COASTAL EVOLUTION

Late Quaternary shoreline morphodynamics

Edited by
R. W. G. CARTER
University of Ulster at Coleraine

C. D. WOODROFFE
University of Wollongong,New South Wales

a contribution to IGCP Project 274: Coastal Evolution in the Quaternary

CAMBRIDGE
UNIVERSITY PRESS
PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE
The Pitt Building, Trumpington Street, Cambridge CB2 1RP, United Kingdom

CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, United Kingdom
40 West 20th Street, New York, NY 10011-4211, USA
10 Stamford Road, Oakleigh, Melbourne 3166, Australia

0Cambridge University Press 1994

This book is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 1994

First paperback edition 1997

Printed in the United Kingdom at the University Press, Cambridge

Typeset in Times 10/13pt

A catalogue record for this book is availablefrom the British Library

Library of Congress Cataloguing in Publication data

Coastal evolution: Late Quaternary shoreline morphodynamics/

edition by R. W. G. Carter, C. D. Woodroffe.
p. cm.
“A contribution to IGCP Project 274: Coastal Evolution in the Quaternary.”
Includes bibliographical references.
ISBN 0 521 41976 X
1. Paleogeography-Quaternary. 2. Coast changes. I. Carter,
Bill. (R. W. G.) 11. Woodroffe, C. D. 111. IGCP Project 274.
QE501.4.P3C63 1994
551.4’57-dc20 94-7763 CIP

ISBN 0 521 59890 7 paperback

ISBN 0 521 41976 X hardback
 Contents

List of contributors page ix

Dedication J.D. Oford xi
Foreword 0. van de Plassche xvii
1 Coastal evolution: an introduction R. W.G. Carter and
C.D. Woodroffe 1
2 Morphodynamics of coastal evolution P.J. Cowell and
B.G. Thorn 33
3 Deltaic coasts J.R. Suter 87
4 Wave-dominated coasts P.S.Roy, P.J. Cowell,
M.A. Ferland and B.G. Thorn 121
5 Macrotidal estuaries J. Chappell and C.D. Woodroffe 187
6 Lagoons and microtidal coasts J.A.G. Cooper 219
7 Coral atolls R.F. McLean and C.D. Woodroffe 267
8 Continental shelf reef systems D. Hopley 303
9 Arctic coastal plain shorelines P.R. Hill, P. W. Barnes,
A. Hkquette and M-H. Ruz 34 1
10 Paraglacial coasts D.L. Forbes and J.P.M. Syvitski 373
11 Coastal cliffs and platforms G.B. Griggs and A.S. Trenhaile 425
12 Tectonic shorelines P.A. Pirazzoli 45 1
13 Developed coasts K. F. Nordstrom 477
Index 51 1

vii
 Coastal evolution: an introduction
R.W.G. CARTER AND C.D. WOODROFFE

‘ifthe environment is the theatre, then evolution is the play’

G. Evelyn Hutchinson

Studies of coastal evolution examine and explore the reasons why the position
and nature of the shoreline alter from time to time. Although this type of
approach has been practised for generations - by geomorphologists, geologists
and engineers -events over the last two decades have brought a new immediacy
to the subject. Generally there has been realisation that many of the world’s
coastlines are under ‘threat’ (see, for example the recent US Geological Survey
Publication Coasts in crisis (Williams, Dodd & Gohn, 1990) or the
Intergovernmental Panel on Climate Change (IPCC) report Global climate
change and the rising challenge of the sea (1992)) and that environmental
change is the consequence of human occupation of shorelines, to which
adjustment (of some kind) is inevitable. Specifically, the spectre of rising sea
levels has induced a strong political response as well as raising inevitable
questions among scientists as to the state of our knowledge and understanding
(processes that are not always convergent). It would not be hard to conclude
that our knowledge is woefully thin. Despite many excellent studies from a
wide range of environments, we are still well short of understanding how a
coastline will respond or react to secular variations in forcing functions such as
sea-level rise, storm intensity and magnitude variations or shifts in the sea state
pattern. The commonest conclusion is to predict flooding or coastal erosion,
yet such processes are clearly only part of much broader responses, which need
to be viewed over a range of scales, in both space and time. It is important to
remember that erosion simply releases sediment for deposition elsewhere, and
that sediment budgets need to be considered in order to explain coastal evolution.

Current practices and paradigms

Most coastal scientists concerned with the processes, products and patterns of
shoreline evolution are aware of the need to investigate the fundamental issues

1
2 R.W.G. Carter & C.D. Woodrofe
which control sediment transport. This can be done directly, by the association
of fluid flows and fluid flow structures to the movement of individual particles
or mass aggregates, or indirectly through the interpretation of sediment
textures, structures and bedforms. A few studies combine both approaches, but
regardless of which approach is used there are many drawbacks, particularly in
developing robust, generic models of coastal behaviour that might be useful in
predicting future evolution - certainly of the type requested by planners and
shoreline managers. Part of the problem relates to the current database; with
very few exceptions data are sparse, particularly on long-term change.
Commonly, one is forced to use information of dubious or unreliable quality,
in which error is often acknowledged but seldom analysed. All too often a
coastal-change study will use a few historical maps, perhaps one or two aerial
photographs and some ground surveys, developing a discursive discussion
from this inherently unstable mix of information. If coastal change is obvious
and explainable then this approach can be relatively powerful and indicate
some underlying symptomatic process. However, in nature such examples are
rare, although, paradoxically, they may be relatively abundant in the coastal
literature, simply because both authors and editors would view them more
favourably. No-one wishes to publish a paper in which the information
(however carefully collected) fails to support a clear conclusion.
A second problem perhaps relates to a common scientific inability to break
with convention - this type of reluctance to shift to new ‘paradigms’ has been
discussed by Kuhn (1962). For many years, coastal evolution studies have
tended to be stereotyped, assuming that there is some kind of observable and
straightforward explanation for most changes (the example of the Bruun Rule
is discussed below). The universal presence of simple linkages is manifestly
not the case (just consider how many studies fail to find such links). This
problem is not confined to coastal studies; recently, Schumm (1991) has
examined a wide range of geomorphic examples and extracted many common
failings of this type, which lead inevitably to misinterpretation. As a matter of
some urgency, researchers concerned with coastal evolution should consider
the alternative models, even if there are few supporting data. The ideas of non-
linear response, stochastic development, deterministic chaos, catastrophism
and criticality all deserve investigation.
One, almost Pavlovian, response to the dilemmas listed above, is to collect
more and more data. There is some rationale in this as the bigger database may
well provide better insights, although more data do not equal better data.
Indeed, consistency of information, especially between different coastal workers
and their projects, is rarely questioned, yet it is evident that a modicum of
effort in standardisation might pay dividends in terms of better interpretation.
 Coastal evolution: an introduction 3
Another recent trend (epitomised by the International Geological
Correlation Programme Project 274 which provided the impetus for this book)
has been to both widen the context of coastal studies and to move towards
integration between the two dominant spheres of research activity, namely
process studies of wave-morphology interaction and Holocene-scale
investigations of environmental change (especially charting the course of
former sea levels). This consideration of the co-adjustments of process and
form is termed morphodynamics (see Chapter 2).
The widening of context arises from the view of the shoreline as part of a
much larger system, within which the parts cannot be easily divorced.
Obviously, there are critical links between the oceans and the shore, but the
exact nature of these was practically unexplored until studies, initiated by
Swift, Hayes, Pilkey, Wright and others along the Gulf and eastern US
seaboards in the 1970s and 1980s, began to reveal the very intimate
connections between the beach, estuary, nearshore, shoreface and shelf. These
have a profound effect on sediment exchange (see Wright, 1987, for review).
Often, this type of sediment budget has developed on a long-term (Holocene or
Pleistocene - -lo3 to lo6 years) time scale, although there is no guarantee that
recent accelerations in the rates of global change will not reactivate and re-
establish such mass movements. A further important broadening of scope
comes from the view that terrestrial environments play a profound role in
coastal evolution. In some ways, this view has been implicit in studies since
the mid-nineteenth century, but nonetheless the increasing amount of
empirical evidence, especially where off-land sediment yields have been
decimated by such processes as shoreline ‘protection’ and river regulation, has
brought a new recognition to the subject. Again, time scale is important, as
there is evidence that such processes as sediment amount, transport and
delivery to the coast may vary by several orders of magnitude both through
natural, climatically controlled factors as well as through those mediated by
human activity. For example, Mediterranean coastal evolution shows distinct
episodes, which may be related not only to regional climatic change, but also
to anthropogenic influences such as agricultural expansion and dam building
(Vita-Finzi, 1964; Sestini, Jeftic & Milliman, 1989).
Over the last 15 000 years, first-order mediation of coastal evolution has
been related to global sea-level history, which in turn is coupled to global
climatic change and, to a lesser extent at this time scale, to tectontic activity
(Carter, 1992). There are strong latitudinal differences in climate response as
well as hemispheric ones, with the result that predictability of events is low.
Nonetheless, it is possible to speculate that, for example, fluctuations in the
glacial mass balance of Antarctica would have a worldwide impact. Anderson
4 R. W.G. Carter & C.D. Woodrofle
& Thomas (1991) suggest such ice-balance changes may well lead to rapid,
episodic sea-level variations around the world, which can be detected in
coastal sediments and morphology from a range of environments, assuming
one is looking for such evidence.
Coastal evolution also has an importance within the biological sciences. At
one level coasts need to be viewed as functional ecosystems playing an
important role in the recycling of nutrients and minerals (Brown &
McLachlan, 1990). It is not too far fetched to describe some coastal
environments, such as estuaries, as ‘reactors’, involved in energy production,
which is then exported to sustain adjacent, energy-deficient areas. For this
function alone it is necessary to maintain coastal dynamics. There is, however,
a broader function, related to long-term biological diversity and genetic
evolution. Woodroffe & Grindrod (199 1) indicate that sea-level change and
the consequent extent of intertidal habitat has been a major control on the
distribution of mangrove forests, and perhaps other intertidal communities,
which in turn gives rise to biodiversity. As such we have a vested interest in
coastal evolution as one means of providing global stability! Deceleration of
sea-level rise around 6000 years ago may have triggered Predynastic
agriculture in Egypt (Stanley & Warne, 1993a), setting the scene for the dawn
of western civilisation.
There is an opposite tendency to proselytise ideas which at one stage proved
interesting, but which through time become largely unsustainable. Such
concepts often have a large number of adherents, who are reluctant to abandon
what, in truth, is a simple, easily applicable procedure. There are plenty of
examples of such occurrences in the Earth Sciences - readers are referred to
Ginsburg’s (1973) book Evolving Concepts in Sedimentology for a series of
case studies and Dury (1978) for a more cynical view focusing on
geomorphology. Two examples in coastal evolution studies are the formula
devised by Bruun (1962), which claims to predict coastline retreat under sea-
level rise, and the ‘equilibrium profile concept’, which has a far longer history,
but which was calculated by Dean (1977, 1991; Dean, Healy & Dommerholt,
1993) to be modelled by the expression y = P X * ‘ ~ . (It is worth noting that both
Bruun and. Dean are engineers, with an interest in quantifying clients’
problems - an absolute value will always beat statistical caution.)
The original 1962 paper of Bruun’s is almost certainly a citation ‘classic’ as
it is widely and persistently quoted. Unfortunately, in the intervening 30 years
the basic idea has never been fully vindicated (certainly not in the field), for the
rather obvious reason that coastlines do not behave in a strict two-dimensional
sense. Although Bruun himself (Bruun, 1988) has tried to widen the rule’s
applicability, it remains in most situations an inadequate oversimplification.
 Coastal evolution: an introduction 5

The equilibrium profile debate is a far more complex one. There is an

intuitive feeling that if a wave-formed slope (like a hillslope) is given enough
‘time’ it will come into some kind of equilibrium with the forces applied to it.
Most authors cite Cornaglia (1891) as being the first to introduce the concept.
However, as with the Bruun problem, a wave-formed slope is subject to
multidirectional forces of considerable spatial and temporal variability, so that
the simple ‘time to equilibrium’ models of, for example, Schumm & Lichty
(1965) and Carson & Kirkby (1972) cannot apply (see Fig. 1.la).
Alternatively, one anticipates a noisy model, in which there is constant
movement across and along the coast.

I I
time ---f time

Ii d Holocene

time time

Figure 1.1. ( a ) the concept of time as expressed by Schumm & Lichty (1965). Cyclic
time is represented by the evolution of landform attributes over geological time; graded
time is a shorter timescale, over which landforms may be considered to be in a steady
state. (b) An example of dynamic equilibrium of reef islands on a Pacific atoll (after
Bayliss-Smith, 1988). Coarse-grained islands, termed motu, react to the passage of a
storm (hurricane) through an increase in size as rubble ramparts are formed; these break
down and are redistributed during the periods between storms. Cays, sandy islands, are
eroded by storms, but gradually recover after the passage of the storm. In mid-Holocene
times sand cays may have been in a steady state, and motu may have been undergoing
gradual net enlargment. During the late Holocene, as a result of less-frequent storms,
lower wave energy and reef productivity, less island vegetation and rising sea level,
islands may have less capacity to recover and may gradually become smaller.
6 R. W.G. Carter & C.D. Woodroffe
The recent work of Wright and others mentioned above (Wright, 1987,
1993; Young & Pilkey, 1992; Wright et al., 1991) is very important in this
context, as it begins to provide extensive data and new insights into the
problem of long-term nearshore response of sandy beach and barrier systems.
Pilkey et al. (1993) provide a recent review and analysis of the equilibrium
slope debate, and interestingly show the almost complete inapplicability of the
Dean formula to field data.
In many cases there is likely to be a dynamic equilibrium. One such
example is shown in Fig. 1.lb. On Pacific atolls there are two types of islands;
motu are islands composed of coral shingle and boulders on the higher-energy
reef flats, and cays are sandy islands in less-exposed situations. These respond
differently to the impact of tropical storms, with a net input of coarse material as
rubble ramparts on motu, but storm cut on cays. Motu subsequently alter with
redistribution of material between storms, while cays rebuild between storms.
Bayliss-Smith (1988) has suggested that in mid-Holocene times these reef
islands may have existed in a steady state, with storm effects averaging out over
time, but in the late Holocene, under different conditions of storm occurrence
and with reduced vegetation cover, the islands may be exhibiting a dynamic
equilibrium, but gradually reducing in size (Fig. 1.lb; see also Chapter 7).

History of coastal evolution

For as long as shorelines have been occupied it is likely humans have taken
note of coastal changes. The oral traditions (dreamtimes) of Australian
aboriginals, perhaps extending back tens of millennia, identify clearly the early
Holocene transgression and the consequences of the rising sea in destroying
food sources and disrupting communications (Flood, 1993). Biblical evidence
also emphasises the impact of coastal change, particularly associated with
extreme events (Bentor, 1989). Geomorphological reconstructions indicate
how the site at Troy, and other cities referred to in the Iliad, have altered (Kraft,
Kayan & Erol, 1980). In this, and subsequent Old World periods, shoreline
changes were especially noteworthy where they interfered with commercial or
defensive activities around river mouths, harbours or ports (e.g. Inman, 1978;
Masters & Flemming, 1983).
The development of geographical studies of coastlines in the mid-
nineteenth century quickly acknowledged the need to understand how coastal
changes occurred. The apotheosis of this early movement came with the works
of Gilbert on Lake Bonneville, and Davis on Cape Cod, and later the studies of
Johnson (1919), of which his book The New England - Acadian Shoreline,
published in 1925, represents perhaps the finest example of the genre. In this
 Coastal evolution: an introduction 7

Johnson attempts to argue from morphological evidence how a complex

shoreline has evolved. In the years immediately after Johnson’s monograph
was published, J. Alfred Steers began his career in Cambridge. Steers’ impact
on British - and to some extent Australian - coastal research was significant.
Steers (along with Vaughan Lewis) began a detailed investigation of coastal
changes around the British Isles, culminating in the publication of The
Coastline of England and Wales in 1946 (later extensively revised in 1964).
This work frequently refers to coastal evolution, as for instance in Steers’
interpretation of the evolution of Dungeness in southern England. Steers was
also responsible, along with biologists such as Frank Oliver, for the initiation
of long-term studies on Scolt Head Island off the Norfolk coast, and the
maintenance of records which have proved invaluable to present-day studies.
It is somewhat invidious to name only a few individuals, as many people
have been influential in the development of coastal evolution studies.
However, honourable mention should be made of Axel Schou who published
his monograph Det Marine Forland immediately after the Second World War.
This remarkable contribution - sadly in Danish - provides a concise and
detailed exposition of the evolution of the Danish coast. Although Schou
lacked much of the chronological support that modern coastal evolutionists
take for granted, he still managed to develop coherent arguments, which stand
the test of time. Slightly later comes the contribution of Andrt Guilcher in
France. Again, working under wartime restrictions, Guilcher was to bring a
broader perspective to coastal evolution studies through his oceanographic
(and to a lesser extent geological) interests. Guilcher’s text Coastal and
Submarine Morphology (published in French in 1954 and in English in 1958)
marked an important diversification, which has continued to find echoes in
many contemporary studies. Finally, it would be remiss not to mention the
Russian work of V. P. Zenkovich. Although occasional Russian studies had
appeared in the English literature, it was the 1967 translation and publication
of Zenkovich’s book which provided the first real insight into the parallel (and
in places, divergent) developments in coastal evolution studies between the
western and communist scientists. Many of Zenkovich’s examples clearly
found their original impetus in the pre-Revolution studies of Gilbert, Davis and
Johnson, but nonetheless a very distinctive praxis had developed. For example,
work on sediment tracing to ascertain sediment transport rates was progressing
in the 1930s, well before similar studies were attempted in the USA or UK.
Zenkovich’s own ideas on shoreline evolution - often based on spectacular
examples from inland sea coasts - are still worthy of consideration, although
many were completed by the 1950s.
During the 1950s and early 1960s, coastal research was still influenced by a
8 R. W.G. Carter & C.D. Woodroffe
number of very active individuals, such as R.J. Russell who began the Coastal
Studies Institute at Louisiana State University (Russell, 1967), C.A.M. King in
Britain (King, 1972), and C.A. Cotton in New Zealand (Cotton, 1974). After
the 1960s, it becomes harder to link influential individuals with specific
developments. There are two reasons for this. One, coastal studies in general
were expanding and, two, there was a shift of emphasis towards process
research - and to a more limited extent towards an emphasis on management
of shorelines. However, a major shift took place in the mid- to late-1970s with
the emergence of the sea-level ‘issue’. Sea-level research had been gaining
momentum throughout the 1960s under the guidance of Rhodes Fairbridge,
Walter Newman and Art Bloom, and workers were beginning to use the
Carbon-14 timescale (basically the last 40000 years) as a foundation for
assessing the time/depth course of sea-level changes for a wide variety of
environments from submerging coral atolls to Arctic raised shorelines. Bloom,
in particular, was a key figure in setting up the global networks of sea-level
researchers and was the coordinator of the first IGCP Project 61 on sea-level
changes. The scares of the late 1980s and early 1990s, which suggested that
sudden massive rises of sea level might occur as a consequence of global
warming (with estimates of several metres) rapidly provided an unprecedented
public and political awareness of coastal studies. Researchers who had worked
for many years on coastal problems were suddenly ‘in demand’, both by the
media and funding organisations. Needless to say they were joined by many
others who recognised the opportunity to present themselves as coastal
‘experts’. In some respects the normal scientific caution was ignored, as many
reports - often illustrated with lurid and misleading maps - indicated areas
likely to be flooded or submerged by rising water. Yet, the positive side has
been a strong resurgence of coastal evolution studies, often from an
interdisciplinary perspective. The completion of IGCP Project 200 under the
direction of Paolo Pirazzoli and its successor IGCP Project 274 under Orson
van de Plassche synergised coastal evolution research on a worldwide basis.
Now, new research programmes such as the Land-Ocean Interactions in the
Coastal Zone (LOICZ) core project of the International Geosphere Biosphere
Programme (IGBP) (Holligan & de Boois, 1993) are providing clear signals as
to future research directions.

Coastal behaviour
The coast ‘behaves’ in a wide variety of ways, depending on time scale, length
scale, geological structure, tectonic setting, sediment type and availability, sea-
level position, wave and current processes, and the adjacent terrestrial and
oceanic environments (Fig. 1 . 2 ~ )The
. coast must be considered as a system;
 Coastal evolution: an introduction 9
a
rook1

b Beach H
State A
I H

liquefaction during storm

Beach
State B
post-storm

storm

Hk!&x%2 time
7

intransitive threshold

time -
Figure 1.2. ( a ) Morphodynamic responses on different shorelines and coastal settings
occur at different rates. Reaction time also depends upon the sediment size, and
lithification. Thus, beaches respond over time scales of months, while barriers respond
over time scales of decades or longer (after Carter, 1988). (b) Beach states A and B
represent opposite ends of a continuum of beach states. A is accreted; however, a large
storm may cause beach liquefaction and result in the eroded state B. This appears as an
intransitive threshold over the time scale of months, with fluctuations occumng in
response to regular wave processes (H, wave height). This is a metastable equilibrium
in which there are apparently two equilibrium states. (After Chappell, 1983.) ( c ) Over
the decadal time scale, the beach may recover from state B and the beach may again
become accreted. This is the case at Moruya, southeast Australia (after Thom & Hall,
1991), where storms in the 1970s caused extensive beach cut, but in the 1980s the beach
recovered to its pre-storm volume (see also Fig. 2.11).
10 R.W.G. Carter & C.D. Woodroffe
its morphodynamics involve complex mutual coadjustment of processes and
forms. For example, on many coasts, shoaling and nearshore transformation of
waves reduce the energy incident on the beach by 95 to 99%. The frictional
turbulence in the wave zone is largely ‘lost’, although at some levels a
proportion leads to sediment entrainment and movement. In places, the coast is
able to become organised, inasmuch as relatively resistant structures are built
up which assist or alter the modes of energy, mass or information transfer.
Perhaps the best known of these organisations is that inherent in beach
morphodynamics (Wright & Thom, 1977; Wright et al., 1979; Wright & Short,
1984; Short, 1991). From the mid 1970s, various models of beach stage were
developed; basically, this involved exploring the morphological sequences
exhibited by sandy beaches in terms of transition (the movement from one
characteristic form to another). The appropriateness of this method can be
judged by both the volume of research and the fact that ‘stage’ could be linked,
quite successfully, to a wide variety of physical and, more recently, biological
phenomena (McLachlan & Erasmus, 1983). The approach also thus allows a
categorisation of sedimentary coasts at relatively short time scales - storm-
fairweather oscillations, tidal cycles and seasonal, climatically induced
variations (Fig. 1.2).
The morphodynamic approach relies on predictability along certain
environmental gradients, such as tidal range, wave exposure and sediment
type; morphodynamic behaviour varies in a deterministic manner. It is feasible
to draw out invariants - for example, Huntley & Short (1992) identify the
influence of sediment fall velocity on rip current spacing, which appears, at the
site they examined, to be independent of wave period. Fig. 1.2b shows two
end-members of a continuum of beach states. State A is accreted; however,
under a severe storm, beach sediment may undergo liquefaction, and the beach
may be cut back to an eroded beach state. The storm exceeds an intransitive
threshold (Chappell, 1983); under the usual range of non-storm wave
conditions the beach fluctuates around one state or the other. This represents an
example of metastable equilibrium. Viewed over a longer term, however, there
may be broad periods during which the beach oscillates around the erosional
state with gradual replenishment of the beach to the accreted state (Fig. 1 . 2 ~ ) .
Reaction to the storm occurs rapidly, but relaxation and beach recovery take
considerably longer, as is shown by long-term beach monitoring at Moruya in
southeast Australia (Thom & Hall, 1991; see also Chapter 2).
As a coast evolves towards an organised morphodynamic system, then it
begins to control its own environment. It may well move from a less-stable to
a more-stable one, capable of absorbing a far greater range of energy, sediment
or informational inputs (Fig. 1.2). Excellent examples of this type of system
are provided by gravel (coarse clastic beaches) which evolve towards stable
 T

morphosedimentary forms (see Fig. 1.4), in which process, forms and

sediments are coupled in both the longshore and the cross-shore directions to
provide very efficient absorbers across a wide range of energy conditions
(Carter & Orford, 1993).
The essential morphodynamic behavioural models of beaches and surf
zones offer great scope for understanding the short-term (days, weeks, months)
behaviour of the coast. Set transitions may be compared from place to place
and even along extensive stretches of coast, in the manner described by Short
(1992) for the Netherlands. However, the morphodynamic model does not
necessarily hold for longer periods or larger stretches of coastal development.
One alternative is to view the coast as a series of discrete cells, which are open
to energy input, but relatively closed to inter-sediment transfers. At the
smallest scale (perhaps 10’ to 103m shoreline) cells merge with
morphodynamic units, but at a meso-length scale ( lo3 to lo5m) the cell is
significantly larger. Over time incident wave and current forces within a cell
may come into balance such that erosion and deposition are effectively
cancelled by equipotential gradients of littoral power. However, totally
balanced cells are rare as there is almost always some loss or gain of sediment
to the system. Negative budget cells (such as estuaries, where there is a net sink
of sediment in the estuary; see Fig. 1.3a, and Chapter 5 ) are characterised by
erosion (and shoreline transgression), although some may indicate ‘arrested’
development as the cell becomes armoured. A positive sediment budget (such
as a delta, where riverine sediment reaches the coast; see Fig. 1.3b, and
Chapter 3) will see gradual accumulation of material within the cell, which
may in time renew inter-cell leakage and re-establish longshore transport. It is
common to find cells that have (or had) positive sediment budgets evolving
extensive beach ridge plains, based on persistent supplies from the shelf or
from fluvial sources (Fig. 1.3b). Over Holocene time scales it is clear that
many cells have shifted from positive to negative budgets, so that ones that
were formerly accumulating are now eroding. Interestingly, some cells,
especially those near circulatory fluvial/marine systems (at river and estuary
mouths), may evolve as mass sediment attractors. On the east coast of the USA
and elsewhere this results in huge ebb and flood tidal-delta systems, satiated
with long phases of sediment input. Such systems are clearly too over-fit in the
sense that they bear little relationship (in terms of size) to current inlet
hydraulics. In southeast Australia, fluvial input is stored both in estuaries and
along the shore in the form of massive inshore ramp structures (Roy, 1984).
Cell development has allowed the gradual leakage of material from estuaries
onto the shelf where it has accumulated, in what might be justifiably described
as fossil forms (see Chapter 4). On many downwind ocean coasts, leakage and
storage is often landward into coastal dunes. The gradual transfer of sediment
 Longshore Headland + Inputs Headland
movement
+ outputs
Longshore
movement r2 Cliff erosion
,\ \ I: L /
/
/ / I I=.
River
Onshore/ I
offshore Tidal niver
movement prchnnnnW sediment

b Barrier
overwash

Swash-
aligned
Dune transgressive
formation
shore,ine

Figure 1.3. Sediment budgets and cell circulation on estuarine and deltaic coasts. ( a ) Riverine sediment may be deposited into an estuarine
embayment, which may also receive sediment from the seaward direction through tidal processes. Note the swash-aligned sandy barrier which
retreats landwards (transgresses) through washover and aeolian processes. ( h ) On a deltaic coast, riverine sediment contributes to coastal
sediment budgets, and is moved up the drift-aligned coast, giving rise to a beach-ridge plain. Dashed lines represent rocky coast; stipples
represent sandy coast.
 Coastal evolution: an introduction 13
from the marine to the aeolian system may be relatively slow, with net
movement only a fraction of gross transport, but over a geological time span it
can be hugely effective, especially if fuelled by a positive-turned-negative
sediment budget, allowing the release of large quantities of sediment. On the
coast of northwest Ireland, the present shelf and shoreface are almost devoid of
large sand bodies as all the sediment has moved inland to form dunes and sand
sheets (Carter & Wilson, 1993).
There are many different types of cell, including those associated with fixed
and free boundary conditions, zeta-bays (log-spiral bays shaped by wave
refraction patterns) formed under conditions of oblique, but persistent swell,
spit cells maintained by counteracting waves and currents, and so on (Carter,
1988). The dominant attribute of cell structures is that they function as longer-
term energy absorbers based on net sediment transport potential.
The cell structure may be extended to embrace the onshore and offshore
boundaries, particularly onto the shoreface and inner shelf, and where
subaerial sediment storage (in beach ridges, dunes and wetlands) occurs.
Long-term exchange of material across and along the coast within defined cells
provides one firm basis for analysis of secular coastal change. At a scale of 10'
to 102km cell definition breaks down, as the process field becomes too
variable. At this length scale, the size of individual weather systems
(particularly tropical cyclones or hurricanes) may be equal to or smaller than
coastal compartments. A pertinent example of this is provided by the western
Gulf of Mexico, where the Louisiana-Texas arc is over 1000km, while
individual hurricane landfalls may only occupy a tenth of this zone. As Keen
& Slingerland (1993) have demonstrated, the impact at this scale is to create a
far more complex pattern of sediment removal and redistribution which cannot
be easily reconciled with any cellular structure. In this context sediment is
being slowly transferred from the riverine terrestrial to the shelf via coastal
processes. Evolution of the shore depends on both the balance between input
and output of sediment, and a more complex probabilistic relationship linking
storms to sediment transport.
In recent years, some attention has focused on the ideas of Large-Scale
Coastal Behaviour (LSCB) (Terwindt & Battjes, 1991; Stive, Roelvink & de
Vriend, 1991; de Vriend, 1992a, 1992b; see also Chapter 2), although the term
is relative, as some author's viewpoints tend to reflect their previous scale of
coastal research. The LSCB concept implies that coastal behaviour may be, in
some way, non-linear. The last decade has seen a proliferation of possible non-
linear models, ranging from deterministic to stochastic to chaotic, and it is not
difficult to envisage situations that might fit certain coastal contexts (e.g.
Phillips, 1992). There is reasonable evidence, for example, that the structure of
14 R.W.G. Carter & C.D. Woodroffe
turbulence is deterministically chaotic, and as a number of coastal phenomena
develop under turbulent flow, this should (or could) impart a similar spatial
pattern to the landforms. Coastal dunes are an obvious example; transverse
dunefields, such as the Alexandria Dunefield east of Port Elizabeth in South
Africa (Illenberger & Rust, 1988), are chaotic in terms of form. Certainly,
given free expression, there is a myriad of superimposed sand forms at a
variety of length scales, which may result in a chaotic landscape.
A bigger problem arises in trying to argue for complex non-linear behaviour
over time, although in terms of understanding coastal evolution this is
probably a key issue. Even if a system is driven by a relatively persistent
forcing function - say sea-level rise - there is no guarantee the response will be
straightforward. The classic earth science example of this phenomenon is
where inter- and intra-plate stresses are resolved into a complex pattern of
earthquakes, which have now been shown to be chaotically predictable
(Turcotte, 1991). It may be simplistic to assume that sea-level rise will result
in predictable responses. A more likely 'scenario' is that a simple forcing
function will lead to a variety of responses, depending to a large extent on the
ability of the coastline to absorb the stress. Thus, initially a cliff coast will
resist (or at least show few signs of alteration), while a diverse, sedimentary
coast may well show a range of rapid adjustments. In time, however,
thresholds may be exceeded, so that widespread cliff failures occur; to the
casual observer (and the scientist), these may be simply unexplainable.
Adjustment thus proceeds in a stepwise fashion. The behaviour may also jump
from one mode to another; Forbes et al. (1991) described the behaviour of a
gravel barrier in eastern Nova Scotia, which prior to 1954 had built-up in situ,
yet after 1954 (the date marks the passage of a major, but not exceptional
storm) began to migrate landward at rates of over l o m a - ' . Carter & Orford
(1993) have speculated that this type of jump behaviour is not unusual and in
fact approximates to a fold catastrophe model (Zeeman, 1977) in which one
slow variable (sea-level rise) is associated with one or two fast variables (wave
field variation, sediment supply). Existence of such behaviour has implications
for adjoining environments, which may have to undergo controlled phases of
change, associated with the dominant morphological structure. Antecedent
morphology is almost certainly an important variable in this respect (see
Orford, Carter & McCloskey, 1993).

Single events with a geological time scale return period

There remains the possibility, at the extreme end of the frequency/magnitude
curve, that single, independent or isolated events may have an unduly
 Coastal evolution: an introduction 15
important significance to coastal evolution. In the past few years evidence has
been mounting from the North Sea (Long, Smith & Dawson, 1989), southwest
Britain (Foster et al., 1993), the Mediterranean (Andrade, 1992), the Pacific
Northwest of North America (Atwater, 1987; Atwater & Moore, 1992), Japan
(Minoura & Nakaya, 1991) and Australia (Young & Bryant, 1992; Bryant,
Young & Price, 1992) that exceptional tsunamis may be capable of massive
reorganisation of the coast. (It needs to be emphasised that tsunamis are
relatively common on decadal scale (Watanabe, 1985), but in these cases the
events can probably be assimilated into the high-frequency end of the spectrum
rather in the manner of hurricanes.) This discussion centres on what might be
described as ‘mega-tsunamis’, occurring so rarely that they are beyond the
time scale of perhaps even secular fluctuations of Holocene sea levels. In many
cases they centre on major epipeiric Earth movements, explosive volcanic
eruptions (Santorini, Krakatoa) or perhaps meteorite impacts (i.e. Bilham &
Barrientos, 1991). The environmental context of such perturbations is
obviously difficult to assess objectively - although some authors see them as
‘driving forces’ for global change (Rampino & Self, 1992). For example,
Young & Bryant (1992) consider that they have been able to identify an event,
around 100000 years ago, which devastated the coast of southeast Australia.
They believe this may have been caused by a massive submarine landslip in the
Hawaiian Islands 105 000 years BP, creating a group of tsunami waves which
radiated circumferentially across the southwest Pacific. Young & Bryant are
able to point to a wide range of what at first appear to be anomalous shore
features, including rock channels across headlands, scoured and fluted rock
platforms at a variety of levels, poorly-sorted high-level boulder deposits
(some boulders weigh over 100 tonnes), and hummocky ‘dump’ deposits in
sheltered areas. Taken individually such features would elicit as many
explanations as there are geomorphologists in a field party, but put together
they begin to reveal a pattern repeated over several hundred kilometres of
shoreline. More controversially, Young & Bryant argue that this single
spectacular event may have completely removed the outer coastal barriers and
initiated a new phase of barrier, dune and beach development along the coast.
Should this be the case, then the many long-held ideas about coastal evolution
in this area will need to be revised (see Young et al., 1993).
A more general point about the single, almost geologically isolated event is
that the coast has to be in a receptive or sensitive condition in order to retain
the record. The widespread evidence for the Storegga slide tsunami in the
North Sea (Dawson, Long & Smith, 1988; Long et al., 1989) around 7200
years BP coincides with the peak Holocene transgression in Scotland, so that
the deposit - basically a sand layer - stood an excellent chance of preservation.
16 R.W.G. Carter bi C.D. Woodroffe
Moreover, the local timing is clearly important, as the arrival of a tsunami at
high tide is likely to leave more evidence than if one arrives at low water.
Similarly, a salt marsh or coastal lake only records storm surges or hurricane
events in localised sites and when conditions are suitable for deposition of
discrete sand or shell layers (i.e. Ehlers et al., 1993; Liu & Fearn, 1993).

Models of coastal evolution

There have been various attempts to model the way that coasts have developed,
ranging from highly subjective denudation chronologies proposed by W.M.
Davis and those geomorphologists who viewed coasts as developing as part of
a geographical cycle, to highly quantitative computer models (Lakhan &
Trenhaile, 1989). A major objective of IGCP 274 was to develop models of
coastal evolution. In most cases these are conceptual morphodynamic models
based on detailed radiometric-dating reconstructions of what has occurred in
the past, though as our understanding of processes involved increases it may be
possible to develop simulation models, such as the shoreface translation model
applied to the coast of New South Wales (see Chapters 2 and 4). Below we
describe two conceptual models of coastal evolution from our own studies,
from disparate field areas, and illustrate how several common themes may be
identified between them.

