and chemical composition description and interpretation below), and are

Fig. 4. (Opposite.) (A) MORB normalized, multi-element variation diagrams of conglomerate debris of volcanic rocks.
(1) MORB metabasalts, (2) LAT metabasalts, (3) boninitic metabasalts and debris of gabbroic rocks, (4) trachyandesites/
dacites, and (5) granites and rhyolites. The order of the elements and the values of the normalizing constants are from
Pearce (1983), except for Cr, which is from Pearce (1980). (B) Discriminant plots for conglomerate debris of ( 1) volcanic
rocks, and (2) and (3) mafic volcanic rocks. Discriminant plot (B . 1 ) is from Winchester & Floyd (1977) , whereas
discriminant plots (B.2) and (B.3) are from Pearce ( 1980).
248 R. Ravnas and H. Fumes

accordingly interpreted to have been derived from metabasalts. The high K content, together with
such deposits. the inferred primary texture and mineralogy jointly
Angular clasts of pebbly mudstone and black suggest that these rocks belong to the potassic ig­
shales are geochemically similar to the olistoliths of neous rock suite (Foley et at., 1987; Wilson, 1989),
pebbly mudstone and interbedded black shales of probably representing shoshonites transitional into
the melange succession, respectively. Furthermore, leucite-tephrites (Bates & Jackson, 1987).
black shales are present only in conglomerate The original tectonic setting of these metabasalts
deposits resting with erosive basal contacts on black is difficult to assess. The high contents of Ti02 and
shale, thus suggesting the intrabasinal provenance of Nb, and to a lesser extent Zr, typical of ocean island
this clast type. Well-rounded pebbles of calcareous basalts (O!B), argue against a subduction-related
rocks have not been analysed geochemically. setting (e.g. Perfit et al., 1980; Morris & Hart,
Although an extrabasinal origin of this clast type 1983). However, high Ti02 and Nb values are
cannot be excluded, they may have been produced reported from alkaline basalts of back-arc, intraplate
by erosion of lava-flow basalts with calcite-filled settings undergoing extension (e.g. Thorpe et a/.,
interstices now present as large olistoliths in the 1984; Briggs et a/., 1990) and subduction-related
melange. ultrapotassic rocks (Thompson, 1977; Edgar, 1980;
Thompson eta/., 1984). The MORB normalized, multi­
element variation diagrams resemble closely the
Olistoliths of volcanic rocks
pattern shown by continental flood-basalts (CFB)
Geochemical analyses of samples from the different (e.g. Thompson et a/., 1984; Wilson, 1989), and
olistoliths of lava-flow basalts and andesites (Table differ from those of OtB by their lack of pronounced
3) show considerable variations, allowing subdivision positive Nb anomalies. Provided these metabasalts
into alkaline, mildly alkaline to subalkaline MORB­ represent potassic rocks, the chemical composition
like and island arc basalts (Fig. 5 & Table 4). In indicates a transitional character between lamproites
addition there are some transitional types. Isotope and potassic/ultrapotassic rocks of active orogenic
analyses were performed on samples from olistoliths zones (i. e. group I and III potassic rocks of Foley
of ignimbrite and alkaline and subalkaline MORB-like et at. (1987)), whereas the content of Nb relative to
basalts/andesites, and the results are presented in Zr strongly suggests a subduction-related origin (see
Table 5. The Es, and ENct values are calculated for discussion in Thompson & Fowler, 1986; and their
several ages (470, 450 and 430Ma) based on the Fig. 9). The calculated ENct value (Table 5) coincides
interpretation that the Kalvag Melange forms part with values reported from both OtB, potassic rocks,
of the Caledonian island arc/marginal basin terrains island arc and active continental margin volcanics.,
(see discussion below). The high E5, values of the as well as some rift-related continental flood-basalts
lava-flow basalts/andesites indicate contamination, (e.g. Faure, 1986; Wilson, 1989; and references
possibly from (a) syn- or post-lithification hydro­ therein). Accordingly, rift-related, continental
thermal solution(s) from which the calcitic infills of within-plate or back-arc, or alternatively subduction··
vesicles and interstices precipitated (Ravnas, 1991). related potassic volcanics provide the best analogues
As the volcanic-rock olistoliths occur scattered in for these alkaline metabasalts.
the olistostromal groundmass and do not form parts Subalkaline MORB-like basalts are present as com··
of a continuous succession, the geochemistry of the pound accumulations of greenschist facies aphyric to
different volcanic-rock olistolith types are described slightly phyric, vesicular, subaerial aa and pahoehoe,
separately, followed by a brief discussion of their and subaquatic, pillow lava-flows. The flow-units
possible primary tectonic setting. range in composition from basaltic to andesitic, with
Alkaline basalts comprise subaerial compound the pahoehoe- and pillowed-type flows representing
flows of pillowed and scoriaceous, porphyritic and the former. Saussuritisized plagioclase and abundant
vesicular, greenschist facies metabasalts. Phenocrysts calcite-filled amygdules are present in all flow types.
are augite/diopside, saussuritisized plagioclase, Despite some variations in geochemical signatures
primary sanidine and microcline pseudomorphs, between different olistoliths (see Fig. 5 and dis··
probably after leucite. Calcite-filled amygdules are cussion below), all metabasalts classify as sub··
common. The chemical composition, characterized alkaline MORB-like (Fig. SB, Table 3). The MORB
by high content of alkali/alkaline earth metals, Ti02, normalized, multi-element variation diagram pat­
Zr and Nb, denotes a strong alkaline affinity of these terns (Fig. SA), however, indicate a slight alkaline
Table 3. Representative geochemical analyses of volcanic rock olistoliths

Sample no. Location Rock type Si02_ Ti02 Al203 Fe203" MnO MgO CaO Na20 K20 P205 LOt Total v � � N � � � & v � m � � � � Th

47,25 2,88 21,65 9,31 0,18 3,88 7,91 1,80 4,61 0,54 1,00 100,01 272 239 45 70 35 154 143 680 38 257 47 70 34 841 15
RB-8 Sleneset Alk.basalt
51,83 2,47 19,68 5,16 0,18 2,35 13.75 0,95 3,36 0,49 1,13 100,24 193 188 21 100 42 87 56 567 30 175 37 40 54 637 20
RB-9 Sleneset Alk.basalt
52,36 2,49 19,67 5,49 0,23 2,19 13,69 0,96 2,53 0,51 1,93 100,11 199 178 32 125 57 106 34 519 29 174 35 34 36 550 33
R9-12 Sleneset Alk.basalt
2,59 19,97 4,96 0,20 2,42 9,92 1,71 5,02 0,53 1,94 99,92 230 198 26 89 13 101 74 456 31 202 37 56 41 870 36
R9-13 Sleneset Alk.basalt 52,60
53,62 2,33 18,05 5,23 0,18 2,52 10,33 0,89 6,41 0,63 1,52 100,18 199 169 25 72 12 93 101 409 34 180 36 38 41 905 14
R9-14 Sleneset Alk.basalt

53,74 2,30 23,19 6,27 0,06 3,63 4,49 2,94 3,08 0,30 7,31 100,01 292 440 34 158 101 209 47 376 35 111 17 26 34 241 0
R9-52 Kalvegj.n. MORB-Iike
R9-53 Kalvegj.n. MORB-Iike 54,56 2,16 22,84 6,27 0,06 3,42 4,40 3,49 2,46 0,31 6,69 99,98 286 439 30 128 49 119 41 371 37 104 17 29 41 167 14 CJ
R9-54 Kalvegj.n. MORB-Iike 58,44 1,96 19,64 8,74 0,13 3,99 0,00 0,88 5,86 0,38 9,40 100,01 255
268
455
409
41
34
160
178
83
17
146
121
127
98
265
591
36
38
113
129
17
20
23
26
29
45
546
510
10
�
3,19 3,86 0,45 7,80 100,02
R9-55
R9-56
Kalvegj.n.
Kalvegj.n.
MORB-Iike
MORB-Iike
53,44
57,57
2,17
2,00
22,65
20,74
7,71
7,00
0,12
0,21
3,69
3,47
2,74
2,33 2,97 3,23 0,50 12,54 100,03 238 430 28 224 18 120 76 514 42 105 17 17 46 307 s.
�

95 26 13 309
::!
MORB-Iike 50,13 1,60 18,15 12,54 0,16 7,02 6,17 1,84 2,27 0,14 3,26 100,03 283 28 52 21 20 123 40 316 36
RB-48 Botnan.
RB-50
RB-51
Botnan.
Botnan.
MORB-Iike
MORB-Iike
52,64
51,15
1,58
1,59
19,60
18,55
10,12
12,04
0,08
0,11
5,90
8,40
3,93
3,31
4,99
4,04
0,96
0,64
0,21
0,20
3,70
4,70
100,00
100,03
210
229
647
633
52
60
277
266
84
74
107
114
20
12
407
338
29
26
87
90
32
29
24
12
178
113
[
RB-54 Botnan. MORB-Iike 52,62 1,64 20,32 10,58 0,20 4,30 3,04 3,51 3,50 0,31 7.87 100,01 249 437 49 239 86 122 78 404 38 102 11 34 38 431
!}
R9-58 Botnan. MORB-Iike 50,33 1,72 19,39 11,05 0,10 8,12 4,06 3,77 1,24 0,21 3,42 99,99 243 742 59 306 52 124 25 380 29 98 10 36 20 220 9
!.:)
0,10 6,51 6,29 4,08 1,39 0,21 1,33 100,01 198 560 47 212 56 99 29 388 26 87 9, 35 20 254 10
R9-59 Botnan. MORB-Iike 51,90 1,53 18,49 9,50
R9-60 Botnan. MORB-Iike 54,05 1,63 19,80 8,94 0,06 5,69 4.97 3,56 0,85 0,41 6,50 99,96 216 574 41 247 67 108 16 446 30 84 10 19 27 160 �
...,

RS-95 Mulen Trans. bas. 51,24 1,12 18,87 7,55 0,35 3,78 15,61 0,75 0,12 0,28 3,84 99,66 247 156 27 41 38 88 3 454 26 48 4 20 28 61 20
�
RB-223 Minnet Trans. bas. 50,60 1,91 18,59 12,07 0,24 9,17 2,78 3,22 1,11 0,31 2,33 100,00 229 302 59 93 28 104 40 446 29 157 10 70 11 787 11
�
R9-89
R9-90
Mulen
Mulen
Trans. bas.
Trans. bas.
55,44
53.15
0,99
1,21
15,78
18,05
10,71
10,09
0,29
0,26
5,10
5,64
6,53
6,06
4,60
4,29
0,25
0,87
0,27
0,29
1,99
1,07
99,94
99,91
189
203
257
290
47
43
85
56
48
88
110
89
5
24
221·
361
26
27
81
103
6 34
35
27
27
264
1251
12
�
;:;:·

RB-18
RS-19
Sjeneset
Sj0neset
tAT
tAT
51,00
49,09
1,03
1,09
18,14
21,03
10,67
11.90
0,30
0,24
5,45
6,35
9,69
4,27
2,95
1,89
0,53
4,00
0,25
0,19
0,72
6,71
100,01
100,03
168
262
235
77
47
53
94
32
16
63
104
141
15
111
367
447
25
20
64
55
30
24
15
17
117
384 0 �-
RB-20 Sjeneset tAT 50,45 1,01 17,66 12,93 0,26 7,13 5,82 1,76 2,78 0,28 4,53 100,06 242 54 58 26 39 131 82 329 19 44 36 21 264 8 �
RB-21 Sjeneset tAT 50,63 1,14 18,25 11.42 0,22 6,56 7,22 3,13 1,26 0,27 5,37 100,09 239 50 49 25 50 130 32 475 22 51 23 16 195 13 ;j
19,41 11,39 0,22 6,74 7,05 2,49 1,84 0,21 4,19 100,02 234 64 50 31 44 140 78 248 17 48 27 30 448 1 ""
R9-1 Sjeneset tAT 49,63 1,04 �
;::s
68,52 0,24 15,37 4,89 0,03 3,21 0,28 1,22 2,15 0,01 2,70 98,62 28 2 18 201 63 140 94 400 59 47 1742 22 11 !:)
Mulevik lgn.matrix
RB-102
4 155 45 174 47 353 4 18
;::s
0.23 14,61 3,45 0.03 1.95 0,65 4,70 1,18 0,02 1,55 99,30 27 4 13 50 49 682 24
70,92
RS-103
RS-104
Mulevik
Mulevik
lgn.matrix
lgn.matrix 70,87 0,23 14.53 3,52 0,03 2.00 0,64 3,96 1,18 0,02 1,49 98,47 28 1 10 5 101 43 146 62 383 5 42 46 719 17 11 �
0,22 14,14 4,91 0,02 3,16 0,28 2,76 1,24 0,02 2,08 98,87 30 2 18 4 161 35 215 115 390 10 70 58 841 22
RB-105 Mulevik lgn.matrix 70,05
0,02 3,24 0,97 4,81 0,67 0,02 1,83 101,40 29 8 18 6 179 16 378 81 348 5 49 48 395 16
R9-84 Mulevik lgn.matrix 71.70 0,20 12,73 4,84

83,54 0,13 7,98 2,56 0,02 0,41 0,14 4,20 0,02 0,01 0,56 99,57 25 12 61 37 48 217 19 40 20 11
R9-82 Mulevik Rhy. core
73,18 0,23 14.18 2,02 0,02 0,31 0,46 7,59 0,04 0,02 0,65 �8.70 30 7 86 65 73 352 68 66 16 27 11
RB-40 Mulevik Rhyolite

66,58 0,22 15,07 6,14 0,05 3,63 0,35 1,40 2,87 0,02 2,56 - 98,88 28·- 24 155 80 101 57 392 54 34 3319 10
RB-98 Mulevik Ash-layer
65,47 0,24 16,36 6,04 0,05 4.07 0,29 1,12 3,74 0,00 2,75 100,20 28 23 159 106 87 72 414 46 42 4180 4
R9-85 Mulevik Ash-layer

LOI, loss on ignition; F�o; , total iron as Fe203 + 1 . 1 x FeO; Listed values of major oxides of mafic volcanic rocks are recalculated values to 100% at LOI = 0.

�
250 R. Ravnas and H. Fumes

A
Q)/>;�
COMENOITE
B
100
PANTELLERITE I

�� c.<
Olislolilh Slenesel- Alkaline
i

/'/-
'< c�
\PHONOLITE
. \
•
10 . . '
'

'............
0.1
- ... ... ..

ZrfTI02
0.1

--�
®
0,01
A NDESITE I BASALT <b,.�'o
•
100
Olistolith Kalvegjerdsneset Subalkaline MORB-Iike (KO)
� ., ,�
�
-- -,t- --
ALKALI·
BASALT
SUB-ALKALINE BASALT
10

0.001 +------+---1---i
O.Ql 0.1 10

Nb/Y

1000 ®
0.1

100
® ,/'"'
I
Olislolilh B otnanesesl- Subalkallne MORB-Iike (BO)
,
,
,
10 ,
I
I
I

\I
ppmCr 100

� • f,
1 .t.6 1: WPB \
' h I
I I\
'
I
I I\ • I
' IAT
0.1
m '
a: I I \ I
0 I
I I ', I
:::E
G) I \ :I
+-------��---'�-�--�---i
100
'
-" 10
()
0
a: 10 100
10 ppmY

10
®
----- ......
'
\
I
I
0.1

\
\
100 ® \
\
Olistolilhs of IAT metabasahs TI02(%) 1 I
'

0.1

10 100 1000

ppmZr

• 01. alkaline metabasalts

0 01. subalkaline/MORB-Iike
metabasalts (KO)
01. subalkaline/MORB-Iike
metabasalts (80)

0.1
o 01. transitional metabasalts
• 01. IAT metabasalts
O.Ql +--+---+--+--1----'<
Sr K Rb Be Nb Ce p Zr Ti y Cr
 Geochemical data for determining provenance 2S1

Table 4. Classification of metabasalts of mafic volcanic rock olistoliths according to the discriminant plots of Fig. 58

Discriminant plot A B c D E

Nb/Y- Zr/Ti02 Alkaline Subalkaline basalt Subalkaline andesite/ Subalkaline andesite/ Subalkaline andesite/
basalt basalt basalt
Y - Cr MORB/WPB MORBIWPB MORB/WPB MORB/WPB !AT
Zr-Ti02 (%) WPB (MORB) MORB/WPB MORB (!AT) All !AT (MORB)
Total WPB MORB? MORB? Transitional !AT

A, Large olistolith of alkaline metabasalts exposed at Sleneset; B, large olistolith of subalkaline MORB-Iike metabasalts
exposed at Kalvegjerdsneset (KO); C, large olistolith of subalkaline MORB-like metabasalts exposed at Botnaneset (BO);
D , smaller olistoliths of transitional metabasalts; E , smaller olistoliths of !AT metabasalts.

Table 5. Isotope data from volcanic-rock olistoliths

470 Ma 450Ma 430 Ma

Sample ENd Esr EN d Esr ENd Esr

AI kaline basalt 7.08 71.26 6.82 72.38 6.5 6 73.49

MORB-like (80) basalt 8.13 1 87 . 8 6 7.99 1 87 . 87 7.86 1 87 . 89
MORB-like (BO) basalt andesite 4.14 69.43 4.06 70. 65 3.98 7 1 . 88
Ignimbrite matrix 7.08 -50.81 6.9 6 -45.55 6.74 -40.3

BO, large olistolith of subalkaline MORB-like metabasalts exposed at Botnaneset.

character, more akin to om or CFB. The lack of and Range) basalts respectively (Faure, 1986; Wilson,
positive Nb anomalies and in som e cases slight 1989) . The lower ENd value of the latter may, how­
relative depletion ofNb (Fig. SA), and the relatively ever, indicate that the more felsic volcanics have
low values of Ti02, P205, Zr and Y jointly favour a experienced higher degrees of crustal contamination
CFB-type rather than an OIB (Thompson et al., 1984; compared with the basaltic types (Faure, 1986 ) .
Wilson, 1989 ) . Low values of Ti, Nb, Zr and Y Transitional basalts chiefly consist of scoriaceous
are generally attributed to crustal contamination and subordinate massive, aphyric or porphyritic aa
or reflect subduction-related magma genesis (Perfit metabasalts and basaltic meta-andesites. Rare
et al., 1980; Saunders et al., 1980; Cox & Hawkes­ phenocrysts are saussuritisized plagioclase and FeTi­
warth, 1984; Thompson etal. , 1984). Moreover, the oxides. Calcite- and minor quartz-filled amygdules
subalkaline MORB-like metabasalts also display geo­ are common, and locally small sandstone xenoliths
chemical characteristics similar to subduction-related are present. Similar to the alkaline and subalkaline
potassic and high-K calc-alkaline basalts (Foley MORB-like basalts, the transitional types also display
et al. , 1987; Wheller et al. , 1987; Stolz et al. , 1990). some geochemical signatures that are characteristic
The isotope data from a basaltic pahoehoe (ENct = 8) of alkaline OIB and/or MORB (Fig. S). However,
and an andesitic aa (ENd = 4) flow-unit are indicative the pronounced negative Nb anomalies (Fig. SA)
of MORB/back-arc basin basalts and oceanic island strongly argue for a subduction-related setting.
arc/some continental rift-related (northern Basin Although most geochemical evidence favours an

Fig. 5. (Opposite. ) (A) MORB normalized trace-element diagrams for volcanic rock olistoliths. See Fig. 2 for location and
stratigraphical position of different olistolith types of mafic volcanic rocks. (1) alkaline metabasalts, (2) subalkaline MORE­
like metabasalts of olistolith located at Kalvegjerdsneset (higher stratigraphical position compared to subalkaline MORB-like
metabasalts of olistolith located at Botnaneset), (3) subalkaline MORB-like metabasalts of olistolith located at Botnaneset,
(4) transitional type metabasalts, (5) !AT metabasalts, and ( 6) acidic volcanic rocks. The order of the elements and the
values of the normalizing constants are from Pearce ( 1983) , except for Cr, which is from Pearce (1980) . (B) Discriminant
plots for olistoliths of ( 1 ) volcanic rocks, and (2) and (3) mafic volcanic rocks. See Fig. 2 for location and stratigraphical
position of different olistolith types of mafic volcanic rocks. Discriminant plot (8 . 1 ) is from Winchester & Floyd ( 1977) ,
whereas discriminant plots (B .2) and (8.3) are from Pearce ( 1980).
2S2 R . Ravnas and H. Fumes

original back-arc setting, some element abundances As noted above, the mineralogical and geo­
and element ratios (e.g. high AI, low Ni, low FeO/ chemical similarities between the granitic and
MgO) are more characteristic of calc-alkaline basalts rhyolitic conglomerate debris and the olistoliths of
and island arc tholeiites. Basalts associated with acidic volcanic rocks suggest a common protolith.
incipient back-arc rifting or initial stages of back-arc The higher K, Rb, Zr, Nb, Y, and Th content of the
spreading are thus thought to be the best analogues rhyolitic rocks and the higher plagioclase content of
of these m etavolcanics (Bruhn et at., 1978; Saunders the granitic rocks may accordingly reflect fractional
etal., 1979; Weaver et at., 1979; Tarney eta/., 1982). crystallization, and particularly plagioclase fraction­
Island arc basalts (IAT) comprise scoriaceous, ation (Taylor et a/., 1968; Saunders et al., 1979).
porphyritic aa flows with phenocrysts of saussuritis­ Notably, the different volcanic-rock olistolith
ized plagioclase and minor FeTi-oxides. The MORB types are present at certain stratigraphical levels.
normalized, multi-element variation diagrams (Fig. Olistoliths of IAT lava-flow basalts and acidic volcanic
SA) show typical IAT character with negative Nb rocks are found at a lower stratigraphical level,
anomalies and relative depletion of Ti, Zr, Y and although generally above 'mega-olistoliths' of
Cr, thus testifying the subduction-related (i.e. island terrigenous sediments, whereas transitional and
arc or active continental margin) affinity of these subalkaline MORB-like metabasalts of increasingly
metabasalts (see also Fig. SB). alkaline affinity are found at successively higher
Acidic volcanic rocks comprise massive, por­ stratigraphical levels (Fig. 2). However, the olistolith
phyritic rhyolite and rhyolitic ignimbrite, with an consisting of alkaline metabasalt makes an important
inferred co-ignimbrite ash layer. Phenocrysts or exception. Whereas volcanic-rock olistoliths gener­
crystaloclasts are quartz (commonly blue), zoned, ally are found in the upper part of the melange
twinned and 'chessboard' sodic plagioclase, sanidine, succession, the olistolith of alkaline metabasalt
myrmekite, quartz/alkali-feldspar spherulites and occurs in the lower part of the melange succession in
inferred glass shards. In addition there are pumice an interval containing olistoliths of rhyolite, bedded
and lithic fragments present in the ignimbrite. Geo­ chert, pebbly mudstones and shallow-marine
chemical analyses of the ignimbrite matrix, rhyolitic sediments.
core of individual ignimbrite flows , ash-layer and Most major and trace elements of the mafic vol­
massive rhyolite, suggest that these acidic volcanic canic rock olistoliths seem to display 'gradational'
rocks are co-magmatic (Fig. S; Ravnas, 1991). variations from alkaline through subalkaline MORE­
The geochemical and isotope composition of these like to transitional types, and finally to !AT-type
silicic rocks favour a subduction-related original tec­ basalts, particularly demonstrated by their Fe, Mg,
tonic setting (Tarney et a/., 1977; Saunders et a/., K, Ti, Nb , Zr and Y content. All metabasalt types
1979; Pearce et a/., 1984; Faure, 1986; Wilson, have high AI and similar Rb, Ba and Sr contents
1989). The pronounced depletion of Nb, P and Ti (Fig. SA & Table 3). Moreover, all volcanic rock
(Fig. SA) indicate derivation from a depleted IAT olistoliths, including the alkaline metabasalts, show
source (Pearce et a/., 1984; Holm, 198S), whereas evidence of a primary subduction-related setting.
positive anomalies of Zr and Y are more typical of These points suggest that the protoliths of the mafic
silicic rocks from tholeiitic associations (Pearce et volcanic rock olistoliths may be genetically related
a/., 1984, and references therein). The isotope com­ and represent parts of a continuous volcanic series.
position, which is close to that of MORB (Table S), Below it is argued that the volcanic rock olistoliths
excludes magma generation by partial melting of represent parts of a volcanic succession in which the
continental crust and suggests little influence of alkaline and the IAT types represent the recorded
melted sediments in the magma source region end-members.
(Faure, 1986). The acidic volcanic rock olistoliths
are, accordingly, interpreted to represent island arc/
Olistoliths of sedimentary rocks
active continental margin volcanics erupted at a
considerable distance from the trench, i. e. behind Geochemical analyses have been performed on
the arc with respect to the trench. Alternatively, samples from all types of sedimentary-rock olisto­
these volcanics may have been derived from a slightly liths. However, only the lithic-volcanic arenites from
contaminated MORB source, representing oceanic the olistoliths of shallow-marine sediments and deep­
back-arc, rift-related acidic volcanics (Gill et a/., marine sandy turbidites will be discussed in some
1984). detail here.
 Geochemical data for determining provenance 253

The lithic-volcanic arenites consist of angular to provided by the framework grains and the gravel
subrounded quartz (blue or colourless), plagioclase population of the interbedded pebbly sandstones
(saussuritisized, zoned, twinned or 'chessboard'), and conglomerates. The type of framework grains
microline, sanidine, lithic-volcanic grains, lithic­ and the polymodal distribution of the gravel popu­
sedimentary grains, minor iron oxides, ilmenite, lation of most debris types strongly suggest that
calcite (locally dolomite), apatite, epidote and volcanic and sedimentary rocks now present as vol­
accessory sphene and zircon. The lithic-volcanic canic rock olistoliths and extrabasinal conglomerate
grains represent massive or foliated metabasalts, debris of the channel-fill deposits formed a m ajor
m eta-andesites, acidic volcanics and minor myrme­ source for these lithic-volcanic arenites. The geo­
kite. The lithic-sedimentary grains represent chert chemical composition of the lithic-volcanic arenites
or quartz schist and minor inferred intrabasinal sand­ in fact can be explained by mixing of detritus derived
stones and mudstones. Interbedded pebbly sand­ by erosion of such rocks alone (Fig. 7). Moreover,
stones and polymictic conglomerates show a similar the deviations in element and element ratio ranges
gravel admixture to the channel-fill conglomerates. compared with those suggested for basins of island
Bhatia ( 1 983) and Bhatia & Crook ( 1986) sug­ arc settings (Bhatia, 1983; Bhatia & Crook, 1986;
gested a simplified plate-tectonic classification of see Table 6) should be expected according to this
continental m argin and oceanic basins based on the interpretation (Fig. 7). Alternatively the apparent
nature of the crust from which the sediments were andesitic composition of the lithic-volcanic arenites
derived. Four types of plate tectonic setting were (Fig 7A), may suggest that andesitic igneous rocks
recognized: oceanic island arcs, continental island originally formed a larger volumetric proportion of
arcs, active continental margins and passive margins. the provenance region. Notably, there appears to be
The geochemical signatures of the lithic-volcanic no systematic stratigraphical variations in mineral­
arenites (Table 6) clearly suggest derivation from ogical and/or geochemical composition, suggesting
island arcs, either continental or oceanic (Fig. 6). that the different source rock types were all present
According to the classification schemes of Bhatia in the provenance region(s) throughout the life-span
( 1983) and Bhatia & Crook ( 1986) most m ajor of the depositional basin.
elements favour a continental ist'and arc as prov­
enance, whereas most trace elements and ratios
Groundmass deposits
argue in favour of an oceanic arc. The 'oceanic
island arc' setting includes sedimentary basins The m atrix of the olistostromal groundmass deposits
adjacent to intra-oceanic island arcs or island arcs is geochemically similar to the m udstones in the
partly formed on thin continental crust, whereas the olistoliths of shallow-marine deposits. The m atrix is,
'continental island arc' type includes sedimentary however, enriched in K, Fe, Mg, V, Cr, and to a
basins adjacent to island arcs formed on well­ lesser degree Zr and Nb, and depleted in Si, Ca and
developed continental crust or on thin continental Na compared with the lithic-volcanic arenites. These
margins (Bhatia, 1983 ) . A comprehensive data set effects are attributed to a larger volumetric pro­
on the geochemistry of detritus from an intra-oceanic portion of original m ud-sized p articles (clays and
arc was provided by Hiscott & Gill ( 1992) . The m icas) in the groundmass deposits (Roser & Korsch,
lithic-volcanic arenites of the sedimentary rock 1986; Argast & Donnelly, 1987). This supports the
olistoliths of the Kalvag Melange differ from the interpretation that the finer grained, more distal
volcaniclastic deposits discussed by Hiscott & Gill mem bers of the shallow-marine deposits provided
( 1 992) by their higher content of Nb, and, although most of the m aterial of the melange' s groundmass
less pronounced, also Zr, La and Th, which is here (Ravnas, 1 99 1 ) . The enrichment of fines in the olisto­
interpreted as the signature of the addition of stromal matrix can be ascribed to remoulding, during
alkaline and/or continental source rocks (see below). which most of the sandier deposits remained as
A m agmatic arc (Dickinson, 1985 ; Valloni, 1985) is clasts/olistoliths, or, perhaps more likely, to incor­
also suggested as the source for these arenites by the poration of shelf and slope mud into the ambient
modal analyses, although such analysis is in this case mud slurries or debris-flows from which the olisto­
hampered by post-depositional alteration effects, stromal 'groundmass' was deposited.
e.g. metamorphic recrystallization and subsequent The 'matrix' of the channel-fill conglomerates and
partial obliteration of primary grain boundaries. the gully- and channel-fill turb idites is, on the other
Further constraints on the provenance are hand, mineralogically and geochemically similar to
 N
Ul
.,.

Table 6. Representative geochemical analyses of sedimentary-rock olistoliths

Sample no. location Olistolith type Si02 Ti02 Al203 Fe203" MnO MgO CaO Na20 1<20 P205 LOI Total v Cr Co Ni Cu Zn Rb Sr y Zr Nb Ce Nd Ba La Th

63,41 0,84 14,92 6,99 0,09 3,81 4,16 2,90 1 ,84 0,11 1 ,73 1 00,81 109 144 30 34 36 91 87 144 35 242 13 89 45 361 37 10
R8-1 1 0 Annevika Shallow marine sed.
R8-124 Annevika Shallow marine sed. 65,97 0,87 1 1 ,93 6,19 0,09 4,62 6,09 3,13 1,13 0,12 0)1 1 00,84 1 04 154 25 35 10 79 52 166 28 203 11 49 46 458 24 3

R8-1 1 5 Annevika Shallow marine sed. 54,64 0,91 9,60 5,21 0,21 3,80 18,09 2,16 0,04 0,17 6,65 101 ,47 94 151 15 42 27 78 0 220 41 203 10 24 50 27 36 13

R8-1 1 9 Annevika Shallow marine sed. 64,55 0,88 13,56 6,58 0,08 4,83 5,17 2,83 1 ,53 0,12 0,82 1 00,96 105 135 27 39 23 90 65 211 29 228 13 56 48 305 29 19

R8-1 1 7 Annevika Shallow marine sed. 56,85 0,80 12,63 5,28 0,11 5,12 10,17 1 ,59 2,36 0,14 5,81 1 00,87 96 137 20 46 12 91 103 345 29 209 13 54 42 328 38 11 :;.;;,
R8-122 Annevika Shallow marine sed. 64,39 0,91 1 3,07 6,41 0,08 4,67 4,97 2,83 1 .88 0,13 0,80 100,14 1 04 143 26 36 22 85 81 202 32 282 14 70 39 373 39 6

R8-127 Annevika Shallow marine sed. 65,76 0,96 1 1 ,83 7,54 0,10 4,86 4,57 2,35 1 ,70 0,12 0,73 100,51 134 193 31 37 20 79 84 199 32 211 7 50 27 ,244 29 8 :;.;;,
Shallow marine sed. 67,43 0,89 12,07 6,55 0,09 4,03 3,92 3,44 1 , 34 0,13 0,75 100,64 107 153 26 35 25 78 67 200 29 262 12 84 44 259 44 10 .,
R8-129 Annevika
"<:
;:s
64,63 0,88 14,27 6,88 0,07 3,76 2,10 3,24 2,35 0,12 1 ,50 99,82 116 140 30 30 21 88 109 117 31 252 14 83 33 376 39 16 !':lo
R9-10 Varenesvika Shallow marine sed. "'
2,23 0,13 1 ,48 1 00,27 114 136 30 30 19 93 107 117 36 300 15 106 51 439 45 11
.,
R9-11 Varenesvika Shallow marine sed. 63,40 0,96 14,65 7,01 0,08 4,01 2,80 3,51
6,57 2,62 1 ,84 0,13 0,90 1 00,48 101 112 23 29 23 83 85 138 34 242 13 75 50 369 48 16
R8-213 Varenesvika Shallow marine sed. 64,42 0,86 12,95 6,16 0,09 3,93 ;:s
38
R8-21 1 Varenesvika Shallow marine sed. 63,71 0,94 1 5,26 6,98 0,08 3,88 2,25 3,03 2,42 0,15 1 ,64 1 00,33 117 139 31 34 20 95 118 111 296 17 101 50 439 41 15 "'-

�
0,08 3,67 2,17 2,46 1 ,63 0,11 1,60 1 00,23 117 145 29 29 22 81 79 108 31 252 14 86 46 214 35 8
R9-7 Varenesvika Shallow marine sed. 68,01 0,86 12,90 6,73
0,09 3,78 2,71 2,95 1 ,43 0,10 1 .46 1 00,48 112 143 27 33 33 83 70 127 27 184 8 80 43 255 41 9
R9-4 Varenesvika Shallow marine sed. 67,63 0,76 1 3,02 6,54

R8-31 Langeneset Deep marine turb. 68,45 0,86 1 1 ,13 7,95 0,13 3,85 4,15
4,75
4,75
2,82
0,50
0,82
0,10
0,12
0,97
0,88
1 00,51
99,44
132
147
185
201
31
34
35
39
18
22
71
75
20
40
179
135
25
28
126
137
7
5
42
38
30
20
123
168
12
15
13
12
�
...,
64,95 0,96 1 1 ,93 7,95 0,13 4,15
R8-32 Langeneset Deep marine turb.
97 125
;:s
2,93 2,11 0,13 0,81 99,57 113 145 28 35 19 76 27 170 9 40 31 301 12 16
R8-33
R8-34
langeneset
Langeneset
Deep marine turb.
Deep marine turb.
68,06
64,36
0,71
0,99
12,76
14,27
6,32
7,04
0,07
0,09
3,54
3,81
2,13
2,40 3,59 2,28 0,14 1 ,03 99,99 121 158 31 39 22 93 1 02 121 38 326 15 76 38 385 35 18 �
7,27 3,83 2,44 4,29 2,25 0,14 1 ,21 99,04 126 159 31 37 28 94 99 122 36 303 16 72 45 455 31 4
R9-61 langeneset Deep marine turb. 62,04 1,00 14,49 0,08
12,50 5,98 0,07 3,44 2,18 2,82 1 ,96 0,10 1 ,06 1 00,08 112 147 26 29 21 74 92 129 24 160 8 38 34 299 20 11
R9-62 Langeneset Deep marine turb. 69,27 0,69
1 1 ,79 6,18 0,08 3,33 2,66 2,69 1 , 54 0,11 0,92 1 00,68 111 150 26 33 18 74 80 127 24 215 9 70 48 283 28 7
R9-63 langeneset Deep marine turb. 70,57 0,80
6,04 0,09 3,10 2,18 5,10 1 ,46 0,12 1,31 101 ,90 107 156 25 31 28 77 62 110 30 298 12 72 51 274 37 7
R9-64 Langeneset Deep marine turb. 69,69 0,86 1 1 ,94
0,08 3,52 2,00 3,38 1 ,99 0,12 1 ,19 1 00,39 114 161 29 33 20 83 89 117 29 256 14 83 39 322 30
R9-65 Langeneset Deep marine turb. 67,37 0,86 13,31 6,57
0,10 3,69 3,46 2,91 0,99 0,11 0,86 1 00,60 118 169 27 34 21 71 50 157 28 1 92 11 72 47 289 31
R9-66 Langeneset Deep marine turb. 70,07 0,85 1 1 ,27 6,29

LOI, loss on ignition; F�O�', total iron as F�03 + 1 . 1 x FeO.

 (b) (c)
1 ,4 (a) 6 -.- I
. I
� . I
.,. ..... - -- ...0,
I
I
1,2 -
-- 0,3 s+ D
I
..
_
I

-
" /,. A /
A;n��. c::.". ' J
I
.... •' / I
..
.. ... / I

�·t
• r" I
,/ "
,' B ',, ....( - -...�
/" -.... .... /
"'ib",,- '
," . . .... ,' I I
I
••
,' D� a - ./
�·
I I
� 0,8 _ Do • •- - - - - N
Q 0,2 ,'-'\ ..... � I
I
r-- -
....
.... ... _ - - .-
- - ,'

-- -'- - --
,
I

/ c
.... ... -
�-:..
',
/ �oo,_-·.., � c;, I

: B
N'
0
,.,..
'\
oD � ,' C :- - ' o � 3 I -- -- --�
0 � 0 •
I
," \:!.
/ B : - -� --·------,
...
i= 0,6 I 0
N
) o oaD �7 \
I
< :
/,- ,_'- -'--- - - - - - -' "' .,.. / ,,
g2 !(
..,
('- o: _,.,.
-
',,
j, .. ..-
0,4 /D "/
"1
c ,
/
0,1 / /� ""
, , ..._ _
..
; < -''),'-•'/ \',� 9::-. oa 0" A
,�
C'l
/
/" D ./ o• ' "'
"
... ... ... ....... -•- - o- - - "
""
,
:_ _
"
0
' ..... _ _ /,"
"
,
.
" g..
0,2 ,_ ....
,
"'
;:!
0 0 +---�--�---+--��-� 0 +---�--�---+--� r;·
�
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
Fe203 MgO (%) + Fe203 MgO (%) + Fe203 MgO +
!}
s

(d) (e) C)>

60
...,
Th

»' �
50 � �
v
� s·
40 �·
"<:::
(3
E '<:
15: 30 "'
;::
-
.. , ....... ..,...
.....
c /: 1-
• Shallow marine sediments
.,
/./
;::

·r...
/
�
�
_, ... .. ""'!. ', ,'
20 __ o Deep marine turbidites
0. D
• B \'-'...
• '\
o · 0 \ "'�
o. .
',_ __
,
'
.-
10 - ...
\ /
_ "'
__
_
__
•
... :c• � D-��- -
( .....
A __ _
_ _ _ ... ...
..... 0
__ _
0 !"' I I I I I I ,'
0 5 10 15 20 25 30 Co Zr/ 1 0
Th ppm

Fig. 6. Discriminant plots of major and trace elements from lithic-volcanic arenites and wackes from sedimentary rock, olistoliths. The outlined fields are: A,
oceanic island arcs, B, continental island arcs; C, active continental margins, D , passive margins (Bhatia, 1983; Bhatia & Crook, 1986). The 'shallow-marine
sediments' and 'deep-marine turbidites' are lithic-volcanic arenites and wackes from olistoliths of shallow-marine deposits and deep-marine turbidites respectively. N
U\
Discriminant plots (a) to (d) are from Bhatia ( 1983), whereas (e) is from Bhatia & Crook ( 1986). U\
 256 R. Ravnas and H. Fumes

A
COMENDITE
B 10
PA NTELLERIT \
I

\ PHONOLITE - - - - - .......
' '
'
.... \
•
RHYOLITE
0 I
' , .._
.• • . 0 \
0.1 <• -- \
\
0 \
\

�--.\ . \
ZrfTi02 Ti02 (%) 1
WITHIN I
•
0
• \ PLATE
-/ ••
- - -r - - • • I \ LAVAS

· �����<0� -v""c·:·�.:�:
A NDESITE I BASALT Q}/" NEPHELINITE
'
ARC LAVAS I \ \
• \
- -%s-�-��slmg> A LKALI· 0
0 •• , .
': \

\
· /" \".
•• 1•: :r '\\
BASALT

·
lo I

....··
SUB·ALKA LINE BASALT BONINITES

0.001 0.1 �------�--+---+--;

0.01 0.1 10 10 100 1000

Nb/Y ppm Zr
-

c 1 000
• sediments

o mafic vole. ol.

-1
• acidic vole. ol.
D
o

J
400 mafic debris
•

acidic debris
350

300
0
ppm Cr 100 250
0
o
o
I\ ppm V 200 0 g
00 1: WPB \
• /1 I 150
IAT
I\ I
/ \ 0 "I 100
\\ I \ I
o
1 \ 1 50
' 'I • : I
10 +-----------------�·- --'--�
'----�'�� · --�
10 1 00 10 15 20

ppm Y ppm Ti/1000

Fig. 7. Discriminant plots of lithic-volcanic arenites from sedimentary rock olistoliths and suggested source rocks (i.e.
conglomerate debris and volcanic rock olistoliths). The filled squares labelled 'sediments' represent lithic-volcanic arenites
of the sedimentary-rock olistoliths, matrLx of the channel-fill conglomerates, gully- and channel-fill turbidites and the
olistostromal groundmass. Note how the lithic-volcanic arenites and the groundmass deposits in all discriminant diagrams
plot in an intermediate position between the suggested source rocks. Discriminant plot (A) is from Winchester & Floyd
( 1 977), (B) and (C) are from Pearce ( 1980), and (D) is from Shervais ( 1982). See text for further discussion.

the lithic-volcanic arenites contained in olistoliths PROVENANCE AND

of shallow-marine sediments and of deep-marine TECTONIC SETTING OF
turbidites. The channel- and gully-fill de posits are THE KALVAG MELANGE
accordingly interpreted to represent resedimented,
originally shallow-marine sands and grave ls. On the basis of geochemical d ata, most volcanic­
and sedimentary-rock olistoliths unequivocally
favour a subduction-related setting for their original
depositional environment. The great thickness of
 Geochemical data for determining provenance 257

the melange succession (> 2500 m) and its facies basalts (Fig. 2) , are interpreted to represent the
assemblage and their distribution, all suggest depo­ latest stage(s) of this inferred continuous magmatic
sition in the lower part or at the base of a fault­ evolution. Even though the single olistolith of
controlled, steep submarine slope that was subjected potassic, alkaline basalt is found at a lower strati­
to persistent mass-failure processes (Ravnas, 199 1 ) . graphical level, the alkaline volcanics are thought to
The type, abundance and geochemical signature of represent the oldest record of this assumed volcanic
olistoliths of volcanic rocks (basaltic, andesitic and rock series (Ravnas, 199 1 ) .
acidic) and first-cycle , marine, volcaniclastic rocks The magmatic evolution suggested b y the volcanic­
indicates that the resedimentation zone was on the rock olistoliths can best be explained by a gradual
submarine slope of a volcanic arc. development from continental intraplate, rift-related
The stratigraphical distribution of the various potassic magmatism, possibly associated with con­
types of olistolith lithology with the shallow-marine tinental flood-basalts, through incipient stages of
material in the lower part of the melange, and the rifting to initial back-arc spreading, and finally to
deep-marine turbidites, cherts and volcanics in the island arc calc-alkaline and tholeiitic magmatism.
upper part (Figs 2 & 3 ) , suggest that the melange The progressively increasing influence of subduction­
was formed by progressive 'unroofing' of the volcanic related processes in the magma source region during
terrane (Fig. S). This interpretation is further the final stages of the recorded magmatic evolution
supported by the vertical changes in the geochemical is demonstrated by the more pronounced negative
affinity of the volcanic rock olistoliths. The general Nb anomalies.
lack of pre-emplacement tectonic deformation of Magmatic development is therefore related to an
olistoliths argues against an origin of the melange area undergoing extension and rifting of continental
as multiple slides tapping a terrain of tectonically crust behind an active continental margin. The rifting
juxtaposed blocks. The occurrence of clasts and is thought to finally have separated a volcanic arc,
olistoliths of sedimentary and volcanic rocks within detached from the mainland by a back-arc basin or
the same debris-flow units suggests that the volcanic marginal sea (see discussion below) . A comparable
debris represents volcanic rocks that either were system would be the present Bransfield Strait, which
interbedded with the sedimentary deposits and/or is a narrow basin separating the South Shetland
constituted the substratum upon which the sediments Islands from the Antarctic Peninsula, and which is
were originally deposited. thought to have formed by recent back-arc extension
Progressively older deposits of the volcanic arc behind the South Shetland volcanic arc (Weaver
margin are inferred to have been uplifted by tectonic et al. , 1979; Garrett & Storey, 19S7 ) . The formation
faulting (thrusting) and resedimented during sub­ and evolution of the Middle Jurassic to Lower Cre­
sequent gravitational mass movement (Fig. SA). The taceous 'rocas verdes' of southern Chile (Bruhn
resulting 'olistolith-composition stratigraphy' in et al., 197S; Saunders et al., 1979) may constitute
the melange succession (Fig. 2) would thus be an another analogue. Shortly after the formation of the
inversion of the original stratigraphy of the volcanic/ back-arc basin, a thick pile of first-cycle, lithic­
clastic sedimentary apron of the volcanic arc. volcanic terrigenous clastic materials and possible
Accordingly, the olistoliths at the lowest strati­ intercalated volcanics were deposited in inferred
graphical level in the melange succession are thought fault-controlled basin(s) (Fig. SA; Ravnas, 199 1 ) .
to represent the resedimented youngest portion of The most favourable setting for the Kalvag
the volcanic-arc edifice, whereas those present at Melange is thought to be along the volcanic arc
successively higher stratigraphical levels would margin of the back-arc basin. The rhyolite and
represent progressively older portion of the arc rhyolitic ignimbrite olistoliths are interpreted to have
margin. been derived from acidic volcanics associated either
The apparently systematic geochemical variations with the incipient rifting to initial back-arc spreading
of the mafic volcanic-rock olistoliths from alkaline stages, or alternatively to reflect the inferred final
through subalkaline MORB-like to transitional types, island arc magmatism. Based on the assumption that
and finally to IAT-type basalts (Fig. 5 & Table 3) are the granitic pebbles and boulders represent the
interpreted to reflect (parts of) the gradual magmatic plutonic equivalents of these acidic volcanics, the
evolution of the melange's provenance region. The presence of such material as conglomerate debris
IAT-type basalts, occurring at a lower stratigraphical suggests that the volcanic arc itself was deeply
level than the transitional and subalkaline MORB-like eroded.
 258 R. Ravnas and H . Furnes

A B

Sea level

Retrogressive
slumping (7)
bottleneck
s l u m p scars

� Intraoceanic Active/
a rc-related magmatism /inactive

Sandy
� Active spreading MORB affinity

�
turbidites
Active spreading B o n i n itic affinity

� Obducted slices of 1. 2 and 3

Extensive

�
s l u m ping/sliding
of shallow· Subduction related magmatism
marine deposits

� Laurentia

� Baltica

Toquima - Table Head faunas

0
•
Mixed T·TH/Baltic faunas
Hirnantian faunas

Continued thrusting
and tectonic uplift
causes further
steepening of the slope
a n d renewed mass-failure.
Slidi ng/slumping of:
Deep-marine turbidites
Large Bedded chert
olistoliths Volcanic rocks
of shallow· Deposition of channe l-fill
marine conglomerates and
deposits gully·fill turbidites

Fig. 8. (A) Suggested depositional model for the Kalvag Melange and geological evolution of its provenance and tectonic
setting. 1, Deposition of the back-arc basin, volcanic arc margin along which the Kalvag Melange originated, 2, Initial
destabilization and gravitational resedimentation of the volcanic arc apron. 3-4, Deposition of the lower (3) and upper (4)
part of the melange. See text for further discussion. (B) Evolutionary model for the Lower Ordovician ophiolite complexes
and island arc sequences, and the Upper Ordovician to Lower Silurian rift-related sequences. 1 - 2 , Initiation of subduction
and development of immature island arc systems. 3, Build-up of mature island arcs that became colonized by shallow­
marine, Toquima- Table Head (Laurentian) faunas. 4, Accretion of the island arc system to the Laurentian continental
margin. 5-6, Rifting of the active continental margin and formation of a late Ordovician to early Silurian marginal basin
inhabited by Hirnantian and Holorhynchus (linked Laurentian and Baltic) faunas. From Pedersen et al. (1992).