Gravel barrier evolution in Nova Scotia

Nova Scotia provides a wide range of coastal environments (Stea et al., 1992).
Among the most distinctive is the drumlin coast of the Eastern Shore, where
late-Wisconsinan deposits are being eroded and transported under a rapidly
rising sea level (3.8 m d y e a r since 1917) (Orford et al., 1992). As the sea level
rises so drumlins are eroded in turn; the average ‘life’ of each eroding drumlin
is probably about 2000 years (Carter et al., 1990). The drumlins comprise a
heterogeneous mix of coarse and fine sediments and as they are attacked by
waves, released material is dispersed according to transportability. This leads
to the development of a transport pathway network (which is interactive),
moving different types of sediment onto the shelf, along the coast or inland
through the estuaries or coastal dunes. Because each drumlin is in a different
stage of evolution there is a competitiveness between them to control the
transport pathway network, and indeed, such control is sequestered only
reluctantly as the drumlin sediment is eventually depleted and the sea rises
finally to drown it. Distinct morphosedimentary architectures and morpho-
dynamic responses can be recognised for both sandy barriers (Nichol & Boyd,
 Coastal evolution: an introduction 17
1993) and gravel barriers (Carter & Orford, 1993) along this coast (see also
Chapters 4 and 10).
Mobile gravel released from the drumlin is moved rapidly by waves along
the drumlin flanks within what is a very abrupt wave power gradient; wave
power may decrease by 90% in only a few hundred metres (Carter et al., 1990).
At first these flanking beaches are immature in the sense that the sediments
show few signs of sorting or profile development (Fig. 1.4). However, in time

Figure 1.4. Evolutionary stages of gravel barriers in Nova Scotia (based on Orford
er al., 1991, and Carter & Orford, 1993). Three phases, linked with drumlin erosion,
which provides a source of sediment, can be recognised. In the initial phase (I) barriers
are drift-aligned; some may become drift-arrested (Ia) and can break down (Ib). Further
sediment supply leads to a phase that is swash-aligned (11). In some cases barriers can
become swash arrested (IIa). In the third phase there are either concatenated (IIIa), or a
metastable equilibrium between massive solitary barriers (IIIb), where the gravel
barrier becomes heightened, or a stretched and overstepped barrier (with a submerged
barrier left behind) (IIIc).
18 R. W.G. Carter & C.D. Woodroffe
an organisation begins to evolve. During the initial stages continued erosion of
the source leads to the extension of the flanking barrier, which in time will lose
contact with the drumlin and tend towards an equilibrium with the incident
waves. This leads to a switch from drift-aligned barriers to swash-aligned
barriers (Orford, Carter & Jennings, 1991). The gravel barrier then enters a
new phase of evolution, that of a transgressive feature. In time, the barrier
becomes stretched and a breach may occur. This breach may well provide a
threshold for dramatic change, causing the realignment and breakdown of the
swash-orientated barriers, and allowing the penetration of wave energy into
previously sheltered back barrier zones. The swash-aligned barriers also
breakdown under sea-level forcing. In time, there may be a catastrophic switch
from a heightened, massive solitary barrier to an overstepped barrier (Carter &
Orford, 1993) where elements of the upper and lower barrier can be
disengaged (Forbes et al., 1991).
Over the relatively short time scale of the last two millennia, the gravel
barriers of Nova Scotia show a surprising range of behaviour, linked to both
morphodynamic and morphosedimentary responses to changing sea level,
sediment availability and local wave field variability (see Fig. 1.4). Many of
these responses can be found in the records of adjacent environments, such as
salt marshes (Carter et al., 1992; Jennings, Carter & Orford, 1993.) Overall,
the changes are decidedly non-linear, but within periods of decades it is
possible to detect linearity between some variables (Orford et al., 1991).

Tidal river and floodplain evolution in northern Australia

A number of northward-flowing rivers drain the sandstone uplands of the
Arnhem Plateau and low rolling laterite hills of the Top End of the Northern
Territory. These rivers are tidal as they enter macrotidal van Diemen Gulf,
with tidal influence extending up to 100 km upstream as they meander across
broad Holocene alluvial/estuarine plains. The tropical climate of this area is
distinctly seasonal, with a dry season from June to October followed by a
monsoonal, wet season from November to May.
Recent morphostratigraphic work on the macrotidal rivers and their
floodplains has built upon initial observations on the Ord River in Western
Australia (Wright, Coleman &Thorn, 1973; Thom, Wright & Coleman, 1975).
The most detailed morphodynamic reconstruction has been undertaken on the
South Alligator River, but a similar pattern of development is indicated for
other neighbouring river systems (Woodroffe et al., 1989, 1993; Chappell,
1993, see also Chapter 5).
The basic sequence of coastal evolutionary events can be divided into three
phases (Fig. 1.5). Phase 1 occurred when the prior valley was flooded by a
 Coastal evolution: an introduction 19

Upland

Mangrove

Palaeochannel
and track of river
Floodplain
(including
and backwater
saline mudflat
swamp)

SINUOUS kilometres
TRANSGRESSIVE BIG SWAMF CUSPATE
PHASE PHASE PHASE

Figure 1.5. Estuarine plains flanking the macrotidal estuaries of the Top End of the
Northern Territory, Australia can be shown to have undergone three phases of
Holocene evolution. The first phase is the transgressive phase, 8000-6800 years BP,
during which the prior valley was inundated by rapidly rising postglacial sea-level rise
and organic muds were deposited beneath landward-migrating mangrove forests. The
second phase, the big swamp phase, 6800-5300 years BP (on the South Alligator River),
was a period of widespread mangrove forests formed on vertically accreting intertidal
mudflats. This was replaced by a phase during which the river adopted a sinuous or
cuspate meandering form, reworking part of the floodplain, which accreted vertically
under freshwater wetland vegetation (from Woodroffe er al., 1993).

rapidly rising postglacial sea level which reached around 14m below the
present level about 8000 years BP. This transgressive phase is represented by
highly organic muds or peat and landward extension of mangrove forests,
which grade upwards into vertically accreted estuarine muds deposited
beneath mangroves.
Deceleration of sea level rise around 6000 years BP, plus a rapid and
widespread growth of mangroves (both within the newly-forming estuaries
and geographically along the coast of northern Australia), resulted in what has
been termed the ‘big swamp’ phase (Woodroffe, Thom & Chappell, 1985).
This ‘big swamp’ phase lasted from 6800 to 5300 years BP in the South
Alligator estuary, and occurred contemporaneously in many other
embayments in northern Australia (based on numerous I4C dates - see
Woodroffe, Mulrennan & Chappell, 1993).
20 R. W.G. Carter & C.D. Woodroffe

Termination of the big swamp phase was rapid and it was replaced by
vertical aggradation of alluvial plains beneath freshwater wetlands. Rapid
progradation of the coastal plains associated with these estuarine systems
occurred 5000-3000 years BP, followed by much slower progradation and the
episodic development of chenier ridges. The channel morphology of the
meandering river, and the extent to which it has reworked the floodplain (see
Fig. l S ) , varies between rivers in response to fluvio-tidal parameters, and is
examined in Chapter 5.

Similarities between these models

These models of very different landform suites from disparate parts of the
world have several features in common. First, both are ‘systems’ which need
to be assessed in terms of time and space, and which respond to marine and
terrestrial processes. The systems become organised, absorbing and adjusting
to environmental change, especially in becoming buffered against fluctuations
in ambient energy, for example storms. For example, the gravels are able to
develop specific clastic architecture, which is more resistant to energy
perturbation than loose aggregates and the shoreline changes from the
reflective to the dissipative end of the spectrum with time, while the expansion
of shoreline vegetation (especially mangroves) on the tropical shoreline
performs a similar function. As the organisational level increases (an entropy
defying process, reliant on extracting and using incident energy), so the system
becomes more robust, often developing quite sophisticated feedbacks between
one sub-system and another. This leads to a situation of homeostasis, which is
hard to break down, although such breakdown may ultimately be achieved by,
for example, allogenic influences such as failures of sediment supply or extreme
climatic conditions. Under such conditions the systems become stressed, the
gravel barriers thin and eventually breach, the finer siltklay deposition of the
tropical estuaries fails and perhaps ceases, and, in time, the system becomes
sensitive and suitable for sudden change, such as the rapid salinisation seen on
several of the estuarine floodplains (see Chapter 5). The switch in
sedimentation patterns in the estuaries of the northern Australian rivers marks
transitions from one system state to another. The trigger for change in the Nova
Scotia gravel systems appears to be storms, but only if antecedent conditions
are suitable. Adjacent beaches and barriers have been observed to respond
differently to what, in many cases, are isotropic sea-state conditions.
Three factors are important in both examples. First, the role of relative sea-
level change; sea level rose rapidly flooding prior valleys in northern
Australia, but stabilised around 6000 years ago, enabling extensive mangrove
 Coastal evolution: an introduction 21
forest development. In contrast to this largely eustatically determined relative
sea-level history, the Nova Scotian coast has experienced glacioisostatic
adjustment with a more individualistic pattern of relative sea-level change that
includes ongoing rapid sea-level rise to which gravel barriers continue to
respond. The second factor is sediment supply. In the case of the Nova Scotian
coast, sediment is supplied from the erosion of drumlins; input rates are highly
variable in space and time, and exert important morphodynamic controls on
barrier evolution. In northern Australia, muds have infilled the estuarine
embayments largely from a seaward source which appears to have become
depleted in recent millennia. Floodplains have since been reworked to different
extents, depending upon catchment size and differing riverine sediment inputs.
The third factor is geological inheritance. In Nova Scotia this is the drumlin
landscape which is being reworked. In northern Australia, it is the prior valley
topography which influences estuarine morphodynamics. These three factors,
relative sea level, sediment supply and geological inheritance can also be seen
to be important controls on coastal evolution in a range of settings in the
chapters that follow.
The two examples above also serve to illustrate the approaches which can
be used to construct conceptual models of coastal evolution and the scope and
applicability of such models. Both models were developed on the basis of
detailed field observations of morphosedimentary architecture, morpho-
stratigraphy and morphodynamic response. They were developed on the basis
of comparison of adjacent sites, and each has been found to have elements
which may be applied outside the area of derivation; thus, the gravel barrier
studies in Nova Scotia show many elements of similarity to gravel-dominated
coasts in Ireland (Carter et al., 1989), whereas the pattern of Holocene
estuarine infill and coastal progradation in northern Australia is similar to that
experienced through many parts of Southeast Asia (Woodroffe, 1993).

Human impact and coastal evolution

To ignore the role of humans and their impact on coastal evolution would be
fallacious. Like it or not, almost all coastal systems have been influenced by
human intervention and the last two centuries of evolution have been
dominated by attempts to control the shoreline. Coastal resources are seen as
valuable (for tourism, minerals, agriculture, aquaculture, etc.), although it is
only since the 1970s that any serious resource conservation has been
attempted, and often then on a limited scale. National initiatives are now
emerging, for example in the USA, Australia, France, Spain, Denmark and the
UK (Ruddle et al., 1988), and further efforts are being made to bring
22 R. W.G. Carter & C.D. Woodroffe

developing countries (which perhaps have the most coastal resources still to
lose) into various local and regional plans (Gomez, 1988; Paw & Thia-Eng,
1991). However, it is proving very difficult to break free of existing
governmental and administrative structures, and in many places local interests
still have over-riding influence.
There are two ways humans have affected coastlines. The first is purely
exploitative, and the second is by trying to control processes and thus change.
One can see such actions both as advertent and inadvertent. Very often the
consequences of any particular action are unforeseen and may unfold over a
long period of time. Legally, such actions may well fall outside any normal
statute of limitation (say 25 years) and, in any case, it may be impossible to
‘prove’ cause and effect.
The history of coastal evolution is linked inextricably with the evolution of
coastal catchments and drainage basins, and their ability to deliver sediment to
the shoreline. Work by Millimam, Meade, Walling and others has shown that
the rate of sediment yield from catchments has varied systematically over the
last few centuries (Milliman & Syvitski, 1992), parallelling deforestation and
agricultural expansion (and in places, contraction). The type and volume of
sediment depends on the lithology, tectonics, climate and vegetation of the
drainage basin. The Yellow River in southern China, which flows through
thick loessic deposits, produces over 1000 X lo6 t a-I of mainly fine silt, which
both supplies the rapidly expanding and subsiding delta and produces dense
turbidity currents taking sediment to deeper water (Wright et al., 1990). In
contrast, the Brahmaputra drains a tectonically active mountain region
providing a much coarser sediment yield, although much of the material most
suitable for coastal evolution tends to be sequestered by the subsiding, alluvial
delta (Barua, 1991). In general, smaller, high-relief catchments deliver coarser
sediment to the coast, partly due to the nature of the weathering process and
partly because of the relatively high competence of flows in the lower reaches
of the channel, enabling larger clasts to be moved.
The wholesale perturbation of coastal sediment budgets worldwide at local,
regional and global scales is a matter for international concern. The
contributions of both large (Milliman & Meade, 1983) and small rivers
(Milliman & Syvitski, 1992) seem equally important, although data,
particularly for the latter, are sparse. A conservative estimate would suggest a
total annual sediment input to the coast around 1016-’7t, of which maybe 10 to
15% would be suitable for ‘coastal’ accumulation. In addition there are great
worldwide disparities; for example, continental Europe yields only small
amounts of sediment relative to southeast Asia.
The net result of changing catchment conditions has been a major influx of
 Coastal evolution: an introduction 23

sediment to the coast over the last few hundred years. The timing of this event
has varied from place to place, but the manifestation of the impact has been
remarkably similar, with the progradation of river mouth deltas, the
augmentation of littoral drift, the infilling of estuaries, the construction of
beach ridge plains and the formation of dunes. In many places, such as the Nile
or Ganges deltas, the expansion has been in the order of kilometres, creating
large tracts of new land which has been voraciously settled, so that it now
supports some of the densest rural settlement in the world (Stanley & Warne,
1993b) . These increases in coastal sedimentation brought problems, not least
the need to abandon harbours as waterways became unnavigable. This led to
the development of dredging techniques and later to harbour protection works,
such as jetties.
In many places, the influx of coastal sediment masked the slow but much
longer-term tendency towards coastal recession. In the twentieth century, far
more attention has been paid to catchment management, as it was becoming
obvious that the spread of soil erosion and loss of soil fertility could not
continue unabated. This has been augmented through the extensive
construction of dams. Although dams have been constructed throughout
recorded history, their large-scale impact on the environment has only become
apparent in the last few decades (Jeftic, Milliman & Sestini, 1992; Stanley &
Warne, 1993b). Dams are being constructed for a variety of reasons, including
power generation, regulation of water supply and, all too often, politics. The
chief impact on the coast has been to reduce dramatically (often by several
orders of magnitude) the amount of sediment. This has led, inevitably, to an
increase in coastal recession. For example, in the Nile Delta, where the
sediment supply has almost completely ceased following the completion of the
Aswan High Dam in 1964 (Stanley & Warne, 1993b), this amounts to tens if
not hundreds of metres a year (Milliman, Broadus & Gable, 1989). Coping
with such erosion is not easy and the general policy in Egypt is to abandon the
shoreline in favour of constructing solid defences inland.
The artificial removal of sediments from the coast is a worldwide
phenomenon, practised on both large and small scales. Common reasons for
removing shoreline materials include precious-mineral mining (diamonds,
titanium, gold), building aggregates and agricultural soil improvers. In some
examples, such practices have direct impacts on the shoreline; Msangi,
Griffiths & Banyikwa (1988) recount how coastal sediments along the
Tanzanian coast were used for concrete in beachfront hotel construction,
resulting almost immediately in a severe erosion problem. Elsewhere the
consequences are not so obvious or fast.
The potential impact of engineering works on coastal evolution is
24 R.W.G. Carter & C.D. Woodroffe
enormous, especially if coupled to sea-level change. Universal response to
these alterations and fluctuations in the sediment budget has been to defend the
coast using engineering. The scale and ingenuity of coastal engineering can be
extraordinary. In the USA long stretches of the east coast are wholly artificial
and have been through several phases of engineering over the last 120 years.
The south and east coast of Britain is heavily 'defended' with an extensive
mixture of old and modern protective works, many now in need of
replacement, while the Belgium, Dutch, German and Danish coast is almost
entirely groyned and has long stretches of sea-walls (see, for example, Ehlers,
1990). Recently, Miossec (1991) has produced a major summary (in French) of
comparative approaches to coastal defence in Europe, which indicates how
strongly regional and national policies have influenced the type of artificial
shoreline that has been developed. Chapter 13 explores the concept of
developed coasts, drawing examples particularly from New Jersey.

Overview
The chapters which follow develop the issues of Late Quaternary coastal
evolution introduced above, and examine specific sites where detailed
morphodynamic studies have been undertaken. Chapter 2 examines the
concept of coastal morphodynamics; it reviews the approaches adopted by
Wright & Thom (1977) in a benchmark paper, and extends these in the light of
recent developments in the field.
Subsequent chapters examine specific coastal types. Each chapter
demonstrates general principles by a detailed study of particular coasts on
which it is possible to reconstruct something of the evolution of landforms, and
then examines the extent to which such models can be extrapolated to other
parts of the world. It is clear that even the detailed case studies require further
research to refine the models; nevertheless, the extension of models of coastal
evolution developed in one location to another geographical setting,
recognising that that entails altered boundary conditions, has been very much
a central focus of IGCP Project 274.
It has been recognised for some time that the morphology of clastic coasts
responds to the relative dominance of river, wave and tidal factors (Wright &
Coleman, 1973). This view has been extended by Boyd, Dalrymple & Zaitlin
(1992) and Dalrymple, Zaitlin & Boyd (1992), and is shown schematically in
Fig. 1.6. Deltas are examined in Chapter 3, with emphasis on the Mississippi
deltaic plain. Chapter 4 describes wave-dominated coasts, with particular
attention to the development of sandy barriers in southeast Australia, a location
in which coastal evolution studies have been assisted by the application of a
 Coastal evolution: an introduction 25

Wave/Tide Power

Figure 1.6. A ternary diagram showing how clastic shorelines may be classified on the
basis of the extent to which they are dominated by river, wave or tidal factors (after
Boyd et al., 1992, and Dalrymple et al., 1992).

simulation model of shoreline behaviour. Chapter 5 examines the morpho-

dynamics of macrotidal estuaries, concentrating on tidal rivers and floodplains
in northern Australia, where adjacent systems may vary in terms of important
boundary conditions. Chapter 6 considers microtidal and lagoonal coasts,
comparing such settings along the African coastline with similar settings
elsewhere.
Carbonate coasts, such as reefs, where calcareous sediment is produced,
transported, deposited, lithified and eroded, show some significant differences
from clastic coasts. Chapter 7 examines coral atolls, comparing mid-ocean
atolls from the Indian and Pacific Oceans. Chapter 8, on the other hand,
describes the evolution of continental shelf reef systems, with particular
emphasis on the Great Barrier Reef, where the influence of the continent in
terms of water quality can be considerable.
High-latitiude coasts differ from low- and mid-latitude coasts, both because
of climate and hence processes, but also because glacioisostatic responses are
still ongoing. Chapter 9 examines Arctic coastal plains, those polar shorelines
on which permafrost is an important influence on sediments, and sea ice is
seasonally extensive. Chapter 10 examines paraglacial coasts, using the
example of the Nova Scotian shoreline to illustrate the diversity of landforms
which can result from reworking of a complex, glacially influenced landscape.
Rocky coasts tend to respond more slowly than soft coasts to perturbations.
Chapter 11 describes cliffs and shore platforms, demonstrating that a variety of
26 R. W.G. Carter & C.D. Woodrofe
controls operate to influence rate of cliff retreat. Chapter 12 considers tectonic
shorelines, those coasts on which vertical movement is, or has been,
significant. Tectonic coasts cover a wide range of different types of coast, and
vertical movment may be expressed morphologically in different ways.
Finally, as emphasised above, consideration of coastal evolution without
the human factor overlooks an increasingly important variable influencing
processes and directions of change. Chapter 13 discusses developed coasts,
using the highly urbanised and built up coast of the eastern United States as an
example.
These chapters represent only a selection of the studies undertaken within
the scope of IGCP Project 274. While they demonstrate something of our
understanding of the morphodynamics of Late Quaternary shorelines, there are
also many new questions raised as to how coasts adjust their form and evolve
through time.

Acknowledgements
Much of this chapter was written by Bill Carter in the weeks before his death
in July 1993. His inspiration lies not only behind this chapter, but in the
shaping of this book, in the initiatives that the IGCP Project 274 has taken, and
in new directions that coastal evolution studies will continue to take. Julian
Orford has kindly read and commented on a draft of this chapter. David Martin
and Richard Miller drew the diagrams.

References
Anderson, J.B. & Thomas, M.A. (1991). Marine ice-sheet decoupling as a mechanism
for rapid, episodic sea level change; the record of such events and their
influence on sedimentation. Sedimentary Geology, 70, 87-1 04.
Andrade, C. (1992). Tsunami generated forms in the Algarve Barrier Islands. Science
of Tsunami Hazards, 10, 21-4.
Atwater, B.F. (1987). Evidence for great Holocene earthquakes along the outer coast
of Washington State. Science, 236,9424.
Atwater, B.F. & Moore, A.L. (1992). A tsunami about 1000 years ago in Puget Sound,
Washington. Science, 258, 1614-17.
Barua, D.K. (1991). The coastline of Bangladesh: an overview of processes and
forms. Coastal Zone '91, American Society of Civil Engineers, 2284-301.
Bayliss-Smith, T.P. (1988). The role of hurricanes in the development of reef islands,
Ontong Java Atoll, Solomon Islands. Geographical Journal, 154, 377-9 1.
Bentor, Y. K. (1989). Geological events in the bible. Terra Nova, 1, 326-38.
Bilham, R. & Barrientos, S. (1991). Sea-level and earthquakes. Nature, 350, 386.
Boyd, R., Dalrymple, R. & Zaitlin, B.A. (1992). Classification of clastic coastal
depositional environments. Sedimentary Geology, 80, 139-50.
 Coastal evolution: an introduction 27
Brown, A.C. & McLachlan, A. (1990). Ecology of sandy shores. Amsterdam: Elsevier.
Bruun, P. ( I 962). Sea level rise as a cause of shore erosion. American Society of Civil
Engineers, Proceedings and Journal, Waterways and Harbors Division, 88,
1 17-30.
Bruun, P. (1988). The Bruun rule of erosion: a discussion on large-scale two and three
dimensional usage. Journal of Coastal Research, 4, 626-48.
Bryant, E.A., Young, R. and Price, D.M. (1992). Evidence of tsunami sedimentation
on the southeastern coast of Australia. Journal of Geology, 100,753-65.
Carson, M.A. & Kirkby, M.J. (1972). Hillslope form andprocess. Cambridge
University Press.
Carter, R.W.G. (1988). Coastal environments: an introduction to the physical,
ecological and cultural systems of coastlines. London: Academic Press.
Carter, R.W.G. (1992). Sea level changes, past, present and future. In Applications of
Quaternary research, ed. J.M. Gray, Quaternary proceedings, vol. 2, pp. 1 1 1-32.
Cambridge: Quaternary Research Association.
Carter, R.W.G., Forbes, D.L., Jennings, S.C., Orford, J.D., Shaw, J. & Taylor, R.B.
(1989). Barrier and lagoon coast evolution under differing relative sea-level
regimes; examples from Ireland and Nova Scotia. Marine Geology, 88, 221-42.
Carter, R.W.G., Orford, J.D., Forbes, D.L. & Taylor, R.B. (1990).
Morphosedimentary development of drumlin flank barriers with rapidly rising
sea-level, Story Head, Nova Scotia. Sedimentary Geology, 69, 117-38.
Carter, R.W.G., Orford, J.D., Jennings, S.C., Shaw, J. & Smith, J.P. (1992). Recent
evolution of paraglacial estuary conditions under conditions of rapid sea-level
rise: Chezzetcook Inlet, Nova Scotia. Proceedings of the Geologists
Association, London, 103, 167-86.
Carter, R.W.G. & Orford, J.D. (1993). The morphodynaimcs of coarse clastic beaches
and barriers: a short- and long-term perspective. In Coastal morphodynamics,
ed. A. Short. Journal of Coastal Research, Special Issue, 15, 158-79.
Carter, R.W.G. & Wilson, P. (1993). The geomorphological, ecological and
pedological development of coastal foredunes at Magilligan Point, Northern
Ireland. In Coastal dunes: form and process, ed. K.F. Nordstrom, N.P.Psuty &
R.W.G. Carter, pp. 129-57. Chichester: John Wiley & Sons.
Cornaglia, P. (1 89 1). Sul regime della spiagge e sulla regulazione dei porti. Turin:
GB Paravia. 569pp.
Chappell, J. (1983). Thresholds and lags in geomorphological changes. Australian
Geographer, 15,357-66.
Chappell, J. (1993). Contrasting Holocene sedimentary geologies of lower Daly River,
northern Australia, and lower Sepik-Ramu, Papua New Guinea. Sedimentary
Geology, 83, 339-58.
Cotton, C.A. (1974). Bold coasts: annotated reprints of selected papers on coastal
geomorphology, 1916-1969. Wellington: A.H. & A.W. Reed.
Dalrymple, R.W., Zaitlin, B.A. & Boyd, R. (1992). Estuarine facies models:
conceptual basis and stratigraphic implications. Journal of Sedimentary
Petrology, 62, 1030-43.
Dawson, A.G., Long, D. & Smith, D.E. (1988). The Storegga slides: evidence from
eastern Scotland for a possible tsunami. Marine Geology, 82, 271-6.
Dean, R.G. (1977). Equilibrium beach profiles: US.Atlantic and Gulf Coasts.
Department of Civil Engineering, Ocean Engineering Report 12. Newark, DE:
University of Delaware.
Dean, R.G. (1991). Equilibrium beach profiles: characteristics and applications.
Journal of Coastal Research, 7,53-84.
28 R.W.G. Carter & C.D. Woodroffe

Dean, R.G., Healy, T.R. & Dommerholt, A.P. (1 993). A ‘blind-folded’ test of
equilibrium beach profile concepts with New Zealand data. Marine Geology,
109,253-66.
de Vriend, H.J. (1 992a). Mathematical modelling and large-scale coastal behaviour.
Part 1. Physical processes. Journal of Hydraulic Research, Special Issue,
Maritime Hydraulics, 7 2 7 4 0 .
de Vriend, H.J. ( 1992b). Mathematical modelling and large-scale coastal behaviour:
Part 2. Predictive models. Journal of Hydraulic Research, Special Issue,
Maritime Hydraulics, 74 1-53.
Dury, G.H. (1 978). The future of geomorphology. In Geomorphology: present
problems and future prospects, ed. C. Embleton, D. Brunsden & D.K.C. Jones,
pp. 263-74. London: Oxford University Press.
Ehlers, J. (1990). The morphodynamics of the Wadden Sea.. Rotterdam: Balkema. 397pp.
Ehlers, J., Nagorny, K., Schmidt, P., Stieve, B. & Zietlow, K. (1993). Storm surge
deposits in North Sea salt marshes dated by ‘34Csand I3’Cs determination.
Journal of Coastal Research, 9, 698-701.
Flood, J. (1993). Archaeology of the dreamtime. Sydney: Harper Collins.
Forbes, D.L., Taylor, R.B., Orford, J.D., Carter R.W.G. & Shaw, J. (1991). Gravel
barrier migration and overstepping. Marine Geology, 97, 305-1 3.
Foster, I.D.L., Dawson, A., Dawson, S., Lees, J.A. & Mansfield, L. (1993). Tsunami
sedimentation sequences in the Scilly Isles, southwest England. Science of
Tsunami Hazards, 1 1 , 3 5 4 6 .
Ginsburg, R.N. (1973). Evolving concepts in sedimentology, Studies in Geology, vol.
21, Johns Hopkins University. 191 pp.
Gomez, E.D. (1988). Overview of environmental problems in the East Asian Seas
region. Ambio, 17, 166-9.
Guilcher, A. (1958). Coastal and submarine morphology. New York: John Wiley and
Sons.
Holligan, P.M. & de Boois, H. ( 1 993). Land-ocean interactions in the coastal zone:
science plan. Stockholm: The International Geosphere-Biosphere Programme.
Huntley, D.A. & Short, A.D. (1992). On the spacing between observed rip currents.
Coastal Engineering, 17, 21 1-25.
Illenberger, W.K. & Rust, I.C. (1988). A sand budget for the Alexandria coastal
dunefield, South Africa. Sedimentology, 35,5 13-2 I .
Inman, D.L. (1978). The impact of coastal structures on shorelines. In Coastal zone ’78,
pp. 2265-7 1. New York: American Society of Civil Engineers.
IPCC Response Strategies Working Group (1992). Global climate change and the
rising challenge of the sea. Canberra: IPCC.
Jeftic, J.D., Milliman, J.D. & Sestini, G. (1992). Climatic change and the Mediterranean:
environmental and societal impacts of climatic change and sea-level rise in the
Mediterranean Region. Sevenoaks: Edward Arnold. 673 pp.
Jennings, S.C., Carter, R.W.G. & Orford, J.D. (1993). Late Holocene salt marsh
development under a regime of rapid sea-level rise: Chezzetcook Inlet, Nova
Scotia. Implications for the interpretation of palaeo-marsh sequences. Canadian
Journal of Earth Sciences, 30, 1374-84.
Johnson, D.W. (1919). Shore processes and shoreline development. New York:
Prentice Hall.
Johnson, D.W. (1925). The New England-Acadian shoreline. New York: Wiley. 608 pp.
Keen, T.R. & Slingerland, R.L. (1993). Four storm-event beds and the tropical
cyclones that produced them: a numerical hindcast. Journal of Sedimentary
Petrology, 63, 2 18-32.
King, C.A.M. (1972). Beaches and coasts, 2nd edn. London: Edward Arnold. 570 pp.
 Coastal evolution: a n introduction 29
Kraft, J.C., Kayan, I. & Erol, 0. (1980). Geomorphic reconstructions in the environs
of Ancient Troy. Science, 209, 776-82.
Kuhn, T.S. (1962). The structure of scientific revolutions. Chicago: University of
Chicago Press.
Lakhan, V.C. & Trenhaile, A S . , eds. (1989). Applications in coastal modeling.
Amsterdam: Elsevier. 387 pp.
Liu, K-B. & Fearn, M.L. (1993). Lake sediment record of late Holocene hurricane
activities from coastal Alabama. Geology, 21,793-6.
Long, D., Smith, D.E. & Dawson, A.G. (1989). A Holocene tsunami deposit in
eastern Scotland. Journal of Quaternary Science, 4,61-6.
McLachlan, A. & Erasmus, T. (1983). Sandy beaches as ecosystems. The Hague:
Junk. 757 pp.
Masters, P.M. & Flemming, N.C. (1983). Quaternary coastlines and marine
archaeology: towards the prehistory of land bridges and continental shelves.
La Jolla: Scripps Institute of Oceanography.
Milliman, J.D., Broadus, J.M. & Gable, F. (1989). Environmental and economic
implications of rising sea level and subsiding deltas: the Nile and Bengal
examples. Ambio, 18, 340-5.
Milliman, J.D. & Meade, R.H. (1983). Worldwide delivery of river sediments to the
oceans. Journal of Geology, 91, 1-21.
Milliman, J.D. & Syvitski, J.P.M. (1992). Geomorphic/tectonic control of sediment
discharge to the ocean: the importance of small mountainous rivers. Journal of
Geology, 100,525-44.
Minoura, K. & Nakaya, S. ( 1 99 1 ). Traces of tsunami preserved in inter-tidal lacustrine
and marsh deposits: some examples from northeast Japan. Journal of Geology,
99,265-87.
Miossec, A. (1991). Defense des cBres etprotection du littoral. UniversitC de Nantes.
Msangi, J.P., Griffiths, C.J. & Banyikwa, W.E. (1988). Man’s response to change in
the coastal zone of Tanzania. In The coastal zones: man’s response to change,
ed. K. Ruddle, W.B. Morgan & J.R. Pfafflin, pp. 37-60. Osaka, Japan:
National Museum of Ethnology.
Nichol, S.L. & Boyd, R. ( 1 993). Morphostratigraphy and facies architecture of sandy
barriers along the eastern shore of Nova Scotia. Marine Geology, 114, 59-80.
Orford, J.D., Carter, R.W.G. & Jennings, S.C. (1991). Coarse clastic barrier
environments: evolution and implications for Quaternary sea-level
interpretation. Quaternary International, 9, 87-1 04.
Orford, J.D., Carter, R.W.G. & McCloskey, J. (1993). A method of establishing
mesoscale (decadal to sub-decadal) domains in coastal barrier retreat rate from
tide-gauge analysis. In Large scale coastal behaviour ‘93,ed. J.H. List, pp.
155-8. US Geological Survey Open File Report 93-38 1.
Orford, J.D., Hinton, A.C., Carter, R.W.G. & Jennings, S.C. (1992). A tidal link
between sea-level rise and coastal response of a gravel-dominated barrier in
Nova Scotia. In Sea level changes: determination and effects. Geophysical
Monograph, 69,71-9.
Paw, J.N. & Thia-Eng, C. (1991). Climate change and sea level rise: implications on
coastal area utilisation and management in south-east Asia. Ocean & Shoreline
Management, 15,205-32.
Phillips, J.D. (1 992). Nonlinear dynamical systems in geomorphology: revolution or
evolution. Geomorphology, 5 , 2 19-29.
Pilkey, O.H., Young, R.S., Riggs, S.R., Smith, A.W., Wu, H. & Pilkey, W.D. (1993).
The concept of shoreface profile equilibrium: a critical review. Journal of
Coastal Research, 9, 255-78.
30 R. W.G. Carter & C.D. Woodroffe
Rampino, M.R. & Self, S. (1992). Volcanic winter and accelerated glaciation
following the Toba super-eruption. Nature, 359, 50-2.
Roy, P.S. (1984). New South Wales estuaries: their origin and evolution. In Coastal
geomorphology in Australia, ed. B.G. Thom, pp. 99-121. Sydney: Academic
Press.
Ruddle, K., Morgan, W.B. & Pfafflin, J.R. (1988). The coastal zones: man’s response
to change. Osaka, Japan: National Museum of Ethnology.
Russell, R.J. (1967). River plains and sea coasts. Berkeley: University of California
Press. 173 pp.
Schou, A. (1945). Det marine forland: geografiske studier over Danske
Fladkystlandskabers Dannelse og Formudvikling samt traek af disse Omraaders
Kulturgeograf, med saerlig Hensyntagen ti1 Sjaelland. Folia Geog. Danica.
vol. 4 , 2 3 6 ~ ~ .
Schumm, S.A. (1991). To interpret the earth: ten ways to be wrong. Cambridge
University Press.
Schumm, S.A. & Lichty, R.W. (1965). Time, space, and causality in geomorphology.
American Journal of Science, 263, 110-9.
Sestini, G., Jeftic, L. & Milliman, J.D. (1989). Implications of expected climatic
changes in the Mediterranean region: an overview. UNEP Regional Seas
Reports and Studies, Kenya, UNEP. 46 pp.
Short, A.D. (1991). Macro-meso tidal beach morphodynamics: an overview. Journal
of Coastal Research, 7,417-36.
Short, A.D. (1992). Beach systems of the central Netherlands coast: processes and
morphology and structural impacts in a storm-driven multi-bar system. Marine
Geology, 107, 103-37.
Stanley, D.J. & Warne, A.G. (1993a). Sea level and initiation of Predynastic culture in
the Nile delta. Nature, 363,435-8.
Stanley, D.J. & Warne, A.G. (1993b). Nile Delta: recent geological evolution and
human impact. Science, 260, 628-34.
Stea, R.R., Mott, R.J., Belknap, D.F. & Radtke, U. (1992). The pre-late Wisconsinan
chronology of Nova Scotia, Canada. Geological Society of America, Special
Paper, 270, 185-206.
Steers, J.A. (1946). The coastline of England and Wales, Cambridge University Press.
Stive, M.J.F., Roelvink, D.J.A. & de Vriend, H.J. (1991). Large-scale coastal
evolution concept. In Proceedings 22nd International Conference on Coastal
Engineering, pp. 1962-74. New York: American Society of Civil Engineers.
Terwindt, J.H.J. & Battjes, J.A. (1991). Research on large-scale coastal behaviour. In
Proceedings 22nd International Conference on Coastal Engineering, pp. 1975-83.
New York: American Society of Civil Engineers.
Thom, B.G. & Hall, W. (1991). Behaviour of beach profiles during accretion and
erosion dominated periods. Earth Surface Processes and Landforms, 16, 1 13-27.
Thom, B.G., Wright, L.D. & Coleman, J.M. (1975). Mangrove ecology and deltaic-
estuarine geomorphology, Cambridge Gulf-Ord River, Western Australia.
Journal of Ecology, 63, 203-22.
Turcotte, D.L. (1991). Earthquake prediction. Annual Review of Earth and Planetary
Sciences, 19, 263-8 1.
Vita-Finzi, C. (1964). Synchronous stream deposition throughout the Mediterranean
area in historical times. Nature, 202, 1324.
Watanabe, T. (1985). Comprehensive bibliography on tsunami of Japan. University
Press of the University of Tokyo. 260 pp. (In Japanese.)
Williams, S.F., Dodd, K. & Gohn, K.K. (1990). Coasts in crisis. U S . Geological
Survey Circular. 32 pp.
 Coastal evolution: an introduction 31
Woodroffe, C.D. (1993). Late Quaternary evolution of coastal and lowland riverine
plains of Southeast Asia and northern Australia: an overview. Sedimentary
Geology, 83, 163-75.
Woodroffe, C.D., Chappell, J., Thorn, B.G. & Wallensky, E. (1989). Depositional
model of a macrotidal estuary and floodplain, South Alligator River, Northern
Australia. Sedimentology, 36,737-56.
Woodroffe, C.D. & Grindrod, J. (1991). Mangrove biogeography: the role of
Quaternary environmental and sea level change. Journal of Biogeography, 18,
479-92.
Woodroffe, C.D., Mulrennan, M.E. & Chappell, J. (1993). Estuarine infill and coastal
progradation, southern van Diemen Gulf, northern Australia. Sedimentary
Geology, 83,257-75.
Woodroffe, C.D., Thom, B.G. & Chappell, J. (1985). Development of widespread
mangrove swamps in mid-Holocene times in northern Australia. Nature, 317,
71 1-13,
Wright, L.D. (1 987). Shelf-surfzone coupling: diabathic shoreface transport. Coastal
Sediments '87, 25-40.
Wright, L.D. (1993). Micromorphodynamics of the inner continental shelf a middle
Atlantic Bight case study. Journal of Coastal Research, SI 15, 93-124.
Wright, L.D., Boon, J.D., Kim, S.C. &List, J.H. (1991). Modes of cross-shore
sediment transport on the shoreface of the Middle Atlantic Bight. Marine
Geology, 96, 19-5 1.
Wright, L.D., Chappell, J., Thorn, B.G., Bradshaw, M.P. & Cowell, P. (1979).
Morphodynamics of reflective and dissipative beach and inshore systems:
southeastern Australia. Marine Geology, 32, 105-40.
Wright, L.D. & Coleman, J.M. (1973). Variations in morphology of major river deltas
as functions of ocean wave and river discharge regimes. American Association
of Petroleum Geologists Bulletin, 57, 370-98.
Wright, L.D., Coleman, J.M. & Thorn, B. G. (1973). Processes of channel
development of a high-tide range environment: Cambridge Gulf-Ord River
Delta, Western Australia, Journal of Geology, 81, 15-41.
Wright, L.D. & Short, A.D. (1984). Morphodynamic variability of surf zones and
beaches: a synthesis. Marine Geology, 56, 93-1 18.
Wright, L.D. & Thom, B.G. (1977). Coastal depositional landforms: a
morphodynamic approach. Progress in Physical Geography, 1,412-59.
Wright, L.D., Wiseman, W.J., Yang, Z-S., Bornhold, B.D., Keller, G.H., Prior, D.B.
& Suhayda, J.N. (1990). Processes of marine dispersal and deposition of
suspended silts off the mouth of the Huanghe (Yellow River). Continental Shelf
Research, 10, 1-40.
Young, R. & Pilkey, 0. H. (1992). A critical review of the engineering concept of
beach and shoreface profile of equilibrium. Geological Society ofAmerica,
Abstracts with Programs, 24, 74-5.
Young, R.W. & Bryant, E.A. (1992). Catastrophic wave erosion on the southeastern
coast of Australia: impact of the Lanai tsunami ca 105 ka ? Geology, 20,
199-202.
Young, R.W., Bryant, E.A., Price, D.M., Wirth, L.M. & Pease, M. (1993). Theoretical
constraints and chronological evidence of Holocene coastal development in
central and southern New South Wales, Australia. Geomorphology, 7,3 17-29.
Zeeman, E.C. ( I 977). Catastrophe theory. New York: Addison-Wesley.
Zenkovich, V.P. (1967). Processes ofcoastal development. Edinburgh: Oliver & Boyd.
 2
Morphodynamics of coastal evolution
P.J. COWELL A N D B.G. THOM

Introduction
Coastal evolution is the product of morphodynamic processes that occur in
response to changes in external conditions (Wright & Thom, 1977). Coastal
morphodynamics is defined as the ‘mutual adjustment of topography and fluid
dynamics involving sediment transport’ (Wright & Thom, 1977) or,
alternatively, the ‘dynamic behaviour of alluvial boundaries’ of fluid motions
(de Vriend, 199 1b). Sediment transport provides the time-dependent coupling
mechanism by which this adjustment occurs (Fig. 2.1). Fluid dynamics drive
sediment transport resulting in morphological change over time. Progressive
modification of topography in turn alters boundary conditions for the fluid
dynamics, which evolve to produce further changes in sediment-transport
patterns and their depositional products. Sediment properties and abundance
affect the process through their influence upon sediment transport and
sediment budgets respectively.
The essential properties of coastal morphodynamic processes are
attributable to the feedback loop between topography and the fluid dynamics
that drive sediment transport producing morphological change (Fig. 2.1). The
feedback can be either negative or positive. Negative feedback confers
properties of self regulation in response to minor perturbations (Wright &
Thom, 1977). Positive feedback signifies growth of an instability and confers
properties of self organisation, which results in new modes of operation
(Waldrop, 1992; Phillips, 1992). Feedback reversal marks thresholds in
morphodynamic behaviour.
A fuller appreciation now exists of the complexity inherent in these
morphodynamic processes (de Vriend, 199 1b) following recent developments
in non-linear dynamics (Gleick, 1988; Waldrop, 1992; Phillips, 1992). The
complexity derives from the morphodynamic feedback that is responsible for
state-determining behaviour or, to use the new language of chaos theory,

33
34 P.J. Cowell & B.G. Thorn

FLUID MOTIONS TOPOGRAPHY

(PLUS BIOPHYSICAL &
& GEOCHEMICAL
STRATIGRAPHY
PROCESSES)

SEDIMENT TRANSPORT

Figure 2.1. Primary components involved in coastal morphodynamics. The feedback

loop between form and process is responsible for fundamental complexity in coastal
evolution. At signifies the time dependence inherent in morphodynamic evolution of
coasts.