Despite some slight variations in geochemical Figs 4A & 5B). However, metabasalts of boninitic
composition , the IAT- and MORB-type metabasaltic affinity have not been observed as olistoliths in the
debris, in addition to the calcareous material, may melange. Boninitic rocks within the Scandinavian
have been derived from intercalated mafic volcanics Caledonides have so far been reported only from the
of the proposed volcanic arc apron, representing Lower Ordovician ophiolite complexes/island arc
eroded IAT- and transitional-type basalts, respect­ sequences (e.g. see Pedersen et at., 1988; Pedersen
ively (compare multi-element variation diagrams of & Fumes, 199 1 ) , which thus provide the only likely
 Geochemical data for determining provenance 259

source for this debris type among the terranes preted to have formed part of the crust upon which
exposed along the orogen today. Moreover, ongoing the volcanic arc was founded, thus suggesting that
research has revealed the presence of large accumu­ the volcanic arc was built on a sliver of continental
lations of gabbroic rocks of boninitic affinity associ­ crust.
ated with one of these complexes (Fumes &
Pedersen, 1994) .
Characteristically, the Lower Ordovician ophiolite REGIONAL CORRELATIONS
complexes/island arc sequences show a long-lived AND INTERPRETATIONS
magmatic development (c. 500-470 Ma) , ranging
from typical MORB to IAT, boninites, calc-alkaline and The Scandinavian Caledonides, being dominated
alkaline metabasalts and differentiates (Pedersen by thrust-emplaced tectonic units, is made up of
et al., 1988; Pedersen & Fumes, 199 1 ) . It is note­ accreted suspect and exotic terranes (sensu Coney
worthy that the relatively flat MORB normalized et al., 1980), representing respectively the outer
pattern displayed by the MORB metabasaltic conglom­ part of the continent margin (suspect terranes) and
erate debris (Fig. 6a) is typical for the metabasalts 'oceanic' or island arc/ophiolite complexes, meta­
from the axis sequence of the Lower Ordovician morphic complexes and continental lithosphere con­
ophiolite complexes, thereby suggesting an alter­ sidered to have been derived from outboard of the
native source for this debris type . Based on mineral­ early Palaeozoic continent Baltica (exotic terranes)
ogical and geochemical data the trachyandesite/ (Stephens & Gee, 1985 , 1989; Roberts , 1988). The
dacite debris are also assumed to have been derived development of marginal basins and associated
from the Lower Ordovician island arc sequences. magmatism within the outboard terranes of the
The sedimentary cover to these ophiolite complexes, Scandinavian Caledonides are summarized in a
consisting of interbedded metamorphosed volcanics, number of papers (e.g. Pedersen & Fumes, 1991 ;
phyllites and chert, may thus represent a probable Pedersen et al., 1992). Two generations of ophiolite
source for the quartzite and possibly also the sub­ complexes and spreading related mafic magmatism
arkosic conglomerate debris (Ravnas, 199 1 ) . are now established within the Caledonian­
The back-arc basin within which the volcanic arc Appalachian orogenic belt; these are of early
apron originally formed received sediments from Ordovician and late Ordovician to early Silurian
two different source rock associations or provenance age, respectively (e.g. see Pedersen & Fumes, 199 1 ;
regions: the newly formed volcanic arc and an ?older Pedersen e t al., 1992, and references therein) .
ophiolite complex/island arc sequence. As no The Lower Ordovician ophiolite complexes
systematic vertical variations in mineralogical and/ and island arc sequences of the Scandinavian
or geochemical composition have been observed , Caledonides are thought to reflect the build-up of an
either in the sedimentary-rock olistoliths or in island-arc- marginal-basin system within the Early
the olistostromal 'groundmass', sediments were ap­ Palaeozoic Iapetus Ocean (Pedersen & Fumes, 199 1 ;
parently shed continuously from both source rock Pedersen e t al., 1992, Pedersen & Dunning, 1995).
associations into the original receiving basin(s). This Pedersen & Dunning ( 1995) suggest a continuous
suggests that sediments were derived either (i) con­ development from an immature island arc system
tinuously from two areally distinct provenances; comparable to the present Palau- Kyushu ridge of
(ii) from a volcanic arc formed by arc splitting of the the western Pacific into a mature island arc similar
Lower Ordovician ophiolite complexes/island arc to the present Sunda- Banda arc. The Upper
sequences; or (iii) from a deeply eroded volcanic Ordovician-Lower Silurian ophiolite and associated
arc separated from and formed by extension of an igneous complexes are related to rifting of an active
active continental margin onto which the Lower continental margin and the development of small,
Ordovician ophiolite complexes/island arc sequences fault-controlled marginal basins (Tucker et al., 1990;
had already been emplaced. Based on the isotope Pedersen & Fumes, 199 1 ; Pedersen et al., 199 1 ,
character of the volcanic-rock olistoliths combined 1992) . Of these Upper Ordovician t o Lower Silurian
with the present knowledge of the development of marginal basins, only that represented by the
marginal basins within the outboard terranes of the Solund- Stavfjord Ophiolite Complex developed
Scandinavian Caledonides, the latter is preferred into an oceanic basin (Fumes et al., 1 990) .
(see discussion below) . The accreted or obducted The relationship between the Lower Ordovician
ophiolite complexes/island arc sequences are inter- ophiolite complexes/island arc sequences and the
260 R . Ravnas and H. Fumes

Upper Ordovician to Lower Silurian magmatic Ordovician to Lower Silurian rift-related igneous
complexes is poorly known. However, a line of complexes and their cover, is further strengthened
evidence suggests that the Upper Ordovician to by the close spatial relationships with the Solund ­
Lower Silurian marginal basins formed by extension Stavfjord Ophiolite Complex and the suggested
along a continental margin onto which Lower tectono-stratigraphical position of the Kalv�lg
Ordovician ophiolite complexes/island arc sequences Melange (Fig. 1) (Fumes et al., 1990; Hartz et a!.,
had already been accreted or emplaced (Fig. 8B) 1 994) . On the basis of isotope data obtained from
(Pedersen & Fumes, 199 1 ; Pedersen et a!., 199 1 , the sedimentary cover (single zircon dates) , the
1992; and references therein). Based on faunal marginal basin represented by the Solund-Stavfjord
and isotopic data, the Lower Ordovician ophiolite Ophiolite Complex also received sediments from a
complexes/island arc sequences are interpreted to composite terrane, which included the Lower Ordo­
have been accreted onto the Laurentian margin of vician ophiolite complexes/island arc sequences
the Iapetus Ocean in the early to mid-Ordovician (Pedersen & Dunning, 1993 ) . Moreover, the inferred
(Bruton & Bockelie , 1980; Pedersen et a!., 1988, rift-related alkaline and subalkaline MORB-like vol­
199 1 , 1992; Pedersen & Fumes, 199 1 ) . canic rocks present as olistoliths in the melange are
A major problem in the reconstruction of the geochemically similar to metabasaltic dykes in the
origin and evolution of the Kalvag Melange is the Hs;;yvik Group of the middle tectonic unit on Atls;;y
lack of age constraints and rather poorly defined further south (Ravnas, 1991 ) . Although the origin
tectono-stratigraphical relationship to other early of these dykes is still uncertain (Fumes eta!., 1990) ,
Palaeozoic marginal basin complexes/sequences of they may possibly be related to the same rifting
the Scandinavian Caledonides. The geochemical event inferred from the volcanic-rock olistoliths of
characteristics of volcanic-rock olistoliths from the the Kalvag Melange above .
Kalvag Melange compare well with various types of The lower part of the melange is interpreted to
volcanic rocks from the Lower Ordovician ophiolite reflect the initial destabilization of the volcanic arc
complexes/island arc sequences reported by Fumes slope and gravitational failure of the sedimentary
et a! . ( 1986) and Sivertsen ( 1992) . These volcanics wedge, whereas the upper part reflects repeated
are now regarded as representing the late, evolved mass-failure processes that finally lead to resedimen­
stages of the build-up of a mature island arc tation of the deeper levels of the original arc apron
(Sivertsen, 1992; Pedersen & Dunning, 1995 ) . (Fig. SA) . The spatial distribution of facies assem­
However, the combined geochemical and isotope blages and olistolith types , together with the large
signature of the volcanic-rock olistoliths, and par­ thickness of the melange succession , thus suggest
ticularly that from the rhyolitic ignimbrite olistolith , major fault-related uplift of the volcanic arc margin.
do not compare favourably with any known Tectonic uplift and steepening of the basin margin
magmatic rocks from the Lower Ordovician com­ may be attributed to at least two mechanisms, which
plexes. Moreover, based on geochemical data and may not be mutually exclusive. The Andaman Sea
the relatively undeformed nature of the alkaline and has been advocated as a modern analogue for the
subalkaline MORB-like volcanic-rock olistoliths, these formation and evolution of the Solund- Stavfjord
olistolith types are not thought to have been derived Ophiolite Complex (Fumes et a!., 1990). Tectonic
from a hypothetical oceanic island or seamount that faulting may, accordingly, be related to transpression
shed material into the trench along the subduction due to large-scale strike-slip faulting along the island
zone (e.g. see Ogawa, 1985; Pautot et a!., 1987) arc, similar to the transcurrent movements along the
above which the Lower Ordovician ophiolite com­ present Andaman-Nicobar Ridge (Curray et a!.,
plexes/island arc sequences formed. 1979; Hla Maung, 1987) . Alternatively, the tectonic
The suggested setting and evolution of the back­ uplift may be attributed to thrusting during the
arc, depositional basin of the Kalvag Melange closing stages of the back-arc basin history. In the
resembles closely that proposed for the late Ordo­ latter case the Kalvag Melange may be coeval with
vician to early Silurian marginal basins, e.g. the the Sunnfjord Melange, which is interpreted to
Solund-Stavfjord Ophiolite Complex and the have been formed during obduction of the Solund ·­
Sulitjelma Gabbro (Fumes et al., 1990; Pedersen & Stavfjord Ophiolite Complex, i . e . during the closing
Fumes, 199 1 ; Pedersen et a!., 199 1 , 1992 ) . A cor­ stage of this Late Ordovician to Early Silurian
relation of the proposed volcanic arc apron along marginal basin (Berg, 1988; Andersen et a!., 1990;
which the Kalvag Melange originated with the Upper Alsaker & Fumes, 1994, Osmundsen & Andersen,
 Geochemical data for determining provenance 261

1994). In either case, continued uplift led to BATES, R.L. & JACKSON , J.A. ( 1987) Glossary of Geology.
oversteepening of the basin margin , gravitational American Geological Institute, Alexandria, Virginia,
788 pp.
instability and deposition of the Kalvag Melange as
BERG, T. ( 1988) Sedimentology of the Her/and Group
a series of successive gravitational resedimentation (Lower Silurian) on Atl¢y, Sunnfjord, western Norway.
events which tapped progressively deeper levels of Unpublished Cand. Scient. thesis, University of Bergen,
the volcanic arc apron. Final emplacement of the 217 pp.
BHATIA, M.R. (1983) Plate tectonics and geochemical com­
melange onto the continental margin probably
position of sandstones. J. Geol., 91, 6 1 1 -627.
occurred during the closure of the Iapetus Ocean as BHATIA, M.R. & CROOK, K.A.W. ( 1986) Trace element
a result of the continent- continent collision of the characteristics of graywackes and tectonic setting dis­
Laurentian and Baltic shields in the late Silurian crimination of sedimentary basins. Contrib. Mineral.
(Bryhni & Sturt, 1985; Ziegler, 1985) . Petrol. 92, 181- 193.
BHATIA, M.R. & TAYLOR, S.R. ( 1981) Trace-element
geochemistry and sedimentary provinces: a study from
the Tasman Geosyncline, Australia. Chern Geol., 33,
AC KNOWLEDGEME N T S 1 15 - 125.
B LAT!", H . , MIDDLETON, G.V. & MuRRAY, R. (1980) Origin

Financial support for this study has been provided of Sedimentary Rocks. Prentice-Hall, Englewood Cliffs,
NJ , 782 pp.
through grants from the Norwegian Research BREKKE, H. & SoLBERG, P.O. (1987) The geology of Atl!'ly,
Council for Science and Humanities (NAVF) , Sunnfjord, western Norway. Bull. Norg. geol. Unders.,
project code D .4 1 .3 1 . 147 , and the University of 410, 677-690.
Bergen. We are grateful to Drs W. Nemec and R . B . BRIGGS, R . M . , UrriNG , A.J. & GIBSON, I.L. (1990) The
origin of alkaline magmas in an intraplate setting near a
Pedersen for their encouragement, helpful criticism
subduction zone: the Ngatutura basalts, North Island,
and discussions during the course of this study. We New Zealand. J. Volcano/. Geotherm. Res., 40, 55-70.
thank Drs J. Winchester, R.M. Hiscott and A . G . BRUHN, R.L., STERN , C.R. & DE Wn, M.J. (1978) Field
Plint for thorough reviews o f a n earlier draft of this and geochemical data bearing on the development of a
manuscript, and J. Ellingsen for help in preparing Mesozoic volcano-tectonic rift zone and back-arc basin
in southernmost South America. Earth planet. Sci. Lett.,
the illustrations. This work represents publication 41, 32-46.
no. 155 in the International Lithosphere Project BRUTON, D . L. & BocKELIE, J . F . ( 1980) Geology and
(ILP) . paleontology of the H!'llonda area, western Norway - a
fragment of north America? In: Proceedings of the Inter­
national Geological Program - Caledonide Orogen
Project 27 (Ed. Wones, D . R . ) , pp. 41 -47. Department
REFERENCES of Geological Sciences, Virginia Polytechnic Institute
and State University, Blacksburg, V.A . , Memoir 2.
ALSAKER, E . & FURNES, H. (1994) The geochemistry of BRYHNI, I. & LYSE, K. (1985) The Kalvag Melange
the Sunnfjord Melange: sediment mixing from different Norwegian Caledonides. In: The Caledonide Orogen ­
sources during obduction of the Solund-Stavfjord Scandinavia and Related Areas (Eds Gee, D . G . & Sturt,
Ophiolite Complex, Norwegian Caledonides. Ceo/. B . A . ) , pp. 417-427. Wiley, Chichester.
Mag., 131, 105 - 12 1 . BRYHNI, I. & STURT, B.A. ( 1985) Caledonides of
ANDERSE N , T . B . & JAMTVEIT, B . (1990) Uplift o f deep southwestern Norway. In: The Caledonide Orogen ­
crust during orogenic extensional collapse: a model based Scandinavia and Related Areas (Eds Gee, D . G. & Sturt,
on field studies in Sogn-Sunnfjord region of Western B . A . ) , pp. 89- 109. Wiley, Chichester.
Norway. Tectonics, 9, 1097 - 1 1 1 1 . BRYHNI, I . , BOCKELIE, J . F . & N YSTU E N ,J . P . ( 1981)
ANDERSEN , T . B . , SKJERLIE, K . P . & FURNES, H. (1990) Excursions in the Scandinavian Caledonides. Uppsala
The Sunnfjord Melange, evidence of Silurian ophiolite Caledonide Symposium (UCS), 1 5 1 pp.
accretion in the West Norwegian Caledonides. J. geol. CAMERON, W.E., McCuLLOCH, M.T. & WALKER, D . A .
Soc. London, 147, 59-68. (1983) Boninite petrogenesis: chemical and Nd-Sr
ARGAST, S . & DONNELLY, T.W. (1987) The chemical isotopic constraints. Earth planet. Sci. Lett., 65, 75-89.
discrimination of clastic sedimentary components. CAN N , J.R. (1970) Rb, Sr, Y , Zr and Nb in some ocean
J. sediment. Petrol., 57 , 813-823. floor basaltic rocks. Earth planet. Sci. Lett., 10, 7 - 1 1 .
BASALTIC VOLCANISM STUDY PROJECT ( 1981) Basaltic CAs, R.A.F. & WRJGHT, J.V. (1988) Volcanic Successions,
Volcanism on the Terrestrial Planets. Pergamon Press, Allen & Unwin, London, 528 pp .
New York, 1286 pp. CoiS H , R.A. (1977) Ocean floor metamorphism in the
BASU, A. ( 1985) Influence of climate and relief on Betts Cove Ophiolite, Newfoundland. Contrib. Mineral.
compositions of sands released at source areas. In: Petrol., 60, 277-302.
Provenance of Arenites (Ed. Zuffa, G . G . ) , pp. 1 - 18. CoNEY, P . J . , JoNES, D . L. & MoNGER, J.W.H. (1980)
NATO ASI Series, Series C : Mathematical and Physical Cordilleran suspect terranes. Nature, 288, 329-333.
Sciences, Vol. 148. Reidel, Dordrecht. Cox, K.G. & HAwKESWORTH , C.J (1984) Relative contri-
262 R . Ravnas and H . Fumes

butions of crust and mantle to flood basalt magmatism, Gasl!ly Intrusion, West Norway. Norsk geol. Tidsskr. ,
Mahabaleshwar area, Deccan Traps. Philos. Trans. R. 69, 209-224.
Soc. London, Ser. A, 310, 627-64 1 . FURNES, H . , SKJERLIE, K . P . , PEDERSE N , R.B . , ANDERSEN,
CRAWFORD, A . J . (1989) Boninites and Related Rocks. T.B . , STILLMAN, C.J . , SuTHREN , R.J . , eta/. (1990) The
Unwin Hyman Ltd, London, 465 pp. Solund-Stavfjord Ophiolite Complex and associated
CROOK, K.A.W. ( 1974) Lithogenesis and geotectonics, the rocks, west Norwegian Caledonides: geology, geo­
significance of compositional variations in flysch arenites chemistry and tectonic environment. Geol. Mag. , 127,
(graywackes) In: Modern and Ancient Geosynclinal 209-224.
Sedimentation (Eds Dott, R.H. & Shaver, R.H.) Spec. GARRETI, S .W. & STOREY, B.C. ( 1987) Lithospheric exten­
Pub!. Soc. econ. Paleontol. Mineral. , Tulsa, 19, sion on the Antarctic Peninsula during Cenozoic sub­
304-310. duction. In: Continental Extensional Tectonics (Eds
CURRAY, J . R . , MOORE, D . G . , LAWVER, L . A . , EMMEL, F . J . , Coward, M . P . , Dewey, J.F. & Hancock, P.L.), Spec.
RAITI, R . W . , HENRY, M . , e t al. ( 1 979) Tectonics o f the Pub!. Geol. Soc. London , 28, 419-43 1 . Geological
Andaman Sea and Burma. In: Geological and Geo­ Society of London, Bath.
physical Investigations of Continental Margins (Eds GILL, J . B . , STORK, A.L. & WHELA N , P.M. (1984) Vol­
Watkins, J . , Montadert, L. & Dickinson, P . ) , Mem. canism accompanying back-arc development in the
Am. Assoc. petrol, Geol . , Tulsa, 29, 189- 198. southwest Pacific. Tectonophysics, 102, 207-224.
DICKINSON , W . R. (1985) Interpreting provenance relations GoviNDARAJU , K. (1984) 1984 compilation of working
from detrital modes of sandstones. In: Provenance of values and sample description for 170 international
Arenites (Ed. Zuffa, G . G . ) , pp. 333-361 . NATO ASI reference samples of mainly silicate rocks and minerals.
Series, Series C: Mathematical and Physical Sciences Geostand. News/. , 8, 1 -9 1 .
Vol. 148. Reidel, Dordrecht. HARTZ, E . , ANDRESEN, A. & ANDERSEN, T . B . ( 1994)
DICKJNSON , W.R. & SuczEK, C.A. (1979) Plate tectonics Structural observations adjacent to a large scale
and sandstone compositions. Bull. Am. Assoc. petrol. extensional detachment zone in the hinterland of the
Geol. , 63, 2-3 1 . Norwegian Caledonides. Tectonophysics, 231 , 123- 127.
DICKJNSON , W.R. & VALLONI , R . ( 1980) Plate settings and HAuGHTON , P.D.W. & HALLIDAY, A.N. (1991) Signifi­
provenance of sands in modern ocean basins. Geology, cance of a late Caledonian igneous complex revealed
8, 82-86. by clasts in Lower Old Red Sandstone conglomerates,
DoRSEY, R.J. (1988) Provenance evolution and unroofing central Scotland. Geol. Soc. Am. Bull. , 103, 1476- 1492.
history of a modern arc-continent collision: evidence HELMOLD, K.P. ( 1985) Provenance of feldspathic
from petrography of Plio-Pleistocene sandstones, sandstones - the effect of diagenesis on provenance
Eastern Taiwan . J. sediment. Petrol. , 58, 208-218. interpretations. In: Provenance of Arenites (Ed. Zuffa,
D u N N I N G , G.R. & PEDERSEN , R.B. ( 1988) U/Pb ages of G . G . ) , pp. 139- 163. NATO ASI Series, Series C :
ophiolites and arc-related plutons of the Norwegian Mathematical and Physical Sciences Vol. 148. Reidel,
Caledonides: implications for the development of Dordrecht.
Iapetus. Contrib. Mineral. Petrol., 98, 13-23. HlcKEY, R.L. & FREY, F.A.F. (1982) Geochemical
EDGAR, A.D. (1980) Role of subduction in the genesis of characteristics of boninite series volcanics: implications
leucite-bearing rocks: discussion. Contrib. Mineral. for their source. Geochim. Cosmochim. Acta , 46,
Petrol. , 73, 429-43 1 . 2099-2115.
EWART, A. (1982) The mineralogy and petrology of H1scorr. R.N. ( 1984) Ophiolitic source rocks for Taconic­
Tertiary- Recent orogenic volcanic rocks with special age flysch: trace-element evidence. Geol Soc. Am. Bull.,
reference to the andesite-basaltic compositional range. 95, 126 1 - 1267.
In: Andesites: Orogenic Andesites and Related Rocks Hlscorr, R.N. & GILL, J.B. (1992) Major and trace
(Ed. Thorpe, R.S.), pp. 26-87. Wiley, Chichester, element geochemistry of Oligocene to Quaternary
697 pp. volcaniclastic sands and sandstones from the Izu- Bonin
FAURE, G. ( 1986) Principles of Isotope Geology. Wiley, arc. In: Proceedings of the Ocean Drilling Program,
New York, 589 pp. Scientific Results, Vol. 126. (Eds Taylor, B . , Fujioka, K . ,
FoLEY, S . F . , VENTURLLI, G . , GREEN , D . H . & TosCANI , L. et a/.), pp. 467-485 . Ocean Drilling Program, College
(1987) The ultrapotassic rocks: characteristics, classifi­ Station. TX.
cation, and constraints for petrogenic models. Earth Sci. HLA MAUNG (1987) Transcurrent movements in the
Rev. , 24, 8 1 - 134. Burma-Andaman Sea region. Geology, 15, 9 1 1 -912.
FURNES, H. & PEDERSEN , R.B. (1994) The Lyngen HoLM, P.E. (1985) The geochemical fingerprints of differ­
Magmatic Complex: geology and geochemistry. ent tectonomagmatic environments using hygromag;­
Abstracts 21 :a Nordiska Geologiska Vintermdtet, Lulea matophile element abundances of tholeiitic basalts and
1994, p. 58. basaltic andesites. Chemical Geology, 5 1 , 303-323.
FURNES, H . , BREKKE, H . , NORDAS, J. & HERTOGE N , J. HoSSACK , 1 .R. (1984) The geometry of listric faults in the
( 1986) Lower Palaeozoic convergent plate margin vol­ Devonian basins of Sunnfjord, W. Norway. J. Geol.
canism on Bl!lmlo, southwest Norwegian Caledonides: Soc. London, 141 , 629-637.
geochemistry and petrogenesis. Geol. Mag., 123, INGERSOLL, V.R. & SucZEK, C.A. (1979) Petrology and
123 - 142. provenance of Neogene sand from Nicobar and Bengal
FURNES, H . , PEDERSEN , R.B . , SUNDVOLL, B . , TYSSELAN D, fans, DSDP sites 2 1 1 and 18. J. sediment. Petrol. , 49,
M. & TuMYR, 0. (1989) The age, petrography, geo­ 1217- 1228.
chemistry and tectonic setting of the late Caledonian JoHNSSON, M.J. & STALLARD, R.F. (1989) Physiographic
 Geochemical data for determining provenance 263

controls on the composition of sediments derived from Leg 2. Earth planet. Sci. Lett. , 83 , 300-312.
volcanic and sedimentary terrains on Barro Colorado PEARCE, J.A. (1980) Geochemical evidence for the genesis
Island, Panama. J. sediment. Petrol. , 59, 768-78 1 . and eruptive setting of lavas from Tethyan ophiolites.
KoLDERUP, N.H. ( 1921) Der Mangeritsyenit und In: Ophiolites (Ed. Panayitou, A . ) pp. 261 - 272. Geo­
umgebende Gesteine zwischen Dalsfjord in Spndfjord logical Survey Department, Cyprus, 781 pp.
im westlischen Norwegen. Bergen Museum A. rbok , PEARCE, J.A. (1983) The role of sub-continental lithosphere
1920-21 (5) 71pp. in magma genesis at destructive plate boundaries.
KoLDERUP, N.H. (1928) Fjellbygningen i kyststrpket mellem In: Continental Basalts and Mantle Xenoliths (Eds
Nordfjord og Sognefjord. Bergen Museum A. rbok, Hawkesworth, C.J. & Norry, M.J.), pp. 230-249. Shiva
Naturvitenskapelig Rcekke, I , 222 pp. Geology Series, Nantwich.
LARU E , D . K . & SAMPAYO, M . M . ( 1990) Lithic-volcanic PEARCE, J . A . , HARRIS, N.B.W. & TINDLE, A.G. (1984)
sandstones derived from oceanic crust in the Franciscan Trace element discrimination diagrams for the tectonic
Complex of California: 'sedimental memories' of source interpretation of granitic rocks. J. Petrol. , 25, 956-983.
rock geochemistry. Sedimentology, 37 , 879-889. PEDERSE N , R . B . & DUNNING, G.R. (1993) Provenance of
MARSAGLIA, K . M . , INGERSOLL, R.V. & PACKER, B .M . turbiditic cover to the Caledonian Solund-Stavfjord
(1992) Tectonic evolution o f the Japanese islands as ophiolite from U - Pb single zircon dating. J. geol. Soc.
reflected in modal compositions of Cenozoic forearc and London, 150, 673-676.
backarc sand and sandstone. Tectonics, 1 1 , 1028 - 1044. PEDERSE N . R.B. & DUNNING, G.R. (1995) Age relations
McBRIDE, E .F. (1985) Diagenetic processes that affect between ophiolite and island arc complexes: new U/Pb
provenance determinations in sandstone. In: Provenance ages from the ophiolitic terrane of SW Norway. Contrib.
of Arenites (Ed. Zuffa, G . G . ) , pp. 95- 1 13 . NATO ASI Mineral. Petrology, (in press) .
Series, Series C: Mathematical and Physical Sciences PEDERSE N , R . B . & FURNES, H. ( 1991) Geology, magmatic
Vol. 148. Reidel, Dordrecht. affinity and geotectonic environment of some Caledonian
McLENNAN , S . M . , TAYLOR, S . R . , McCuLLOCH, M.T. & ophiolites in Norway. J. Geodyn. , 1 3, 183-203.
MAYNARD, J . B . ( 1990) Geochemical and Nd- Sr isotopic PEDERSEN , R . B . , FURNES, H . & DUNNING, G.R. (1988)
composition of deep-sea turbidites: crustal evolution and Some Norwegian ophiolite complexes reconsidered.
plate tectonic associations. Geochim Cosmochim. Acta , Norges geol. Unders. Spec. Pub/. , 3, 80-85.
54, 2015-2050. PEDERSE N , R.B . , FURNES, H. & DUNNING, G.R. (1991) A
MoRRIS, J . D . & HART, S . R. ( 1983) Isotopic and incom­ U/Pb age for the Sulitjelma Gabbro, North Norway:
patible element constraints on the genesis of island arc further evidence for the development of a Caledonian
volcanics from Cold Bay and Amak Island, Aleutians, marginal basin in Ashgill-Llandovery time. Geol. Mag. ,
and implications for mantle structure. Geochim. 128, 141- 153.
Cosmochim. Acta , 47, 2015-2033. PEDERSE N , R.B . , BRUTON , D.L. & FURNES, H . ( 1992)
MuTn, E . (1985) Turbidite systems and their relations to Ordovician ophiolites, island arcs and faunas in the
depositional sequences. In: Provenance of Arenites (Ed . Scandinavian Caledonides. Terra Nova , 4, 217-222.
Zuffa, G . G . ) , pp. 65 -93. NATO ASI Series, Series C : PERFIT, M . R . , GusT, D . A . , BENCE, A . E . , ARcuLus, R.J.
Mathematical and Physical Sciences Vol. 148. Reidel , & TAYLOR, S.R. (1980) Chemical characteristics of island
Dordrecht. arc basalts: implications for mantle sources. Chern.
NORMARK, W.R. (1985) Local morphologic controls and Geol. , 30, 256-277.
effects of basin geometry on flow processes in deep PETTIJOHN, F.J . , POTTER, P.E. & SIEVER, R. (1987)
marine basins. In: Provenance of Arenites (Ed. Zuffa, Sedimentary Rocks. Springer-Verlag, New York,
G.G.), pp. 47-64. NATO ASI Series, Series C: 553 pp.
Mathematical and Physical Sciences Vol. 148. Reidel, RAvNAS, R. ( 1991) A sedimentological, geochemical and
Dordrecht. provenance study of the Kalvag Melange, western
NORRISH, K. & HUTTO N , T.T. (1969) An accurate X-ray Norwegian Caledonides. Unpublished Can d . Scient.
spectroscopic method for the analysis of a wide range of thesis, University of Bergen, 2 12 pp.
geological samples. Geochim. Cosmochim. Acta, 33, RAYMOND, L.A. ( 1984) Classification of melanges. In:
431 -453. Melanges: their Nature, Origin and Significance
NORTON, M.G. (1986) Late Caledonian extension (Ed . Raymond , L.A.), Geol. Soc. Am. Spec. Paper,
in western Norway: a response to extreme crustal Boulders, 198, pp. 7-20.
thickening. Tectonics, 5, 195-204. REUSCH, H . ( 1903) Forsteninger i fjeldet paa Frpyen.
OGAWA, Y. (1985) Variety of subduction and accretion Naturen, 7 , 160.
processes in Cretaceous to Recent plate boundaries Rlcc1 LuccHI, F. (1985) Influence of transport processes
around southwest and central Japan. Tectonophysics, and basin geometry on sand composition . In: Provenance
112, 493 -518. of Arenites (Ed. Zuffa, G . G . ) , pp. 19-45. NATO ASI
0SMUNDSEN , P.T. & ANDERSEN, T.B. ( 1994) Caledonian Series, Series C: Mathematical and Physical Sciences
compressional and late-orogenic extensional defor­ Vol. 148, Reidel, Dordrecht.
mation in the Staveneset area, Sunnfjord, Western RoBERTS, D . (1988) The terrane concept and the
Norway. J. struct. Geol. , 16, 1385- 1401 . Scandinavian Caledonides: a synthesis. Bull. Norges
PAUTOT, G . , NAKAMURA, K . , HUCHON, P . , ANGELlER, J . , geol. Unders. , 41 3 , 93-99.
BouRGOIS , J . , FuJIOKA, K . , et al. (1987) Deep-sea ROBERTSON, A.H.F. & HEN DERSON , W.G. ( 1984) Geo­
submersible survey in the Suruga, Sagami and Japan chemical evidence for the origins of igneous and sedi­
trenches: preliminary results of the 1985 Kaiko cruise, mentary rocks of the Highland Border, Scotland. Trans.
264 R. Ravnas and H. Fumes

R. Soc. Edinburgh: Earth Sci. , 75, 135- 150. Deep. Sea Trenches and Back-arc Basins (Eds Talwani,
ROSER, B.P. & KORSCH, R.J. ( 1986) Determination of M. & Pitman, W.C., III), pp. 367-378. Maurice
tectonic setting of sandstone- mudstone suites using Si02 Ewing Series, Vol. 1 , American Geophysical Union,
content and K20/Na20 ratio. J. Geol. , 94, 635-650. Washington, DC.
SAUNDERS, A.D. & TARNEY, J. (1979) The geochemistry of TARNEY, J . , WEAVER, S . D . , SAUNDERS, A . D . , PANKHURST,
basalts from a back-arc spreading center in the East R.J. & BARKER, P.F. ( 1982) Volcanic evolution of the
Scotia Sea. Geochim. Cosmochim. Acta, 43, 555-572. northern Antarctic Peninsula and Scotia arc. In:
SAUNDERS, A . D . , TARNEY, J . , STERN , C.R. & DALZIEL, Andesites: Orogenic Andesites and Related Rocks
I.W . D . (1979) Geochemistry of Mesozoic marginal basin (Ed. Thorpe, R.S.), pp. 371-402. Wiley, Chichester.
floor igneous rocks from southern Chile. Geol. Soc. Am. TAYLOR, S . R . , EwART, A. & CAPP, A.C. (1968) Leuco­
Bull. , 90, 237-258. granites and rhyoHtes: trace element evidence for frac­
SAUNDERS, A.D . , TARNEY, J. & WEAVER, S . D . ( 1980) tional crystallization and partial melting. Lithos, 1 ,
Transverse geochemical variations across the Antarctic 179 - 186.
Peninsula: implications for the genesis of calc-alkaline THOMPSON, R.N. (1977) Primary basalts and magma
magmas. Earth planet. Sci. Lett. , 46, 344-360. genesis. III. Alban Hills, Roman Comagmatic Province,
ScHWAB, F.L. (1975) Framework mineralogy and chemical Central Italy. Contrib. Mineral. Petrol. , 60, 9 1 - 108.
composition of continental-margin type sandstone. THOMPSON, R.N. & FOWLER, M . B . (1986) Subduction­
Geology, 3, 487-490. related shoshonitic and ultrapotassic magmatism : a
SERANNE, M. & SEGURET, M. ( 1987) The Devonian basins study of Siluro-Ordovician syenites from the Scottish
of western Norway: tectonics and kinematics of an Caledonides. Contrib. Mineral. Petrol., 94, 507-522.
extending crust. In: Continental Extensional Tectonics THOMPSON , R . N . , MORRISON , M . A . , HENDRY, G.L. &
(Eds Coward, M . P . , Dewey J.F. & Hancock, P.L.) PARRY, S.J. ( 1984) An assessment of the relative roles
Spec. Publ. Geol. Soc. London, 28, pp. 537-548. Geo­ of crust and mantle in magma genesis: an elemental
logical Society of London, Bath. approach. Philos. Trans. R. Soc. London, Series A, 297 ,
SHERVA!S , J.W. (1982) Ti-V plots and the petrogenesis of 409-455.
modern and ophiolitic lava. Earth Planet. Sci. Lett. , 59, THORPE, R . S . , FRANCIS, P.W. & O'CALLAGHAN, L. (1984)
101 - 1 18. Relative role of source composition, fractional crystal­
SIVERTSEN , J.-E. ( 1992) Stratigraphy and geochemistry of lization and crustal contamination in the petrogenesis of
the extrusives of the Torvastad Group: marginal basin Andean volcanic rocks. Philos. Trans. R. Soc., London,
deposits associated with the Karm¢y Ophiolite Complex, Ser. A, 310, 675-692.
SW Norway. Unpublished Cand. Scient. thesis, Univer­ TucKER, R . D . , BoYD, R. & BARNES, S.-J. ( 1990) A U - Pb
sity of Bergen, 1 1 8 pp. age for the Rana intrusion, N. Norway: new evidence of
SKAR, 0 . , Fu R N ES, H. & CLAESSON, S . (1994) Middle basic magmatism in the Scandinavian Caledonides in
Proterozoic orogenic magmatism within the Western Early Silurian time. Norsk geol. Tidsskr. , 70, 229-239.
Gneiss Region, Sunnfjord, Norway. Norsk. geol. VALLONl, R. (1985) Reading provenance from modern
Tidjskrift. , 74, (in press) . marine sands. In: Provenance of Arenites (Ed. Zuffa,
SMITH, G.A. , CAMPBELL, N . P . , DEACON , M.W. & G . G . ) , pp. 309-332. NATO ASI Series, Series C :
SHAF!QULLAH, M. (1988) Eruptive style and location of Mathematical and Physical Sciences Vol. 148. Reidel ,
volcanic centers in the Miocene Washington Cascade Dordrecht.
Range: Reconstructions from the sedimentary record . VALLON!, R. & MEZZARDl, G . ( 1984) Compositional suites
Geology, 16, 337-340. of terrigenous deep-sea sands of the present continental
STAUD!GEL, H. & HART, S.R. (1983) Alteration of basaltic margins. Sedimentology, 31, 353-364.
glass: mechanism and significance for the oceanic crust­ WEAVER, S . D . , SAUNDERS, A . D . , PANKHURST, R.J. &
seawater budget. Geochim. Cosmochim. Acta, 47, TARNEY, J. (1979) A geochemical study of magmatism
37-50. associated with the initial stages of back-arc spreading.
STEPHENS, M . B . & GEE, D . G . (1985) A tectonic model for Contrib. Mineral. Petrol., 68, 1 5 1 - 169.
the evolution of the eugeoclinal terranes in the central WHELLER, G . E . , VARN E, R . , FODEN , J . D . & ABBOTT, M.J.
Scandinavian Caledonides. In: The Caledonide Orogen ­ (1987) Geochemistry of Quaternary volcanism in the
Scandinavia and Related Areas (Eds Gee, D . G . & Sturt, Sunda-Banda arc, Indonesia, and three-component:
B . A . ) , pp. 953 -978. Wiley, Chichester. genesis of island-arc basaltic magmas. In: Tectonic
STEPHENS, M . B . & GEE, D . G . (1989) Terranes and Controls on Magma Chemistry. (Eds Weaver, S . D . &
polyphase accretionary history in the Scandinavian Johnson, R.W.) J. volcano/. Geotherm. Res. , 32, 137-
Caledonides. In: Terranes in the Circum-Atlantic 160.
Paleozoic Orogens (Ed. Dallmeyer, R.D.). Geol. Soc. WILSON, M. (1989) Igneous Petrogenesis: a Global Tectonic
Am . Spec. Pap . , Boulder, 230, 17-30. Approach . Unwin Hyman, London, 466 pp.
STOLZ, A . J . , VARN E, R . , DAVIES, G . R . , WHELLER, G.E. & WINCHESTER, J.A. & FLOYD , P.A. ( 1977) Geochemical
FoDEN , J . D . (1990) Magma source components in an discrimination of different magma series and their dif­
arc-continent collision zone: the Flores-Lembata sector, ferentiation products using immobile elements. Chern.
Sunda arc, Indonesia. Contrib. Mineral. Petrol. , 105, Geol., 20, 325-343.
585-60 1 . ZIEGLER, P.A. (1985) Late Caledonian framework
TARNEY, J . , SAUNDERS, A.D. & WEAVER, S . D . (1977) of western and central Europe . In: The Caledonide
Geochemistry of volcanic rocks from the island arcs and Orogen - Scandinavia and Related Areas (Ed. Gee,
marginal basins of Scotia Sea region . In: Island Arcs, D . G . & Sturt, B . A . ) , pp. 3 - 19. Wiley, Chichester.
Spec. Pubis int. Ass. Sediment. (1995) 22, 265-281

Differential subsidence and preservation potential of

shallow-water Tertiary sequences, northern Gulf Coast Basin, USA

MARC B . EDWARDS
5430 i s, H o uston, Texas77096-4020, USA
D umfr e

ABSTRACT

Growth faulting, which accompanied shelf-edge progradation and filling of the northern Gulf Coast
Basin, resulted in partitioning of the basin margin into fault blocks with contrasting subsidence rates.
Study of correlative sections in juxtaposed fault blocks reveals that contrasting subsidence rates can
result in strongly differing facies patterns in neighbouring areas. This complicates the task of predicting
sandstone reservoir occurrence and properties. Shallow-water clastic sections from Eocene to Miocene
in age were investigated using extensive well-log observations, supplemented with micropalaeontology
and seismic profiles.
All depositional environments involve an ongoing, complex interplay between sedimentation and
erosion at different time- and physical scales. In certain settings, a greater subsidence rate causes the
preservation of certain facies that would otherwise have been eroded at lower subsidence rates by
processes inherent to the environment. The critical subsidence rate that separates preservation from
non-preservation is termed the preservation potential threshold for a particular depositional facies.
Examples are provided for progradational mouth-bar facies in a deltaic setting (Wilcox), and storm­
deposited shoreface-shelf muds in a prograding shoreline setting (Frio).
Where rates of subsidence are even greater, the growth fault may produce a topographical scarp at
the surface, which will influence the disposition of depositional environments. Here, the concept of
preservation thresholds is not adequate to account for the observed facies changes. Rather, the presence
of the surficial scarp as the surface manifestation of the subsurface fault causes the preferential
development and preservation of channel activity in the topographic lows, and progradational environ­
ments with channel bypass in the topographical highs . An example is provided for a series of prograding
stacked deltas (Miocene).
These concepts may help to focus attention on the role of subsidence in constraining the appearance
of the sedimentary record .

INTRODUCTION

In growth-faulted regions, such as the northern Gulf Exploring in such regions, It IS common to drill
Coast Basin of Texas and Louisiana, contempor­ into fault blocks in which strata of a particular age
aneous faulting structurally offset stratigraphical have not been penetrated previously. Typically, little
surfaces shortly after their formation. Through time, is known about how the downthrown section will
the sediment on the upthrown block (footwall) sub­ differ from the comparatively well-known equivalent
sided at a lower rate than sediment on the down­ section on the upthrown block. This paper describes
thrown block (hangingwall). This setting provides several examples of characteristic changes that take
the opportunity to study the relationship between place across growth faults , and attempts to explain
the preserved stratigraphical record and changing the observed changes.
subsidence rates, while other important variables, The most obvious manifestation of the effect of
such as eustatic sea-level , sediment supply and growth faulting on sedimentation is an increase in
depositional environment, remain comparatively thickness of a genetic unit from the upthrown to the
unchanged. downthrown blocks. Expansion ratios (downthrown

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0

265
266 M.B. Edwards

thickness divided by upthrown thickness; Thorsen, unchanged. Hence it is possible to identify those
1963) as great as 10 have been documented. I n changes that most likely relate to subsidence rate.
addition t o thickness, other properties that vary Sequence stratigraphy has heightened interest in
across the faults include net sandstone, percentage the temporal significance of certain types of strati-·
sandstone and log facies development. The way in graphical surfaces. Peculiarly, emphasis has been
which these properties change across growth faults given to surfaces that are admittedly not isochronous
will be described in the examples below. (e .g. sequence boundaries; see Posamentier &
The interpretation of three stratigraphical units, Weimer, 1993) as opposed to surfaces that are (e .g.
the Palaeocene -Eocene Wilcox Group, the Oligo­ flooding surfaces; see Galloway, 1989). The require-·
cene Frio Formation, and the Lower Miocene (no ment that sequence boundaries everywhere separate
formal nomenclature) of the Texas-Louisiana Gulf older from younger sediments adds to the demands
Coast all illustrate contrasting relationships between placed on the stratigrapher to resolve the age of a
growth-faulting and sedimentation patterns. One stratigraphical section. Thus, the lessons learned
important relationship concerns the relative rates of while studying the effects of growth faulting can be
subsidence and sediment supply in shallow-water applied to the problems of attempting to distinguish
clastic depositional systems, in which sea-level is the surfaces formed by normal environmental processes,
major constraint on base level (Wheeler, 1964). If such as channel migration (termed 'source diastems';
sediment supply is much greater than subsidence, Swift et al., 1991), from those formed by imposition
then significant relief is unable to develop where the of external relative sea-level controls, such as 'incised
fault trace emerges at the depositional surface. How­ valleys' (which are components of sequence
ever, where differential subsidence rates are suf­ boundaries).
ficiently high, significant topographical relief can be
created, which can then influence sedimentation
patterns. STUDY AREA AND SCOPE
Sedimentologists and stratigraphers commonly
invoke mechanisms such as changes in sea-level, Examples from diverse geographical locations andl
subsidence rate and sediment supply to explain the geological ages (Fig. 1) have been chosen to illus­
distribution of sedimentary facies and the character­ trate the principles set forth in this paper: the
istics of vertical profiles. In ancient sediments it is Palaeocene-Eocene Upper Wilcox Group of South
generally impossible to validate these interpretations Texas, the Oligocene Frio Formation of South
independently or to isolate their respective effects. Texas, and the Lower Miocene of southwest
However, growth faults that rise to the depositional Louisiana. The data set consists primarily of well
surface cause contrasting subsidence rates in juxta­ logs at a scale of 1 in. =100ft, supplemented by
posed areas, while other conditions are relatively seismic and micropalaeontological data. With cali-

TEXAS LOUISIANA
San Patricio
County

Live Oak
County
Fig. 1. Index map to the location of
the four examples presented in this
paper : 1, Upper Wilcox in Zapata
County; 2, Upper Wilcox in Live
GULF
Oak County, 3, Frio in Nueces and
OF San Patricio Counties; 4, Lower
MEXICO Miocene in Vermilion Parish.
General location of regional sandy
shelf-edges are indicated in stippled
100 Miles
patterns for Wilcox, Middle Frio
and Lower Miocene trends. (From
100 Km
fig. 2, Winker & Edwards, 1983.)
 D ff
i erent a
i l subsd
i ence 267

bration from whole cores, it is possible to make structural basins (Sloane, 1971), probably largely
fairly reliable inferences about lithology from the controlled by subsurface salt movement. Detailed
electric logs (Fig. 2). In this paper, diagrams show mapping indicates that deltas were supplied by large
only the spontaneous potential (SP) curves, but all distributary channels and incised valleys.
logs were examined and interpreted by integrating Numerous subsurface studies in this region have
the SP with the induction logs. shown that the major sand-bearing stratigraphical
The Upper Wilcox of South Texas (areas 1 and 2, units on the contemporaneous shelf and upper slope
Fig. 1) was deposited in a series of shelf-edge delta can be divided into regressive packages, or cycles,
lobes as part of the Rosita delta system (Edwards, that can be correlated from tens to hundreds of
1980, 1981). A variety of sand geometries reflecting miles along strike (e .g. Curtis & Picou, 1978). These
a varied depositional environment were identified. packages are bounded by transgressive shales that
Strike continuous, upward-coarsening sand bodies, are associated with marine flooding surfaces or maxi­
appear to represent wave-dominated shorelines, mum flooding surfaces. Recent studies tying planktic
such as strandplains, whereas sections with scattered, foraminiferal assemblages to a chronostratigraphical
blocky and fining upward sandstones suggest depo­ framework suggest that the cycles have durations of
sition along wave-influenced deltas with pronounced about 100-200 ka (Edwards, 1990; Mitchum & van
mouth-bar development. Wagoner, 1991). A recent attempt to relate Lower
The Frio of South Texas includes an interdeltaic Miocene oxygen-isotope cycles to well-log cycles
embayment dominated by stacked and prograding suggested a dominant periodicity of 100 ka (Ye
barrier bar and strandplain sandstones that have et al., 1993). In the examples that follow, most of the
pronounced strike continuity (Boyd & Dyer, 1964; correlation markers shown are thought to bound
Galloway et al., 1982a, b). Sandstones pinch out into cycles of this order (see Figs 5, 9, 10, 11 & 14).
lagoonal and continental mudstones and siltstones
up-dip, and grade down-dip into storm-deposited
interbedded siltstones and mudstones of the shelf. EFFECTS OF GROWTH FAULTING
The Lower Miocene of southwest Louisiana was ON SEDIMENTATION
deposited as a series of stacked delta lobes at the
initiation of the major sediment influx during the A considerable literature covers many aspects of the
Miocene (Curtis, 1970). Intense contemporaneous setting and effects of growth faulting on sedimen­
structural activity focused sediment into a series of tation . Curtis & Picou (1978) placed the major

1
growth-faulted trends of the Gulf Coast Basin into
an offlapping delta model. Winker & Edwards (1983)
......
HC suppression of SP
::::::: Hydrocarbons
examined the delta model in a shelf margin setting
and pointed out some of the ways in which this
Sandstone � setting differs from a platform setting. We can
envision growth faults as part of the extensional

I �
�"""" Increasing
head region of a large gravity driven slope failure
silt content
that also includes a contractional toe region with
Shale upwards folds and thrusts in deep water. Study of salt tec­
tonics in the past few years has focused on the role
of salt in terms of both lateral and vertical flowage.
Growth faults in salt-dominated areas can form in
response to: the evacuation of deep salt into extrud­
ing diapirs, the evacuation of salt sheets, or combi­
nations of salt-driven and slope-driven gravity
SP INDUCTION
systems (VendeviUe & Jackson, 1992).
The most obvious effect of growth faulting on
sedimentation is the change in thickness (Thorsen,
Fig. 2. Sample electric log showing response of SP and 1963). Aside from such thickness changes, the
induction resistivity logs to lithology. In this paper, only
character of a stratigraphical unit can appear
the SP curves are shown, but in all cases both sets of curves
were utilized in the stratigraphical studies upon which this unchanged, or can show significant changes across a
paper is based. (Modified from Edwards, 1984.) growth fault. In either case, it is important to attempt
268 M.B. Edwards

to predict sandstone properties for the purpose of

Jonnell gas 2900m J onnell gas
hydrocarbon exploration and production. The fol­
lowing sections present examples of various depo­
1 Ramos 1-- 9500' --11 Yzaguirre

sitional responses across growth faults.

Expansion only

Stratigraphical sections with large expansion ratios

usually display significant stratigraphical changes
across a growth fault. A noteworthy exception occurs
in the Upper Wilcox in Zapata Country. A fault
with an expansion ratio of greater than three, shows
no obvious changes in percentage sandstone or log
facies (Fig. 3). These sandstones were illustrated in
"
.><
stratigraphical and structural sections (see Edwards,
1981, Figs 10 & 11).
.><
" £
£ E
Comparison between the upthrown and down­ ro
E
"'
"'
thrown blocks is facilitated by changing the vertical (") �

scale of one of the logs, in order to make correlation � 0

0 0
ro
markers subparallel (Fig. 3). This procedure of

t
0 (")
'double datuming' attempts to reconstruct the strati­ c:
c: 0
graphy as if all of the locations had been subject to 0 '£
·� Q)
!/)
the same subsidence rate. In this and the following Q)
!/) a.
examples, no attempt has been made to decompact a. :0
:0 c:
the sections in order to compensate for greater a. ::
::J 0
0
compaction of shale relative to sandstone. Con­
trasting amounts of compaction could influence
present-day thickness where lithology changes sig­
nificantly across the fault. However, these sections
are overpressured, which has resulted in compara­
tively small amounts of shale compaction.
Inspection of the two logs (Fig. 3) indicates that
percentage sandstone and log facies are largely unaf­
fected by the fault, except for the presence of a high­
frequency signal in the middle part of the down-dip
well. In this example, the proportion of net sand in
the two wells is the same as the expansion ratio .

Facies changes and channel erosion

Fig. 3. Stratigraphical dip section across a large growth
Extensive mapping in the South Texas Upper Wilcox fault in Zapata County, South Texas. The vertical scales of
(Fig. 1) reveals the presence of stacked delta lobes stratigraphically equivalent section both upthrown and
downthrown to the fault have been adjusted to compensate
with mappable distributary channels (Edwards, for differential subsidence. Note the similarity of the SP
1980, 1981; Winker & Edwards, 1983). A set of logs, despite the expansion ratio of 3.45. Similar 'double­
eight wells arranged in a dip section (Fig. 4) has datumed' sections are also shown in Figs 5, 1 1 & 13.
been selected to show thickness and log facies (Modified from Edwards, 1984.)
changes (Fig. 5).
In an unfaulted area, with a uniformly increasing (Fig. 5), a striking change in log facies occurs at the
subsidence rate down-dip (flexure), a gradual change up-dip growth fault (between wells 2 and 3), with
in log facies would be expected, with a change from more subtle changes at the other faults. The effects
distributary channels up-dip, through mouth bar to of growth faults on stratigraphical preservation can
distal mouth bar down-dip. However, in this example be observed by comparing and contrasting adjacent
 Dfi ferent a
i l subsd
i ence 269

with 'foreshortening' of dipwise facies gradients (the

Texas Louisiana INDEX
MAP distance along dip from proximal to progressively
distal facies) rather than down-dip translation of
facies belts by continued progradation into the basin.
A schematic facies preservation diagram (Fig. 6)
enables prediction of the depositional sequences
that can be juxtaposed across a growth fault in a
prograding delta. At the top (Fig. 6A) is a subsidence
5ooo· rate graph . The duration of the progradational phase
I I
1500 m is suggested to have lasted about 1000 yr, although it
could have been much shorter or longer. The growth­
fault curve (solid line) shows subsidence rates that
reflect the thickness changes observed due to growth
faulting and roll-over (up-dip thickening on the
downthrown block toward the fault). The vertical
steps show the locations of individual growth faults.
Shown for reference is the flexure curve (dashed
line), which assumes a linear increase in subsid­
Live Oak County
ence rate down-dip. Arrows labelled U and D are
Fig. 4. Index map showing the location of wells depicted in explained below.
Fig. 5, Live Oak County, Texas (see Fig. 1 for context). Figure 6B illustrates some of the facies relation­
ships and inferred environments of deposition. After
transgression and abandonment of the previous delta
subsidence-normalized wells that are in the same lobe, the cycle begins by progradation of the delta
versus different fault blocks. For example, in the front, with superimposed higher frequency trans­
Luling regressive cycle (Fig. 5B), wells 1 and 2 are gressive -regressive cycles. The progressive basin­
very similar, as are wells 3 and 4. However, wells 2 ward shift of environments down-dip has been shown
and 3 are clearly different. by the down-dip termination of channel erosion
The preserved components of each progradational surfaces and mouth bars. Coastal-plain facies are
deltaic cycle (Fig. 5) include mouth bar and distal depicted as blocky sandstones, although they are in
mouth bar deposits, typically arranged in upward­ reality complex intercalations of various channel
coarsening successions; distributary channel deposits and bay deposits. There is a total of 5 0-200ft
represented by blocky sandstones and upward-fining (15 -60m) of subsidence at the up-dip end, whereas
successions; and coastal plain and interdistributary at the down-dip end there is 25 0-2000ft (75 -600 m)
bay deposits, which consist of mixed sandstones and of subsidence . It is assumed that distributary chan­
shales in a variety of patterns, usually thinner than nels are 5 0-200ft deep regardless of location,
the overall regressive succession . although it is likely that they would decrease in
Subsidence creates space for the preservation of depth down-dip due to channel bifurcation and dis­
facies beneath erosion surfaces. In this example, charge through crevasses.
major fluvial erosion is represented by the base of The subsidence curves can be used to estimate
the blocky and upward-fining sandstones. Subsid­ preservation of vertical sections on either side of a
ence allows the deposition and burial of pro­ growth fault. At a growth fault, the regionally aver­
gradational facies beneath the depth of erosion aged subsidence rate does not occur. Instead, much
attained by channels. This relationship becomes higher subsidence rates occur on the downthrown
critical in up-dip areas, where the decreasing thick­ block, and much lower rates occur on the upthrown
ness of the space available for preservation of block. On the upthrown block (see Fig. 6A), the
progradational facies approaches the depth of dis­ lower subsidence rates resulted in a vertical section
tributary channels. Up-dip of this point, the Slick that resembles that developed up-dip. This is shown
and Luling genetic regressive successions are charac­ by the arrow U, which is projected to the left until it
terized by amalgamated sandy channel-fill deposits. intersects the flexure curve, where the appropriate
Subsidence rate also affects regional facies trends. lower subsidence rate would have occurred in a non­
High subsidence rates favour stacking of delta lobes growth-faulted setting (depicted by flexure curve in
270 M.B. Edwards

Downdip
Updip

2 3 4 5 6 7 8

- -100m

91m - -500'

- -1000'

\F - -1500'
A - -500m
u
L

19
� T
'
5
-;--
-2000'
A Paleostructure
579m 614m

B Subsidence normalized

2 3 4 5 6 7 8

PD PD
F
L
T

Fig. 5. Thickness and facies changes in two stacked delta complexes, the older Luling (A and B) and the younger Slick (C
and D) units in the Upper Wilcox of South Texas (see Figs 1 & 4 for locations) . For both deltas the upper panels (A and C)
show a palaeostructural cross-section with the top to the delta complex as the upper datum, and growth faults that were
active during delta formation. The lower panels (B and D) are 'double-datumed' sections in which vertical scales were
adjusted to normalize for differential subsidence. Generalized deltaic facies are identified using electric log characteristics.
CH, channel; MB, mouth bar; PD, prodelta. No horizontal scale.

Fig. 6A and facies distribution in Fig. 6B). At the down into the facies diagram below. This procedure
intersection point, the arrow is then extended down predicts that the vertical sections on either side of
into the facies diagram, in order to determine the the growth fault would show considerable differences
vertical section that corresponds to that subsidence in facies profiles that could not be explained solely
rate. as a function of changing palaeogeographical lo­
A similar procedure can be carried out for the cation. The thickness of the upthrown and down­
downthrown block by projecting the downthrown thrown profiles could then be expanded or contracted
subsidence rate to the right (arrow D, Fig. 6A) until to restore their present-day relative thicknesses.
it intersects the flexure curve, and then extending it However, this method may predict excessive lateral
 Different a
i l subsd
i ence 271

Updip Downdip

2 3 4 5 6 7 8

- -100'

- -200'

- -300'

- -400'

- -500'

- -600'

C Paleostructure

D Subsidence normalized

2 4 5 6 7 8

:><:
u
::::;
PO (f)

Fig. 5. (Continued.)

facies changes because the distance between the two suggest that sand was supplied to the Frio shoreface
schematic wells (below arrows U and D, Fig. 6) will by wave-driven alongshore transport in the fore­
be greater than the actual distance between two shore/upper shoreface, rather than by prograding
wells on either side of a growth fault. deltas (Martin, 1969).
A typical well-log in the Frio shows an upward
transition from shelf mudstone to strandplain sand­
Facies changes and storm/wave erosion
stone over a depth range of about 5 000 ft. Subdivision
The South Texas Frio Formation (Fig. 7) illustrates of the section into component cycles reveals that
shorelines that prograde as a line source (i . e . wave­ the thickness of the cycles increases markedly with
dominated deltas, barrier islands and strandplains), greater depth, from less than 5 0 ft to almost 5 00ft
rather than point sources (such as river-dominated (Fig. 8). This suggests that structurally deeper sedi­
deltas). It is based on complex stratigraphical and ments were deposited at much greater subsidence
structural relationships determined from approxi­ rates than shallower Frio sediments. Rigorous cor­
mately 600 well-logs (Weise et a!., 1981; Bebout relation of hundreds of well-logs in the area allows
et a!., 1982; Edwards, 1986), coupled with micro­ the tracing of these varied cycles across growth
palaeontological studies that indicate neritic to faults. Palaeostructural cross-sections (Figs 9 & 10)
coastal environments throughout the section, with show the position of growth faults and their effect on
overall shallowing upwards (Martin, 1969, 1970). thickness and log facies.
The sandy, wave-dominated Frio shoreline was The analysis of log facies changes is presented in a
bordered by muddy coastal plains up-dip, muddy set of four well-logs, each in different fault blocks
slope deposits down-dip, and by major deltaic depo­ (Fig. 11). The log correlations look questionable at
centres along strike (Galloway et a!., 1982a, b). their original depth-scales, but when normalized for
These palaeogeographical relationships strongly differential subsidence, the correlation of the indi-
272 M.B. Edwards

A Subsidence rates (thickness/arbitrary unit of time)

0' 1-��;===�---- --�� ----1�� ��---- ------------------------�_0'

Growth faults \
. _1 U Upthrown
- -- =
I= -
� -- -
----
---J / Downthrown
. _
� F lexure rr:- / .=
-
- -
I=
---�
- -

50'-�
1=
200' 15-60m
D

\"-0II0"e
r = 250'-
2000'
--
u D 75-600m
B Facies
Updip Downdip

a
o E
0 0
� �
b �
� ....

Fig. 6. Schematic diagram of facies preservation relationships in a deltaic environment characterized by significant channel
erosion. (A) Two subsidence rate relationships are shown, a straight dashed line indicating uniformly increasing subsidence
rate with distance as in a flexure, and a complex solid line indicating abrupt increases in subsidence rate down-dip across
growth faults and decreasing subsidence rate up-dip within some fault blocks due to roll-over. (B) Facies preservation
trends and characteristic SP logs. Proximal to distal facies trends have been exaggerated. The figure can be used as a kind
of nomograph to estimate facies changes across growth faults in this setting. See text for additional explanation. (Modified
from Edwards, 1984.)

vidual cycles is much more obvious. Figure 11 illus­ production that can be achieved despite the thin,
trates the change in log facies from thick sand bodies ratty character of the sands. The large amount of
up-dip, to thin interbedded sandstones (siltstones) interbedded mudstone was presumably derived from
and mudstones down-dip. The palaeostructural sec­ the adjacent deltaic depocentres.
tions emphasize that the facies changes take place A schematic diagram (Fig. 12) suggests how abrupt
abruptly across the growth fault. However, the cor­ facies changes can develop across growth faults,
relatability of the individual cycles across the fault without the necessity of an abrupt change in depo­
suggests that significant sea-floor relief was not main­ sitional environment. The model proposed in Fig. 12
tained at the location of the fault. relies upon differential subsidence to produce a con­
Core study of the down-dip interbedded facies trast in the sediments preserved, even though the
(Berg & Powell, 1976) suggested that the sandstones depositional processes on both sides of the fault are
were deposited by turbidity currents. Studies of the virtually identical. Unfortunately, the available data
modern Texas shelf, however, suggested that shelf do not allow many aspects of the model to be
sands are deposited from storm-generated geo­ resolved until additional core data are obtained .
strophic flows that produce turbidite-like beds The dynamics of shoreface -shelf systems have
(Morton, 1981). Transport is oblique to the shore­ been discussed by many authors (e .g. Swift &
line, but the resulting thin beds have enormous Thorne, 1991), however, a considerable number of
strike continuity (Snedden & Nummedal, 1991). issues remain contentious. For the present dis­
The continuity is consistent with the excellent gas cussion, it suffices to emphasize the key points. The
 Dff
i erent ai l subsd
i ence 273

shales down-dip. It seems likely that the coarse­

Louisiana
Texas grained shelly sandstones up-dip were deposited in a
foreshore and upper shoreface setting and are com­
posed of wave-reworked and amalgamated storm
beds. The ratty interbedded section down-dip was
deposited on the lowermost shoreface to shelf and
consists of interbedded storm-deposited sands and
muds. The abrupt facies change at the fault remains
N to be explained. The important distinction is that

t
mud layers are preferentially preserved on the more
rapidly subsiding down-dip block. The mud could
have been deposited from suspension either by turbid
nearshore, semi-permanent currents, or by waning
10 Miles flow following a storm event. In either case, it is
Nueces I important that the mud was not eroded subsequently,
10 Km
either by storms or fair-weather wave activity.
Kleberg The Frio shelf system can be regarded in terms of
profiles of equilibrium and changing base levels.
Fig. 7. Index map showing location of cross-sections Areas of greater subsidence rate permitted less sedi­
A-A' and B-B' in Figs 9 & 10, and general location of
Frio investigation around Corpus Christi Bay. ment to be reincorporated by storm currents and
bypassed on to the shelf, and hence more sediment
was allowed to accumulate. If eustasy and sediment
precise roles of fair-weather versus storm conditions supply are held constant, then the high subsidence
in transporting sediment on the upper shoreface is rates on the downthrown block effectively raise
often unclear and may vary from one shoreline to the local base level (increase accommodation ) ,
another. The Frio shoreline prograded steadily increasing the probability that storm-deposited muds
through time due to the high sediment supply, and would be preserved. At some critical subsidence
significant wave transport along shore and storm rate, sand beds stop being amalgamated, and instead
transport to the inner shelf. are separated by thin mud layers. At this point, the
Logs and whole cores indicate a progression from reservoir characteristics of the sand body change
relatively coarse-grained homogeneous sandstones drastically.
up-dip, through interbedded thick sandstones and The most rapidly subsiding areas immediately
shales, to thinly interbedded 'ratty' siltstones and down-dip of the growth fault would probably have

A 8 c
7000' -

1 1
�

0
8000'-

9000' -
J A
�
� 1
E
�
� E E

�
"'
"' 0 (")
"'
a: t:. :::.
Fig. 8. Example of Frio log from
Corpus Christi Bay showing vertical
u..
3000m - ] 8 0
"'
0 0
10 000' - N (")
"'
...
"'
changes in cycle thickness,

� j l j
suggesting increasing subsidence
rate with depth. Small portions of
the log have been enlarged to 11 000' -
illustrate cycle styles and thickness.
The numbers next to logs A, B and
C refer to the thickness of the log 12 000' -
] c

segments illustrated.
274 M. B. Edwards

NW 30 31 32 33 34 35 37 38 SE
GULf GULf CITIES SER !c SUNRAY ATlANTIC CHERRYVILLE ATlANTIC TENNECO
2 ST TR 47 2 ST TR 53 1 ST TR 52 1 ST TR 51 1 ST TR 81 1 ST TR 81 6 ST � 470 1 �
w IA TANG I '
A -1> -1> -1> -1> -1> -1> A
(8560') (8678') (8670') (8740') (8885') (8983') (10360') (10038')
�;�o���o����o���o��!�O·�=-+- ----.!i�:':n='-------t---.,2��0� OAlUM
HORIZON 10

I
INDEX MAP
E
oo
oo
I{)

PALEOSTRUCTURAL DIP SECTION

IN FRIO FORMATION, NUECES CO., TX.
DAlUM: LOCAL MARKER IN FRIO
V. E. = 32 X

Fig. 9. Frio palaeostructural section A-A' (see Fig. 6 for location and Fig. 1 for setting) showing an overall regressive
stratigraphical interval bounded by marine flooding surfaces expanding dramatically across two growth faults, with
concomitant facies changes. (Modified from Edwards, 1986.)

served as an effective sediment trap, especially as ence occasionally results in a significant topographi­
the local onshore-directed subsidence gradient was cal scarp being developed where the fault intersects
opposed to the obliquely offshore-directed sediment­ the sea-floor. This is illustrated by the Lower
transport gradient. Storm-deposited sands typically Miocene of southwestern Louisiana (Fig. 1), where
pinch out down-dip, leading to the development of a large depositional basin formed as a result of the
combination traps in off-structure positions. sudden and rapid removal of subsurface salt, and
then stabilized when the salt body was fully evacu­
ated (Edwards, unpublished).
Facies changes at faults that develop
Thin stratigraphical units in this area were mapped
topographical relief
using almost 2000 well-logs, micropalaeontological
The balance between sediment supply and subsid- data, and seismic data. The maps show the devel--
 Different ial subsd
i ence 275

39 40 41 42 43 44 3
BRITlSH-AMERICAN CITlES SERVlCE HAMON HAMON HAMON TENNECO RENWAR I< PHILLIPS sE
#1 ST TR 12 13 ST TR 10 #1 ST TR 15 #3 ST 8#2 ST 8#2 ST TR 5 #1 HOGG

i} f/1 i} f/1 B'

(8185') (8288') (8549') (9400') (9121') (9029') (9021') (9220') (8905')

HORIZON 20

B E
1 oo
INDEX MAP oo
*-...... 2 1[)..--
*
\ 39 40
N
*-0
\ 41
42 43 PALEOSTRUCTURAL DIP SECTION

1
* ....
44
10,000' *·-¢-.0 NUECES AND SAN PATRICIO COS., TX
3
3000 m '---- DATUM: LOCAL MARKER IN FRIO
0 B' V. E. = 32 X

Fig. 10. Frio palaeostructural section B-B' (see Fig. 6 for location and Fig. 1 for setting) showing dramatic expansion of
two sandy regressive cycles, each bounded by marine flooding surfaces, across a growth fault, with concomitant facies
changes. Note excellent roll-over structure with expansion and improved development of sandstones up-dip to the
northwest into the growth fault. (Modified from Edwards, 1986.)

opment of channel sandstone bodies that traverse other settings (e.g. Hopkins, 1987; Leeder &
the up-dip stable shelf and trend toward the down­ Alexander, 1987).
dip basin, where large quantities of sand were depo­ The up-dip areas (Fig. 13B) appear to be com­
sited on the downthrown side of the fault (Fig. 13). prised largely of shallow-water mudstones and local­
The pattern is repeated in several successive units ized mouth-bar sandstones, which were deposited
(Fig. 14). while sea-level was relatively high . The development
The coincidence of the structural boundary and of narrow incised valleys allowed the preservation of
the facies boundary indicates that the basin was a these fine-grained deposits, and contrasts with the
topographical as well as a structural low, which not Wilcox example, described above, in which there
only collected sediment but also attracted distribu­ was extensive scouring at the bases of distributary
tary channels and incised valleys from the adjacent channels. The apparently stable channels suggests a
highs (Fig. 13B). Similar relationships between struc­ relative fall in base level, due either to eustasy or to
tures, relief and facies have been recognized in local structural uplift beneath the upthrown block
276 M.B. Edwards

Updip Down dip Updip Down dip

2 3 4 Datum 2 3 4 Datum

Top J Top J

Paleostructure

2000' Base J
610m
A B 230' 405' 500' 2000'

Fig. 11. (A) Frio schematic palaeostructural section of one stratigraphical unit ('J') with well-log segments selected from
four wells. (B) Frio stratigraphical section with vertical scales adjusted to normalize for differential subsidence across
growth faults. The correlations suggest that thin digitated SP facies up-dip have closely time-equivalent ratty/serrated SP
facies down-dip. Figures 9 & 10 demonstrate that the facies changes occur at growth faults.

due to movement of deep salt. The down-dip basin­ environment. As noted above, absolute subsidence
fill attests to the huge sediment volumes that were rate has to be evaluated in the context of the other
bypassed through the relatively small valleys. variables that affect base level: eustatic fluctuations,
and the amount and calibre of sediment supply (e.g.
Swift & Thorne, 1991). For example, in the deltaic
PRESERVATION POTENTIAL setting discussed above, distributary channels are a
THRESHOLDS significant source of erosion. With sufficiently low
subsidence rates, migrating channels will remove all
The above examples raise the question as to why or most of the progradational facies, resulting in
some of those faults that are not associated with amalgamated multistory channel sand bodies. When
depositional topography show dramatic facies the threshold subsidence rate is exceeded, progra­
changes, whereas others do not. It is customary to dational facies will be preserved . Another example
express the significance of a growth fault in terms of is the balance of erosion and deposition on the
its growth ratio, as this can be measured directly in shoreface . A threshold value of subsidence, in the
the subsurface data. However, this ratio gives no context of the other controlling variables, separates
information about the absolute rates of subsidence regimes in which mud layers will be eroded from
on either side of the fault. It seems likely that a those in which they are preserved. Analogous thres­
major factor controlling facies development is the holds could be postulated for other environments
absolute subsidence rate. and sedimentary features.
Along a continuum of subsidence rates, will be
rates that separate domains that are associated with
the preservation or non-preservation of a particular CONCLUSIONS
facies component, or sedimentary feature (Fig. 15).
These specific subsidence rates can be referred to as 1 The interplay between erosional and depositional
preservation potential thresholds for a particular processes in sedimentary environments is controlled
sedimentary feature in a particular depositional to a large extent by subsidence rates. The latter
 Dff
i erent ial subsd
i ence 277

Bay/ Barrier/
l
FORESHORE
SHOREFACE
Storm
SHELF

lagoon strandplain erosion _

A Stable setting

B Unstable setting

Prograding wave-dominated shoreline

Fig. 12.Schematic cross-sections showing wave-dominated shoreline progradation in (A) stable and (B) unstable settings.
Characteristic well-logs (SP only) are shown. (A) This section indicates the presence of wave grading shaping the
foreshore and upper shoreface, and storm events resulting in erosion of the upper shoreface and deposition of a storm
couplet on the shelf. (B) The enhanced subsidence rates in the unstable setting allows for the preservation of the muddy
portion of the storm couplet from erosion, whether by wave grading or a subsequent storm event. The enhanced
preservation seems to have resulted from greater subsidence rates, which modified local base levels. (Modified from
Edwards, 1984.)

A Topography B Topography
healed by controls
sedimentation sedimentation

Fig. 13. Schematic block diagram

contrasting: (A) normal
sedimentation across growth fault
Amalgamated
without development of
channel
topographical relief, versus (B)
complex
growth fault with associated
topographical scarp and bypassing
of coarse sediment across up-dip
Expansion with drastic facies changes
block to rapidly subsiding down-dip
across growth fault
block. Based on a study of the
Lower Miocene. (Modified from Expansion without facies changes
Edwards & Tuttle, 1993.) across growth fault
278 M. B. Edwards

- Downdip- A B
Hawthorne Texaco - Downdip-
1 Delcambre E lle Blanc B
135-04E-001 135-04E-Q49
(12281') (12036') Fault/channel relationships
137'/42m 143'/44m

Jllt lllf
1 Fault buried;
no channels in area
,04

168'/51m 316'/96m

Jllt�JIIr
2 Fault active but channels
not closely related
to fault traces

108'/33m 283'/86m

�II).: (IIi
3 Fault active with tight
control on channel location.
Channel updip and downdip

153'/47m 743'/226m

ill}.} ill[
4 Fault active with tight
control on channel location.
Channel in downdip well only

132'/40m 780'/238m

5 Fault active with tight

control on channel location.
Channel in downdip well only

411'/125m -1700'/520m

111r� �111
6 Fault active with tight
control on channel location.
Channel in downdip well only

Fig. 14. (A) Palaeostructural cross-section showing change of Lower Miocene section across a prominent growth fault that
had topographical relief developed during part of its history. The section was subdivided into six units for mapping and
correlation, of which the deepest unit has no clear lower boundary in this well. (B) The six units are shown with vertical
scales adjusted to normalize for differential subsidence across the fault. For each layer, the original thicknesses are shown,
and the expansion ratio is shown between the logs.
 Dff
i erent ial subsd
i ence 279

Examples of different fault/subsidence relationships

0)(1) - (1)
c-
·- :J ro
�

1
V>-
0
"' Vl "'
"' Fault A Fault Fault 2 Fault 3 Fault 4
� .0
(.)
c
(.) "' "'
.!: "0
'iii D
.0
::J
Vl
D u

Preservation threshold: - ----------- --- ------- - -- --------- -- ----- - -- ---- -

storm-deposited mud
D u

]\�1: : 1\�
Preservation threshold: -
- ------- - --- -- - - - - - --- - - -- ----
progradational facies
D D

Channel inc;sion
threshold:
+
-0 -
u
�--
\\
--
J _i��_]__
-- - --
\\
---- -

-I
Presence of features

Preservation of

:I I I I I I I I I I I
storm-deposited mud • I§! •
Preservation of
progradational facies • I§! • I§! •
Updip bypass due to I§!
channel incision

Fig. 15. Representation of preservation potential thresholds in growth-fault settings. The vertical axis shows increasing
absolute subsidence rate, but is schematic with no scale. Most of the stratigraphical units depicted in this study are
considered to be 'fourth order', and hence have a duration of the order of approximately 100 ka. Preservation thresholds
are depicted for progradational facies at a lower rate than that for storm-deposited mud in the transitional lower shoreface
to shelf environment. Five different fault situations indicate how various combinations of depositional environment and
subsidence rates determine whether particular facies will or will not be preserved across specific growth faults. Note that
expansion ratio is related to relative rather than absolute subsidence rate. If blocks on either side of a growth fault are on
the same side of the preservation threshold, then facies patterns will be similar with regard to the particular process
threshold, although thickness will be different. Settings that represent examples described in this paper are fault A, Lower
Miocene of southwestern Louisiana; fault 1, amalgamated sandstones on both sides of fault common in up-dip areas; fault
2, Wilcox in Live Oak County, Texas; fault 3, Frio in South Texas; fault 4, thick sections with similar facies on both sides of
the growth fault, common in down-dip areas.

affects preservation potential via thresholds that major facies, and not just the preservation potential
separate domains of facies preservation versus of sensitive component facies.
erosion. 4 The documented effects of preservation thresholds
2 Subsidence rates, through their effect on preser­ on reservoir architecture indicate that it would be
vation potential, influence facies composition, useful to be able to predict preservation patterns on
including geometry, bedding characteristics, fabric, unexplored fault blocks in growth-faulted basins.
palaeontology and seismic response. The effect of However, the present study indicates that expansion
subsidence rates is readily demonstrated in the Gulf ratios are generally inadequate to make such predic­
Coast Basin due to the presence of growth faulting. tions: absolute subsidence rates are required, but
In tectonically stable basins, the effect of subsidence are often difficult to obtain or predict.
rate on facies composition may not be determined
readily.
3 Where subsidence rates were sufficient to create ACKNOWLEDGEMENTS
topographical scarps at the depositional surface, the
effect of faulting on sedimentary facies is greater, as The Wilcox and Frio work was initiated while I
there is then a feedback effect on the location of was with the Bureau of Economic Geology of the
280 M. B. Edwards

University of Texas. I thank my colleagues there EDWARDS, M.B. (1984) Stratigraphic and structural analysis
for their generous support, especially Don Bebout, of growth-faulted regions using well logs: a workshop.
Houston, Texas. Unpublished lecture notes and
Bonnie Weise and Bill Galloway. Competent assist­
problems.
ance was provided by Rick Schatzinger, Jim Lockley, EDWARDS, M.B. (1986) Sedimentary effects of differential
Down Downey, Susan Hallam, Steve Mann and subsidence in Frio shoreface-shelf sediments, Gulf
Doug Wilson. Contract funding which led to some Coast Tertiary. Houston geol. Soc. Bull., 28, 10-14.
EDWARDS, M . B . (1990) Stratigraphic analysis and reservoir
of these results was provided by the US Department
prediction in the Eocene Yegua and Cook Mountain
of Energy and the Gas Research Institute. This early Formations of Texas and Louisiana. In: Sequence Stra­
and much of my subsequent work on Gulf Coast tigraphy as an Exploration Tool (Ed. Armentrout, J .M.),
onshore stratigraphy has been supported by many pp. 151-164. Gulf Coast Section, Soc. econ. Paleontol.
oil and gas companies, for which I express my Mineral. 1 1th Ann. Res. Conf.
EDWARDS, M.B . & TuTTLE, J.L. (1993) Regional Sequence
appreciation. The Miocene work benefited from the
Stratigraphy and Exploration Potential of the Lower
micropalaeontological expertise of J . Loyd Tuttle Miocene of Southwest L ouisiana. Proprietary industry
and seismic data from Geophysical Pursuit, Inc. study (unpublished) 121 p.
Yvonne Bowlin ably assisted in all parts of this GALLOWAY, W . E . (1989) Genetic stratigraphic sequences

study. in basin analysis I: architecture and genesis of flooding;­

surface-bounded depositional units. Bull. Am. Assoc.
This manuscript has been improved by the reviews petrol. Geol. , 73, 125-142.
of William R. Dupre, William C. Ross and the GALLOWAY , W.E., HoBDAY , D . K. & MAGARA, K. (1982a)
Editor Guy Plint. In addition, many colleagues , too Frio Formation of the Texas Gulf Coast Basin - depo­
numerous to mention, have discussed the concepts sitional systems, structural framework, and hydrocarbon
origin, migration, distribution, and exploration potential.
discussed herein, offering anything from constructive
Tex. Univ. Bur. econ, Geol. Rep. Invest., 122, 78 p.
criticism to disbelief. GALLOWAY, W.E., HOBDAY, D . K. & MAGARA, K., (1982b)
Finally, I thank Harold Reading for picking me up Frio Formation of the Texas Gulf Coast Plain: depo­
at Gatwick after my first trans-Atlantic journey, sitional systems, structural framework, and hydrocar­
and for attempting to teach me to drive on the bon distribution. Bull. Am. Assoc. Petrol. Geol. , 6�>,
649-688.
correct side of the road. HoPKINS, J.C. (1987) Contemporaneous subsidence and
fluvial channel sedimentation: Upper Mannville C Pool,
Berry Field, Lower Cretaceous of Alberta. Bull. Am.
Ass. petrol. Geol. , 71, 334-345.
REFERENCES
LEEDER, M.R. & ALEXANDER, J. ( 1987) The origin and
tectonic significance of asynunetrical meander-belts.
BEBOUT, D.G., WEISE, B.R., GREGORY, A.R. & EDWARDS, Sedimentology, 34, 217-226 .
M.B. ( 1982) Wilcox sandstone reservoirs in the deep MARTIN , G.B. (1969) The subsurface Frio of South Texas:
subsurface along the Texas Gulf Coast. Tex. Univ. Bur. stratigraphy and depositional environments as related
econ. Geol. Rep. Invest., 117, 125 pp. to the occurrence of hydrocarbons. Trans. Gulf Coast
BERG, R . R . & PowELL, R . R. (1976) Density-flow origin Assoc. Geol. Soc. , 19, 489-499.
for Frio reservoir sandstones, Nine Mile Point Field, MARTIN , G.B. (1970) Depositional history: key to explo­
Aransas County, Texas. Trans. Gulf Coast Assoc. Geol. ration. Oil Gas J. , January 12, 98-106.
Soc. , 26, 310-319. MiTCHUM, R.M. & VAN WAGONER, J.C. (1991) High­
BoYD, D.R. & DYER, B.F. ( 1964) Frio barrier bar system frequency sequences, and their stacking patterns:
of south Texas. Trans. Gulf Coast Assoc. Geol. Soc., 14, sequence-stratigraphic evidence of high-frequency
309-321. eustatic cycles. Sediment. Geol. , 70, 131-160.
CuRTIS, D.M. ( 1970) Miocene deltaic sedimentation, MoRTON, R . A . (1981) Formation of storm deposits by
Louisiana Gulf Coast. In : Deltaic Sedimentation ­ wind-forced currents in the Gulf of Mexico and the
Modern and Ancient (Eds Morgan, J.P. & Shaver, R.H.), North sea. In: Holocene Marine Sedimentation in the
Spec . Pub!. Soc. econ. Paleontol. Mineral, Tulsa, 15, North Sea Basin (Ed. Nio, S.D.), Spec. Pubis int. Ass.
293-308. Sediment., No. 5, pp. 385-396. Blackwell Scientific
CuRTIS, D .M. & Picou , E.B., JR. (1978) Gulf Coast Publications, Oxford.
Cenozoic; model for application of stratigraphic concepts PoSAMENTIER, H . W . & WEIMER, P. ( 1993) Siliclastic
to exploration on passive margins. Trans Gulf Coast sequence stratigraphy and petroleum geology - where
Assoc. Geol. Soc. , 28, 103- 120. to from here? Bull. Am. Assoc. petrol. Geol. , 77 ,
EDWARDS, M.B. (1980) The Live Oak delta complex : an 731-742.
unstable shelf-edge delta in the deep Wilcox trend of SLOANE, B.J. (1971) Recent developments in the Miocene
South Texas. Trans Gulf Coast Assoc. Geol. Soc., 30, Planulina gas trend of south Louisiana. Trans Gulf Coast
71-79. Assoc. Geol. Soc. , 21, 199-210.
EDWARDS, M.B. (1981) Upper Wilcox Rosita delta system SNEDDEN , J.W. & NU MMEDAL, D. (1991) Origin and
of South Texas: growth-faulted shelf-edge deltas. Bull. geometry of storm-deposited sand beds in modern sedi­
Am. Assoc. petrol. Geol. , 65, 54-73. ments of the Texas continental shelf. In: Shelf Sand and
 Dff
i erent ial subsd
i ence 281

Sandstone Bodies: Geometry, Facies and Sequence Stra­ Bur. econ. Geol. Rep. Invest. , 209, 60 p.
tigraphy (Eds Swift, D.J.P., Oertel, G.F., Tillman, R . W. WEISE, B.R., EDWARDS, M.B . , GREGORY , A.R . , HAMLIN,
& Thorne, J.A.), Spec. Pubis int. Ass. Sediment., No. H.S. , JIRIK, L . A. & MORTON , R.A. (1981) Geologic
14, pp. 283-308. Blackwell Scientific Publications, Studies of Geopressured and Hydropressured Zones in
Oxford. Texas: Test-well Site Selection, Final Rep ort. Texas
SwtFT, D.J.P. , PHILLIPS , S. & THORN E , J .A. (1991) Sedi­ University Bureau of Economic Geology, unpublished
mentation on continental margins, IV: lithofacies and contract report prepared for Gas Research Institute,
depositional systems. In: Shelf Sand and Sandstone 308 pp.
Bodies: Geometry, Facies and Sequence Stratigraphy WHEELER, H.E . ( 1964) Baselevel, lithosphere surface, and
(Eds Swift, D.J.P. , Oertel, G.F., Tillman, R.W. & time-stratigraphy. Geol. Soc. A mer. Bull., 75, 599-6 10.
Thorne, J.A.), Spec. Pubis int. Ass. Sediment., No. 14, WINKER, C. D. & EDWARDS, M.B. (1983) Unstable progra­
pp. 89- 152. Blackwell Scientific Publications, Oxford. dational clastic shelf margins. In: The Shelfbreak: Critical
SwtFT, D.J.P. & THORNE , J.A . (1991) Sedimentation Jnte1Jace on Continental Margins (Eds Stanley, D.J. &
on continental margins, I: a general model for shelf Moore, G.T.), Spec. Publ. Soc. econ. Paleontol. Min­
sedimentation. In: Shelf Sand and Sandstone Bodies: eral . , Tulsa, 33, 139-157.
Geometry, Facies and Sequence Stratigraphy (Eds Swift, YE, Q . , MAn-HEWS, R.K., GALLOWAY, W. E . , FROHLICH, C.
D.J.P., Oertel, G.F., Tillman, R.W . & Thorne, J.A.), & GAN , S. (1993) High-frequency glacioeustatic cyclicity
Spec. Pubis int. Ass. Sediment . , No. 14, pp. 3-3 1 . in the Early Miocene and its influence on coastal and
Blackwell Scientific Publications, Oxford. shelf depositional systems, NW Gulf of Mexico Basin.
THORSE N , C. E. (1963) Age of growth faulting in southeast In: Rates of Geologic Processes: Tectonics, Sedimen­
Louisiana . Trans. Gulf Coast Assoc. Geol. Soc., 13, tation, Eustasy and Climate (Eds Armentrout, J.M.,
103- 1 10. Bloch, R. & Olson, H.C.), pp . 287-298. Gulf Coast
VENDEVILLE, B.C. & JACKSON , M.P.A. (1992) The rise and Section, Soc. econ. Paleontol. Mineral. 14th Ann. Res.
fall of diapirs during thin-skinned extension. Tex. Univ. Conf.
 Sequence and Seismic Stratigraphy
in Facies Analysis

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
Spec. Pubis int. Ass. Sediment. (1995) 22, 285-303

Seismic-stratigraphical analysis of
large-scale ridge-trough sedimentary structures
in the Late Miocene to Early Pliocene
of the central North Sea

J O E CAR T WRI G HT
Department of Geology, Imperial College of Science, Technology and Medicine,
Prince Consort Road, London SW7 2BP, UK

ABSTRACT

This paper describes the geometry, seismic-stratigraphical characteristics and palaeogeographical setting
of a system of ridges and troughs measuring approximately 5-20km long, by 1-3km wide by
100-300 m high, developed in the mudstone-dominated Upper Miocene and Lower Pliocene of the
central North Sea. The ridge-trough system occupied a base-of-slope to basin floor setting centred on
UK Quadrant 22 at the time of deposition. The ridge crests are oriented NNW-SSE and are almost
straight in plan view. The troughs are characterized by incision at the base, followed by drape, and
subsequent onlap fill. The ridges and troughs have a distinctly aggradational character.
Two possible mechanisms are proposed to explain the development of these structures. The
geometry and scale of the ridge-trough structures is similar to that described for mud waves formed on
deep-marine sediment drifts by the action of oceanic bottom currents. By analogy, it is possible that the
ridge-trough structures could have formed in response to bottom currents flowing along the axis of the
central North Sea, linking the Norwegian-Greenland Sea with the Bay of Biscay. This mechanism
requires a marine connection in the southern North Sea, and direct evidence for this possible connection
has been removed by erosion during the Pleistocene. An alternative mechanism is that the ridge-trough
system formed in response to downslope-directed currents of unspecified type developed on the pro­
delta slope. This mechanism is appealing in that it explains the orientation of the ridges and troughs
(orthogonal to the nearest delta front) , and, most significantly, explains the significant incision observed
along the troughs. The proximity of the ridge-trough structures to a rapidly prograding Neogene delta
(the Skaggerak Delta) suggests a close link between pro-delta processes and the formation of these
structures. The necessary confinement of the current system may have been provided by slope gullies,
but conclusive evidence of this on the regional seismic data is lacking at present.

INTRODUCTION

This paper describes a set of large-scale sedimentary described from the Cenozoic of the North Sea. They
structures that are developed in Upper Miocene to are on average 5-20 km long, and have a fairly
Lower Pliocene mudstones in the central North Sea. regular spacing of 1-3 km. They have not been
The structures are a series of ridges and troughs that described in any previous publications on the North
are distributed throughout an elliptical area centred Sea Basin, and directly analogous structures have
on UK Quadrant 22 (Fig. 1). The ridge-trough not, apparently, been described from any other
structures formed in a physiographic base-of-slope­ basin.
basin-floor setting, although the maximum water The lack of attention given to these extraordinary
depths were relatively shallow, between 200 and structures can probably be attributed to the fact
3 00m. The ridge-trough structures are by far the that they are contained in an interval that has no
largest type of discrete sedimentary structure so far economic potential from a petroleum perspective.

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0

285
286 J. Cartwright

The quality and quantity of petrophysical, litho­ spacing of 25 km was used to map the distribution
logical and biostratigraphical data available for the of the ridge-trough system (Fig. 1). The area of
Neogene interval in the North Sea is greatly inferior interest is located in the axis of the Cenozoic Basin,
to that available for the more prospective Palaeogene close to the region of maximum total Cenozoic:
and older formations. This contrast in the well data­ subsidence and sediment accumulation (Ziegler,,
base explains why there have been so few studies 1982; Joy, 1993).
on the Neogene of the North Sea, and contrasts
markedly with the vast literature available on the
Cenozoic Basin evolution
Palaeogene (see Lovell (1990) for a comprehensive
bibliography). The Cenozoic North Sea Basin is widely regarded as
The main aims of this paper are to describe the a type example of a post-rift thermal sag basin
seismic facies characteristics of the ridge-trough (Dewey, 1982; White, 1989). Rifting ceased at
structures with particular emphasis on their internal the end of the Jurassic along most of the con­
reflection configurations and external morphology, stituent grabens of the North Sea Rift (Ziegler,
and to place these structures in a palaeogeographical 1982; Bertram & Milton, 1989; Cartwright, 1991).
context. The paper concludes with some preliminary After a period of infilling of remnant depositional
ideas on possible depositional models. topography in the Early Cretaceous and early part
of the Late Cretaceous, the rift axis and rift shoulders
began a phase of regional sag-type subsidence that
GEOLOGICAL SETTING persists to the present-day (Joy, 1993). Uplift of the
hinterlands bordering the North Sea Basin com-·
A grid of regional two-dimensional seismic data menced at approximately the same time as regional
covering the central North Sea, with an average line subsidence (Watson, 1985; Rundberg, 1989), and

····
•
· ·····
·- ��
� !a····· · ·· ·· "············ · ···.:
· .
NORTH• ···············•
. -�-, .. .··········.
� --····
. ...
.......

: : :
: : :

. . . . . !. . . . .
\• • • �·· ·
15
····'· . . .
·
58°N
,

, r,,i-·::i:·''.
i •
· �·
�.. ,.. ......... t ..

. . .. . .
.
.

..
\ \ .......
', l ,..•:

\� _,
..
.....
\ ..... :.-" ;
.
\
� ·,("'.. """.. ; . : . .. . .... . ..

l<F',
. . . .. ... ..

�
.. . . .
· .
········ ········· ··•.······· ·········
• ·· . .

r--.
\ .

BlJ Study Area Distribution of ridge/trough

features
16 Quad Numbers

Fig_ 1. Location map of the study area for the detailed mapping of the ridge-trough system in the central North Sea. Also
located are the two line drawings of regional seismic lines illustrated in Fig. 2.
 Ridge structures, central North Sea 287

sediment supply changed from dominantly carbonate Pleistocene interval is dominated by a large deltaic
material during the Late Cretaceous, to dominantly system, which prograded from the eastern basin
clastic material throughout the Palaeogene and margin in Denmark, crossing the Danish Sector en
Neogene (Ziegler, 1982; Lovell, 1990). route to the central North Sea Basin (Fig. 2). This
Basin-margin deltaic systems were established large deltaic complex is referred to in this paper as
along certain sectors of the basin at different times, the Skaggerak Delta. This name was chosen because
presumably in response to changes in the location of the main axis of sediment supply has been mapped
the main centres of marginal uplift. Uplift of the along a NE-SW course running through the present­
Scottish Massif in the Late Cretaceous to Early day Skaggerak (Morgan, 1992).
Cenozoic (Watson, 1985) is generally considered to The scale and influence of the Skaggerak Delta
pre-date the main phase of uplift of Fennoscandia, on the Neogene fill of the basin is illustrated
which probably dates from the Oligo-Miocene in two schematic sections based on interpret­
(Holtedahl, 1960; Spjeldnaes, 1975). One result of ations of regional seismic lines calibrated with the
this diachroneity in uplift of the eastern and western released well-data (Fig. 2). The Skaggerak Delta
marginal areas is that the pattern of infill of the is represented on these two sections by large,
Cenozoic is asymmetric. The main progradational westward-building sigmoidal and tangential­
clinoforms built out from the western margin for the oblique clinoforms above the regional middle
Palaeocene to Oligocene interval, with sediment Miocene marker (MM). The magnitude of the clastic
mainly supplied from the Scottish Massif and the input from the east is particularly apparent in the
Shetland Platform (Parker, 1975; Rochow, 1981; way that the clinoforms cross the basin axis and
Lovell, 1990). A major reorganization of the clastic downlap or onlap the western margin of the basin.
supply into the basin occurred at approximately the These two sections also illustrate the asymmetric
beginning of the Middle Miocene, and progradation pattern of basin-fill in the Cenozoic, with contrasting
switched to a dominantly easterly source. Over large westerly and easterly sources of sediment supply in
areas of the Norwegian, Danish, and German sectors the Palaeogene and Neogene, respectively.
of the North Sea, the Middle Miocene to Late The Late Miocene to Early Pliocene ridge-trough

(a)
WSW ENE

(b)

4km
Q._
_____ _.7pm
� Late Miocene - Early Pliocene
Ridge - Trough System

Fig. 2. Line drawings of two regional profiles across the central North Sea, showing the elements of the Mesozoic rift, and
the configurations of the post-rift succession. The Late Miocene to Early Pliocene ridge-trough system is located at the
base of slope of large-scale prograding clinoforms associated with the Skaggerak Delta. MM, middle Miocene marker;
uM, upper Miocene; eP, lower Pliocene; Ku, upper Cretaceous.
288 J. Cartwright

system occupies a position in the basin axis in the although limited by the vertical seismic resolution of
two sections illustrated in Fig. 2. The ridge-trough the order of 1 0-15 m in the Neogene interval, is
system is associated with the toesets of major west­ none the less sufficient to give some preliminary
ward prograding clinoforms of the Skaggerak Delta. indications of the processes responsible for the devel­
Progradation and aggradation of the clinoforms is opment of the ridge-trough structures.
matched in part by aggradation and southwesterly Detailed interpretation and mapping of the ridge­
migration of the ridge-trough system. trough structures was carried out on a study area of
approximately 400 km2 in Quadrant 22, of the UK
Sector of the North Sea. This study area is located in
EXTERNAL MORPHOLOGY AND the central portion of the larger region in which the
INTERNAL CONFIGURATION OF ridge-trough structures are developed (Fig. 1). The
THE RIDGE-TROUGH SYSTEM area was selected for detailed study because well­
and seismic data were readily available. Mapping
was based on a seismic grid with an average line
Database and methodology
spacing of 1 km, which was sufficient to allow indi­
The descriptive parts of this paper are based largely vidual ridges and troughs to be correlated without
on interpretations of reflection seismic data. Litho­ difficulty. The seismic data is a late 1980s vintage,
logical data used in this study are restricted to cut­ and the quality and resolution of the Late Miocene
tings descriptions compiled in completion logs for to Early Pliocene interval are excellent.
over 100 released well-data in the central North Sea,
and a limited amount of petrophysical data. Unfor­
External morphology
tunately, the biostratigraphical control of many of
these released well-data is very poor, in part due to The external morphology of the ridge-trough struc­
lack of sampling and in part due to drilling methods tures is illustrated with reference to a representative
(Gradstein et at., 1992). In addition, the lack of seismic section from the study area (Fig. 3). The
velocity information means that accurate well-to­ section is oriented NE-SW, approximately orthog­
seismic calibration has not been possible. Conse­ onal to the long axes of the ridge-trough structures
quently, detailed lithostratigraphical correlations in the study area. The base of the main ridge-trough
have not been successful. The analysis is therefore system in the study area is picked at a low-frequency
based on a seismic-stratigraphical approach, which reflection that appears to mark the upper limit of a

Fig. 3. Seismic profile from the study area showing the cross-sectional geometry of the ridge-trough system. Note the
diachroneity evident in the abandonment of active ridges from right to left. Also note the asymmetry of the post­
abandonment infill of the troughs. Section oriented SW-NE. Datum at top of section is 600ms TWTT (two-way travel
time).
 Ridge structures, central North Sea 289

precursor set of ridges and troughs (reflection A on

Internal reflection configurations
Fig. 3). The upper boundary is picked wherever a
clear onlap can be seen in a trough (e.g. reflection B The typical internal configurations of the ridges and
on Fig. 3). The ridges have an external form that is troughs are illustrated in Fig. 4, a seismic section
best described as mounded (Mitchum et at., 1977). parallel to that shown in Fig. 3, but reproduced at
The intervening troughs have a channel-like profile. an enlarged scale to enhance the resolution. The
The ridge-trough structures have crest-to-crest ridges are composed of low- to moderate-amplitude
spacing of between 1 and 3 km, and vary from reflections with moderate continuity. The lowermost
almost perfectly symmetric to slightly asymmetric in reflections are concordant with the base, but there
form. The crest-to-trough amplitude of the structures is increasing convexity exhibited by successively
defined using the upper and lower bounding reflec­ shallower reflections, such that the maximum relief
tions varies between 1 00 and 200m. The ridges have between ridges and adjacent troughs is at the upper
broad crests with fairly steep flanks, with dips of up bounding reflection (e.g. at reflection A).
to 7° (uncorrected for compaction). Numerous reflection terminations can be observed
Throughout the study area there is clear evidence within the ridges illustrated in Fig. 4, particularly
that the system of ridges and troughs is diachronous towards the margins of individual ridges. Some of
from northeast to southwest. This diachroneity can these reflection terminations may be indicative of
be seen, for example, in Fig. 3 in the contrasting genuine stratal truncation, pinch-out or downlap.
topography and external form of the reflections However, considering the probable distortive effect
comprising the ridge-trough system across the on ray-paths of the troughs, and the possibility of
section. The relief on the shallow, high amplitude multiples being generated in the overlying high­
reflections that can be seen draping the ridges on the amplitude reflections, it is possible that many of the
northeastern side of the section (e.g. reflection C) reflection terminations are artefacts produced by
increases to the southwestern side of the section. interference or tuning. Genuine downlap is observed
Reflection C changes from a post-abandonment only rarely within the ridges, and consequently the
drape of the ridge in the northeast, into an internal most common diagnostic feature used to define a
reflection of the ridge-trough structure in the constructional sedimentary mound (Mitchum, 1985)
southwest. This indicates a younging of the active cannot be confidently applied to the ridges illustrated
growth of the ridges to the southwest. in Figs 3 & 4.

1: . . .... . 1. �Ni:-
,,-"''IHI•I�����

0 1km

Fig. 4. Enlarged scale seismic section oriented at right angles to the axes of the ridge-trough structures showing truncation
of events and downlap at the margins of the ridges and the aggradational configurations of stratal reflections within both the
ridges and the troughs. Datum at top of section is 600 ms TWIT.
290 J. Cartwright

The troughs imaged on the section in Fig. 4 are oriented seismic section (Fig. 5), located several
characterized by an almost concordant set of high­ kilometres to the northwest of Fig. 4. Figure 5 shows
amplitude reflections whose lower contact is difficult an example of a trough that has a clearly recognizable
to define. The overall configuration of the troughs is basal reflection, with a classical channel profile
almost equivalent to an upside-down version of the (trough A). Erosional truncation can be seen on
ridges. No obvious incision can be seen at the base both margins of the trough, and there are sufficient
of the lower reflections defining the troughs, and genuine reflection terminations to define the position
direct correlation between reflections from the of the basal erosion surface. This basal surface
troughs to adjacent ridges is hampered by changes in (reflection X) passes laterally into the neighbouring
reflection character and continuity. However, at ridges, and can be correlated with the erosional
least some of the low-amplitude internal reflections bases of the two neighbouring troughs (points B
of the ridges pass directly into the concave, high­ and C). Above the erosional base in trough A, the
amplitude reflections of the troughs. It is possible overlying reflection is concordant with reflection X,
that the contrast in reflection amplitude between and above this there is a clear onlap of the trough
reflections in the troughs and ridges is due to lateral margin by three subhorizontal reflections. The evol­
variations in lithofacies. For example, some form of ution implied by these reflection configurations is:
lag deposit in the troughs could explain the higher (i) incision, (ii) drape, and (iii) onlap fill. Taking the
amplitudes. However, it is also possible that the truncation and onlap relationships together, trough
variation is due to tuning and the concave shape A can only reasonably be interpreted as an erosional
of the reflectors. At the upper levels of the ridge­ feature. A similar pattern of erosion, drape and
trough system, the relief across the troughs is onlap can be seen on the other two troughs in Fig. 5.
gradually diminished by the development of onlap The variability in the detailed configurations of
fill (e.g. Fig. 4, point X). Thus, although the troughs the three troughs seen on the section in Fig. 5 is
seen in Fig. 4 have the general morphology sugges­ typical of that observed throughout the study area.
tive of some type of channel, there is little conclusive Incision is invariably observed along at least part of
evidence that the troughs on this section formed by the axial extent of every trough. However, the
an erosive process. degree of incision, and the clarity with which the
A clearer indication of a possible erosional origin incision is expressed, varies widely, both from trough
of the troughs can be seen on another NE - SW to trough, and along any individual trough. The

Fig. 5. Seismic section oriented at right angles to long axes of ridge-trough structures, showing contrast in reflection
configurations of three troughs, and evidence of erosional truncation at their bases . Datum at top of section is
600ms TWTI.
 Ridge structures, central North Sea 291

troughs in Fig. 4 are typical of those in which basal puzzling aspects of the ridge- trough system illus­
incision is difficult to define from the cross-sectional trated in this figure is that there is a combination of
reflection configurations, whereas trough A in Fig. 5 features that are diagnostic of both constructive and
is a type example of a trough whose seismic destructive process elements. Included for com­
expression most clearly conforms to that of an ideal parison in Fig. 7 are two idealized line drawings
incised profile. through a definitive constructional mound and an
Figure 6 is a comparison of two representative incised channel separating interchannel ridges based
seismic sections located along the axes (NW-SE) of on criteria outlined in Mitchum et al. ( 1977) and
a ridge and trough pair. The internal configurations Mitchum (1985).
of the ridge and trough are comparable, with a The incision observed at the base of at least a
consistent parallel form with respect to the lower portion of every trough clearly points to the activity
and upper bounding reflections. There is no evidence of an erosive current regime for at least part of the
of any significant thickness variation along the axes time represented by the growth of the ridge-trough
of either the· ridges or the troughs, and no evi­ structures. In contrast, the increasingly convex­
dence of internal downlap, progradation, onlap or upwards configurations of the reflections within the
erosional truncation in the along-axis direction of ridges is interpreted as evidence that the ridge struc­
either the ridge or the trough. ture is at least partly constructional. The common
development of drape configurations in both ridges
and troughs indicates that the constructional and
Interpretation
erosive components of the system alternated with
The reflection configurations of the ridge-trough a dominantly pelagic or hemi-pelagic mode of
system observed in cross-section are summarized in deposition.
a schematic line drawing in Fig. 7. One of the most Applying the criteria outlined by Mitchum (1985)

NW
SE

�:: :;:�: :::--�:::=�::::::;:;::_,,>M,�I" :·' ·�,;,::::�:;:;;;;;2::1:=:==���-��-�....;_,�,

'II''.!I!INI'"'I�-1!::!.;:;¥:· -;;;'P!!'P:O "'" '"'"'t• I
tiMrll:....... . ..Ill

0 1km
I I

Fig. 6. Comparison of two longitudinal seismic sections along a ridge (A) and adjacent trough (B). Note the absence of
erosional truncation, onlap, downlap or progradation in the internal reflections of either the ridge or the trough.
292 J. Cartwright

- --- ---

BASE MIDDLE
0 1 km MIOCENE MARKER

EROSIONAL CHANNEL INTER-MOUND TROUGH

(NON-EROSIONAL)

Fig. 7. Summary diagram showing the main stratigraphical features of the ridge-trough system observed from the study
area. Idealized sections across an erosional channel and a pair of constructional mounds shown beneath main figure for
comparison. Note that the ridge-trough structures have affinities with both these contrasting idealized cases. Note also
that with low-resolution seismic data it may be difficult to distinguish the two cases, particularly If multiples obscure the true
stratal configurations at the ridge margins.

for the recognition of different types of construc­ the relatively closely-spaced seismic grid ( 1-2 km)
tional mounds, it is difficult to make a conclusive allowed correlation of individual ridges and troughs.
argument that the ridges are primary constructional Structural contour mapping of specific reflections
features. Neither is it reasonable to interpret the within the Late Mioc�ne to Early Pliocene interval
ridges as purely remnant relief produced by erosion was attempted initially, but correlating horizons
along the troughs. Without the benefit of detailed across the grid proved to be too difficult, mainly
lithological calibration of the seismic data it is due to loss of reflection continuity in the troughs.
impossible to determine the extent to which erosive Consequently it was decided to map the individual
and depositional processes alternated or dominated trough axes because these are narrower than
throughout the evolution of the ridge-trough the ridges and are therefore less likely to be
system. miscorrelated.
The map of trough axes in the study area is
presented in Fig. 8. The most important features of
MAPPING THE this map are all well constrained by the seismic grid.
RIDGE-TROUGH SYSTEM These are:
1 the orientation of the trough axes is confined to a
Mapping of the ridge-trough system was under­ narrow range from 320°N to 340°N;
taken in order to determine the spatial relationship 2 the troughs are linear to gently sinuous, and die
between ridges and troughs and to determine their out abruptly in both directions along their axes;
orientation relative to the major physiographic 3 there is no interconnection between individual
features of the basin. Detailed mapping of indi­ troughs - they neither bifurcate nor anastomose in
vidual ridges and troughs was restricted to the small either direction;
study area (Fig. 1) because it was only there that 4 the length of those troughs that die out along
 Ridge structures, central North Sea 293

Seismic lines
N
··············

+
-�- Axis of Trough

• --·• Termination of Trough

-<a£• Shallowing of Depression
(decreasing relief in
direction shown)

0 2

Fig. 8. Map of the trough axes in the study area. Also shown is the seismic grid used for the mapping. Note the
discontinuous array of trough axes in both an upslope and downslope direction (regional palaeoslope to the northwest).

their axes in both a northwesterly and southeasterly Fig. 5) clearly indicates that currents strong enough
direction within the confines of the mapped area to erode muddy sediments flowed along the trough
varies from 2 to 10 km - one trough is over 20 km axes. Because it has not been possible to correlate
long, and continues outside the limits of the seismic the erosional bases of individual troughs across the
grid; study area, the extent to which erosional activity in
5 the trough axes are fairly consistently spaced over the troughs was synchronous is not known. For
the area, with the intertrough spacing varying example, did the basal incision along the troughs
between 1 and 3 km. take place at approximately the same time through­
One of the most surprising aspects of Fig. 8 is out a narrow zone, a broad zone, or was it confined
the lack of any interconnection between adjacent to a single trough at any one time?
troughs. The incision observed in the troughs (e.g. The overall diachroneity in the growth of the
294 J. Cartwright

system from northeast to southwest evident on the (A-B of Fig. 8) with terminations in both directions
seismic data (e.g. Fig. 3) and on the regional sections along-axis are presented in Fig. 9. This diagram
(Fig. 2) suggests that only a few of the trough axes illustrates the subtle changes in internal configuration
indicated on Fig. 8 were active at any one time. This of the trough, with variable depth of incision and a
interpretation is based on the assumption that, general decrease in relief towards both ends of the
because the depositional unit containing the ridge­ trough. At the limit of the trough, only a vague
trough structures is diachronous as a whole, then it indication of an erosive feature can be discerned,
is probable that the time of initial incision of troughs and beyond this point, the trace of the trough axis
was also diachronous. If this is correct, then Fig. 8 projects directly into a ridge structure.
should not be viewed as a palaeogeographical
reconstruction, but as an overall progression estab­
lished during several million years of deposition and
PALAEOGEOGRAPHICAL SETTING OF
erosion.
THE RIDGE-TROUGH STRUCTURES
Another important feature of the organization
of the troughs is the observation that the troughs
die out in both a northwesterly and southeasterly A palaeogeographical reconstruction of the Early
direction. Of the 20 troughs partly or wholly mapped Pliocene North Sea Basin is presented in Fig. 10.
within the study area, 13 die out along their axes The reconstruction is based on a basin-wide cor­
at one end of the trough within the mapped area, relation of a seismic horizon marking the end of the
and five die out at both ends along the trough axis period in which the ridge-trough system formed .
(Fig. 8). A series of line drawing cross-sections This horizon is dated as approximately 4 Ma, i.e .
constructed at intervals along one of the troughs Early Pliocene. The depositional shoreline break

I
I : UNE1

� �
13------------ ---------- -------------0 LINE1 l
I Be i i
I f i i
G----------L·-·---r·+------�------o UNE 2

\I I ! !
:
I :
I I : :
! 1 i i

� = ..

. . .
it ti ��::
/ i
I :

..
.
:I
iI
• I :
/ :t
! u
---
1
: �
1
G- --- -- - --- ----- ------- ----
I ---
�
{J UNE 5
:
�
,:
; X I ; 1;
i c�: I /i Ii

JJJ< I
Fig. 9. Line drawings of seismic
· · sections oriented SW-NE along the
.... .
trough A-B (location in Fig. 8) ,
showing the change in reflection
o • , p �•• configurations of the trough along
./ii \ i its axis. Note that the trough dies
/ t
G ··- ----1- ---·-----··-···--·-··1-- d UNE7 out towards the north-northwest
/ \i \ and merges into a neighbouring
I I
ridge structure.
 Ridge structures, central North Sea 295

was defined at the break of slope between topsets identified by mapping regions dominated by pro­
and foresets for clinoforms along the progradational gradational clinoforms. The palaeobathymetry of
sections of the basin margin (Posamentier & Vail, the basin was estimated by measuring the relief on
1988). Progradational shorelines were differentiated clinoforms using topset surfaces as mean sea-level
from retrogradational or static shorelines by defining projection datums.
the extent of marine onlap for the Late Miocene to Figure 10 shows that the North Sea Basin was a
Early Pliocene interval. Major depocentres were narrow marine gulf or seaway in the Early Pliocene,

Actively prograding shoreline at Position of depositional shoreline

4ma \9ma break for Skaggerak Delta at
Passively onlapped shoreline at \ various times
4ma I
Progradation route of Skaggerak
Bathymetric contour in metres
Delta
I
�istribution of Ridge-Trough System
�bandoned at 4ma Other major distributary systems
Active at 4ma

t
.
�·
,,
,:
,,
i :
i :
...... ................\� ........