'sensitive dependence upon initial conditions'. This, together with the strongly
stochastic boundary conditions, gives rise to Markovian inheritance of
antecedent morphology by succeeding morphodynamic states. Such
inheritance is subject to additional memory effects associated with response-
time characteristics inherent in morphological change. These characteristics
are attributable to the finite rate of volume displacement involved in any
sediment transport that produces morphological change.
Evolution of coastal landforms is inherently non linear and time dependent
as a consequence of the combined effects of morphodynamic feedback and
Markovian behaviour. Details of coastal evolution are therefore unpredictable,
unrepeatable and irreversible, when subject to stochastic boundary conditions
over the long term (decades to millennia). By definition, coastal evolution is
also nonstationary in the statistical sense, making extrapolation based upon
geostatistical relationships unreliable, especially given the complicating
effects of non linearity stemming from the interconnectedness of processes
(Swift & Thorne, 1991; Schumm, 1991). This complexity underlies the
problem of large-scale coastal behaviour (LSCB) which signifies that coastal
evolution is the outcome of multiplicative probabilities and therefore largely a
question of historical accident (Terwindt & Battjes, 1991; Stive, Roelvink &
de Vriend, 1991; de Vriend, 1991a, 1992a,b). Application of classical,
 Morphodynamics of coastal evolution 35
reductionist methods of prediction and explanation based upon separate
analysis of component processes is of limited use under these circumstances
(Schumm, 1991). Nevertheless, new techniques are emerging for prediction
and explanation of LSCB based upon inverse modelling principles that allow
both a measure of quantitative rigour and the incorporation of uncodified
qualitative expertise (see page 64). These models provide a general
understanding and explanation of coastal evolution. Future advances in the
mathematics of non-linear dynamics may also yield ways of undertaking more
conventionalforward modelling predictions for LSCB (de Vriend, 1991b).
The ‘chicken-and-egg’ nature of the mutual interaction between coastal
topography and processes, as articulated formally by Wright & Thom (1977),
involves changes to coastal landforms over a broad range of time and space
scales. These dimensions tend to vary together, due to scale-related variation in
response times. Scales at which morphodynamic processes operate are
grouped into four classes in Fig. 2.2 based upon considerations similar to those
presented by Stive et al. (1991) and de Vriend (1991b). ‘Instantaneous’ time

Time
Scale
millennia

centuries

decades

years

seasons

days

hours

seconds

0.1 1 10 100

Length Scale (km)

Figure 2.2. Definition of spatial and temporal scales involved in coastal evolution, with
typical classes of sedimentary features used as illustrations throughout the text.
36 P.J. Cowell & B.G. Thom
scales involve the evolution of morphology during a single cycle of primary
forcing agents, such as waves or tides. The coupling between fluid flows and
morphology relates directly to physical processes causing sediment transport.
Morphodynamic adjustment can be followed in detail as a real-time process.
Coastal evolution at event scales is a response to processes operating across
time spans ranging from that of a single event, such as a storm, through to
seasonal variations in environmental conditions. Morphological changes
comprise time-averaged effects of instantaneous processes during a single
fluctuation in boundary conditions. Depositional products correspond to event
beds, which form the fundamental sedimentary units in lithofacies and
depositional sequences (Swift, Phillips & Thorne, 1991a). Engineering (and
historical) time scales involve composite evolution over many fluctuations in
boundary conditions, each of which entails many cycles in the fundamental
processes responsible for sediment transport. Evolutionary outcomes depend
upon both the nature and sequencing of fluctuations in boundary conditions.
The dependence on individual fluctuations becomes less important toward
geological scales where, over millennia, evolution occurs more in response to
mean trends in environmental conditions. At geological scales longer than
millennia, fluctuations in boundary conditions again become significant,
primarily driven by the effects of Milankovich cycles (Hays, Imbrie &
Shackleton, 1976). The successively larger length scales in Fig. 2.2 represent
the spatial counterpart of the time-related variability. The smallest scales range
from grain dimensions to bedform lengths (which even so, may be in the order
of lo2 metres when associated with low-frequency water motions such as
tides). At these scales, spatial uniformity or simple, systematic variation (e.g.,
over the length of a single bedform) allows localised consideration of
processes. In contrast, the successively larger length scales in Fig. 2.2 are
associated with an increasing degree of spatial variation (and possibly non
homogeneity). Larger morphologies may incorporate a more extensive
hierarchy of subordinate forms, signifying a greater level of spatial complexity
in related geomorphic process (Schumm, 1991).
Morphodynamic studies during the past two decades have been concerned
predominantly with instantaneous and event scales. However, the focus of this
chapter is on coastal evolution over engineering and geological time spans.
Such a focus is not only in keeping with the themes throughout this book; it
also reflects the emphasis given at present to the study of large-scale coastal
behaviour. The purpose of this chapter is to: (a) explore the essential properties
of coastal morphodynamics, especially in relation to coastal evolution at
engineering through to geological time scales, and (b) suggest an approach to
morphodynamic modelling at these time scales. The chapter is divided into
 Morphodynamics of coastal evolution 37

four sections dealing with (i) the organisation of morphodynamic systems,

(ii) the properties of morphodynamic processes; (iii) the cumulative nature of
coastal evolution attributable to morphodynamic behaviour; and (iv) the nature
of LSCB and how it can be dealt with in modelling coastal evolution.

Morphodynamic systems: definitions

The purpose of this section is to outline the conceptual basis for understanding
the complexity in coastal morphodynamics. The organisational properties and
vernacular of general systems theory are employed to define the relationships
between constituent processes. These relationships involve formal
representation of the structural and functional logic of interconnections
between processes (von Bertalanffy, 1968; Bennett & Chorley, 1978; Phillips,
1992). This approach allows for a manageable reduction of the problem into
self-contained elements that are simple enough to analyse in isolation, whilst
maintaining the context for each of the elements as well as their significance to
the overall problem. Moreover, it provides a necessary methodology when
interactions between component processes are not weak enough to be
neglected and when relations describing behaviour of the components are not
linear (von Bertalanffy, 1968). An overview of the general application of
systems dynamics to coastal modelling is given by Lakhan & Trenhaile (1989).

Boundary conditions
The most fundamental step in defining a morphodynamic system is to identify
the spatial and process boundaries of the system.

Spatial boundaries
Spatial boundaries are defined in Fig. 2.3. They most appropriately conform,
for the purposes of studying coastal evolution, to the conventions set out by
Inman & Brush (1973). These boundaries correspond to the limits to which
coastal processes have extended during the Quaternary due to the large sea-
level fluctuations, ranging over more than 100 metres vertically, associated
with expansion and contraction of Quaternary ice sheets (Hays et al., 1976;
Chappell, et al., 1982; Shackleton, 1987).
The landward limit encompasses the coastal depositional and marine-
erosion surfaces formed during the interglacial highstands of the present and
Quaternary sea levels. The seaward limit is defined by the edge of the
continental shelf, which typically occurs in water depths from 100 to 200
metres (Inman & Brush, 1973). The reason that the coastal zone is so much
38 P.J. Cowell & B.G. Thorn

Figure 2.3. Spatial boundaries of the coast (adapted from Inman & Brush, 1973).
h represents water depth; -h represents elevation of land.

wider than the present-day littoral zone is twofold. First, the coastal water mass
is usually distinct from the ocean water mass in terms of its physical and
dynamic properties (Johnson & Rockcliff, 1986; GESAMP, 199 1); ocean
flows typically have vertical length scales an order of magnitude greater than
coastal waters so that, for ocean flows, the continental slope effectively is the
coast (e.g., Bane, 1983; Johnson & Rockliff, 1986; Brink, 1987; Blanton et al.,
1987), although surface manifestations of ocean currents can intrude into the
coastal ocean (Cresswell, et al., 1983; Blanton et al., 1987; Huyer et al., 1988).
Second, the lowest sea levels probably placed littoral processes close to the
continental-shelf edge in many parts of the world on several occasions during
the Quaternary (Chappell et al., 1982; Shackleton, 1987). Moreover, shorelines
have swept across mid- and inner-continental shelf zones many times
throughout most of this period. The evolutionary sequences for large-scale
coastal deposits, such as barrier and deltaic plains, have histories covering such
time spans ( lo3 years). The early development of these features at lower sea
levels has, in part, determined present-day morphology through the Markovian
inheritance that is characteristic of morphodynamic processes.

Process boundaries
Process boundaries for the coastal system correspond to the environmental
conditions in Fig. 2.4. Exogenous inputs from the environment are responsible
for the geographic variation among coasts. The environment comprises
 Morphodynamics of coastal evolution 39

ATMOSPHERE 8 OCEAN FORCING' FRESH WATER FROM LAND

COASTAL SYSTEM
COASTAL FLOWS
WAVES, CURRENTS a WINDS
ENERGY
TOPOGRAPHY
LOSSES

BOUNDARY LAYER
FLOWS
TERRIGENOUS
SEDIMENTS

SEDIMENT TRANSPORT

SEDIMENT
LOSS STRATIGRAPHY

SEDIMENT BALANCE

AUTOGENIC
SEDIMENT
GAINS 8 EROSIONlDEPOSITION
LOSSES

ENVIRONMENTAL CONDITIONS

Figure 2.4. Structure and function of the morphodynamic model for the coastal system.
The dashed arrows represent input-output between the coastal system and the
environment. Sediment transport is ,highlighted because it provides the coupling
mechanism between fluid dynamics and morphological change in the direction of
forcing (feedforward); it thereby exerts frequency-response modulation.

climatic and geological controls (Davies, 1980). The geological controls

govern the physiographic setting as well as the abundance and properties of
sediments. The regional climates of the ocean and atmosphere determine the
energy regime. The energy regime is responsible for introducing uncertainty
into morphodynamic processes, and therefore coastal evolution, because it is
stochastic.
Tectonic setting and denudational history of continental margins
throughout the Cainozoic are responsible for the primary geological controls
(Inman & Nordstrom, 197 1) which determine continental-shelf dimensions
and terrestrial sediment supply (Davies, 1980). Over Quaternary time scales,
Milankovich cycles (Hays et al., 1976) drove interaction between climate and
geology causing glacio- and hydroisostatic adjustments that are responsible for
the variations in relative sea level; a boundary condition of fundamental
importance in coastal evolution from geological down to engineering time
spans (Roy & Thom, 1987; Thom & Roy, 1988). Process boundary conditions
dominating at time scales typically less than one year include oceanic and
40 P.J. Cowell & B.G. Thom
atmospheric inputs, although the importance of inter-annual and inter-decadal
variations, such as the El Nifio Southern Oscillation (ENSO) (Deser &
Wallace, 1987), are becoming increasingly evident (Bryant, 1983; Komar &
Enfield, 1987; Braatz & Aubrey, 1987; Thom & Hall, 1991). Ocean forcing
results from: (i) direct and indirect intrusion of ocean-boundary currents and
eddies (Curray, 1960; Johnson & Rockliff, 1986; Blanton, et al., 1987;
Flemming 1988; Lee et al., 1989); (ii) astronomical and meteorological tides
in internal and barotropic modes (Baines, 1986; Brink, 1987; Pugh, 1987;
Blanton et al., 1987; Griffin & Middleton, 1992); (iii) swell waves (Inman &
Brush, 1973; Davies, 1980; Wright, 1976); and (iv) other sources of variation
in the ocean water level (Komar & Enfield, 1987; Braatz & Aubrey, 1987).
Winds transport sediment in coastal dunes (Hsu, 1987; Nordstrom, Psuty &
Carter, 1990). They are also responsible in the coastal ocean for generation of
wind waves (Bretshneider, 1966; Davies, 1980; Bishop & Donelan, 1989), and
contribute to Eckman currents and internal flows (Csanady, 1977, 1982; Swift
et al., 1985; Atkinson et al., 1989; Lee et al., 1989; Wright et al., 1991; Griffin
& Middleton, 1992), as well as coastally trapped waves (Hamon, 1966;
Huthnance, Mysak & Wang, 1986; Griffin & Middleton, 1992).
External inputs of fresh water as run-off from the land (Fig.2.4) are
estimated globally at 1.1X 10' m3s- I , carrying an estimated sediment load of
530 t s-' (Inman & Brush, 1973). These sediment inputs are supplemented by
autogenic (mainly biogenic) production which yield peats and other organic
accumulations in protected areas (Davies, 1980; Frey & Basan, 1985; Nichols
& Biggs, 1985), with carbonate sediment supply predominating in arid and
tropical environments (Davies, 1986).
Sediment losses from the coastal system in Fig. 2.4 mainly comprise fine
material which is transported over the shelf break, out onto the abyssal plains
fronting continental margins, especially during glacial low sea levels (Thorne
& Swift, 1991a). Energy output (losses) mostly involves heat, being the
ultimate product of dissipation in fluid and sedimentary processes, although
some reradiation of energy back into the open ocean can be expected,
particularly for lower-frequency water motions such as the tides. The global
rate of energy dissipation in the coastal ocean is estimated to total 5 X 1O9 kW,
or 0.17 W m-' of continental-shelf surface area (Inman & Brush, 1973).

Organisation of morphodynamic systems

The process linkages, shown by the arrows in Fig. 2.4, chart the input-output
relationships between each of the sub systems in the morphodynamic model.
These relationships show that, to know about morphological change, it is
 Morphodynamics of coastal evolution 41
necessary to know the sediment budgets during the time period of interest
throughout the region. To know about sediment budgets, a model is required
for sediment dynamics, and another for the near-bed boundary-layer flows
which entrain and carry the sediments. In turn, the flows driving the boundary
layer must be understood. Practical theory for each of these components is
required before morphodynamic modelling can be undertaken. Identification
of which flows are important, and how they might be modelled best, has
occupied most of the effort in coastal-process studies over the past two decades
(eg, Wright, 1987, 1993; Wright et al., 1991) and is by no means complete
(de Vriend, 1991b). The development of principles for integrated, time-
dependent modelling of morphodynamics has occurred only recently. These
principles, and their numerical implementation employing constituent models
for each of the above processes, are reviewed by de Vriend (1991a) and
GESAMP ( 1 99 I).
Sediment-transport processes
These are highlighted in Fig. 2.4 because of their importance in providing the
primary coupling mechanism between topography and fluid dynamics.
However, it is the spatial budget, resulting from sediment-transport patterns,
that causes erosion and deposition. This is described exactly by the continuity
equation for sediment transport,

in which the h is the bed elevation (h = - 2 ) at each point ( x , y ) on the

topographic surface in the coordinate system defined in Fig.2.3, and E
accounts for sediment density and porosity. Each of the terms inside the
brackets is depth integrated; Cis the total concentration of sediment suspended
above the bed, qX and sYare the sediment-mass fluxes normal and parallel to
the shore respectively, and V (per unit surface area) represents autogenic
sediment production and losses, as well as engineering interventions such as
beach nourishment and sand mining. (A term like Valso is required to include
the effects of subsidence and isostatic or tectonic movements.) Generally, q,r
and qY are the important terms, except under special conditions such as during
the formation of tempestites (Niedoroda, Swift & Hopkins, 1985) when aC/&
is critical, or in low-energy and biologically rich environments where aV/&
dominates (Frey & Basan, 1985; Nichols & Biggs, 1985; Davies, 1986).

Topography and stratig raphv

A primary objective of coastal morphodynamics is to predict coastal
topography at event time scales, and its evolution over engineering and
42 P.J. Cowell & B.G. Thom
geological time scales. Topography is therefore highlighted in Fig. 2.4, along
with stratigraphy. Stratigraphy is the integrated effect of topographic change
through time in depositional environments. Stratigraphic sequences therefore
become a partial record of coastal evolution (Kraft & Chrzastowski, 1985;
Nummedal & Swift, 1987; Haq, Hardenbol & Vail, 1987; Boyd, Suter &
Penland, 1989; Swift et al., 1991a; Thorne & Swift, 1991a, Vail et al., 1991).
Topography itself represents a time integration of morphological changes, as
inferred by the time loop, At in Fig.2.4. Therefore, coastal evolution is
expressed as

h = ho + dt

starting with an initial topography ho. Topography controls coastal evolution in

two ways. First, it governs boundary conditions for the fluid dynamics.
Second, it determines the accommodation space available for deposition in
relation to the rate at which sediments are supplied to the coastal cell from
fluvial and biogenic sources, or from along the coast (Thorne & Swift, 1991a).

Morphodynamic systems - general concepts

The primary process-response linkages in Fig. 2.4 follow the forcing sequence
from the environmental inputs, through the fluid and sediment dynamics, to
drive morphological evolution. However, the opportunity for morphodynamic
adjustment stems from the all-important feedback loop between topography
and fluid dynamics (Wright & Thom, 1977; Hardisty, 1987).

Properties of morphodynamic processes

Topographic feedback provides the essence of morphodynamic behaviour and
it arises through the effects of sediment transport (Fig. 2.1). Feedback is
responsible for making evolution of coastal morphology partially self
determining, since the operation of the system, and its state at any instant, is
affected by the antecedent topography; ie, there is ‘state dependence’ (Bennett
& Chorley, 1978). The state of the system is defined by the values of all
internal (state) variables at an instant. It is the inheritance by morphodynamic
processes of their own past that is responsible for the so-called ‘sensitivity to
initial conditions’ (Phillips, 1992).
The essential properties of morphodynamic processes comprise: (i) self
regulation (equilibrium tendencies); (ii) selfforcing (which leads to thresholds,
self organisation and regime changes); (iii) Murkovian inheritance (which
introduces uncertainty - Bennett & Chorley, 1978, p. 322); (iv) hysteresis
 Morphodynamics of coastal evolution 43

(which causes a filtered response by morphology to changes in boundary

conditions); (v) non linearity; (vi) nonstationarity; and (vii) nonhomogeneity.
Feedback is responsible for each of these, except for hysteresis and
nonhomogeneity. However, it is self regulation and self organisation,
attributable to state-determined feedback, that bear the hallmark of morpho-
dynamic processes.
Because of these properties, complexity is inherent in coastal evolution.
The non linearity involves interdependence of morphodynamic states through
space and time (i.e., they are ‘stateldetermined’). The spatial interdependence
is complicated further by nonhomogeneity, which generally exists in
morphodynamics since the coast is a transition zone between the land and deep
ocean. Because of time dependence, evolutionary sequences feature
cumulative rather than uncorrelated, random changes, which can result in
strong nonstationarity. This compounds the filtering effects of hysteresis
which occur due to the finite time required for the redistribution of sediment
volumes in response to changing environmental inputs.

Feedback mechanisms
Morphodynamic feedback fundamentally occurs through the control that the
solid boundary (i.e., the topographic surface) exerts over flows, across all
scales from continental-shelf circulations (e.g., Csanady, 1982; Cannon &
Lagerloef, 1983; Pettigrew & Murray, 1986) to shoaling wind waves (e.g.,
Svendsen & Buhr Hansen, 1978; Cowell, 1982; van Rijn, 1990). This control
mechanism involves (i) the continuity principle for fluids; and (ii) the bottom-
boundary condition of no flow at the bed. (The latter is not to be confused with
the spatial and system boundary conditions.)
The fluid continuity principle is a statement that flows over an irregular
boundary must experience spatial changes in their velocities, pressures, and/or
densities in order to conserve their fluid mass. The continuity principle for
fluids is expressed as

(2.3)

where p is the fluid density, and u, v, w are the velocity components in the
x , y , z directions respectively (see Fig. 2.3).
The bottom-boundary condition (at z = -h) stipulates that the velocity
vector V = 0. This condition arises because there can be no effective flow
through the solid boundary for velocity components (w)perpendicular to it,
whilst friction causes tangential components (u, v) to vanish at the boundary.
The boundary-normal condition (w= 0) prevails to preserve fluid continuity in
44 P.J. Cowell & B.G. Thom
the absence of cavitation over a bed that is impervious (or only weakly
infiltrated). The tangential condition (u = v = 0 ) must exist to avoid shear at
the wall becoming infinite in viscous fluids. The wall friction also provides an
opposing force acting against pressure and inertia in flows. It affects the
overall hydrodynamic behaviour through the momentum balance. This is
expressed by the Navier-Stokes equation (in vector form), as

p - = grZd - + (ciirl v) X V 2 = -griTd(p + p g z ) + p o b (2.4)

at
where p is pressure, g is gravity, and p is viscosity. The topographic
roughness, upon which the strength of the friction force depends, therefore
provides an additional feedback mechanism via the boundary layer (Fig. 2.4)
(Csanady, 1982; Pettigrew & Murray, 1986). Apart from the role friction plays
in the momentum balance (far-right term in eqn. (2.3)), it is responsible for
cumulative dissipation of energy from flows. Friction is also enhanced as a
result of loose-bed roughness (e.g., Grant & Madsen, 1982). This contributes
to the feedback between sediment dynamics and boundary layer flows (not
shown in Fig. 2.4).

Negative feedback and equilibrium

Negative feedback is a damping mechanism that acts against departures from a
morphodynamic state. It is responsible for self regulation and is therefore a
stabilising process. Self regulation applies to the behaviour of a system in
response to small perturbations. It is the mechanism by which equilibrium is
re-established (Wright & Thom, 1977). Although equilibrium tendencies are a
manifestation of self regulation, morphological hysteresis, together with the
stochastic nature of boundary conditions, can prevent steady-state conditions
from being attained. Therefore, morphodynamic adjustment is modulated by
frequency-response characteristics which reflect the way changes in sediment
masses lag behind variations in environmental inputs (see page 58).
An example of self regulation due to negative feedback in morphodynamic
processes, operating over time scales of minutes to hours, is provided by the
interaction between edge waves and beach cusps. Subharmonic edge-wave
amplitudes are progressively suppressed by the growth in topographic relief of
beach cusps that the edge waves themselves produce (Guza & Bowen, 1981).
The plan-form adjustment of shorelines to changes in wave direction and
period is an example of self regulation occurring over time spans of hours to
days (Davies, 1958; Komar, 1973, 1976). Self regulation also occurs at
engineering and geological time scales. Examples include maintenance of the
equilibrium profile on the upper shoreface over time (Dean, 1991; Kotvojs &
 Morphodynamics of coastal evolution 45
Cowell, 199l), despite episodic perturbations due to storms (Vellinga, 1986),
and the infill of the Waddensea (Netherlands), in which the rate of deposition
appears to be in dynamic equilibrium with the local sea-level rise (Eysink,
1991).
Equilibrium is the product of negative feedback and represents a
morphodynamic state that is stable for a given range of environmental
conditions. It is now recognised that equilibrium can take three forms: (i)
steady state; (ii) periodic; and (iii) ‘chaotic’ (Phillips, 1992). Steady-state
equilibrium in coastal evolution can either mean that sediment flux averages to
zero through time, or that there is no gradient in the sediment flux vector, in
which case the topography acts as a sediment pathway rather than a source or
sink. Therefore, equilibrium may be expressed in terms of sand-mass
continuity (eqn. (2.1)) as

(2.5)

The spatial terms in eqn. (2.5) define source and sink zones, such that

As Wright et al. (1991) point out, whether or not equilibrium exists depends
upon the length of time over which eqns. (2.5) or (2.6) are integrated (i.e., it is
a question of time scale). For example, net cross-shore transport of sediments
during a single storm can be cancelled out by successive storms (Vincent,
Young, & Swift, 1981). Moreover, morphodynamic equilibrium is unlikely at
any instant because of the stochastic nature of boundary conditions and finite
morphological response times. Nevertheless, we can assume from the physical
imperative expressed in eqn. (2.5), that coastal evolution is driven by a
disequilibrium-stress in the continuity of sediment dynamics.
There are two approaches to analysis of equilibrium, entailing (i) real-time
morphodynamics and (ii) regime morphodynamics. The first of these involves
dynamic modelling of fluid-flows, sediment-transport fields and their co-
evolution with the morphology through time until steady state is attained.
Because of the complexities involved, especially in the sediment dynamics
(Huntley & Bowen, 1990), simulations have traditionally employed physical
modelling (e.g., Dalrymple, 1985; Sunamura, 1989), but increasingly
computer-based numerical studies are being undertaken into morphodynamic
evolution in shallow water (de Vriend, 1987c), in the surf zone and on the
shoreface (Coeffe & Pechon, 1982; Huthnance, 1982; Stive, 1987; Boczar-
Karakiewicz & Davidson-Arnott, 1987; Steijn, et al., 1989), as well as on the
46 P.J. Cowell & B.G. Thom
inner shelf (Boczar-Karakiewicz & Bona, 1986; Boczar-Karakiewicz, Bona &
Cohen, 1987; de Vriend, 1987b, 1990; Boczar-Karakiewicz, Bona & Pelchat,
1991), and in estuaries (de Vriend, 1987a; de Vriend et al., 1989). A dynamic
steady-state occurs within this approach when net sediment transport becomes
either spatially uniform or zero. Under these conditions, the flow field
becomes stationary in the statistical sense (i.e., time-average velocities remain
constant from one time interval to the next), and time-average morphological
change ceases. However, geographic variation in hydrodynamic conditions
and geologically inherited effects produce site-specific differences in
equilibrium form (Pilkey et al., 1993).
Regime morphodynamics sets out conditions necessary for equilibrium in
terms of a bulk representation of processes. Thorne & Swift (1991b) elaborate
the approach in relation to the ‘regime’ concept in hydraulics and the ‘grade’
concept in geomorphology. Equilibrium morphology is determined directly
without consideration of initial or intermediate states and employs related
surrogate quantities for sediment transport. A clear example of this approach is
the model for shoreline equilibrium (Davies, 1958; Komar, 1973), which
involves determination of shoreline alignments that cause a balance in wave-
driven alongshore sediment transport or, in the case of deltaic shorelines, a
balance between alongshore transport and the discharge rate of sediments at
the river mouth (Komar, 1973). Regime equilibrium models exist for beach and
shoreface profiles based upon force balance (Bowen, 1980; Bailard & Inman,
1981) and wave-energy dissipation rates (Bruun, 1954; Dean, 1977, 1991;
Chappell, 1981). In reality, these formulations are guided by physical insights
that are generally incomplete (Pilkey et al., 1993) and rely heavily upon
empirical tuning (Kotvojs & Cowell, 1991). The result is therefore little
different from relying on inductive insights to formulate equilibrium
relationships directly from an empirical basis, as in the case of dimensional
scalings available for the morphologies of tidal inlets (O’Brien, 1969; Bruun,
1978) and tidal basins (e.g., Eysink, 1991).

Positive feedback, thresholds and self organization

Positive feedback produces self-forcing morphodynamic behaviour which is
necessarily of limited duration and unstable (Wright & Thom, 1977). Positive
feedback is illustrated by the infilling of deep estuaries with marine sediments
when a positive tidal asymmetry exists. The rate of infilling grows as the
estuary becomes shallower since the tidal asymmetry also increases as the
friction and shoaling effects are enhanced by the reduced water depths
(Dronkers, 1986). Eventually, the infilling of the estuarine basin produces the
development of tidal flats, as exemplified in the formation of the ‘Big Swamp’
 Morphodynamics of coastal evolution 47
phase of the Holocene history of the South Alligator River (see Fig. 1.5). This
marks a reversal in feedback from positive to negative. As the tidal flats
become more extensive, and a larger volume of the tidal flow occurs across
them, the positive asymmetry of the tide progressively decreases so that the
estuarine morphology approaches steady state as sediment imports and exports
equilibrate (Chappell & Thom, 1986; Woodroffe et al., 1989). The relatively
brief duration of the ‘Big Swamp’ phase (22% of the stillstand period) is
indicative of the transient nature of morphodynamic processes governed by
positive feedback. The reversal in feedback in this case itself constitutes a
threshold (Wright & Thom, 1977).
Thresholds occur when changing environmental inputs, or changes in the
mechanical strength of sediment masses, drive morphodynamic state variables
toward limiting values for the operational range of existing self-regulation
mechanisms (Schumm, 1979; Roberts, Suhayda & Coleman, 1980). Beyond
these critical values, transformation of state occurs such that a new set of
processes operates with a new set of state variables. Examples of coastal-
process thresholds at instantaneous scales include the often-used illustration of
breaking-wave transformations (e.g., Thom, 1975; Chappell, 1983), the
transitions from rippled-bed to flat-bed sand transport under water at a critical
shear stress (Nielsen, 1992) and the onset of sediment motion at a critical shear
velocity in the sediment-dynamics subsystem (Carstens, Nielsen & Altinbilek,
1969; Nielsen, 1990).
At larger (event) scales, changes in beach state between reflective and fully
dissipative extremes, or a variety of intermediate states, each with its own
characteristic bar morphology and surf-zone hydrodynamics, represents a
discontinuous range of morphodynamic processes operating over a continuous
range of input wave parameters at a given site (e.g., Wright et al., 1979; Wright
& Short, 1984; Wright, Short & Green, 1985). For example, it is well known
that barred intermediate and dissipative beaches respond to rising wave
conditions through an increasing width of the surf zone associated with a
shifting of bars offshore (Shepard, 1950; Bascom 1954; Sunamura, 1989).
However, reflective beaches accommodate such conditions through an
increase in height of the beach step so that waves can propagate to the
shoreline without breaking (Hughes & Cowell, 1986). The transition between
the two modes occurs at a threshold wave height when waves begin to break
seaward of the beach. Other examples of thresholds at event scales include the
breaching of barriers by tidal-inlet formation during extreme storms (Coates &
Vitek, 1980), lateral switching of over-extended deltaic distributary channels,
triggered by floods when hydraulic gradients become too low, and various
types of mass movement, including diapiric intrusions (Roberts et al., 1980).
Examples of state thresholds at engineering to geological time scales include
48 P.J. Cowell & B.G. Thorn
transformations between barrier, tidal-delta and innershelf sand bodies,
especially under conditions of rising sea level (Swift, Phillips & Thorne,
1991b; Roy et al., 1992), and the alternation between mud progradation and
coarse, carbonate chenier-ridge deposition on mangrove coasts in tropical
Australia, where threshold conditions involve the occurrence of major storms
with return intervals exceeding that required for critical levels of shell
production on the shoreface (Rhodes, 1982; Chappell & Grindrod, 1984).
Where input variables (environmental conditions in Fig. 2.4) are non-
stationary, as in the case of climate and sea-level change, morphodynamic
processes may go beyond their stable range. This is exemplified in coastal
evolution by barriers overstepped during rising sea levels. Under certain
circumstances, a critical point is reached at which the landward, upslope
migration of a barrier fails to keep pace with the rate of sea-level rise (Penland,
Suter & Boyd, 1985). The process, shown in Fig. 2.5, has been observed on the
Nova Scotian coast (see Chapters 1, 4 and 10) where rising sea level caused

a
MSL at step 10

remnant core of
over-stepped barrier stranded on inner shelf
Figure 2.5. Schematic representation of barrier translation in response to a 10 metre sea-
level rise, in increments of l metre, which corresponds to c. 1500 years for the last post-
glacial marine transgression. Two modes of coastal evolution are shown: ( a )continuous
roll-over, and (b)the occurrence of a threshold resulting in barrier overstepping. MSL
represents mean sea level.
 Morphodynamics of coastal evolution 49
overstepping and remobilisation of a barrier superstructure landwards (Forbes
et al., 1991; Orford, Carter & Forbes, 1991). A sudden landward shift in barrier
position occurred and the barrier core was left offshore to form a shoreface
ridge in the manner proposed (contentiously) for such ridges on the US New
Jersey coast (Stubblefield, McGrail & Kersey, 1984; Swift, McKinney and
Stahl, 1984; Rine et al., 1991).
Self organisation, which is the subject of recent developments in non-linear
dynamics, is driven initially by positive feedback (Waldrop, 1992; Phillips,
1992). It involves the evolution of morphology toward a new equilibrium
regime following creation of a new system or an extensive shift in process
boundary conditions. Large glacio-eustatic sea-level changes and major
engineering interventions, for example, often have these consequences (at
geological and engineering scales respectively). Self organisation is, however,
an inevitable consequence of morphodynamic feedback at all scales. It has
been simulated recently at instantaneous scales with the spontaneous
development of beach cusps (Werner & Fink, 1993). Emergence of cusps as a
global pattern on the beach face resulted from morphodynamic feedback
involving the interaction of swash, sediment transport and time-dependent
topographic changes on the beach face. The simulated feedback initially was
positive, causing growth in small topographic irregularities, but then switched
spontaneously to negative feedback as cusps became well developed,
providing a stabilising mechanism for cusp survival (Werner & Fink, 1993).
Self organisation at event scales is exemplified in nature by the reworking of
dune-sand masses into beach profiles of emergent equilibrium form during
storm surges (Wang, 1985; Vellinga, 1986). A heuristic example at
engineering time scales (months to years) is provided by the partial, artificial
closure of the Rhine/Meuse and Eastern Scheldt estuaries in The Netherlands
(Steijn et al., 1989; de Vriend et al., 1989), where longitudinal, shore-normal
ebb-tide shoals were reorientated to form sand bars with shore-parallel
alignment due to the increased importance of waves among the constituents in
the shore-normal velocity field. These changes involved the self organisation
of antecedent, estuarine-tidal dominated morphologies into shoreface
morphologies tending toward dynamic equilibrium with waves.
Self organisation at geological time scales is illustrated in Fig. 2.6 by the
Holocene evolution of broad estuarine sediment sinks into channelised
pathways due to progressive infilling involving morphodynamic adjustment to
stillstand conditions (Roy, Thom & Wright, 1980). The infilling results (in
accordance with eqn. (2.6)) from sharp, negative gradients in transport fluxes
for marine and fluvial sediments entering the estuary. The transport gradients
are associated with expansion of tidal and stream flows within the estuarine
50 P.J. Cowell & B.G. Thorn

coastal washover. estuarine fluvial sand, flood plain silts,

barrier tidal delta mud muddy sand swamp deposits
sand sand

Figure 2.6. Evolution of bay-barrier estuaries on the south-east Australian coast

involves self organisation in which progressive infilling over the Holocene ultimately
leads to sediment bypassing corresponding to a steady state in the estuarine sediment
volume (from Roy, 1984).
 Morphodynamics of coastal evolution 51
basin. The scale of the expansion diminishes as the basin fills until transport
gradients become zero and sediment bypassing is established, signifying
overall steady state conditions. The evolutionary sequence is similar to that
described for South Alligator River, despite vast differences in environment as
well as detail. Both examples illustrate how self-organising behaviour
involves a sequence of positive feedback, which drives morphodynamic
change toward a new state, followed by negative-feedback processes, which
stabilise the sedimentation regime.