Fig. 10. Palaeogeographical reconstruction for the Late Miocene to Early Pliocene. The North Sea is shown as a semi­
enclosed marine seaway. Marine connection to the Norwegian and Greenland Sea was open via a narrow strain between
two deltas. The marine connection to the south is uncertain (see text). The position of the depositional shoreline break of
the Skaggerak Delta is shown for various time intervals, and illustrates the rapid advance of this major depositional system.
296 J. Cartwright

with a NNE-SSW trending elliptical central basin In summary, the ridge-trough system in the
narrowing to the north and south. The bathymetry Early Pliocene occupied a base-of-slope-basin-fioo:r
estimated from clinoform relief indicates that maxi­ position, in water depths of 300-400 m, along the
mum water depths of 300-400 m were developed in southern portion of a narrow marine basin that was
a basin-plain region centred on Quadrant 22 of the being rapidly infilled by a major delta prograding
UK Sector (ct. Figs 1 & 10). Palaeobathymetric east to west and either narrowing or closing any
constraints from the limited sampling of microfauna possible marine connection to the south.
in wells in the central North Sea are imprecise, but
are broadly consistent with the magnitude of water
depth interpreted from the clinoform geometry. DISCUSSION
Water depths derived from an analysis of subsidence
of the central North Sea are in close agreement with Any explanation of the genesis of the ridge-trough
the values presented here (Joy, 1993). system needs to take account of the following key
The region of active ridge-trough structures is observations.
shown to be confined to the southern end of the 1 The ridge-trough system developed during a
basin plain in the Early Pliocene. The full extent of short interval of time (c. 1-2 Myr) spanning the
the ridge-trough structures (active and abandoned) Late Miocene to Early Pliocene. No remotely similar
is also indicated in Fig. 10, in order to show the features have been documented from comparable
overall southwestward migration of the ridges and base-of-slope depositional systems in the North Sea
troughs from the Late Miocene to the Early Pliocene. at any other period in the Cenozoic.
The most probable marine connection into the 2 The dimensions of the ridges are large in com­
North Sea was at the northern end of the basin, with parison to most constructional bedforms. They
the Norwegian-Greenland Sea (Ziegler, 1982). are of the order of 10 km long, 1-3 km wide, and
Whether there was an additional connection to the 100-200 m thick.
Bay of Biscay to the south is not clear. Neogene 3 The orientation of the long axes of the ridges and
strata are absent throughout much of the south­ troughs range from 320°N to 340°N.
western part of the North Sea (south of this study 4 Ridges and troughs exhibit a variety of construe-·
area), and a major erosional unconformity is tiona) and erosional characteristics, from stratal
developed near the base of the Late Pleistocene thickening towards the ridge axes (constructional),
(Cameron et al., 1987). Hence it has not been to truncation of ridge flanks and incision along
possible to extend the palaeogeographical recon­ trough axes (erosional).
struction into the southern North Sea and clarify the 5 There is no evidence of lateral accretion of indi­
nature of any possible marine connection. vidual ridges. The internal reflection configurations
Four main depocentres prograded into the North indicate that ridges grew by aggradation: abandon-­
Sea in the Early Pliocene. The courses of distributary ment was followed by a migration of active growth
systems supplying these depocentres are sche­ to new ridges to the southwest (of the inactive
matically illustrated in Fig. 10, to give some indi­ ridges).
cation of probable source areas, although it should 6 The troughs do not interconnect and, where
be noted that there is little direct evidence concerning mapped to their full extent, die out along both axial
the exact positions of the distributary systems. In the directions.
extreme north of the basin, two large deltaic systems 7 The zone of ridge growth migrated from northeast
prograded towards each other simultaneously, to southwest with time.
narrowing the marine connection to the Norwegian­
Greenland Sea. In the south (south of 57°30'), clino­
Depositional models
forms of the Skaggerak Delta dominated the infilling
of the basin. The rapid westward progradation of
Mudwaves formed by the action of bottom currents
the Skaggerak Delta is indicated on Fig. 10 by the
mapped positions of the depositional shoreline break Large-scale, constructional muddy waveforms
through the Late Miocene to Early Pliocene interval. formed by the long-term action of bottom currents
A relatively minor depocentre was confined to an have been described only from slope and abyssal­
embayment in the otherwise passively onlapped plain settings in oceanic basins (see Pickering et al.
southwestern margin of the basin. (1989) for a comprehensive review), but there is no
 Ridge structures, central North Sea 297

reason why similar structures should not be found in the two oceanic basins), it is difficult to reconcile
locations other than oceanic basins, provided that this mechanism with the often significant incision
the necessary current regime was operative for observed along many of the troughs. In particular, it
a sufficient period. The dimensions and internal is difficult to see how bottom currents could have
stratification of the ridge-trough system are com­ produced the incision of tens of metres of fine­
pared with a compilation of line drawings of mud­ grained sediments along the troughs in a north­
waves deposited by bottom currents in various northwest orientation (i.e. transverse to the long
oceanic localities (Fig. 11). The relief of these mud­ axis of the basin and hence the likely flow route for
wave examples is of the order of 10-20 m, i.e. bottom currents), and on a longer term lead to a
considerably less than the c. 100-200 m relief of the general southwesterly migration of the entire ridge­
ridges in the central North Sea. However, the shape, trough system.
internal reflection configurations and spacing of some In conclusion, the interpretation of the ridges as
of the mudwaves are comparable with the ridge­ some form of bottom-current deposit cannot be
trough structures. discounted, particularly in view of the uncertainties
Better grounds for comparison can be made with regarding a southerly marine connection, but the
large-scale bedforms associated with a sediment drift observation of incision along the troughs argues
in the Northeast Atlantic, described by Kidd & Hill against this depositional model.
(1986). The mudwaves developed on the Feni and
Gardar Drifts were studied by these authors using a
Pro-delta current regime
combination of sparker and GLORIA sidescan sonar
data. The resulting map of the mudwaves (Fig. 12) An alternative model for the development of
shows a pattern with some similarity in scale and the ridge-trough system is suggested by two
organization to that exhibited by the ridges and observations:
troughs from the central North Sea study area, 1 the approximately orthogonal relationship be­
although there is less regularity to the mudwave tween the axes of the ridges and troughs (NNW­
organization than to the ridge-trough system SSE to NW-SE) and the strike of the nearest
(Fig. 8). An interesting feature of the map shown in depositional shoreline break (NE-SW);
Fig. 12 is that the mudwave crests close at both ends 2 the skewed distribution of the ridge-trough struc­
of their long axes. This is in contrast to the ridge­ tures to that sector of the basin plain immediately
trough system where the mapping showed that it is adjacent to the toe of the Skaggerak Delta, which
the troughs that terminate in both directions along­ was by far the most significant Mio-Pliocene pro­
axis. It is not clear whether this attribute is diagnostic gradational system in the North Sea.
of mode of origin. In common with the ridges in the The evidence of erosional truncation and incision
North Sea, the mudwaves on the Feni and Gardar in the troughs (Fig. 5) implies that currents flowing
Drifts show no evidence for lateral migration on an along the trough axes were of sufficient strength to
individual basis. The long axes of the mudwaves are result in the cumulative erosion of tens of metres of
thought to be oriented obliquely to the bottom­ muddy sediment. If the troughs were foci for strong
current flow direction. currents flowing along the trough axes, it is reason­
The limited basis for comparison means that it is able to infer that their origin was directly upslope,
hard to make an unequivocal case that the ridge­ either at the delta front or on the upper reaches of
trough system in the central North Sea was the result the pro-delta. Because the nearest delta front to the
of bottom-current activity. The similarity with the ridge-trough system was along the southeastern
published descriptions of mudwaves associated with margin of the basin, it is reasonable to argue that
sediment drifts is sufficient to at least consider the ridge-trough system developed in response
bottom currents as a possible driving mechanism for to a current regime associated with the rapidly
the development of the ridge-trough system. prograding Skaggerak Delta.
However, although a bottom-current mechanism If some form of pro-delta current regime was
may offer a reasonable explanation for both the responsible for the development of the ridge-trough
distribution of the ridge-trough system in the basin system, several questions need to be addressed.
axis· and the long-term migration of the zone of 1 What was the specific nature and flow regime of
active ridge-trough structures towards the south­ the current system?
west (i.e. along the likeliest route for currents linking 2 How and where did the currents originate?
298 J. Cartwright

0 1km

Fig. 1 1 . Line drawings of high-resolution shallow seismic images of mud waves from the Central and Southern Atlantic
showing typical dimensions and internal configurations. Taken from Pickering eta/. (1989). Line drawing of ridge-trough
structures displayed at the bottom for comparison.
 Ridge structures, central North Sea 299

N
0 KILOMETRES 5

1
r-------7'"--.:--� so·N

Fig. 12. Detailed bathymetric map

over a field of sediment waves so•
located on the Feni Drift, Rockall
Trough, North Atlantic. Note that
the sediment wave crests (stippled)
are discontinuous. Taken from Kidd
& Hill (1986). 30'

3 How did these currents result in the stratal patterns the Late Miocene to Early Pliocene, given that the
observed on the seismic data? Skaggerak Delta was first established in the Early
4 In the absence of any evidence linking similar Eocene, and continued prograding until the Early
ridge-trough structures to the other deltas along Pleistocene?
the basin margin, why was this current system 6 Why did the locus of current activity migrate
uniquely developed in connection with the Skaggerak towards the southwest from the Late Miocene to the
Delta? Early Pliocene?
5 Why was this current regime operative only during The postulated downslope current system must
300 J. Cartwright

have been organized in linear, confined flow paths spacing of these slope gullies is closely comparable
with lateral dimensions of 1-2 km in order to to those exhibited by the troughs in the central
produce the regular, isolated and linear troughs North Sea, but the erosional relief of the Scotian
mapped in the study area. To explain the lack of gullies is only of the order of 10 m. It is possible,
interconnection between the troughs and the obser­ however, that even shallow erosional gullies could
vation that the troughs die out at both ends of their provide the necessary confinement and guide the
axes (Figs 8 & 9), it is necessary to invoke downslope flow paths from the upper reaches of the pro-delta to
currents that were either of a finite extent in the flow the toe region. Minor erosional features have been
direction, or currents whose strength varied along observed on the pro-deltaic foresets of the Skaggerak
the flow direction. Both of these conditions would Delta on regional seismic data (Morgan, 1992), but
be expected to apply to a downslope pro-delta the wide spacing of the seismic grid available at this
current system. Any downslope currents would reach time makes any direct correlation of these features
their maximum velocity at the base-of-slope, and with the ridges and troughs in Quadrant 22 extremely
would gradually lose velocity along the horizontal tenuous.
basin floor. The finite longitudinal extent of the A direct relationship between the distributary
incision observed along the troughs could, therefore, channels on the delta front and confined axes of flow
be interpreted as a result of the current exceeding a at the base of slope is appealing in that it may
critical erosive velocity threshold for the muddy explain the overall southwesterly migration of the
sediments as the currents accelerated downslope in a ridge-trough system from the Late Miocene to the
northwesterly direction. Then, as the power of the Early Pliocene. If downslope currents originated at
current decreased as flow proceeded along the basin the outlets of the major distributary channels on the
floor, the degree of incision would also decrease delta front, possibly during one of the glacially forced
until the trough was no longer recognizable as such relative lowstands of sea-level (Haq et al., 1987),
(Fig. 9). then the locus of trough development would prob­
One of the problems with this pro-delta current ably have been confined to the sector of the basin
model is the large distance from the likely point of floor directly downslope from the major channels. If
origin of the currents at the head of. the slope, to for any reason the channels avulsed at the delta
the toe region where the ridge-trough structures front, then the locus of active ridge-trough growth
formed. A question that cannot be resolved as yet is would be expected to migrate accordingly (Fig. 13B) .
how sufficient flow confinement would have been A depositional model linking the evolution of the
maintained from the delta front, where the currents Skaggerak Delta directly to the development of the
may have originated, for example, as some form of ridge-trough system also provides a satisfactory
turbid underflow (Bates, 1953), over a distance of explanation for the aggradational nature of the
tens of kilometres to the northwest at the base of stratal patterns interpreted on the seismic data
slope, where incision commenced. (Fig. 7). The Late Miocene to Early Pliocene is
The extent to which the Coriolis effect would have known to have been a period of rapid basin sub­
deflected downslope flowing currents originating at sidence (Kooi & Cloetingh, 1989; Joy, 1993), and
or close to the delta front is also a problem with this this led to increased aggradation of the basin-margin
depositional model. If a modern system, such as the deltas (Morgan, 1992). The aggradational character­
Huanghe (Wright et al., 1990), represents a reason­ istics observed for the troughs and ridges can be
able analogue, then along-slope deflection should be viewed as a response to the aggradation of the
expected (Hill, 1984). It is also difficult to envisage deltaic depositional system as a whole, including the
a mechanism whereby the downslope flow would foreset and bottomset depositional units, and any
maintain its confined flow pattern once it began to downslope current system originating at the delta
decelerate at the lower break of slope. One method front.
of providing the necessary flow confinement would The pro-delta current mechanism can thus explain
be if there was a pro-delta channel or gully system many of the observations documented in this paper,
linking the distributary channels at the delta front but a major unresolved question is why the ridge-­
with the troughs at the base of slope (Fig. 13A). trough structures were restricted to the Late Miocene
Linear gully systems have been observed and to Early Pliocene in particular? What was so unique
mapped on the present-day Scotian slope of eastern about this period that either a pro-deltaic downslope
Canada (Piper & Sparkes, 1987). The scale and current system or a bottom-current system should
 Ridge structures, central North Sea 301

A NE
Pro-delta "' -

Topset
} aggradation
Trough

'
'�
SW

ll----.,--
Bottomset

500m --,
� C u r rent flow path

0 10 20km

Approximate scale

Palaeogeography time t1 Palaeogeography time t2

B

N
'

Avulsion of major
distributary
c h a n nel

Progradation of
depositional
shoreline break

Fig. 13. (A) Evolutionary reconstruction of the development of the ridge-trough system based on a pro-delta current
regime. Flow confinement from the delta front is provided by a system of slope gullies. Maximum flow velocity is reached at
the base of slope, and the trough incision begins at this position. Deceleration along the basin floor results in the
elimination of the troughs in a north-northwest direction. (B) Migration of the ridge-trough system is achieved by avulsion
and channel switching at the delta front.

have been operative for this specific period, and not Pliocene (Lidmar-Bergstrom, 1982) may be part of
throughout the entire Cenozoic? The major climatic the solution. The resolution of this question will
change that took place with the onset of the glacial­ require considerable additional research, not least
interglacial cycles in the Late Miocene to Early of which would be a comprehensive study of the
302 J. Cartwright

morphological evolution and palaeogeographical necessary confinement, but much additional map­
setting of the Skaggerak Delta throughout the ping is required to confirm this suggestion.
Cenozoic.
The impetus for additional work on the commer­
cially unprospective Neogene interval of the North ACKNOWLEDGEMENTS
Sea Basin may be provided by the enigmatic linear ·

sandbodies of the Eocene in the North Sea. The I would like to thank Fina Exploration Ltd for
troughs described in this paper may be examples financial support and Fina and partners for access to
of large sedimentary features that resulted from a well and seismic data. Thanks also to colleagues at
process involving the transport and confinement of Fina for valuable discussions, namely Rod Laver,
reservoir-prone units from the delta-top into deep­ Dave Fassom, Mick Cope and Joe Staffurth. The
water environments. If the troughs were a locus for manuscript was greatly improved as a result of the
sand transport and deposition, then the resultant many suggestions made by the reviewers, Chris
reservoir geometry would be a linear body compar­ Baldwin and Phil Hill, and the Haroldesque editorial
able in scale and thickness to reservoir sands devel­ skill of Guy Plint. Thanks also to Howard Johnson
oped on the Eocene slope and basin floor. and Dick Selley for suggestions for improving the
manuscript. Finally, my sincerest thanks to Harold
for putting me on the right track over the last 15
CONCLUSIONS years (derailments are my responsibility !). Tony
Brown and Jubrail produced the figures.
1 The ridge-trough structures of the central North
Sea are unique features in the context of the
Cenozoic stratigraphy of the North Sea, with REFERENCES
apparently no direct analogues in the published
literature. BATES, C.C. ( 1953) Rational theory of delta formation.
2 The linearity of the structures and the regularity Bull. Am. Assoc. petrol. Geol. , 37, 2 1 19-2 162.
BERTRAM, G.T. & MILTON, N.J. ( 1989) Reconstructing
of their spacing suggest that they either formed in
basin evolution from sedimentary thickness; the import­
a current regime that was highly oblique or even ance of palaeobathymetric control, with reference to the
transverse to their long axes, or in one that was North Sea. Basin Res. , I, 247-257.
parallel to their long axes. CAMERON , T.D.J. , STOKER, M.S. & LoNG , D. ( 1987) The

3 Two contrasting current regimes are regarded as history of Quaternary sedimentation in the UK sector
of the North Sea Basin. I. geol. Soc. London, 144,
the basis for a genetic model: (i) bottom currents 43 -58.
flowing along the NNE-SSW basin axis, or (ii) pro­ CARTWRIGHT, J.A. ( 199 1 ) The kinematic evolution of the
deltaic currents flowing downslope in a NW-SE Coffee Soil Fault. In: The Geometry of Normal Faults
direction from the delta front of the actively pro­ (Eds Roberts, A.M. , Yielding, G . & Freeman, B . ) ,
Spec. Pub!. geol. Soc. London, No. 56, pp. 29-40.
grading Skaggerak Delta.
Geological Society of London, Bath.
4 Although both current regimes can provide satis­ DEWEY, J.F. ( 1982) Plate tectonics and the evolution of the
factory explanations of the dimensions and mor­ British Isles. I. geol. Soc. London, 139, 371 -412.
phology of the ridge-trough structures, the often GRADSTEIN , F.M . , KRISTIANSEN, l . L. , LOEMO, L . &
KAMINSKI, M. ( 1992) Cenozoic foraminiferal and dino­
pronounced incision observed in the troughs is dif­
flagellate cyst biostratigraphy of the central North Sea.
ficult to reconcile with bottom currents flowing Micropalaeontology , 38, 101 - 137 .
obliquely across the long axes of the troughs. The HAQ, B.U. , HARDENBOL, J. & VAIL, P. R . ( 1 987) Chron­
incision is compatible with currents flowing along ology of fluctuating sea levels since the Triassic. Science,
the trough axes. Hence the preferred model is the 235, 1 156- 1 167.
HILL, P.R. ( 1984) Facies and sequence analysis of Nova
pro-delta current regime.
Scotian slope muds: turbidite vs 'hemipelagic' deposition.
5 A weakness in the pro-delta current model is the In: Fine-grained Sediments: Deep Water Processes and
lack of a conclusive explanation as to how the Facies (Eds Stow, D. A . V. & Piper, D.J.W. ) , Spec .
necessary confinement in the current system was Pub!. geol. Soc. London, No. 1 5 , pp. 3 1 1 -318. Geo­
initiated and maintained down the slope and across logical Society of London, Bath.
HoLTEDAHL, 0. ( 1960) On supposed marginal faults and
the basin floor so as to produce the linearity observed the oblique uplift of the land mass in Cenozoic time.
for the troughs. Slope gullies observed on foresets Norges geol. Unders. , 208, 35 1 -357.
of the Skaggerak Delta may have provided the JoY, A.M. ( 1993) Post-rift subsidence of the North Sea
 Ridge structures, central North Sea 303

Basin. Unpublished PhD thesis, University of London, PICKERING, K.T . , HISCOTr, R.N. & H EIN , F.J. ( 1989) Deep
248 pp. Marine Environments, Clastic Sedimentation and
Kmo, R.B. & HI LL, P.R. ( 1 986) Sedimentation on mid­ Tectonics . Unwin & Hyman, London, 416 pp.
ocean sediment drifts. In: North Atlantic Palaeocean­ PIPER, D.J.W. & SPARKES, R. ( 1987) Proglacial sediment
ography (Eds Summerhayes, C.P. & Shackleton, N.J.), instability features on the Scotian slope at 63 degrees W.
Spec. Publ. geol. Soc. London, No. 2 1 , pp. 87 - 102 . Mar. Geol. , 76, 1 5 - 3 1 .
Geological Society o f London, Bath. PoSAMENTIER, H.W. & VAI L , P.R. ( 1988) Eustatic controls
KoOI, H. & CLOETINGH, S. ( 1989) Intra-plate stresses and on clastic deposition I I : sequence and system tracts
the stratigraphic evolution of the North Sea Central models. In: Sea-level Changes: an Integrated Approach
Graben. In: Extensional Tectonics and Stratigraphy (Eds Wilgus, C.A. , Hastings, B.A. , Kendall, C. G.St
of the North A tlantic Margins (Eds Tankard, A.J. & C. , Posamentier, H.W. , Ross, C.A. & Van Wagoner,
Balkwill, H.), Mem. Am. Assoc. petrol. Geol. , Tulsa, J.C.), Spec. Publ. Soc . econ. Paleontol. Mineral. , Tulsa,
46, 541 -558. 42, 125 - 154.
LIDMAR-BERGSTRO M , K. ( 1982) Pre-Quaternary geo­ RocHow, K.A. ( 198 1) Seismic stratigraphy of the North
morphological evolution in southern Fennoscandia. Sver. Sea 'Palaeocene' deposits. In: Petroleum Geology of the
geol. Unders. Ser. C. , 785, Arsbok 75 , 6. Continental Shelfof North- West Europe (Eds Illing, L.V.
LovELL, J.P.B. ( 1990) Cenozoic. In: An Introduction to the & Hobson, G. D.) , pp. 255-266. Institute of Petroleum,
Petroleum Geology ofthe North Sea (Ed. Glennie, K.W), London.
pp. 273-293. Blackwell Scientific Publications, Oxford. RuNDBERG, Y. ( 1989) Tertiary sedimentary history and
MITCHU M , R.M. ( 1985) Seismic stratigraphic expression of basin evolution of the Norwegian North Sea: an integrated
submarine fans. In: Seismic Stratigraphy, II: an Integrated approach . Dr. Ing. thesis, University of Trondheim.
Approach to Hydrocarbon Exploration (Eds Berg, O.R. SPJELDNAES, N. ( 1975) Palaeogeography and facies dis­
& Woolverton, D.G . ), Mem. Am. Assoc. petrol. Geol. , tribution in the Tertiary of Denmark and surrounding
Tulsa, 39 , 1 17 - 138. areas. Norges Geol. Unders. , 316, 289- 3 1 1 .
MrrcH U M , R.M. , VAIL, P.R. & SANGREE, J.B. ( 1 977) Strati­ WATSON , 1 . ( 1985) Northern Scotland as an Atlantic-North
graphic interpretation of seismic reflection patterns in Sea divide. J. geol. Soc. London , 142, 22 1 -243.
depositional sequences. In: Seismic Stratigraphy ­ WHITE, N.J. ( 1989) Nature of lithospheric extension in the
Applications to Hydrocarbon Exploration (Ed. Payton, North Sea. Geology, 17, 1 1 1 - 1 14.
C.E.), Mem. Am. Assoc. petrol. Geol. , Tulsa, 26, WRJGHT, L . D. , WISEMAN, W .J . , YAN G , Z.S. , BORHOLD,
107 - 134. B . D . , KELLER, G.H. , PRJOR, D.B. eta!. ( 1990) Processes
MoRGAN, C. ( 1992) The advance of the Baltic Delta. of marine dispersal and deposition of suspended silts off
Unpublished MSc thesis, University of London, 86 pp. the modern Huanghe (Yellow River). Continent. Shelf
PARKER, 1 . R. ( 1975) Lower Tertiary sand development Res. , 10, 1 - 40 .
in the central North Sea. In: Petroleum and the ZIEGLER, P. A. ( 1982) Geological A tlas of Western and
Continental Shelf of N. W. Europe (Ed. Woodland, Central Europe . Shell International Petroleum BY, The
A.W.), pp. 447-452. Applied Science Publishers, Hague, 130 pp.
London.
Spec. Pubis int. Ass. Sediment. (1995) 22, 305-327

Millstone Grit cyclicity revisited, II:

sequence stratigraphy and sedimentary responses
to changes of relative sea-level

O L E J . M A R T I N S E N* J OH N D . C O L L I N S ONt
and B R I A N K. H O L DS W O RTH:j:

*Norsk Hydro Research Centre, N-5020 Bergen, Norway;

tCollinson Jones Consulting, Westgate House, Hill's Lane, Shrewsbury SYI 1 QU, UK; and
:j: Department of Geology, Keele University, Keele,
Staffordshire ST5 5BG, UK

ABSTRACT

The Namurian (Upper Carboniferous) Millstone Grit cyclothems of northern England were formerly
interpreted as the results of deltaic progradational episodes formed during lowstands of glacial-eustatic
sea-level, punctuated by widespread fossil-bearing marine bands formed during sea-level highstands.
New research, based on detailed fieldwork in the Askrigg B lock and Craven Basin area, has led to a
reinterpretation of Millstone Grit cyclothem stratigraphy. Although the broad framework of a cyclothem
bounded by goniatite-bearing marine bands is retained, a more complex internal archi tecture is
recognized .
The tectonic framework of northern England in the Carboniferous evolved from an active extensional
phase in the Dinantian to a thermal sag phase in the Namurian. As a response, two different marine­
band-bounded sequence types developed in the Askrigg Block and Craven Basin area: (i) a 'block-edge'
sequence, formed when there was significant differential bathymetry between earlier active fault blocks
and adjacent deep basins, and (ii) a ramp sequence, formed when clastic fill and decaying differential
subsidence had reduced the bathymetric contrast between the Askrigg Block and the Craven Basin to a
gently inclined ramp. The main differences between these two sequence types are: (i) the formation and
preservation of fine-grained highstand strata, which is a particular featu re in the 'block-edge' sequence;
(ii) formation of a falling-stage systems tract, which occurred only in the ramp sequence; (iii) the
formation of a lowstand turbiditic wedge in the 'block-edge' sequence; and (iv) the greater thickness of
the 'block-edge' sequence. These differences can be attributed to varying basin bathymetry and
available accommodation as a response to the evolving tectonic and palaeogeographical situation.

INTRODUCTION

Recent advances in stratigraphical methods, par­ 1992; Maynard, 1992). This area is particularly
ticularly in sequence stratigraphy (e.g. Embry & favourable for an appraisal of the consequences of
Podruski, 1988; Van Wagoner et a/., 1988; Galloway, sequence stratigraphical analysis particularly because
1989; Cross & Gardner, 1991), have caused new of the highly refined Namurian biostratigraphical
light to be shed upon the interrelationships of stra­ framework (Fig. 2; Holdsworth & Collinson, 1988).
tigraphy and sedimentology (ct. Walker (1990, 1992) In particular, based on a detailed data set from
for a critical discussion). One basinal succession for the Askrigg Block and Craven Basin area in northern
which new interpretations based on sequence stra­ England (Martinsen, 1990), we discuss: (i) the
tigraphical analysis are being presented and debated goniatite-bearing marine bands and their relation to
is the classic Namurian (Upper Carboniferous) sea-level curves; (ii) the architectural arrangement
Millstone Grit of northern England (Fig. 1; cf. of Millstone Grit cyclothems; (iii) the sedimentary
Martinsen, 1990, 1993; Read, 1991; Collinson et al., responses to relative sea-level change; and (iv) the

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
305
306 O.J. Martinsen, J.D. Collinson and B. K. Holdsworth

ALSTON BLOCK Fig. 1. Location map of northern

England and simplified structural
BLF
map of the study area. For a
N STAINMORE BASIN detailed structural map of northern
England, see Gawthorpe eta/.
(1989). The location of cross­
(Cross-section
in Fig. 6a) sections in Figs 6 & 8 are shown,
and the location of localities
ASKRIGG mentioned in the text are indicated
BLOCK 4, 6
by numbers: 1, Lad Gill,
Stonesdale; 2, Kettlesbeck, near
2 Settle; 3, Cononley Beck, Airedale;
0 (Cross-se . 4, Backstean Gill, Nidderdale; 5 ,
Lancaster Ctton in F'
tg. 6b) Horn Crag Q uarry, Silsden; 6,
CRAVEN BASIN general area of several localities on
Cross-section in Askrigg Block used in Fig. 8. For
Figs 6c and 8)
grid references, see Figs 6 & 8.

PF, Pennine Fault; BLF,

Bu tterknowle-Lunedale Fault.

existence of a Namurian major-scale cyclicity. schemes include those of Riley (1985) and Riley
The aim is to extend existing stratigraphical models et al. (1987).
for the transition from the Askrigg Block into the There are about 60 marine bands in the Namurian
Craven Basin (Martinsen, 1993) and to provide the succession, most containing their own goniatite
impetus for a thorough re-evaluation of Millstone species (see Fig. 2; Holdsworth & Collinson, 1988).
Grit cyclicity. The average duration of a Millstone Grit cycle is
about 180 000 yr, although significant variability
probably existed (Maynard & Leeder, 1992).
NAMURIAN GEOLOGICAL Most marine bands have an interbasinal extent
FRAMEWORK (Ramsbottom et al., 1978). Some marine bands
are particularly widespread and extend over
much of Europe, in fact, a few can be traced from
Stratigraphy
western Ireland to the Urals (Wagner et al., 1979;
The concept of sedimentary cyclicity is not new in Martinsen, 1990).
Carboniferous geology and dates back to Phillips Other essential work for establishing a reliable
(1836), who recognized that the Carboniferous Namurian stratigraphical framework was carried out
Yoredale cyclothems of northern England consist of in numerous mapping projects organized by the
repeated lithological successions and that there are British Geological Survey. Perhaps the most sig­
lateral variations within these. nificant of these is the classic Rossendale Anticline
For the Namurian, the detailed stratigraphy is Memoir (Wright et al., 1927), where the authors not
dependent largely on the detailed goniatite bio­ only presented a reliable stratigraphical scheme
stratigraphy established by the work of Bisat (1914, based on goniatites, but also showed a sophisticated
1924) and Hudson (1934, 1939, 1941, 1944). Because understanding of how a particular Millstone Grit
of inadequate exposures, complex structure, and cyclothem operated in terms of depositional setting
local, inconsistent stratigraphical naming of coarser and relative sea-level change.
grained lithologies, the goniatite-bearing marine
bands are vital for stratigraphical analysis. The later
Structure and regional depositional patterns
work of Ramsbottom and co-workers (Ramsbottom
et al., 1962, 1978; Ramsbottom, 1 977, 1979) has The central and northern England area (the Central
been instrumental in consolidating and expanding Province) was in early Carboniferous time domi­
Bisat's and Hudson's work. Holdsworth & Collinson nated by a pronounced differential bathymet1ry
(1988) presented a goniatite-based biostratigraphical related to carbonate build-ups on tilted fault blocks
scheme. It presents a refined subdivision of Namurian (Gutteridge, 1989), some of which were granite­
stratigraphy and will be used in the following. Other cored (Miller & Grayson, 1982; Fig. 3A). The
 Millstone Grit cyclicity 307

MINOR CYCLE GONIATITE SPECIES

H2c(ii) Ht. prereticulatus

z
<X: H2c(i) V. eostriolatus
f=
a: H2b H. undulatum
0
"- .
� H2a(ii) H. smithi-undulatum
<X:
H2a(i) H. smithi/Hd. proteus

H1b(iv) H. aff. beyrichianum

H1b(iii) H. aff. beyrichianum

z
H1b(ii) H. beyrichianum
<X:
a: H1b(i) H. beyrichianum
(j) UJ
::J
e Q H1a(iv) Homoceras sp. nov.
0
�c :I:
u H1a(iii) I. subglobosum
0
-2 H1a(ii) I. subglobosum
co
u
Q; H1a(i) I. subglobosum
Q.
Q.
::J E2c2(iii) N. nuculum
-
c
co E2c2(ii) N. nuculum
·c:
::J
E E2c2(i) N. nuculum
co
z E2c1 N. stellarum
.....
0
t E2b3 Ct. nititoides
z
<X:
co
"-
0 E2b2(ii) F. holmesi
Fig. 2. Excerpt of the goniatite a:
UJ E2b2(i) Ct. nitidus
band stratigraphy established for C!l
(f)
the Namurian. Based on z E2b1(iii) C. subplicatum Grp.
a:
Holdsworth & Collinson (1988). <X:
E2b1(ii) Ct. edalensis/Ct. bisati

C, Cravenoceras; E2b1(i) C. subplicatum Grp.

Ct, Cravenoceratoides;
E, Eumorphoceras; E2a(iii) E. yatesae
F, Fayettevillae; E2a(ii) E. ferrimontanum
H, Homoceras;
Ht, Homoceratoides; E2a(i) C. cowlingense
N, Nuculoceras;
Pendleian (part) E1c C. malhamense
V, Valites.

shallow-water limestone platforms were separated Fault (Fig. 1). The Askrigg Block is cored by the
from adjacent deeper water basins by steep sub­ Wensleydale Granite (Fig. 3A), which contributed
marine slopes (Fig. 3A; Collinson, 1988; Gawthorpe significantly to maintaining the Askrigg Block as a
et al., 1989). Most authors agree that this configur­ structural high throughout Carboniferous time. The
ation resulted from N -S extension north of the Askrigg Block was tilted to the north during the
northwards migrating Variscan front (Leeder, 1982, main extensional phase, and a half-graben basin, the
1988; Leeder & McMahon, 1988). Active extension Stainmore Basin, was formed along its northern
probably ceased in early Namurian time, causing a edge. This basin is bounded to the north by the
change from mainly fault-controlled subsidence to Butterknowle- Lunedale Fault, which also marks
thermal sag, which progressively decreased during the southern end of the Alston Block. The Stainmore
the Namurian (Fig. 3B; e.g. Leeder & McMahon, Basin ceased to subside differentially by early
1988). Namurian time (based on uniform thickness data of
The area discussed in this paper, the Askrigg overlying strata across basin), and throughout the
Block to Craven Basin transition, lies at the northern rest of the Namurian, had no significant impact on
margin of the central Province, and the two depositional patterns (Martinsen, 1990). Therefore,
components are separated by the North Craven for the cyclothems investigated in the present paper,
308 0.1. Martinsen, J.D. Collinson and B . K. Holdsworth

A
END-DINANTIAN BLOCK-AND-BASIN BATHYMETRY
CENTRAL PROVINCE N-S CROSS-SECTION
N s

30km Rossendale 25km 30km

Block

Wensleydale
Granite Only relative vertical scale

B TRANSITION OF SOUTHERN END OF ASKRIGG BLOCK

N S N s
E1c-E2a(iii); pronouced 'block-edge' E2b2(i)-Kinderscoutian: gently-inclined ramp
with turbidite-fronted deltas with shallow-water deltas

Inactive
Craven Fault system
Only relative
Ca. 40km
vertical scale

Fig_ 3. (A) North-south cross-section across the Central Province to show the end-Dinantian block-and-basin
physiography. Modified from Collin son (1988). (B) Cartoon illustrating the change of the southern end of the Askrigg
Block from a pronounced 'block-edge' to a ramp in the Namurian. Based on Martinsen (1993).

the Stainmore area is considered part of the Askrigg bathymetric break occurred, which was a first-order
Block. control on early Namurian depositional patterns.
Although the Namurian basin-filling sediments of To the south, turbidite sedimentation was import­
the Central Province were partially deposited in ant in the earliest Namurian cycles (Pendleian and
deep water and therefore are expected to be largely earliest Arnsbergian; Baines, 1977; Collinson,
marine, it is important to realize that the basin 1988; Martinsen, 1993). With time (from latest
complex formed in an intracratonic setting far Arnsbergian), this bathymetric break transformed
removed from the nearest ocean. In fact, the into a gently inclined, southerly dipping ramp
contemporaneous Mid-European Ocean probably (Fig. 3B), resulting from decreased differential sub­
lay as far south as the present position of Portugal sidence rates due to transition to thermal sag and
(Anderton et at., 1979; Ziegler, 1986, 1988), possibly progressive fill of the Craven Basin. Consequently,
connected to the Central Province through the the potential for significant turbidite sedimentation
Variscan foredeep in central Europe. The Central immediately south of the Craven Faults declined,
Province may also have had a connection to the and shallow-water deltaic sediments dominated
Canadian Maritime Provinces (Ziegler, 1988). This deposition in the later Namurian cycles of this area
intracratonic setting and isolation from the ocean is (cf. Martinsen, 1993).
important for explaining marine bands and Millstone Further south in the Central Province, in the
Grit cycle salinity variations (Collinson, 1988; Pennine Basin, basinal shales dominated deposition
Holdsworth & Collinson, 1988). until the mid-Namurian (Kinderscoutian). These
The initial Namurian depositional systems pro­ shales were succeeded first by turbidite-fronted
graded from the north and northeast into the deltas and then by shallow-water deltas (Walker,
Central Province. At the southern end of the 1966; Collinson, 1969, 1988). Although this tran­
Askrigg Block, across the Craven Faults at the sition occurred at a much later stage compared with
transition to the Craven Basin, a pronounced the northern margin of the Craven Basin, it was
 Millstone Grit cyclicity 309

perhaps not a result of ongoing, prolonged differ­ glaciations in Gondwana during the Carboniferous
ential subsidence in the south. Rather, because the (Caputo & Crowell , 1985; Veevers & Powell, 1987).
clastic supply was mainly from the north, the feeding Although goniatites constitute the stratigraphically
systems did not manage to prograde into the area most useful faunal phase of marine bands, it is
until the middle part of the Namurian. Basin-filling not uncommon for various other faunal elements
was therefore delayed in the Pennine Basin compared to occur in distinct faunal phases stacked one
with the Craven Basin, probably as a result of above the other. Based on detailed analysis of a
the clastic supply being absorbed by depocentres cored condensed basinal succession at Ashover,
beneath the southern North Sea. Derbyshire, an ideal succession of faunal phases
In the Staffordshire Basin, a slightly different has been constructed (Ramsbottom et al., 1962;
situation existed. There, protoquartzitic turbidites Holdsworth & Collinson, 1988; Fig. 4). In the
were supplied from the southern Midland land mass Ashover boreholes the total succession from base
in several minor cycles from as early as E1b Namurian to the Marsdenian is of basinal mudstones
(Holdsworth, 1963; Trewin & Holdsworth, 1973). In with no coarse clastics. The Ashover succession
the Widmerpool Gulf to the east, the same situation was deposited on part of the Derbyshire Massif, a
seems to have taken place (Aitkenhead, 1977). mudstone-draped Dinantian carbonate platform,
Although sand was supplied to these areas at this isolated from clastic supply by a partial moat of
early stage, the supply was insufficient to fill the deep-water gulfs. The isolation from clastic supply
available accommodation space, and complete basin means that there is virtually no change of lithofacies
filling did not occur until the Marsdenian R2b(ii) between marine bands and intervening barren mud­
Ashover/Roaches Grit was supplied from the north stones. In such circumstances it is very difficult to
(Jones, 1980; Collinson, 1988). conclude that marine bands are purely reflections of
sediment starvation and condensation.
The presence or absence and the changes of fauna
GONIATITE-BEARING MARINE BANDS in a succession like Ashover almost certainly result
AND RELATIONSHIP TO from changes in the overlying water column, in
SEA-LEVEL particular its salinity. Of the faunal elements present
in the sequence, only the thick-shelled goniatites
The marine bands, which allow such a refined bio­ are thought to reflect fully marine salinity (Fig. 4).
stratigraphical breakdown of the Namurian, are The other phases are thought to record progressively
virtually the only fossiliferous sediments in Namurian less saline conditions. The implication is that the
basinal successions. Fossiliferous intervals range marine bands do not reflect faunal concentration but
from a few centimetres up to around 2 m in thickness represent the arrival of fauna that had been absent
and generally constitute only a small proportion of from the basin during intervening times. From the
the total Namurian succession (Collinson, 1988; succession of faunal phases, the arrival of a fully
Holdsworth & Collinson, 1988). Sediments between marine fauna appears to be associated with pro­
marine bands are virtually devoid of body fossils, gressively increasing salinity (Fig. 4)·. Because many
irrespective of whether the sediments are fine­ of the faunal elements are nektonic or pelagic, it is
grained, deep-basinal mudstones, thick sequences of not clear whether the salinity changes applied to the
turbidite sandstones or progradational deltaic suc­ whole water column or whether they affected only
cessions (Fig. 4). Where the intervening sediments the photic zone, where most of the fauna swam.
are deltaic progradational successions, the marine Whatever the extent of desalination, it is apparent
bands occur closely above an abandonment surface that the time of maximum salinity in the basin is
or just below the base of the next upward-coarsening most likely to coincide with or occur very slightly
pro-delta unit, leaving a clearly asymmetric cyclo­ after the peak of the eustatic sea-level curve. If one
them motif (Fig. 4). These observations, coupled considers the wider palaeogeographical context of
with the widespread nature of many marine bands the Namurian basins of the British Isles (see above),
(see above) suggests an external control on the it is clear that they occupied an enclosed intra­
occurrence of goniatite-bearing marine bands. Most continental setting, removed by considerable dis­
Namurian researchers have believed this mechanism tance and topographical complexity from the nearest
was glacial-eustasy, a theory which is increasingly ocean. Connections between the ocean and the
supported by growing evidence for widespread British basins were therefore probably long and
310 O.J. Martinsen, J.D. Collinson and B . K. Holdsworth

E1c minor E 1 c minor Ideal

cycle in cycle in faunal/salinity
Eustatic curve Craven Basin Staffordshire cycle
F-R (Ca. 800 m) Basin (based on Ashover cycles)
(Ca. 25-90 m)

E2a(i) 6 E2a(i) 6 Goniatite

Delta top sands Barren shale 5 AID
4 Spat
3 Lingula
Delta slope Turbidite sands 2 Planolites
heterolithics and ironstones 1 Fish

E_j_
SIMPLE, BARREN
MINOR 1 Fish

T
CYCLE 2 Planolites

Barren s ale
3 Lingula
4 Spat
Calcareous 5 AID
siltstone 6 Goniatite Ideal
6 E1c 5 faunal/salinity
4 oscillation

1
3

L AEN

Fig. 4. Minor cycle stratigraphy showing variations of the basic cyclicity, from the ideal faunal/salinity cycle erected from
Ashover borehole data (cf. Ramsbottom et al. , 1962; Holdsworth & Collinson, 1988), to thicker, sand-prone cycles from
the Staffordshire Basin and the Craven Basin. The probable relation to the eustatic curve is also shown. (Modified
from Holdsworth & Collinson, 1988.)

A, Anthracoceras; D, Dimorphoceras (thin-shelled goniati tes) .

tortuous. That, coupled with the fact that the British Hormuz are all relatively sediment free.
basins were supplied by very large volumes of fresh Simple but conservative calculations based on
water, which delivered an abundant supply of clastic likely river discharges (approximately 1 x 10 4m3 s-1,
sediment, makes it easy to envisage how fluctuations based on Kinderscout Grit channels; Collinson,
in eustatic sea-level would have induced changes in 1969 ) and basin water volumes (around 1 x 10 13 m3,
the salinity of basinal waters. During lowstands, the based on an assumed areal extent of the basins of
cross-sectional areas of the connecting channels 5 x 10 4km2 and a depth of 200 m) suggest that river
would be much reduced, especially if they were of water could have replaced basinal water in a matter
low relief. Their main function would then be to of 30-35 yrs if the process was 100 % efficient. Given
accommodate the discharge of fresh water from the the average duration of a minor cycle was 180 000 yrs,
basins to the ocean. During highstands, expanded desalination seems likely even at much lower
cross-sectional areas would have allowed oceanic efficiencies. In order for basinal waters to have
waters to enter the basins and thereby increase become fully saline after a period of lowstand, and
salinity. Salinity changes are recorded from the Red thereby allowing the deposition of a marine band, it
Sea as a result of Pleistocene/Holocene changes in is clear that some threshold of eustatic level must
sea-level, although in this case lowstand led to hyper­ have been exceeded and that sea-level must have
salinity because of the climate and Jack of river remained above the threshold for a period of time.
discharge (Thunell et al., 1988 ). The fact that sal­ The marine bands, therefore, on this basis, are
inities appear to have oscillated throughout the thought to record periods of time at and around
Namurian within rather constant limits suggests that eustatic maxima. Given that there was probably
the straits which connected the basins with the ocean some lag in the response of salinity to eustatic level,
were rather stable and not subjected to major it is likely that the interval of marine band deposition
tectonic change or sedimentary choking. Partial may be skewed slightly later about the eustatic peak
analogues at the present day, such as the Skagerak­ (Fig. 4).
Kattegat, the Hudson Straits and the Straits of From their positions close to the bases of pro-
 Millstone Grit cyclicity 311

gradational deltaic cyclothems it is likely that marine occurrence in relation to eustatic oscillation differ
bands are associated with periods of relative sedi­ only slightly from those of conventional sequence
ment starvation and of rapid deepening. If marine stratigraphical theory, they illustrate ways in which
bands are considered only in this context then the principles of sequence stratigraphy must be
it could be suggested that they record the time adapted in settings other than those in which they
of maximum rate of rise in relative sea-level, i.e. were formulated initially.
around the R inflection point of the Exxon sequence
stratigraphical model (Jervey, 1988; Posamentier
et at., 1988). M ILLSTONE GRIT
However, we believe this to be unlikely. If we are CYCLOTHEM ARCHITECTURE:
correct in believing that full marine salinity was RESPONSE TO
necessary to support a fauna including thick-shelled RELATIVE SEA-LEVEL CHANGE
goniatites, a view supported by the common occur­
rence of Radiolaria in this faunal phase and (as far
Introduction
as is known) no other, then, given the ocean-distant
setting of the Millstone Grit basins, it seems likely Since Wright et al. (192 7) published the classic
that such salinities could have been achieved only Rossendale Anticline memoir, the established view
close to the eustatic maxima. This would normally has been that Namurian cyclothems represent more
be expected to significantly post-date the R inflection or less continuous progradational episodes formed
point. during falling or low relative sea-level, punctuated
It could even be inferred that, in some cases, parts by marine transgressions, highstands of sea-level
of the curve between the R inflection point and and deposition of goniatite-bearing marine bands
the eustatic maximum may be represented by (Fig. 5). In the following, two Millstone Grit cyclo­
the inferred lower salinity phases of fauna that thems are discussed in order to illustrate what has
occur within marine bands below the thick-shelled been interpreted as a different and more complex
goniatites. In other cases, there is compelling evi­ internal architecture. The two cyclothems were
dence to place the response to the R inflection point selected based on outcrop quality, and because they
clearly below the marine band. In the E1c- E2a show different lateral facies and thickness develop­
cycles in Staffordshire (Trewin & Holdsworth, 1973), ments. It is focused on the cyclothem development
units of 'transitional shale', unfossiliferous and from the Askrigg Block and into the northern part
lacking carbonate, invariably separate a sideritic of the Craven Basin. The depositional history of
protoquartzitic turbidite unit from calcitic marine each cyclothem is reviewed briefly (for a full dis­
siltstones below and goniatite-bearing marine bands cussion of this aspect for the lower Namurian
above (Fig. 4). Such shales can readily be interpreted (Pendleian -Chokierian) cyclothems, see Martinsen,
as recording periods of rapid adjustment of environ­ 1990, 1993).
ment when conditions favoured neither marine fauna In the following, we take a less rigid view concern­
or mineralogy nor influx of delta-derived clastics. ing the use of the word 'sequence' than some existing
The transitional shales below a marine band probably definitions. In the American Geological Institute
closely approximate the R infection point, whereas definition, for example (Bates & Jackson, 1987),
the overlying transitional shales mark the earliest 'sequence' is constrained to a depositional unit
stages of eustatic sea-level fall. bounded by unconformities. However, as there is in
In cyclic deltaic sequences, maximum conden­ existence more than one sequence stratigraphical
sation, predicted as a response to the R inflection model (Embry & Podruski, 1988; Van Wagoner
point, is most likely recorded by heavily bioturbated et a/., 1988; Galloway, 1989; Cross & Gardner,
flooding surfaces or coal seams, which commonly 1991), and the models differ in what type of strati­
terminate sandstone sequences. Such surfaces graphical surface is chosen as a boundary, we use
commonly are separated from the overlying marine the term 'sequence' for the stratigraphical unit that
band by a thin interval of mudstone. This mudstone is practical to use in a given basin. The most practical
is therefore thought to be the record between the R sequence type is chosen based on an objective
inflection point and the sea-level maximum of the evaluation of the nature and extent of various strati­
marine band. graphical surfaces. The most easily defined, most
Whilst these modifications regarding marine band extensive surfaces, based on detailed facies analysis,
312 0.1. Martins�n, J.D. Collinson and B . K. Holdsworth

,_ Goniatite
band
Fluvio-deltaic distributary
channel systems Barren
mudstones

� Goniatite
band
Slope channels feeding
turbidites

Continuous progradation
with successive clinoforms
downlapping base-of-cycle
goniatite band

Base-of-slope turbidites related to Adjacent goniatite bands

gravitational instability or underflows only separated by thin shale
during progradation in basinal areas

Coarsening-up
siltstone

t# Goniatite band
m

Fig. 5. The traditional Mill stone Grit cycle interpretation as a result of d elta progradation (turbidite-fronted delta
example). See text for discussion.

are chosen as sequence boundaries. To avoid conflict

The E2a(i) cyclothem
with allostratigraphy, the same type of surface is
picked throughout the succession studied.
Boundaries and thickness
Goniatite-bearing marine bands are considered to
represent the most significant stratigraphical levels The Cravenoceras cowlingense goniatite band forms
in the Namurian of the Askrigg Block to Craven the base of the E2a(i) cyclothem (Fig. 2), and can be
Basin succession, because of their lateral extent and found across most of the Askrigg Block, as well as in
distinctive faunal characteristics (Martinsen, 1993). the western and easternmost parts of the Craven
They represent maximum transgression between Basin (cf. Martinsen, 1990). The top of the cyclothem
successive progradational episodes and compare with is marked by the base of the Eumorphoceras bisul­
the genetic stratigraphical sequence boundaries of catum ferrimontanum band, which forms the base of
Galloway (1989). Marine bands have a thickness the overlying E2a(ii) cyclothem. This band is found
and are not surfaces in a strict sense. However, throughout most of the Craven Basin, and on the
when we describe them in the following, we see their northwestern side of the Askrigg Block.
mid-point as a 'surface', representing peak marine­ On the Askrigg Block, the cyclothem is approxi­
band conditions. mately 70 m thick in the northwest, whereas it thins
Detailed facies analysis and regional understand­ to around 40 m in the southeast (Fig. 6A-C). In the
ing of marine bands suggest that they represent the Craven Basin it is highly variable in thickness. In the
units across which the most significant palaeogeo­ northeast of the basin it is apparently less than 1 m
graphical changes occurred in the Namurian suc­ thick due to condensation (further discussion below),
cession. The marine bands are therefore chosen whereas in the Lancaster Fells in the west, the
as sequence boundaries. Within a marine-band­ E2a(i) cyclothem approaches 150 rn in thickness
bounded sequence, however, the idea is to establish (Fig. 6A; cf. also Moseley, 1954, 1956).
a detailed stratigraphical architecture, using both
other stratigraphically significant surfaces (e.g.
Depositional model and tectonism
lowstand erosion surfaces or unconformities) and
systems tracts. Detailed studies (cf. Martinsen, 1993) suggest the
following depositional model for the E2a(i) cyclo-
 Millstone Grit cyclicity 313

them (Fig. 7). A phase of mudstone deposition wedge to develop there (see above), because there
occurred on the Askrigg Block. A zone of barren was still a significant bathymetric contrast between
mudstones were succeeded by a phase of normal the block and the basin.
marine conditions represented by a fossiliferous
interval with the Cravenoceras cowlingense fauna.
E2a(i) cyclothem architecture and sequence
A slight coarsening upward trend within the mud­
stratigraphy
stones overlying the fossiliferous level suggests
progradation of a clastic shoreline from the north. The Cravenoceras cowlingense marine band itself is
A drop of relative sea-level probably followed, and interpreted to represent the eustatic maximum or
a rapid basinward shift of facies is interpreted shortly thereafter (discussion above; Fig. 6A- C),
to have taken place, causing establishment of the whereas the thin, barren mudstones below the
shoreline at the southwestern margin of the Askrigg marine band probably represent maximum trans­
Block. From this shoreline, sediment was supplied gression (maximum flooding).
down into the Craven Basin as turbidity currents The fine-grained deposits above the marine band
and suspended silts, leading to the establishment of (eustatic maximum), and directly below the major
a slope apron (Fig. 7). fluvial complexes, but separated from these by a
During the ensuing rise of relative sea-level, fluvial major erosion surface across which a major basin­
aggradation took place within the incised palaeo­ ward shift of facies occurs, are interpreted as late
valleys (Upper Howgate Edge Grit) and eventually highstand deposits (Fig. 6A & C). The mudstones
led to the deposition of a fluvial sheet sandstone are interpreted to conform to the upper part of the
(the Tan Hill Grit; Figs 6A & 7; Martinsen, 1993). falling limb on the eustatic curve, and on the relative
Deposition corresponded with continued pro­ sea-level curve prior to initiation of the fall, but at a
gradation and fill of the basin, with the development time when the rate of rise had decreased to the point
of a deltaic plain in the northwestern parts of the where clastic systems could prograde back into the
Craven Basin. The deltaic plain was subsequently area from the north.
flooded and transgression took place. In the initial The erosion surface underneath the Upper
phase of the transgression, shallow-water conditions Howgate Edge Grit fluvial complex (Fig. 6A) is
persisted on the northwestern parts of the Askrigg interpreted as a lowstand erosion surface. Based on
Block, allowing for sporadic, high-frequency pro­ the interpretation that a prominent basinward shift
gradations of small delta lobes (Figs 6A & 7). of facies takes place across the surface, it is also
Finally, the relative sea-level rise overwhelmed interpreted as an Exxon-type sequence boundary.
sediment supply, causing transgression and establish­ The surface is unlikely to correlate with the base of
ment of offshore shelf conditions where mud­ the turbiditic slope sediments in the Craven Basin.
stone sedimentation dominated. The conditions This is because incision probably evolved with time,
were shallow enough for occasional storms to and the final position of the incision surface most
rework benthic fauna into a shell bed, the Lad Gill likely corresponds to some cryptic level within
Limestone. the upper parts of the turbidite slope succession
Fluvial channels were confined largely to the (Fig. 6A). Therefore, the downlap surface below the
western parts of the block, leaving the eastern part turbidite deposits in the Craven Basin is more likely
relatively starved of sediment (Fig. 7). The cause for to record initiation of sea-level fall rather than cor­
this was probably increased subsidence along the relate directly with the final position of the lowstand
Dent Line, a long-lived lineament bounding the erosion surface below the Upper Howgate Edge
Askrigg Block in the west (cf. Underhill et al. , Grit on the Askrigg Block. The stratigraphical level
1988). The Dent Line probably experienced minor corresponding to the ultimate lowstand of relative
reactivation during the early Namurian (Martinsen, sea-level is considered as the boundary between a
1993). The lack of sediment supply to the eastern lower and upper part of the lowstand systems tract
parts of the block caused sediment starvation in the (Fig. 6A). Architecturally, the unit below this surface
eastern parts of the Craven Basin (Figs 6C & 7), is a wedge in the Craven Basin, and the correlative
because most sediment was supplied to the basin of this lowstand wedge on the Askrigg Block is the
across the western part of the Askrigg Block. This lowstand erosion surface.
continued sediment supply to the western parts of When relative sea-level started to rise, and the
the Craven Basin caused the slope apron or turbiditic incised valleys started to fill, turbidite sedimentation
314 O.J. Martinsen, J.D. Collinson and B . K. Holdsworth

NW Askrigg Block Ca. NW Craven Basin

Sea-level curves Lad Gill (SE 888046) 50km Kettlesbeck (SD 7 43632)
(Lad Gill section)

- Wave-influenced
IFS
3 shallow-water deltaic
- sediments

j, Braided, fluvial
1 channel complexes
1 and overbank fines Upper E
LST "'
c.
-- -
- --- �

- - Q)
�
Lower LST g_
c"O
0 Q)
·- "'
� 0
"' c.
EX
� Q)
F R F
��
.,-o
-"'
Relative Eustatic co.o
Distal part of "0-

highstand delta complex

E
:J
.0
Q)
0
Lower sequence boundary a:
(Cravenoceras cowfingense marine band)

A ms

Sea-level curves NW Craven Basin NE Craven Basin

+----- 4okm -----•
Kettlesbeck (SD 743632) Cononley Beck (SD 986468)
Relative Eustatic

River mouth
bar

E
"'
c.
� Upper sequence boundary
Q)
� and superimposed initial
.2- flooding surface
--o Upper
c Q)
0 "'
·-0
LST

�
�c.
co X
E.,
0� --
u..-o --
Q) "' -- 10
-.0
"' - Lowstand, turbiditic - -- Edge Marine
"
slope wedge - Band
E _
:J 0 - (E2a (ii))
.0 Lower m
Q) m fs
0
a:
LST

Lower sequence boundary

(base E2a(i)) and conformable
lowstand surface
Cravenoceras cowfingense
marine band (E2a(i))

F R F R
.. .. .. .. B
 Millstone Grit cyclicity 315

Sea-level curves SE Askrigg Block NE Craven Basin

Backstean Gill (SE 058723)
+----- 25km ____ .,. Cononley Beck (SD 986468)
Relative Eustatic
F R F R
,._____.. ,._____..
40

Upper sequence boundary

Lower sequence boundary All surfaces merge into

Cravenoceras cowlingense marine band condensed section in basin
c (base E2a(i) cycle) on Askrigg Block

Fig. 6. (Opposite. ) (A) Stratigraphical correlation and proposed sea-level curves for the cross-section of the E2a ( i )
sequence from the northwestern part of the Askrigg Block into the northwestern part of the Craven Basin. (Opposite.) (B)
Like (A), but cross-section from the northwestern part of the Craven Basin to the northeastern part of the basin. (Above. )
(C) Like (A) , but correlation from the southeastern part of the A skrigg Block into the northeastern part of the Craven
Basin. For section locations, see Fig. 1.