Cumulative evolution and state dependence

Coastal evolution is a cumulative process because the morphological outputs
are included amongst the inputs for the next cycle of change (Fig. 2.1). This
section outlines the principles involved in cumulative, state-dependent
evolution, which fundamentally entail non linearity in the process-form
interaction (Fig. 2. l), Markovian inheritance and response-time effects.
Although cumulative evolution operates at all scales, its significance in the
development of major depositional landforms at longer, geological scales is
paramount. An example of such state-dependent evolution in the Quaternary
history of the south-east Australian coast is given in Fig. 2.7 with respect to the

Figure 2.7. Generalised stratigraphic section of the Newcastle Bight embayment,

100 km north of Sydney on the south-east coast of Australia, based upon core logs and
radiocarbon dates. The inner and outer barriers are of Last-Interglacial and Holocene
ages respectively. (After Thom et al., 1992.)
52 P.J. Cowell & B.G. Thom
120 ka and Holocene high sea levels and the post-glacial marine transgressions
that preceded them. These episodes both caused deposition of transgressive
and regressive sediment sequences, each of which presents a different
substrate to their successors. The stratigraphy shows that the earliest deposits
occur over substrates which are broader and shallower (see Chapter 4).
Shorefaces associated with these deposits were also more protected from wave
energy since they generally lie within the landward recesses of embayments on
this coast. Later shoreface sediments were deposited under altered
morphodynamic conditions since substrates became progressively steeper,
deeper and more exposed. The geometric changes produced an increasing
accommodation volume per unit width of shoreface as high-stand shorelines
were relocated progressively further seaward (Fig. 2.7). Recent thermo-
luminescence dating of deposits in another embayment, 100 km further north,
indicates that there may have been at least four barrier sequences with ages of
240 ka, 120 ka, 90 ka and Holocene (Roy er al., 1992). Each sequence remains
largely intact, each represents intermediate states in the evolution of the coast,
and each has presented a different substrate during successive transgressions
(Fig. 4.12).

Non linearity
Non linearity is an inevitable consequence of the feedback effects (Phillips,
1992) that make coastal evolution cumulative (Figs. 2.1 and 2.4). Non linearity
‘means that the act of playing the game has a way of changing the rules’
(Gleick, 1988, p. 24). The evolving morphological outputs (eqn. (2.1)), which
comprise the boundary conditions for fluid dynamics (via eqns. (2.3) and
(2.4)), are included amongst the inputs for the next cycle of change.
Morphological change in coastal evolution therefore ‘cannot be explained
simply by studying the ‘inputs’ individually’ (Wright & Thom, 1977, p. 418),
since morphodynamic processes themselves vary according to the antecedent
system state.
Such non-linear processes are neither additive nor homogeneous (Bendat &
Piersol, 1971, p.37), which translates respectively to: (i) the old expression that
‘the whole is more [or less] than the sum of the parts’; and, more strictly, (ii)
the output of the system is not in constant proportion to the input. Thus, a
system in which the output, Y is a function of its input, X such that
Y = G(X) (2.7)
is non linear if
G =f(X) # constant (2.8)
 Morphodynamics of coastal evolution 53
in which the gain G, resulting from operation of the system (i.e., the process),
is itself dependent upon the system state. Applying eqn. (2.7) to coastal
evolution, Y might be topography and X the flows forcing sediment transport.
Morphodynamic feedback (Fig. 2.1) prescribes that X changes as Y evolves,
resulting in modification of sedimentation processes (eqn. (2.8)). Changes in G
may be continuous for non-linear systems but thresholds can also produce
discrete changes (e.g., at the onset of motion in sediment dynamics). In
general, non linearity arises if representative relationships involve terms
containing powers, products of the output, and/or dependent variables (Bennett
& Chorley, 1978, p. 37).
Non linearity of morphodynamic processes is inherent in the overall
structure of coastal systems since the interactions between the morphodynamic
components (Figs. 2.1 and 2.4) are not weak (von Bertalanffy, 1968). For
example, the evolving substrate in the Newcastle Bight (Fig. 2.7) resulted in
shorefaces becoming progressively more exposed to waves and other open
coast processes as successive shorelines were displaced seaward within the
coastal embayment. Apart from consequent changes in cross-shore flow and
sedimentation regimes, alongshore-transport budgets changed profoundly as
the coast evolved from being laterally bounded toward an open coastline
configuration. Chronostratigraphic evidence from other parts of south-east
Australia indicates that such changes provided the primary controls over the
Quaternary evolution of this coast (Roy et al., 1992), implying strong
variations in G (eqn.(2.7)) and thus marked non linearity in the overall
evolution of the morphodynamics (see Chapter 4).
Moreover, many of the subsystems themselves are strongly non linear by
nature (de Vriend, 1991a). This is especially the case for hydrodynamic
processes, which in general behave like wave motions, with different flow
categories occupying distinctive frequency bands (e.g., Mooers, 1976). These
motions usually become increasingly non linear as water depth diminishes
toward the shore, provided flows have horizontal length scales smaller than the
continental-shelf width. These flows undergo shoaling toward the coastline,
resulting in a steepening of water surface slopes, which causes increasing
interaction between fluid-dynamic processes (Ursell, 1953). The effect of
water-surface slope also becomes important for flows with very long wave-
lengths, such as tides and shelf waves, if they are constricted within straits,
bays or estuaries.
The significance of non-linear hydrodynamics in producing sediment
transport in a coastal morphodynamic system is that the behaviour of particular
flows changes as their non linearity grows. In addition, they gain the capacity
to interact with other non-linear flows to produce mutual changes involving a
54 P.J. Cowell & B.G. Thorn
transfer of energy between flows, often at different frequencies (Guza &
Thornton, 1981; de Vriend, 1991a). This is an important attribute for the
selective growth of resonant motions. Such motions and interactions are
known to characterise the shallow-water hydrodynamics of surf zones leading
to complex bar-trough topographies (Bowen & Guza, 1978; Huntley &
Bowen, 1975; Chappell & Wright, 1979; Wright et al., 1979; Holman, 1981;
Huntley, Guza & Thornon, 1981; Wright, Guza & Short, 1982; Mei, 1985).
Non linearity in some hydrodynamic subsystems is fundamental to coastal
morphodynamic adjustment. For example, the non-linear properties of ‘finite-
amplitude’ gravity waves are deemed responsible for onshore sand transport in
several models for profile morphodynamics of the shoreface and surfzone
(e.g., Bowen, 1980; Bailard & Inman, 1981; Roelvink & Stive, 1991). It also
provides one explanation for the formation of surfzone bars (Boczar-
Karakiewicz & Davidson-Arnott, 1987). Moreover, the non-linear interactions
within and between component morphodynamic processes often require the
full non-linear equations to be solved so that detail and accuracy in the
hydrodynamic model is sufficient to meet the needs of the sediment transport
model (Nielsen, 1990; de Vriend, 1991a; GESAMP, 1991, pp. 39-40). At a
more general level, self-regulating non-linear systems possess greater
resistance to change than do linear systems which, ‘given a slight nudge, tend
to remain slightly off track’ (Gleick, 1988). This inherent stability is
attributable to negative feedback loops that epitomise ‘control systems’ and
enable them to return to their former state after a perturbation. An example of
this is the recovery of the surfzone profile to its pre-storm position following
its upward and landward translation in response to raised water levels during a
storm (Vellinga, 1986; Sumamura, 1989; Dean, 1991).
Most of the preceding examples are from processes operating at the shorter
time scales in Fig. 2.2. This reflects a greater understanding of details in non-
linear dynamics at such scales than is available at present for morphodynamics
operating over either engineering or geological time spans. Nevertheless, the
principles can be expected to apply at all scales. Moreover, the very
‘interactiveness’ responsible for non linearity is likely to cause small-scale
effects to be compounded into larger-scale coastal evolution.

Markovian inheritance
The effect of stochastic variation in external inputs (Fig. 2.4) is superimposed
upon the state-dependent control of coastal evolution. Coastal evolution
thereby proceeds through Markovian inheritance (e.g., Fig. 2.8). In general,
each morphological state in an evolutionary sequence has a set of states to
 Morphodynamics of coastal evolution 55
which it may further evolve during the next step in the succession (e.g.,
Goodman & Ratti, 1979, p. 196). The set of succeeding states is limited by the
position of the existing state within the Markov chain. Selection of the new
state from the set of possibilities depends upon ensuing conditions created by
the stochastic, environmental inputs. In principle therefore, probabilities can
be ascribed to state selection based upon: (i) the transfer function representing
system processes (eqn. (2.7)); (ii) the state of the system immediately before
the change; and (iii) probability distributions for input variables. State
transitions in nature may be continuous through time. In addition, an unlimited
number of possibilities in the set of outcomes may exist for each transition
(e.g., a new bed elevation, hj = hi+Ah). In practice, however, the set of
outcomes must be approximated by a finite number of discrete states with
transitions occurring on an event basis, such as in the case of Swift, Ludwick
& Boehmer’s (1972) Markov process model for progressive sorting of
sediments across the shoreface. In this model different storm-event histories
result in different sediment-dispersal patterns. Nevertheless, there are
problems in nature where the potential transitions are both discrete and finite
in number. Examples include Sonu & James’ (1973) markov chain for the
beach face, and models for surfzone bars by Wright & Short (1984) and
Sunamura (1989), each of which contain a total of only six or eight states
overall, arranged in cycles of erosion and accretion.
Murkovian inheritance at larger scale is illustrated in Fig. 2.8 by the
evolution of a coastal sand body during a marine transgression. Locally, the
sand body may exist in one of three states: a barrier, mainland beach or inlet.
There are three potential state transitions at each time step, and transition
between any of the states is possible (see state-transition table in Fig. 2.8). The
burrier may continue to ‘roll over’ as the sea level rises (Leatherman, 1983) or
undergo transformation into a muinland beach if overstepping occurs (Penland
et al., 1985) or if the substrate becomes too steep. Local transition from barrier
to inlet may occur either through the arrival of an inlet that is migrating
alongshore, or if the barrier is breached during a storm event. Restoration of a
barrier may occur locally at a subsequent step with further migration of the
inlet away from the site, or through shoaling and atrophy of the inlet.
Transition from mainland beach to either barrier or inlet requires flooding of
the low area behind the beach as sea level rises, in accordance with Hoyt’s
(1967) detachment mechanism, provided that the substrate is not too steep (see
Chapter 4).
The potential course of coastal evolution entails a myriad of possibilities
which increase over time (at rate of s”, where s is the number of states at each
step and n is the number of time steps). In this simplest of Markov trees, each
56 P.J. Cowell & B.G. Thom

state-transition probability table

to:

from:

@ mainland beach; @ barrier; inlet;

Figure 2.8. (a) Markov-tree representing stochastic, state-dependent evolution of a
coastal sand body during a marine transgression over 7 discrete time steps. Three
possible states exist, signified by the circled letters in the tree: mainland beach (M),
barrier (B), and tidal inlet (I). The dotted circles (states) and transition lines map the
myriad of possible sequences, of which only one can be realised (hypothetically, the
solid circles and lines). Inset (b) details the probabilities (p) associated with each type
of state transition. For any transition CP,=1, whereas through time, probabilities
multiply such that P+ 0.

state can undergo transition to any other state. Nevertheless, each evolutionary
trajectory through Fig. 2.8 is unique. The general geomorphology of beaches,
barriers and inlets may be identical at every occurrence, unless such
geomorphology is sensitive to antecedent conditions. However, different
evolutionary sequences result in unique differences overall. Stratigraphies
underlying each of the geomorphological elements are the product of recent
antecedent states (e.g., Fig. 2.7). Residual surfaces left by the landward retreat
of sand bodies also depend upon the state sequence. For example, a succession
of inlet states leaves behind shoal-retreat-massif and shelf-valley morphology
on the inner shelf, whereas a barrier sequence leaves a ravinement surface
(Swift et al., 1991a). The sequencing of barrier and inlet states through time
therefore determines the character of the residual innershelf surface.
 Morphodynamics of coastal evolution 57
The probability that any specific evolutionary sequence will occur becomes
progressively smaller with the passage of time since successive transition
probabilities multiply. Thus, if all three transitions in Fig. 2.8 have equal
probabilities ( P = 0.3), the probability for any sequence after just three time
steps is a mere P = 0.04. The hypothetical sequence illustrated in Fig. 2.8 never
persists for more than three steps as any one sand-body type. Accordingly the
residual surface and sand-body stratigraphy would be complex. Coastal
evolution is therefore essentially a matter of historical accident. Joint
dependence upon intermediate states and stochastic inputs means that
evolutionary history is unlikely to repeat itself exactly, even through closed
Markov cycles such as those for the beach face and surf zone., For example, in
Sonu & James’ (1973) closed-cycle Markov simulation of the beach face, there
were only six states, four of which could undergo transition to two other states,
while a single transition was possible in the case of the two end members. In
Sonu & James’ computer simulations, the minimum number of six transitions,
in a single cycle of erosion and accretion, occurred only 6% of the time.
However, one cycle went through 98 transitions before completion. The
median number of transitions per cycle was 16, for which there exists 81
permutations in transition sequencing!
The possibility of geomorphological convergence (e.g., Davies, 1980, p. 7)
seems to be denied by the diversity of coastal evolutionary potential
epitomised by Fig. 2.8. Schumm (1991) suggests that geomorphological
convergence is illusory since landform histories may appear convergent upon
superficial inspection, but reveal idiosyncratic differences when subjected to
closer examination. Nevertheless, the possibility of geomorphological
convergence remains if ‘absorbing’ Markov chains exist. ‘Absorbing’ chains
possess at least one ‘absorbing state’, which can be reached from any other
‘non-absorbing state’, but from which transitions to another state do not exist
(e.g., Goodman & Ratti, 1979, p. 316). Evolutionary sequences become locked
into a single state (or fixed cycle of states) under such circumstances. Residual
surfaces and stratigraphies are then predictable, regardless of initial or
antecedent conditions. This type of Markov chain possesses stationarity as
does the ‘simple’ illustration in Fig. 2.8. In general, however, probabilities
ascribed to the state transitions are likely to change, along with the
morphodynamic processes, as the coast evolves. The resulting nonstationarity
is an important cautionary consideration in application of time-series analysis
to evolutionary sequences (Jenkins & Watts, 1968; Bennett & Chorley, 1978).
Similarly, the interpretation of stratigraphy based on the ‘Law of
Uniformitarianism’ must take nonstationarity into account (Kraft &
Chrzastowski, 1985).
58 P.J. Cowell & B.G. Thom
Finite response-time effects
Morphodynamic response times compound Markovian behaviour since the
input-event sequencing, upon which this behaviour depends, also affects the
strength of hysteresis. Morphodynamic adjustment is constrained by
frequency-response characteristics which reflect the finite time required for
sediment-volume transfers to occur. The relaxation time, TR needed for the
morphological change (Ah)necessary to attain equilibrium, following a shift in
boundary conditions, is related to sediment transport rates (Melton, 1958;
Hardisty, 1987), through sediment continuity (eqn. (2. l)), by

Relaxation times of coastal sand bodies in south-east Australia (examined in

more detail in Chapter 4) were in the order of lo3 years (Fig. 2.9) following
onset of stillstand conditions c. 6000 years ago (Fig. 2.10). Barriers, which

b
MSL

Figure 2.9. Holocene relaxation times indicated in chronostratigraphy from south-east

Australia for: ( a ) prograded barriers; and ( b ) flood-tide deltas. The isochrons are
labelled in years BP. (Adapted from Nielsen & Roy, 1981 .)
 Next Page
Morphodynamics of coastal evolution 59

Figure 2.10. Response times for topographic change in a sand barrier during the last
phase of the Holocene marine transgression and stillstand. A hypothetical sea-level
curve is plotted in ( a ) within an envelope based on radiocarbon dates from south-east
Australia (Thom & Roy, 1985) and the Huon Peninsula, Papua New Guinea (Chappell
& Polach, 1991). Response times are plotted relative to the time constant ( r ) for : ( b )
continuous variations in rate of sea-level rise; (c) a step response (comparing two values
of 7')following a change from transgressive to stillstand conditions; and (d) an impulse
response, during the stillstand, following a sudden event of short duration, such as
severe storm (comparing two values of 7'). The abscissa in (c) and (d)is time (relative
to 7')after the respective events, and <h> denotes time-average bed elevation; g is the
gain;fis the forcing and T is the time elapsed since the input event.
 3
Deltaic coasts
J.R. SUTER

Introduction
Deltas are broadly defined as coastal accumulations of sediment extending
both above and below sea level, formed where a river enters an ocean or other
large body of water (Fig. 3.1) . The key element in this definition is the
presence of a fluvially derived point source of sediment (Boyd, Dalrymple &
Zaitlin, 1992). This chapter concentrates on the Late Quaternary evolution of
marine deltas.
The term ‘delta’ was first applied about 450 BC by the Greek historian
Herodotus, who noted the similarity of the Greek letter to the shape of the
accumulation of sediment around the mouth of the Nile River. Despite the fact
that many of the world’s deltas do not show this particular morphology, the
term has remained.
Deltas are extremely important depositional systems, both ecologically
and economically. They often contain extensive wetlands, whose high
biological productivity makes them vital nursery grounds for fisheries. Just as
significant are the agricultural activities supported by the fertile soils of the
world’s deltas. Many of the world’s largest ports are located on distributaries
of major deltas. Huge amounts of coal, oil, and natural gas, still the principal
fuels of the world economy, are derived from subsurface deltaic deposits.
The purpose of this chapter is to examine the geomorphology and
sedimentology of modern deltas and provide a synthesis of the evolutionary
elements. This will be done in several steps: 1. Examination of the processes
responsible for deltaic deposition; 2. discussion of basic deltaic environments,
3. development of idealized, end-member models of deltaic sedimentation;
4. discussion of the evolutionary implication of these concepts. Although these
will be wholly derived from examination of modern deltas, this approach can
be usefully applied to ancient deltas in a sequence-stratigraphic context.
Finally, the chapter examines interactions of human activities and deltaic

87
88 J.R. Suter

Alluvial
Valley
Coastal
Plain

Coastal
Plain
%:
Plain
Distributary
Mouth Bars
Lower
Delta
Plain Abandoned
Delta

Delta Front

Prodelta

Figure 3.1. Basic environments of a fluvially dominated delta.

processes to show how an understanding of such processes can contribute to a

more enlightened approach to coastal management.

Deltaic processes
Marine deltas form where a fluvial system delivers sediment faster than
marine processes can rework it. Continued sediment supply results in
progradation of the shoreline and the characteristic protuberance of the delta
(Fig. 3.1). Deltaic sediments accumulate in three main environments: the delta
plain, dominated by fluvial processes; the delta front, reflecting river-marine
interaction, and the prodelta, which is fully marine. The expression of these
environments is the result of the interaction of many forces (Fig. 3.2). These
 Deltaic coasts 89

FLUVIAL INPUT
1

17
4 RNER-
15 DOMINATED

6
11
I
9
WAVE- TIDE-
11 DOMINATED DOMINATED 1b
10
14 a ia

WAVE ENERGY TIDAL ENERGY

MISSISSIF'PI 10 COLORADO
BRAZOS 11 REONE
GANGES-BRAHMAPUTRA 12 sxoFRANcISco
ATCHAFALAYA 13 KLANGWGAT
MAEAKAM 14 COPPER
NIGER 15 DANUBE
MEKONG 16 FLY
ORINOCO 17 PO
NILE 18 EBRO
Figure 3.2. The process-based deltaic classification. Positions of specific deltas are
plotted on the basis of general delta-front morphology, which reflects the dominant
processes acting upon the system (after Galloway, 1975).

include the nature of the fluvial system (drainage basin, climate, discharge,
sediment calibre and load), river-mouth processes, basin morphology and
tectonics, marine reworking, and evolutionary factors such as autocyclic delta
switching or eustatic fluctuations.

Fluvial input
Fluvial input to a delta is largely a function of the rate and seasonality of
discharge, and the amount and calibre of the sediment load. These are
controlled by the drainage basin climate, tectonics, and lithology.
Climate determines the system's discharge, and affects the rate of physical
and chemical weathering, governing the amount and calibre of the sediment
90 J.R. Suter
load supplied. The seasonality of rainfall also affects the uniformity of
discharge, with important implications for the type of fluvial systems
delivering water and sediment to the river mouth. Climatic factors also greatly
influence the deposition and preservation of organic material on the delta
plain.
Sediment load and calibre are determined by the nature of the drainage
basin. An active tectonic upland leads to a coarser-grained fluvial system, such
as in the Ganges-Brahmaputra River; a more quiescent drainage area contributes
to a finer sediment supply, as in the Mississippi River. Some aspects of deltaic
deposition are a direct function of the type of sediment supplied to the marine
receiving basin.

River-mouth factors
Fluvial-marine interaction at the river mouth is a primary control on
deltaic deposition. As the river waters mix with those of the receiving basin,
the inflow expands, diffuses and loses competence, resulting in deposition.
The sedimentation pattern is influenced by the nature of the river waters,
amount and calibre of the sediment load, and many basinal factors, such as
water depth, bathymetry, tidal range (and thus the strength of tidal
reworking),wave energy, and oceanic currents. These interactions produce
three primary forces controlling deposition at the river mouth: 1. the inertia of
issuing river water and associated turbulent diffusion; 2. friction between the
effluent and the sea bed, and 3. buoyancy resulting from density contrasts
(Fig. 3.3). Most considerations of river-mouth processes draw heavily on the
work of Bates (1953). This ‘rational theory’ of delta formation (Fig. 3.3), was
a pioneering application of hydrodynamic principles to geological problems.
Wright (1977, 1985) provides excellent reviews of river-mouth processes.
Where fluvial input is large, density contrasts near the river mouth are
usually small, and flow is referred to as homopycnal (Fig. 3 . 3 ~ )The
. primary
forces acting on the jet flow are inertial, causing the effluent to spread as an
axial turbulent jet (Fig. 3.3). Coarse-grained sediment is deposited radially
near the river mouth as bars with steeply dipping foreset beds, separating
lower-angle topset and bottomset beds (Fig. 3 . 4 ~ ) .Such deposits were first
described in lacustrine settings by Gilbert (1885), and are now commonly
termed Gilbert-type deltas. This behaviour is not common in marine
deltas, although some homopycnal behaviour may occur during periods of
high discharge.
Hyperpycnal conditions occur when the inflow is more dense than the
receiving basin water (Fig. 3.3b). Incoming waters flow along the bottom as
 Deltaic coasts 91

a
f THOROUGH MIXING
\
DISTRIBUTARY~ 3
- J
-+
AXIAL JET FLOW

-AS GILBERT-TYPE DELTA

SUSPENDED LOAD
SETTLES RAPIDLY

AXIAL CROSS-SECTION
PLAN VIEW b

2- d

/:-%
AXIAL CROSS-SECTION I BED LOAD
PLAN VIEW I C

---__

LOW DENSITY HIGH

Figure 3.3. Modes of interaction between fluvial and marine waters at distributary
mouths: ( a ) homopycnal flow, (b)hyperpycnal flow, (c) hypopycnal flow (after Bates,
1953, and Fisher et al., 1969).
92 J.R. Suter

A. Inertia-dominated

B. Friction-dominated

Subaerial Levee

Subaerial Levee

C. Buoyancy-dominated

Subaerial Levee

Coarser Sand Silty Sand

I 0Finer Sand 0Silt and Clay
Figure 3.4. Idealized depositional patterns associated with distributary-mouth bars:
( a ) inertia-dominated, ( b ) friction-dominated, and (c) buoyancy-dominated (after
Wright, 1977).

turbidity currents, bypassing the shoreline. Such behaviour is seen in the silt-
rich Huanghe delta of China (Wright et al., 1988), and is often associated with
coarser-grained systems characterized by flash floods and sediment gravity
flows (Orton & Reading, 1993).
 Deltaic coasts 93
Most marine deltas form under hypopycnal conditions, in which the river
effluent is less dense than the saline receiving basin (Fig. 3 . 3 ~ )Turbulent
.
mixing results in deposition of the sediment load, with the coarser material
being deposited near the river mouth. Finer sediments are transported further
offshore and deposited from suspension as the plume disperses. When outflow
is into relatively shallow water, friction between the sea bed and the base of the
outflow causes deposition of a ‘midground’ distributary-mouth bar bordered by
bifurcating channels (Fig. 3.4b). Shallower delta platforms and coarser-grained
sediment loads favour this friction-dominated river-mouth style. Outflow into
deeper water results in lessened bottom friction, and subsequent dominance of
buoyant forces. Under these conditions, fresh water spreads as a narrow
expanding plume above a salt-water wedge, which may extend inland up the
distributary channel for a considerable distance. This process results in
elongate distributary-mouth bars which extend a considerable distance into the
basin, as shown by the Balize complex of the Mississippi delta (Figs. 3.4c, 3.5).
The above processes are idealized end members of the river-mouth
spectrum. Real-world deposits display a combination of inertial, frictional, and
buoyant processes, influenced by variations in discharge and characteristics of
the receiving basin.

Main Pass

North Pass

Head of
Passes Fresh Marsh
Salt Marsh

Kilometers

Southwest Pass
Gulf of Mexico

Figure 3.5. The river-dominated Balize lobe of the Plaquemines-Modern complex of

the Mississippi delta, showing the elongate ‘birdfoot’ morphology.
94 J.R. Suter
Marine reworking
Once a fluvial system has delivered sediment into a marine basin, waves,
tides, and oceanic currents begin to rework the sediment. Basin tectonics and
morphology greatly influence marine processes. An open-ocean basin has a
greater fetch, and thus the potential for greater wave energy, making a wave-
dominated delta (Figs. 3.2, 3.6) more likely. An elongate, enclosed basin can

Rio SBo Francisco

ATLANTIC OCEAN

FRESHWATER SWAMP

HOLOCENE BEACH RIDGES MANGROVE SWAMP

HOLOCENE DUNES PLEISTOCENE BEACH RIDGES

FLUVIAUDISTRIBUTARY TERTIARY UPLANDS

Figure 3.6. The wave-dominated mouth of the SPo Francisco River of Brazil. This
feature is often cited as a type-example of wave-dominated deltas, although beach
ridges on either side of the river may have different provenances. Despite such
complications, this may still be considered a delta because of the protruding shoreline,
fluvial sediment source, and coeval delta plain (after Dominguez et al., 1992).
 Deltaic coasts 95

amplify tidal currents, resulting in tide-dominance (Figs. 3.2, 3.7). River-

dominated deltas, such as the Mississippi (Figs. 3.2, 3 3 , tend to occur
in tectonically quiet basins with more limited marine energy. Rivers entering
open-ocean settings are subject to the full effects of oceanic processes,

TIDAL LIMIT

ACTIVE
DISTRIBUTARIES CHENIER
PLAIN

DELTA FRONT

PRODELTA

Non-tidalDelta Plain Tidal Sand

RidgeslBercher

Tidal Delta Plain

Figure 3.7. Idealized tide-dominated deltaic environments (after Dalrymple, 1992).

96 J.R. Suter

including waves, tides, storms, and semi-permanent ocean currents. These

rework the sediment to a greater or lesser degree. The Niger delta (Allen, 1970;
Oomkens, 1974) is an excellent ‘mixed-energy’ example. The world’s largest
river, the Amazon, which carries an extensive sediment load, is unable to
establish a subaerial delta, although its subaqueous platform extends for several
hundred kilometres offshore and alongshore (Nittrouer et al., 1986).
In wave-dominated settings, breaking waves cause immediate mixing of
fresh and salt water (see Chapter 4). Typically, freshwater flow velocity
decelerates rapidly. A bar niay form in the immediate vicinity of the
distributary mouth, often supplemented by landward migrating swash bars.
These bars are constantly reworked by wave action, and sediment is
moved alongshore, usually being deposited as a series of beach ridges centred
around distributary mouths (Fig. 3.6). Favoured directions of wave approach
can result in asymmetric beach ridges, and may cause the progradation of a spit
across the river mouth, resulting in channel flow oblique or parallel to shore.
Finer-grained sediments are carried offshore to form the subaqueous portions
of the deltaic platform, which is usually areally restricted as compared with
more fluvially dominated deltas. Some well-documented examples of wave-
dominated deltas include the Brazos, Nile, Senegal, Rhone, and Siio
Francisco (Figs. 3.2, 3.6).
Tidal currents at the river mouth cause mixing of the basinal and incoming
waters and thus lessen density stratification. Tidal currents may be stronger
than the fluvial input, dominating sediment dispersal and moving the land-sea
and river-marine interfaces horizontally and vertically (Wright, 1985).
Distributary mouths of tide-dominated deltas (Figs. 3.2,3.7) are usually funnel
shaped (see Chapter 5). Under such conditions, sediments tend to be deposited
as linear sand ridges, flanked by different channels for ebb and flood tides.
These tidal ridges replace the normal distributary-mouth deposits. Tidal
influence can reach a considerable distance up the river itself, resulting in
landward transport of coarse-grained material. Finer-grained sediments are
carried either offshore by ebb flows, or upriver by flood tides. These may
accrete as marginal or interchannel flats. Many of the tide-dominated or tide-
influenced deltas are found in tropical areas, such as the Mekong,
Klang-Lagat, and Mahakam deltas of southeast Asia. These commonly have
delta plains dominated by mangroves or palms in the areas of tidal influence,
which grade into more typical delta plains landward (Fig. 3.7).
Prevailing oceanic currents or geostrophic storm flows may also distribute
deltaic sediments. The general effect of such currents is to align deposition
more alongshore (Coleman & Prior, 1982). Sediments may be transported
great distances away from the river mouth, and not be considered part of the
 Deltaic coasts 97
deltaic system at all. Numerous rivers, including the Mississippi,
Amazon-Orinoco and Yangtze, have deposited extensive marginal chenier
plains (Augustinus, 1989).

Syndepositional deformation
Sediment redistribution and modification of deltaic sediments may occur
by contemporaneous mass movement. This is the result of high rates of
deposition, producing elevated water contents and excess pore pressures in
sediments. Biogenic gas produced by decomposing organic matter adds fluid
pressure and decreases internal strength. If sufficient stress is applied by
waves, currents, or sediment loading, failure may occur. Modern deltas which
show such contemporaneous slumping include the Mississippi, Huanghe,
Niger, Amazon, Nile, Orinoco, and Magdalena (Coleman & Prior, 1982; Lu
et al., 1991; Wright et al., 1988). Such mass movement may take numerous
forms, including small-scale mudflows and gullies, diapiric intrusions
(‘mudlumps’), growth faults, and shelf-edge failures (Coleman & Prior, 1982;
Coleman, Prior & Lindsay, 1983). Syndepositional deformation is responsible
for much downslope transport of sediments in modern deltas, and was
common in Late Pleistocene lowstand deltas (Lehner, 1969; Winker &
Edwards, 1983; Suter & Berryhill, 1985).

Autocyclicity and allocyclicity

Evolutionary factors affecting deltaic systems take two major forms.
Autocyclic processes are those that are a part of the system itself, while
allocyclic processes, such as climatic changes or eustatic fluctuations, act
independently of deltaic deposition.
The classic autocyclic process is delta switching, in which an alluvial
system or distributary is abandoned or avulsed for a more hydraulically
favoured route. This occurs in some form in virtually all deltas, but the best
known examples of this process lie in the Mississippi delta of the Gulf of
Mexico (Figs. 3.5, 3.8). Here, delta switching occurs at two scales. During a
period in which the main trunk of the Mississippi River occupies a particular
position, a number of delta lobes prograde by a series of distributary
channels, which are successively abandoned. Abandonment initiates the
transgressive phase of the delta cycle (Scruton, 1960). The abandoned delta
subsides, and coastal processes rework the seaward margin (see Penland,
Boyd, & Suter, 1988). New lobes form nearby. These lobes stack together to
form a delta complex (Frazier, 1967). Seven major deltaic complexes during
98 J.R. Suter

TECHE BERN A R 0

LAF~URCHE

G u u o f Mexico
Kilometm

Pleistocene TST D e l t a C o m p l e x e s
Subaqueous Sand Bodies
HST D e l t a C o m p l e x e s
Barrier Shoreline

Figure 3.8. Distribution of deltaic complexes and associated transgressive deposits

built during the last 7000 years in the Mississippi River delta plain. Line X-X' shows
location of cross-section shown in Fig. 3.12. TST = transgressive systems tract;
HST= highstand systems tract. (From Boyd et al., 1989; after Frazier, 1967.)

the Holocene mark trunk stream avulsions of the Mississippi River. One is
ongoing as the river tries to switch to a new course down the Atchafalaya
(Fig. 3.8). Lobe switching is typical of river-dominated deltas which have a
single favoured distributary path, and results in the accumulation of a large,
complex delta plain (Fig. 3.8).
Delta switching may also occur by channel extension, in which several
distributaries originate from a common point, taking turns at being the active
conduit. Series of beach ridges oriented about the various river mouths are a
typical result. The Rhone delta is an excellent example of this process
(Oomkens, 1970). Avulsions far updip on the delta plain may also initiate
switching, as in the Ganges-Brahmaputra delta.
Allocyclic processes dramatically affect deltaic deposition. Climatic
variations may alter discharge or sediment supply as discussed above. Eustatic
fluctuations also have major impacts. Falling sea level forces the deltaic
system to prograde rapidly across the emerging continental shelf, and may
result in entrenchment of a particular distributive network. Rising sea level
 Deltaic coasts 99
creates greater accommodation space, favouring aggradation or retrogradation
over progradation. In a later section it will be shown that different deltaic styles
characterize each portion of a eustatic cycle.

Constituent facies: deltaic successions

Broadly, deltas can be divided into subaerial, transitional, and subaqueous
components. Within these major subdivisions, individual environments occur
(Fig. 3.1). The delta plain is entirely subaerial, the delta front is transitional
between fluvial and marine environments, while the prodelta is entirely
subaqueous. Each of these major environments can be recognized in some
form in all delta types, although a host of specific differences are present which
are caused by the dominant forcing factors operating on the particular system.