MFP, Mirk Fell palaeosol ( Elc) ; UHEG, Upper Howgate Edge Grit; THG, Tan Hill Grit, LG Lst, Lad Gill Limestone;
MFS, maximum fl ooding surface; LES, lowstand erosion surface; IFS, initial flooding surface; SB, sequence boundary.
Systems tracts; HST, highstand; LST, lowstand; TST, transgressive.

may have gone on for some time. The time for Secondly, the effects of a particular sea-level
cessation of turbidite sedimentation within the E2a(i) change may reach only a surprisingly short dis­
cycle may have been a result of elimination of the tance upstream from the coast. Saucier (1981; cf.
differential bathymetry (and accommodation space) review in Schumm, 1993) suggested that the effects
at the southern margin of the Askrigg Block. How­ of the approximately 120-m sea-level lowering
ever, during times of rising sea-level, rivers are more during the late Wisconsin lowstand extended only
likely to aggrade and trap sediment, particularly in 370 km upstream in the Mississippi river valley, in
their downstream reaches (Schumm et al., 1987; contrast to the 1530 km estimated by Fisk (1944).
Schumm, 1993), which in turn may reduce the poten­ Also Blum (1994) showed similar ·effects for the
tial for voluminous turbidity current generation. Colorado River of Texas.
Several factors complicate the correlation of major Both points are important for the E2a(i) cyclo­
stratigraphical surfaces from the basinal part of a them. If there was a lag time between the initial sea­
succession upstream to a significant distance into level rise and the fluvial response by aggradation,
fluvial systems. Firstly, there is likely to be a lag in the fluvial deposits on the northwestern part of the
the time the relative sea-level behaviour is recorded Askrigg Block may not be coeval, but perhaps be
in upstream reaches. For instance, it is known from later than those in the basin or at the edge of the
field data as well as flume experiments that upstream Askrigg Block, 60 km to the south. This is similar to
reaches may incise while downstream reaches back-filling during relative sea-level rise, where
aggrade (Shepherd, 1979; Schumm, 1993). This deposition starts at sea-level and with time progresses
suggests that sequence boundaries commonly may back up the alluvial tract. Such a relationship cannot
have very complex histories and are not strictly be verified from the existing data set. An even more
correlative over large distances in upstream­ refined chronostratigraphy is needed to work out
downstream profiles. such details.
316 O.J. Martinsen, 1. D. Collinson and B. K. Holdsworth

A B

Condensation in NE part
of basin (lateral setting,
no supply from block)

Bay deltas on
SE block,
D

t
�
I
condensation
Rapid coastal transgression, but minor,
in
shallow-water deltas briefly interrupted
NW part of
retrogradation
basin
Salinity
increase and
faunal reintroduction
Distributary channels in
Start of prolonged period of
basin, braided fluvial style ..,.
on block leaving sheet
sandstone (Tan Hill Grit)
slow deposition and condensation
in basin and on the block
..a
�
9

Fig. 7. Palaeogeographical summary diagram of the E2a(i) cyclothem related to different stages of sea-level.
A, Eustatic maximum; B, lowstand; C, rising sea-level; D, rapidly rising sea-level.

It is fairly certain that the effects of the E2a(i) sea­ Edge Grit lies directly on offshore mudstones. The
level fall extended as far upstream as the north­ sheet sandstone occurs above (the Tan Hill Grit),
western part of the Askrigg Block (provided the within the part of the cyclothem where relative sea­
equivalent shoreline during the sea-level minimum level probably was rising steadily and significant:
lay at the southern end of the Askrigg Block), alluvial aggradation occurred (Fig. 6A). During very
because the facies interpretation clearly indicates rapid rates of relative sea-level fall, the only possi­
that a sea-level fall and basinward shift of facies bility for the fluvial system is to incise, particularly
occurred (Fig. 6A; Martinsen, 1993). when the substrate is fine-grained sediments with
On the southeastern part of the Askrigg Block, pronounced cohesive stability and little possibility to
the abrupt transition from offshore mudstones to compensate by channel-belt widening and increased
burrowed and rooted, thin crevasse splay deposits bed roughness (Schumm, 1993). The resulting fluvial
probably records the lowstand phase (Figs 6C & 7). morphology of the Upper Howgate Edge Grit is
Overbank deposition was probably limited during therefore likely to be different from the braided,
the early parts of the lowstand period when most amalgamated model of Shanley & McCabe (1993).
sediment was transferred directly to the block-edge. This may have been the case for the Upper Howgate
During later aggradation, more sediment may have Edge Grit of the E2a(i) cyclothem. Only when
been supplied to floodplain or interchannel areas, relative sea-level had risen further, perhaps to where
but the proposed tectonic tilting of the Askrigg the channels were no longer confined by an incised
Block to the west would have prevented substantial valley did the braided, sheet-like morphology of the
volumes from reaching the southeastern parts of the Tan Hill Grit develop (Fig. 7).
block (Fig. 7). The thin deltaic coarsening upward succession on
On the Askrigg Block, the incised Upper Howgate top of the Tan Hill Grit in the northwestern part of
 Millstone Grit cyclicity 317

the Askrigg Block suggests that a flooding surface Basin, only one or two bands are developed at any
occurs between the two units (Fig. 6A). This is the one location (e.g. Wilson, 1977), leaving a great
first indication of inundation and transgression within deal of ambiguity as to which exact cyclothem the
the E2a(i) cyclothem. It also suggests that the described rocks belong. For the present purpose, it
boundary between a lowstand and transgressive is described as 'the E2c2 cyclothem'.
systems tract should be put at this surface (see Based on facies analysis (Martinsen, 1990), map­
above). Moreover, the mudstones and goniatite­ ping (Stephens et al. , 1953), and local stratigraphical
bearing Lad Gill Limestone on top of the deltaic studies (Wilson, 1960, 1977), there is a high degree
succession indicate further transgression and delta of confidence in correlating the described cyclothem
abandonment. These data support a transgressive between the southeastern part of the Askrigg Block
systems tract interpretation of the 8-m deltaic suc­ and the northeastern parts of the Craven Basin, over
cession, perhaps including the 4-5 m of mudstones a N-S distance of approximately 25 km. Therefore,
below the Lad Gill Limestone. the exact age and correlation of the described cyclo­
In the basin, there is no evidence for coarser them with the minor cycle chart (Fig. 2), is of no
grained, transgressive sediments. Instead, the thin importance for the present purpose.
deltaic succession is inferred to thin basinwards and The E2c2 cyclothem varies considerably in thick­
merge with the mudstones on top of the distributary ness. It is 5 -36 m in the southeastern corner of the
channel complex (Fig. 6A). In effect, the trans­ block, around 34 m in the northeastern part of the
gressive systems tract in the basin is represented by a Craven Basin, only 12 m in the north-central parts of
condensed section, which probably merges with the the basin, and around 40 m thick in the northwestern
condensed shales at the base of the overlying E2a(ii) parts of the basin (Fig. 8).
cyclothem (Fig. 6A). On the Askrigg Block, the cyclothem base is
In the southeastern parts of the block and in the interpreted to be the erosive base of the Lower
eastern parts of the Craven Basin, there is little Follifoot Grit, whereas in the basin, the base of the
evidence for a transgressive phase (Fig. 6C). Only, cyclothem is one of the three N. nuculum bands
the intense burrowing on top of the crevasse sand­ (Fig. 8). Martinsen (1993) interpreted the underlying
stones indicate delta abandonment and trans­ E2cl cyclothem to be entirely eroded on the block
gression. The burrowed level is sharply overlain by by the Lower Follifoot Grit, which also erodes into
dark, offshore mudstones (Fig. 6C). the E2b3 cyclothem (Fig. 8). The cyclothem top is
The E2a(ii) marine band, the E. bisulcatum interpreted to coincide with the base of a multi­
ferrimontanum band caps the E2a(i) cyclothem. Like cyclothem condensed interval, which encompasses
its E2a(i) counterpart, the C. co wlingense band, it is at least nine minor cycles (E2c2(i or ii)- H1b(iii)),
interpreted to represent the peak of the eustatic and which can be traced across the entire Craven­
curve. Dark mudstones underlie the marine band. It Askrigg region (Fig. 8; ct. Martinsen, 1990).
is considered likely that maximum transgression
occurred at some point during deposition of these
Depositional model and tectonism
barren mudstones.
In the southeast part of the Askrigg Block, the
Lower Follifoot Grit is a trough cross-stratified
The E2c2 cycl othem
sandstone, which has been interpreted to represent
deposition in low-sinuosity, braided river channels
Boundaries and thickness
(Martinsen, 1993; Figs 8 & 9). Geometrically, it has
One of three upper Arnsbergian cyclothems with two modes of occurrence. For the most part, it is a
Nuculoceras nuculum at the base (i.e. E2c2(i, ii or tabular sheet sandstone recognizable over several
iu); cf. Fig. 2) is described as a counterpart to the tens of square kilometres (Fig. 8; Wilson, 1960).
E2a(i) cyclothem. There is uncertainty concerning However, locally, two well-defined storeys of sand­
the exact age of both the interpreted basal marine stone are observed. The lower one is very lenticular,
band and the cyclothem itself, because all the three forming ribbon-shaped bodies up to 15 m thick and
nuculum bands are not developed everywhere. On hundreds of metres wide (Fig. 8). These sandstone
the Askrigg Block, no N. nuculum bands are found , bodies are clearly incised into the underlying depo­
probably as a result of occlusion or erosion (see sits. The upper storey is much more extensive
below). In the northeastern part of the Craven lateraLly, forming a pronounced sheet. The Lower
318 0.1. Martinsen, J.D. Collinson and B . K. Holdsworth

SE Askrigg Block ---- 20-25krn ---+--• NE Craven Basin Sea-level curves

N S Relative Eustatic
Great Whernside West Scrafton Beldin G i l l S l i pstone Crag Thrope Edge !Horn Crag Quarry, Silsden F R F R
SE 0074 SE 082824 SE 095799 SE 1 38821 SE 1 06752 SE 053480 ---- ----

Scar House Beds

(E2b3)

35m
Lowstan d erosion Mu lti-cycle
su rface condensed section
(coincides with (contains maxi m u m flooding
sequence boundary) s u rface and top seq u e n ce Highstand
boundary of this cyclothem) systems tract
0

Fig, 8, Cross-section showing stratigraphical correlation and proposed sea-level curves of the studied E2c2 cyclothem from
the southeastern part of the Askrigg Block to the northeastern part of the Craven Basin. See Fig. 1 for section location.

FSST, falling stage systems tract; LST, lowstand systems tract; MFS, maximum flooding surface; LES, lowstand erosion
surface; FRS, forced regression surface.

Follifoot Grit is abruptly overlain by a multi­

E2c2 cyclothem architecture and
cyclothem condensed section, which contains gonia­
sequence stratigraphy
tites of the lsohomoceras subglobosuml Homoceras
beyrichianum marine bands of H1a- H1b age (Fig. The Nuculoceras nuculum goniatite band probably
2; Wilson, 1960). represents maximum eustatic sea-level (discussion
In the basinal section above Silsden in Airedale above). Consequently, the dark mudstones that
(Fig. 8), to the south, offshore mudstones occur overlie the goniatite band in the basin are interpreted
directly above the Nuculoceras nuculum marine band to represent late highstand deposits (Fig. 8), laid
at the base of the cyclothem (Stephens et al. , 1953). down when the nearest shoreline was still far to the
These mudstones are only a few metres thick, and north.
are then abruptly overlain by a 26-m-thick, entirely In the basin , the abrupt contact between mud­
sandy succession (Fig. 8). This sandstone succession stones and the delta sandstones, and the close
has been interpreted as friction-dominated and juxtaposition of the goniatite band and the delta
wave-influenced delta-mouth-bar deposits (Fig. 9; sandstones (only separated by a few metres; Fig. 8),
Martinsen, 1990). The mouth bar sandstone is suggest that a relative drop of sea-level took place,
growth-faulted to the depth of the base of the sand­ represented by the lower boundary of the sandstone.
stone, suggesting that the fault soles out within the One interpretation of the lower boundary is as a
underlying mudstones. The mouth bar sands are lowstand surface of erosion or Exxon-type sequence
overlain by an erosively based, very coarse-grained boundary formed during the most rapid rate of sea­
pebbly sandstone, which is entirely trough cross­ level fall. The delta sandstones above the lowstand
stratified (Fig. 8). The sandstone is interpreted as a surface would further be interpreted as lowstand
braided fluvial complex, and is eventually overlain sediments, suggesting that the entire delta is a
by fossiliferous mudstones of the overlying multi­ lowstand delta.
cyclothem condensed section (cf. Stephens et al. , An alternative interpretation is that the sharp
1953; Wilson, 1977). transition from the mudstone to the deltaic sandstone
was produced by accretionary forced regression
during relative fall of sea-level, making the delta
 Millstone Grit cyclicity 319

A B

Falling-stage (forced
regression) delta with
growth faults

N N

r coastal transgression

Salinity
increase and
faunal
Start of prolonged period of reintroduction
slow deposition and condensation
in basin and on the block

Fig. 9. Palaeogeographical summary diagram of the studied E2c2 cyclothem related to different stages of sea-level .

A, E ustatic maximum; B, falling-stage; C, lowest/ slowly rising sea-level; D , rapidly rising sea-level.

sandstones part of a falling stage systems tract the inferred falling-stage delta in the basin (Fig. 8),
(Fig. 8; e.g. Plint, 1988; Hunt & Tucker, 1992; but perhaps more likely, it represents the incision
Posamentier et al. , 1992). This interpretation is related to relative lowstand and correlates with the
supported by the fact that the braided, fluvial sand­ erosion surface between the deltaic sandstones and
stone overlying the deltaic rocks is so much coarser the very coarse fluvial sandstones in the basin. This
than the deltaic sandstones, shows entirely different interpretation implies that erosion below the Lower
palaeocurrents and therefore does not seem to be Follifoot Grit on the Askrigg Block involved both
related to the deltaic sandstones (i.e. it is not their initial formation during the falling stage and further
distributary channel). The erosion surface below the development and modification during the relative
fluvial sandstones could represent the relative sea­ lowstand.
level lowstand and thus equate with an Exxon-type If this interpretation is correct, the fluvial sand­
sequence boundary (Fig. 8). stones on the block and in the basin were deposited
On the Askrigg Block, the erosion surface below during a relative rise of sea-level (Fig. 8). Questions
the lenticular channels of the Lower Follifoot Grit as to whether the shoreline prograded during filling
probably marks an extensive erosion surface related of the incised fluvial channels, or if enough sediments
to falling and low relative sea-level (Fig. 8). were trapped on the alluvial plain to enhance trans­
Martinsen (1993) suggested that extensive incision gression at the contemporaneous shoreline cannot
took place on the block at this time, and the under­ be answered due to inadequate exposure. In the
lying E2c l cyclothem was largely eroded. The latter case, the fluvial interval should be termed part
incision phase may correlate with the deposition of of a transgressive systems tract. However, the lack
320 O.J. Martinsen, J. D. Collinson and B. K. Holdsworth

of evidence for transgressive deposits, and analogy post-extensional subsidence but also of decreasing
with the E2a(i) cyclothem suggests that it is more overall subsidence (cf. Leeder & McMahon , 1988).
likely that progradation continued up until or very One similarity between the two cyclothems is the
close to cyclothem abandonment. In that case, the proposed prolonged absence of clastic supply from
fluvial rocks should be termed part of the lowstand the time of maximum transgression (thin shales
systems tract (Fig. 8). below marine bands), until maximum eustatic
It is likely that sea-level fall across a slowly sub­ highstand (marine bands; see Figs 6 & 8). This
siding area like the Askrigg Block would enhance encompasses one-quarter of a eustatic curve, and,
incision and stabilize single channels or 'valleys', depending on the rate of subsidence, probably
particularly if incision was rapid and occurred in a corresponds to the middle part of the rising, extended
fine-grained substratum (see Schumm, 1993). Only limb on a relative sea-level curve. Therefore, it
during sea-level rise and consequent increasing seems that clastic supply was absent for a prolonged
accommodation would the incised channels be period compared with conventional sequence strati­
'overflooded' and fluvial deposition take place out­ graphical models (cf. Van Wagoner et al. , 1990).
side the incised channels. If grain size, discharge and The decreasing overall subsidence is important for
gradient were favourable for a braided fluvial style explaining why late highstand deposits are not pre­
to develop at this time, the resulting fluvial sediments served on the Askrigg Block in the E2c2 cyclothem
would probably be sheet-like. in contrast to the E2a(i) cyclothem within which
Provided the fluvial rocks of the E2c2 cyclothem highstand deposits are relatively thick. A lower rate
form part of a lowstand systems tract, no trans­ of subsidence at this time probably enhanced and
gressive systems tract seems to have developed amplified the fall of relative sea-level on the Askrigg
above, because there is an abrupt transition to Block causing deeper incision and erosion. In
marine mudstones with goniatites, both in the addition, a lower subsidence rate during the relative
Craven Basin and on the Askrigg Block (cf. Wilson , highstand generated less new space for sediment
1960, 1977). This transition suggests a rapid rise of accumulation, which gave less accommodation and a
relative sea-level, flooding and cyclothem abandon­ thinner highstand succession to start with.
ment with no time for development of a 'late' Another contrast is the interpreted development
progradational interval, like that of the E2a(i) cyclo­ of a lowstand, turbiditic wedge during E2a(i) time
them, at least in the study area. whereas a falling stage, forced regression delta
Therefore, in summary, within the E2c2 cyclo­ evolved during E2c2 time. These depositional
them, falling-stage sediments are interpreted to patterns follow closely the shelf-edge and ramp
predominate in the basin , whereas on the block, models of van Wagoner et al. (1988). However, in
lowstand systems tract sediments dominate above addition , in the higher subsidence situation of the
a complex lower erosional surface. Highstand E2a(i), the response time of the sedimentary system
sediments were probably eroded on the block, to falling relative sea-level was longer. This longer
whereas in the basin they are restricted to a thin response time may have slowed regression during
interval immediately overlying the Nuculoceras E2a(i) time. The opposite was probably the case for
nuculum marine band (Fig. 8). Transgressive sys­ the E2c2 delta. A similar conclusion was reached by
tems tract sediments are condensed, and probably Holdsworth & Collinson (1988, p. 138), where the
occur as a thin zone between the fluvial sediments response time of sand deposition to lowered sea­
and the overlying multi-cyclothem, goniatite-bearing level was related to bathymetry, the latter being
condensed section (Fig. 8). a combined effect of compaction of previously
deposited sediments, eustasy and subsidence.
Sequence model
Choice of sequence stratigraphical model
Introduction
Local sequence stratigraphical models should fulfil
The E2a(i) and E2c2 cyclothems illustrate different the same requirements as facies models (Reading,
stratigraphical architectures as responses to con­ 1986; Walker, 1992). Thus, a sequence strati-·
tJ·asting physiographic relationships between the graphical model for the E2a(i) and E2c2 cyclothems
Askrigg Block and the Craven Basin. The contrast will have to serve several purposes: (i) act as a norm
evolved as a result not only of decreasing differential, so that the model can be compared with sequence
 Millstone Grit cyclicity 321

models from other areas; (ii) the model should most other Millstone Grit cyclothems deposited near
be rigorous enough to be a framework for future the basin margin (Wright et al. , 1927; Stephens
studies; (iii) it should aid prediction of sequence et al. , 1953; Wilson, 1960; Collinson, 1988;
stratigraphical architecture in adjacent areas; Holdsworth & Collinson, 1988; Martinsen, 1993),
and (iv) it should be a basis for interpretation of there is a pronounced lithological asymmetry (see
Namurian sequence stratigraphy. Figs 6 & 8; cf. also modelling by Collier et al. , 1990).
The models described here are considered to fulfil The asymmetry developed in the thicker Millstone
these requirements, and they are modified versions Grit cyclothems contrasts with the symmetry of the
of the general models of Galloway (1989) and Exxon basinal cycles at Ashover (Rams bottom et al. , 1962;
(e.g. Jervey, 1988). Several local factors prevent the Holdsworth & Collinson, 1988; Fig. 4). The Ashover
general models from being applied as templates in faunal cycles were probably controlled by salinity
the present situation, and modified versions are (see above; Holdsworth & Collinson, 1988).
required. First, erosion surfaces at the bases of the Although salinity probably fluctuated symmetrically
incised fluvial complexes are laterally discontinuous with time (Fig. 4; discussion above), the supply of
and cannot be correlated with confidence between coarser sediments did not. Consequently, rather
individual outcrops (e.g. through palaeosols), nor in than salinity, sand deposition must have been related
a proximal- distal direction (cf. Martinsen, 1993). to other mechanisms, such as supply variations and
Secondly, goniatite-bearing marine bands are the subsidence. Therefore, relative sea-level rather than
most laterally continuous stratigraphical horizons eustasy controlled the succession in the Askrigg
(see above), and are interpreted to be the levels Block and Craven Basin area. Because of the added
across which the most important palaeogeographical influence from subsidence, relative sea-level curves
reorganizations and stratigraphical breaks occur are characterized by longer periods of rise (e.g.
(cf. Figs 7 & 9). The erosion surfaces are recognized Fig. 6A). The sedimentary expression and character
as lowstand surfaces of erosion and compare with of the succession at Ashover would, therefore, be
Exxon-type sequence boundaries, but they are not expected to be different from the Askrigg Block to
considered as the most prominent breaks in the Craven Basin succession, even if sediment input was
succession, nor are they laterally continuous enough similar.
to be reliable for correlation (Martinsen, 1993). For Given the inherent nature of a glacial world, with
example, the lowstand erosion surfaces are not re­ major glaciations in Gondwana (Caputo & Crowell,
cognizable in the Craven Basin (Figs 6A-C & 8). 1985 ; Veevers & Powell, 1987), glacigenic sea-level
Thus, the goniatite bands are the most sensible falls tend to be much slower than corresponding sea­
stratigraphical levels to pick as sequence boundaries. level rises due to slow ice-cap build-ups and rapid
Thirdly, since lowstand wedges of sediment occur melt-downs (e.g. Williams, 1988). As pointed out by
both in the E2a(i) and E2c2 cyclothem, the Galloway Read (1991), although periods of eustatic sea-level
(1989) model cannot be applied in a straightforward fall were relatively short, the rate at which sea-level
way because such units are not emphasized in that fell must have been high enough to overcome sub­
model. sidence rate and produce a relative sea-level fall.
However, Xue & Galloway (1993) described low­ During these periods, sandy systems filled the basins.
stand prograding complexes from both shelf-edge The rate of relative sea-level rise must have been
margins and ramp margins in a study from the extremely rapid, particularly in basinal areas,
Songliao Basin in China, thus extending the original where it was amplified by high subsidence rates.
Galloway (1989) model. The Xue & Galloway (1993) Consequently, in low-lying coastal plains, which
model is therefore a reasonable comparison for the probably formed an important part of the Namurian
models presented herein for the E2a(i) and E2c2 systems (cf. Holdsworth & Colbnson, 1988), the
cyclothems, with the added complexity of a forced rate of coastal transgression must have been high,
regressive unit/falling stage systems tract for the with little possibility for the sedimentary systems to
E2c2 cyclothem. leave significant sandy deposits behind. Also, the
generally observed lack of significant wave or tidal
action decreased the possibility for significant
Cyclothem asymmetry, glacio-eustasy and
reworking of the previously deposited sediments.
spatial development of systems tracts
These relationships suggest that the asymmetric
In the E2a(i) and the E2c2 cyclothems, as well as in lithological response to relative sea-level change was
322 0 .J. Martinsen, J. D. Collinson and B. K. Holdsworth

inversely related to the time needed to deposit the 'mesothems'. The mesothem theory has been subject
facies. The thick, coarsening upward successions in to severe criticism and its validity questioned (see
the basins were probably deposited in a relatively Holdsworth & Collinson (1988) and Martinsen
short time, whereas the thin condensed mudstones (1990) for a thorough discussion). It is proposed that
and marine bands record most of the time rep­ major cyclicity exists, but in a different form than
resented by each Millstone Grit cyclothem. mesothems.
This asymmetry contributed to the areal distri­
bution of systems tracts for both the 'block-edge'
Nam urian maj or-scale cyclicity
and ramp situation. The rapid transgression resulted
in a very thin transgressive systems tract, with, at the Within the Namurian E1c- H2c succession of the
most, only a thin shallow-water delta on top as a last Craven-Askrigg region, three regionally correlative
sedimentary response to the accelerating sea-level stratigraphical intervals can be defined. In each are
rise. The transgressive systems tract developed a several minor cycles, which are condensed, represent
recognizable thickness only on the Askrigg Block. non-deposition and/or are entirely fine-grained
The highstand systems tract of the two cyclothems, (Fig. 10). Interval 1 extends from cyclothem E2b1(i)
which is mudstone-dominated in the study area, to cycle E2b2(ii) and contains five minor cycles,
is probably better developed to the north (discuss­ interval 2 extends from cyclothem E2c2(ii) to
ion above), closer to the palaeoshoreline. The ex­ cyclothem H1b(iii) and contains nine minor cycles,
pectedly slower rate of glacio-eustatic sea-level and interval 3 extends from cyclothem H2a(i) into
fall resulted in a slower rate of relative sea-level fall, the Kinderscoutian (R1) and contains five minor
and a well-developed lowstand systems tract in the cycles to the base of the Kinderscoutian.
Craven Basin. However, even in an ideal situation, Common to these three, multi-cyclothem con­
with no along-strike displacement and variability of densed intervals, is that they are sand-free and only
systems tracts, no one vertical section is expected to small thicknesses of fine-grained sediments were
show a complete development of a marine-band­ deposited. Only locally within one section can evi­
bounded sequence. The systems tracts are displaced dence for subaerial exposure be documented, and
either down-dip or up-dip with relation to each erosion has not been proven. Therefore, these fine­
other (Figs 6A & 8), and their complete presence in grained intervals are mainly of subaqueous origin.
any one vertical section is unlikely. With the added There are ambiguities to some of the correlations,
complexity of a strike-variable development of particularly on the Askrigg Block where few goniatite
systems tracts (see Fig. 6A -C; compare cross­ bands occur. Nevertheless, the general framework
sections), a common situation along any shoreline seems to hold, not least because the intervening
(Martinsen, 1994), it is self-evident that only very cycles are dominated by sandstones, which them­
few two-dimensional cross-sections would show a selves can be correlated regionally both because
complete development of systems tracts. of adjacent goniatite faunas, classic mapping and
their distinctive depositional environments (cf.
Martinsen, 1990).
MULTI-CYCLOTHEM Based on these multi-cyclothem condensed mud­
CONDENSED SECTIONS AND stone intervals, it is suggested that three major
MAJOR SCALE CYCLICITY cycles of deposition can be defined within the E1c­
H2c succession (Fig. 10). Major cyclothem 1 (MC1)
comprises the E1c- E2a(iii) minor cycles, bounded
Background
below by the sub-E1c unconformity on the Askrigg
Although the basic cyclothem pattern ('minor cycles' Block (ct. Dunham & Wilson, 1985) and the base of
of Holdsworth & Collinson, 1988) of the Millstone the Cravenoceras malhamense goniatite band in
Grit has been established since Wright et a!. 's (1927) the basin, and above by the base of the E2b1(i)
classic Rossendale Anticline Memoir, the issue of a Cravenoceras subplicatum group marine band. This
cyclicity at a larger scale has been more actively band forms the base of the lowest multi-cyclothem
debated recently (cf. also Read (1991) for a dis­ mudstone interval. All the cycles in MC1 contain
cussion of cyclicity at several orders in the Namurian sandstones.
of the Midland Valley of Scotland). In 1977, Major cyclothem 2 (MC2) encompasses minor
Ramsbottom argued that Millstone Grit cyclothems cycles E2b1(i)- E2c2, and is bounded below by
could be grouped into larger scale cycles, termed the base of the Cravenoceras subplicatum group
 Millstone Grit cyclicity 323

H l bliv)

SE Askrigg Block .----- 1 5-20km --------. NE Craven Basin

Base E2c2(i?)­
unconformity

MC2

Base E 1 c­
unconformity

\ E1c

Craven Fault System

\
� 1 50
\
MC1

t om

• Multi-cycle mudstones
III:t!Il Dominantly sandstones
D Other mudstones and siltstones
Minor Major
cycles cycles

Fig. 10. Proposed division of the Namurian Elc-H2b stratigraphy into major cyclothems. See text for discussion.

marine band, and above by the base of the second

Maj or c ycl othem architecture and genesis
Nuculoceras nuculum marine band at the base of the
E2c2(ii) cyclothem, which also forms the base of In only one major cyclothem (MC2) is there indi­
the second multi-cyclothem condensed interval. cation of a systematic organization of the contained
Whereas the lower five cycles in MC2 are sandstone­ minor cycles. The E2b3, E2c1 and the one sandy
free, the upper three contain sandstones (see below). E2c2 cyclothem show an increasingly progradational
Major cyclothem 3 (MC3) contains ten minor character (cf. Martinsen, 1993). In addition,
cycles comprising E2c2(ii) to H1b(iv). The lower Martinsen (1993) argued that the lowstand erosion
boundary is the base of the second Nuculoceras surface below the E2c2 Lower Follifoot Grit on the
nuculum marine band at the base of the E2c2(ii) Askrigg Block was, in addition to the unconformity
cyclothem, whereas the top boundary is defined at below the E1c Grassington Grit, the most prominent
the base of the Homoceras smithi/ Hudsonoceras within the E1- H2 part of the Millstone Grit in this
proteus goniatite band at the base of the third multi­ area. Therefore, it is probable that the sequential
cyclothem condensed interval. Only the uppermost evolution of the MC2 cycles reflects progressively
cyclothem, H1b(iv), contains sandstones, whereas lower sea-level with time. Such a pattern is compar­
the underlying nine make up the second multi­ able to a progradational sequence-set architecture,
cyclothem condensed interval. as described by Mitchum & Van Wagoner (1990),
and may reflect longer term sea-level changes than
324 0.1. Martinsen, J.D. Collinson and B . K. Holdsworth

those at minor cycle scale. However, none of Askrigg Block and Craven Basin area have been
the other proposed major cycles show a systematic re-evaluated in terms of architectural and strati­
stacking pattern. graphical response to relative sea-level change.
Although there is no direct reason to argue against 2 Goniatite-bearing marine bands are thought to
sea-level changes as the main control on the major represent the highest eustatic sea-level evidenced by
cycles, in most of the multi-cyclothem condensed salinity data. This interpretation contrasts with the
sections, the basic cyclicity is still present, with conventional sequence stratigraphical models where
faunal bands being interbedded with barren mud­ faunal concentrated bands are considered to
stones (e.g. Wilson & Thompson, 1959; Arthurton represent maximum transgression and flooding.
et al. , 1988). This suggests that lowstands were still Instead, thin , condensed shale below the marine
reducing salinities below the critical marine-band bands probably represent maximum rate of rise of
threshold. In addition , there is evidence for subaerial sea-level.
exposure in the E2b1(i)- E2b2(ii) condensed section 3 The E2a(i) cyclothem developed when there
in the Colsterdale Marine Beds on the eastern side was a significant bathymetric contrast between the
of the Askrigg Block (Wilson & Thompson, 1959). Askrigg Block and the Craven Basin. Prolonged
Therefore, it appears that there are no major changes condensation took place, and clastic deposition did
to the sedimentation and cyclicity patterns beside not occur until the late part of the highstand of sea­
the decreased thickness and the lack of sand. level. Subsequent lowstand sediments are developed
Possibly, one would expect a change of the cyclicity both on the Askrigg Block as fluvial systems and
patterns, particularly in the condensed sections (the in the basin as a turbiditic slope apron. The trans­
assumption then being that these would represent a gressive systems tract is very thin due to rapid trans­
particularly high sea-level), if major-scale sea-level gression and condensation in the basin.
changes caused the major cyclicity. 4 The E2c2 cyclothem developed at a time when
Major changes of sediment supply are more likely there was a gently inclined ramp basinward of the
to have caused the major cyclicity. Within the Askrigg Block. Prolonged condensation and/or non­
E2c2(ii)- H1b(iii) cyclothems, which form a con­ deposition took place until sea-level started to fall
densed section, particularly in the eastern part of and a forced regression delta formed. This delta was
the Craven-Askrigg area, the mudstones thicken incised during the subsequent sea-level lowstand ,
dramatically towards the west, include turbidite and fluvial deposition took place o n the Askrigg
sandstones and reach 165 m in thickness (Wilson Block. No underlying highstand or overlying trans­
et al. , 1989). This relationship suggests that a major gressive systems tract sediments are preserved or
switch of sediment supply to the west occurred within were deposited.
the E2c2(ii)- H1b(iii) time interval, and probably 5 Major cycles of deposition are interpreted to exist
explains the condensation in the Craven -Askrigg in the Namurian of the Askrigg Block and Craven
area. It is suggested also that the other multi­ Basin area. The cycles are bounded by condensed,
cyclothem condensed sections were formed by large­ multi-cyclothem mudstone levels. Except in one
scale switches of sediment supply. To the east, over case, the major cycles show no systematic architec­
the southern North Sea, large Namurian depocentres ture, suggesting they were not related to major
are known to have been present. However, the changes of relative sea-level. Instead, they prob­
stratigraphical subdivision in this area is yet not ably represent major avulsive shifts of sediment
refined enough to test the hypothesis that those depocentres.
areas were alternative sites of sediment accumu­
lation . Therefore, a sequence stratigraphical scheme
at major cyclothem scale is not considered to be ACKNOWLEDGEMENTS
possible at present.
We wish to thank numerous colleagues who have
inspired and guided our thinking about the Millstone
CONCLUSIONS Grit both in the UK and in Ireland, particularly
Harold Reading, Trevor Elliott, Roger Walker and
1 To illustrate the lateral and temporal develop­ Colin Jones. Ole Martinsen wishes to thank Total
ment of Namurian Millstone Grit cyclothems , the Norway for sponsoring 3 years of doctoral research,
Arnsbergian E2a(i) and E2c2 cyclothems in the Norsk Hydro for sponsorship of 3 years of post-
 Millstone Grit cyclicity 325

doctoral research , and Frank Heggernes and W . A. Read. Geol. J. , 27, 173 - 180.
Karianne Krag Kjos for providing new ideas about CRoss, T.H. & GARDNER, M . H . ( 1991) Base level,
accommodation space and high-resolution sequence
the northwest quarter of the Askrigg Block. The
stratigraphy of the Ferron Sandstone (Cretaceou s).
referees Charles Bristow, Robert L. Gawthorpe and Utah. In: Proceedings of the NUNA Conference on
Peter J. McCabe are thanked for reviewing the High-Resolution Sequence Stratigraphy (Eds Leckie,
manuscript. D . A . , Posamentier, H.W. & Lovell, R.V.V.), pp. lO­
l l . Geological Association of Canada, St J ohn's.
DuNHAM, K . C. & WILSON , A . A . ( 1985) Geology of the
Northern Pennine Orefield : Volume 2, Stainrnore to
REFERENCES
Craven. Econ. Mem. geol. Surv. G.B., Eng. Wales, 247
pp.
AITKENHEAD, N. ( 1977) The Institute of Geol ogical EMBRY, A.F. & PODRUSKl, J.A. (1988) Third-order depo­
Sciences borehole at Duffield, Derbyshire. Bull. geol. sitional sequences of the Mesozoic succession of Sverdrup
Surv. G. B. , 59, 1-38. Basin. In: Sequences, Stratigraphy, Sedimentology:
ANDERTON , R., BRlDGES, P.H., LEEDER, M.R. & Surface and Subsurface (Eds James, D . P. & Leckie,
SELLWOOD, B.W. ( 1979) A Dynamic Stratigraphy of the D.A . ) , Mem. Can. Soc. Petrol . Geol . , Calgary, 15,
British Isles: a Study in Crustal Evolution . George Allen 73 -84.
& Unwin, London, 301 pp. FISK, H.N. ( 1944) Geological Investigation of the Alluvial
ARTHURTON, R.S . , JoHNSON, E.W. & MuNDY, D.J.C. Valley of the Lower Mississippi River. Missi ssippi River
( 1988) Geology of the country around Settle. Mem. Commission, Vicksburg, MS, 78 pp.
geol. Surv. G.B., Eng. Wales, Explan. Sheet, 60, GALLOWAY, W.E. (1989) Genetic stratigraphic sequences
147 pp. in basin analysis. I : architecture and genesis of flooding­
BAINES, J.G. ( 1977) The stratigraphy and sedimentology of surface bounded depositional units. Bull. Am. Assoc.
the Skipton Moor Grits (Namurian Elc) and their lateral petrol. Geol. , 73, 125 - 142.
equivalents. Unpubli shed PhD thesi s, Keele University. GAWTHORPE, R.L., GuTIERIDGE, P. & LEEDER, M . R. (1989)
BATES, R.L. & JACKSON, J . A. (1987) Glossary of Geology, Late Devonian and Dinantian basin evolu tion in northern
3rd edn. American Geological Institute, Alexandria, England and North Wales. In: The Role of Tectonics in
Va, 788 pp. Devonian and Carboniferous Sedimentation in the British
BISAT, W.S. ( 19 14) The Millstone Grit Series between Isles (Eds Arthurton, R.S., Gutteridge, P. & Nolan,
Ma sham and Great Whernside. Proc. Yorks. geol. Soc. , S.C.), Occas. Pubis Yorks. geol. Soc., 6, 1-23 .
19, 20- 24. GuTIERlDGE, P. (1989) Control s on sedimentation in a
BISAT, W.S. ( 1924) The Carboniferous goniatites of the Brigantian intrashelf basin (Derbyshire). In: The Role of
north of England and their zones. Proc. Yorks. geol. Tectonics in Devonian and Carboniferous Sedimentation
Soc. , 20, 40- 124. in the British Isles (Eds Arthurton, R.S., Gu tteridge, P.
BLUM, M.D. ( 1994) Genesis and architecture of alluvial & Nolan, S.C.), Occ. Pubis Yorks. geol. Soc., 6,
stratigraphic sequences: a late Quaternary example from 171- 187.
the Colorado River, Gulf Coastal Plain of Texas. H oLDSWORTH, B.K. ( 1963) Pre-fluvial, autogeosynclinal
In Siliciclastic Sequence Stratigraphy: Recent Develop­ sedimentation in the Namurian of the Sou thern Central
ments and Applications (Eds Weimer, P. & Posamentier, Province. Nature (London) , 199, 133-135.
H.W.) Mem. Am. Assoc. Petrol. Geol., 58, 259-283. HOLDSWORTH, B.K. & COLLINSON , J.D. (1988) Millstone
CAPUTO, M.V. & CROWELL, J.C. ( 1985) Migration of glacial Grit cyclicity revisited. In: Sedimentation in a Synoro­
centers across Gondwana during the Paleozoic Era. genic Basin Complex: the Upper Carboniferous of
Geol. Soc. Am. Bull. , 96, 1020 - 1036. Northwest Europe (Eds Besly, B.M. & Kelling, G.),
COLLIER, R.E. LL., LEEDER, M . R. & MAYNARD, J . R. pp. 132-152. Blackie, Glasgow.
(1990) Transgressions and regressions: a model for the HuDSON, R. G.S. (1934) The Mill stone Grit succession
influence of tectonic subsidence, deposition and eu stasy, south of Harrogate. Trans. Leeds geol. Assoc. , 5,
with application to Quaternary and Carboni ferous 118-124.
examples. Geol. Mag. , 127, 117-128. HuDSON, R.G.S. ( 1939) The Millstone Grit succession of
CoLLINSON, J.D. ( 1969) The sedimentology of the the Simonseat Anticline. Proc. Yorks. geol. Soc. , 23,
Grindslow Shales and the Kinderscout Grit: a del taic 3 19.
complex in the Namurian of northern England. J. HuDSON, R.G .S. (1941) The Mirk Fell Beds (Namurian
sediment. Petrol. , 39, 194-221. E2) of Tan Hill, Yorkshire. Proc. Yorks. geol. Soc. , 24,
COLLINSON , J . D. ( 1988) Controls on Namurian sedimen­ 259-289.
tation in the Central Province basins of northern HuDSON, R.G.S. ( 1944) The faunal succession in the
England. In: Sedimentation in a Synorogenic Basin Ct. nitidus zone in the mid-Pennines. Proc. Leeds Philos.
Complex: the Upper Carboniferous of Northwest Europe Lit. Soc. , Sci. Sect. , 4, 233-242.
(Eds Besly, B.M. & Kelling, G.), pp. 85 -101. Blackie, HuNT, D. & TucKER, M . E. ( 1992) Stranded parasequences
Glasgow. and the forced regressive wedge systems tract: deposition
COLLINSON, J.D., HOLDSWORTH, B.K., J oNES, C.M. & during base-level fall. Sediment. Geol. , 8 1 , 1-9.
MARTINSEN, O.J. (1992) Discussion of 'The Millstone JERVEY, M.T. (1988) Quanti tative geological modeling of
Grit (Namurian) of the southern Pennines viewed in the siliciclastic rock sequences and their seismic expression.
light of e ustatically controlled sequence stratigraphy by In: Sea-Level Changes: an Integrated Approach (Eds
326 0.1. Martinsen, J.D. Collinson and B. K. Holdsworth

Wilgus, C. K . , Hastings, B.S. , Kendall, C.G.StC., Part II: the Mountain Limestone District. John Murray,
Posamentier, H.W. , Ross, C.A. & Van Wagoner, J.C.), London, 253 pp.
Spec. Pubis Soc. econ. Paleontol . Mineral . , Tulsa, 42, PuNT, A.G. (1988) Sharp-based shoreface sequences an d
47-69. 'offshore bars' in the Cardium Formation of Alberta:
JONES, C.M. (1980) Deltaic sedimentation in the Roaches their relationship to relative changes in sea-level. I n :
Grit and associated sediments (Namurian R2b) in the Sea-Level Changes: an Integrated Approach (Eds Wilgus,
southwest Pennines. Proc. Yorks. geol. Soc. , 43, 39-67. C . K . , Hastings, B. S . , Kendall , C.G.StC . , Posamentier,
LEEDER, M . R. (1982) Upper Palaeozoic basins of the H.W. , Ross, C.A. & Van Wagoner, J.C.), Spec. Pubis
British I sles: Caledonide inheritance versus Hercynian Soc. econ. Paleontol. Mineral. , Tulsa, 42, 357-370.
plate margin processe s. J. geol. Soc., London, 1 39 , PosAMENTIER, H.W. , JERVEY, M.T. & VAIL, P . R. (1988)
479-491. Eu static controls on clastic deposition I - conceptual
LEEDER, M . R. (1988) Recent developments in Carbonifer­ framework. In: Sea-Level Changes: an Integrated
ous geology: a critical review with implications for the Approach (Eds Wilgus, C.K. , Hastings, B.S . , Kendall,
British I sles and N.W. Europe. Proc. Geo/. Assoc. , 99, C.G .StC . , Posamentier, H.W. , Ross, C.A. & Van
73 - 100. Wagoner, J . C . ) , Spec. Pubis Soc. econ. Paleontol.
LEEDER, M . R. & McMA HON , A . H. (1988) Upper Carbon­ Mineral . , Tulsa, 42, 1 09-124.
iferou s (Silesian) basin subsidence in northern Britain. POSAMENTIER, H.W. , ALLEN , G.P. , JAMES, D . P . & TESSON ,
In: Sedimentation in a Synorogenic Basin Complex: the M . (1992) Forced regressions i n a sequence stratigraphic
Upper Carboniferous of Northwest Europe (Eds Besly, framework: concepts, examples, and exploration signifi­
B.M. & Kelling, G . ) , pp. 43-52. Blackie, Glasgow. cance. Bull. Am. Assoc. petrol. Geol. , 76, 1687-1709.
MARTINSEN, O.J. (1990) Interaction between eustasy, RAMSBOITOM, W . H.C. (1977) Major cycles of transgression
tectonics and sedimentation with particular reference to and regression (mesothems) in the Namurian. Proc.
the Namurian Elc-H2c of the Craven-Askrigg area, Yorks. geol. Soc. , 41, 261-291.
northern England. Unpubl. Dr. Sci. thesis, University of RAMSBOITOM, W.H .C. (1979) Rates of transgression and
Bergen, Norway, 495 pp. regre ssion in the Carboniferous of NW Europe. J. geol.
MARTINSEN, O.J. (1993) Namurian (late Carboniferous) Soc. , London, 1 36 , 147-154.
depositional systems of the Craven-Askrigg area, RAMSBO'ITOM, W.H.C. , RHYS, G . H . & S M ITH , E . G . (1962)
northern England: implications for sequence strati­ Boreholes in the Carboniferou s of the Ashover district,
graphic models. In: Stratigraphy and Facies Associ­ Derbyshire. Bull. geol. Surv. G. B. , 19, 75 - 168.
. ations in a Sequence Stratigraphic Framework (Eds RAMSBOlTOM, W.H.C. , CALVER, M . A . , EAGAR, R.M.C. ,
Posamentier, H.W. , Summerhayes, C . P . , Haq, B. U . & HODSON , F. , HOLLIDAY, D.W. , STUBBLEFIELD, C. et a/.
Allen, G.P.), Spec. Pubis int. Ass. Sediment. , 18, (1978) A correlation of Silesian rocks in the British I sles.
247-281. Spec. Rep. geol. Soc. , London, 10, 81 pp.
MARTINSEN, O.J. (1994) Sequence stratigraphy, three READ, W.A. (1991) The Millstone Grit (Namurian) of the
dimensions and philosophy. In: Sequence Stratigraphy: sou thern Pennines viewed in the light of e ustatically
Advances and Applications for Exploration and Pro­ controlled sequence stratigraphy. Geol. J. , 26, 157-165.
duction in Northwest Europe (Eds Steel, R.J. et a/.) , READING, H.G. (1986) Facies. In: Sedimentary Environ­
Spec. Pubis Nor. petrol. Soc . , 5 , 23 - 29. ments and Facies (Ed. Reading, H.G . ) . Blackwell
MAYNARD, J. R. (1992) Sequence stratigraphy of the Upper Scientific Publications. Oxford, pp. 4 - 19 .
Yeadonian of northern England. Mar. petrol. Geol. , 9 , RILEY, N.J. (1985) Asturoceras and other dimorphoceratid
197 - 207. ammonoids from the Namurian (E2) of Lancashire.
MAYNARD, J . R. & LEEDER, M . R. (1992) On the periodicity Proc. Yorks. geol. Soc. , 45, 219-224.
and magnitude of Late Carboniferou s glacio-eustatic RILEY, N.J . , VARKER, W . J . , OWENS, B . , HIGGINS, A . C &
sea-level changes. J. geol. Soc., London, 149, 303-311 . RAMSBOTTOM, W.H.C. (1987) Stonehead Beck, Cowling,
MILLER, J. & GRAYSO N , R.F. p 982) The regional North Yorkshire, England: a British proposal for the
context of Waulsortian facies in northern England. In: Mid-Carboniferous boundary stratotype. Cour. Forsch. ­
Symposium on the Palaeoenvironmental Setting and Inst. Senckenberg, 98, 159-177.
Distribution of Waulsortian Facies (Eds Bolton, K. , SAUCIER, R.T. (198 1 ) Current thinking on riverine
Lane, H . R. & Lemone, D . V . ) , pp. 17-33. El Paso processes and geologic hi story are related to human
Geological Society/University of Texas at El Paso. settlement in the southeast. Geosci. Man, 22, 7-18.
MITCHU M , R.M. JR. & VAN WAGONER, J.C. (1990) High­ ScHUMM, S.A. (1993) River response to baselevel change:
frequency sequences and e u static cycles in the Gulf of implications for sequence stratigraphy. J. Geol. , 101 ,
Mexico Basin. In: Sequence Stratigraphy as an Explo­ 279-294.
ration Tool: Concepts and Practices in the Gulf Coast. SCHUMM, S . A . , MOSLEY, M . P. & WEAVER, W.E.
Programs and Abstracts, 11th Annual Research Confer­ (1987) Experimental Fluvial Geomorphology. Wiley,
ence, Gulf Coast Section, Society of Economic Paleon­ New York, 413 pp.
tologists and Mineralogists, Houston, pp. 257-267. SHANLEY, K.W. & Mc CABE , P . J . (1993) Alluvial archi tec­
MoSELEY , F. (1954) The Namurian of the Lancaster Fells. ture in a sequence stratigraphic framework: a case hi story
Q. J. geol. Soc. , London, 109, 423-454. from the Upper Cretaceous of southern Utah, USA. In:
MoSELEY, F. (1956) The geology of the Keasden area, west The Geological Modelling of Hydrocarbon Reservoirs
of Settle, Yorkshire. Proc. Yorks. geol. Soc. , 30 , and Outcrop Analogues (Eds Flint, S.S. & Bryant, I . D . ) ,
331-352. Spec. Pubis int. Ass. Sediment. , No. 15, p p . 21-56.
PHILLIPS, J. (1836) Illustrations ofthe Geology of Yorkshire, Blackwell Scientific Publications, Oxford.
 Millstone Grit cyclicity 327

SHEPHERD, R.G. ( 1979) River channel responses to bedrock J. sediment. Petrol. , 36, 90- 1 14 .
lithology and stream capture, Sandy Creek Drainage, WALKER, R.G. ( 1990) Facies modeling and sequence stra­
central Texas. In: Adjustments of the Fluvial System tigraphy. J. sediment. Petrol. , 60, 777-786.
(Eds Rhodes, D . D . & Williams, G . P . ) , pp. 255-276. WALKER, R . G . ( 1 992) Facies, facies models and modern
Kendall/Hunt, Dubuque, Ia. stratigraphic concepts. In: Facies Models - Response to
STEPHENS, J.V., MITCHELL, G . H . & EDWARDS, W. ( 1953) Sea-Level Change (Eds Walker, R . G . & James, N . P . ) ,
Geology of the country between Bradford and Skipton. p p . 1 - 14 . Geological Association o f Canada, St John's.
Mem. geol. Surv. G. B. , Eng. Wales, Explan. Sheet, 69, WILLIAMS, D .F. ( 1 988) Evidence for and against sea-level
180 pp. changes from the stable isotopic record of the Cenozoic.
THUNELL, R . C . , LOCKE, S . M . & WILLIAMS, D . F. ( 1 988) In: Sea-Level Changes: an Integrated Approach (Eds
Glacio-eustatic sea-level control on Red Sea salinity. Wilgus, C.K . , Hastings, B . S . , Kendall , C.G .StC . ,
Nature, 334, 601 -604. Posamentier, H.W. , Ross, C . A . & Van Wagoner, J . C . ) ,
TREWIN , N.H. & HoLDSWORTH, B . K . ( 1 973) Sedimentation Spec. Pubis Soc. econ. Paleontol. Mineral. , Tulsa, 42,
in the lower Namurian rocks of the North Staffordshire 31 -36.
Basin. Proc. Yorks. geol. Soc. , 39, 3 7 1 - 408. WILSON , A.A. ( 1960) The Millstone Grit Series of
UNDERHILL, J . R . , GAYER, R . A . , WOODCOCK, N . H . , Colsterdale and neighbourhood, Yorkshire. Proc.
DoNNELLY, R . , JOLLEY , E . J . & STIMPSON, I . G . Yorks. geol. Soc. , 32, 429-452.
( 1 988) The Dent Fault System, northern England - WILSON , A.A. ( 1977) The Namurian rocks of the Fewston
reinterpreted as a major oblique-slip fault zone. J. geol. area. Trans. Leeds geol. Assoc. , 9, 1 -42.
Soc. London, 145, 303 - 3 1 6 . WILSON , A . A . & THOMPSON, A.T. ( 1 959) Marine bands of
VAN WAGONER, J.C. , POSAMENTIER, H.W., MITCHU M , R.M. Arnsbergian age (Namurian) in the south-eastern portion
JR. , VAIL, P.R., SARG , J.F., LouTIT, T.S., et al. ( 1 988) of the Askrigg Block, Yorkshire. Proc. Yorks. geol.
An overview of the fundamentals of sequence stra­ Soc. , 32, 45-67.
tigraphy and key definitions. In: Sea-Level Changes: an WILSON, A . A . , BRANDON , A . & JOHNSON, E . W . ( 1 989)
Integrated Approach (Eds Wilgus, C . K . , Hastings, B . S . , Geological Context of the Wyresdale Tunnel Methane
Kendall, C.G .StC . , Posamentier, H . W . , Ross, C . A . & Explosion . Geological Survey of Great Britain,
Van Wagoner, J.C.), Spec. Pubis Soc. econ. Paleontol. Technical Report, WA/89/30, HMSO, London.
Mineral . , Tulsa, 42, 39-45. WRIGHT, W . B . , SHERLOCK, R.L. , WRAY , D . A . , LLOYD, w.
VAN WAGONER, J . C . , MITCHU M , R.M. JR . , CAMPION , K . M . & ToNKS, L.H. (1927) The geology of the Rossendale
& RAHMANIAN , V . D . ( 1 990) Siliciclastic sequence stra­ Anticline. Mem. geol. Surv. G. B . , Eng. Wales, 182 pp.
tigraphy in well-logs, cores and outcrops. Am. Assoc. XuE, L. & GALLOWAY , W.E. (1993) Genetic sequence
petrol. Geol. Meth. Explor. Ser. , 7, 55 pp. stratigraphic framework , depositional style, and hydro­
VEEVERS, J.J. & POWELL, C.McA. ( 1 987) Late Paleozoic carbon occurrence of the Upper Cretaceous QYN
glaciations in Gondwanaland reflected in transgressive­ formations in the Songliao lacustrine basin, northeastern
regressive depositional sequences in Euramerica. Geol. China. Bull. Am. Assoc. petrol. Geol. , 76 , 1792- 1808.
Soc. Am. Bull. , 98, 475-487. ZIEGLER, P . A . ( 1 986) Geodynamic model for Palaeozoic
WAGNER, R . H . , HIGGINS, A.C. & MEYEN, S . V . (eds) crustal consolidation of western and central Europe.
(1979) The Carboniferous of the USSR. Occas. Pubis. Tectonophysics, 126, 303-328.
Yorks. geol. Soc . , 4 , 247. ZIEGLER, P .A. ( 1 988) Evolution of the Arctic-North
WALKER, R . G . ( 1966) Shale Grit and Grindslow Shales: Atlantic and the Western Tethys. Mem. Am. Assoc.
transition from turbidites to shallow-water sediments petrol. Geol. , 43, 196 pp.
in the Upper Carboniferous of northern England.
 Facies Analysis in
Reservoir Sedimentology

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
Spec. Pubis int. Ass. Sediment. (1995) 22, 331-354

Productive Middle East clastic oil and gas reservoirs:

their depositional settings and origins of their hydrocarbons

Z I A D R. B E Y D O U N
American University of Beirut, Lebanon

ABSTRACT

The Middle East region contains 67% of the world's reported recoverable oil reserves and 30% of its
reported recoverable gas reserves. The greater bulk of these are housed in carbonate reservoirs of Late
Permian-Miocene age, with sandstone reservoirs of that age range being minor and restricted to
specific areas of the region. The predominance of carbonates since the latest Permian is related to the
plate tectonic position of the region, located principally in low latitudes and warm climatic belts. By
contrast, from the Infra-Cambrian to the Early Permian, the Middle East region was located principally
in middle to high southern latitudes in cool-wet to cold-dry climatic belts, when clastic sedimentation
predominated and carbonates and evaporites were deposited mainly during brief excursions into lower
latitudes and/or during short episodes of aridity in higher latitudes. Only fairly recently has the
petroleum potential of the Palaeozoic clastic reservoirs begun to be properly appreciated after discover­
ies in Oman demonstrated ample sourcing from extensive organic-rich Infra-Cambrian to Early
Cambrian sediments, with both vertical and lateral migration into Cambro-Ordovician and Permo­
Carboniferous sandstone reservoirs. More recently still, important discoveries in Central Arabia have
confirmed sourcing of Permo-Carboniferous sandstone reservoirs from regionally deposited organic-rich
Lower Silurian shales related to transgressive global sea-level rise. Either of these sources, where
overmature, have most probably also provided most of the region's vast Permian Khuff carbonate
reservoir gas reserves.
The depositional environments and relevant petrophysical properties of the productive sandstone
reservoirs of the region are reviewed by stratigraphical position and age and the source rocks charging
these are briefly discussed. It is concluded that tremendous petroleum potential exists in the Middle East
for substantial new Palaeozoic sourced and reservoired hydrocarbons. However, unlike the extensive
lateral facies continuity of the post-Palaeozoic source-reservoir-seal relationships and the giant traps
of the very wide Arabian shelf region (resulting in almost predictable accumulation trends), the
Palaeozoic clastic reservoirs and seals generally exhibit rapid lateral and vertical facies changes coupled
with the generally smaller sizes of local fault-block traps and/or low-amplitude fold traps. Consequently,
the distribution of hydrocarbon accumulations in Palaeozoic sequences is less easily predicted, and in
some areas is critically dependent upon seal distributions. Provided source and 'kitchen' distributions
can be better understood and migration paths to traps better constrained, major additional reserves in
Palaeozoic sandstone reservoirs will be established. Additional discoveries in post-Palaeozoic sandstone
reservoirs are likely to be mainly restricted to the identification of hidden rift plays similar to those of
Yemen and the Euphrates area of Syria.