Delta plain
The delta plain is a mosaic of distributary channel and interchannel
environments. Interchannel areas may consist of a variety of bays, lagoons,
estuaries, flood plains, lakes, tidal flats and creeks, marshes, swamps, dune
fields, or salt flats, depending on the type of delta and the climate. In large
deltaic systems like the Mississippi, the active delta plain area is small in
comparison to the area of abandoned deltaic plain, which continues to evolve
under transgressive conditions (cf. Penland et al., 1988; Fig. 3.9).
The delta plain is subdivided into the upper and lower delta plains. The
upper delta plain is essentially a fluvial environment. Deposition is by fluvial
processes, both within and between channels. Vegetation can be quite variable,
but a simple distinction is that the upper delta plain is mostly a freshwater
environment and usually has more diverse vegetation.
The lower delta plain is typically a brackish to saline environment and is the
scene of more active deposition, mostly by crevassing or overbanking from the
deltaic distributaries, modified by input of marine sediments from storms.
Because the natural levCes are usually more poorly developed in the lower
delta plain, such crevassing is more frequent, as witness the subdeltas of the
Balize complex of the Mississippi delta (Coleman & Prior, 1982). Vegetation
is necessarily salt tolerant, and is confined to those plants that can withstand
prolonged inundation, such as Spartina grasses or mangroves, depending
on climate. In arid settings, the lower delta plain can be a zone of salt flats or
active dune fields, especially on wave-dominated deltas with their extensive
beach-ridge plains (Fig. 3.6).
The coarsest-grained deposits within delta plains are associated with
 ACTIVE DELTA SLSU
EROSIONAL HEADLAND
WITH FLANKING BARRIERS

1
r
REGRESSIVE
ENVIRONMENTS
Distributary
Fresh Marsh
z
4
dl Beach Ridge
c

TRANSGRESSIVE
ENVIRONMENTS
Subaerial
Barrier Sands
STAGE 2 Subaqueous
TRANSGRESSIVE Barrier Sands
INNER-SHELF SHOAL
BARRIER ISLAND ARC Sand Sheet
Salt Marsh
Recurved Spit
Shell Reef
Tidal Channel

Figure 3.9. A geomorphic model of the evolution of a lobe/complex in the Mississippi delta (from Penland et al., 1988).
 Deltaic coasts 101

the distributary channels. Processes within upper delta-plain channels are

essentially fluvial and facies consist of fining-upward successions with
erosional bases. On the lower delta plain, distributaries are more influenced by
marine processes, such as tidal fluctuations and salt-water wedge intrusion
(Fig. 3.3). Abandoned distributary channels may develop into estuaries, and be
filled by marginal marine deposits. For example, Bayou Lafourche, an active
distributary of the Mississippi River until about 300 years ago (Frazier,
1967), remains largely open and is essentially a brackish estuary (Fig. 3.8). The
proportion of distributary channel fill to other deltaic facies is largely a
function of the degree of fluvial dominance of the delta. Wave-dominated deltas
usually have fewer distributaries than their river-dominated counterparts and a
correspondingly larger proportion of interchannel and marginal marine facies.
The delta plains of tide-dominated deltas are complexes of swamps, marshes,
tidal flats and channels, reflecting the mix of fluvial and tidal processes (Fig. 3.7).
Another aspect of deltaic plain sedimentation is the deposition of organic
lithofacies, often used as analogues for coal deposits in the rock record. Many
ancient coal deposits are considered to have formed in deltaic settings (e.g.,
Horne et al., 1978, Ryer; 1981), although this view is undergoing some re-
evaluation (McCabe, 1984, 1991; McCabe et al., 1989; Haszeldine, 1989;
Kosters, Chmura & Bailey, 1987; Kosters, 1989; Kosters & Suter; 1993, Scott,
1989). However, organic lithofacies can form an important component
of deltaic deposits in modern settings. Peats can form throughout the upper and
lower delta plain. Because peat accumulation depends upon organic
productivity, subsidence, and lack of clastic sediment supply, higher quality
peats have typically been associated with the upper delta plain or in abandoned
portions of deltaic complexes. Brackish and salt marshes on the lower delta
plain tend to produce organic-rich muds or high-sulphur peats, due to
admixture with salt water and frequent clastic influx. Climatic conditions are
of paramount importance. Substantial differences in plant type and water table
level are found from high to low latitudes and from arid to humid regions.
Some peats may form during active deltaic deposition as raised mires (e.g.,
Coleman, Gagliano & Smith, 1970; Styan & Bustin, 1983). These deposits are
elevated above levels of river flooding and clastic influx, and are not coupled
to the groundwater table (Diessel, 1992).

Delta front
Most of the active deposition in a delta takes place in the transition zone from
the fluvial to the marine environment, referred to as the delta front. Usually,
the coarsest material is deposited at the mouths of the distributaries as bars,
102 J.R. Suter
variously called distributary-mouth, channel-mouth, stream-mouth and
middle-ground bars (Fig. 3.4; see discussion above). In river-dominated deltas,
finer-grained materials travel further offshore in suspension, or by traction and
gravity-currents. Progradation gives the deltaic deposit its characteristic
upward-coarsening signature, in grain size, bed thickness, and scale of
sedimentary structures.
Offshore of the distributary-mouth bars is the marine portion of the delta
front. Wave-dominated delta fronts differ very little from standard shoreface
environments, although the latter would have a more diverse microfauna and
more bioturbation. In the immediate area of the distributary mouth there may
be a subdued distributary-mouth bar. However, these deposits can be
completely reworked by wave energy following abandonment of the
distributary. Good documentation of stratigraphic successions of some wave-
dominated deltas can be found in Bernard et al. (1970), Oomkens (1970),
Dominguez & Wanless (1991), Chen, Warne, & Stanley (1992), and Stanley &
Warne (1993).
The tide-dominated delta fronts are complex mazes of tidal sand ridges
and channels, grading offshore into open shelf environments. Less information
is available about these stratigraphic successions than for other deltaic types.
Allen, Laurier & Thouvenin (1979) and Gastaldo & Huc (1992) offer some
stratigraphic information about the Mahakam delta, and Meckel (1975)
provides some on the Colorado delta of the Sea of Cortez. Coleman & Wright
(1975) showed idealized vertical successions for tide-dominated deltas.

Prodelta
Entirely subaqueous, the prodelta is the finest-grained portion of a delta.
Sediments here are deposited mostly from suspension, or from dilute turbidity
current flows. In most cases, the prodelta merges imperceptibly landward into
the delta front, and seaward into the open shelf environment (Fig. 3.1).
Prodelta deposits are usually laterally continuous, and show less lithological
variation than the delta front. The finest sediments are found at the greatest
depths, and usually a coarsening-upward signature is present. Parallel
laminations are the dominant sedimentary structure. Beds are thinnest at
the base, thickening upward.
Relatively slow or intermittent deposition can permit marine organisms to
colonize the sediments of the prodelta. These communities are most abundant
on the seaward margins and in the basal sections of prodelta deposits.
Landward and upward, greater stress by increasing sedimentation rates results
in less bioturbation.
 Deltaic coasts 103
Deltaic models
Models are commonly used in geology as idealized simplifications of complex
natural environments and processes. Models may be scale representations,
mathematical constructs, visualizations, inductive, or actualistic (Reading,
1986a). Facies models are general summaries of a depositional system, which
can be used to interpret ancient deposits, guide future observations, and predict
facies distributions and process-responses (Reading, 1986b; Walker, 1992).
Deltaic facies models have long been used in the energy and minerals
industries, where decisions of great economic consequence must be made
routinely with limited data. Today, such concepts are finding increasing use
in hydrogeology, environmental studies and in coastal management.
Deltaic models and classifications have undergone a considerable
refinement since the days in which the B a k e complex of modern Mississippi
delta (Fig. 3.5), the familiar ‘birdfoot’ delta, was used as an exploration model
for virtually all deltaic deposits. As more deltas have been investigated and
understanding of sedimentary processes has evolved, it has become
increasingly difficult to synthesize the existing database into the simple,
idealized forms alluded to above. A multitude of processes and variables
interact in the deposition of deltas and a very large number of types are
possible. Although every delta is different to some extent, there are common
characteristics which can be used to define useful classification schemes which
have some predictive capability. The challenge is to prepare models that both
capture this variety and synthesize the available information into a useful
form.
Deltas have been classified and modelled on the basis of their alluvial
feeder system (e.g., Holmes, 1965; McPherson, Shanmugan, & Moiola, 1987),
general morphology (Galloway, 1975), tectonic and physiographic settings
(Ethridge & Wescott, 1984), position in a eustatic cycle (Suter & Berryhill,
1985; Suter, 1991), thickness distribution patterns (Coleman & Wright, 1975;
Wright, 1985), dominant grain size (McPherson et al., 1987; Orton, 1988;
Orton & Reading, 1993), and numerous combinations of all of these factors
(e.g., Postma, 1990). The references cited above are by no means a complete
listing. Excellent recent reviews of this subject can be found in Elliot (1986,
1989), Nemec (1990), and Orton & Reading (1993). To capture the variety of
deltaic systems, many of these classifications are quite detailed and complex,
requiring large amounts of information for accurate ‘pigeon-holing’ of a
particular deposit. The predictive capability of such detailed models is quite
high, but applicability is limited to those deposits which share many similar
characteristics and are dominated by the same forcing factors. In other words,
in order to use the models as a predictor of facies or an interpretive guide, one
104 J.R. Suter
must already know the facies distributions and forcing factors involved. If a
specific detailed model is chosen improperly it can do more harm than good.
Such detailed comparisons are more properly referred to as analogues,
applicable to single systems rather than the more general term facies model.
Amongst sedimentologists, the most widely accepted deltaic facies models
have been process based. Building upon earlier efforts (e.g., Fisher et al., 1969;
Morgan, 1970; Wright & Coleman, 1973), Galloway (1975) proposed a
simple, elegant ternary classification which differentiated modern deltas into
end members based upon delta-front morphology (Fig. 3.2). Deltas are
considered to be fluvially or river dominated if they have sufficient sediment
input to prograde the shoreline significantly and show little modification by
marine basinal processes. If marine reworking is sufficient to redistribute
sediments away from the river mouth, deltas are classified as wave or tide
dominated, depending upon the morphology. This classification allows the
recognition of dominant processes acting upon a given deposit, and predicts
within a broad framework what the resulting style of deltaic deposition will
be.
This simple scheme does have some drawbacks. Classifications dependent
upon delta-front morphologies tell us little about the deltaic plain. The
classification ignores the grain size or calibre of sediment input, a critical
factor in determining deltaic type. Orton & Reading (1993) have modified the
delta triangle into a prism, with its long axis being various dominant grain-size
populations. Interested readers are referred to this paper, which contains a
thorough treatment of many of the factors alluded to above. However, despite
its greater complexity and useful modifications and extensions of the original
idea, this approach ultimately classifies deltas as river, wave, or tide
dominated.
At present there is much debate about the wave- and tide-dominated apices
of the delta triangle (Bhattacharya & Walker, 1992; Dominguez, Martin &
Bittencourt, 1987; Walker, 1992). Many wave-dominated ‘deltas’ are actually
composite features formed by longshore drift of sediments from other sources.
The distributary system provides some sediment input, and also acts as a
barrier or ‘groyne’ to the longshore drift. Excellent examples of this type of
depositional system are the beach-ridge plains of the Brazilian coast, such as
the ,920 Francisco delta, in which different provenances can be demonstrated
for beach ridges on either side of ‘distributary’ channels (Dominguez et al.,
1987; Fig. 3.6). Numerous other deltas, e.g., the Orinoco (van Andel, 1967),
the Brazos (Bernard et al., 1970; Suter & Morton, 1989) and the Rhone
(Oomkens, 1970) derive some of their delta-front sediments by longshore drift.
These deposits might be considered as prograding strandplains (Curray,
Emmel, & Crampton, 1969; Dominguez & Wanless, 1991; Boyd et al., 1992).
 Deltaic coasts 105

However, similar beach-ridge plains are also seen in some lobes of the
Mississippi delta, which are overall river dominated (Penland et al., 1988).
Strandplains are strike-fed depositional systems which are welded onto a pre-
existing coastal plain, whereas a true delta has a fluvial point source and a
coeval delta plain (cf. Boyd et al., 1992).
Tide-dominated deltas present special problems due to their possible
confusion with tide-dominated estuaries (Bhattacharya & Walker, 1992;
Dalrymple, 1992; Walker, 1992; see Chapter 5). The critical point of distinction
lies in the dominant sediment source, which in the estuarine case may be non-
fluvial (Boyd et al., 1992), and in the morphology of the shoreline. In the Bay
of Fundy, perhaps the type example of a tide-dominated estuary, the updip
fluvial source (the Schubenacadie and Salmon rivers) is small in comparison to
the sediment being derived from erosion of the valley walls (Dalrymple, pers.
commun., 1989). However, examples such as the Ganges-Brahmaputra
(Wright, 1985; Umitsu, 1993) in the Indian Ocean, the Mahakam (Allen et al.,
1979; Gastaldo & Huc, 1992) and Mekong (Kolb & Dornbusch, 1975) deltas
in the South China Sea and the Colorado (Meckel, 1975) delta in the Sea
of Cortez show that it is entirely possible for the shoreline to be prograded by
fluvially derived sediment, whilst showing a tidally modified or dominated
morphology. Such systems are properly considered deltas.
A final shortcoming of process-response models relates to deltaic evolution.
Deltas evolve throughout their existence, and reflect processes extant at a
given time, which may not be representative of the entire deltaic cycle. A lobe
or complex of the Mississippi delta evolves from fluvial dominance during its
deposition, to wave dominance during its transgressive phases (Fig. 3.9)
(Russell, 1936; Fisk, 1944; Scruton, 1960). Penland et al. (1988) discussed this
process in detail, showing how the reworking of deltaic deposits during the
abandonment phase changes the nature of even the deltaic plain. Use of models
as predictive tools, either for subsurface exploration or coastal management,
requires an understanding of these processes.
Deltaic style also varies throughout a eustatic cycle (Winker & Edwards,
1983; Suter & Berryhill, 1985; Elliot, 1989; Suter, 1991). Shelf-phase or shoal-
water deltas, the typical modern variety, can differ dramatically from their
shelf-margin, or deeper water counterparts, which are more commonly
deposited during eustatic lowstands (Figs. 3.10 and 3.1 1). Quaternary deltaic
facies models which incorporate their evolution throughout complete eustatic
cycles are more apt to be applicable to the ancient record than those
which concentrate solely on the current highstand version. Most ancient deltaic
deposits are really composite systems which record the evolution of
depositional systems through a variety of phases of basin subsidence, eustasy,
and sediment supply.
106 J.R. Suter

Figure 3.10. General model of deltaic deposition during eustatic fall in the Gulf of
Mexico. Rates of progradation are enhanced by absolute sea-level fall, and ensuing
fluvial incision removes much of the delta plain (after Suter & Berryhill, 1985). Shelf-
phase delta is characterized by low accommodation, rapidly eustatically enhanced
progradation, thin, widespread, low-angle clinoforms, and extensive fluvial incision.
Shelf-margin delta is characterized by high accommodation and subsidence,
aggradation, thick localized wedge, and steeper clinoforms.

Figure 3.1 1. Location of Late Quaternary incised fluvial systems and associated
lowstand deltaic systems in the northwestern Gulf of Mexico (after Suter & Berryhill,
1985).
 Deltaic coasts 107
Quaternary deltaic evolution
Early deltaic models were constructed largely on the basis of the
subaerial geomorphology of a few well-studied deposits around the world. The
advent of high-resolution seismic reflection profiling coupled with the
movement of the petroleum industry into the offshore made available a new
database on deltaic deposits on the continental shelf and shelf margin (Moore
& Curray, 1964; Lehner, 1969; Winker & Edwards, 1983; Suter & Berryhill,
1985). These data allow the consideration of deltaic evolution during a
complete eustatic cycle, which can be used to construct models in the context
of the new field of sequence stratigraphy.
Sequence stratigraphy is the study of rock relations within a chrono-
stratigraphic framework of genetically related strata, bounded by surfaces of
erosion or nondeposition, or their correlative conformities (van Wagoner et al.,
1990). These successions are termed depositional sequences, and contain
smaller genetic units. Systems tracts are linkages of contemporaneous
depositional systems (Posamentier & Vail, 1988; van Wagoner et al., 1990).
These are defined objectively on the basis of types of bounding surface,
position within a sequence, and parasequence or parasequence set stacking
patterns (van Wagoner et al., 1990). Lowstand, transgressive and highstand
systems tracts accumulate during various phases of a eustatic cycle.
Parasequences are relatively conformable successions of genetically related
beds or bedsets bounded by marine-flooding surfaces or their correlative
surfaces (van Wagoner et al., 1990). Flooding surfaces are boundaries between
individual parasequences. Progradation and abandonment of a single delta
lobe creates a parasequence (Boyd, Suter & Penland, 1989; Kosters & Suter,
1993; van Wagoner et al., 1990), and related groups of deltaic lobes, i.e. a delta
complex (Frazier, 1967), can be considered a parasequence set (Kosters &
Suter, 1993). The stacking patterns of these parasequences and parasequence
sets can be related to glacio-eustatic fluctuations.
Quaternary deltaic evolution in the northwestern Gulf of Mexico has
followed a predictable if complicated pattern, controlled by relative sea level,
nature of the receiving basin, and the rate of sediment influx (Suter, 1991).
During a glacio-eustatic cycle, sea level fluctuates in a roughly sinusoidal
fashion, divisible into falling, lowstand, rising, and highstand phases. Each of
these divisions shows a particular style of deposition (Suter & Berryhill, 1985;
Suter, 1991). During falling stage, deltaic systems prograde rapidly across the
previously submerged shelf. Accommodation space is relatively low, and
these shelf-phase deltas (Suter & Berryhill, 1985) are relatively thin and
stacked progradationally (Fig. 3.10). Much of the present continental shelf
is covered by these deposits. Because of eustatic fall, these deposits do not
108 J.R. Suter
undergo the full deltaic cycle, lacking a fully developed transgressive phase.
Such systems become progressively more wave dominated during the overall
fall, as frictional attenuation of wave energy is lessened by the decreasing shelf
width. As eustatic fall continues, fluvial systems incise into the deltas,
removing much of the pre-existing deltaic sediments (Fig. 3.1 1). Fluvial
incision and subaerial weathering processes create an erosional unconformity
on the Pleistocene surface. This incised lowstand surface is a Type-1 sequence
boundary in the nomenclature of sequence stratigraphy (van Wagoner et al.,
1990).
At eustatic lowstand, deltaic systems are fixed at the heads of the incised
valleys, at or near the shelf margin. Greater accommodation space from
increased water depth, higher subsidence and sea-floor gradients creates a
distinctly different deltaic style (Suter & Berryhill, 1985; Berryhill & Suter,
1986). These shelf-margin deltas (Fig. 3.11) reach thicknesses of over 150 m
during a single glacio-eustatic cycle, and tend to stack aggradationally,
comprising the lowstand systems tract. Syndepositional deformation processes
are common, and serve as a primary mechanism for downslope sedimentation.
Shelf-margin deltas are a primary mechanism of basin-filling, and are
preferred hydrocarbon exploration targets. Numerous petroliferous deposits in
the Tertiary of the Gulf of Mexico have been interpreted as shelf-margin deltas
(Edwards, 1981; Morton et al., 1991).
Eustatic rise causes retrogradation into the incised valleys. Fill within
these laterally restricted containers is largely controlled by sediment supply.
High sediment supply systems choke their valleys with fluvial sediments,
while lower sediment supply systems have more complicated fill, reflecting
the creation of estuarine conditions. Thin ( < l o m) lacustrine and bayhead
deltas are deposited under these conditions. These deposits are characteristic of
the transgressive systems tract. In the Gulf of Mexico these deposits are mostly
fluvially dominated. In areas with greater tidal ranges, tide-dominated or
influenced deltas may be deposited due to amplification of tidal energy by the
embayments (Allen, 1991; Dalrymple, Zaitlin, & Boyd, 1992).
Once the valley has been filled, deltaic style changes. Sufficient sediment
supply results in the progradation of shelf-phase deltas during stillstands in the
overall rise. These backstep, or stack retrogradationally, with each successively
younger delta being thinner and occurring further landward (Fig. 3.12). This
reflects a diminishing sediment supply as well as relative sea-level rise. This
retrogradational stacking pattern defines the transgressive systems tract (Boyd
et al., 1989).
Completion of a eustatic cycle results in highstand conditions similar to
those now extant. Larger fluvial systems, such as the Rio Grande (Fig. 3.1 l ) ,
 Deltaic coasts 109
Houma Bayou
Navigational Petit
c.nai Caillou ides
X Dernierer X
SL 0

100
0 50 100
Distance (km)

Figure 3.12. The Holocene Mississippi delta can be subdivided into a number of
eustatic systems tracts. Location of cross-section shown in Fig. 3.8 (after Boyd
et al., 1989).

Colorado and Brazos maintain relatively small, wave-dominated shelf-phase

deltas. These deltas are currently in transgressive phases. Most other streams
debouch into partially filled valleys, i.e., the estuaries of the modern coast, as
bayhead deltas. Such deposits are mostly fluvially dominated (Fig. 3.13).
Much of the area remains in the transgressive systems tract.
The Mississippi delta itself presents a somewhat different picture. During
the Wisconsinan lowstand, the Mississippi River incised a deep alluvial valley
across the continental shelf (Fisk, 1944), which ultimately breached the shelf
margin as a submarine canyon (Bouma & Coleman, 1985; Feeley et al., 1990;
Weimer, 1990). The youngest lobe of the Mississippi Fan received sediment
through this canyon, largely by mass-movement processes (Coleman, Prior, &
Lindsay, 1983; Goodwin & Prior, 1989).
The melting of the Wisconsinan glaciers brought relative sea level from
about 125 m below present (Berryhill, 1986) to about 20m below present by
about 9000 years BP (Frazier, 1974), and brought huge volumes of meltwater
and sediment down the river (Emiliani, Rooth & Stipp, 1978; Trainor &
Williams, 1991; Kolla & Perlmutter, 1993). The river filled its incised valley
and began depositing a number of progressively shallower shelf-phase deltas
(Boyd et al., 1989; Fig. 3.12). Backstepping of these deposits indicates that
sediment volume supplied by the river was unable to keep pace with the overall
eustatically driven transgression, and the deltas were probably developed
during stillstands. As rates of relative sea-level rise increased these earlier
Figure 3.13. Two new bayhead deltas of the Mississippi River are forming in Atchafalaya Bay: the Atchafalaya (right) and the Wax Lake (left)
lobes. The Atchafalaya, subaerial since 1973, shows substantial modification by the ship channel dredged down its main distributary. The Wax
Lake lobe, formed by an artificial outlet of the Atchafalaya river, displays a classic fluvially dominated morphology (photo courtesy of 0. Huh,
taken December 1990).
 Deltaic coasts 111

shelf-phase deltas were themselves transgressed and reworked (Frazier, 1967,

1974; Penland et al., 1988). Shelf-phase delta complexes onlap the Pleistocene
Type 1 unconformity with Holocene coastal sedimentary facies, comprising
the transgressive systems tract.
The culmination of eustatic rise occurred around 3 4 0 0 0 years BP in the
Mississippi delta and resulted in the retreat of the coastline to the mouth of the
Mississippi alluvial valley. The exact timing of the eustatic stillstand is
debatable and may not have begun until about 2500 years BP (Penland, Suter &
McBride, 1987). The subsequent development of the Holocene Mississippi
delta consists of a series of prograding delta complexes that began at the early
St. Bernard shoreline and advanced more than 150 km southeast
(Frazier, 1967) to the position of the currently active Balize lobe of the Modern
delta. Determination of the highstand systems tract of the Mississippi is based
on the change from retrogradational to progradational stacking. Three new
delta complexes have formed during the highstand, the Lafourche (active
3500-400 years BP), the Plaquemines-Modem (active 1000 years BP to present),
and the Atchafalaya (active c. 1950-present). The central region of the
Mississippi River delta plain contains the transgressive systems tract of
the Teche-Maringouin delta overlain by the highstand systems tract of the
Lafourche delta. The transgressive systems tract is separated from the
highstand systems tract by a regional ravinement surface (Penland et al.,
1987). During the highstand, sediment supply has easily outstripped regional
subsidence and any minor eustatic rise. Stacking of the St. Bernard, Lafourche
and Modern delta complexes has produced a progradational parasequence set,
making up the highstand systems tracts still in the process of formation. The
Atchafalaya complex is in the incipient phase of prograding over the older
Teche-Maringouin deltas.
Each of these complexes has evolved in a similar fashion, though the
process has been carried to different extents. During the initial phase of new
delta creation, the incipient lobe debouches onto foundering deposits of the
pre-existing delta. This primarily results in lacustrine delta deposition (cf. Tye
& Coleman, 1989). Once the lakes on the upper delta plain are filled, the new
bayhead deltas are deposited in the bays and estuaries created by subsidence
and transgression of the abandoned delta plain. The present Atchafalaya and
Wax Lake deltas (Fig. 3.13) typify this phase (Roberts et al., 1980; van
Heerden & Roberts, 1988). Continued progradation fills the bay and extends
seaward as a shelf-phase delta, typified by the currently abandoned Lafourche
complex (Fig. 3.8). Further sediment supply and progradation can then create
the circumstance which exits today in which the highstand delta has prograded
to the shelf margin (Figs. 3.5, 3.8).
112 J.R. Suter

Although the evolution of individual deltaic systems throughout the Late

Quaternary differs in detail, the above model provides a reasonable general
synthesis for many Late Quaternary deltas. Broadly similar evolutionary
histories are shown by the Tiber (Bellotti, Tortora, & Valeri, 1989), Ebro (Farran
& Maldonado, 1990), Niger (Oomkens, 1974), Nile (Chen, Warne, & Stanley,
1992; Stanley & Warne, 1993), Rhone (Tesson et al., 1990), Saloum (Ausseil-
Badie et al., 1991), and Mahakam deltas (D. Nummedal, pers. commun., 1993).

Deltas and coastal management

Human impacts upon coastal zones have increased dramatically through the
last half of the twentieth century. This is partly a result of increasing
population pressure and utilization of coastal regions, and partly due to the
ongoing sea-level rise These impacts are keenly felt in deltaic areas. A full
discussion of the complex issues of coastal management of deltaic areas is
beyond the scope of this chapter, but we can consider a few examples of how
deltaic processes directly impact human affairs.
Deltas form and evolve in a process-response mode. The natural
geomorphology of a given delta is the result of complex interactions of
sediment supply, relative sea-level changes, and marine reworking. Human
interference with any of these factors inevitably alters the form and evolution
of the delta. For example, widespread forest clearing and agricultural
utilization has led to dramatically increased sediment supply in some deltas.
It is estimated that heavy farming has increased the sediment load of the
Huanghe River of China by an order of magnitude over its early Holocene
level (Milliman et al., 1987). Such variations in sediment supply will
obviously have dramatic impacts on deltaic development.
The Nile delta has changed markedly in the Holocene and continues to
evolve rapidly. This evolution has been controlled by eustatic fluctuations,
climatic variations, wave reworking, neotectonics, and human activities
(Stanley, 1988). For millennia, Egyptian civilizations have depended on the
Nile River floods for irrigation and nutrient supply for agriculture and for
fisheries. Construction of the Aswan dams, commencing in 1902 and
culminating in the closure of the High Aswan dam in 1964, has completely
shut off the fluvial sediment supply to the coast. As a result, the Nile delta has
now entered an abandonment phase and is essentially a wave-dominated
coastal plain (Stanley & Warne, 1993; Sestini, 1989).
Stanley & Warne (1993) extrapolated coastal changes affecting the Nile
delta to predict its future evolution. Current deltaic degradation will continue
well into the twenty-first century, with potentially alarming consequences for
 Deltaic coasts 113
the population of Egypt. Deterioration of the delta could be slowed by massive
responses, such as construction of artificial wetlands, freshwater management,
sewage recycling, pollution control, and emplacement of coastal protection
structures (Stanley & Warne, 1993). Funding for such heroic undertakings is
an open question and, at current levels of population growth, perhaps even
such measures would be insufficient.
Several environmental geological problems of huge economic importance
confront residents of the Mississippi delta area. European settlement began
along the river in the early eighteenth century. Early settlers were aware of the
dangers of flooding, but took an active approach to make the delta plain more
conducive to their lifestyles, building levees and communities along the course
of the main distributary. Successive floods over the next several centuries
resulted in even higher and more extensive levees. A devastating flood in 1927
resulted in the completion of an extensive artificial levCe system to keep the
river in its banks and prevent the overflow of waters onto the communities of
the delta plain.
However, through construction of the levees, the natural processes of the
delta were drastically altered. Overbank flooding and sediment supply to the
delta plain were eliminated. Subsidence of the abandoned lobes of the
Mississippi delta is quite rapid, up to almost 2 c m a - ' . (Penland et al., 1988).
Depleted sediment supply from overbanking has accelerated the long-term
degradation of the deltaic plain, and the very existence of many communities
is threatened. Political struggles are underway to determine the response.
Initial plans focused on building artificial structures to maintain shorelines and
protect communities. More recently, solutions are being proposed which
attempt to work with the natural systems, ranging from small scale artificial
breaches in the levee systems to wholesale diversion of the Mississippi River
to a new route. These attempt to ameliorate land loss by providing new
sediment to subsiding areas.
Another major coastal management problem in the Mississippi delta relates
to the process of delta switching. Heavy settlement and industrial development
has occurred along the Mississippi River, including the cities of Baton Rouge
and New Orleans and one of the world's largest petrochemical complexes.
However, the current hydraulically favoured distributary is not the Lower
Mississippi itself, but the Atchafalaya River, located some 200 km to the west
(Fig. 3.8). Avulsion to this more efficient course is only prevented by
two massive River Control structures built by the US Army Corps of Engineers
above Baton Rouge, some 300 km north of the mouth of the Mississippi River.
At present, the control structures restrict flow down the Atchafalaya to 30%,
funnelling the remainder to the Lower Mississippi.
114 J.R. Suter
Should the control structures ever prove inadequate, dire economic and
human consequence would ensue, most of which probably cannot be
accurately predicted or appreciated. The city of New Orleans would be
deprived of its fresh water supply as the salt water wedge migrated up the then
abandoned lower Mississippi River. The industrial complex located along the
river would be deprived of fresh water for its operation. Land loss in the
eastern delta plain would accelerate to an even greater extent, increasing
the threats to coastal residents. Valuable wildlife habitat would be lost.
Navigation along the current Mississippi River would suffer. Various coastal
communities in the path of the ‘new’ distributary would require protection or
relocation. Transportation and other infrastructure would have to be repaired
or replaced. At the same time, renewed deltaic sedimentation and progradation
in western Louisiana would create a new landscape, a process already
underway (Penland & Suter, 1989).
These examples serve to illustrate the effects of deltaic processes on
human affairs. As people continue to migrate towards the coastal zone, it is
inevitable that further population increases and economic development will
increase demands upon deltaic areas and resources. Understanding of deltaic
processes is of paramount importance to minimizing deleterious impacts, and
managing the use of the shorelines.

Acknowledgements
I thank Exxon Production Research for permission to publish this paper. A
review by and discussions with Kevin Bohacs were very helpful. Numerous
ideas contained within this paper, particularly those dealing with Quaternary
sequence stratigraphy and deltaic evolution, have been developed through
collaborations and discussions with Henry Berryhill, Ron Boyd, Elisabeth
Kosters, Robert Morton, Dag Nummedal and Shea Penland. I also thank the
editors of this book, the late Bill Carter and Colin Woodroffe, for their patience
throughout the production of the manuscript.

References
Allen, G., Laurier, D. & Thouvenin, J. (1979). Erude sedimentologique du delta de
la Mahakam. Paris: TOTAL, Compagnie Francaise des Petroles, Notes et
Memoires 15. 156 pp.
Allen, G.P. (1991). Sedimentary processes and facies in the Gironde estuary: a
recent model for macrotidal estuarine systems. In Clastic tidal sedimentology,
ed. D.G. Smith, G.E. Reinson, B.A. Zaitlin & R.A. Rahmani, pp. 29-39.
Canadian Society of Petroleum Geologists Memoir 16.
 Deltaic coasts 115
Allen, J. R. L. (1970). Sediments of the modern Niger delta, a summary and review.
In Deltaic sedimentation modern and ancient, ed. J. P. Morgan, pp. 138-5 1.
Society of Economic Paleontologists and Mineralogists Special Publication 15.
Augustinus, P.G.E.G. ( 1 989). Cheniers and chenier plains: a general introduction.
Marine Geology, 90, 2 19-30.
Ausseil-Badie, J., Barusseau, J. , Descamps, C., Salif Diop, E. H., Giresse, P. &
Pazdur, M. (1991). Holocene deltaic sequence in the Saloum Estuary, Senegal.
Quaternary Research, 36, 178-94.
Bates, C. C. (1953). Rational theory of delta formation. American Association of
Petroleum Geologists Bulletin, 37, 21 19-62.
Bellotti, P., Tortora, P. & Valeri, P. (1989). Sedimentological and morphological
features of the Tiber Delta. Sedimentology, 36, supplement.
Bernard, H. A,, Major, C. F., Parrott, B. S. & LeBlanc, R. J., Sr. (1970).
Recent sediments of Southeast Texas. Texas Bureau of Economic Geology
Guidebook, 1 1.
Berryhill, H. L. (1986). Sea level lowstand features off southwest Louisiana. In Late
Quaternary facies and structure, northern Gulfof Mexico, ed. H. L. Berryhill,
pp. 225-40. American Association of Petroleum Geologists Studies in Geology
23.
Berryhill, H. L. & Suter, J. R. (1986). Deltas. In Lute Quaternary facies and structure,
northern Gulf of Mexico, ed. H. L. Berryhill, American Association of
Petroleum Geologists Studies in Geology 23.
Bhattacharya, J. & Walker, R.G. (1992). Deltas. In Facies models: response to sea
level change, ed. R. G. Walker & N. P. James, pp. 157-77. Geological
Association of Canada.
Boyd, R.L., Suter, J.R. & Penland, S. (1989). Relationship of sequence stratigraphy
to modern sedimentary environments. Geology, 17, 926-9.
Boyd, R. L., Dalrymple, R. W. & Zaitlin, B. A. (1992). Classification of clastic
coastal depositional environments. Sedimentary Geology, 80, 139-50.
Bouma, A. H. & Coleman, J. M. (1985). Mississippi Fan: Leg 96 program and
principal results. In Submarine fans and related turbidite systems, ed. A. H.
Bouma, W. R. Normark, & N. E. Barnes, pp. 247-52. New York: Springer-
Verlag.
Chen, Z., Warne, A. & Stanley, D. J. (1992). Late Quaternary evolution of
the northwestern Nile delta between the Rosetta Promontory and Alexandria,
Egypt: Journal of Coastal Research, 8, 527-61.
Coleman, J. M., Gagliano, S. M. & Smith, W. G. (1970). Sedimentation in a
Malaysian high-tide tropical delta. In Deltaic sedimentation: modern &
ancient, ed. J. P. Morgan & R. H. Shaver, pp. 185-97. Society of Economic
Paleontologists and Mineralogists Special Publication 15.
Coleman, J. M. & Prior, D. B. (1982). Deltaic environments. In Sandstone
depositional environments, ed. P. A. Scholle, & D. Spearing, pp. 129-78.
American Association of Petroleum Geologists.
Coleman, J. M., Prior, D. B. & Lindsay, J. F. (1983). Deltaic influences on shelf
edge instability processes. In The shelflreak: critical inteiface on continental
margins, ed. D. J. Stanley & G. T. Moore, pp. 121-37. Society of Economic
Paleontologists and Mineralogists Special Publication 33.
Coleman, J. M. & Wright, L. D. ( 1 975). Modern river deltas: variability of processes
and sand bodies. In Deltas: models for exploration, ed. M. L. Broussard, pp.
87-98. Houston, TX: Houston Geological Society.
Curray, J. R., Emmel, F. J. & Crampton, P. J. S. ( 1 969). Holocene history of a
strandplain, lagoonal coast, Nayarit, Mexico. In Coastal lagoons: a symposium,
116 J.R. Suter
ed. A. A. Castanares & F. B. Phleger, pp. 63-100. Mexico: Universidad
National Autonoma.
Dalrymple, R.W. ( 1 992). Tidal depositional systems. In Facies models: response to
sea level change, ed. R. G. Walker & N. P. James, pp. 195-218. Geological
Association of Canada.
Dalrymple, R.W., Zaitlin, B.A. & Boyd, R. (1992). Estuarine facies models:
conceptual basis and stratigraphic implications, Journal of Sedimentary
Petrology, 62, 1030-43.
Diesel, C.F.K. ( 1992). Coal-bearing depositional systems. New York: Springer-
Verlag. 721 pp.
Dominguez, J. M. L., Bittencourt, A. C. S. P. & Martin, L. ( 1 992). Controls on
Quaternary coastal evolution of the east-northeastern coast of Brazil: roles of
sea-level history, trade winds and climate. Sedimentary Geology, 80, 2 13-32.
Dominguez, J. M. L., Martin, L. & Bittencourt, A. C. S. (1987). Sea-level history
and Quaternary evolution of river mouth-associated beach-ridge plains along
the east-southeast Brazilian coast: a summary. In Sea-1evelJluctuations C?
coastal evolution, ed. D. Nummedal, 0. H. Pilkey & J. D. Howard, pp. 115-27.
Society of Economic Paleontologists and Mineralogists Special Publication 41.
Dominguez, J. M. L. & Wanless, H. R. (1991). Facies architecture of a falling sea-
level strandplain, Doce River coast, Brazil. In Shelfsand and sandstone bodies:
facies and sequence stratigraphy, ed. D. J. P. Swift, G. F. Oertel, R. W. Tillman
& J. A. Thorne, pp. 259-8 I . International Association of Sedimentologists
Special Publication 14.
Edwards, M. B. (1981). Upper Wilcox Rosita delta system of south Texas: growth-
faulted shelf edge deltas, American Association of Petroleum Geologists
Bulletin, 65, 1574-85.
Emiliani, C., Rooth, C. & Stipp, J.J. (1978). The late Wisconsinan flood into the Gulf
of Mexico. Earth and Planetary Science Letters, 41, 159-62.
Ethridge, F. G. & Wescott, W. A. (1984). Tectonic setting, recognition and
hydrocarbon reservoir potential of fan-delta deposits. In Sedimentology of
gravels and conglomerates, ed. E. H. Koster & R. J. Steel, pp. 217-35. Memoir
Canadian Society of Petroleum Geologists 10.
Elliot, T. ( 1 986). Deltas. In Sedimentary environments and facies, ed. H. G. Reading,
pp. 1 13-54. Oxford: Blackwell Scientific Publications.
Elliot, T. (1989). Deltaic systems and their contribution to an understanding of basin-
fill successions. In Deltas: sites and traps for fossil fuels, ed. M. K. G.
Whateley & K. T. Pickering, pp. 3-10. Oxford, Blackwell Scientific
Publications. Geological Society Special Publication 41.
Farran, M. & Maldonado, A. ( 1 990). The Ebro continental shelf Quaternary
seismic stratigraphy and growth patterns. Marine Geology, 95, 289-3 12.
Feeley, M. H., Moore, T.C., Jr., Loutit, T.S. & Bryant, W.R. (1990). Sequence
stratigraphy of Mississippi Fan related to oxygen isotope sea level index.
American Association of Petroleum Geologists, 74,407-24.
Fisher, W. L., Brown, L. F., Jr., Scott, A. J. & McGowen, J. H. ( 1 969). Delta systems
in the exploration for oil and gas: a research colloquium, The University of
Texas at Austin, Bureau of Economic Geology. 2 12 pp.
Fisk, H. N. ( 1 944). Geological investigation of the alluvial valley of the lower
Mississippi River. Vicksburg: Mississippi River commission. 78 pp.
Frazier, D. E. (1967). Recent deltaic deposits of the Mississippi delta: their
development and chronology. G u y Coast Association of Geological Societies,
Transactions, 17, 287-3 15.
 Deltaic coasts 117
Frazier, D. E. ( 1974). Depositional episodes: their relationship to the Quaternary
stratigraphic framework in the northwestern portion of the Gulf basin, Austin,
Texas, Bureau of Economic Geology, Geologic Circular 74-1. 28 pp.
Galloway, W.E. ( 1 975). Process framework for describing the morphologic &
stratigraphic evolution of deltaic depositional systems. In Deltas, models
for exploration, ed. M. L. Broussard, pp. 87-98. Houston, TX: Houston
Geological Society.
Gastaldo, R. A. & Huc, A. (1992). Sediment facies, depositional environments,
and distribution of phytoclasts in the Recent Mahakam River delta, Kalimantan,
Indonesia. Paliaos, 7, 574-90.
Gilbert, G.K. (1885). The topographic features of lake shores. United States
Geological Survey, Annual Report, 5 , 69-1 23.
Goodwin, R. & Prior, D. B. (1989). Geometry and depositional sequences of the
Mississippi Canyon, Gulf of Mexico. Journal of Sedimentary Petrology, 59,
3 18-29.
Haszeldine, R. S . ( 1 989). Coal reviewed: depositional controls, modern analogs,
and ancient climates. In Deltas: sites and traps for fossil fuels, ed. M. K. G.
Whateley & K. T. Pickering, pp. 3-10. Oxford: Blackwell Scientific
Publications. Geological Society Special Publication 41.
Holmes, A. (1965). Principles ofphysical geology, 2nd edn, London: Thomas Nelson
and Sons. 1288 pp.
Horne, J. C., Ferm, J. C., Caruccio, F. T. & Baganz, B. P. (1978). Depositional models
in coal exploration and mine planning in Appalachian region. American
Association of Petroleum Geologists Bulletin, 6 2 , 2 3 7 9 4 1 I .
Kolb, C. R. & Dornbusch, W. K. (1975). The Mississippi and Mekong Deltas: a
comparison. In Deltas, models for exploration, ed. M. L. Broussard, pp.
193-208. Houston, TX: Houston Geological Society.
Kolla, V. & Perlmutter, M. A. (1993). Timing of turbidite sedimentation on the
Mississippi Fan. American Association of Petroleum Geologists Bulletin, 77,
1129-41.
Kosters, E. C. ( 1989). Organic-clastic facies relationships and chronostratigraphy of
the Barataria interlobe basin, Mississippi delta plain. Journal of Sedimentary
Petrology, 59,98-113.
Kosters, E. C., Chmura, G. L. & Bailey, A. (1987). Sedimentary and botanical
factors influencing peat accumulation in the Mississippi Delta. Journal of
Sedimentary Petrology, 144,423-34.
Kosters, E. C. & Suter, J. R. (1993). Facies relationships and systems tracts in the
Late Holocene Mississippi delta plain. Journal ofSedimentary Petrology,63,727-33.
Lehner, P. (1969). Salt tectonics and Pleistocene stratigraphy on continental slope
of northern Gulf of Mexico. American Association of Petroleum Geologists
Bulletin, 53, 243 1-79.
Lu, Z., Suhayda, J., Prior, D. B., Bornhold, B. D., Keller, G. H., Wiseman, W. J.,
Wright, L. D. & Yang, Z. S. (1991). Sediment thixotropy and submarine mass
movement, Huanghe delta, China. Geomarine Letters, 11, 9-1 5.
McCabe, P. J. ( 1 984). Depositional environments of coal and coal-bearing sequences.
In Sedimentology of coal and coal bearing sequences, ed. R. A. Rahmani, &
R. M. Flores, pp. 13-42. International Association of Sedimentologists Special
Publication 7.
McCabe, P.J. (1991). Geology of coal: environments of deposition. In Economic
geology, U S . , The geology of North America volume P-2, ed. H. J. Gluskoter,
D. D. Rice, & R. B. Taylor, pp. 469-82. Geological Society of America.
118 J.R. Suter
McCabe, P. J., Breyer, J. A., Kosters, E. C., Chmura, G. L. & Bailey, A. (1989).
Discussion of sedimentary and botanical factors influencing peat accumulation
in the Mississippi Delta. Geological Society of London Journal, 146, 877-80.
McPherson, J. G., Shanmugan, G. & Moiola, R. J. (1987). Fan-deltas and braid deltas:
varieties of coarse-grained deltas. Geological Society of America Bulletin, 99,
331-40.
Meckel, L. D. (1975). Holocene sand bodies in the Colorado delta, Salton sea,
Imperial County, California. In Deltas, models for exploration, ed. M. L.
Broussard, pp. 239-66. Houston, TX: Houston Geological Society.
Milliman, J. D., Qin Y., Ren, M. & Saito, Y. (1987). Man's influence on the erosion
and transport of sediment by Asian rivers: the Yellow River (Huanghe)
example. Journal of Geology, 95,75 1-62.
Moore, D. G. & Curray, J. R. (1 964). Sedimentary framework of the drowned
Pleistocene delta of Rio Grande de Santiago, Nayarit, Mexico. In Developments
in sedimentology, I, Deltaic and shallow marine deposits, ed. L. M. J. U. van
Straaten, pp. 277-8 I . Amsterdam: Elsevier.
Morgan, J. P. (1970), Depositional processes and products in the delta environment,
In Deltaic sedimentation modern & ancient, ed. J. P. Morgan, pp. 3 1-47.
Society of Economic Paleontologists and Mineralogists Special Publication 15.
Morton, R. A., Sams, R. H. & Jirik, L. A. (1991). Plio-Pleistocene depositional
sequences of the southeastern Texas continental shelf geologic framework,
sedimentary facies and hydrocarbon distribution. Texas Bureau of Economic
Geology Report of Investigations 200. 90 pp.
Nemec, W. (1990). Deltas: remarks on terminology and classification. In Coarse-
grained deltas, ed. A. Colella & D. B. Prior, pp. 3-12. International
Association of Sedimentologists Special Publication 10.
Nittrouer, C. A., Kuehl, S.A., DeMaster, D.J. & Kowsmann, R.O. (1986). The
deltaic nature of Amazon shelf sedimentation. Geological Society of America
Bulletin, 97,444-58.
Oomkens, E. (1970). Depositional sequences and sand distribution in the postglacial
Rhone delta complex. In Deltaic sedimentation modern & ancient, ed. J. P.
Morgan, pp. 198-2 12. Society of Economic Paleontologists & Mineralogists
Special Publication 15.
Oomkens, E. (1974). Lithofacies relations in the Late Quaternary Niger Delta
complex. Sedimentology, 21, 195-22 1.
Orton, G. J. (1988). A spectrum of Middle Ordovician fan deltas and braidplain deltas,
North Wales: a consequence of varying fluvial clastic input. In Fan deltas:
sedimentology and tectonic settings, ed. W. Nemec & R. J. Steel, pp. 3-13.
London: Blackie & Son.
Orton, G. J. & Reading, H. G. (1993). Variability of deltaic processes in terms of
sediment supply, with particular emphasis on grain size. Sedimentology, 40,
475-5 12.
Penland, S., Boyd, R.L & Suter, J. R. (1988). Transgressive depositional systems of
the Mississippi delta plain: a model for barrier shoreline and shelf sand
development. Journal of Sedimentary Petrology, 58,932-49.
Penland, S. & Suter, J. R. (1989). The geomorphology of the Mississippi River
Chenier Plain. Marine Geology, 90,231-58.
Penland, S., Suter, J. R. & McBride, R. A. (1987). Delta plain development and sea
level history in the Terrebonne coastal region, Louisiana. In Coastal sediments
'87, pp. 1689-705. New Orleans, LA: Water Ways Division, American Society
of Coastal Engineers.
 Deltaic coasts 119
Posamentier, H.W. & Vail, P.R. (1988). Eustatic controls on clastic deposition. 11.
Sequence and systems tract models. In Sea-level changes: an integrated approach.
ed. C.K. Wilgus, B.S. Hastings, C.G. St C. Kendall, H.W. Posamentier, C.A.
Ross & J.V. van Wagoner, pp. 125-1 54. Society of Economic Paleontologists
and Mineralogists, Special Publication 42.
Postma, G. (1990). Depositional architecture and facies of river and fan deltas:
a synthesis. In Coarse-grained deltas, ed. A. Colella, & D. B. Prior, pp. 13-27.
International Association of Sedimentologists Special Publication 10.
Reading, H. G. (1986a). Facies. In Sedimentary environments and facies (2nd edn),
ed. H. G. Reading, pp. 4-19. London: Blackwell Scientific Publications.
Reading, H. G., ed. (1986b). Sedimentary Environment and Facies (2nd edn).
London: Blackwell Scientific Publications. 615 pp.
Roberts, H. H., Adams, R. D. & Cunningham, R. H. W. (1980). Evolution of sand-
dominant subaerial phase, Atchafalaya delta, Louisiana. American Association
of Petroleum Geologists Bulletin, 64, 264-79.
Russell, R. J. (1936). Physiography of lower Mississippi River delta. In Reports on
the geology of Plaquemine and St. Bernard Parishes, Louisiana. pp. 3-199.
Louisiana Department of Conservation, Geological Bulletin 8.
Ryer, T.A. (1981). Deltaic coals of the Ferron sandstone member of Mancos
shale: predictive model for Cretaceous coal-bearing strata of Western Interior
American Association of Petroleum Geologists Bulletin, 65,2223-340.
Scruton, C. (1960). Delta building and the deltaic sequence, In Recent sediments,
Northwest Gulfof Mexico, 1951-1958, ed. F. Shepard et al., pp. 82-102.
American Association of Petroleum Geologists Special Publication.
Scott, A. C. (1989). Deltaic coals: an ecological and palaeobotanical perspective. In
Deltas: sites and traps f o r fossil fuels, ed. M. K. G. Whateley & K. T.
Pickering, pp. 3-10. Oxford: Blackwell Scientific Publications. Geological
Society Special Publication 41.
Sestini, G. (1989). Nile delta: a review of depositional environments and geological
history. In Deltas: sites & traps f o r fossil fuels, ed. M.K.G. Whateley & K.T.
Pickering, pp 99-127. Oxford Blackwell Scientific Publications, Geological
Society Special Publication 41.
Stanley, D.J. (1988). Subsidence in the northeastern Nile delta: rapid rates, possible
causes and consequences. Science, 240,497-500.
Stanley, D.J. & Warne, A.G. (1993). Nile delta: recent geological evolution and
human impact. Science, 260, 628-34.
Styan, W.B. & Bustin, R.M. (1983). Sedimentology of Fraser River Delta peat
deposits: a modern analogue for some deltaic coals. International Journal of
Coal Geology, 3, I0 1-43.
Suter, J. R. (199 I). Late Quaternary chronostratigraphic framework, northern Gulf
of Mexico. In 12th Annual Research Conference Proceedings. Gulf Coast
Section, Society of Economic Paleontologists and Mineralogists.
Suter, J. R. & Berryhill, H.L. (1985). Late Quaternary shelf-margin deltas, northwest
Gulf of Mexico. American Association of Petroleum Geologists Bulletin, 69,
77-9 I .
Suter, J. R. & Morton, R. A. (1989). Coastal depositional systems, northwest Gulf
of Mexico. 28th International Geological Congress Field Trip Guidebook
T370. 52 pp.
Tesson, M., Gensous, B., Allen, G.P. & Ravenne, C. (1990). Late Quaternary
deltaic lowstand wedges on the Rhone continental shelf, France. Marine
Geology, 91, 325-32.
120 J.R. Suter
Trainor, D. M. & Williams, D. F. (1991). Quantitative analysis and correlation of
oxygen isotope records from planktonic and benthonic foraminifera and well
log records from OCS Well G I267 No. A- 1, South Timbalier Block 198, north-
central Gulf of Mexico. Gulf Coast Section-SEPM 12th Annual Research
Conference Proceedings, pp. 363-77.
Tye, R. S. & Coleman, J. M. (1989). Depositional processes and stratigraphy of
fluvially dominated lacustrine deltas: Mississippi delta plain. Journal of
Sedimentary Petrology, 59,973-96.
Umitsu, M. ( 1 993). Late Quatenary sedimentary environments and landforms in the
Ganges Delta. Sedimentary Geology, 83, 177-86.
van Andel, T.J. (1967). The Orinoco delta, Journal of Sedimentary Petrology,
37,297-3 10.
van Heerden, I. L. & Roberts, H. H. (1988). Facies development of Atchafalaya
Delta, Louisiana: a modern bayhead delta. American Association of Petroleum
Geologists Bulletin, 72,439-53.
van Wagoner, J.C., Mitchum, R.M., Campion, K.M. & Rahmanian, V.D.
( 1990). Siliciclastic sequence stratigraphy in well-logs, cores, and outcrops.
American Association of Petroleum Geologists, Methods in Exploration Series
7 , 5 5 pp.
Walker, R. G. (1992). Facies, facies models, and modern stratigraphic concepts. In
Facies models: responses to sea level change, ed. R. G. Walker & N. P. James,
pp. 1-14. Geological Association of Canada.
Weimer, P. ( 1990). Sequence stratigraphy, facies geometries, and depositional history
of the Mississippi Fan, Gulf of Mexico. American Association of Petroleum
Geologists Bulletin, 74,425-53.
Winker, C. D. & Edwards, M. B. (1983). Unstable progradational clastic shelf
margins. In The shelfbreak: critical interface on continental margins, ed. D. J.
Stanley, & G. T. Moore, pp. 139-57. Society of Economic Paleontologists and
Mineralogists Special Publication 33.
Wright, L. D. (1977). Sediment transport and deposition at river mouths: a
synthesis. Geological Society of America Bulletin, 88, 857-68.
Wright, L. D. (1985). River deltas. In Coastal sedimentary environments, ed. R. A.
Davis, Jr., pp. 1-76. New York: Springer-Verlag.
Wright, L. D., Wiseman, W. J., Bornhold, B. D., Prior, D. B., Suhayda, J., Keller, G.
H., Yang, Z. S. & Fan, Y. B. (1988). Marine dispersal and deposition of Yellow
River silts by gravity-driven underflows. Nature, 332, 629-32.
 Wave-dominated coasts
P.S. ROY, P.J. COWELL, M.A. FERLAND AND B.G. THOM

Introduction
Wave-dominated sedimentary coasts comprise accumulations of detrital sand
and gravel-sized material which undergo high levels of physical reworking,
interspersed with periods of burial before finally being deposited as the coastal
deposits we see today (Davis & Hayes, 1984). Quite commonly sediments tend
to be of clean sand and gravel, often quite well sorted and abraded, containing
relatively high proportions of more resistant minerals and rock types such as
quartz, chert and heavy minerals. Waves and wave-induced currents are the
dominant mechanisms for moving and depositing sand on shorefaces and
beaches of the open coast, although winds, river discharge, tidal currents and
Ekman flows variously act as transporting agents landward of the beach, in
estuaries and seaward of the shoreface. In relation to the shoreface and beach,
open coastal types are determined by four factors: (i) substrate gradient, (ii)
wave energy versus tidal range; (iii) sediment supply versus accommodation
volume (Swift & Thorne, 1991); and (iv) rates of sea-level change. At one
extreme are steep, high-energy, sediment-deficient coasts that have bedrock
cropping out as headlands, with negligible sand at their base and relatively
deep water offshore (autochthonous, accommodation-dominated coast of
Swift & Thorne, 1991). At the other extreme are low-gradient, low-energy
coasts that are typically muddy with a coastal fringe of wetland vegetation.
Here, incident wave action is dissipated over very shallow offshore gradients
such as those associated with deltaic environments at river mouths
(allochthonous, sediment-supply dominated coast of Swift & Thorne, 1991;
see Chapter 3). But even here, rare high-energy events such as cyclones can
cause episodes of wave reworking leading to the formation of cheniers.
The objectives of this chapter are to examine the evolution of wave-
dominated coasts through the Late Quaternary, especially the Holocene, and to
demonstrate how present coastal geomorphologies and stratigraphies are

121
122 P.S.Roy et al.
predicated upon this evolution. The focus therefore is on principles involving
the four controls enumerated above, and on how, as indicated by Pilkey et al.
(1993), geological inheritance plays a much more important role than
suggested in some previous reviews. Nonetheless, reviews such as Swift,
Phillips & Thorne (199 1a) provide a comprehensive basis for the concepts
presented here, which are drawn from case studies in southeast Australia
(24.5" to 38" S ) . The variety of substrate slopes that have existed on this shelf
and coast at different Quaternary sea levels furnish backgrounds against which
different types of wave-dominated coastal development can be evaluated. The
southeast Australian shelf presents a substrate ranging from very low angle
(0.1-0.2") on the outer shelf to relatively steep (>1 .Oo) immediately seaward of
the present shoreline (Wright, 1976), with very steep slopes (>5") occurring
locally. In contrast, gradients on other well-studied coasts, such as those of the
US Atlantic and Gulf of Mexico, are very low angle (<0.1"), immediately
seaward of the present coast. The natural variability of the southeast Australian
coast and shelf therefore affords valuable experimental control. At lower sea
levels in the Late Quaternary, a low-gradient, open coast prevailed. At sea
levels approaching those of today, the coast is much steeper with a wide range
of open and embayed coastal sectors (Roy & Thom, 1981).
Two approaches are applied in order to develop the concepts in this chapter.
The first is based upon extensive field investigations of the Quaternary
geology on the southeast Australian coast and continental shelf, involving
surface sediment sampling, vibrocores, side-scan sonar, and high-resolution
seismic reflection data. Granulometric analyses and radiocarbon dates of
numerous vibrocore samples provide the necessary ground-truthing for
interpreting the seismic records. These investigations include
morphostratigraphic surveys of present-day barriers (Roy & Stephens, 1980;
Roy & Thom, 198 1; Roy & Crawford, 1981; Thom, 1984; Thom et al., 1992),
shoreface (Gordon & Roy, 1977; Roy 1985), and inner shelf (Roy & Crawford,
1980; Roy & Thom, 1981; von Stackelberg, 1982; Field & Roy, 1984; Cowell
& Nielsen, 1984; Hudson, 1985; Roy, 1985; Cowell, 1986; Hudson & Ferland,
1987; Roy &Hudson, 1987; Ferland, 1986,1987, 1988,1990; Roy & Ferland,
1987), as well as former environments at lower sea levels that are partly
preserved on the present inner and mid shelf areas (Kudrass, 1982; Roy 1985;
Roy & Hudson, 1987; Roy et al., 1992; Bickford et al., 1993; Roy & Keene,
1993). One of the problems of investigating coastal evolution is that much of
the field evidence is not preserved, as deposits have been reworked and
modified through the action of evolutionary processes. Computer modelling
therefore provides the second approach underpinning the principles examined
in this chapter. The fundamental rationale for this computer modelling is based
 Wave-dominated coasts 123

on principles of large-scale coastal behaviour (Terwindt & Wijnberg, 1991)

and is outlined in Chapter 2 (also see Cowell, Roy & Jones, 1991, 1992).
The scope of this chapter does not extend to consideration of the role of
tectonics and continental-margin setting in basin evolution (see Chapter 12).
Relationships at this scale are assumed to be implicit in variables such as
substrate slope and the degree to which coasts are laterally bounded by
headlands or drift divides. Similarly, climatic variables are ignored as are the
influences of tides and wind-driven currents. Furthermore, coastal evolution is
primarily examined in terms of its two-dimensional behaviour in the vertical
plane perpendicular to the coastline using a computer simulation model
(Cowell et al., 1992). The model expresses the dynamic complexity of the
coast in simple shape terms employing the principle of conservation of sand
mass. Barrier shape is empirically defined by active surfaces (shoreface and
barrier top) that reflect ambient conditions and thus can act as a surrogate for a
site-specific set of hydrodynamic processes and sediment properties. The role
of coastline configuration in plan form is of first-order importance to the
evolution of wave-dominated coasts, especially given the significance of
littoral sand transport in the surf zone (Wright et al., 1991) and its effects on
shoreline adjustment (Komar, 1976; Gordon, 1988). The effects of shoreline
configuration are examined through their implicit control over coastal-profile
response, via the alongshore sediment budget which, in turn, primarily
depends upon coastal plan form and the direction of incident wave energy
(Davies, 1958; Komar, 1976).
Description of the lithofacies found on wave-dominated coasts is provided
only to the extent required in considering evolutionary processes. Such
descriptions can be found elsewhere (e.g., Swift, 1976a; Greenwood & Davis,
1984; Davis, 1985, 1992; Niedoroda, Swift & Hopkins, 1985; Swift et al.,
199 la). Present-day studies of barrier beaches and tidal inlets are invariably
dominated by process measurements and historical reconstructions. While
these studies are invaluable in documenting the mechanics of short-term
coastal change, they do not address the broad questions of barrier genesis and
evolution over geological time spans of one or more glacial-interglacial cycles.
The principles presented here are, for the most part, grossly simplified.
Notably in this respect, coastal evolution in response to marine transgression is
presented in relation to the high rates of postglacial sea-level change that
occurred 20 to 7 kaBP. While it may be possible to draw implications for the
evolution of wave-dominated coasts subject to slower rates of sea-level rise,
such as that occurring along the present US Atlantic coast, substantial
differences exist due to local variations in sediment budgets. It is probably
helpful to keep these differences in mind when comparing concepts developed
124 P.S. Roy et al.
in this chapter with models presented by others such as D.J.P. Swift (e.g., Swift
& Thorne, 1991; Thorne & Swift, 1991a, 1991b) and J.C. Kraft (e.g., Kraft,
197 1; Kraft & John, 1979; Belknap & Kraft, 1985). The question of time scale
underpinning different models for evolution of wave-dominated coasts is also
important. Many aspects of previous ideas are based upon the long-term and
large-scale perspective of basin infill studies, which incorporate modern-
analogue processes into geological time scales (e.g., Haq, Hardenbol & Vail,
1987; Vail e t a l . , 1991; Swift, Phillips & Thorne, 1991b). The critical question
here is whether pre-Quaternary coasts experienced glacial-interglacial sea-
level oscillations of anything like the magnitude and frequency of those of the
last 700 ka. Its answer will determine the extent to which lessons from the
Holocene can be extrapolated into the geological past. The perspective of this
chapter, however, concerns the much more rapid sea-level changes
characteristic of Late Pleistocene and early Holocene conditions that are more
immediately relevant to the antecedents of present-day wave-dominated
coasts.

Background
In the past, wave-dominated coasts have been described in terms of beaches
(Zenkovich, 1967; King, 1972), the nature of which was seen to be solely the
product of contemporary wave processes (Pilkey et al., 1993). In contrast, and
in keeping with the evolutionary focus of this book, our emphasis is
geological: we propose to use the term ‘coastal sand barrier’ as the basic
depositional element for wave-dominated coasts. The concepts of geological
inheritance, as well as morphodynamics, are used to describe their nature.
Within this framework, the beach is seen as only part of a larger entity.
Following the classifications of Curray (1 969) and Boyd, Dalrymple, & Zaitlin
(1992), barriers are defined as elongated, shore-parallel sand bodies, extending
above sea level, and consisting of a number of sandy lithofacies including
beach, dune, shoreface, tidal delta, inlet and washovers. Together they form a
continuum of barrier morphologies (Fig. 4.1). Barrier islands with wide
lagoons are the end member on low-gradient, open coasts (Davis, 1992), but as
substrates progressively steepen, the lagoon becomes narrower so that the
special case of a mainland beach constitutes the other end member, with
negligible backbarrier morphologies (Fig. 4.1). Between these end members
lie bay barriers (King, 1972, p. 502) which reflect the degree to which the coast
is laterally bounded (Fig. 4.1 ). Embayedness generally relates to the extent of
bedrock outcrop. This can be expected to become more frequent where
regional substrates are steeper, as is the case in southeast Australia where the
 Wave-dominated coasts 125
headland spit

Figure 4. I . Both sediment-rich and sediment-starved settings exist on wave-dominated

coasts. These range from low-lying coastal strand plains and barrier island/lagoon
systems to rocky coasts with bay barriers and estuaries in embayments and rock reefs
on the sea bed off headlands. The insets show in more detail the settings and
stratigraphies of headland and barrier island spits. (ssb: shelf sand body; bb: bay barrier;
bi: barrier island; td: tidal delta; sp: strand plain)

sea intersects erosional bedrock landforms comprising drowned valleys and

interfluves that form headlands (Fig. 4.1) (Roy & Stephens, 1980; Roy &
Thom, 1981). Submarine deposits of shelf sand seaward of the shoreface zone
are residuals of coastal deposits formed during lower phases of sea level and
are reworked to varying degrees by waves and shelf flows (Everts, 1987;
Niedoroda et al., 1985; Rine el al., 199 1).
Wave-dominated coasts correspond mainly to Swift & Thorne’s (1991)
‘accomodation-dominated’ coastal settings in which basins receiving
sediments are volumetrically large compared with the rate at which sediment
is supplied to them. Nevertheless, many ‘supply-dominated’ (allochthonous)
coasts are presently building sandy deltas with wave-dominated littoral zones;
the S5o Francisco in Brazil (see Fig. 3.6) is an example (Wright, 1985). The
main distinction between barriers in accomodation- as opposed to supply-
dominated coastal settings is the thickness of their lithofacies, especially on the
shelf (Fig. 4.2). The former are associated with autochthonous shelves that
have sandy sediments and experience extensive reworking (Fig. 4 . 2 and ~ 6). In
126 P.S. Roy et al.

headland
SL

wide thin barrier

Qeotte
b
inner shelf
sands

narrow
thick barrier
rock reefs
C

river
SL

alluvium
marine
muds

Figure 4.2. Barrier shape is influenced by embayment geometry and accommodation

space. Even where barrier sand volumes are the same, very different surface
morphologies are produced in embayments with shallow, gently sloping substrates ( a )
compared with embayments with deeper and steeper slopes ( b ) , that probably
experience higher energy waves than (a).In supply-dominated (allochthonous) settings
(c) with the same wave regime as (b),a similar-sized barrier sand body shows markedly
different stratigraphic relationships. The small arrows indicate the toe of the shoreface
which, because the barrier sand bodies are the same size but energy levels are different,
occurs in deeper water, further offshore in (b) compared to (a); in ( c ) the toe of the
shoreface is indeterminate. The width of each section are in the order of 5-10 km and
the maximum water depths are 10-30 m. SL =sea level.
 Wave-dominated coasts 127
contrast, supply-dominated regimes produce allochthonous shelves
characterised by more rapid sedimentation and thick muddy sequences in
which barrier sand bodies may be encased (Fig. 4 . 2 ~ ) .However, Swift &
Thorne’s (1 99 1) reliance on contemporary processes to discriminate between
different types of coastal deposits takes insufficient account of the influence of
inherited geological phenomena or antecedent topography (Belknap & Kraft,
1985; Pilkey et al., 1993; Riggs & Cleary, 1993). As Fig. 4 . 2 ~and b shows,
even with the same sediment supply, variations in accommodation space
(specifically, in substrate slope) can produce different coastal geometries and
stratigraphies.
Table 4.1 is an alternative framework based on inherited landscape, but
without implying a specific tectonic setting (see also Short, 1988). This
tripartite division is a way of grouping coastal deposits in plan form; features
on coastal plain (unbounded) coasts generally have the longest dimensions
coast-parallel while those on embayed (bounded) or cliffed coasts are smaller
and more variable lithologically. There are also cross-sectional differences,
between the three classes of coasts, in the slope or gradient of the pre-existing
land surface on which contemporary coastal deposits have formed. In general
terms, slopes are steepest on cliffed or protruding coastal sectors and flattest on
coastal plain coasts; slopes are most variable on embayed coasts.
For more than a century the origin of barriers and barrier islands has been
debated (e.g., Hoyt, 1967; Schwartz, 1973; Otvos 1979). Emergent shoals,
submergent bars and spit progradation were some of the early hypotheses.
Some workers recognised the importance of the postglacial marine
transgression (e.g., Curray, 1964; Kraft, 1971; Leatherman, 1983), and now it
is widely acknowledged that, rather than being solely the product of
interglacial highstands of the sea, barriers have also existed at lower sea levels
(Swift, 1975a; Hovland & Dukefoss, 1981; Field & Trincardi, 1991; Roy et al.,
1992). Vibrocores that encountered relict estuarine sediments in present-day
shelf environments provide evidence for the existence of barrier lagoons when

Table 4.1. Regional coastal depositional morphologies

Coastal plain coasts Embayed coasts Protruding or

(unbounded) (bounded) cliffed coasts

Linear barriers Bay barriers Cliff-top dunes

Barrier islands Headland spits Mainland beaches
Shelf ridge fields and Inner shelf sand sheets Headland - attached shelf
inner shelf sand sheets sand bodies
128 P.S. Roy et al.

estuarine and
fluvial deposits
present day barriers
present SL

0-6ka former transgressive

barriers
8ka
1Oka

T
I
c.50m
c.5km
I
12ka relict estuarine
muds

Figure 4.3. Continental shelf coring investigations such as those in southeast Australia
encounter relict estuarine sediments in the subsurface with ages which indicate they
formed as the shelf was inundated by the postglacial marine transgression. Estuarine
and lagoonal environments only exist in the protection of barriers which means that
these latter features must also have existed at times of lower sea levels. The lack of open
marine deposits behind barriers at the present coast accords with their onshore
migration as sea level rose, rather than by in situ accretion when sea level stabilised.

sea level was lower (Sanders & Kumar, 1975; Field & Duane, 1976; Williams,
1976; Kudrass, 1982; Colwell & Roy, 1983) (Fig. 4.3). Furthermore, the
absence of open marine sediments on the mainland shore behind barriers
indicates that, rather than emerging in place as offshore bars, most barriers
migrated shorewards during the postglacial marine transgression from an
initial position on the outer shelf (Hoyt, 1967; Swift, 1975a; Field & Duane,
1976). Thus, modern models of barrier formation must accommodate sea-level
oscillations of the order of lOOm every 100000 years or so over the last
400000 years (Chappell & Shackleton, 1986). For non-tectonic coasts, this
means that shorelines were lower than present levels for more than 90% of
Quaternary time (and occupied a modal zone between -25 and -75 m; Roy &
Thom, 1981) and that two of the major variables in barrier formation - shelf
morphology and sediment supply - are essentially imposed on contemporary
barriers through inheritance.
There is an important distinction between the flux of sediment on wave-
dominated coasts and the movement of barrier sand masses during sea-level
transgressions, stillstands and regressions. As distinct from headland spits,
which grow alongshore (Fig. 4.1, inset), barriers translate across the shelf in a
shore-normal sense as sea levels change, although both alongshore and shore-
normal sediment fluxes are involved in their translation. The emplacement of
barriers is seen as a complex response to changes in the rate of sea-level rise,
substrate gradient, sediment supply and energy conditions. However, the
 Wave-dominated coasts 129
geological continuity (and longevity) of barrier shorelines may be interrupted
in particular cases: high latitude coasts that were permanently ice covered
during glacials are one example (see Chapter 9) and shallow, land-locked seas
(e.g., the North Sea and Gulf of Thailand) that were exposed as dry land during
sea-level lowstands are another. But even in these cases, wave action re-
established its dominance as the ice retreated and sea levels rose with the
postglacial marine transgression. Overall, however, wave-dominated deposits
of various types have existed either continuously or for substantial periods of
time on most of the world’s coasts throughout the Quaternary. It is not
surprising, therefore, that the rapid eustatic changes that have characterised
this geological period have had an important influence on the style of present-
day coastal development.
More is known about coastal development under stable or slowly changing
sea levels than when sea level is changing rapidly, because we are presently in
an interglacial period. Periods of changing sea level were more common in the
past, but our understanding of the events accompanying them is more
speculative and relies on indirect techniques such as simulation modelling
supported by limited field evidence at specific localities. In Fig. 4.4a, various
types of wave-dominated deposits are arranged according to: 1. whether
relative sea level (RSL) is rising, falling or stable, and 2. the morphology of the
land surface over which the sea is moving. However, because many
continental shelves tend to be more rugged on their inner parts and planar
elsewhere, low-gradient coastal morphologies were much more common at
times of lower sea level than they are today. This general relationship is shown
schematically in Fig. 4.4b for southeast Australia and indicates that, for most
of the Quaternary, coastal settings during both marine transgressions and
regressions have been low-gradient and planar with linear barrier shorelines.
Cliffed and embayed coasts probably represented only a minor proportion of
the region’s coastal settings over this time period.
The following sections use field studies from southeast Australia, together
with the results of computer modelling, to describe in morphodynamic terms
the evolution of wave-formed coastal features under conditions of marine
transgression, marine regression and prolonged stillstand. The section on the
evolution of wave-dominated coasts under stillstand conditions has the most
direct bearing on the present-day coasts, which mostly have been subjected to
the relatively stable sea levels of the late Holocene compared with those of the
preceding 100000 years. This section is therefore presented in more detail.
Sections on sea-level transgression and regression are presented first, however,
because these are the antecedents of the Holocene stillstand. Again, because
the marine transgression is the more recent of the two antecedents, more is
130 P.S. Roy et al.

a Local substrates steepening

Coastal plain Embayed Protruding and

coasts coasts cliffed coasts

SL transgressive barrier islands transgressive sand

sheets
rising and bay barriers
erosion surfaces

SL erosional
regressive barrier strand plains reworking
falling

SL barrier islands bay barriers mainland beaches

headland spits
stable shelf sand ridges inner sand shelf sand bodies
sheets

b
Present Cliffed Embayed Coastal interglacial
sea level coasts coasts plains highstands

Present inner
shelf zone
c. 20-60m
Present mid
shelf zone
c.60- 120m
Present outer
shelf zone
glacial
c. 120- 150m sb
lowstands
0 1
proportion of SE Australian coasts

Figure 4.4. Field diagrams indicating (a) relationships between regional morphologies,
changing sea levels and the occurrence of various types of wave-formed coastal
deposits, and ( b )the relative proportion of various coastal morphologies as functions of
relative sea level and of geological time (glacial-interglacial cycles). The “sb” in the
bottom left corner denotes the situation during lowstands on shallow shelves (e.g., the
US Atlantic shelf) when the sea drops below the shelf break and encounters the
relatively steep upper slope.
 Wave-dominated coasts 131

known about coastal evolution during rising sea levels. These two sections are
therefore presented in the reverse order of their chronology in nature, to
convey a firmer foundation for principles of coastal evolution driven by
changing sea levels.

Marine transgression
This section focuses on coastal evolution during a rapid marine transgression,
such as that of the postglacial marine transgression or that predicted under
some scenarios for Greenhouse-induced sea-level rise (Roy & Thom, 1987;
Thom & Roy, 1988). The effects of changing substrate slopes, sediment
budgets and backbarrier accommodation volumes are also examined.
Low-gradient coasts, whether laterally unbounded or embayed, that are
exposed to wave action at times of rising sea level are characterised by
transgressive barriers (Fig. 4%). These correspond to Swift’s ‘erosional
shoreface retreat’ model and are primarily made up of tidal inlet and washover
deposits. Transgressive barriers are essentially transitory features that maintain
themselves in dynamic equilibrium with rising sea level by the landward
transfer of sand, eroded from the shoreface, to backbarrier settings (Fig. 4.5b).
The shape and dimensions of such transgressive barriers have been
reconstructed from investigations of coasts presently experiencing slow
relative sea-level rise (e.g., US and Canadian Atlantic coasts - Kraft & John,
1979; Boyd, Bowen & Hall, 1987) and from more stable coasts where drilling
and dating of prograded barriers have delineated their transgressive facies that
were deposited just before sea level stabilised (Roy & Thom, 1981; Thom,
1984; Roy et al., 1992). On high-energy coasts, transgressive barriers have
dimensions of 15-25 m thick and 2-3 km wide and extend 2 4 m above mean
sea level depending on the storm surge level (Thom, 1984). Southeast
Australian studies suggest that their cross-sectional areas are in the order of
10000-30 000m2(Roy &Thorn, 1981; Roy etal., 1992).
Calculations based on modelling a constant morphology during barrier
translation while sea level is rising at around 15 mm a-* (its average rate
during the postglacial marine transgression in southeast Australia) involves
eroding as much as 200 m3 m-’a-’ of sand from the front of the barrier and
transporting it to the rear by tidal inlet and overwash processes. The volumes
are much less than this on low-wave-energy coasts (Cowell & Nelson, 1991).
Volumetric changes far in excess of this are thought to occur on the shoreface
during single storms on this coast (Thom & Hall, 1991). Thus, there would
seem to be ample energy available to shift barriers landward even while sea
132 P.S. Roy et al.