I N T RO D U C T IO N

The geological boundaries of the Middle East region, west Iran-southeast Turkey, lying, respectively,
as discussed in this survey, are those of the Arabian southwest and south of the Zagros and Bitlis
Plate and the adjacent Levant-Sinai fragment to its (Taurus) sutures, together with the Levant states
northwest (Fig. 1). Geographically, this encompasses and including the Sinai Peninsula but not Egypt
the countries of the Arabian Peninsula and south- proper (Fig. 1).

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
331
332 Z. R. Beydoun

Jo• oo·
LEGEND

r::o-o-ol Palaeozoic @
40° oo· CJ Quaternary-Recent
� (Infra-Cambrian in Oman)

t1lJlll /
Cenozoic C
© k;:���::�·�l PreCambrian

!:::::>:I Mesozoic @ f"V\'VVl Mesozoic-Cenozoic

� (Ophiolites & Basalts}
Igneous Rocks

Approximate extent of currently

hydrocarbon-productive basins
with predominance of sandstone
reservoirs and their ages P.M. C.
(see legend for key)

35° 00'
N.B. Small lener (e.g. pl denotes
only minor production or potential.

--::::::;::; Main fractures or thrusts

New Oceanic Crust

Jo• oo·

20° 00'

SUDAN
··
.r
(
.

1s•oo·

�
)

118506923
40° 00'

Fig. I. Geological sketch map of the Middle East.

 Middle East clastic oil and gas reservoirs 333

The Middle East region, as outlined above (includ­ and southeast Turkey and to the minor production
ing the Gulf of Suez rift basin), at the start of 1992 of Jordan and Israel (Beydoun, 1988). Cumulative
contained a reported 665.6 billion (B) barrels Middle East oil production from all reservoirs since
of recoverable oil and 1325.8 trillion (T) cubic the first Iranian discovery has totalled about 175.3 B
feet (tcf) of recoverable gas (88.51 B tonnes and barrels, making the ultimate recoverable oil reserves
3
37.54 T m respectively) (Oil and Gas Journal, 30 for the region around 841 B barrels of oil (Oil and
December 1991). This is equivalent to 67.2% of the Gas Journal, 30 December 1991). Cumulative gas
World's reported recoverable oil reserves and 30.3% production figures are not discussed here because all
of the reported recoverable gas reserves. In 1991, gas production in the region is at present for local
the region accounted for 28.5% of the World's usage (desalination, power generation, fertilizers,
annual oil production (Oil and Gas Journal. , 30 feedstock, light industry) or for reservoir pressure
December 1991); this figure, however, is determined maintenance by reinjection; some gas is still being
more by economic and engineering constraints than flared during oil and gas separation, pipeline oper­
by limitations on reservoir potential and productive ations and refining. No gas as yet is being exported,
capacity. Oil initially was discovered in the region in although plans to export are well advanced in several
1908, at Masjid-i-Suleiman in southwest Iran in the countries of the region and exploration for non­
Oligo-Miocene carbonate reservoir of the Asmari associated gas to establish additional reserves has
Formation, and that field and others went on pro­ been stepped up.
duction shortly afterwards. Commercial oil was Most of these immense reserves are housed in
found subsequently in Iraq, at Kirkuk in 1927 in a reservoirs predominantly of Permian to Oligo­
similar carbonate reservoir. Oil initially was found Miocene age, of which 80% are carbonate; the bulk
in Egypt at Gemsa field, located onshore on the of these reserves (98.7% of the oil and 98.8% of the
western side of the Gulf of Suez in 1909, again in a gas) are located along the prolific northeast Arabian
carbonate reservoir of Miocene age. This was fol­ shelf. This region had formed the Arabian passive
lowed by the discovery of the first clastic production Tethyan margin of Gondwana (and subsequently of
in the Gulf of Suez at Hurghada field near to Gemsa Afro-Arabia) from the initial cratonization of the
in 1915, from Cretaceous 'Nubia' sandstone reser­ Arabian Shield in the late Proterozoic until the
voirs as well as from a Miocene carbonate reservoir. collision and suturing with Eurasia. Initial collision
On the eastern side of the Arabian Peninsula, the commenced in the medial Tertiary and continued to
next discovery was made in Bahrain in 1932 in the the present, overprinting the Zagros foreland basin
Cretaceous sandstone and carbonate reservoirs of on the outer edge of this margin; the northeast
the Awali field, and in 1938 in eastern Saudi Arabia margin extends from northwest Iraq through south­
at Dammam in Upper Jurassic carbonate reservoirs west Iran and the eastern Arabia-Gulf region to
and also in Kuwait at Burgan in Cretaceous sand­ northern central Oman (Fig. 1).
stone reservoirs. World War II interrupted further The current hydrocarbon-producing reservoirs of
exploration activity, but shortly after it was over a the region and those being developed for production
remarkably rapidly expanding exploration pro­ are shown by country in Table 1. The clastic reser­
gramme commenced in many different parts of the voirs are shown in Fig. 2 and the principal areas of
region and continues to the present; discoveries clastic reservoir production are indicated on Fig. 1.
have been made in every single country of the region Apart from the relatively long established major
except Lebanon, where the protracted phase of civil production areas of the Gulf of Suez ('Nubia'
strife commencing in the early -mid- 1970s has cur­ Palaeozoic-Mesozoic, and Miocene, constituting
tailed all exploration activities. The last country to Egypt's most prolific hydrocarbon province) and the
join the producers club was Yemen in 1984. Basra -Kuwait-northeast Saudi Arabia-Bahrain
The magnitude of the many discoveries and devel­ Cretaceous reservoir areas near the head of the
opments varies considerably, from the 'giant' Arabian-Persian Gulf, the remainder of the import­
(0.5 -4.5 B barrels estimated ultimate recoverable ant producing areas constitute new additions to the
(EUR) reserves of oil or oil equivalent) and 'super­ list, with the Oman Palaeozoic discoveries, many
giant' (> 5 B barrels) fields of Iran, Iraq, Kuwait, dating from the mid- 1970s, the Syria and Yemen
Saudi Arabia, Qatar, the UAE and Oman, to the Mesozoic discoveries, from the mid-1980s, and the
major fields of Syria and Yemen and the Gulf of Central (Saudi) Arabian Upper Palaeozoic discover­
Suez region, to the relatively modest fields of Bahrain ies, from the late 1980s. Other minor clastic reservoir
334 Z. R. Beydoun

Table 1. Petroleum producing formations of the Middle East. Expanded and updated from Beydoun (1991), reproduced
by permission of AAPG

Periods, Epochs, Ages

GULF OF ISRAEL/
(Modified fonn Dunnington 1967) YEMEN JORDAN S.E. TURKEY SYRIA
SUEZ PA L ES T INE

MIQ. MIQ.PLIOCENE
PLIOCENE M · U MIOCENE
L-M MIOCENE Belayim: 0 JOJTibe: O.G.
OLIGO. OGLIQ. Kareem: 0 Chilou: O.G.
MIOCENE Rudeis: 0
Nukhul: 0
EOCENE/ U • M EOCENE Habshiya : 0. Jaddala : 0.
PALAEOCENE L.EOCENE·PALAEDCENE (Gulf of Aden)
Sinan : 0.
MAESTRICHTIAN Sudr: 0 Garzan: 0 Shiranish: O.G.
u U.CAMPANIAN Amah: O.G.
L.CAMPAN-/SANTDNIAN Beloka : 0. ''Massive"
Matulla: 0
c CONIACIAN Karabogaz : 0.
R
E TURONIAN Ajlun Gp: 0. Karababa : 0. Limestone : 0.
T (Hummar & +
A M CENOMANIAN Raha: 0 & Na'ur) Derdere: o. "Judea" : 0.
(Mardin)
c
ALBIAN
E
0 APnAN Nubia "A"
u Qishn
("Malha"): 0
s BARREMIAN Clastics: 0
L
HAUnERIVIAN Heletz: 0 Rutbah : 0.

VALANGINIAN
BERRIASIAN Naifa: O.G.
Alit: O.G.
PORTLANDIAN-
J M. KIMMERIDGIAN
u u
R L. KIMMERIDGIAN
A
s OXFORDIAN/CALLOVIAN Amran : G. Nir Am: 0
s BATHONIAN Brur: O.G
M
I BAJOCIAN Kohlan: 0 (Zohar: G)
c
LIASSIC
L
L. LIASSIC Butmah : 0.
Camurlu: G--+-

Kurachine : O.G.
TRIASSIC
(Mulussa) : G.O.
p PERMIAN
A PERMQ.CARBONIFEROUS
L Markadah : G.
CARBONIFEROUS Nubia "C"
A
E SILURIAN-DEVONIAN Katin : G.O.
Dubaidib: G
0
z CAMBRQ.ORDOVICIAN
0
I
CAMBRIAN I INFRA·
c
CAMBRIAN

0 =Oil G = Gas, as predominant production G after 0 = Gas alone in some fields

Condensate in many of the non-Permian Gasfields

hydrocarbon production in the region occurs in concerned in the interpretive detail reached by some
Turkey (Palaeozoic), Jordan (Palaeozoic) and Israel of the other contributors or, indeed, that which
(Cretaceous). The identified or inferred source rocks might be considered adequate by the dedicatee of
that charge these Middle East reservoirs, both car­ this volume. Nevertheless, the review is offered
bonate and clastic, are indicated in Fig. 2 and those with the intention of drawing out the important
that have a known relevance to sourcing the clastic distinctions necessary in viewing the hydrocarbon­
reservoirs are discussed briefly in the text, together productive clastic sediments of the prolific Middle
with the environmental characteristics and some East region in terms of age, plate tectonics,
relevant petrophysical properties of the reservoirs. palaeolatitude, climate and broader depositional
In a survey and review of this scale, no attempt environments.
has been made to describe the clastic sediments Numerous oil and gas field names are mentioned
 Middle East clastic oil and gas reservoirs 335

Table 1. (Continued. )

KUWAIT & UNITED

OFFSHORE CENTRAL & SAUDI QATAR-
DIVIDED ARAB
N. IRAQ S.W. IRAN IRAN S. IRAQ ARABIA BAHRAIN OMAN
ZONE + OFFSHOR E EMIRATES

Fars: G.
Belayim/Rudois :
"Main" O.G. Asmari: O.G. Ahwaz/Asmar i "Ghar": 0. Ghar: 0 G.O.
limestone : (Kalho<) G.O. (Rod Sea basin)
AvanaiJaddala Jahnum: 0. Dammam : O.G. Pabdeh: G.
/Aaliji: 0. U. Radhuma : 0.

Shiranishl Gurpi : 0. Hartha : 0.

"Pilsner" :0.
llam : 0.
Magwa: 0.
(Mishri� Mishrif: 0. Mishrif: 0. Mishril: 0.

Komotan/ Servak : 0. Rumaila : 0. A hmadi/

Dokan I 0. Ahmadi : 0. Ahmadi: 0. Khatiyah: 0 G.
"Qamchuqa" : Wara: 0. Wara: 0. Wara : 0. Wara: 0 Natih:
Mauddud: 0. Mauddud: 0. Mauddud: 0. Maud d ud: 0 Mauddud : 0. o.
Burgan: 0. Nahr Umr : 0. Burgan: 0. Burgan : 0. Nahr Umr: 0.
Thamama Thamama
Dariyan: G. Shuaiba: 0. Shuaiba: 0 (Shuaiba) : 0. (Shuaibo) : 0.
L. Qamchuqa Biyadh : Kharaib: 0
and 0. Zubair: 0. Zubair : 0. (Zubair) :
0. Thamama: 0.
Garagu 0. Bubwaib: 0.
Fahliyan : Fahliyan : 0.
G. (Yamama) Minagish :0. Yamama : 0.
Yamama : 0. (Ratawi) Sulaiy: 0.

Hith: 0.
: 0.
Arab Arab A�: 0. Arab A�: G.O.
Arab D : 0 Arab D I Arab I
Arab D : 0. Jubaila : 0. Darb: 0. Darb. 0.
Hanila : 0.
Tuwaiq: 0.
Najmah : 0. Dhruma: 0. Araej: 0. Araej : 0.
Mafraq: 0

Marra!: 0. Marrat: 0.

Kurachino : 0.
Kangan: G.
Dalan : G. Dalan: G. ' Khuff: G. Khuff: G. Khuff: G. Khuff: G. Khuff: O,G.
Unayza: O,G. Haushi : 0.
Pre-Khuff : G. (G harif, -
AI Khlata)
Haima: O,G.

Sara: 0 Haima: 0, G
Sajir: 0 (Haradh, Amin,
M ahawis/Andam)
Ara: 0.
Buah: G.

in the text but not given precise location on Fig. 1 (Permian) the Arabian Plate, including the Levant­
because of the scale involved. A better overview of Sinai fragment, was attached to the African Plate,
field locations (with the exception of those in the which together formed part of Gondwana and was
Levant-Sinai fragment) can be obtained from the located in the southern hemisphere; it was part
enclosure map in Beydoun ( 199 1), which was also of the long and very wide shelf adjacent to the
issued with Bull. Am. Assoc. petrol. Geol., 75(7) for 'northern' margin of Gondwana and bordered the
July 1991. Palaeo-Tethys ocean. This wide shelf expanse was
covered intermittently by shallow epeiric seas along­
side lowlands consisting of the cratonized Arabo­
P A L A EO G EO G R APH I C A L A N D Nubian and other shields; the areal extent of these
P A L A EO C L I M A T I C S Y NOP S I S shelf seas changed in response to succeeding trans­
gressions and regressions. Following plate tectonic
From the latest Proterozoic to the latest Palaeozoic reconstructions by Scotese et at. ( 1979) and Scotese
336 Z. R. Beydoun

z
«

<<<
,i II l :�
ffi < <<
:!;
0

��!,� IIU
-�
"
:0

<
��
....
:z:

:0
-=�

<
0

�
<n

�I t ! ! ��� ,
�<
< < :<6& <
<<;;_ <�&I, <
; Q)
I

<<<<<
m

""

< li <II
! <-<�1? I�
� tt
�....
jl<: IJI ' li �: ��
<< I< ' II'\ ' I�
t::
:0

'
w

' '
.,;

1
<<<< << I< 1!11 !l � l·.-:1
!"'

�I<� !f ·� IJ: iI ll '

, ,
«

I
ii' �
>-

<
<n

� i
I "'

l� I i �
�< ' .1 1�:1
� l I

J? hjt �� ! i
� Q)

.i·.
>
�. ..,
u
::l

I
� 'C

!X\
·• 0
Q:
-�,

e� <
N
w

,F

i
:0
<n
If.·�:-:

: l .,:\ � 1111
�

1<<
�
"
:0
-� 1-:-::.i!:·:·:·:·

a:<<< li Ji f!� llI 1_1

r-- ·············••d

�;
1: > : :

I
\1< 87
z

� :�

K!
. Q)

J�
>- u

:;

I �
0
"'
�
:o l � :o l � 'C
Q)

i
:;::
..,
!�
w
w
ili ffi
a: a:
a:
�
c:
u
:l:
0 � � � Q)

·�
0 0
� :0 � g :0 g :!:!
ii
3N3�03N 13N3�03'o'l'o'd sno3::J'<>'l3�-. ::JISSvl:lnr l
 Middle East clastic oil and gas reservoirs 337

& McKerrow (1990), the Arabian Plate (within the influence of warmer climatic belts occurred before
borders defined above) from latest Precambrian the Late Permian, allowing warm water carbonate
to Early Silurian had its 'long axis' (Fig. 1 inset) shelf sedimentation, with or without accompanying
oriented) approximately NE-SW, with its present aridity and deposition of evaporites, for relatively
northern extremity in the southwest (Fig. 3). limited spans of time. For example, during latest
Palaeolatitudinal positions during this time Proterozoic to Early-Middle Cambrian time, evap­
changed through the following steps (Fig. 3): orites and carbonates were deposited in the Gulf
region and Oman (latitudes 5° -15°S) (Huqf and
Latest Precambrian between 7°S (Oman)
Hormuz Groups) and carbonates (Middle Cambrian
and 2SOS (Sinai)
Burj Formation) in the Levant (25°S); during the
Late Cambrian between 30°S (Yemen)
Devonian and Early Carboniferous, Jauf Formation
and 55°S (Sinai)
limestones were laid down in Saudi Arabia (latitude
Middle- Late Ordovician between 28°S (Oman)
22°-20°S) and Markada and Harur Formations car­
and 52°S (Sinai)
bonates in Syria and Iraq (latitudes 15°-28°S).
Late Ordovician between 35°S (Oman)
Finally, in the Late Permian, the Khuff cyclic car­
and 58°S (Sinai)
bonates and evaporites, which were laid down over
Early Silurian between 29°S (Oman)
most of eastern Arabia (S0-30°S) during a rise in
and 50°S (Sinai)
sea-level following the end of Gondwanan Permo­
The 'long axis' orientation changed with the advent Carboniferous glaciation heralded the Mesozoic
of the Devonian and continued to shift orientation predominance of low-latitude carbonates (Fig. 3).
for the remainder of the Palaeozoic, as tabulated During the Late Ordovician, the present part of
below, with the position of the present northern central and western Arabia formed the northernmost
extremity ( PNE) indicated for each new orientation. extent of the Ordovician (Saharan) glaciation, lying
(Fig. 3): at about 50°S (Zarqa and Sarah Formations of Saudi
Arabia), and during Late Carboniferous to Early
Devonian orientation E-W, PNE west
Permian, the present southern part of Arabia
Early Carboniferous orientation NW-SE, PNE
(Oman-Yemen and southwest Saudi Arabia­
northwest
Ethiopia) again underwent an episode ofglaciation
Permian orientation N-S, PNE north
(the AI Khlata Formation of the Haushi Group of
Palaeolatitudinal positions during this time span Oman and its correlatives in Yemen and Ethiopia).
changed as follows (Fig. 3): However, considering that this glaciated strip of
Arabia (and adjacent Afghanistan-India to its
Early Devonian between 22°S (Zagros
south) was located in middle latitudes during that
suture) and 39°S (Yemen)
phase (average 42°-45°S) the glaciation there was
Early Carboniferous between 14°S (southeast
probably of uplifted (upwarped) areas prior to
Turkey/northern Iraq) and
Permo-Triassic rifting of this part of Gondwana.
37°S (Yemen)
During the Mesozoic, the Arabian Plate region
Late early Permian between 28°S (southeast
moved to tropical latitudes, and its orientation ('long
Turkey/western Iraq) and
axis') was similar to that of the present. In the Early
47°S (Yemen)
Triassic it was still wholly in the southern hemisphere
Late Permian between l2°S (southeast
(0°-25°S), but from the Early Jurassic onwards it
Turkey/northern Iraq) and
straddled the equator, extending to no more than
36°S (Yemen)
20°S but generally averaging about equal areas north
This sequence of changes would have placed the and south of the equator (Beydoun, 1991). In the
Middle East region throughout most of the Palaeo­ Cenozoic the move northwards accelerated and from
zoic mainly in temperate middle to high latitudes, the late Palaeocene onwards it lay totally in the
where a humid continental climate with large con­ northern hemisphere, its northern tip not extending
trasts in seasonal temperatures would have prevailed northwards beyond about 35°N. In consequence,
and where clastic sedimentation would have pre­ warm-climate carbonate sedimentation has been
dominated. The region drifted into the edge of the predominant from the Late Permian onwards,
warm and dry (arid) climate belt towards the end of accompanied from time to time by episodes of
the Palaeozoic. Only sporadic migration into or aridity; the deposition of associated clastics was
338 Z. R. Beydoun

3o• s

30° s

so• s so•
c D

o• o•

NUB - Nubia IR - Iran IND - India A. LATE CAMBRIAN

AFG - Afghanistan MAD - Madagascar B. MIDDLE-LATE ORDOVICIAN
TUR - Turkey SOM - Somalia (LLANDEILIAN - CARADOCIAN)
c. LATE ORDOVICIAN (ASHGILLIAN)
D. EARLY SILURIAN (LLANDOVERIAN)
E. LATE EARLY DEVONIAN (EMSIAN)
F. EARLY CARBONIFEROUS (VISEAN)
G LATE EARLY PERMIAN (ARTINSKIAN)
H. LATE PERMIAN (KAZANIAN)

Fig. 3. Palaeozoic plate tectonic reconstructions. Arabia and surroundings. Modified with permission from Scotese eta/.
(1979); Scotese & McKerrow (1990).
 Middle East clastic oil and gas reservoirs 339

principally in the form of transgressive shales or correlatives; other potential Palaeozoic source rocks
pelagic marls, some of which form major source are of local significance only.
rocks and/or seals. Episodes of coarse clastic depo­ Because Palaeozoic formations, and especially the
sition were restricted to either arid conditions in pre-Devonian ones, are stratigraphically much older
specific climatic rain-shadow belts (Triassic) or to and, generally, deeper as exploration targets in a
major fluvio-deltaic regimes draining uplifted regions region fortunate in its profusion of prolific Tertiary
and confined to specific time slices and areas. and Mesozoic carbonate-reservoired oil accumu­
Examples are the Lower and Middle Cretaceous lations and in Late Permian carbonate-reservoired
systems of the southern Iraq-Kuwait and north­ gas accumulations, the widespread search for pre­
eastern Saudi Arabian region, the Gulf of Suez area Permian targets had to await changes in global
and the Levant; the Oligo-Miocene systems of economic and energy demands before being
the same Kuwait-Basra area; the clastic deposits initiated. The first such exploration was carried out
associated with initiation of rifting and subsequent in Oman, where the Palaeozoic sediments are not
marginal infill (Marib-Jauf in Yemen, Gulf of Suez): deeply buried in the southeastern part of the country,
the regressive/basal transgressive coarse clastic and later in other parts of the region lacking younger
deposition following regional uplift (end Jurassic­ reserves ( Jordan), before the latest successful ven­
early Cretaceous in various sectors): or finally, the tures in Saudi Arabia. Based on these recent very
Mio-Pliocene molasse associated with the rising major discoveries in central Arabia in Upper Car­
Zagros-Taurus orogens of southwest Iran, Iraq and boniferous to Lower Permian sandstone reservoirs
southeast Turkey. and the identification of the source rock for these
It is these varied palaeolatitudinal settings of the substantial accumulations as being the Lower
Middle East Arabian Plate region that have localized Silurian shales (see next section), the potential for
the development of clastic (reservoir) facies from finding considerable new Palaeozoic reserves else­
the Permian onwards but permitted their extensive where in the Middle East region is particularly
regional development prior to the Late Permian. promising. Potentially, overall the Middle East
Post-Palaeozoic deposition of carbonates over the reserves could well double in the coming two
wide Arabian shelf region gave rise to very areally to three decades of intensive exploration of the
extensive reservoir facies, periodically alternating stratigraphically deeper levels, provided due
with widespread source-seal units, thus enabling understanding of the palaeogeographical/subcrop
the generation and trapping of the very large hydro­ distribution of regional source sediments, source
carbon reserves that are located principally in the 'kitchens' and maturation levels are arrived at to
northeastern part of the region, in large to 'giant' direct the search (Beydoun, 1991).
structural traps with predictably identifiable 'kitchen'
areas and migration pathways. By contrast, the gen­
erally quite rapid lateral facies changes of the P A L A EO ZO I C C L A STI C R E S E RVO I R S
predominantly clastic Palaeozoic sequences and
particularly the relative paucity of thick efficient At present, the most important Palaeozoic clastic
(unbreached) seals, precludes short migration path­ reservoir production in the Middle East region is
ways from source to reservoir into trap, and often that in central and southern Oman (Fig. 1), which
permits considerable vertical migration until an first commenced regular production in 1975. Heavy
effective (regional) seal is encountered. Moreover, oil of 19°-22° API gravity had been discovered in
the distribution of Palaeozoic source sediments has southern Oman (Dhofar Province) as early as 1956,
been dependent largely on deposition in tectonically by the Cities Service Company at Marmul, but it was
controlled basins or during major transgressive never developed because of the heavy and viscous
episodes, which subsequent tectonic events have nature of the oil and the low oil prices prevail­
obscured or stripped off. So far, only two major ing prior to 1973, which made the venture non­
regional source facies have been identified: the Infra­ commercial; the age of the reservoir had not then
Cambrian to Lower Cambrian Huqf Group of Oman been established precisely nor had the significance
and the Gulf region and its correlatives, which are of the oil sourcing been appreciated or properly
now generally deeply buried beneath later sediments understood.
and are mainly overmature, except in Oman, and The sharp rise in oil prices in 1973 coincided with
the Lower Silurian Qusaiba/Safiq shales and their the early phase of serious exploration for older
340 Z.R. Beydoun

deeper reservoirs in central and southern Oman The Mahwis Formation of South Oman comprises
(additional to the Cretaceous reservoirs of central waterlain sheet-flood sands and conglomerates,
Oman, which had been on production since the mid- probably again derived from the west (Heward,
1960s), and this gave added impetus for accelerated 1989). It passes northwards into the coastal andl
exploration and development following the first dis­ storm-dominated shallow-marine sediments of the
coveries of light oil in Palaeozoic sandstone reser­ Andam Formation, which consists of fine sandstones,
voirs. First commercial production began in 1975, shales and thin ooidal and stromatolitic limestones
from the fields in the Ghaba basin, and there are in the middle part. The sediment package, of Late
now at least 33 fields producing from Palaeozoic Cambrian to Early Ordovician age on the basis of its
reservoirs in southern and central Oman, which had trilobite and other fauna, represents the first marine
cumulatively produced some 890 million barrels of transgression in the Palaeozoic of Oman (Hughes
oil until the end of 1991. There are several billion Clarke, 1988).
barrels of recoverable oil in these reservoirs. The Ghudun Formation reservoir (Haima Group),
although an excellent reservoir, is only of minor
regional significance, being of importance in only
Clastic reservoirs in Oman
one South Oman field.
There are two major clastic reservoir groups in
Oman. The older group comprises continental and
Haushi Group
shallow-marine sandstones of the Haima Group,
which range in age from Early Cambrian for the For the Haushi Group reservoirs, the lower or the
oldest (Amin Formation) to the Late Cambrian to AI Khlata Formation reservoir consists of a complex
Early Ordovician (Andam Formation); the second of clastic lithologies with considerable lateral and
group comprises glaciogenic sandstones and shallow­ vertical changes in both facies and in thickness; this
marine and fluvial-channel sandstones of the Haushi is particularly the case in South Oman, where these
Group, which ranges in age from Late Carboniferous changes occur over very short distances (Levell
(AI Khlata Formation) to Early Permian (Gharif et a/., 1988; Heward, 1990). Lithologies range from
Formation) (de Ia Grandville, 1982; Hughes Clarke, coarse-fine diamictites, conglomerates (particularly
1988; Sykes & Abu Risheh, 1989). at the base), gravels, pebbly sands, sands, silts, silty
shales and shales that are often varved. Boulders
can reach up to 10 m in diameter and are of very
Haima Group
varied composition, including granite, granophyre
The oldest oil-bearing Haima Group clastic sedi­ and volcanics, as well as Infra-Cambrian sedimentary
ments in Oman are those of the Karim and Haradh rocks of the Huqf Group. The AI Khlata is a glacially
Formations (the Lower Haima of South Oman; generated sediment package and was deposited by a
Hughes Clarke, 1988). The Haradh is of greatest series of glacial advances and retreats from the
significance as a reservoir and comprises chert-rich southwest towards the northeast, as evidenced by
cross-stratified sands with abundant synsedimentary striations and 'whale backs' in glacial pavements in
deformation features. It probably represents a outcrop (Levell et at. , 1988). The reservoirs are
braided alluvial apron shed from highlands to the predominantly glaciofluvial and glaciolacustrine
southeast, with the abundant chert being derived (Heward, 1990). The AI Khlata correlates with simi­
from the underlying formations of the Huqf Group lar deposits present elsewhere in southern Arabia
(Heward, 1989). and Ethiopia (Beydoun, 1988; McClure, 1980; Levell
The Amin Formation reservoir is predominantly et a/. , 1988).
an intercalation of aeolian dune, sabkha and water­ The overlying Gharif Formation reservoir of the
lain sands. It is remarkable for the generally small Haushi Group consists of a series of fluvial deposits
size of the dunes that deposited it and reflects a derived from the southeast. The arkosic character is
period of arid continental conditions of widespread due to recycling of granitic material from AI Khlata,
extent. Conglomerates derived from a deformation presumably being stripped off from the uplift that
front to the west of Oman interfinger with these finally rifted between Oman- Afghanistan-India to
desert deposits. The Amin Formation probably cor­ the southeast. The formation fines and becomes
relates, at least in part, with the widespread Saq more shaley to the northwest; its deposits reflect an
(sandstone) Formation of Saudi Arabia. ameliorating climate, most being red-beds, locally
 Middle East clastic oil and gas reservoirs 341

with calcrete and frequent occurrences of silicified generally have limited lateral extent and are often
wood in outcrop, with growth bands indicating a thin and easily breached. Migration of Huqf sourced
seasonal climate. Marine sands and limestones oil has, thus, often moved vertically up as high as the
occur to the north and west and reflect the interplay Carboniferous-Permian reservoirs of the Haushi
between marginal-marine -coastal-plain deposits Group, where sealing is provided by glaciogenic or
and continental red-bed facies (Hughes Clarke, shallow-marine-floodplain shales and mudstone
1988). The formation generally correlates with the deposits. Furthermore, where these have locally
upper Unayza Formation of Saudi Arabia discussed become breached or have proved inadequate, verti­
below. cal migration has continued in central Oman to
The considerable volumes of hydrocarbons in charge the Mesozoic carbonate reservoirs of the
place, housed in the Palaeozoic clastic reservoirs of largest Oman fields (Sykes & Abu Risheh, 1989).
the Oman region, are demonstrated to have been The subject and wider implications of Palaeozoic
abundantly sourced from Huqf Group source sedi­ oils in Middle East exploration has been discussed
ments at least 550 Ma in age (but also of older Upper and reviewed in more detail by Beydoun ( 1991).
Proterozoic age) straddling the Infra-Cambrian to The AI Khlata reservoirs are determined by a
Early Cambrian boundary and terminating with number of facies belts whose distribution is partly
a widespread evaporite unit- mainly salt (Gorin influenced by local salt dissolution (of the underlying
et al., 1982; AI Marjeby & Nash, 1986; Grantham Huqf Group Ara Formation salt); where subsurface
et al., 1988; Mattes & Conway Morris, 1990). A few salt dissolution occurred during the glaciation
fields in western Oman, however, adjacent to the episode, lakes were formed with generally shaley
eastern side of the Rub al Khali basin, have oil deposits. Along the margins of the lakes, glaciogenic
accumulations in Haushi Group clastic sediments deltas gave rise to mixed sequences of sands and
that have been geochemically typed to Silurian (Safiq shales (multiple reservoir-seal couplets) and consti­
and equivalent) source rocks (Grantham et al., 1988; tute the most prospective facies belt. Within the
Beydoun, 1991). main South Oman salt basin, terrestrial deposition
The intermittent uplift of the southeast Arabian dominated, with much waste rock (diamictites)
plate margin since mid-Palaeozoic times, and par­ and erratic reservoir character (Levell et al., 1988).
ticularly since the Silurian, when the region became Deltaic deposits are occasionally incised by sand­
relatively positive, is the principal reason why these stone-filled channels and rapid lateral and vertical
very old source rocks in South Oman remain in the variations in lithology and stratigraphy make it diffi­
oil window and continue to generate oil; they lie, on cult to estimate and indeed develop all recoverable
the whole, at moderate depths and seemingly have oil reserves from the AI Khlata reservoir. The over­
not been more deeply buried. The relatively heavy lying transgressive Gharif Formation sands often
nature of some of the oil is considered to be due to provide a peripheral stratigraphical trap component
subsequent transformation, although some may be reservoir to the structural trap formed by drape over
ascribed to generation from still relatively immature Haima Group sediment pods left by salt withdrawals
sediments at shallow depth. Hydrocarbon charge and dissolution (AI Marjeby & Nash, 1986).
from these old source rocks into Palaeozoic clastic Porosities and permeabilities of the South Oman
reservoirs occurs from beneath the terminal Ara Palaeozoic clastic reservoirs are generally good and
Formation (Huqf Group) salt edge and takes place range from 15 to 35% and 100 mD to over 10 D in
in conjunction with progressive subsurface salt the friable and poorly cemented reservoir levels (de
removal and dissolution; this leads to the formation Ia Grandville, 1982; Levell et al., 1988).
of residual cores composed of basin-fill bodies of the
overlying post-salt Palaeozoic clastic deposits (AI
Clastic reservoirs in Saudi Arabia
Marjeby & Nash, 1986). Short- to medium-distance
lateral migration into the Palaeozoic clastic deposits, In central Saudi Arabia, several billion barrels of
and/or subsequent long-distance vertical migration high gravity (c. 43° API), low-sulphur Palaeozoic oil
within these clastic deposits, occurs until stopped by (and gas) have been discovered recently in the region
an effective seal in trapping geometry; these seals south of Riyadh (Fig. 1), principally in the Unayza
are predominantly shales or argillaceous sediments. Formation sandstone reservoirs of ?latest Carbon­
In the clastic depositional facies patterns that charac­ iferous to Early Permian age (AI Laboun, 1986;
terized the Palaeozoic of the region, shale seals Husseini, 1992; McGillivray & Husseini, 1992), but
342 Z.R. Beydoun

also including some oil housed in Devonian and Other reservoir sandstones in this region of Arabia
Ordovician sand reservoirs. Oil to potential source include Ordovician glacially related sandstones and
rock correlations have identified the principal source Devonian ?continental sandstones. The hydrocarbon
of these hydrocarbons to be the basal, organically charge to the trap from the Silurian Qusayba shales
rich shales of the Qusayba Member of the Lower is controlled by juxtaposition of reservoir to source
Silurian Qalibah Formation (Abu-Ali et al., 1991; through faulting and/or fluid communication with
Husseini, 1992; McGillivray & Husseini, 1992; the main Unayzah reservoir. Stratigraphical pinch
Mahmoud et at., 1992). The Qalibah Formation was out of the Unayzah reservoir against the basal
previously unrecognized as a separate formation and Khuff seal unconformity may affect the trap form
was included in the middle part of the Tabuk Forma­ significantly.
tion (Powers, 1968). The first significant Palaeozoic The Lower Silurian Qusayba shales (and the
sandstone reservoir hydrocarbon discovery was equivalent Safiq in Oman) are widespread across
made in 1979 in Qirdi (south of the Khurais field, Arabia and reflect the first transgressive cycle
then the westernmost established oilfield in Saudi related to global sea-level rise consequent on the
Arabia) in an Upper Permian pre-Khuff (carbonate) deglaciation occurring at the end of the Late Ordo­
Formation sandstone sequence and in the same year vician to Early Silurian 'Saharan' glaciation. Similar
pre-Khuff gas was also tested in the 'supergiant' shales are widely distributed throughout the Middle
Ghawar field (Husseini, 1992). Other discoveries East and North Africa, but subsequent post- Middle
followed in deep tests on the Khurais trend south of Devonian tectonism in the Middle East and uplift
that field, but at depths in excess of 4000 m. It was have removed them from elevated areas; conse­
not until 1989 that the major discoveries in the quently, their subcrop palaeogeographical distri­
previously unexplored areas were made, and at bution in the· region beneath the Late Permian
moderate depths. Khuff seal needs to be better controlled. The
The Unayza Formation is a siliciclastic succession Qusayba shales are, however, seemingly present
of fluvial, coastal-plain, deltaic and shallow-marine everywhere in the basins of Saudi Arabia and at
facies whose age may extend back to Late Carbon­ depths where they are sufficiently mature to generate
iferous (AI Laboun, 1986), but is principally of oil and gas (Husseini, 1992); this includes those
Early to Late Permian age; it appears broadly cor­ portions of the Arabian Peninsula adjacent to Jordan
relatable with the Gharif Formation of Oman whose and southern Iraq (AI Laboun, 1986). Oil generation
deposition is associated with the Late Carboniferous and up-dip migration from some of the Arabian
to Early Permian deglaciation episode that affected depocentres commenced as long as 160 Ma (Abu-Ali
Gondwana, including the southern part of Arabia et a!., 1991).
(see earlier). Distribution of the varied Unayza
facies is controlled by palaeogeography, sea-level
Clastic reservoir in the Gulf Region
fluctuations and by local and regional structure
(Ferguson & Chambers, 1991). The complex inter­ In other areas of the Gulf region, gas had been
play of these factors coupled with erosion during discovered in pre-Khuff clastic reservoirs in Abu
sea-level lowstands, has resulted in a variety of Dhabi and Bahrain and non-commercial gas found
reservoir types, varying through poorly sorted coarse in onshore Qatar. These gas discoveries are as yet
silt-sand of tentative marine origin, accumulated in undeveloped although potentially producible should
fault bound depressions; fine to pebbly cross-bedded market forces warrant it; but although the gas is
sandstone of alluvial origin; a variety of paralic generally H2S free, appreciable amounts of associ··
sandstone facies, including shoreface-foreshore, ated nitrogen and C02 cast doubt on commerciality.
delta channel and coastal-plain channel sands (it is The principal accumulations in these countries
these paralic facies that have proved to be the most occur in Unyiza-equivalent clastic deposits, and the
prolific reservoirs); the final facies is a transgressive sequences consist of continental fluvio-deltaic sands,
lag of unpredictable distribution but consisting of silts and shales with several reservoir sandstone
clean fine- to very coarse-grained reservoir quality intervals separated by shale bodies. The source of
sandstone (Ferguson & Chambers, 1991). Porosities the gas is generally assumed to be the Silurian
in these facies are highly variable but average 20%, shales, which, however, have not been penetrated in
although permeabilities of several darcies are not any Abu Dhabi well (Ali & Sildawi, 1989; Loutfi
uncommon (Husseini, 1992). & El Bishlawy, 1989), although they may have
 Middle East clastic oil and gas reservoirs 343

been partly penetrated m Bahrain (Samahiji & sequences with stringers and intercalations of shales,
Chaube, 1987). claystones and siltstones; lateral and vertical con­
tinuity varies rapidly, however, being the main
reason why only three or four of the dozen or more
Clastic reservoirs in Syria
wells drilled over the (?)imprecisely defined sub­
In Syria, a number of gas accumulations occur in surface feature to date are producers (NRA Jordan,
anticlinal structures in the central part of the country 1989). Good permeabilities, however, are developed
(southern Palmyrid basin) (Fig. 1). These are cur­ locally, as evidenced by the reported frequent mud
rently undergoing delineation drilling in preparation losses during drilling. Light but non-commercial oil
for commercial production. The reservoir facies is has been found in the same reservoir in well Wadi
sandstone, occurring in intervals separated by clay­ Sirhan 4, some 250 km to the southwest (NRA
stones, thin shales and intercalated dolomites of the Jordan, 1989).
Upper Carboniferous Markada Formation; there
are rapid vertical and lateral lithological variations.
Clastic reservoirs in southeast Turkey
The depositional environments are probably shallow­
marine to ftuvio-deltaic, with provenance of clastic In southeast Turkey, several small accumulations
material from erosion of the adjacent uplifts (Rutbah of gas and condensate have been discovered in
High and ?Aleppo High) created by an 'Hercynian' Devonian age clastic sediments of the Katin Forma­
tectonic event. Periodically reduced clastic supply tion, below the southern front of the Taurus thrust
allowed carbonate deposition, commensurate with belt (Fig. 1). The reservoir facies consists of clean
palaeolatitudinal location in tropical waters. Log shallow-marine sands, which are oil impregnated
porosities in the clastic sediments are reported to nearby at outcrop (Hazro sands, previously thought
range between 5% and 16% . The source of the gas to be Permian in age; Beydoun, 1991). These reser­
(predominantly methane but with C2 and C3 fractions voir sands are considered to have been sourced by
and varying small amounts of nitrogen) has been the Ordovician-Silurian organic-rich Bedinan For­
inferred to be from shales within the Markada itself; mation shales (locally defined as the Dadas Forma­
however, based on regional considerations, it seems tion) with TOCs up to 5% (Harput & Erturk, 1991);
more likely to be from unconformably underlying these also contain reservoir sand levels (Handof
Silurian (Tanf Formation) shales in a similar manner Member) that have contained strong but non­
to the Unayza accumulations of Central Arabia. commercial hydrocarbon shows in a number of wild­
Although not proven to underlie this area, these cat wells in the southeast Turkey region. The
shales are extensively distributed in the region and cumulative production from these Palaeozoic reser­
have been encountered in a number of widely voirs has been modest to date; however, geochemical
scattered wells (Lababidi & Hamdan, 1985 ; work in the area by a number of operators confirms
Beydoun, 1988, 1991). that the Bedinan Formation has been the source
charging a number of oil-productive Cretaceous car­
bonate reservoirs in southeast Turkey in areas where
Clastic reservoirs in Jordan
the Middle Cretaceous often unconformably over­
In adjacent Jordan, the Risha field, close to the Iraqi lies the Palaeozoic (Mardin-Bozova High). It is
border (Fig. 1) , is producing gas from a clastic thought likely that accumulations in older clastic
reservoir of Late Ordovician age; this is the Dubaidib Palaeozoic reservoirs, such as those found in the
Sandstone Formation of the Khreim Group lying Barbes and Katin fields, could be discovered with
directly beneath the organic-rich Lower Silurian better subsurface well control and seismic mapping.
(Mudawara Formation) shales, which appear to act
both as source and seal for this reservoir (NRA
Clastic reservoirs in the Gulf of Suez
Jordan, 1989). This area of northeast Jordan was an
Early Palaeozoic basin that accumulated a thick In the petroliferous Gulf of Suez rift basin of Egypt,
succession of Palaeozoic sediments, several thousand on the south western edge of the Levant-Sinai
metres of which (Early Cambrian-Silurian) are still fragment of the Arabian Plate, the bulk of the oil
preserved following inversion and erosion in the reserves are housed in sandstone reservoirs, the
Late Palaeozoic. The Ordovician reservoir sequence majority being of Miocene age, but a number of
here consists of shallow-marine cyclic sandstone fields also produce from 'Nubia' sandstone reservoirs
344 Z. R. Beydoun

of Late Palaeozoic and Cretaceous ages. The Palaeo­

Clastic reservoirs in the
zoic reservoirs belong to the Nubia 'C' unit, which is
Bahrain- Neutral Zone- Kuwait - Basra region
of ?Devonian-Carboniferous age (the Nubia 'B'
consists of Carboniferous black shales, which are The 'Bahrain zone' in the Awali field consists of
regarded as a likely source for some of the oil by several productive Middle Cretaceous sandstone-­
some authors- Ayouty, 1990). The Nubia 'C' siltstone (Wara, Nahr Umr) and carbonate (Ahmadi,
depositional facies in the Gulf of Suez region is Mauddud) reservoirs. Subsequently, the discovery
principally shallow marine with deltaic and fluvial and development of large oil reserves from equiv­
intercalations over much of the area but with alluvial alent age sandstone reservoirs occurred in Kuwait in
deposits in the southern part; the facies became 1938 (Wara and Burgan Formations), in Lower Cre ­
predominantly continental to shallow marine towards taceous sandstone reservoirs in the Basra area o:f
the end of Nubia 'C' time in the Late Carboniferous southern Iraq at Zubair in 1949 (Zubair Formation)
to Permian (Klitzch, 1990). and in Albian age sandstone reservoirs (Safaniya/
About six or seven oil fields produce from the Khafji Members of the Wasia Formation) in the
Palaeozoic sandstone reservoir, including the now northeastern part of Saudi Arabia and the adjacent
depleted Hurghada field in the west, where the bulk part of the Saudi-Kuwaiti Divided (Neutral) Zone
of the oil was housed in the Nubia 'C' reservoir and in 1951 (among others, the Safaniya-Khafji field ­
which produced during its life span (1915-1969) 41 the largest offshore oil field in the world).
million barrels of oil. Other Palaeozoic reservoir The oil (and gas) fields of the southern Iraq ­
fields include the Ras Gharib, July, Ramadan, Ras Kuwait-Divided (Neutral) Zone-northeast Saudi
Badran and Sidky fields, where reservoir porosities Arabia onshore and offshore region, were discovered
and permeabilities are fair to good and contain oils and developed since the late 1940s and earliest 1950s.
having gravities that range between 24° and 40° API They produce (or are being prepared to produce)
(Ayouty, 1990). from these Middle and Lower Cretaceous sandstone
The oil in the Palaeozoic and younger reservoirs reservoirs, and although only 23 in number, they
of the Gulf of Suez is considered by many to be constitute (both in production and recoverable
sourced mainly from organic-rich Early Miocene reserves) by far the most important clastic reservoir
shales and marls (Rudais-Karim Formations) oil province of the whole Middle East region .
brought into the right juxtaposition of source to The recoverable hydrocarbon volumetrics of this
reservoir in tilted fault-block traps or stratigraphical province alone involve a sustainable production of
pinch outs. Additionally, and perhaps alternatively, several million barrels of oil per day and many tens
Late Cretaceous and Eo-Palaeocene sediments also of billions of barrels of remaining recoverable
contain ample organic carbon, enough to generate reserves (Beydoun, 1988, 1991), well in excess of the
the oil in Cretaceous and Neogene reservoirs oil totals for the entire North Sea Basin !
(Ayouty, 1990), and the Carboniferous Nubia 'B' The sandstone reservoirs housing these vast
shales cannot be ignored as another possible source reserves represent the deposits of a huge fluvio-·
to be considered in the calculation of total oil charge deltaic-littoral-marginal marine depositional
volumetrics. system, which commenced in early Cretaceous
(Hauterivian-Aptian) time, when this area of the
Middle East lay athwart the humid equatorial belt ,
M E S O Z O I C C L AS T I C R E S E R V O I R S 2°-3° to north and south of the equator, and ter-­
minated in the Albian to early Cenomanian, when
Commercial oil production from Mesozoic sandstone the area drifted to between so and l0°N (Dercourt
reservoirs was first established in the Middle East et al., 1986; Beydoun, 1991). This fluvial system
region on the western side of the Gulf of Suez, in (a virtual 'Nile sized' river complex) transported
what are now depleted fields (Hurghada, 1915; Abu immense quantities of clastic detritus northeastwards
Durba, 1918; etc.) in 'Nubian' sandstones of Cre­ from the Arabian -Nubian shield highlands, and
taceous age (now known as the Nubia 'A') (Ayouty, from the exposed Palaeozoic clastic sediments over­
1990). Sustained and continuing commercial oil lapping the shield, on to the shallow and very wide
production from sandstone reservoirs in the Middle Arabian passive margin shelf. The delta complex
East region, however, was first made in Bahrain in prograded during sea-level lowstands but retreated
1932 with the discovery of the Awali field. during transgressive high sea-level stands when shelf
 Middle East clastic oil and gas reservoirs 345

carbonates or fine clastic sediments were deposited of siltstone and shale and exhibits rapid lithological
between the coarser lowstand sandstones. A distribu­ changes both laterally and vertically (Powers, 1968).
tary channel system also extended the delta complex In its lower part, in the Kuwait region, it includes a
to the Baghdad area in Central Iraq (Jawad Ali & large lenticular sand body enclosed by grey, lignitic
Aziz, 1993). shales (Owen & Nasr, 1958; Dunnington et al . ,
The repetitive nature of the Zubair Formation 1959) which indicate a return t o fiuvio-deltaic depo­
deltaic cycle (Hauterivian-Aptian and correspond­ sitional conditions. This sandstone body interfingers
ing to the subsurface Biyadh Formation of Saudi with littoral-shallow-marine facies to the southeast
Arabia), which is one of the major productive reser­ (Bahrain) and with deeper marine pro-delta facies
voirs of the area, recently has been documented to the north and northeast (southern Iraq); sub­
qualitatively (Khaiwka, 1989): the cycle includes aerial delta deposits, tidal-fiat and coastal-plain
pro-delta, delta front and delta plain, with marine deposits dominate to the west, with the submarine
shelf facies concluding the cycle. Subaqueous levee, portion of the delta being located east of the present
distributary channels , distributary mouths, distal coastline.
bars, freshwater lake, mud-fiat, marsh, subarea) The Wara reservoir contains significant volumes
levee and lobe abandonment facies have all been of oil in Bahrain, in the Divided (Neutral) Zone and
recognized and have a strong repetitive vertical in northeast Saudi Arabia and southern Kuwait,
pattern. where clean permeable sands are common. The
Rapid facies changes provide the lithological con­ Wara clastic cycle is terminated by a transgressive
trols on the variable hydrocarbon distribution in the marine shale facies of the Ahmadi Formation,
lower regional (Zubair) clastic reservoir; it provides which forms the regional seal to the Wara reservoir
the major production in southern Iraq but is a (Brennan, 1990; Samahiji & Chaube, 1987). In the
relatively minor reservoir in Kuwait and the north­ Basra area the Wara is not distinguishable as a
east Saudi Arabian fields. Potentially important sandstone sequence, the coeval interval being a facies
new oil production from the Zubair reservoir was of black silty shales and siltstones with a thin oil­
reported as being established recently in the East saturated sand (Dunnington et al., 1959) .
Baghdad oilfield of central Iraq (Beydoun, 1988). Reservoir porosities and permeabilities for these
The Zubair Formation is overlain by the trans­ different Lower and Middle Cretaceous sandstone
gressive carbonate cycle of the Shuaiba Forma­ reservoirs vary and data are generally sparse, but
tion, which concludes with a terminal reefal good reservoir properties are commonly present.
build-up phase (Dunnington et al., 1959; Powers, (OAPEC, 1985 ; Brennan, 1990, 1991) . In southern
1968). Iraq, a petroliferous belt of Albian Nahr Umr clastic
The Albian (Nahr Umr and Burgan Formations) sediments coincides with a marginal-marine depo­
clastic sequences of the Basra-Kuwait area and sitional area (Ibrahim, 1983) but production from
their equivalents in northwest Saudi Arabia and the this reservoir is subordinate to that from the Zubair
Neutral Zone (Khafji and overlying Safaniya reservoir.
Members of the Wasia Formation) initiate a new The hydrocarbons in the Lower and Middle Cre­
prograding fiuvio-deltaic-littoral-shallow-marine taceous clastic reservoirs of the 'head of the Gulf '
depositional cycle following the Shuaiba (transgress­ region of the Middle East are believed sourced from
ive) carbonate cycle, similar to that of the preceding the Kazhdumi. Formation organic-rich sediments
early Cretaceous Zubair cycle. This Nahr Umr to in the Kazhdumi basin immediately to the east
Burgan progradational episode grades up into the (Brennan, 1990) and/or from a 'deeper' Lower Cre­
widespread shallow-shelf carbonate sedimentation taceous source ( Ibrahim, 1981) within the Gotnia
of the Mauddud Formation, which terminates the depositional basin of Murris (1980) (Sulaiy and Chia
Albian. Gara/Sarmord and Garau source) with vertical
The early Cenomanian Wara clastic depositional migration and charging from the Zubair into the
episode that followed the Mauddud was of more Nahr Umr reservoir of southern Iraq ( Ibrahim,
limited duration and areal extent, covering Kuwait 1981). This question of sourcing of the various
and eastern Arabia to Bahrain, and representing a Middle East reservoirs and the regional distribution
brief terrigenous interlude in an otherwise marine of source rocks has been reviewed and discussed
Middle Cretaceous depositional phase. The Wara in more detail by Beydoun (1988, 1991) and by
consists of a sequence of sandstones with interbeds Beydoun et al. (1992).
346 Z. R. Beydoun

other Late Jurassic rifts on an E-W trend (Mills,

Clastic reservoirs in Yemen
1992), with a deep lying pre-rift sediment package
Two new and potentially major hydrocarbon pro­ underlying a Kimmeridgian-Tithonian synrift se­
ductive Mesozoic clastic reservoir regions have been quence, which includes an analogue of the organic­
established in the Middle East since 1984. These rich Madbi source sequence; a series of important
are in the Marib -Jauf rift basin and the Wadi discoveries have been made in this area, the first in
Hadhramaut basin of Yemen in southwest Arabia. December 1990; estimated recoverable reserves to
The Marib-Jauf basin now has about seven oil­ date are conservatively estimated at about 250 million
fields on production, or being developed for pro­ barrels.
duction, where the oil and gas/condensate reserves The succession in this Hadhramaut-Jeza rift basin
are housed in sandstone reservoirs of latest Jurassic area is that described by Beydoun ( 1964, 1966) and
age. The discovery field was Alit, which is also the Beydoun & Greenwood ( 1968) for the typical non­
biggest. In 1991, this basin's oil production averaged evaporitic Jurassic Amran Group of Hadhramaut
about 190 000 barrels per day and recoverable overlying the Liassic to Middle Jurassic clastic
reserves were estimated at about 3-4 B barrels (Oil Kohlan Sandstone Formation ; this consists of
and Gas Journal, 30 December 1991); the gas (with pre-rift Shuqra Formation platform carbonates
condensate) reserves are estimated to run into sev­ and synrift Madbi and Naifa (Kimmeridgian ­
eral trillion cubic feet and means for developing Tithonian-Berriasian) organic-rich shales and marls
these reserves are under consideration. The Upper and argillaceous limestones, respectively. No evap­
Jurassic subsurface succession within this rift basin orites equivalent to the Sabatain Formation are
differs from that outcropping nearby in that a thick developed here and hence there is no efficient seal
synrift sequence of sediments was laid down ; this on top of the Jurassic, as in the adjacent Marib -­
NW-SE trending basin started to subside rapidly in Jauf and Shabwa rift basins. Exploration wells have
the post-Oxfordian, with in excess of 3500 m of shown some unexpected subsurface differences in
sediments being deposited in about 6 Ma (Maycock, the post-rift Cretaceous stratigraphy from the out­
1987). The Callovian-Oxfordian Amran (Shuqra) crop sequence described by Beydoun & Greenwood
platform carbonates here are overlain by a thick and ( 1968). An early Cretaceous platform carbon­
predominantly shaley sequence followed by carbon­ ate unit occurs beneath the Lower Cretaceous
ate turbidites rapidly deposited following graben (Barremian-Aptian) Qishn Formation consisting of
inception. This sequence, informally referred to as carbonates and alternating shales unconformably
the 'Lam Formation', is correlative with the organic­ resting on the terminal Jurassic Naifa Formation and
rich Madbi Formation of the outcrop area to the separating it from the Qishn ; this sequence has been
southeast (Beydoun, 1964; Beydoun & Greenwood, informally named the Saar Formation of Hauterivian
1968) and is the principal generative source for the to early Barremian age (Mills, 1 992) and appears to
Marib-Jauf hydrocarbon accumulations ; the 'Lam' have been laid down in an E-W shallow marine
is overlain by the 'Alif Formation', which is the main embayment. Additionally, the Qishn itself exhibits a
reservoir and consists of a shallow-marine-deltaic­ substantial development of clastics derived from
fluviatile succession of clean sands that are of bordering and still positive uplifts to the marine
Tithonian age (Huurdeman et al . , 1989) and are embayment, below its main Barremian-Aptian
succeeded by sandstones and shales of braided flu­ carbonate unit, which are not developed in the
viatile and deltaic systems, in turn followed by the main reference surface section for the formation
Sabatain (Main Salt) evaporite sequence, also of within the Tawilah Group (Beydoun & Greenwood,
Tithonian age (Maycock, 1987; Huurdeman et al., 1968) located near Mukalla along the coast about
1989) ; these form an excellent regional hydro­ 150 km due south of the first discovery. These
carbon seal. Excellent reservoir porosities and per­ clastic sediments in the area of recent discoveries
meabilities have been recorded ranging from 16% to (Hadhramaut-Jeza rift) are not only well developed
18% and 500- 1200 mD, with the oil being of but have good reservoir qualities. In this area, the
39°-41 ° API gravity (Huurdeman et al . , 1989). hydrocarbons were generated in the organic-rich
Some 300km to the east of the Marib-Jauf rift and well-developed Madbi source rocks in the
basin (and about 225 km east-northeast of its Late adjacent graben(s) and migrated into the suitably
Jurassic carbonate-reservoired Shabwa hydrocarbon juxtaposed older Kohlan (Lower-Middle Jurassic)
basin extension) recent exploration has revealed sandstone reservoir and also, in the absence of an
 Middle East clastic oil and gas reservoirs 347