C
washover deposits

shoreface
erosion

ravinement surface

tidal lniet deposits

relict estuarine muds

Figure 4.5. Transgressive barriers characterised most coasts when sea levels were
rising on low-gradient substrates. They were composed almost entirely of tidal delta (a)
and washover ( c ) deposits that retrograded into estuarine/lagoonalenvironments as sea
level rose. Landward migration was through a process of erosional shoreface retreat (b)
as the barrier adjusted to changing sea level. The sea bed exposed by the retreating
barrier is an erosion or ravinement surface.

level is rising relatively rapidly. As much, or even more, time is available for
sediment transfers when the rate of marine transgression slows down.
However, another factor - foredune development - may alter barrier dynamics
under these conditions. While sea level is rising rapidly, aeolian processes only
have time to build low, discontinuous foredunes on the barrier surface that
provide little impediment to storm washovers (Fig. 4.5~).As the rate of sea-
level rise slows, more dunes build above storm surge levels and the frequency
of washovers is reduced thus retarding barrier translation. The development of
a continuous foredune leads to in situ submergence as sea level continues to
rise, and to erosion of the barrier superstructure which, in planform, shows a
progressive narrowing from both the ocean and the lagoon side. Leatherman
(1983) identified this historical pattern on Fire Island, New York. Sanders &
Kumar (1975) present controversial evidence from the Long Island shelf that
 Wave-dominated coasts 133

this process has led to in-place drowning and overstepping of barriers under
conditions of slow sea-level rise. However, it is more likely that landward
barrier translation recommences when the sea rises to a level that effectively
narrows the barrier and once again allows storm washovers (Leatherman,
1979; 1983). It follows from this discussion that barrier translation is fairly
continuous while sea level is rising relatively rapidly but may become
intermittent, with hiatuses of many hundreds of years, when it slows down.
While external conditions of substrate slope, wave regime, sediment budget
and rate of RSL rise are maintained, it seems reasonable to assume that
parameters such as shoreface profile shape, inlet geometry, backbarrier width
and relief on the barrier surface remain approximately constant. Under these
conditions, the rate of shoreface translation is simply a function of the substrate
slope and the speed at which RSL is rising; computer modelling suggests that
landward translation of the transgressive barrier is accomplished without
eroding the substrate (Fig. 4.5b) and that its basal contact corresponds to the
former land surface. Later, an erosional ravinement surface (the product of
wave reworking at the toe of the shoreface) is developed on the shelf after the
transgressive barrier has passed but contemporary estuarine sediments may be
preserved in topographically low areas related to antecedent topographies
(Belknap & Kraft, 1985) (Fig. 4 . 5 ~ ) .

The role of substrate slope

Changes to the external parameters influence transgressive barriers in ways
that simulation modelling suggests are subtle and, in some cases, unexpected.
The most sensitive parameter seems to be substrate slope - a geologically
inherited property. Changes to the coastal sand budget that mostly arise from
local variations in littoral transport rate are generally less important when sea
level is changing rapidly.
The types of behaviour of a transgressive barrier translating over a range of
substrates are listed in Fig. 4.6, based on simulation modelling of a typical
southeast Australian barrier (Cowell et al., 1991). The modelling experiments
show that for a constant RSL rise, the rate of shoreface retreat, the amount of
sand recycled each year and the cross-sectional area of the transgressive barrier
all decrease as the substrate steepens from very gentle gradients of 0.1 " to
about 0.8" (Fig. 4.6a, b and c). On slopes steeper than approximately 0.8", for
barrier geometries applicable to southeast Australia, the model predicts a
Bruun-like response involving a net offshore sand transfer as RSL rises (Fig.
4.6d and e ) . Shoreface changes on 'steep' gradients involve a continuing input
of sand either from coastal erosion (standard Bruun Rule; Dean & Maurmeyer,
134 P.S. Roy et al.
Substrate Barrier Sediment Tnnifrr Coastline
barrier coastline slop Volume (m31m) translation
(dcwet) (m'lm)
a sand body translation (+onshore. -0Kshore) Late@)
SL rising o.l 23275 + 12120( 180) SSS(8.8)

0.2 19788 +5339(80) ZSS(4.2)

b
0.4 12935 +1966(29) 143(2.1)

C 0.6 6053 +823(12) 95(1.4)

0.8 915 +334(5); -79(1) 71(1.0)

d
1.0 0 -253(4) 57(0.85)

e 1.1 0 -597(9) 38(0.57)

2.0 0 -804(12) 29(0.43)

INSET

Shoreface dimensions 4.0 0 -1144(17) 14(0.21)

Xs:25m Ys:1500m
backbarrier width:500m
6.0 0 -1277(19] 9(0.13)

10.0 0 -1372(20) 4(0.06)

Figure 4.6. Changing substrate slopes affect depositional styles on coasts undergoing a
marine transgression. On gentle slopes ( a and b), rates of coastal recession are high and
there is a mass onshore transfer of sand with the migration of transgressive barriers. In
contrast, on relatively steep substrates (d and e ) the coast is in encroachment mode and
rates of coastal recession are slow. Sediment moves offshore to the lower shoreface,
and is lost to the barrier which consequently loses volume. In (c) onshore and offshore
sand movements are approximately balanced. (Each illustration shows equal amounts
of erosion and deposition). The table documents simulation results of the shoreface
translation model (inset) for a range of substrate slopes. Sand transfer and coastline
translation rates are for a 1.Om rise in sea level; those in brackets are annual rates if the
sea was rising at 1.5 dcentury.

1983) or by littoral transport (i.e. a sediment budget imbalance). The amount

of sand deposited offshore increases as the substrate slope steepens but if the
external sand supply is starved or cut off, the inshore sea bed becomes a
drowned, scoured surface with exposed bedrock - a situation that occurs off
many southeast Australian headlands (Ferland, 1990).
Changes in the rate at which RSL rises not only control shoreface
translation but also influence the amount of work that waves expend on the
shoreface and beach. They also influence the shape of the barrier, especially
the shoreface profile, although little is known about this (Pilkey et al., 1993).
On very gentle gradients (<O. 1") bottom-friction effects reduce incident wave
energy levels (Wright, 1976) such that equilibrium shoreface profiles and
barrier widths (and volumes) become smaller thus reducing the quantity of
 Wave-dominated coasts 135
1 lOOm 380m
140m

SL rising

tess
0.4O

Figure 4.7. A marine transgression on a substrate that progressively flattens in a

landward direction (0.4' to <0.05') inevitably leads to an increase in the rate of coastal
recession (from 140 to I lOOm m-' rise in sea level). Frictional effects of the shoaling
sea bed tend to diminish incident wave energy, and also shoreface and barrier
dimensions. On extremely shallow shelves, this culminates in submerged sand banks.
Reduction in barrier size is accomplished by shedding sand which, if not dispersed by
littoral currents, forms a trailing-edge sand sheet (tess) on the shelf surface.

sand transferred landward per unit time. This trend is illustrated in Fig. 4.7 on
a flattening substrate; a similar effect is produced on a constant slope by an
acceleration in the rate of RSL rise (rates as high as c. 20-25 mm a-' were
probably achieved during the postglacial marine transgression - Fairbanks,
1989). As indicated in Fig. 4.6a, the quantities of sand that are redistributed on
very low-gradient substrates can be very large and may exceed the transporting
capacity of the incident waves. In response, the barrier probably reduces size
to accommodate a more rapid shoreface translation, leaving behind a trailing-
edge sand sheet (Fig. 4.7). With slow rates of RSL change, more time exists for
waves to mould the barrier shoreface into an 'equilibrium' configuration and to
increase the width of the backbarrier (tidal delta and washover) deposits.
Under quasi-stillstand conditions the shoreface profile becomes wider and
extends to deeper water depths than when RSL is changing rapidly. Therefore,
barriers attain their maximum dimensions under prolonged stillstand
conditions, as discussed in the section on stable sea levels.
In many respects, the pattern produced by a transgression over a relatively
steep slope (Fig. 4.8) corresponds to Swift's (1976a) trailing-edge sand sheet
that he associates with low-gradient shelves. In southeast Australia,
transgressive sand sheets of the type shown in Fig. 4.8b occur on relatively
steep substrates and are composed of uniformly fine sand, typical of the lower
shoreface. They are found off headlands and steep sectors of coast, and form
the lower unit of shelf sand bodies (SSBs) (Field & Roy, 1984; Roy, 1985;
Ferland, 1986, 1990) (Fig. 4 . 8 ~ )The
. age of these transgressive deposits in the
Sydney region is shown in Fig. 4 . 8 ~ the
; overlying unit was deposited under
stillstand conditions and is discussed in a later section.
136 P.S. Roy et al.

a steep
cliffed coast

SL

mid shelf
seds.
regressibe sand
' (7-0kal
transgressive 'rock reefs
marine sand
(14~7ka)
> 10

C time (14C years BP)

PRESENT
DAY
erosion
SL rising

-
0 transgressive deposits
regressive deposits
+ mid shelf sediments

Figure4.8. Shelf sand bodies that occur along steep and deep sections of coast in
southeast Australia are made up of transgressive and regressive units ( a ) deposited
subaqueously during the postglacial marine transgression and the following stillstand.
At all stages of their formation, the direction of net sand movement was seaward (b).
Radiocarbon dates (environmentally corrected) on shell fragments from the two
depositional units in the Sydney shelf sand body, and from mid shelf sediments lying
immediately seaward of it, are plotted in (c).
 Wave-dominated coasts 137

SL rising

erosion
C
large estuary
inner shelf
sand
backbarrier
sand
substrate

tess

Figure 4.9. Simulation modelling predicts that variations in the marine sand budget will
cause a transgressive barrier to erode the sea bed if there is a net loss of sand (case b)
and to cause deposition of a trailing-edge sand sheet (tess) if there is a net addition of
sand (case c) (The sediment budget is balanced in case a ) . The trailing-edge sand sheet
in case c is composed of backbarrier sand, but subsequent marine reworking may
change its character to a marine deposit and re-arrange its sheet-like morphology into
sand ridges.

The role of the nearshore sediment budget

The predominance of ‘coastal plain, open drift’ coastal settings in Fig. 4.4b
indicates that transgressive barriers existed on low-gradient continental
shelves for much of the Quaternary, at times of lower sea levels. On these
shelves, alongshore sediment movement was essentially unimpeded and
operated largely independently of onshore sand transfers. However, as Boyd et
al. (1987) recognised on the highly indented, glaciated coast of Nova Scotia,
local changes in sediment supply during transgressions strongly influence
barrier behaviour.
Fig. 4.9 exemplifies a number of situations involving sediment
budget imbalances during a marine transgression on a relatively gently sloping
substrate. In cases a, b and c, the littoral-transport differential (the difference
between littoral drift rates at various places along the coast) is important, not
the total drift rate. In the case of a balanced budget (Fig. 4 . 9 ~the
) dimensions
of the transgressive barrier and its rate of landward translation remain constant
with time provided that the substrate gradient and rate of RSL rise are
unchanged. A ravinement surface is formed by waves reworking the inner
shelf surface, but with a balanced sediment budget the ravinement surface
corresponds closely to the original land surface. With a negative sand
138 P.S. Roy et al.
imbalance (Fig. 4.9b), and the same boundary conditions as Fig. 4.9a,
landward barrier translations occur at a similar rate but the toe of the shoreface
erodes into the underlying substrate and, in so doing, generates additional sand
to maintain transgressive barrier volume. However, the cross-sectional
dimensions of this barrier are significantly smaller than in Fig. 4 . 9 ~ .The
reverse is true for a positive sand imbalance (Fig. 4 . 9 ~ )Excess
. sand, originally
deposited in backbarrier environments, is left as a sand sheet on the shelf
surface after the transgression passes and the transgressive barrier itself is
considerably larger than in the case of Fig. 4 . 9 ~For
. example, a drift imbalance
of 5m3m-' length of coasdyear will produce about 3.5m of erosion/
deposition on a low angle shelf (0.4" slope) undergoing transgression. This
drift imbalance is equivalent to a difference between littoral drift inputs and
outputs of 50000m3a-' on a lOkm long sector of coast experiencing a RSL
rise of 10mma-l. The negative imbalance causes a 40% decrease in barrier
size; the positive imbalance produces a 55% increase in its size. Despite this,
rates of coastline retreat are dominated by the marine transgression and are
only slightly modified (+ 10%) by imbalances in the sand budget. These
results are for steady-state conditions, which the simulation experiments show
are only achieved after 30 to 50 time steps, the equivalent of many thousands
of years in nature. They also show that similar coastal impacts can be produced
by changing the rate of RSL rise as by varying the sand budget.

Lagoons, estuaries and accommodation space

The preceding discussion focused on shelf sand budgets. Fig. 4.10, on the other
hand, illustrates the situation where an estuary or lagoon, located behind a
transgressive barrier, is receiving sediment at the same time as sea level is
rising. In Fig. 4.10~1,the estuary is infilling at a much slower rate than sea level
is rising (the result of a small sediment supply or very rapid sea-level rise),
which gives rise to a thin layer of backbarrier sediments on the sea bed after the
barrier has translated landward. With estuarine sedimentation occurring at
almost the same rate as sea level is rising (Fig. 4.10b), the estuary becomes
wider as the barrier translates somewhat more slowly over the estuarine
deposits which form a relatively thick deposit on the inner shelf. Compared
with Fig. 4.10a, the barrier in Fig. 4.10b becomes smaller in the process of
attaining steady-state conditions. The sand it loses to the inner shelf veneers
the muds to some extent, although the sea bed undoubtedly experiences
erosion during severe storm events. Slower translation of the barrier shoreface,
due to rapid infilling of the estuary (equivalent to a delta-margin environment),
leads to a protruding coastline which may be exposed to marine erosion. This
 Wave-dominated coasts 139
1200171
a lagoon or
estuary
SL rising

95Om
b thin coarse
sand layer

mud
estuarine mud 8ubstraie

1040111
C
SL

Figure 4.10. The model simulates the effect of lagoonal mud sedimentation (i.e., a
fluvial sediment budget gain) on the same transgressive barrier as shown in Fig. 4.9. A
small amount of mud deposition (case a ) has little effect on barrier size or coastal
behaviour, but with mud accumulating almost as fast as sea level is rising (case b), the
barrier decreases in size and recession slows down. In case c, a deficit in the marine
sand budget is superimposed on case b). In all cases lagoonal muds conformably
blanket the substrate beneath and to seaward of the barrier; after the barrier passes,
wave reworking ensures that the sea bed is soon veneered by a coarse lag (the inner
shelf sand sheet). (The arrows and distances indicate comparative translations for the
same increment of sea-level rise).

is modelled in Fig. 4 . 1 0 ~as a shelf sand loss that causes the shoreface to erode
into the mud substratum, the barrier to decrease in volume and the rate of
shoreline translation to increase. In this scenario, the estuarine sediments are
muds, which are dispersed by wave action; if they contain sand it would be
added to the shoreface and incorporated in the transgressive barrier which
would, again, slow the rate of shoreface retreat.
Lagoons and estuaries also act as long-term sediment sinks for marine sand
(Swift & Thorne, 1991). The importance of accomodation volume to the rear
of the barrier is demonstrated by Fig. 4.10b (where the estuary is infilling
rapidly), which shows a reduction in the translation rate as the estuary becomes
shallower. The overall behaviour is indicative of the state-dependent evolution
outlined in Chapter 2, with the accommodation volume providing a self-
regulating feedback mechanism. In nature, the physical character of the
estuary mouth is controlled by a dynamic balance between littoral and tidal
sand movements on the open coast and in the tidal inlet, respectively (Roy,
 P.S. Roy et al.
1984; Nichols & Biggs, 1985; Dalrymple, Zaitlin, & Boyd, 1992). Usually
inlets in bay barriers are fixed against bedrock outcrops or headlands and do
not form extensive tidal inlet deposits (see section on stable sea levels).
However, in the case of barrier island chains, local imbalances in the littoral
sand budget may lead to inlet migration and the formation of recurved spits at
the ends of the adjacent barrier. Under these conditions large quantities of sand
become stored in abandoned channel and flood tide delta deposits. Where the
channels are deep (> 10 m) these deposits may make up the bulk of the barrier
sand mass (Hoyt & Henry, 1967; Moslow & Heron, 1978).

Marine regression
All barrier sand bodies that prograde in a seaward direction are regressive in
nature sensu stricto; however, in this chapter, we restrict usage of the term
‘regressive barrier’ to features deposited by waves under conditions of falling

abandoned
highstand estuary

a former high SL

SL failing

regressive beach and

shoreface sand

deposition
SL
erosion
advancing
toe siopa
shoreface profile

Figure 4.1 1. A regressive barrier that progrades under falling sea levels produces a
tabular, gently seaward-inclined sand deposit 10-20 + m thick with an erosional base
(a). As sea level falls, the inner shelf surface erodes and sand moves onto the shoreface
(b).The surface of the regressive barrier forms a wide strand plain but without estuaries
comparable to those formed under conditions of stable or rising sea level.
 Wave-dominated coasts 141
a Inset

20m

Figure 4.12. ( a ) The Tuncurry embayment in central NSW (lat. 32” S) showing the
distribution of barriers onshore and on the inner shelf. ( b )A shore-normal cross-section
(dashed line in a ) through the various barriers showing their dimensions and age
groupings. The pre-Holocene ages are based on thermoluminescence measurements by
the University of Wollongong documented in Roy et al. (1992); the numbers in
brackets indicate the numbers of samples dated. The inset shows the group ages of the
various barrier systems (shaded) and the estimated position of the sea when they formed
in relation to the sea-level curve (dashed line) of Chappell & Shackleton (1986). (c)
Shallow marine seismic ‘Uniboom’ profile (Line 12 in Roy et al., 1992) through the
axis of the Tuncurry embayment. The interpretation of a drowned (regressive) barrier
system is based in part on the concordance between the attitude of its dipping reflectors
and the present shoreface profile (inset).
142 P.S.Roy et al.
RSL - the so-called ‘forced or eustatic regression’ (Fig. 4.1 1). Similar deposits
when formed at times of stable, or even slowly rising, RSL are termed
‘prograded barriers’ and are discussed in the next section.
Very little is known about the effects of marine regression, particularly at
the scale associated with the onset of glacial periods. The little that is known
comes from field evidence on contemporary coasts that are undergoing slow
uplift due to tectonic or isostatic processes (see Chapter 12). Regressive
barriers appear to be features of low-gradient, relatively sediment-rich shelves
that have not experienced excessive erosion during the postglacial marine
transgression. The principles presented in this section are based upon field data
from a mid shelf barrier complex drowned in 30 to 50m of water north of
Sydney, Australia (Roy et al., 1992) (Fig. 4.12), supplemented by the results of
computer modelling. Regressive barriers preserved on present-day shelves are
not widely reported but where they have been identified in southeast Australia,
they seem to be as wide or wider than their stillstand counterparts (Fig. 4 . 1 2 ~ )
and contain almost as much sand as the onshore barriers. In the Forster-
Tuncurry area, the regressive (drowned) barrier shown in Fig. 4 . 1 2 ~and b is up
to 7km wide and 15m thick. Highstand barriers onshore are almost 10 km
wide and range in thickness from 12 to 23m (Roy et al. 1992).
Thermoluminescence dating of both the highstand and shelf barriers indicates
a history of intermittent progradation spanning the last 260 ka with at least two
regressive barrier systems forming during Late Pleistocene interstadials
(isotope stages 5a and 3) (Fig. 4.12, inset). The drowned barrier on the inner
shelf prograded during the period 60-40 ka when RSL was falling slowly at
about 0.5 m per thousand years (see the inset in Fig. 4.12). Its base is erosional,
in places intersecting bedrock (Fig. 4.12~);the upper surface is thought to have
been modified during passage of the postglacial marine transgression which
caused 3 to 5 m of erosion (Roy et al., 1992).

The role of substrate slope

Results of a simulated regression are illustrated in Fig. 4.13 for both gentle and
steeper-gradient shelves. In the former case (Fig. 4 . 1 3 ~and b), as RSL falls,
the highstand barrier experiences shoreface progradation with sand eroded
from further offshore. Erosion occurs on the lower shoreface and in a zone
extending seawards of its toe (the ‘toe slope’ zone shown in Figs. 4.1 l b and
4.13b). With continued RSL fall, the regressive barrier builds seawards into
the excavated zone and reconstitutes the former sea-bed morphology (Fig.
4.13b). Shoreface reflectors in the seismic profile of the drowned barrier
shown in Fig. 4 . 1 2 ~have very similar geometry to the onshore barriers and to
 Wave-dominated coasts 143
gentle substrates steeper substrates
I

Figure 4.13. Mechanisms for regressive barrier formation, as sea levels fall on gently
sloping shelf surfaces, involve erosion of the sea bed and the onshore movement of sand
to prograde the shoreface. The inset shows a closed system with no net gains or losses
of sand to the barrier. The simulation model (a) suggests that the sea bed experiences
erosion for distances of many kilometres seaward of the barrier in high-energy, open-
coast environments such as southeast Australia. In (b),the regressive barrier lies on the
newly-formed erosion surface and represents a zone of massive in situ reworking
equivalent to the barrier’s thickness. The model also shows that steady-state conditions
are only achieved very slowly during a marine regression. On a relatively steep, sandy
substrate (c) falling sea level causes extensive reworking of the sea bed and a
progressive offshore movement of sand. On slopes that are not too steep, a thin veneer
of coarse beach sand may be left stranded as the sea receeds. A marine regression
completely reworks a shelf sand body (d)transporting much larger quantities of sand
offshore than is the case in (c).

the present shoreface profile (Fig. 4.12c, inset). Wellner, Ashley & Sheridan
(1 993) report similar seismic patterns from the New Jersey inner shelf (USA)
and also attribute them to interstadial barriers. Furthermore, the erosional
nature of the base of the drowned barrier in the Forster-Tuncurry area is
explained by the initial phase of offshore erosion that precedes barrier
progradation in the simulation model (Fig. 4.13). Initially, progradation of
regressive barriers was slow but gradually increased until it reached a steady
state when the surface of the regressive barrier tracked the former sea bed. As
144 P.S. Roy et al.
shown previously, steady-state conditions take thousands of years to achieve in
the model; in nature they probably are never reached before environmental
parameters (such as sea level or substrate slope) change.
As is the case for marine transgressions, substrate slope plays a critical role
in controlling barrier size, as well as the rate of coastal recession, as RSL falls.
For the same wave conditions, the largest regressive barriers are predicted to
occur on the most gently sloping surfaces. In reality, however, bed friction
tends to reduce incident wave energy and also barrier size in such cases. The
best-developed regressive barriers should occur on moderately low-gradient,
but high-energy, shelves when RSL is falling slowly, thus allowing maximum
time for the shoreface profile to fully develop. The southeast Australian case
study illustrated in Fig. 4.12 seems to exemplify these conditions. Like
transgressive barriers, barriers formed during regressions recycle their local
sand budget as RSL falls (Fig. 13a and b) and, while they are affected by
littoral sand supplies, they are not dependent on them for their existence.
Regressive barriers should therefore be widespread globally on relatively
energetic, low-gradient and sand-rich shelves.
In general terms, regressive barriers can be expected to form on slopes of
<lo, but simulation studies depicted in Fig. 4 . 1 3 ~and d suggest that as
substrates steepen the sand flux reverses and becomes directed dominantly
offshore. As shelf gradients steepen, increasingly large amounts of sand are
deposited beyond the toe of the shoreface. Fig. 4 . 1 3 and~ d which exemplifies
a marine regression on a relatively steep slope, illustrates two cases: a planar
substrate sloping at > 1 O (Fig. 4 . 1 3 ~and
) the same substrate with a convex shelf
sand body formed under highstand conditions (Fig. 4.134. In the former case,
however, a thin beach deposit, possibly composed of coarse sand and enriched
in heavy minerals, is left behind as the shoreface retreats; a larger amount of
sand is transferred offshore in the latter case (Fig. 4.13d) and the SSB is
completely reworked. Initially, the regression causes massive erosion of the
existing substrate but, eventually, steady-state conditions are approached at
which time reworking is focused on the body of fine sand deposited at the toe
of the shoreface (Fig. 4.13d). Because, in nature, steeper substrate slopes are
found off headlands or protruding sectors of coast with high wave energy and
littoral transport rates, it is unlikely that the offshore sand accumulations will
remain intact during a regression. Rather, they will be reworked and added to
the downdrift coastal sediment budget. A case study where sand eroded from a
SSB is incorporated in regressive barriers forming contemporaneously in an
adjacent embayment is discussed in the conclusions to this chapter.
 Wave-dominated coasts 145
Lugoons and estuaries during marine regressions
It is clear from Figs. 4.1 1 and 4.13 that lagoons and estuaries are spatially
restricted during marine regressions. The combined effects of a massive
onshore movement of sand, falling water levels and fluvial deposition reduced
estuary size to a minimum and changed their morphologies from the common
‘barrier lagoons’ of today to the more constricted ‘river estuary’ type at the
mouths of the larger and more active rivers (Roy, 1984; Wright, 1985) (Fig.
4.1 la). Furthermore, there was little opportunity for the formation of the coast-
parallel topographic depressions landward of the shoreface that characterise
coastal evolution during transgressions over gentle substrates. Estuarine
waters therefore would have been restricted mainly to the seaward extremities
of distributary channels and estuarine habitats would thus be greatly curtailed
during the long periods of the Quaternary that RSL was falling. This has
special implications for estuarine habitats and species: perhaps only those
estuarine species that were able to adapt to more marine conditions on the open
coast would survive (see Woodroffe & Grindrod, 1991 for a similar discussion
about mangroves).

Stable sea-level conditions

Cessation of barrier transgression and barrier stabilisation is a function of sand
supply in relation to RSL. In principle, a large sand influx while sea level is
still rising could cause a barrier to stabilise and grow upwards in place.
However, these conditions probably occur rarely and then only near active
river mouths (Swift, 1976a). While sea level is rising, most river mouths are
constantly being inundated and the offshore sea bed is deepening, thus cutting
off a direct supply of river sand to the coast and reducing onshore movement
of sand from the adjacent sea bed. Conditions favourable to barrier
progradation occur only when rates of sea-level rise slow or stabilise. In the
case of some southeast Australian barriers, coastal progradation was triggered
when sea level stabilised at around 6.5 ka. At this time, the offshore sea bed
began slowly to equilibrate, thus providing an influx of shelf sand; a process
that has continued for thousands of years on this coast (see Fig. 2.9) (Roy,
Thom & Wright, 1980; Thom, Bowman & Roy, 1981; Thom & Roy, 1985).
Changing RSL, substrate slopes and sediment budgets all contribute to
coastal evolution under rising sea levels, but as the marine transgression slows,
and RSL eventually stabilises, the role played by imbalances in sediment
budgets becomes dominant. This is illustrated in Fig. 4.14, which shows the
simulated translation of a transgressive barrier experiencing constant drift
imbalances (negative and positive) during a marine transgression that slows
146 P.S. Roy et al.
time (ka)
a
n b
E
Y
P
-I
v) $
c5
3

5P
al
P

A SL rise (m/oentury)

2
1
3

Y
E

-0
4-

Q progradation
0

Q erosion

time (ka)

Figure 4.14. As the rate of RSL rise slows, sediment budget differentials play an
increasingly important role in coastal change. ( a ) Shows a theoretical marine
transgression from 25 ka to 6 ka BP (solid line) with the rate of SL rise slowing down
 Wave-dominated coasts 147
down and then stabilises (stillstand). The sea-level history is shown in Fig.
. sand imbalance modelled in Fig. 4.14b (? 5 m3m-'RSL rise a-I)
4 . 1 4 ~The
is equivalent to a 10km long embayment with a differential of 50 000 m3a- '
between alongshore sand inputs and outputs. During most of the marine
transgression (in this case from 25 ka to 10 ka) coastal changes are dominated
by rising RSL and the sediment budget has a very minor impact on rates of
coastal recession (Fig. 4.14~).However, when rates of sea-level rise slowed to
less than about 0.5 m per century, sediment budget effects become more
noticeable and under stillstand conditions they eventually dominate coastal
change. With a positive sediment budget, coastal recession slows rapidly and
is reversed before sea level stabilises (Fig. 4.14b, case 2). Barriers prograde
during the stillstand but the rate of progradation slowly declines as the
shoreface builds into deeper water. Fig. 4.14b (case 3) shows that, with a
negative sediment imbalance, the rate of coastal recession decreases to a
minimum when RSL stabilises (around 20m per century in Fig. 4 . 1 4 ~ )and
remains just below this value as the shoreface erodes into the substrate (Fig.
4.14b, case 3).
The following sections examine stillstand coastal deposits ranging from
sediment-rich to sediment-deficient types. The models are based on the wide
variety of bay barriers recognised along the mainly embayed coast of southeast
Australia, but they have their counterpart on coastal plain coasts elsewhere in
the world (Dickinson, Berryhill & Holmes, 1972; Thom, 1984; Short, 1988).
Each has characteristic lithofacies architecture reflecting sediment budget
imbalances that arose either during the stillstand or were inherited at the end of
the postglacial marine transgression. The various types of stillstand barriers
and inner shelf sand deposits are listed in Fig. 4.15 which also indicates their
inferred sediment budgetdfluxes. In southeast Australia most barriers are
composed of 'marine' or shelf sand - a mixture of relatively mature clastic
grains, with lesser amounts of calcareous detritus, reworked and transported

Caption for Figure 4.14 (cont.).

(dotted line) and a stillstand of the sea after 6ka BP. ( b ) Simulations of two
transgressive barriers responding to the SL changes in ( a ) and experiencing positive
(case 2 ) and negative (case 3) sediment budget imbalances. (The case of a balanced
sediment budget is not shown.) (c) The behaviour of transgressive barriers, in terms of
coastal recession, is plotted against time and rate of sea-level rise for the case of a
balanced sand budget (case I ) , for a barrier gaining sand (case 2 ) , and for a barrier
losing sand (case 3). The lower graph compares cases 2 and 3 to case 1. It shows that,
in the case of a barrier gaining sand, the sediment budget factor had the greatest effect
on coastal behaviour around 8 ka BP, well before sea level stabilised. With the barrier
loosing sand, the effect gradually increased as the rate of sea-level rise slowed and
reached a maximum when sea level stabilised at 6 ka BP, after which it stayed relatively
constant.
148 P.S.Roy et al.

(a) Transgressive Dune Low-gradient shelf, strong onshore

Barrier winds, high energy waves, large sand
(incl. CliffTnp Dunes) supply.

(b) Prograded Barrier Low-gradient, shallow shelf, moderate to

large, continuing sand supply.

(c) Stationary Barrier Moderate-to-low gradient shelf, small

shelf sand supply balances sand losses.
8:estuary

(d) Headland Spit Alongshore sand and gravel supply from

local source.
b:bay

(e) Receded Barrier Moderate-gradient shelf, slow sand loss

exceed sand supply.

(0 Mainland Beach Steep-gradient shelf, continuing sand loss

Barrier from eroding coast.

(g) Inner Shelf Sand Low-gradient shelf, seabed erosion,

Sheet onshore movement of winnowed sand.

(h) Shelf N d g e Field Shallow, low-gradient shelf, in sifu

reworking, offshore movement of
shoreface sand.

(i) Headland-attached Deep, steep-gradient shelf, offshore

Shelf Sand Body movement of shoreface sand.

Figure 4.15. List of wave-dominated coastal deposits formed under quasi-stillstand

conditions and the main factors responsible for their formation. Relative magnitudes of
long-term sand budgets and net directions of sediment fluxes on the shoreface are
indicated by arrows. The jagged arrows in (g), ( h )and (i) indicate irregular transport on
the inner shelf.
 Wave-dominated coasts 149
landward from the continental shelf. Except in a few cases, rivers on this coast
are still infilling their estuaries or are delivering sand to the coast in quantities
that are small in comparison to the littoral sediment flux (Roy & Crawford,
1977; Roy & Thom, 1981). Elsewhere in the world, more active rivers than
those that occur in Australia have delivered large quantities of sediment to the
coast throughout the Holocene. As a result, coasts in these areas are
characterised by barriers at the sediment-rich end of the spectrum of coastal
types listed in Fig. 4.15 (transgressive dune and prograded barriers) and are
composed of immature fluvial sand. The various types of stillstand coastal
deposit are discussed below.

Transgressive dune barriers

The geomorphology and evolution of transgressive dunes are described by
Hesp & Thom (1 990) who, following others, identify three largely
independent factors as being important in their development: rate of sand
supply, wind and wave energy and the effectiveness of sand-binding plants.
Many transgressive dune barriers occur where sand is supplied by waves in
such large quantities to an exposed, windy coast that the local pioneer plants
are inundated and killed before they can stabilise the sand in incipient
foredunes (Short & Hesp, 1982; Hesp & Thom, 1990). They may also result
from the remobilisation of a previously stabilised barrier surface (Davies,
1980); the transgressive dunes at Newcastle Bight, 150km north of Sydney,
are an example (Thom et al., 1992).
Transgressive dune barriers are transitional between barriers with complex
foredunes and non-barrier (desert) coastlines with aeolian dune sheets that, in
migrating inland, have become detached from the shoreline. The cliff-top
dunes described by Short (1988) from the southern Australian coast are
detached remnants of what were formerly much larger transgressive dune
fields. Originally these incorporated sand ramps a hundred metres or more
high connecting coastal sand sources to long-walled parabolic dunes on the
cliff tops (Jennings, 1967). Subsequently, when the marine sand supply was
cut off, the sand ramps and beaches were destroyed - either blown inland or
reworked by waves, or a combination of both. Many cliff-top dunes appear to
have formed in the early Holocene while RSL was still rising (Pye & Bowman,
1984), while others formed at various times in the Pleistocene (Bryant et al., in
press).
Transgressive dunes contain larger quantities of sand per unit length of
coast than other barrier types and occur in the most exposed locations on the
southern Australian coast (Short, 1987). In southeast Australia, where they
150 P.S. Roy et al.

162.5O E

d
8 Smiths Lake

8
MyaU Lake

BHR
Seal Rocks

5 krn

32.5O S
dune ridge crests
beach and active dunes
Holocene dunes
mining excavation Pleistocene barrier
drill holes

NE sw

SL

20m 3 km bedrock

Figure 4.16. Generalised morphology and stratigraphy of the Seal Rocks transgressive
dune field (td), Myall Lakes, southeast Australia. Dashed line shows transect in b. (b)
Topography along the transect. The landward margin of the dune field is marked by
Bridge Hill Ridge (BHR), a large precipitation ridge that reaches elevations of 100 m in
places. Radiocarbon dates on charcoal pieces sampled from buried soil surfaces in a
sand-mining excavations (1. 1265 2 135 years BP; 2. 3025 2 135 years BP; 3.
2485 2 135 years BP; 4. 1975 2 110 years BP; and 5. 2600 t- 110 years BP) suggest
late Holocene phases of transgressive dune construction for the bulk of the large sand
ridge (based on Thom et al., 1992).
 Wave-dominated coasts 151
occur interspersed with other barrier types, their occurrence can be attributed
to locally high rates of sand supply (and trapping) at the downdrift terminus of
a littoral drift system (Chapman et al., 1982) rather than to variations in wind
energy or vegetation (Fig. 4 . 1 5 ~ ) Here
. they form long-walled transgressive
dunes similar to Cooper's (1958) 'precipitation ridges'. The largest examples
are the sand islands of southeast Queensland: Fraser, Moreton and Stradbroke
Islands (24.5-28"S, Roy & Thom, 1981; Stephens, 1982). These are
composite features comprising beach as well as aeolian deposits that formed
episodically throughout much of the Quaternary. A NSW example is Myall
Lakes (Fig. 4.16) where the Bridge Hill Ridge reaches elevations of lOOm
(Thom et al., 1992). Here, radiocarbon dates from buried swamp deposits and
soil horizons on transgressed surfaces are mostly less than 3000 years BP and
probably reflect the migration time for the dunes to travel inland (in this case,
estuary
river delta flood tide delta

stable sea level

Progaded
estuarine muds barrier sand

'.Iomc.l.0 km bansgressive
barrier sand inner shelf
sand sheet

deposition
b SL
littoral drift

shoreward sand
transport

erosion and winnowing

Figure 4.17. Prograded barriers occur in bedrock embayments and on coastal plain
coasts. (a)The transgressive part was deposited in backbarrier environments (tidal delta
and washover facies) as sea level was rising (see Fig. 4.5) (the small arrow indicates the
toe of the shoreface). In autochthonous settings such as southeast Australia, the
regressive (prograded) part of the barrier is composed of sand winnowed from the
adjacent shelf and/or carried alongshore by littoral currents (b). A very different
situation occurs on allochthonous shelves as shown in Fig. 4 . 2 ~ .
152 P.S. Roy et al.

a distance of 2-5 km). Both at Myall Lakes and Evans Head (29.2’s) it is
possible that large transgressive dune fields were initiated by sudden changes
in coastal alignment following the failure (erosion) of isolated rock outcrops
anchoring one end of a barrier system (Roy, 1982). Elsewhere, drilling in
episodic transgressive dunes has encountered older soil horizons with ages of
8000-10000 years BP (Roy & Crawford, 1981; Pye & Bowman, 1984; Cook,
1986), indicating that transgressive dune development commenced in some
areas during the postglacial marine transgression well before barriers began
prograding. In many cases, these dunes mantle bedrock headlands and may
have been associated with former barriers (‘proto barriers’) and sand ramps
that formed at lower sea level (Roy & Crawford, 1981).