efficient seal (Tithonian evaporites of the Sabatain basal Cretaceous respectively (Daniel, 1963b;
Formation being absent), through the overlying suc­ Dubertret, 1963).
cession into the Qishn sands beneath the thick and The 'Rutbah' reservoir of the Euphrates graben is
tight main Qishn carbonate that forms the regional sourced by the Upper Cretaceous, organically rich
seal (Mills, 1992). Good quality oil of 36° API gravity Shiranish-Soukhne Formation(s) marls; these are
is found in the Kohlan reservoir and 30° API oil in phosphatic as a consequence of upwelling, and were
the Qishn clastic reservoir(s) and substantial flow widely deposited pelagic sediments that extend in
rates were tested. Individual zones in the Qishn outcrop from Iraq across the Levant and northern
clastic deposits have flowed in excess of 4875 barrels Arabia through North Africa all along the southern
per day of 29°-33° API gravity with little dissolved Tethyan margin of North Africa. Their organic
gas and low sulphur (Mills, 1992). maturity level is low except where deeply buried
in rapidly subsiding basins, such as the Euphrates
graben, whose inception coincided with the start of
the pelagic sedimentation regime (Beydoun, 1991).
Clastic reservoirs in Syria
Juxtaposition of maturing source sediment with older
In the Euphrates graben basin of eastern Syria, reservoir sandstone facies in tilted extensional
some 25 oilfields discovered since 1984 are now on regime fault blocks in the subsiding grabens resulted
production and collectively were producing about in both oil charging and sealing of these generally
330000 barrels per day of 34°-38° API gravity low good 'Rutbah' sandstone reservoirs (Beydoun, 1988,
sulphur oil in early January of 1992 ( Middle East 1991). The presence of volcanics associated with the
Economic Survey, 35: 16, 20 January 1992). Recov­ Euphrates rifting episode and the initial high palaeo­
erable reserves are reported to be well over 1 B heat-flow have together created locally 'dirty' sands,
barrels. The reservoir is the so-called Rutbah causing production problems that have necessitated
Formation sandstone of Early Cretaceous age remedial pressurization measures in some fields.
(Beydoun, , 1988). The Rutbah type section was
defined originally across the border in Iraq, south of
Clastic reservoirs in other Middle East countries
the Euphrates River, where a 23-m section of vari­
coloured, white and ferruginous unfossiliferous sand­ Elsewhere within the Middle East region, minor oil
stone is exposed, with the basal part probably being production or potential production from Mesozoic
of continental origin and the uppermost part marine; sandstone reservoirs occur in Oman, offshore Qatar
the age was given as Late Triassic to early Ceno­ and in southern Israel-Palestine. In the Oman area,
manian based on the relationship with the dated producible oil has been found in a Lower Jurassic
overlying and underlying formations (Dunnington sandstone reservoir in (several) fields of the Ghaba
et a/., 1959). The correlation of the Rutbah with the basin of central Oman (Alsharhan & Kendall, 1986).
Nahr Umr of southern Iraq as a single diachronous The main production of the Saih Raw! field comes
sandstone deposit was not regarded favourably (see, from the Upper Palaeozoic Haushi Group sandstone
however, Ibrahim ( 1981) who recorrelates the two). reservoirs and, more recently, considerable volumes
The informal ' Rutbah Sandstone' of Syria was for­ of gas were estabbshed in the Lower Palaeozoic
malized as the Cherrife Shale Formation of Late Andam Formation sandstone reservoirs. In 1978,
Jurassic to Early Cretaceous age (Daniel, 1963a). commercial oil was discovered in the 'Marrat'
The type section was the Cherrife- 1 exploration well reservoir (Beydoun, 1988). Hughes Clarke (1988)
in the Palmyrid basin, where the formation comprises redefined the 'Marrat' of Oman and renamed it the
predominantly red-brown fet:ruginous sandy shale Mafraq Formation (only partly equivalent to the
and green pyritic shale with intercalations of quartiz­ Marrat of Saudi Arabia), with the lower Mafraq
itic and calcareous sandstone and streaks of lime­ being continental and locally possessing the charac­
stone. Elsewhere in Syria, the succession includes teristics of a palaeosol, whereas the upper Mafraq is
decomposed basalt and ash and is more varicoloured, shallow marine with variable thickness but including
more sandy and, in certain levels, clayey and lignitic. important clean reservoir sand levels (e.g. at Saih
Daniel ( 1963a) considered the Cherrife Forma­ Raw!) and is regarded as being essentially a basal
tion an extension of the Hathira Formation of clastic unit to the Middle-Upper Jurassic Dhruma
Jordan and the 'Gres de Base' of Lebanon, which Formation; thus the upper Mafraq is younger than
are Upper Jurassic to Lower Cretaceous and the Marrat of Saudi Arabia. Reservoir charge is
348 Z.R. Beydoun

probably from a deeply buried Huqf Group source

Clastic reservoirs in the Levant- Sinai region
that also charges the underlying Haushi Group
reservoirs and not from organically rich units within The Levant-Sinai fragment constituted an integral
the so called 'Marrat', as suggested by Alsharhan part of the Arabian Plate until the Neogene. In this
& Kendall (1986), nor from the overlying Middle region two areas of Mesozoic clastic reservoir oil
Jurassic Dhruma Formation shales that provide the production occur. The first is of minor significance
seal (Beydoun, 1988). and is located in the coastal region of southern
It is appropriate here to note that Beydoun (1988), Israel-Palestine. Here the only sustained oil pro­
in discussing the 'Marrat' accumulation at the Saih duction is from the Heletz-Brur-Kokhav field
Raw! field in Oman (p. 149) , made the passing complex, where the predominant reservoir facies is
remark that deeper exploration in producing fields a Lower Cretaceous sandstone with some carbon­
in Kuwait (p. 185) had established substantial new ate interbeds belonging to the Lower Cretaceous
reserves in lower Jurassic Marrat (assumed as clastic Kurnub Group. The discovery was made in 1955
sediments in consequence of the Saih Raw I discovery on re-entering and deepening a well abandoned in
in Oman and the Saudi Arabian surface section for Aptian sediments in 1948 by the Iraq Petroleum Co.
the Marrat) and other formations. In reality, the subsidiary Petroleum Development (Palestine)
Marrat facies in Kuwait, as in the adjacent por­ because of security conditions. The trap is a faulted
tion of Saudi Arabia, is predominantly carbonate anticline with a major NNE-SSW boundary fault to
(Powers, 1968), changing from the clastic facies the west and a strong stratigraphical trap component
of the surface section to the west, and the new consisting of channel sands and pinch-outs. Facies
Kuwait Marrat reserves are housed in carbonate range from marine in the west to fluvio-continental
reservoir(s). in the east. The source of the oil is not firmly
The minor offshore Qatar Mesozoic clastic reser­ established but appears to be 'basinal' Middle
voir accumulation occurs in an areally limited struc­ Jurassic Barnea Formation carbonates located off­
tural closure over a portion of the northern flank shore to the west, with TOC values of 0.5-2.6% of
of the 'supergiant' Permian Khuff carbonate gas type II kerogen; other source candidates are laterally
accumulation of the areally huge but very gentle equivalent basinal Lower Cretaceous shales filling
North (Dome) field. The reservoir is a thin Nahr canyons cut in the Late Jurassic ( Gilboa et al.,
Umr (Albian) fine-grained sandstone, locally 1990). The principal reservoir intervals are sands,
cemented, and tight to the southeast of the closure but a Lower Cretaceous dolomite also produces.
but shaling out westwards (Wells, 1987). Thus, the The sandstone reservoirs vary in characteristics and
thin and areally limited reservoir sand facies, shaley environmental facies and include three main types :
at the base and top, represents the distal clastic an offshore (marine) environment calcite-cemented
regional facies of the Nahr Umr at this distance from sandstone; a tidal channel and/or lagoonal environ­
the main fluvio-deltaic depocentres of the Basra­ ment sand, calcite or dolomite cemented; a coastal
Kuwait area. The hydrocarbon source for this sand­ aeolian sand, very fine-medium grained, well
stone reservoir (and also for the overlying and sorted, loosely packed, low cement content and with
underlying Cretaceous carbonate permeability porosities reaching 32% (average 24%) and good
pinch-out reservoir accumulations on this same north permeabilities at times exceeding 2000mD; most of
flank of the North field) is the Upper Jurassic Hanifa the recoverable oil is in the coastal environment
bituminous carbonates, as discussed by Murris facies and the oil gravity is 27-31° API. However,
(1980). The migration route was complex ; initially reserves are modest with 38 MM barrels in place and
up the flank of the Qatar Arch under the Upper 19 MM barrels being the estimated recoverable;
Jurassic Hith anhydrite seal whence most of the oil with cumulative production until 1988 standing at
moved towards the then culmination in southern 16.5 MM barrels the field must be close to exhaustion
Iran, but some was retained in palaeostructures or (Gilboa et al., 1990).
stratigraphical traps. Later, during late Tertiary By far the most important Mesozoic clastic reser-­
times, when the North (Dome) structure was formed, voir oil region of the Levant-Sinai fragment of the
this oil remigrated southwards, where the Hith Middle East is the Gulf of Suez. There, at least 10
anhydrite seal had been breached allowing vertical fields produce oil from multiple sandstone reservoirs
access into the Cretaceous reservoirs of the area of Cretaceous age, including Lower Cretaceous
(Wells , 1987). Nubia 'A' sands, Cenomanian sands , Turonian sands
 Middle East clastic oil and gas reservoirs 349

and lower Senonian sands (Ayouty, 1990). The Cenozoic commercial oil production from sandstone
depositional environments charactizing the different reservoirs. These are the Kuwait-Basra-Ahwaz
Cretaceous reservoirs vary. In Late Jurassic to pre­ (Iran) area at the head of the Gulf, and the Gulf of
Aptian time, the Gulf of Suez region was positive Suez rift of Egypt. Everywhere else, Cenozoic pro­
with erosion or non-deposition taking place. By the duction is from carbonate reservoirs, with the
Aptian, the northern extremity was within a delta­ principal prolific producing reservoirs being the
front domain and during the Albian, the entire Oligo-Miocene Asmari Formation limestone of
region was receiving alluvial deposition, with a flood­ southwest Iran and its correlative of north Iraq.
plain environment belt straddling the Gulf and Towards the end of the Oligocene, the present
extending across the adjacent Eastern Desert of head of the Gulf area was situated at about 25°N, a
Egypt and Sinai. During the Cenomanian, the Gulf time coinciding with a global sea-level lowstand.
of Suez, much of Sinai and the northern part of the Much of Arabia was positive and uplifted, with the
Eastern Desert were located in an open-marine exception of a narrow seaway along the line of the
regime, the Cenomanian reservoir sands thus being present Zagros range (Beydoun et al., 1992). A
mainly shallow marine in facies. This situation con­ major river rising in the highlands of Central Arabia
tinued in the Turonian but with the southern Gulf of flowed northeast into the shallow shelf seas of the
Suez area alternating between shallow marine and northern Gulf area, prograding a large delta in the
fluvial -deltaic. By the Coniacian, the alternat­ Kuwait-Basra sector (Ghar Sandstone Formation);
ing shallow-marine and fluvio-deltaic facies had a pattern not unlike that initiating the deposits of the
expanded to include the Egyptian Red Sea and all of Cretaceous Zubair-Burgan-Wara Formations.
the Eastern Desert while the northern part of the From this delta, distal reservoir-quality sand tongues
Gulf of Suez became uplifted as a ridge, with open­ extended into the southwestern part of Khuzestan
marine conditions to its east and north. In the Province in Iran, to form the Ahwaz Sandstone
Santonian, the whole of the Gulf of Suez region Member of the Asmari (carbonate) Formation
was within the open-marine-embayment domain (Setudehnia, 1972; Beydoun et al . , 1992). The
bounded by positive uplifts to the north and south Ahwaz Member is considered stratigraphically
and southeast. In the late Campanian the Gulf of equivalent to the Ghar Formation of Kuwait and
Suez region became part of the phosphatic 'flint' southeast Iraq (Basra area) and is of imprecise
facies or 'shale with bone beds' facies covering much Miocene age (Dunnington et al., 1959). During
of Egypt and extending into the Sinai-Levant region deposition of the lower part of the Asmari carbonates
and across North Africa (Said, 1990). in Khuzestan, sands were being flushed from the
Porosities and permeabilities in these reser­ west into this essentially carbonate shelf area; this
voirs are variable but generally good to fair, and oil situation continued until finally the carbonate facies
gravities vary greatly in the different reservoirs , of the Asmari prevailed (Setudehnia, 1972). The
from 20° to 39° API (Ayouty, 1990). These oils, as Ahwaz sands die out eastwards, but in a number of
with those in the Upper Palaeozoic Nubia 'C' oil fields in southwest Khuzestan, such as Ahwaz,
sandstone reservoir, are all thought to be sourced Mansuri and Marun, where the sands are of near­
principally from Lower Miocene sediments with shore, littoral -shoreline and aeolian environments,
Eo-Palaeocene and Upper Cretaceous sediments they constitute a dominant or significant part of the
being probable additional sources (Ayouty, 1990). Asmari reservoir and are clean, loosely cemented,
However, as stated in the previous section, there is have porosities of 20-30% and permeabilities in the
an alternative view postulating the Carboniferous darcy range (Beydoun et al., 1992). The oil produced
Nubia 'B' black shales as the source, at least for the is of 30-33° API gravity and field reserves are sub­
Palaeozoic and Cretaceous reservoirs. Sourcing from stantial, with exceptional flow rates due to the excel­
the younger units would require fault or uncon­ lent reservoir properties.
formity juxtaposition of source against reservoir and
involve only short-distance migration.
Clastic reservoirs in the Kuwait -Basra-Ahwaz area
In the Kuwait-Basra area a few small shallow fields
CENOZOIC CLASTIC RESERVOIRS produce from the Miocene Ghar sandstone reservoir,
but the oil is heavy and sulphurous, ranging between
Only two areas o f the Middle East have established 18° and 20° API gravity. The fields with producible
350 Z. R. Beydoun

Ghar oil are Zubair near Basra in Iraq, and Bahra Karim Formation is similarly varied in facies but
and Ratga in Kuwait, and the reservoir facies is broadly consists of a lower, characteristically evap­
fluvio-deltaic; the heavy nature of the oil is either oritic member and an upper, fine-coarse-grained
because of biodegradation or of early maturity of clastic member; the fine clastic sediments are shales
source, but the source itself remains debatable and marls, and the evaporite member also contains
(Beydoun, 1988). intercalated shales and marls. The evaporites are
absent locally owing to lateral facies changes, so that
the formation becomes mainly shale-marl in com­
Clastic reservoirs in the Gulf of Suez
position, with sand levels that in marginal areas
In the Gulf of Suez, the principal productive sand­ become the dominant clastic deposits and hence
stone reservoirs of the region are of Miocene age good reservoirs; environments range from shelf­
and contain the bulk of the reserves in many fields. edge through the sublittoral (Hughes et al., 1992).
The main oil-bearing levels are sandstones (although The Belayim is typically evaporite and shale, but in
occasionally some limestones too) in the Nukhul marginal areas and laterally it can become pre­
(oldest), Rudeis and Karim Formations of the dominantly sand, with carbonates developing on
Gharandal Group (Lower to Middle Miocene), submerged palaeorelief; it is mostly of a sublittoral
representing the main Neogene clastic cycle, and in environment, generally shallow to deeper inner sub­
the overlying Middle Miocene Belayin Formation, littoral with localized lagoonal conditions (Hughes
which marks the beginning of the main Miocene et al., 1992). The South Gharib Formation is typically
evaporite cycle represented by the Ras Malaab a thick evaporite body with intercalations of shale
Group (Ayouty, 1990; Hughes et al., 1992). A num­ and fine to very fine-grained, well-sorted sandstones.
ber of thin sand reservoir levels are also present in The Zeit Formation represents a shallow-marine
the overlying Middle -Upper Miocene South Gharib regime with alternating restricted and open con­
Formation and the succeeding mainly Upper ditions (Hughes et at., 1992).
Miocene Zeit Formation (the youngest of the Ras Porosities and permeabilities in these formations
Malaab Group). The main oil-bearing reservoir units are variable but generally good to fair. The reservoirs
occur in the lower four formations (Ayouty, 1990). are sourced by oils generated in the Lower Miocene
The main reservoirs of the Morgan, Belayim Land Rudeis and Kareem Formations, as in the case of
and Marine fields, and July, Shoab Ali and Zeit Bay oils in Mesozoic and older reservoirs (Ayouty, 1990),
fields are in the Rudeis-Karim Formations. Belayim although, as discussed earlier, additional input from
Formation oil-producing reservoirs are found in older sources is not ruled out.
Belayim Land and Marine fields, Morgan and Shoab The northernmost Red Sea has an identical
Ali fields, with less important production coming Miocene succession to that of the Gulf of Suez (as
from the same reservoirs in several other fields. The also the remainder of the Red Sea with recognizable
Nukhul sandstone reservoir is oil productive in the Gulf of Suez units; Hughes & Beydoun, 1992). So
Rudeis, Sidri, Shoab Ali and GS-173 fields, but far, however, only one hydrocarbon field, Barqan,
some production is also obtained from contempor­ has been discovered in the northernmost sector of
aneous carbonate reservoirs deposited as reefs the Arabian side in Saudi Arabia, where light oil,
or platform carbonates laid down on submerged gas and condensate were tested in four sand reser­
palaeohighs. voirs, in Rudeis-Kareem and Belayim Formation
The Nukhul Formation is the basal marine equivalents, but the field remains undeveloped
Miocene unit and reflects an irregula_r reljef in the despite promising flow rates (Beydoun, 1988).
basin floor, leading to a diversity of sediments
represented by bioclastic limestones, sandstones, or
shales with sabkha-type anhydrites present in the CONCLUSIONS
central and southern parts of the Gulf of Suez. The
Rudeis consists of highly variable lithology, thickness The bulk o f oil and gas production i n the Middle
and depositional environments, controlled by syn­ East region comes from Permian to Miocene reser­
depositional fault-block topography and represented voirs, the overwhelming proportion of which are
by marls, limestones or shales with minor sands, carbonates that were deposited when the region lay
willch, however, dominate locally at the expense of mainly in warm low latitudes where marine carbon­
the shales (Hughes et al., 1992). The overlying ate sedimentation dominated; sandstone sedimen-
 Middle East clastic oil and gas reservoirs 351

tation during that time was subordinate and was Rub al Khali basin, but in the Gulf region they
related to particular climatic or tectonic conditions. are now too deeply buried and have become over
Arid or warmer, wet climatic episodes coupled with mature; but the hydrocarbons they had generated in
local or regional uplifts and/or sea-level lowstands earlier geological time have probably charged Lower
resulted in substantial localized aeolian or fluvio­ Palaeozoic reservoirs lying up-dip and awaiting dis­
deltaic deposition. Tectonic phases allowed detritus covery. Elsewhere , exploration for these Infra­
to inundate local, rapidly subsiding rift basins; clastic Cambrian/Early Cambrian sourced hydrocarbons
detritus from developing and rising collisonal ­ outside Oman has been minimal and their potential
compressional mountain belts along the northern remains to be assessed by drilling, especially in the
and eastern peripheries rapidly infilled marginal Rub at Khali basin.
seaways. The Lower Silurian source sediments were depo­
During the pre- Late Permian Palaeozoic, by sited regionally in response to global sea-level rise
contrast, the Middle East region was located in the and they have been geochemically shown to have
southern hemisphere in middle to high temperate sourced the sandstone-hosted oils of central Arabia
(humid-cool) latitudes with only occasional influ­ and, locally, of western Oman. Where they lie deeply
ence from warmer belts , hence the dominance of buried under the Gulf, the consensus of specialist
clastic sedimentation in general and of sandstone opinion is that they are responsible for the vast gas
facies in particula�;. reserves housed in the Upper Permian Khuff car­
The distribution of favourable (hydrocarbon) bonated reservoirs of the region (Beydoun, 1991).
reservoir facies by itself is insufficient for the dis­ Post-Silurian sandstone reservoirs , located up-dip
covery of commercial petroleum fields unless widely from the kitchen areas where these Lower Silurian
distributed organic-rich source facies are present to shales are preserved , are highly prospective in an arc
generate the hydrocarbons. Retention of hydro­ around the Arabian Shield, from the Rub al Khali to
carbons in producible reservoirs also requires southwest Iraq and Jordan passing by central and
impervious seal facies that are suitably formed northwest Saudi Arabia; similar areas in southeast
geometrically (by tectonics or stratigraphy) into a Turkey and Syria up-dip from similar Lower Silurian
trap juxtaposed to the source. Moreover, the dis­ kitchen areas are equally prospective. Many Palaeo­
covery of these accumulations commences with the zoic basins became inverted in the Late Palaeozoic
most easily accessible , invariably the shallowest (gen­ and some of the sedimentary section was stripped
erally also the youngest) reservoirs. In time, older off, so that better control of the palaeogeographical
and deeper accumulations are explored for as the subcrop distributions of these source rocks is first
shallower ones are exhausted or become insufficient required before exploratory drilling successes can
to meet rising demand. In the highly petroleum-rich follow routinely.
Middle East region, the need to explore deeper than The successful search for additional Mesozoic or
the Mesozoic reservoirs was for many years neglected younger sandstone reservoirs in the region depends
because of the huge reserves established in the mainly on the identification of 'disguised' rift basins,
Mesozoic and Tertiary (predominantly carbonate) where source rocks have been buried to maturation
reservoirs. It is only in recent years that attention and thus charge suitably juxtaposed sandstone reser­
has turned to the (deeper) Palaeozoic sequences, voirs, as in the cases of Yemen and the Euphrates
where sandstone reservoirs predominate and recent graben in Syria. The Jordan and other grabens merit
successes in Oman and in Central Arabia have more exploration with these models in mind.
opened up a vast new exploration domain for the Unlike the wide areal continuity exhibited by
whole region. Mesozoic-Cenozoic carbonate reservoir facies and
Exploration drilling results in Oman and in Cen­ evaporitic regional seal facies, the Palaeozoic clastic
tral Arabia have shown that the major new reserves deposits generally show rapid lateral and vertical
of hydrocarbons housed in Cambro-Ordovician changes in continuity of the reservoir-seal couplets
and Permo-Carboniferous sandstone reservoirs are (although exceptions have been demonstrated
sourced from two main regionally distributed source­ in Oman and elsewhere); hence major vertical
rock sequences of Infra-Cambrian/Early Cambrian migration of hydrocarbons , either to dissipation at
and Early Silurian age. the surface or until stopped by younger more effec­
The former are thought to be present over much tive seals and diverted to accumulate in integral
of the Gulf region and over Central Arabia and the traps , can be expected. This factor will, in general,
352 Z. R. Beydoun

affect the size of many Palaeozoic accumulations, BEYDOU N , Z.R. (1966) Eastern Aden Protectorate and
because the traps at this level will be subject to part of Dhufar. In: Geology of the Arabian Peninsula.
U.S. Geol. Surv. Prof. Pap., 560-H, 49 pp.
fault control and thin seals can be breached often.
BEYDOUN, Z. R. (1988) The Middle East: Regional Geology
Nevertheless, the total hydrocarbons generated by and Petroleum Resources . Scientific Press, Beaconsfield,
regionally distributed source rocks are believed to 292 pp.
be enormous and the exploration potential rewards BEYDOUN, Z. R. (1991) Arabian Plate hydrocarbon geology

are believed to be huge potentially. and potential - a plate tectonic approach. Am. Assoc.
petrol geol. Stud. Geol. , 33, 77 pp.
BEYDOUN, Z . R. & GREENWOOD, J.E.G.W. (1968) Aden
Protectorate and Dhufar. In: Lexique Stratigraphique
ACKNOWLEDGEMENTS International, Vol. III, Asie, Protectorat d'Aden et
Dhufar (Ed. Dubertret, L.). CNRS Paris, fasc. 10b2,
128 pp.
The author acknowledges with gratitude the typing
BEYDOUN, Z.R., HUGHES CLARKE, M.W. & STONELEY, R.
and draughting support provided by the Exploration (1992) Petroleum in the Zagros Basin: a late Tertiary
Department of Marathon International Petroleum Foreland Basin overprinted onto the outer edge of a vast
(GB) Ltd in London during the preparation of this hydrocarbon-rich Palaeozoic-Mesozoic passive-margin
paper, and particularly to Mrs Wendy Milligan for shelf. In: Foreland Basins and Foldbelts (Eds Macqueen,
R.M. & Leckie, D.A.), Mem. Am. Assoc. petrol. Geol .,
drawing the figures and Miss Susan Hambrook for Tulsa, 55, 309-339.
typing the manuscript. BRENNAN, P. (1990) Greater Burgan field. In: Treatise of
To Dr M.W. Hughes Clarke, sincere thanks are Petroleum Geology, Atlas of Oil and Gas Fields; Struc­
expressed for his continuing role as the authors' tural Traps I, Tectonic Fold Traps (Compilers Beaumont,
E.A. & Foster, N . H.), pp. 103-128. American Associ­
active 'sounding board' and for critically reading and
ation of Petroleum Geologists, Tulsa.
discussing the contents of the paper and providing BRENNAN, P. (1991) Raudhatain field - Kuwait, Arabian
many constructive and helpful ideas and suggestions. Basin. In: Treatise of Petroleum Geology Atlas of Oil
Thanks are also expressed to Dr R. Crossley for and Gas Fields; Structural Traps V, ( Compilers Foster,
reviewing the paper and for providing a number of N . H. & Beaumont, E.A.), pp. 187-210. American
Association of Petroleum Geologists, Tulsa.
helpful comments and suggestions. Gratitude is also DANIEL, E.J. (1963a) Syrie Interieure. In: Lexique Stra­
expressed to Dr A.P. Heward for providing new tigraphique International, Vol. III, Asie, Liban, Syrie,
insight into the Early Palaeozoic of Oman which has Jordanie (Ed. Dubertret, L.), fasc. 10 cl, pp. 157-291.
greatly improved that section of the paper, and for CNRS, Paris.
DANIEL, E.J. (1963b) Jordanie (Ensemble). In: Lexique
his other helpful remarks and suggestions.
Stratigraphique International, Vol. III, Asie, Liban,
Syrie, Jordanie (Ed. Dubertret, L.) fasc. 10 cl ,
pp. 293-399. CNRS, Paris.
REFERENCES
DERCOURT, J . , ZoNENSHAJN, C.P., Rtcou, L.-E., KAZMtN,
V.G., LE PICHO N , X., KNIPPER, A.L., et al. (1986)
Asu-Au, M.A., FRANZ, U.A., SHEN , J . , MONNIER, F., Geologic evolution of the Tethys belt from the Atlantic
MAHMOUD, M . D . & CHAMBERS, T.M. (1991) Hydrocar­ to the Pamirs since the Lias. Tectonophysics, 1221,
bon generation and migration in the Palaeozoic sequence 241-315.
of Saudi Arabia. Proceedings, 7th SPE Middle East Oil DuBERTRET, L. (1963) Liban, Syrie. In: Lexique Stra­
Technical Conference, November 1991, Bahrain, SPE tigraphique International, Vol. III, Asie, Liban, Syrie,
21376, pp. 345-356. Jordanie (Ed. Dubertret, L.) fasc. 10 cl, pp. 7-155,
Au, A. R. & SILDAWI, S.J. (1989) Hydrocarbon potential CNRS, Paris.
of the Palaeozoic pre-Khuff clastics in Abu Dhabi, D u N N INGTON , H.V., VAN BELLEN, R.C., WETZEL, R. &
U.A.E. Proceedings, 6th SPE Middle East Oil Technical MoRTO N , D.M. (1959) Iraq. In: Lexique Stratigraphic
Conference, March 1988, Bahrain , SPE 18009, pp. 819- International, Vol. III, Asie, Iraq (Ed. Dubertret, L.).
832. CNRS, Paris, fasc. lOa, 333 pp.
ALSHARHAN , A.S. & KENDALL, e.G. St C. (1986) Pre­ FERGUSON , G.S. & CHAMBERS, T.M. (1991) Subsurface
cambrian to Jurassic rocks of Arabian Gulf and adjacent stratigraphy, depositional history and reservoir devel­
areas: their facies, depositional setting and hydrocarbon opment of the early to late Premian Unayzah Formation
habitat. Bull. Am. Assoc. petrol. Geol. , 70, 977 -1002. in central Saudi Arabia. Proceedings, 7th SPE Middle
AYOUTY, M.K. EL- (1990) Petroleum geology. In: The East Oil Technical Conference, November 1991, Bahrain,
Geology of Egypt (Ed. Said, R.) pp. 567-722. Balkema, SPE 21394, pp. 487-496.
Rotterdam. GILBOA, Y. , FLIGELMAN , H. & DERJN, B. (1990) Heletz-­
BEYDOUN, Z.R. (1964) The stratigraphy and structure of Brur-Kokhav field - Israel. In: Treatise of Petroleum
the eastern Aden Protectorate. Overseas Geology and Geology, Atlas of Oil and Gas Fields, Structural Traps
Mineral Resources Bulletin , Supplement 5, HMSO, IV, Tectonic and non-Tectonic Fold Traps (Compilers
London, 107 pp. Beaumont, E.A. Foster, N.H.) pp. 319-345. American
 Middle East clastic oil and gas reservoirs 353

Association of Petroleum Geologists, Tulsa. potential of the Palaeozoic succession in both Tabuk and
GORIN , G.E., RAcz, L.G. & WALTER, M.R. (1982) Late Widyan Basins, Arabia. In: Future Petroleum Provinces
Precambrian-Cambrian sediments of Huqf Group, of the World (Ed. Halbouty, M.T.) Mem. Am. Assoc.
Sultanate of Oman. Bull. Am. Assoc. petrol. Geol. , 66, petrol. Geol., Tulsa, 40, 373-397.
2609-2627. LEVELL, B.K., B RAAKMAN , J.H. & R UTTE N , K.W. (1988)
G RANDVILLE, B.F. D E LA (1982) Appraisal and develop­ Oil-bearing sediments of Gondwana glaciation in Oman.
ment of a stratigraphic and structural trap oilfield with Bull. Am. Assoc. petrol. Geol. , 72, 775-796.
reservoirs in glacial and periglacial clastics. In: The LouTH, G. & BISHLAWY, S. EL- (1989) Geology and hydro­
Deliberate Search for the Subtle Trap (Ed. Halbouty, carbon potential of the de.ep Palaeozoic sequence in Abu
M.T.) Mem. Am. Assoc. petrol. Geol., Tulsa. 37, Dhabi. In: OAPECIADNOC Seminar on Deep Forma­
267-286. tions in Arab Countries: Hydrocarbon Potential and
GRANTHAM, P.J., LUMBACH, G.W.M., POSTHUMA, J., Exploration Techniques, Abu Dhabi, October 1 989,
H uGHES CLARKE, M . W . & W !LLIN K , R.J . (1988) Origin Proceedings Technical Papers, pp. 85-129.
of crude oils in Oman. I. petrol. Geol. , 1 1 , 61-80 . MAHMOUD, M.D., VASLET, D. & H u sSEIN !, M . l. (1992)
H ARPUT, O.B. & E RTURK, 0. (1991) The organic geo­ The Lower Silurian Qalibah Formation of Saudi Arabia.
chemical evolution of the 9th and lOth Districts of SE Bull. Am. Assoc. petrol. Geol. , 76, 1491-1506.
Anatolia. J. Southeast Asian Earth Sci. 5, 421-428. MARJEBY, A . AL- & NASH, D. (1986) A summary of the
H EWARD, A.P. (1989) Early Ordovician alluvial fan geology and oil habitat of the Eastern Flank Hydrocar­
deposits of the Marmul oil field, South Oman. J. geol. bon Province of South Oman. Mar. petrol. Geol. , 3,
Soc. London, 146, 557-565. 306-314.
H EWARD, A.P. (1990) Salt removal and sedimentation in MATTES , B.W . & CoNWAY MoRRis, S. (1990) Carbonate/
southern Oman. In: The Geology and Tectonics of the evaporite deposition in the Late Precambrian-Early
Oman Region (Eds Robertson, A.H.F., Searle, M.P. & Cambrian Ara Formation of southern Oman. In:
Ries, A.C.) Spec. Pub. Geol. Soc. London, No. 49. The Geology and Tectonics of the Oman Region (Eds
pp. 637-652, Geological Society of London, Bath. Robertson, A.H.F., Searle, M.P. & Ries, A.C.) Spec.
H uGHES, G . W. & BEYDO U N , Z.R. (1992) The Red Sea­ Publ. Geol. Soc. London, No. 49, pp. 617-636. Geo­
Gulf of Aden: biostratigraphy, lithostratigraphy and logical Society of London, Bath.
palaeoenvironments. J. petrol. Geol. , 15, 135-156. MAYCOCK, I. (1987) Oil exploration and development in
H uGHES, G.W., AB D I N E , S. & G IRGIS, M.H. (1992) Mio­ Marib/AI Jauf Basin, Yemen Arab Republic. Proceed­
cene biofacies development and geological history of the ings, 5th SPE Middle East Oil Technical Conference,
Gulf of Suez, Egypt. Mar. petrol. Geol. , 9, 2-28. March 1987, Bahrain , SPE 15685 (unpublished).
H uGHES C LARKE, M.W. (1988) Stratigraphy and rock-unit McCLURE, H.A. (1980) Permo-Carboniferous glaciation
nomenclature in the oil-producing area of Interior Oman. in the Arabian Peninsula. Geol. Soc. Am. Bull. , 91,
I. petrol. Geol. , 11, 5-60. 707-712.
H usSEIN!, M.l. AL- (1992) Potential petroleum resources McGILLIVRAY, J.G. & H u ss EIN !, M.I. (1992) The Palaeo­
of the Palaeozoic rocks of Saudi Arabia. Proceed­ zoic petroleum geology of Central Arabia. Bull. Am.
ings, 13th World Petroleum Congress, Buenos Aires, Assoc. petrol. Geol. , 76, 1473-1490.
November 1991 , Vol. 2, pp. 3-13. MILLS, S.J. (1992) Oil discoveries in the Hadramaut: how
H UURDEMA N , A.J.M., BREUNESE, J.N., ASBAHI, A.AL-, Canadian Oxy scored in Yemen. Oil Gas J. , 9 March,
L UTGERT, J.E. & FLORIS, F.J.T. (1989) Assessment of 49-52.
halite-cemented reservoir zones with a combined geo­ MuRRIS, R.J. (1980) The Middle East: stratigraphic evol­
logical/engineering approach; a case study (Yemen). ution and oil habitat. Bull. Am. Assoc. petrol. Geol. , 64 ,
Proceedings, 64th SPE Annual Technical Conference, 597-618.
San Antonio, TX, October, 1 989, SPE 19603, pp. 335- NRA (NATURAL RESOURCES A UTHORITY) J ORDAN (1989)
343. Petroleum Exploration Opportunities in Jordan , Amman,
IBRAHIM, M.W. (1981) Lithostratigraphy and subsurface 39 pp.
geology of the Albian rocks of South Iraq. J. petrol. OAPEC (O RGAN IZATION OF A RAB P ETROLEUM E XPORTING
Geol. , 4, 147-162. Cou NTRIES) (1985) Register of Exploration Wells and
IBRAHIM, M.W. (1983) Petroleum geology of southern Iraq. Arab Petroleum Fields (in Arabic). OAPEC, Kuwait,
Bull. Am. Assoc. petrol. Geol. , 67, 97-130. 500 pp.
J AWAD Au, A. & Az1z, Z.R. (1993) The Zubair Forma­ OwEN , R.M . S . & N ASR, S.N. (1958) Stratigraphy of the
tion, East Baghdad oilfield. J. petrol. Geol. , 16, Kuwait-Basra area. In : Habitat of Oil (Ed . Weeks,
353-364. L.G.), AAPG Symposium , pp. 1252-1278. Tulsa,
KHAIWKA, M.H. (1989) Order and repetitive nature of the Oklahoma.
Zubair deltaic facies, central and southern Iraq and PowERS, R.W. (1968) Arabie Saoudite. In: Lexique Stra­
northern Kuwait. Mar. petrol. Geol. , 7, 38-43. tigraphique International, Vol. III, Asie (Ed. Dubertret,
KUTZCH, E. (1990) Palaeozoic. In: The Geology of Egypt L.). CNRS, Pa1is, fasc. 10b1, 177 pp.
(Ed. Said, R.), pp. 393-406. Balkema, Rotterdam. SAID, R. (1990) Cretaceous palaeogeographic maps. In:
LABABIDI, M.M. & H AMDAN , A . M. (1985) Preliminary The Geology of Egypt (Ed. Said, R.), pp. 439-449.
Lithostratigraphic Correlation Study in OAPEC Member Balkema, Rotterdam.
Countries. Organization of Arab Petroleum Exporting SAMAHUI, J .A. & CHAUBE, A.N. (1987) Evolution of
Countries (OAPEC), Kuwait, 171 pp. Bahrain Field in relation to western Arabian Gulf Basin.
LABOUN, A.A. AL- (1986) Stratigraphy and hydrocarbon Proceedings, 5th SPE Middle East Oil Technical Confer-
354 Z.R. Beydoun

ence, March 1987, Bahrain, SPE 15697, pp. 97-107. J.) Spec. Pub!. Geol. Soc. London, No. 50, pp. 293-
ScOTESE, C.R . & McKERROW, W.S . (1990) Revised world 298. Geological Society of London, Bath.
maps and introduction. In: Palaeogeography and Bio­ SYKES, R.M. & All u RISHEH, A . K. (1989) Exploration of
geography (Ed. McKerrow, W.S. & Scotese, C.R. ) , deep Palaeozoic and Pre-Cambrian plays in the Sultanate
Mem. geol. Soc. London, No. 12, pp. 1-21 . Geological of Oman. In: OAPECIADNOC Seminar on Deep For­
Society of London, Bath. mations in the Arab Countries: Hydrocarbon Poten­
SCOTESE, C.R., BAMBACH, R . K . , BARTO N , C., VAN DER tial and Exploration Techniques, Abu Dhabi, October
Voo, R. & ZEIGLER, A.M. (1979) Palaeozoic base maps. 1989, Proceedings, Country Reports and Case Studies,
J. Geol. , 87, 217-277. pp. E71 -113.
SETU DEHNIA, A.O. (1972) Iran du Sud-Ouest. In: Lexique WELLS, P.R.A. ( 1987) Hydrodynamic trapping of oil and
Stratigraphique International, Vol. III, Asie, (Ed. gas in the Cretaceous Nahr Umr Lower Sand of the
Dubertret, L.) fasc. 9b, pp. 285-376. CNRS, Paris. North Area, offshore Qatar. Proceedings, 5th SPE
STONELY, R. (1990) The Middle East basin: a summary Middle East Oil Technical Conference, March 1 987,
overview. In: Classic Petroleum Provinces (Ed. Brooks, Bahrain, SPE 15683, pp. 17-26.
Spec. Pubis int. Ass. Sediment. (1995) 22, 355-380

The evolution of Oligo-Miocene fluvial sand-body geometries

and the effect on hydrocarbon trapping:
Widuri field, west Java Sea

R AY Y O U N G ,* W. E. H A RM O NYt and T H OM A S B U D I Y E N T Ot
*Maxus Ecuador Inc. , PO Box 650496, Dallas, Texas 75265-0496, USA; and
tMaxus Southeast Sumatra Inc., Five Pillars Office Park,
]alan Letjen MT Haryono 58, PO Box 2759,
Jakarta, Indonesia

ABS T R A C T

Widuri field is a faulted anticline located in a Tertiary back-arc basin at the western end o f the Java Sea
in the Indonesian archipelago. The discovery well was drilled in early 1988 to a depth of 3735 ft subsea
(1139 m SS) and penetrated 170ft (52 m) of net oil pay in late Oligocene and Early Miocene sandstones.
Eight delineation wells, 47 development wells and five sidetracks confirmed a large oil field over an area
of approximately 7500 acres.
The oil is trapped in six separate reservoirs over a vertical interval of 325ft (99 m). The lowermost
reservoir is a thick, immature, fluvial sandstone with multistorey and multilateral stacking, high
permeabilities and a sheet geometry; the uppermost reservoir is a series of thin, lower permeability,
2000 ft (610 m) wide, distributary channel sandstones associated with a tide-influenced delta. The
intermediate reservoirs show a complete gradation in size, geometry and reservoir quality between these
two extremes. This gradual change from an immature fluvial system, through progressively more mature
fluvial systems into deltaic sediments at the top is the result of a gradual relative rise in sea-level.
The systematic upwards decrease in sand-body width is accompanied by a change in the trapping
mechanism, from structural in the two sheet-like reservoirs at the base to stratigraphical-structural in
the upper four reservoirs, where the existing structural closure is enhanced significantly through each
reservoir having a restricted lateral extent and a favourable orientation with respect to structure.
However, the Widuri reservoirs share a common oil-water contact defined by the spill point of the
stratigraphically highest of the sheet-like reservoirs. The reservoirs must therefore be in communication
and, without a stratigraphical component to trapping in every one of the overlying four reservoir
intervals, the field oil-water contact would be significantly higher and the field size very much smaller.
The understanding of sand body geometry and trapping has led to efficient field development (only
two of the original holes needed to be side-tracked owing to lack of pay) and will form the basis for
future reservoir management. From an exploration viewpoint , the Widuri work has provided analogue
models in the search for other stratigraphically trapped oil in the contract area .

I N T R O DU C T I O N

Widuri field is located in the Southeast Sumatra 500 million barrels of oil in late 1989 but virtually all
Production Sharing Contract (PSC) area approxi­ of this oil came from the Sunda basin. It was not
mately 100 miles ( 161 km) north of Jakarta in the until late 1987 that the first oil was discovered in the
west Java Sea (Fig. 1). The PSC comprises approxi­ Asri basin and this discovery subsequently led to the
mately 5300 square miles ( 13 725 km2) and consists establishment of five new fields on the northwest
of two principal areas, the Sunda and Asri basins. flank of the basin (Fig. 2).
Cumulative oil production from the PSC surpassed Widuri field is by far the largest of these accumu-

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
355
35 6 R. Young, W.E. Harmony and T. Budiyento

SOUTH CHINA SEA

Natuna

JAVA SEA

0 500

Miles

Fig. 1. Location of the Pertamina-Maxus , Production Sharing contract area showing the relative position of the Sunda
and Asri basins and the distribution of oil fields.

lations and was discovered in early 1988 when relationship between reservoir geometry and struc­
Widuri-1 was drilled to a total depth of 3735 ft ture at Widuri field and to demonstrate the resulting
subsea (1139m SS) and penetrated 1 70ft (52m) of hydrocarbon trapping configurations.
net oil pay in multiple sandstones of late Oligocene
and Early Miocene age. Eight delineation wells,
47 development wells and five sidetracks (as of mid- BA C K G R O U N D G E O L O GY
1992) confirmed a giant oil field with over 500 million
barrels of original oil in place spread over approxi­ The PSC area is situated in a Tertiary back-arc basin
mately 7500 acres. Recovery efficiency is expected and the history of tectonics in the region is therefore
t9 be excellent and the discovery was the largest in essentially extensional but with some wrench modi­
Indonesia since 1974. The oil is a high pour point, fication (see also Hutchison, 198 6). Structurally, the
low sulphur crude with a gravity of 31 API. First oil
o Asri basin is an asymmetrical syncline or half-graben
was produced in December 1990 and there are with a roughly N-S axial trend (Fig. 2) and a
currently five platforms (A-E, Fig. 3). To optimize maximum sedimentary fill of approximately 15 OOO ft
reservoir management it was decided to complete (4773 m) of Oligocene and younger sediments. The
wells in single zones only and this strategy meant oil accumulations are trapped in upthrown closures
that a detailed and predictive geological model was related to NE-SW trending, down-to-the-northwest,
vital. normal faults on the gently sloping northwest flank
The primary aim of this paper is to show the of the basin. These faults probably originated in
 Fluvial sand-body geometries 357

early Oligocene times during basin formation but

were subsequently reactivated at several times
throughout the Tertiary. Widuri field is a low relief,
three-way-dip, faulted closure (Figs 3 & 4), which
I formed approximately 19Ma (Early Miocene) and
i has not been altered significantly since.
i The stratigraphy in the Widuri field area is shown
i in Fig. 5. Lower Cretaceous, metamorphic basement
i was penetrated only over the A-platform area, which
i is the structurally highest part of the field. Basement
i is onlapped and overlain by upper Oligocene fluvial
I sandstones, shales and coals of the Zelda Member,
Talang Akar Formation. The Lower Miocene Gita
i Member is a fluvio-deltaic succession of sandstones
i and shales with thin interbedded limestones and
i coals. The boundary between the Zelda and Gita
I members is the top of the Oligocene succession and
______ _j is an unconformity over most of the Asri basin. At
Widuri field, however, the boundary is probably a
discontinuity that is associated with an important
field-wide coal marker - Coal 'A' (see Figs 4 & 5).
The Talang Akar sandstones are currently the only
Cl 50 ms
commercial reservoirs in the Asri basin.
10 Kms
The top of the Talang Akar Formation is an
10 Miles unconformity that is generally marked by a relatively
high gamma-ray (uranium) reading (Fig. 6). The
Fig. 2. Asri basin time-structure map on the top Talang overlying Batu Raja Formation has yielded non­
Akar Formation.
commercial oil shows from thin argillaceous lime­
stones and sandstones, but the predominant lithology
is mudstone which acts as a local seal. The regional

Fig. 3. Depth structure map on the

top 34-2 reservoir showing (1) that
Widuri field is a low-relief, faulted
closure, (2) the location of the five
platforms and 62 wells in the Widuri
area and (3) the location of seismic
line 88-196 (Fig. 4).
358 R. Young, W.E. Harmony and T. Budiyento

NNW
Widuri Field SSE
(approx. 3 miles)
TWT
0.7

.2

Fig. 4. Seismic line 88-196 over Widuri field (see Fig. 3 for location).

seal for both the Sunda and Asri basin is claystone The oil is trapped in six different sandstone units
of the overlying Gumai Formation (Bushnell & (from base to top the 36-1, 35-2, 35-1, 34-2,
Temansja, 1986). The remainder of the succession 34-1 and 33-series sandstones) over a vertical
consists of Miocene, shallow-marine shales, lime­ interval of approximately 325 ft (99 m). The reser­
stones and sandstones, which are overlain by pre­ voirs almost certainly share a common oil-water
dominantly alluvial sediments and volcanics of the contact at 3680ftSS (1122 m SS - see below), but
Plio-Pleistocene. there is no evidence for direct, sand-on-sand com­
An adequate source rock for the oil accumulation munication between any two reservoirs, nor any
has not yet been penetrated in the Asri basin. prediction that this might occur. However, the
However, by analogy with the Sunda basin, it is reservoirs must either be in communication or have
believed that lower Oligocene lacustrine shales are been in communication at some time in the past
present in the central part of the basin and that these through faulting, fractures or seepage. So far,
shales generated oil from Early Miocene until at however, the production history indicates that each
least Late Miocene time. reservoir is acting as a separate sealed unit.
Reservoir descriptions are based on over 650ft
(198 m) of core from seven wells. The sandstones
R E S E RV O I R OVE RVI E W vary from coarse- to very fine-grained and both
conglomerate and granule layers are common .
Figure 6 illustrates the full Talang Akar succession Compositionally, the sandstones are mature, with
in the Widuri A-1 well, one of the few wells in the 75% being quartz arenites and the remainder split
field that penetrated each of the main reservoirs. between sublitharenites and subarkoses. There does
 Fluvial sand-body geometries 359

Paleoenvt.
c::
c:: 0
� .<:::
0
·;:;"' Q; -e "' Sub
::2: <.) .0 "'<.) 0 littoral
0
c. E E Lithology :::: �
Q)
Cl UJ 0
Q) e :.:; 0 "'
<{ LL
::2: "0
> "' :::: Q; Q; >
I c. :.:; c:: 5 -E
"'
:J .!: 0 co
(f)
"
Q)
c::
Q)
-
. . ..:
.·
. "�
.
'· . . .
!}
�B
l�
Q;
"--�
"- <Jl c.
·a; c. " v
c. .<::: :::J
:J
"' .0
:J
<Jl

u
Q;
;:0
-�
Q)
'iii --'
-=:,:.:.:::l
--'

\
Parigi
"' - .....--r-(.
� -
� Q)
:0 �
Q;
"0 c.
Q) "' c.
..,. c::
Q) -
::2: -"
"'c:: :::J
'
.f <.) Q)
.

::2:
2 co
<(
Q;
;:0 � ;:··_......
... ..
.
... ....
� �
--' .
.... :

�
�
Gumai -
-
\.
-.:::"'>
��
D
Batu
UJ
Raja = --�

'""
.f
N

Q)
;;;
-"
<{
Cl
c::
-�
<.:)

��
<::::;:!>

r=2.
":
· ·' • •
•

•
• I
v
c::
"' "' "'-
· ·.· . , . .

Q)
<.)
�
"'
(ij
f-
"0
(ij
. ,, •
.2'
0 --' N � �'( � : .
..- :
. . · . .·

�
.... ..... /��
0 ,t,,/,,;,, ·. · ·

/, /,

�: ,<. � ,"', : \;� � \� �

Early to late Basement t,,t,..._t, ..... :,..., l,
Fig. 5. Stratigraphy over Widuri field. Cretaceous ,,/, .. // <.., "
....,,_,,_,,_ / -:./

not appear to be any significant variation in detrital other mineral cements are extremely sparse and the
composition between the different reservoir units sandstones are generally unconsolidated. Reservoir
or between the different sedimentary facies. The quality is usually excellent but the measurement
maturity of the framework mineralogy probably of porosity and permeability from core data has
reflects the provenance (see Folk, 1968) and this is inherent errors because the reservoirs are uncon­
supported by the preponderance of metasedimentary solidated. However, log porosities average 29%
basement in the Sunda and Asri basins (see also (using a 15% cut-off) and permeabilities from
Ben-Avraham & Emery, 1973). production tests on initial completions range up
Clay minerals dominate the non-framework to 30 darcies (D).
mineralogy but usually only account for less than A reconstruction of the palaeogeography of the
5% of the bulk rock volume in the majority of the area during Talang Akar times is shown in Fig. 7
reservoir quality sandstones. Kaolinite predominates (see also Beddoes, 1981). The basin margin is
with subordinate illite and chlorite, and minor defined seismically by the onlap of the Talang Akar
smectite and illite-smectite mixed-layer clay. All Formation on to basement (Fig. 7) and the sediment
 360 R. Young, W.E. Harmony and T. Budiyento

Fig. 6. The type well, Widuri A-1.

transport directions were then defined using interval hydrocarbon trapping that an updated and abbrevi­
isopachs (see also Wight et a!., 1986) and individual ated version is included.
channel isopachs. The Talang Akar clastic sediments
in the Widuri area of the Asri basin were derived
36-1 sandstone
exclusively from a northerly direction.
The 36-1 sandstone is the lowermost of the Widuri
reservoirs and is oil-bearing only over the structurally
RES ERVOIR GEOMETRY AND TRAP higher A- and D-platform areas. Few down-dip wells
penetrated the entire 36-1 sandstone and therefore
A discussion o f the reservoir character, geometry less information is available compared with other
and trap for each of the six main reservoir units reservoirs. However, the reservoir was fully cored in
follows. These descriptions and interpretations are the discovery well, Widuri-1, where it is 43 ft (13 m)
based on a data set that includes a 250-m grid of very thick and rests directly on top of metamorphic base­
. good quality two-dimensional seismic data, a mostly ment (Fig. 8). The base of the sandstone is erosive
full set of log data from approximately 60 wells, and and the top has a sharp contact with the overlying
a total of over 650ft (198 m) of full-hole core from rootleted siltstone and coal. Grain size is coarse to
seven wells. A detailed discussion of the sand-body very coarse and sedimentary structures are domi­
geometries and reservoir model was the subject of a nated by trough cross-bedding in 1-2-ft (0.3-0.6 m )
previous publication (Young et a!., 1991) and it is sets. The sandstone does not show any overall grain­
not the intent of the present article to duplicate this. size trend but at least five sedimentary units can be
However, reservoir geometry is so critical to the recognized based largely on grain size (Fig. 8). Each
 Fluvial sand-body geometries 361

Java Sea

r.::l Sediment source

� areas/Sunda
Shield emergent
(Talang Akar FM
absent)

Sediment
transport
directions

Fig. 7. Local palaeogeography Asri Basin oil

during Talang Akar Formation fields
times.

unit fines upwards slightly and they are interpreted Figure 9B is a depth structure map on top of the
to represent a multistorey fluvial system. 36-1 sandstone. The structural spill point is located
Elsewhere in the field, sidewall cores indicate a approximately 5000ft ( 1525 m) north-northwest of
more variable grain-size with a slightly finer average, the D platform at a depth of 3740ftSS (1140m SS)
and the thickness reaches a maximum of 185 ft (56 m) and the maximum structural closure is shaded. The
in Widuri E-1 (see Fig. 8). Log character is commonly oil- water contact for this reservoir is at 3680 ft SS
blocky but occasionally, as in the case of Widuri-6 ( 1122 m SS) and this plots inside the closing contour.
(Fig. 8), it is clear that the 36-1 is made up of several The trapping mechanism for the 36-1 reservoir is,
different sandstones. However, this distinction therefore, clearly structural but the structure is only
between individual sandstones cannot usually be partially filled. In other words, there is more struc­
traced with any confidence beyond adjacent wells tural closure than oil column.
and the thick, blocky nature of the 36-1 over most of
the field is almost certainly the result of both vertical
35-2 sandstone
and lateral channel amalgamation.
The 36-1 sandstone overlies and onlaps basement The 35-2 sandstone is separated from the 36-1 by an
and the net sandstone isopach (Fig. 9A) defines a average of 20ft (6 m) of mudstone, siltstone and
relatively thick, sheet sandstone, which is more coal. The reservoir has been cored in three wells but
extensive than the available well control in all areas long sections of core that are representative of the
except the extreme southwest of the field. The sand­ 35-2 sandstone field-wide are not available. The
stone is interpreted to be a coarse grained, immature, sandstone, however, is similar to the 36-1, but with
multistorey and multilateral fluvial sheet-sandstone some important differences. Grain size is more vari­
deposited in a high gradient, high bedload river. able and slightly finer (field-wide average is medium
Reservoir quality is excellent; porosities average to coarse grained) but the sedimentary structures
28% and production test permeabilities range from are still dominated by trough cross-beds in sets up to
20 to 30D. 2 ft (0.6 m) thick. Bed contacts are still abrupt and
362 R. Young, W. E. Harmony and T. Budiyento

A 11 000' 6000' 5000' 10 000' A'

Widuri E-1 Widuri-7 Widuri-4 Widuri-1 Widuri-6
20 API 120 20 API 120 20 API 120 20 API 120 20 API 120

50

100
FT

150
Basement
of weathered basement

200

OEP TH Core log 36-1 sandstone

GAMMA RAY
(FT) Widuri-1
API
0 100

LOCATION MAP Rootleted siltstone

and coal

Trough cross-bedded,
pebbly, coarse to
very coarse grained
sandstone in at
least five different
sedimentary units

Metamorphic
basement and
weathered
basement

Fig. 8. 36-1 reservoir: vertical and lateral sedimentary character.

evidence for multistorey and multilateral stacking of dip using log character correlation. The most sig­
sand bodies is common (see Young et at., 1991); in nificant difference to the 36-1, however, is the
one instance, an individual channel sandstone can shape of the sand body. The net sandstone isopach
be traced for up to 3000ft (915 m) along depositional (Fig. lOA) defines a sandstone that is thinner and
 Fluvial sand-body geometries 363

+ + + +

+ + +

,_�
c�
�0
+ + +

,co>be:, �
0�...
c;� �0
�7
,r-.:\;: {, ,
'
+ + +

j 6
�-�-
+
Cl 20'
5000 FT
A

Fig. 9. 36-1 reservoir. (A) Net

sandstone isopach with the oil­
water contact marked. (B) Depth
structure map on the top 36-1
sandstone showing that the oil­
water contact is 60ft above Cl100'
structural closure (shaded). The 5000 FT
structural spill point is marked by a
triangle. a

less extensive. Net sandstone thickness from well The 35-2 sandstone is interpreted to be a multi­
penetrations varies from 0 to 67ft (0-20m) and a storey and multilateral fluvial system that is less
sand body is defined (or predicted from thickness extensive, slightly finer grained, and therefore
trends) that is elongated roughly N-S and has an probably slightly more mature than the 36-1 system.
average width of approximately 11 000 ft (3354 m) However, the sandstone was still probably deposited
with a zero-edge to the west and an inferred one to in a relatively high gradient, high bedload river.
the east. The abrupt change to a westwards depo­ Reservoir quality is excellent, with log porosity
sitional strike at the north end of the sand body is averaging 29% and test permeabilities ranging from
defined by well penetrations to the north and west 20-30D.
outside of Widuri field. Figure lOB is the depth structure map on top of
364 R. Young, W.E. Harmony and T. Budiyento

+ + +

+ + +

+ +

+ +

+ Cl 20'
5000 FT
A

Fig. 10. 35-2 reservoir. (A) Net

sandstone isopach with the oil­
water contact marked . (B) Depth
structure map on the top 35-2
sandstone showing that the
Cl100'
structural closure (shaded) is
5000 FT coincident with the oil-water
B
contact. The structural spill point is
marked by a triangle.

the 35-2 sandstone. The spill point has -remained 6 ft (1. 8 m ) of the 35-1 sandstone but adequately
in approximately the same position but is now at demonstrates the sedimentary character of the
3680 ft SS ( 1122 m SS) and this is coincident with the reservoir. The 35- 1 is an erosively based sandstone,
oil-water contact for this reservoir. The trapping which is trough cross-bedded in 1-2-ft (0. 3-0. 6 m)
mechanism is therefore again clearly structural, but sets, and which averages medium grain size. Intervals
this time the structure is full to spill. of coarse grain size are common towards the base of
the sandstone and fine/very fine grain size common
towards the top, but the sandstone does not fine
35-1 sandstone
upwards systematically and several distinct units
The Widuri E-1 core (Fig. 11) is lacking the basal with abrupt contacts are evident (see Fig. 11). This
 Fluvial sand-body geometries 365

A A'
D-8 D-6 D-5 B-3 A-1 B-6 C-3
1800' 3000' 3150' 4250' 4400' 3900'
20 API 120 20 120 20 API 120 20
API API 120 20 API 120 20 API 120 20 API 120
� L.-.-----1 L.-.-----1 .______. 1------4 � �

CORE LOG 35-1 sandstone

Widuri E-1

LOCATION MAP

Rootleted shale
and coal

3750
Predominantly medium
to coarse grained,.
trough cross-bedded
sandstone
Abundant mudstone
clasts and carbonaceous
3760
debris

Fig. 11. 35-1 reservoir : vertical and lateral sedimentary character.

multistorey character is readily recognizable in Fortunately, these thickness vanatwns could be

all other 35-1 cores and the units are separated mapped and predicted seismically prior to any
occasionally by coals (see Widuri A-1, Fig. 11). development drilling. The 35-1 interval is extremely
The 35-1 sandstone is the most variable of the well constrained between two readily correlatable
Widuri reservoirs (see cross-section A-A', Fig. 11). markers - Coal 'A' at the top and usually a double
The log character is predominantly gamma-ray coal at the base (see cross-section A-A', Fig. 11).
increasing upwards (fining upwards) with a gra­ Acoustic impedance contrasts associated with these
dational top, but is also commonly blocky with a coal markers are responsible for two seismic reflec­
sharp top. The most striking feature of the sand­ tion which envelope the 35-1 sandstone and can be
stone, however, is the rapid lateral thickness changes traced field-wide and beyond. The isochron between
within an overall range of 0-64 ft (0-19.5 m). these two seismic markers, when linked with
366 R. Young, W.E. Harmony and T. Budiyento

well penetrations, proved to be a very effective trending southwest from the D platform is a single­
predictor of net sandstone distribution (see Young storey channel), and sand body widths vary from
et al., 1991). approximately 3000ft (915 m) in the west to 6500ft
The net sandstone isopach (Fig. 12A) demon­ (1982 m) in the east. Finer grain size, thinner and
strates an anastomosing pattern and the 35-1 narrower sand bodies, and more order to the
sandstone is interpreted to be a multistorey and sequence suggest a more mature fluvial system than
multilateral fluvial system, possibly representing the 35-2. Reservoir quality is again excellent, with
anastomosing rivers. Individual channels can be log porosity averaging 29% and test permeabilities
mapped with more confidence and over a larger area ranging from 10 to 20 D.
than those of the 35-2 and 36-1 (e.g. the sand body The depth structure map on top of the 35-1 sand-

Fig. 12. 35-1 reservoir. (A) Net

sandstone isopach with the oil­
water contact marked. (B) Depth
Cl100'
structure map on the top 35-1
5000 FT sandstone showing that the oil­
water contact is 20ft below
a structural closure (shaded).
 Fluvial sand-body geometries 36 7

stone (Fig. 12B) defines a structural spill point a relatively mature fluvial system but a meandering
or closing contour at 3660ft SS (1116 m SS). The river model (Allen, 1965) is discounted on the lack
oil-water contact for this reservoir, however, is of evidence for lateral accretion and on the geometry
the same as for the 36-1 and 35-2 at 3680 ft SS of the reservoir which is not, so much a sheet as a
( 1122 m SS). There is therefore 20ft (6 m) more oil ribbon of sandstone that tapers towards both sides
column than can be accounted for by structural (see Fig. 14A). Similarly, equations relating sand
closure. Returning to Fig. 12A, it is clear that the body thickness to channel depth and channel belt/
structural contour of the oil-water contact does not sand body width (see Leeder, 1973; Collinson, 1978;
close to the northwest but that closure is effected by Lorenz et al., 1985; Fielding & Crane, 198 7),
the sand body pinch-out to the northwest. There is although having a wide margin of error, predict a
therefore a stratigraphical component to trapping much wider sand body if the 34-2 were a high
and a combination stratigraphical-structural trap sinuosity channel deposit.
results. It is tempting to use a Mahakam River (East
Kalimantan, island of Borneo) type model (Allen
et al., 1979; Young & Atkinson, 1993) as the analogy
34-2 sandstone
for the 34-2 reservoir. In this case, the river channel
The 34-2 sandstone rests directly on top of Coal 'A' does not meander significantly but does have a
and the sedimentary character is demonstrated in meandering thalweg. On the inside of every thalweg
core from the B-8 well (Fig. 13). The base is erosive meander loop is a non-emergent lateral bar which is
on top of Coal 'A' and the thin, very coarse grained, attached to the channel side and progrades down­
basal lag contains coal clasts. The overlying sand­ stream. Echo-sounding and dredge sampling suggest
stone consists of predominantly medium- to coarse­ that the Mahakam River immediately above the
grained sandstone in the basal and middle section delta has a width, sinuosity, maximum sand thick­
before passing into fine- and very fine-grained sand­ ness, grain size and bedform distribution that is
stones in the upper part, and finally into rootleted remarkably similar to the 34-2 sandstone. However,
siltstone and coal at the top. Trough cross-bedding irrespective of the exact mechanism of channel fill,
in 1 ft (0. 3 m) sets is the dominant sedimentary struc­ the 34-2 reservoir is interpreted to be a single storey,
ture. Maximum penetrated thickness for the 34-2 fluvial sandstone. Proximity to the sea is suggested
sandstone is 54ft (16.5 m) and in the thicker sections by the maturity of the fluvial system and by the
the log character tends to be blocky, with a gamma­ overlying transgressive limestone. However, evi­
ray increasing (fining upwards) signature at the very dence for marine conditions within the sandstone is
top (see Fig. 13). In the thinner sections, however, absent. Reservoir quality is excellent, with average
the log character is mostly gamma-ray increasing log porosity of 29% and test permeabilities of
(fining upwards). Immediately above the top coal is 10-30D.
the first indication of fully marine conditions in the The depth structure map on top of the 34-2 sand­
Talang Akar Formation. Calcareous mudstone stone is shown in Fig. 14B and the structural spill
grades up into muddy limestone containing skeletal point/closing contour is at a depth of 3630ft SS
debris of bivalves, large foraminifera, solitary corals (1107 m SS). The oil-water contact for the reservoir
and echinoid fragments. This limestone can be is still 3680 ft SS (1122 m SS) and there is again a
recognized throughout most of the field. discrepancy (50ft or 15m) between oil column and
The 34-2 net sandstone isopach (Fig. 14A) out­ structural closure. It is clear on Fig. 14A, however,
lines a sinuous sand body with an average width of that the trapping mechanism is the same as for the
4000ft ( 1220m). The shape and character suggest 35-1. The structural contour of the oil-water contact
that the sand body is fluvial but the exact type of does not close to the northwest but the sand body
fluvial system is less clear. Several distinct sedi­ pinches out in that direction. The trapping mechan­
mentary units can be recognized within the 34-2 at ism again combines stratigraphy and structure.
B-8 (Fig. 13), but these are not as discrete as those
of previous sandstones. Similarly, evidence from
34-1 sandstone
other full-hole core, dense sidewall cores in most of
the other wells, and log character analysis suggests a The base of the 34-1 sandstone lies just above the
broad systematic fining within a single-storey system limestone at the top of the 34-2 sequence and all but
(see also cross-section A- A', Fig. 13). This implies the top few feet of the sandstone was cored in the
368 R. Young, W.E. Harmony and T. Budiyento

A 1600' 1900' 2200' 5300' A'

B-2 B-3 B-4 Widu ri- 7 Widuri-5
20 120 20 120 20 120 20 120 20 120
API API API API API

Top l i m estone

Base coal 'A'

Core log 34-2 sandstone

DEPTH Widuri B-8
(FT)
GAMMA RAY
LOCATION MAP API100
Muddy l i mestone and
0
calcareous m u dstone
Bivalve and fora m i n ifera
fragments

Root leted si ltstone

and coal

Trough cross- bedded

sandstone g rading
from med i u m to coarse
at the base to
fine-very fine at the top

4490 Si MCg
,...,...
Sh�Cl
SA ND I
NO RECOVERY
4500

4510

4520 Coal 'A'

Fig. 13. 34-2 reservoir: vertical and lateral sedimentary character..

B-8 well (see Fig. 15). The sandstone is erosively laminations are common, and abundant mudstone
based and consists of mostly fine- and very fine­ clasts in the lower to middle part of the sandstone
grained, very well-sorted sandstone, which broadly result in increased gamma-ray readings (Fig. 15).
fines upwards giving a predominantly gamma-ray Towards the top of the sandstone (4382. 5-4391f t
increasing upwards log signature. Cross-bedding is MJ?) a heavily bioturbated and interbedded·­
sporadically present throughout but, towards the interlaminated unit of mudstone, siltstone and
top, parallel lamination, ripple cross-lamination and sandstone contains predominantly Thalassinoides
ripple/wavy bedding predominate. Carbonaceous but also Teichichnus type burrows, which are indica-
 Fluvial sand-body geometries 369

\ + +

....____
, )
I

' :

'\ + +
(
)

"-.!
;--s··:
(
+ +

+ + +

+ Cl 20'
5000 FT

Fig. 14. 34-2 reservoir : (A) Net

sandstone isopach with the oil­ a
water contact marked. (B) Depth E
structure map on the top 34-2
reservoir showing that the oil­ CllOO'
n 5000 FT
water contact is 50ft below '\
·.-----.... ,/ \ B
structural closure (shaded).

tive of marine influence. In Widuri-1 the 34-1 was the result of extrapolated well data. Prior to any
also cored and clay drapes, a common fei!ture of development drilling it was recognized that an
tidal sediments, are abundant in the upper part of acoustic impedance contrast between the base of the
the sandstone, where marine acritarchs also have sandstone and the underlying mudstone was respon­
been found. sible for a seismic peak (see Blue seismic horizon,
The net sandstone isopach (Fig. 16A) defines a Fig. 6, for example) which, when mapped, had a
sand body with a maximum penetrated thickness of restricted areal extent. Seismic modelling suggested
49ft ( 15 m) and an average width of 2000 ft ( 610 m). that the sandstone thickness could be estimated
The detailed shape of the contours in Fig. 16A is not from the amplitude of the reflection through the use
370 R. Young, W.E. Harmony and T. Budiyento

A A'
D-11 2150' A-12 1600' Widu ri-1 5000' C-6 3850' B-8
20 API 120 20 API 120 20 API 120 20 API 120 20 API 120

LOCATION MAP DEPTH

Core log 34-1 sandstone
GAMMA RAY (FT) Widuri B-8
API

Biotur bated mudstone,

si ltstone and sandstone
Thalassinoides/
Teichichnus type bu rrows

Fine and very fine

grained sandstone

Calcareous mu dstone
with shell fragments

Fig. 15. 34-1 reservoir: vertical and lateral sedimentary character.

of a tuning curve (Young et al., 1991), and the net Fig. 15). When traced southwards away from the
sandstone isopach map was constructed based on sediment source area, however, Young & Atkinson
reflection amplitude. This map proved to be very (1993) have shown that the entire productive interval
accurate and has changed little after development at Widuri becomes more marine and is dominated
drilling. mostly by 6-24 ft thick, stacked, coarsening upwards
The shape and dimensions of the sandstone, units, which are interpreted to be distributary mouth
together with the log and sedimentary character, bars and various other marginal marine bars. Dia­
suggest a single-storey channel which was reworked chronism of the whole interval is clear and the
by tides in the upper part. Within Widuri field, the overall interpretation is a deltaic succession domi­
interval that is laterally equivalent to the channel nated by tidal and fluvial processes (Young &
sandstone consists of mudstone and siltstone with Atkinson, 1993).
rare, thin sandstones (see cross-section A-A', The 34-1 is therefore interpreted to be a distribu-
 Fluvial sand-body geometries 371

��') \),1 ,..,

+ + +
..
ON N
' '
Co
+ /V ,...../+�u' ,{ +

'•?
r·· ) (/
'· ·-r<'-' + \'
<;:) J
....._._ ___
)
,/
'·---) + + I
i +

,.
(
,....,
!
\
I
I I

I
I 0'
+ + 20' +
I'

(
+ ol
+ + + &J + +
"''
... <.)'
s!
�
oJ
Cl 20'
+ + .t../ 5000 FT

,
-,__ _) A

Fig. 16. 34-1 reservoir. (A) Net

sandstone isopach with the field
oil-water contact marked.
(B) Depth structure map on the top
34-1 sandstone showing that the Cl100'
5000 FT
oil-water contact is 1 00ft below
8
structural closure (shaded).

tary channel and the tidal reworking at the top is those of a modern Mahakam delta distributary
perhaps mostly a result of estuarine conditions after channel (see Allen et at., 1979; Allen, 1985}.
abandonment of the distributary. The apparent Reservoir quality is still very good with average log
absence of delta-front bars may reflect the proximal porosity of 29% and test permeabilities of 5-10 D.
position of the Widuri area with respect to the delta The depth structure map on top of the 34-1 sand­
complex and/or the low preservation potential of stone (Fig. 16B) identifies the structural closing
tidal bars (oriented normal to the shoreline) as the contour at 3580ftSS (1091 m SS). An oil-water
delta prograded. The shape, dimensions and sedi­ contact for the 34-1 sandstone has not been
mentary character of the 34-1 are very similar to penetrated in any well. However, a reservoir simu-
372 R. Young, W.E. Harmony and T. Budiyento

lation study using the geological model described stratigraphical levels over a vertical interval of about:
here and an oil-water contact at 3680 ft SS (1122 m 75ft (23m) immediately above the 34-1 sandstone.
SS) gave a very good match with the available None of the reservoirs has been cored, but Fig. 17
production history. The oil-water contact is there­ demonstrates the log character and limited lateral
fore assumed to be the same as all the underlying extent of the sandstones. Nine separate sandstones
reservoirs. A discrepancy of 100ft (30.5 m) between have been identified so far (Fig. 18A), although this
the oil column and the structural closure can again number could increase or decrease with further
be accounted for by a stratigraphical component to drilling. The sandstones are fine to very fine grained
the trapping, where northwestwards closure is effected with sharp bases and sharp or gradational tops and
by pinchout of the reservoir. usually show a gamma-ray increasing upwards log
signature. Penetrated net sandstone thickness does
not exceed 15ft (4.5 m), with one exception, the
33-series sandstones
33-2, which is 26ft (8 m) in Widuri-6. Sand body
There is less information on the 33-series sandstones widths average 2000 ft (610 m) but well penetrations
than on any of the other reservoirs and they are the of individual sandstones are too few to define geo-·
least understood and most difficult to define. The metries accurately. However, at least some of the
33-series is a group of several different sandstones sandstones are thought to have shoestring geometries
which, for the most part, occupy slighty different and occasionally, as in the case of the 33-2, reflection

A A'
D- 1 3,800' D-4 12,600' C- 2 32
, 00' C-1
20 API 120 20 API 120 20 API 120 20 API 120

Top
0 0
Talang Akar
formation

50 50
FT
FT Top 33-1
sa ndst one
100 100
i e �L-
_ �!_ .!

150 Top 34-1 150

san dst one
inte rval

5000FT
t.----J

Fig. 17. Cross-section showing the electric-Jog character and lateral discontinuity of 33-series reservoirs.
 Fluvial sand-body geometries 3 73

Cl100'
5000 FT

Fig. 18. 33-series reservoirs. (A) Net sandstone isopach with the field oil-water contact marked. (B) Depth structure map
on a middle 33-series limestone marker showing that the field oil-water contact is 150ft below structural closure (shaded).
374 R. Young, W. E. Harmony and T. Budiyento

amplitude at the base of a sandstone can be used to the overlying mudstone is clearly a shallow-marine
map it. deposit with minimal coarse clastic sediment input.
The sand bodies are laterally equivalent to finer However, only 16 ftMD (4. 9 mMD) above the top
grained sediment, which was cored in the C-2 well of the core is a 3 ft (1 m) thick coal.
(Figs 1 7 & 19). The basal 32ft (9. 75m) of the core Using similar reasoning as for the 34-1 sandstone,
consists of carbonaceous mudstone with laminations the 33-series sandstones are interpreted to be single­
of siltstone and very fine-grained sandstone. These storey distributary channel deposits of a tide­
laminations are millimetre-scale and sparse at the influenced delta. Reservoir quality is variable but
base but centimetre-scale and frequent towards the log porosities average 27% and test permeabilities
top. Both bioturbation and ripple cross-lamination are of the order of 1-5 D.
increase upwards and fragments of fish bones, The depth structure map on a marker in the
foraminifera and echinoids are scattered through­ middle of the 33-series interval identifies a structural
out. The basal 27ft (8. 25 m) is a coarsening upwards closing contour at 3530ft SS (1076 m SS, Fig. 18B).
unit which grades into a 5 ft ( 1.5 m) thick, fining The oil-water contact for the 33-series sandstones is
upwards unit. The upward fining trend continues unknown and reservoir simulation studies do not
into highly bioturbated siltstone before a very sharp help because production histories are too short. The
contact at 3899 ftMD marks the top of the unit. The oil-water contact is assumed to be at 3680 ft SS
overlying dark brown, carbonaceous mudstone (1122m SS), the same as the other reservoirs,
contains scattered marine fauna. The coarsening­ although multiple oil-water contacts are possible.
fining couplet is thought to reflect prodelta to lower However, irrespective of the position of the oil­
mouth-bar progradation and abandonment, whereas water contact, the oil column is again greater

DEPTH
GAMMA RAY (FT)
0 API 100 3880 Widuri C-2

Fig. 19. Core log of part of the 33-

series interval in Widuri C-2 (see
Fig. 17 for location of cored
interval).
 Fluvial sand-body geometries 375

Feet
NW
0 1000 2000 3000 SE

Time Feet
1000

3100
1020

3200
1040

1060 3300

1080
3400

1100

3500
1120

3600
1140

1160 3700

1180 3800

1200
3900
1220

1240 4000

1260

4100
1280

1300
4200

1 320

4300
1340

1360
4400

Fig. 20. An uninterpreted seismic line from the three-dimensional volume over the Widuri area. The line is located just
south of Widuri field and shows an oblique view through one of the features that are interpreted to be incised valleys
associated with the top Talang Akar horizon. See Fig. 6 for the stratigraphical location of the seismic markers.

than the structural closure and a stratigraphical mudstone and siltstone with some thin, interbedded
component to the trapping, although less well limestone. A low-energy, marginal marine setting is
defined, is again necessary. envisaged prior to the development of fully marine
The interval between the top 33-1 sandstone and conditions in the overlying Batu Raja Formation.
the top Talang Akar Formation consists of coal,
376 R. Young, W.E. Harmony and T. Budiyento

Average
® * Average
Reservoir grain
Multi storey/ Reasonable @Average ®I@ porosity a n d Aerial distribution Environment Litho-
m u ltilateral maximum width (ft) ratio permeability (to scale) of deposition stratigraphy
size
thickness (ft) range

33
Series
fine
to
very fine
NO/NO 715'-730' 72000'
7 1 :67
to
7 1 : 1 33
27%
1-5
darcies
� <>: 1- w
>-
CC UJ
<LJ
f-' ::::l Z

g f-'� I� cc
w

�
m
fine 29% �u
0 ::;;
34-1 to NO/NO 40' 2000' 1:50 5-10 w
very fine darcies ::;;
j:!
(3 z

([:
0

29% �
34-2 medium ?NO/NO 50' 4000' 1:80 10-30 UJ ::;;
cc cc
darcies ::::> 0
u..
� cc
::;; - ..:
""

1flj
..:
29% ::;; <.:J
w
35- 1 medium YES/YES 60'** 6500'* * 1: 108 10-20 1- z
darcies Ul ..:
...J
>-
Ul
...J cc j:!
..: UJ
m

0
> ::;;
::::> UJ
medium 29% ...J
u.. ::;;
35-2 to YES/YES 65' 1 1000' 1: 169 20-30
coarse darcies UJ
..:
0
cc ...J
::::> UJ
N
�
::;;

"l
28% � :::;
36 -1 coarse YES/YES 175' Sheet Sheet 20-30
darcies

\� __ ;

Fig. 21. Summary and comparison of sedimentary character, geometry and environment of deposition of the Widuri field
reservoirs.
* Using 'reasonable' maximum thickness is an attempt to eliminate thickness anomalies; the parameter is not rigorously
defined but nor is it arbitrary.
* * Only the east channel system was considered ; the west channel system may be structurally controlled.

Wright, 1975, their Fig. 16). The distributary

Discussion
channels of tidally influenced deltas are therefore
The upper part of the productive interval at Widuri not always juxtaposed against delta-front bars and,
(the 34-2, 34-1 and 33-series sandstones) is inter­ once abandoned, they become estuaries.
preted as deltaic based largely on evidence outside These same characteristics (namely, channel
of the Widuri area. Core data from approximately features that erode into marine sediments and are
the same stratigraphical interval in the Sunda basin estuarine in their upper parts), however, could be
and the Ardjuna basin (approximately 70 miles to used to interpret the reservoirs (especially the 34-2
the southeast of Widuri in the adjacent PSC) indicate and 34-1) as incised valley fills, which formed as a
a predominance of tidal and fluvial processes at the response to periods of falling relative sea-level (see
shoreline (Young & Atkinson, 1993). In this setting, e.g. Van Wagoner et al., 1990). It is also reasonable
tidal reworking at the delta front fashioned delta­ to expect that relative sea-level changes caused by
front bars, oriented predominantly normal to eustasy and/or tectonic movement had, at times,
the shoreline, which were thin (up to 25ft thick) a major effect on deposition of the Talang Akar
compared with the channel deposits (approximately Formation over the 7.5Myr period. An example
15-40ft thick). The delta-front bars therefore had of this is at or near the top of the Talang Akar
low preservation potential as the delta prograded Formation, where a horizon characterized by down­
(see e.g. Allen et al., 1979; Allen, 1985; Coleman & cutting is clearly seen on seismic records (Fig. 20).
 Fluvial sand-body geometries 377

Structural
Aerial Oil
owe closing F i l l w.r.t. Trap
Reservoir distribution col u m n
(ft/SS ) contour structure type
('to scale') (ft)
(ft SS)

33
series � 3680 ? 3530 325

Q)
-o �
34-1 � 3680 3580 2 70
c Vl
0 0
�u
- ·�
- .c
� 0.
::0 "'
.o c;;
-o �
u rn

q;�
:o · -
-
<ll 3 �
= u - "'
34-2 3680 3630 240 ·- ::0 Ul !::
"'
LJ._ !::
"'

35-1
rf!!
��'
3680 3660 180

35-2 5'? 3680 3680 110

F i l l ed
to
spill
-

3
�

�1
u

/f Part i a l ly
2
cil
36- 1 3680 3740 55
Fig. 22. Summary of Wid uri field fi l l ed
trapping characteristics.

When mapped, this down-cutting feature has a R E S E RV O I R SUM M A RY

channel form with a roughly N�S orientation, a
width of approximately 1000-3000ft, and a thick­ The more important aspects of the geometry and
ness or 'down-cut' of about 30-lOO ft. An interpret­ character of the Widuri reservoirs are documented
ation of recent three-dimensional seismic coverage in Fig. 21. The productive interval as a whole shows
over the northwest flank of the Asri basin has a gradual and systematic change from base to top.
identified several more of these features, all associ­ Grain size gradually decreases upwards from coarse
ated with the same horizon. Unfortunately none of grained in the 36-1 to fine grained - very fine grained
these features has yet been cored, but the event can in the 33-series reservoirs, and multistorey and multi­
be traced seismically all over the Asri basin and lateral channel systems in the basal three reservoirs
60miles south into the western Sunda basin, where give way to single-channel systems in the upper
it correlates to a karst surface within limestone. The three. The maximum sand body thickness gradually
event therefore has regional extent and is interpreted decreases upwards, as does the average width, and
to be the result of a lowstand of sea-level. There is the thickness/width ratio shows a corresponding
strong seismic stratigraphical evidence that the low­ change. Porosities are similar throughout but test
stand is a response to tectonic reactivation of the permeabilities show a significant decrease upwards.
Asri basin. The change in sand body size, shape and character
In contrast, within the productive interval at is the result of a maturing depositional system
Widuri field there is no such seismic expression and associated with a rise in sea-level between 25. 5 and
no such evidence for regional lowstand surfaces of 22Ma (see Fig. 5; Haq et a/., 1988). The immature
erosion (see Fig. 20). Until further evidence surfaces, fluvial system at the base (36-1 reservoir) is suc­
therefore, the retrogressive stacking of the reservoirs ceeded by progressively more mature fluvial systems
at Widuri field is interpreted to be a response to before deltaic conditions prevail at the top. The
essentially autocyclic switching of sediment supply change in character of the sand bodies therefore
within an overall transgressive systems tract. How­ reflects the change in environment of deposition and
ever, the primary aim of this paper is to demonstrate this is supported by the available palaeontological
trapping configurations resulting from the interaction data from the intervening shales. Rare occurrences
of sand body geometry and structure, and these are of palynomorph taxa commonly associated with a
independent of the mechanism of stacking. brackish water setting have been found as low down
378 R. Young, W.E. Harmony and T. Budiyento

NW SE

SE

Fig. 23. (A) Schematic diagram of

the Widuri field trapping
configurations. (B) A projection of
what the Widuri trapping
configuration would have been if the
34-2 reservoir had been a sheet
sandstone.

as the 35-2 interval and increased frequencies in the is the spill point of the 35-2 reservoir that therefore
overlying 35-1 and 34-2 intervals indicate increasing defines the oil- water contact for the entire field.
proximity to the sea. The lower two reservoirs are structural traps, but
the upper four reservoirs are filled beyond structural
closure and rely on a favourable orientation
TRAP SUM M A R Y (northwestwards pinch-out) with respect to structure
in order to trap oil by a combination of structure and
The important aspects of the trap are summarized in stratigraphy.
Fig. 22. With the possible exception of the 33-series, The trap model is drawn schematically in Fig. 23A
the reservoirs share a common oil-water contact but it is interesting to evaluate how the trapping
and therefore must either be in communication now would change if one of the upper four reservoirs, the
or have been in communication some time after oil 34-2 for instance, had been either a sheet sandstone
migration. The structural closing contour on top of or did not have a favourable orientation with respect
the 36-1 sandstone is below the oil-water contact to structure. In other words, what would have
and the reservoir is only partially filled. The d osing happened if the 34-2 did not have a stratigraphical
contour on top of the 35-2 sandstone is coincident component to trapping? This case is shown in
with the oil-water contact and is therefore full to Fig. 23B. The reservoirs are still in communication
spill. This is an extremely important point because it and so the oil would migrate upwards through the
 Fluvial sand-body geometries 379

lower reservoirs into the 34-2, where it would C.L. Mora and C.D. Atkinson for a critical review
be trapped with respect to a new field oil-water of the manuscript, and to J.F. Mitchell for his
contact as defined by the 34-2 structural spill point technical input in the early stages of the study. We
(3630 ft SS or 110 7 m SS, see Fig. 22). The net effect also acknowledge the help and support of all the
is exaggerated in the schematic diagram of Fig. 23B, other members of the multi-disciplinary team that
but a 50 ft (15 m) higher oil-water contact would was responsible for developing Widuri field.
significantly reduce the volume of oil in the field.
An overall conclusion then is that the trapping at
Widuri is intimately related to reservoir geometry REFERENCES
and that the large size of the field is the result of a
remarkable coincidence of events. A relative rise ALLE N ,G.P. (1985) Delta sedimentation: modern and
of sea-level helped cause a gradual change in the ancient examples in the Mahakam delta basin. Indonesia
Petroleum Association, Field Trip Guide.
environment of deposition which, in turn, produced
ALLEN , G.P. , LAURIER, D. & THOUVENIN , J. (1979) Etude
a systematic reduction in the size of sand bodies, the sedimentologique du delta de Ia Mahakam. Compagnie
upper four of which each had to be oriented favour­ Francaise Petrol. Notes Mem. , 15 , 156 pp.
ably with respect to structure in order to trap oil A LLEN , J.R.L. (1965) A review of the origin and character­
stratigraphically. istics of recent alluvial sediments. Sedimentology , 5,
89-192.
BEDDOES, L.R. (1981) Hydrocarbon plays in Tertiary basins
of Southeast Asia. Energy, 6, 1141-1163 .
APPLICATIONS BEN-A VRAHAM , Z. & E MERY, K.O. (1973) Structural
framework of Sunda Shelf. Bull. Am. Ass. petrol. Geol. ,
57, 2323-2366.
The geological model has contributed to the devel­
B USHNELL, D . C. & TEMANSJA, A.D. (1986) A model for
opment of Widuri field in two important ways: hydrocarbon accumulation in Sunda Basin, West Java
(i) improved drilling success and (ii) improved Sea. Proceedings of the Indonesian Petroleum Associ­
reservoir management. Early recognition of sand ation, 15th Annual Convention, Vol. 1 , pp. 47-75.

body geometries resulted in more efficient develop­ COLEMAN, J. & WRIG HT , L.D. (1975) Modern river deltas:
variability of processes and sandbodies. In: Deltas (Ed.
ment drilling, such that of the 47 development, infill Broussard, M.L.), pp. 99-149. Houston Geological
and step-out wells drilled so far only two lacked Society, Houston, TX.
economic pay, and both of these were side-tracked COLLINSON , J.D. (1978) Vertical sequence and sand body
to successful locations. Similarly, reservoir manage­ shape in alluvial sequences. In: Fluvial Sedimentology
ment has been optimized by having a detailed and (Ed. Miall , A.D.), Mem. Can. Soc. petrol. Geol. ,
Calgary, 5, 577-588.
predictive geological model, which forms the basis F IELDING , C.R. & CRA N E , R.C. (1987) An application of
of reservoir simulation studies. This application will statistical modelling to the prediction of hydrocarbon
become increasingly important as the field matures, recovery factors in fluvial reservoir sequences. In: Recent
water production increases, and further infill drilling Developments in Fluvial Sedimentology (Eds Ethridge ,
F . G. , Flores, R.M. & Harvey, M . D.), Spec. Publ. Soc.
is required to optimize hydrocarbon recovery. econ. Paleontol. Mineral. , Tulsa, 39, 321-327.
From an exploration viewpoint, the work on F oLK, R.L. (1968) Petrology of Sedimentary Rocks.
Widuri field has provided an analogue model in the Hemphill's, Austin, TX, 170 pp .
search for new fields in the PSC area, in which HAQ, B.U . , H ARDENBOL, J. & VAIL, P.R. (1988) Mesozoic
stratigraphy is an important means of trapping and Cenozoic chronostratigraphy and eustatic cycles. In:
Sea-level Changes: an Integrated Approach (Eds Wilgus ,
hydrocarbons. Widuri is by definition a large oil field C. K. , Hastings, B.S. , Kendall, C.G.StC. , Posamentier,
but approximately 75% of its reserves have a strati­ H . W. , Ross, C.A. & Van Wagoner , J.C.) , Spec. Pubis
graphical component to trapping. Any future dis­ Soc. econ. Paleontol. Mineral. , Tulsa, 42, 71-108.
covery of stratigraphically trapped oil could therefore H uTCHISO N , C.S. (1986) Tertiary basins of S.E. Asia ­
their disparate tectonic origins and eustatic strati­
potentially be large. graphical similarities. Bull. geol. Soc. Malay. , 19,
109-122.
LEEDER, M.R. (1973) Fluviatile fining-upward cycles and
AC KNOWLEDGEMENTS the magnitude of palaeo channels. Geol. Mag. , 110,
265-276.
LORENZ, J.C. , H EINZE, D. M. , C LARK, J .A. & S EARLS,
The authors wish to thank the management of C .A. (1985) Determination of widths of meander-belt
Pertamina, Maxus and partners for permission to sandstone reservoirs from vertical downhole data,
publish this paper. Special thanks are extended to Mesaverde Group , Piceance Creek Basin, Colorado.
380 R. Young, W.E. Harmony and T. Budiyento

69, 71 0-721.
Bull. Am. Ass. petrol. Geol. , YOUNG, R. & ATKINSO N , C.D. (1993) A review of Talang
VAN WAGONER, J. , R.M . , CAMPION , K.M. &
MITCHU M , Akar Formation (Oligo-Miocene ) reservoirs in the off­
RAHMANIAN , V . D . (1990) Siliciclastic sequence stra­ shore areas of Southeast Sumatra and Northwest Java.
tigraphy in well-logs, cores and outcrops: concepts In: Clastic Rocks and Reservoirs of Indonesia (Eds
for high resolution ·correlation of time and facies. Atkinson , C.D. , Scott, J. & Young, R.). Indonesian
Am. Assoc. petrol. Geol. Meth. Explor. Ser. , 7, 55 pp. Petroleum Association, Core Workshop Notes, October
WIGHT, A. , SUDARMONO, & lmron, A. (1986) Stratigraphic 11, pp. 177-21 0.
response to structural evolution in a tensional back-arc YOUNG, R. , HARMONY, W . E . , GUNAWAN, J. & BUDIYENTO,
setting and its exploratory significance: Sunda basin, T. (1991) Widuri field, offshore Southeast Sumatra :
west Java Sea. Proceedings of the Indonesia Petroleum sandbody geometries and the reservoir model. Proceed­
Association, 15th Annual Convention, Vol. 1 , pp. 77- ings of the Indonesian Petroleum Association, 20th
100. Annual Convention, Vol. 1 , 385-417.
 Index

Abyssal plain,iron 85-94 Cherty iron formation 139, 140

formations 149-51, 154 Bideford Group 101-3, 110 Chesil Beach 91, 93
Acidic volcanic rocks,geochemistry Black shale facies,Bude Formation Clastic rocks, Upper Jurassic
252 126-7, 130 232-3
Ahwaz, Cenozoic clastic reservoirs description 110-11 Clasts 80-5
349-50 interpretation 120-2 Coastal zone terminology 76-7
AI Khlata Formation 340,341 Black shales, geochemistry 248 Computer maps,preparation 73
Allostratigraphy 72-3 Boss Point Formation 117 Conglomerate debris, Kalvag
Alluvial fan facies association, Bottom-current activity,ridge- Melange
Tryfan Fach Member trough system 296-7 geochemical analyses 245
172-4,184-5 Bovis Formation,Pororari Group geochemistry 241-8
petrography 178 204,208-10 Conglomerates
Alluvial fans 203 palaeocurrent pattern 212-13 grain sizes 80-4
Amin Formation 340 Braided stream association,Tryfan Guaritas sequence 5-7,9-10
Anglesey 178 Fach Member 166-9, Ricinus 62
Animikie Group, iron formations 174-5 Senora de Brezales Formation
139-40, 142 bounding surface with marine 223-7
Arabian Plate 335-9 shelf association 177 shoreface 75
Arenites petrography 178-9 Crackington Formation 104-5,
lithic, geochemistry 247-8 Breccia 198-203 107-10
lithic-volcanic,geochemistry Bude Formation 101-5,130-1 Craven Basin 308-9,312-18
252-3 bedtypes 108-9 Cryptobioturbation 35
Askrigg Block 307-8,312-20 facies analysis 110-26 Culm Basin 102-4
Asri basin 355-8, 377 description 110-20 tectonic setting 105-7
interpretation 120-6 tectonically juxtaposed
facies succession 106-10 successions 107-10
Baby Formation,iron formations description 126
143-6 interpretation 126-30
Back-arc basin 257 Burrowing, Woburn Sands 27-8, Debris flows 186
Bahrain, Mesozoic clastic reservoirs 29,35-6 breccia 203
344-5 Diorite 247
Bars, gravelly shorefaces 85-6 Dune migration, Guaritas sequence
Basalts, geochemistry Camaqua Basin, Brazil 3-4,6 10-11
alkaline 248 Cameros Basin 221
island arc 252 stratigraphy 220, 422
subalkaline MORB-Iike 248- Canadian Shield, iron formations Erosion
51 139 channels 268-70
transitional 251-2 Capel Curig Volcanic Formation ridge-trough system 297
Basra, clastic reservoirs 159, 183 storm/wave 271-4
Cenozoic 349-50 Carbonate rocks,Senora de surfaces 26-7
Mesozoic 344-5 Brezales Formation 227-8 Estuaries 69-70
Beachfaces Cardium Alloformation,Alberta Estuarine environments, essential
bedforms and sedimentary 48-50,73 elements 41-2
structures 92-4 allostratigraphy 49-50
characteristics 95 comparison of allomembers 65
facies successions 94-6 stratigraphy and Facies modelling 187-9
gravelly 76 sedimentology 50-5 Fan-deltas 213, 216
sediment transport 78-9 Cardium Formation 73, 76 Feni Drift 297, 299
Beardmore Volcanic Associated gnivelly shorefaces 80-4, 86- Floodbasin facies association,
Iron Formation 151-2 94 Tryfan Fach Member 169-
Beardmore-Geraldton Clastic Carstone 38 72, 174-7
Associated Iron Cenozoic clastic reservoirs 349- petrography 178-9
Formation 146-9 50 Frio Formation 267, 271-5
Bedfordshire Strait 25 Channel erosion 268-70
Bedforms,gravelly shorefaces Chert,geochemistry 247

Sedimentary Facies Analysis: A Tribute to the Research and Teaching of Harold G. Reading Edited by A. Guy Plint
© 1995 The International Association of Sedimentologists ISBN: 978-0-865-42898-0
381
382 Index

Gabbro 247 Ignimbrite, geochemistry 252 Lower Ordovician ophiolite

Gardar Drift 297 Incised valley complexes/island arc
Gas, Middle East 333 Ricinus 47-9, 66-71 sequences 259-60
Gault Clay 22-3 Interbedded sandstone-shale facies,
Gharif Formation 340 Bude Formation 126-7,
Goniatite-bearing marine bands 130 Magma migration 185
309-11 description 112-17 Mahwis Formation 340
Goniatites 111 interpretation 122-5 Marib-Jauf basin 346
Graphitic shear zones 149 Iron formations 137-9 Marine bands, goniatite-bearing
Gravel deposition, Guaritas case studies 139-53 309-11
sequence 8-10 discussion 153-4 Marine shelf association, Tryfan
Gravel plunge step 90-1 objectives 139 Fach Member 161-6, 174
Gravel wave ripples 85-6 types 138 bounding surface with braided
Gravelly shorefaces 75-6 Island arc basalts, geochemistry stream association 177
Gravels 252 petrography 179-80
clast shape 84-5 Israel, Mesozoic clastic reservoirs Mesozoic clastic reservoirs 344-9
grain size and shape 80-5 348 Metabasalts, Kalvag Melange
textural trends 80-4 241, 246-7
wave-worked 90 Middle East
Groundmass, olistostromal, Kalvag Jordan, Palaeozoic clastic reservoirs Clastic reservoirs
Melange 240-1 343 Cenozoic 349-50
geochemistry 253, 256 Mesozoic 344-9
Growth faulting 265-6 Palaeozoic 339-44
effects on sedimentation 267- Kalvag Melange 240-1 geological boundaries 331-2
76 depositional model 257 oil and gas reservoirs 333-9
Guaritas depositional sequence, geochemistry 241-56 palaeogeographical and
Brazil 3-5, 6 geological setting 238-40 palaeoclimatic synopsis
alluvial facies profile through 242-3 335-9
general features 5-11 provenance and tectonic setting source rocks 336
interpretation of features 256-9 Millstone Grit
8-11 regional correlations and cycle duration 306
interpretation of lateral interpretations 259-61 cyclothem
changes 11-14 Karst surfaces, Senora de Brezales architecture and sea-level
lateral changes 11-14 Formation 228-9 change 311-21
pattern of textural dispersion Kuwait, clastic reservoirs asymmetry 321-2
11 Cenozoic 349-50 E2a(i) cyclothem
palaeocurrent pattern 11 Mesozoic 344-5 architecture and sequence
Gulf of Suez, clastic reservoirs Kyeburn Formation 198 stratigraphy 313-17
Cenozoic 350 boundaries and thickness 312
Mesozoic 348-9 depositional model and
Palaeozoic 343-4 Labrador Trough 143 tectonism 313
Gulf Region, Palaeozoic clastic Laminar calcretes, Senora de E2c2 cyclothem
reservoirs 342-3 Brezales Formation 228 architecture and sequence
Gunflint Formation, iron formation Laminated beds, Bude Formation stratigraphy 318-20
139 112, 123 boundaries and thickness 317
Gwern Gof Tuff 159-60 Late Jurassic unconformities 233- depositional model and
4 tectonism 317-18
Levant-Sinai region, Mesozoic major cyclothems 322-3
Hadhramaut-Jeza rift basin clastic reservoirs 348-9 architecture and genesis
346-7 Lithic arenites, geochemistry 323-4
Haima Group, clastic reservoirs 247-8 major scale cyclicity 322-4
340 Lithic-volcanic arenites, multi-cyclothem condensed
Haradh Formation 340 geochemistry 252-6 sections 322-4
Haushi Group, clastic reservoirs Llewelyn Volcanic Group 158- sequence model 320-2
340-1 60, 183 choice 320-1
Hawks Crag Breccia 199-201, volcanic source 180 Morley Occurrence, iron formations
214-15 London- Brabant land mass 19, 20 152-3
deposition 211 Longshore currents, sediment Muddy siltstone facies, Bude
Hummocky cross-stratification, transport 77-8 Formation 126-7, 129-30
Bude Formation 129 Lower Follifoot Grit 317-19 description 111-13
Hydrocarbon trapping 378-9 Lower Greensand Group, interpretation 122
deposition 19-20 Mudstones
Lower Miocene, Louisiana 266- Guaritas sequence 8
7, 274-6, 278 Ricinus 60-2, 65, 66
 Index 383

deposition 70 pebbly sandstone and massive Rutbah Formation 347

distribution 63-5 or graded sandstone 203-
Tryfan Fach Member 161-6 10
Mudwaves, ridge-trough system geological setting 199 Salinity, and marine bands
296-7 palaeocurrent pattern 212-13 309-11
tectonic implications 213-15 San Leonardo Formation 228-9
Pororari River 211 Sand volcano 120-1
Namurian Basin, Central Province, Preservation potential thresholds Sand-body geometries, Widuri field
30-8 276 377
Namurian stratigraphical representation in growth-fault Sandstone-dominated facies, Bude
framework 306 settings 279 Formation 129
Neutral Zone, Mesozoic clastic Pro-delta current system, ridge­ description 117-20, 122-3
reservoirs 344-5 trough system 297-302 interpretation 124-6
New Zealand 197-9 Puysegur Formation 198 Sandstones
North Sea, ridge-trough system Pyroclastic material 158-9 Bude Formation 112-17
285-6 syneruptive, aggradation 184-5 graded silty 112-13, 123-5
database and methodology 288 Guaritas sequence 7-8, 10
depositional models 296-302 irregularly laminated 117,
external morphology 288-9 Qatar, Mesozoic clastic reservoirs 125-6
geological setting 286-8 348 red, Senora de Brezales
internal reflection configurations Quasi-planar lamination 164 Formation 220-3
289-91 Qusayba shales 342 Ricinus 57-60, 69-70
interpretation 291-2 parallel-laminated 59-60
mapping 292-4 structureless 57-9
palaeogeographical setting of Ramp-fan system, iron formation Tryfan Fach Member
structures 294-6 147-9, 154 conglomeratic alluvial fan
Rhyolite, geochemistry 252 172-4
Rhyolitic braided-stream rhyolitic braided-stream
Oil, Middle East 333 sandstones, Tryfan Fach 166-9
Olistoliths, Kalvag Melange Member 166-9 structure 179
240-1 Ricinus, Alberta 47-9, 70-1 thinly bedded floodbasin
sedimentary rocks Allomember 54 169-72
geochemistry 252-3, 255 comparison with other Widuri field 360-75
representative geochemical allomembers 65 Saudi Arabia, Palaeozoic clastic
analyses 254 cross-sections 71-2 reservoirs 341-2
spatial distribution 244 deposition in estuary 69-70 Scandinavian Caledonides 259
volcanic rock facies 57-62 Sea-level
geochemistry 248-52 lateral and vertical distributions change and Millstone Grit
representative geochemical 62-5 cyclothem architecture
analyses 249 geometry 55-7 311-21
Oman, clastic reservoirs incised valley 56 goniatite-bearing marine bands
Mesozoic 347-8 sand body 56-7 309-11
Palaeozoic 340-1 incised shoreface 65-7 Seaweed, sediment transport 78
Outer shelf and slope, iron interpretation 65-70 Sediment transport processes
formation 143-6, 154 origin of valley 66-8 76-9
Outwash fans 169 palaeogeography 68 Sedimentary rocks, olistoliths,
structural problems 49 geochemistry 252-5
turbidite interpretation 49 Sedimentary structures, gravelly
Palaeocurrents 212-13 valley orientation 70 shorefaces 85-94
Palaeozoic clastic reservoirs Ridge-trough system, North Sea Sedimentation and growth faulting
339-44 285-6 265-6, 267-76
Palestine, Mesozoic clastic database and methodology 288 Senora de Brezales Formation 234
reservoirs 348 depositional models 296-302 clast derivation 230-1
Pennine Basin 308-9 external morphology 288-9 depositional history 230-1
Petroleum producing formations, geological setting 286-8 depositional model 231-2
Middle East 334-5 internal reflection configurations environmental synthesis 231
Planar lamination 164 289-91 erosion 230-1
Porcupine Basin 232-3 interpretation 291-2 facies development, tectonic
Pororari Group 199 mapping 292-4 controls 229-30
depositional systems 211-12 palaeogeographical setting of late Jurassic unconformities,
facies assemblages 199-210 structures 294-6 geometry 228-9
laminated mudstone 210 Rip currents, sediment transport lithofacies 220-7
matrix- and clast-supported 78 stratigraphy 220
breccia 199-203 River, trunk braided 12-14 Sequence stratigraphy 72, 266
384 Index

Shallow shelf, iron formation 139, Tryfan Fach Member 160, 186-9 deposition 211-12
153-4 depositional system 183-4 palaeocurrent pattern 212-13
Shallow-marine sands 18 epiclastic interpretation 180-2 Wave erosion 271-4
Shenley Limestone 38 evidence for volcanism 178-80 Wave motion and sediment
Shorefaces extent of deposit 177-8 transport 77
bedforms and sedimentary facies Widur field 355-6
structures 85-92 associations 161-74 background geology 356-8
characteristics 95 organization and distribution palaeogeography 362
facies successions 94-6 174-8 reservoir
gravelly 75-6 relations 174-7 geometry and trap 360-77
sediment transport processes position of volcanic source 180 overview 358-60
77-8 Tryfan Formation 159-61, 178 summary 377-8
Silica, iron formations 138-9 Turbidite system, Bude Formation stratigraphy 357, 359
Skaggerak Delta 287, 297-300 129-31 trapping characteristics 377,
Slaty iron formation 139, 140 Turkey, Palaeozoic clastic 378-9
Slumped beds, Bude Formation reservoirs 343 Woburn Sands
116-17, 120-1, 125, 131 depositional models 20
Slurried beds, Bude Formation facies 22-3
114-16, 118-19' 125 Unayza Formation 342 Heterolithic Sands 22-3
Snowdon graben 159 Unconformity 72 deposition 38-40
Solund-Stavfjord Ophiolite Underflows 164 description 28-9
Complex 260-1 Upper Howgate Edge Grit 313- interpretation 29-30
South Saskatchewan River 168 16 lithostratigraphical subdivision
Staffordshire Basin 309 Upper Jurassic clastic rocks 232-3 and relationships 20-4
Stainmore Basin 307-8 Upper Wilcox Group 266-70 Orange Sands 22-3
Storm erosion 271-4 deposition 38-40
Submarine extrusive edifice, iron description 26-8
formation 152-3 Vertical-sequence analysis 157-8 interpretation 28
Submarine lava plain, iron Volcanic rock olistoliths, palaeogeographic setting and
formation 151-2 geochemistry 248-52 depositional environments
Submarine rise, iron formation Volcanic terrains 158 19-20
146-9 bounding surface between syn­ palaeogeographical evolution
Subsea fans 104-5, 106-7, 126, and post-eruptive deposits 38-40
129, 130-1 187 Red Sands 22-3
Subsidence 269-70 depositional systems 183-4 deposition 38, 40
rates 276, 279 epiclastic interpretation 180-2 description 35-6
Syria, clastic reservoirs facies associations 161-74 interpretation 36-7
Mesozoic 347 modelling 187-9 sedimentological characteristics
Palaeozoic 343 organization and distribution 24-38
174-8 Silty Beds 22-3
geological framework 159-60 deposition 40
Talang Akar Formation 357, 359, inter-eruptive sedimentation description 34
376 187 interpretation 34-5
palaeogeography 361 sedimentation 185-6 Silver Sands 22-3
Tan Hill Grit 316-17 terminology 158-9 deposition 40
Terrace Bay Clastic Associated Iron Volcanic-associated iron formation description 30-3
Formation 149-51 150-3, 154 interpretation 33-4
Trace Fossils, gravelly shorefaces Volcanism, stratigraphical record stratigraphic framework 18-19
91-2 185-7 Transition Series 22-3
Transgressiveestuarine-embayment Volcano-tectonic uplift 183 description 37-8
system, depositional model interpretation 38
40-2
Tributary alluvial fan 12-14 Walther's Law 157, 187-9
Troughs, nearshore 86-7 Watson Formation, Pororari River Yemen, Mesozoic clastic reservoirs
Trunk braided river 12-14 203-10 346-7