Prograded barriers
Prograded barriers or strand plain complexes (Clifton & Hunter, 1982) (Fig.
4.17) are characterised by multiple, coast-parallel beach or foredune ridges, the
genesis of which has been discussed by Davies (1957), Bird (1976) and Hesp
(1984a,b, 1988). Foredune ridges have an aeolian (foredune) cap on the
accreted beach berm which reflects the occurrence of winds sufficiently strong
to transport beach sand and a climate able to support sand-binding plants to
trap sand on the upward accreting dune surface (Hesp, 1984a). Beach ridges
are low-relief, wave-formed berms that rarely rise more than 3 m above mean
sea level but foredune ridges and swales have greater amplitude (3-5 m) and
crest elevations of 7-10m above sea level (Thom, Polach & Bowman, 1978).
Beach and foredune ridge plains represent coastal progradation that, in
Australia, spans the last 6.5 ka (Fig. 4.18), but in other countries with different
RSL histories, progradation began somewhat later (e.g., Galveston Island,
Texas, USA, Bernard & LeBlanc, 1965, and the Dutch coast, van Straaten,
1965). Prograded barriers are usually well developed near large, active river
mouths on wave-dominated coasts (e.g., Western Africa, Brazil, the Gulf coast
of the USA), but in southeast Australia the sand source, in all but a few cases,
is the offshore sea bed (Fig. 4.17b) (Thom, 1984). In cross-section, regressive
sequences tend to thicken seaward, due to embayment geometry, and become
progressively younger towards the present coast (Thom, 1984). Dimensions
and shapes of typical prograded barriers have been determined by detailed
drilling and dating as exemplified by the Tuncurry Holocene barrier (Roy et
al., 1992) (Fig. 4 . 1 9 ~and b); ground-penetrating radar shows its internal
structure (Fig. 4 . 1 9 ~ ) .Southeast Australian prograded barriers tend to
completely infill bedrock embayments leaving space only for small wetlands
 Next Page
Wave-dominated coasts 153

time(ka calibrated)

p5 1.0 6
woy woy
4 2 0
2. Wonboyn
.-z
L
s
-m
ri 0.5

.-
C
0
c

B o
0
h
3a. Tuncurry 3b. Tuncurry
(width) (volume)

line of best fit

1500 E

4. Fens

SYDNEY
Q South
s
Y central

5. Moruya 6. Shoalhaven

north
central north
south central
south

Figure 4.18. Age structure of selected prograded barriers in NSW based on drilling and
radiocarbon dating, with sample locations plotted as a proportion of total barrier width
(the Tuncurry data is also plotted as a proportion of total barrier sand volume). The
dates are on shell fragments and the problems of mixing are discussed in Roy (1991).
Sites range from moderately low- energy bays (Sites 1 and 2 ) to exposed, high-energy
beaches on the open coast (Sites 5 and 6); histories of progradation show no consistent
trends. Dating of barriers with multiple drilling transects show that shore-normal
progradation occurred roughly simultaneously along the length of the barrier, and not
by alongshore spit progradation. Average rates of barrier progradation (m a-I) are:
Woy Woy 0.57; Wonboyn 0.38; Tuncurry 0.26; Fens 0.35; Moruya 0.34; Shoalhaven
0.24 (based on Chapman et al., 1982 and Thom e t a / . , 1978).
 5
Macrotidal estuaries
J. CHAPPELL AND C.D. WOODROFFE

Introduction
Macrotidal estuaries can be viewed within a continuum of deltaic-estuarine
coastal depositional settings, influenced by riverine processes, wave regime
and tidal energy (Wright & Coleman, 1973; Wright, 1985; Boyd, Dalrymple &
Zaitlin, 1992; see Fig. 1.6). The morphodynamics of individual macrotidal
estuaries are a function of sea-level changes and prior or inherited topography.
The response of estuaries to sea-level changes, past and future, is affected by
tidal range, nearshore wave climate, river inflow and the nature and supply of
sediment, All estuaries assumed their present form during the rise of sea level
that followed the last glacial maximum, about 18 000 years ago. In areas that
are relatively stable, such as northern Australia, estuaries have had similar sea-
level histories, whereas in areas of very rapid crustal uplift or glacioisostatic
response, estuaries are more likely to have experienced highly individual
relative sea-level histories. The variety of estuaries reflects the range of
submerged prior landforms, from relatively straight, steep coastlines through
valleys in differing stages of infill to rock-barred basins. These different types
of prior topography, which are manifest in coasts such as rias and dendritic
drowned valley harbours, influence the interplay of processes which
redistribute sediment to produce estuarine channels, tidal basins, backwaters
and floodplains. Vegetation can modify estuarine morphodynamics through its
effects upon sediment trapping and through its influence on the shear strength
of channel banks.
This chapter is concerned with macrotidal estuaries, mostly in northern
Australia which tectonically is a very stable region, with low sediment yield
(about 5-15 t km-2a-') and broad continental shelves. Rising postglacial sea
level invaded coastal valleys relatively late, usually within the last 7000-9000
years (Thom & Roy, 1985), reaching a peak around 5500-6000 I4C years BP,
about 1 to 2 m above present sea level at most locations on the Australian

187
188 J. Chappell & C.D. Woodroffe
mainland coast (Chappell et al., 1982, 1983). Since then, relative sea level has
fallen slowly and apparently smoothly to its present position (Chappell, 1982,
1987; Nakada & Lambeck, 1989; Lambeck & Nakada, 1990). In this respect,
north Australian estuaries, in common with much of the Old World tropics,
and northern South America, have a mid to late Holocene sea-level history
which differs from north Atlantic and some other regions where a slowly-
rising trend persisted until the present (Pirazzoli, 1991).
Although this chapter is focussed on macrotidal estuaries in north Australia,
other systems are discussed which differ in terms of sediment yield from their
fluvial catchments, nearshore wave climate or tidal regimes, in order to examine
the effects of boundary conditions. Locations of major north Australian systems
which are examined are shown in Fig. 5.1 and their characteristics are listed in
Table 5.1. Morphodynamics, and sedimentary effects of Holocene sea-level
changes, within macrotidal and selected other systems will be examined and
possible effects of future sea-level changes are considered.

Principal boundary conditions

Most of the macrotidal estuaries in tropical northern Australia in the region
near Darwin are tidal rivers which pass through extensive sedge- and grass-
covered plains, seasonally flooded with fresh water during the monsoonal wet
season. Tidal limits of the longer estuaries lie 90-140km inland. They have
wide entrances typically set between prograded coastal plains, adopting the
funnel-like estuarine form typical of macrotidal rivers (Wright, Coleman &
Thom, 1973). Several distinct meandering channel forms, examined below, are
recognised. Infilled palaeochannels are visible on the neighbouring flood-
plains, and backwater swamps, dominated by paperbark trees (Melaleuca spp.),
lie near the margins of the plains. Most of the fluvial rivers which feed into
these estuaries flow only during the wet season, and discharges are approximately
proportional to catchment sizes. On a global scale the sediment yield of their
catchments is low to very low. To appreciate similarities with and differences
from other types of estuaries, some general effects of boundary conditions are
briefly reviewed.

Thefluvial-marine spectrum
The relative effects of tides, wave climate and fluvial dominance upon delta
morphology were clearly delineated by Wright & Coleman (1973). Estuarine
systems, which typically have elements similar to those of deltas without the
same degree of seaward progradation, vary across a comparable spectrum
(Dalrymple, Zaitlin & Boyd, 1992). The effect of wave climate is
 Macrotidal estuaries 189

0 50 100 1340
km

120

Darwin

tidal river plains

low relief
hills
plateaux

Figure 5.1. Locations of the estuarine systems and catchments in northern Australia
which are referred to in this chapter. 1: Daly River; 2: Elizabeth River; 3: Adelaide
River; 4: Mary River; 5: South Alligator River.

demonstrated effectively by the mesotidal Clarence estuary in New South

Wales (NSW), eastern Australia, which enters the high-energy wave regime of
the southwestern Pacific Ocean. The estuary is separated into two fluvio-
estuarine basins by a bedrock barrier and the outer basin is impounded by a
coastal barrier of marine sands which were driven shoreward during
postglacial sea-level rise (see Chapters 2 and 4); in this respect the outer
 Table 5.1. Parameters of selected estuarine systems

TR A, Q, Qb S Ve re k EW
River (m) (km2 X lo3) (m3/s X lo3) (m3/s X lo3) (ty-I) (m3) (ka) (m-7

Daly 6 50 25 1.5 2 x
105 2 x 1010 6X low
S. Alligator 5 9 30 0.8 4 x
104 3 x 109 6X low
Adelaide 3.5 5 10 0.4 2 104
x 2 x 109 4 x 10-5 low
McArthur 3.5 (d) 12 2 0.6 4 x
104 109 2 x 10-5 v. low
Clarence (I)* 1.5 22 0.1 0.6 2 x
105 2 x 109 (n) v. low
Clarence (0) 1.7 23 0.6 0.6 I 05 2 x 109 10-5 high
Sepik (I)* 1( 4 77 1 6 4 x 107 6 X 10" (n) v. low
Sepik (0) 1(d) 77 <I 6 4 x 107 >I010 10-6 int
Yangtse 4 1800 200 30 109 5 x 10" 3 x 10-5 int

TR = Tidal Range (= 2N, see eqn. (4.1)); A, = catchment area; Q, = mean tidal flux at entrance; Qb = bankfull discharge; and S = fluvial
sediment input. V, = volume of Holocene sediment in estuarine system; re = time in which 90% of sediment V, accumulated, based on
stratigraphic and radiocarbon data; k = funnelling coefficient (see eqn. 4.2); Ew = wave energy at coast; I = inner; 0 = outer; and d = diurnal
tides. All values are estimated for mid-Holocene inner basins of Clarence and Sepik systems; (n), funnelling coefficients k are not known.
 Macrotidal estuaries 191
Clarence basin resembles other estuaries and barrier-basins in southeastern
Australia (Roy, 1984). The inner basin of the Clarence, protected by the
bedrock barrier, has no high-energy Holocene coastal features and, during its
development, the basin divided into extensive low-energy backwater swamps
which appear to have been separated by a prograding birdfoot delta (Fig. 5.2).

1 2 3 4 5 6 7
0 5 10 km

OUTER
BASIN
INNER
BASIN

Figure 5.2. Composite estuarine system of the Clarence River, New South Wales. Key:
1. Pre-Holocene land surface, 2. Holocene inner basin sediments: deltaic sands,
brackish estuarine and swamp muds beneath thin floodplain silts, 3. Fluvial IevCe and
chute-channel sands, 4. Holocene outer basin estuarine and marine sands, 5. Coastal
dune barrier complex, 6. Relict Pleistocene barrier, 7. Estuarine channels and lakes. A
bedrock ridge separates inner and outer Holocene sedimentary basins. Estuarine
sediment of the inner basin is dominantly of fluvial origin with deltaic and brackish
backwater lake and swamp sediments of a low wave-energy environment, capped by
fluvial levCe, chute-channel and overbank sediments. The outer basin is estuarine sands
and muds dominantly of marine origin and is impounded by a high-wave-energy sand
barrier at its seaward margin.
192 J. Chappell & C.D. Woodroffe

The Clarence estuary is fed by the largest of the rivers draining to the NSW
coast and represents the fluvial end of a spectrum of NSW sand-barrier estuaries,
which includes saline coastal lakes and mud basins at the other extreme (Roy,
1984). A comparable spectrum is recognisable amongst macrotidal estuaries of
northern Australia, although two fundamental differences exist. Coastal chenier
and beach-ridge plains of the north are not exactly equivalent to the high-wave-
energy, sand-barrier coasts of southeastern Australia, although comparable
longshore transport processes do occur in the northern, lower-wave-energy
regime (Chappell & Thom, 1986). More significantly, the macrotidal, funnelling
estuaries are actively formed and maintained by tidal flows, whereas the NSW
mesotidal estuary entrances are strongly affected by wave-driven littoral drift
and shoal development, and their dimensions are not solely determined by tidal
flows (Nielsen & Gordon, 1981; Roy, 1984).
Significance of the fluvial factor in macrotidal northern Australia is
expressed by differences between the Daly, South Alligator and the Mary
rivers systems, all of which formed within elongate, shallow prior valleys and
have prograded coastal plains (Chappell & Thom, 1986; Woodroffe er al.,
1986, 1989; Woodroffe, Mulrennan & Chappell, 1993; Chappell, 1993).
Catchment and fluvial discharges of the Daly are largest of these three and the
Mary is smallest (Table 5.1). The Daly, which is shown in Fig. 5.3~1,has a very
actively meandering estuarine channel and scroll plain, the South Alligator
estuary (Fig. 5.51) is characterised by relatively stable, cuspate estuarine
meanders, and the formerly large estuary of the Mary is virtually infilled with
marine sediment. As shown later, active meandering in the Daly estuary is
forced by its large fluvial floods, stable estuarine meanders in the South
Alligator reflect statistical equilibrium between fluvial and tidal flows, while
extinction of the Mary system is partly due to its relatively small fluvial input.

Prior topography
The range of forms of mesotidal estuaries identified by Roy (1984) in eastern
Australia, from drowned valley harbours through barrier estuaries to saline
coastal lake, reflects diversity of prior landscapes which were inundated by
postglacial rising sea level. Constrasting morphologies and Holocene
sediments in the two basins of the Clarence River illustrate the effect of prior
topography (Fig. 5.2). The inner basin, protected by a bedrock barrier, contains
estuarine sediments of dominantly fluvial origin and channel features are
fluvially dominated, while the outer basin contains mostly reworked marine
sands and its anastomosing estuary is impounded by a high-wave-energy sand
barrier. A comparable case, where a bedrock barrier divides inner and outer
 Macrotidal estuaries 193

DRILLHOLE COASTAL PLAIN with cheniers

PALAEOCHANNEL MANGROVE FACIES: proven

MANGROVE FACIES: inferred
MANGROVE FRINGED MEANDERS
SCROLL PLAIN of tidal river
PALAEO-ESTUARY
UPLAND

Elizabeth
Downs

Chambers Bay

Adam Bay

Figure 5.3. Coastal plains, channels, palaeochannels and major mid- to late-Holocene
sedimentary facies which underlie the veneer of floodplain sediment, in ( a ) the Lower
Daly River estuary (after Chappell, 1993) and (b) the Adelaide River estuary (after
Woodroffe et al., 1993).

Holocene fluvio-estuarine sedimentary tracts with contrasting facies and

morphodynamic styles, is the microtidal Sepik system in Papua New Guinea
described by Chappell (1 993).
Macrotidal estuaries in northern Australia also vary with prior topography,
ranging from small tributaries of drowned valleys (e.g., Darwin Harbour,
Semeniuk, 1985a, 1985b) through large, funnelling tidal rivers such as the
194 J. Chappell & C.D. Woodroffe
Daly to the unusual case of the Adelaide River estuary, which has a bedrock
barrier at its entrance and a long reach of sinuous meanders which are inherited
from a prior fluvial river (Fig. 5.3b). While some morphodynamic differences
are due to their differing fluvial inputs, the effects of geometry and depths of
their prior valleys are significant.

Fluvial sediment influx

Holocene sedimentation in estuarine systems is strongly influenced by
sediment yield of their fluvial catchments. Processes in north Australian
macrotidal estuaries with large catchments, such as the Daly, resemble
processes in the macrotidal Fly River delta of southern Papua New Guinea, in
the dynamics of mud recirculation, funnel morphodynamics and formation of
elongate mouth-bar islands. However, the sediment yield of north Australian
catchments is very low (about 10 t km-2 a-I) while the yield from tectonically
active New Guinea reaches more than 1000tkm-2a-1. As a consequence,
large areas of estuarine plain in the north Australian systems are underlain by
vertically accreted mangrove sediments derived from sources to seaward,
while Holocene sediments of the New Guinea systems dominantly are of
fluvial origin (Chappell, 1993). Fluvially derived sand and gravel also
contribute to the infill of macrotidal systems along the higher-relief coast of
Queensland (Hacker, 1988). Morphodynamic effects and responses to sea-
level changes arising from these differences are examined below.

Classification of tidal estuaries

End-points in the morphologic spectrum of deltas reflect relative dominance of
tidal, wave-climate and fluvial regimes (see Chapter 3). Estuaries may be
classified in similar terms (Dalrymple et al., 1992), with sediment yield of the
fluvial catchment as a fourth variable. There is probably no unambiguous way
of expressing form and dynamics in terms of these boundary conditions,
because of the effects of prior topography. However, a classification scheme is
presented here, which takes several of the process factors into account and
separates estuaries with different suites of morphodynamics and sedimentary
facies models. The frame is shown in Fig.5.4. The ratio Q,/Qf of tidal
discharge defines the horizontal axis; the vertical axis is defined by the ratio
VJV, of gross fluvial sediment input, to sediment volume of the estuarine
system. Q, is taken as the mean flood-tide inflow across the entrance, Qf is
bankfull flood discharge entering the estuary and V , is the volume of Holocene
sediment within the estuarine basin. Gross fluvial sediment volume (VJ is
 Macrotidal estuaries 195

I I
I I
I I
10 I I
I I

OSEPIKoII '
I
I a
1 I a
I CC
vi
- I I-
Ve

w
0.1 I I >
OCLARoI U
I
I
I
I
I
I
I
zcn
0.01 I
I
I
I i

0.1 1:o 10 100

Qt/Qf
<RIVER DOM. TIDE DOM.>

Figure 5.4. Classification diagram for estuaries which are developed in Holocene
sediment bodies. The horizontal axis is QJQ, where Q, =mean tidal discharge at
entrance and Q,=bankfull fluvial flood discharge at tidal limit. The vertical axis is
VJV, where V, = total volume of fluvial sediment which entered the system during the
period in which 90% of the Holocene sediment accumulated (including sediment which
has bypassed the estuarine system) and V, = volume of the Holocene estuarine sediment
body. Abbreviations: SALL = South Alligator, McA = McArthur, CLAR = Clarence;
,
subscripts = outer basin, = inner basin. The figure is compiled from data in Table 5.1,
Mary River values were calculated using data for the similar-sized Adelaide River.

estimated by multiplying the mean annual sediment yield by the interval of

time (in years) during which 90% of the Holocene sediment body
accumulated. This interval (&in Table 5.1) is used because in many estuarine
systems the bulk of the sediment accumulated within a few thousand years
centred on the culmination of postglacial sea-level rise, with little
sedimentation in the past several thousand years.
196 J. Chappell & C.D. Woodroffe
Tidal dominance increases with Q,/Qf; vertical dashed lines in Fig. 5.4
separate river-dominated, intermediate and tide-dominated regimes. A value of
VflVe=1 separates those systems where most of the input fluvial sediment is
retained in the estuarine basin (VfIVe<1) from those where some fluvial
sediment bypasses the basin and is deposited offshore (VfIVe>1). When V f l V ,
is significantly less than one, fluvial sources are insufficient to account for
sediment in the basin and it is likely that some sediment is derived from
offshore. Fig. 5.4 shows tropical Australian macrotidal systems considered in
this chapter (plus several other estuaries and deltas for comparison) - both
inner and outer basins of the mesotidal Clarence River estuary, the Daly, Mary
and South Alligator macrotidal estuaries, the mesotidal McArthur, the inner
and outer basins of the microtidal Sepik system described by Chappell (1993)
and the macrotidal Yangtse River delta.

Morphodynamics
Morphodynamics of macrotidal estuaries in northern Australia are forced by
strong tidal flows throughout the year and fluvial floodwater flows for a
relatively short part of the monsoonal wet season. Tides tend to be symmetrical
offshore but become progressively asymmetrical with distance up the tidal
rivers. Flood tides have shorter duration and higher flow velocities, including
tidal bores in more extreme cases, than the ebb tides (Woodroffe et al., 1986;
Vertessy, 1990). This flow asymmetry affects sediment movement. Bedload
transport formulae relate transport to a current velocity term raised to some
power 2, with in the range from 3 to 5 in most formulae (Bagnold, 1966;
TAPSM, 1971). Where highest mean tidal flows are directed upstream the net
bedload sediment movement also tends to be upstream, unless tidal asymmetry
is offset by downstream fluvial flood discharges. Vertical profiles of
suspended sediment concentration (SSC) also vary through the tide cycle.
High velocities and shallow water depths early in the flood tide generate high
turbulent mixing through the water column, whereas sediment settling at high-
tide slack water followed by lower ebb velocities leads to formation of a
lutocline with lesser velocities in the underlying, high-SSC water than in the
upper, low-SSC water (Wolanski et al., 1988). Vertically integrated
velocity-density products differ between flood and ebb flows, which may
cause upstream drift of suspended sediment. Salt-wedge structures, which
develop when there is freshwater outflow, also generate upstream sediment
transport in the underlying saltwater inflow. Finally, evaporative water loss
from the estuary in regions with low relative humidity can produce a net water
and suspended sediment flux into the tidal system (Wolanski, 1986; Chappell,
 Macrotidal estuaries 197
1990). Interaction between these various processes is strongly seasonal in
northern Australia.
In those north Australian macrotidal estuaries, such as the Daly and South
Alligator, which are little affected by bedrock topography and lie completely
within bodies of Holocene sediment, competing upstream- and downstream-
directed flows and sediment movements generate a variety of morphodynamic
elements which are reviewed in the following paragraphs. Examples of the
elements described here can be found in the following systems: the Daly
(Fig. 5.3a), the South Alligator (Fig. 5 . 9 , the Adelaide (Fig. 5.3b) and the
Mary (Fig. 5.7).

DRILLHOLE MANGROVE FACIES: proven

MEANDERTRACT MANGROVE FACIES: inferred
UPLAND COASTAL PLAIN

ESTUARINE FUNNEL SINUOUS CUSPATE UPSTREAM

b N S
4
2
AHD-
2
4

6
8
10
12

Marine sand and mud

Mangrove facies Undifferentiated sediments
Fioodpiain clay Pre-Holocene

Figure 5.5. (a) South Alligator estuary, showing funnel, sinuous, cuspate and anabranching
upstream channel segments, plus (b)Holocene sedimentary facies. Tentative isochrons
indicating coastal progradation are shown. AHD = Australian Height Datum, which
approximates mean sea level. (After Woodroffe et al., 1986, 1993.)
198 J. Chappell & C.D. Woodroffe
The estuarine funnel
Long macrotidal estuaries typically are funnel-shaped in planform, tapering
negative-exponentially upstream (Chappell & Woodroffe, 1985), often with a
few large, dog-leg bends in their lower reaches (Figs. 5.3a, 5.5 and 5 . 1 0 ~ ) .
Radiocarbon dating of sediment cores adjacent to these funnels shows that they
can migrate slowly, sometimes accompanied by progradation of the adjoining
coast, but they appear to maintain their overall form through time (Woodroffe
et al., 1986; Chappell, 1993). Estuarine funnels are tidally generated, but there
are different morphodynamic forms of estuarine funnel - flood-tide dominated,
ebb-dominated and progradational. North Australian macrotidal funnels are
flood-tide dominated and we focus on these, although the other forms are
discussed at the end of this section.
The cross-sectional area of the mouth of most estuaries is closely related to
magnitude of the tidal prism (Byrne, Gammisch & Thomas, 1981). The cross-
section at the point where tidally reversing flow becomes negligible is related
to bankfull floods which enter the tidal river (in this chapter, the point where
reversing flow is negligible at low fluvial stages is referred to as the tidal limit,
despite that elevation of the water surface can vary tidally upstream of this).
The following simple analysis indicates that rate of change of channel width
with distance is determined by shear strength of the channel banks, for a given
tidal regime. Total discharge through a given cross-section during a tidal half-
cycle equals tidal prism upstream of the section. For a simple funnel parallel to
the x-axis, tidal limit at x = L, this is approximately represented for a section at
x by

-
L

W.HU.t I(4NW). dx (5.1)

X

where W is width, HU is the mean product of depth (H) and depth-averaged

velocity U , N is tidal amplitude and t is tidal period. Tidal asymmetry and
progression of the tidal wave along the channel are neglected in the
approximation in eqn (5.1) and can be significant. However, within limits of
these assumptions and provided that N, U and H are constant along the length
of the funnel, which is approximately true in the estuarine funnels of north
Australian tidal rivers such as the South Alligator and Daly (Woodroffe et al.,
1986; Vertessy, 1990), then the solution of eqn. (5.1) is

W = Woexp(-kx), k=4N/tHU (5.2)

where W, is width at the entrance, x = 0. The funnelling coeflcient k defines
the rate of change of channel width with distance and eqn. (5.2) shows that this
 Macrotidal estuaries 199

a SINUOUS BEND (Adelaide)

1krn
Mangrove
STABLE CUSPATE B E N D
(S. Alligator)

FLOOD TIDE
DRY SEASON

1krn
depths in m

C MIGRATING MEANDER (Daly)

1977

1km

Figure 5.6. ( a ) Inherited sinuous meanders of the Adelaide River estuary. The pre-
Holocene palaeochannel is buried beneath 8 m of Holocene ‘big swamp’ mangrove and
estuarine channel sediment. Mangroves on insides of bends indicate extent of bend
migration of the last several thousand years. ( b )Cuspate estuarine meanders of South
Alligator river, showing flood-tide course of maximum flow velocities path (solid
arrows). Note large eddies (dashed arrows) around mid-channel shoals. (c) Rapidly
migrating sinuous estuarine meander on Daly River. Dashed lines indicate scroll bars.
(After Vertessy, 1990.)
200 J. Chappell & C.D. Woodroffe
is related to flow velocity as well as tidal amplitude and channel depth.
Constancy of flow velocity throughout a funnel implies that channel banks are
adjusted to a critical shear stress induced by the tidal current. Actually, the
peak shear force which the channel banks can withstand is likely to be directly
related to peak current velocity; tide-cycle current measurements in the South
Alligator and Daly rivers (Woodroffe et al., 1986, Vertessy, 1990, and
unpublished data) indicate that peak velocity in these systems occurs during
flood tides and is proportional to U .
Macrotidal estuarine funnels in northern Australia conform closely to
exponential eqn. (5.2). Data in Table 5.1 show that the funnels of the South
Alligator and Daly have k-values around 5 X m-' and other macrotidal
estuaries in the same region, which lie within similar bodies of Holocene
sediment (East Alligator, West Alligator, and Wildman), have similar values
of k. Mesotidal systems in tropical Australia have lower k-values and the long
reach of inherited sinuous meanders in the Adelaide River estuary has a low k-
value. Macrotidal reaches upstream of the funnel in the South Alligator and
some other systems also have smaller k-values, reflecting their different
morphodynamics.
Not all funnelling estuaries are flood-tide dominated. Tidal creeks which
drain intertidal flats typically have highest flow velocities during ebb tide (due
to efficient tidal runoff from saturated intertidal flats). Provided that the
assumptions underlying eqn. (5.1) apply and that N , U and H are constant, an
ebb-dominated funnel will be exponential. This distinction between flood- and
ebb-dominated funnels is relevant to the Holocene evolution of funnels,
outlined later. Finally, we note that many funnelling tidal creeks, such as occur
in mangrove swamps, show flow velocity diminishing upstream from a high
value at the mouth to zero near the upstream end. These usually occur in
prograded coastal swamps, and funnelling is presumed to reflect continued
morphodynamic adjustment of the mouth cross-section to the tidal prism,
which enlarges with progradation of the swamp system. In these cases the
channel upstream of the mouth is relict rather than in dynamic equilibrium
with tidal flow.

Estuarine meanders
River bends in macrotidal channels quite commonly are different from the
sinuous bends of meandering fluvial rivers in that, seen in planform, the inner
bank forms a sharp cusp opposite the apex of the bend (Fig. 5.6b). Mid-channel
shoals often occur in the broad reaches which intervene between cuspate
 Macrotidal estuaries 20 1
bends. Ahnert (1960) termed these ‘estuarine meanders’ and noted that ebb-
and flood-tide flow paths occur on opposite sides of the mid-channel shoals.
Through measurements in the monsoonal South Alligator River, Vertessy
( 1990) showed that morphodynamics of cuspate, estuarine meanders are a little
more complex. When they occur, mid-channel shoals tend to be centred
beneath large eddies which form in the lee of cusps (Figs. 5.6b, 5.10b), but
shoals sometimes occur at the channel side and some estuarine meanders are
not sharply and symmetrically cuspate. Separation of flood and ebb flow paths
within a tide cycle occurs in some but not all cuspate meanders although, in
monsoonal north Australia, separation occurs between the flood-tide path in
the dry season and the ebb-tide path, enhanced by fluvial floodwater, in the wet
season. Shoals tend to migrate to and fro with the dominant seasonal flow and
cuspate estuarine meanders are likely to be statistically stationary. Aerial
photographic runs of estuarine meanders in the South Alligator, spanning 40
years, show almost no detectable changes. This contrasts with very active
sinuous meanders observed in the Daly macrotidal estuary, described below.
Abandoned sinuous meanders, pinching out opposite cusps of some
estuarine meanders in northern Australia, indicate that some cuspate reaches
formed after cutoff (Vertessy, 1990). Other estuarine meanders are not
associated with cutoff palaeochannels, however, and appear to be freely
developed where sediment entrainment and transport under reversing flows are
statistically equal and opposite. Once formed, stationary cuspate meanders
show no tendency to revert to the sinuous form, which implies that prior
formation of a ‘parent’ sinuous meander may involve a different fluvio-tidal
regime from that which exists today.

Active meanders and scroll plains

The general morphodynamic equilibrium which exists in cuspate estuarine
meanders is not sustained when there is strong net drift of bedload sediment in
one direction. In this case the cusp, which forms by cutoff of a sinuous
meander, acts as a focus for deposition of a point-bar or point-shoal, and the
meander redevelops a sinuous form by continued accretion which generates a
scroll bar (Fig. 5 . 6 ~ )For
. example, a 30 km reach within the strongly funnelled
Daly River estuary in northern Australia is characterised by active meanders,
with migration rates up to 50 m a - ’ , set into a late Holocene scroll plain with
nests of cutoff meanders (Fig. 5 . 3 ~ ) .Scroll-bar ridges show that active
meanders of the Daly prograded in the downstream direction, evidently during
monsoonal floods.
202 J. Chappell & C.D. Woodroffe
The funnelling coefficient k is relatively constant throughout the estuarine
funnel and active sinuous meandering reaches of the Daly, which suggests that
tidal flows dominate over fluvial discharge in determining channel width
despite the fact that fluvial floods cause meanders to migrate. This contrasts
with the cuspate meandering reach of the South Alligator, where channel
widths are locally highly variable but the average channel width is relatively
constant through a reach of about 30 km. This seems anomalous because Q,/Qf
is larger in the South Alligator than in the Daly, and further, detailed analysis
of flow regimes is needed in order to explain this.

Inherited channels
Cuspate and active estuarine meanders are dynamic forms within macrotidal
estuaries which lie within Holocene sediment bodies, and contrast with
inherited fluvial channels which persist through Holocene estuarine
development, although they are modified by tidal flows and originate in prior
riverine plains which may outcrop at the surface or even may be buried by
estuarine sediments. The Adelaide River estuary is a fine example, which
includes a 60km reach of inherited sinuous meanders (Figs. 5.3b, 5 . 6 ~and
5.10~).The prior riverine plain lies beneath about 8 m of Holocene estuarine
and mangrove sediment while the tidal river channel is 10-12m deep and,
except for minor progradation under mangroves on insides of bends, it
generally conforms to the prior fluvial river course. Inherited meanders in
some cases appear to have become cutoff progenitors of cuspate estuarine
meanders, after postglacial sea level stabilised and estuarine plains became
fully developed.

Anabranches and distributaries

Not all tidal river systems are single, funnelling or meandering channels, but
can be compounds of multiple channels formed of anabranches and
distributaries. When present, distributary channels occur in lower reaches of
tidal rivers and extend to the coast, whereas anabranches, which are alternative
channels which diverge only to rejoin further downstream, occur in upper
reaches of tidal rivers (examples in the South Alligator are described by
Woodroffe et al., 1986). Distributary channels are funnelled and typically have
estuarine meanders or dog-leg bends, while anabranches are of rather constant
width and apparently form through avulsion during fluvial floods entering the
upper tidal river.
 Macrotidal estuaries 203

Extinguished systems
Although some macrotidal rivers historically have been rather stable and
changed only by channel migration, sinuous-cuspate meander transition or by
anabranching in late Holocene times, such as the Alligator group in northern
Australia, others changed more dramatically. The Adelaide River estuary
(Fig. 5.36) switched from its prior, highly sinuous meandering course to an
alternative distributary funnel, and back again to the sinuous channel in late
Holocene times (Woodroffe et al., 1993). During the same period, the nearby
Mary River estuary developed two major, meandering funnels (Fig. 5.7).
These later became filled with sediment and died as tidal systems, so that the
fluvial discharge of the Mary River largely dissipated into extensive
swamplands and evaporated with relatively minor discharge to the sea through
lesser, avulsive distributaries which diverted the fluvial flow. The former
estuarine system of the Mary is thought to have infilled with sediment driven
tidally upstream, after fluvial discharge was diverted (Woodroffe et al., 1993;

DRILLHOLE COASTAL PLAIN with cheniers

PALAEOCHANNEL MANGROVE FACIES: proven
DATES ON CHANNEL INFILL 3000 MANGROVE FACIES: inferred

2000
1000-1600
Sampan
Creek 4000
Shady
1200-2600 Camp
1200
Tommycut I900
Creek
3400
3600 3300 Alligator
Head

2
0 km 10

Figure 5.7. Extinguished estuarine system of the Mary River, with times of channel
extinction based on radiocarbon dates (years BP) of sediment infills (after Woodroffe,
et al., 1993).
204 J. Chappell & C.D. Woodroffe
Woodroffe & Mulrennan, 1993). Radiocarbon dates support the chronology of
successive phases of channel activity and infilling in the Mary system, shown
in Fig. 5.7.

Estuarine floodplains
Macrotidal estuarine floodplains in north Australian regions of higher rainfall
typically have a low convex profile, descending at a very low gradient away
from the tidal river with backwater swamps where the floodplain meets the
adjacent hills. Sedges and grasses with low herbs cover the plains and swamp
forests occur near the margins. Tree-lined levtes occur on the banks in upper
reaches of the tidal rivers and mangrove creeks, often draining infilled
palaeochannels, join the main channel.
Today, there is slow tidal invasion of some floodplain areas, shown by
headward-extending tidal creeks colonised by juvenile mangroves (Figs. 5.8
and 5.10d). Owing to convexity of the floodplain, invading tidal waters flow
readily into low-lying areas within reach of these creeks and the freshwater
sedgelands and swamp forests are overtaken by mangrove or by bare saline
mudflats. Attrition of freshwater vegetation and extending saline flats is
occurring on floodplains adjoining parts of the lower reaches of several
estuaries. This is particularly conspicuous in the lower Mary River, where
salinisation of the plains is activated by tidal creek extension, cutting into the
extensive freshwater floodplains (Knighton, Mills & Woodroffe, 199 1 ;
Knighton, Woodroffe & Mills, 1992), and by erosion of sedgeland where scarp
retreat at the interface with saline mudflat leaves small remnants of sedgeland
vegetation within the extending salt flats (Fig. 5.8).

Chenier and beach-ridge plains

Prograded coastal plains with chenier and/or beach ridges adjoin most of those
north Australian estuaries which are associated with substantial bodies of
Holocene sediment (Rhodes, 1982; Chappell & Thom, 1986). The transition
from estuarine plain to coastal plain often is indicated by a chenier or beach
ridge with a I4C age around 5000-6000 years BP, marking the coastline of or
soon after the peak postglacial sea level. Widths of coastal plains adjoining the
larger north Australian estuaries range from about 1 to 20 km and this variation
depends more on depth to the buried prior landsurface than the fluvial
discharge from nearby rivers.
Progradation of the coastal plain causes seaward translation of the entrance
and funnel of an adjacent estuary, which enlarges the tidal prism. In flood-tide
 Macrotidal estuaries 205

1943 S 1950 s
T T

1963 S 1973 S
t T

1989 S
T

T Tommycut Creek
S Sampan Creek
Palaeochannel Boundary

Okm 2 4 8 8 10

Shady Camp

Figure 5.8. Recent headward extension of minor tidal creeks and areas of tributary salt
flats which have formed from prior freshwater floodplain, in lower Mary River area
(after Knighton et al., 1991).
206 J. Chappell & C.D. Woodroffe
dominated systems, the consequent increase of tidal dominance (increased
Q,/Qf) is likely to enhance the tendency for sediment to move landward up the
funnel. This process, associated with coastal progradation, may have
contributed to extinction of the Mary estuary in the past few thousand years
(Fig. 5.7). Radiocarbon chronologies of the subsurface sediments of chenier
plains of most north Australian macrotidal systems indicate most rapid
progradation around 5000-3000 years BP, with negligible net change in the last
2000 years (Woodroffe et al., 1989).

Net sediment movement and estuary development

Sediment movements upstream by tidal transport, downstream by fluvial
flooding and overbank onto the estuarine floodplain (Fig. 5.9) are implicated in
morphodynamics of most of the elements of macrotidal estuaries reviewed in
the preceding paragraphs. The balance between opposed tidal and fluvial
transport, which influences morphodynamics of the middle reaches of
macrotidal estuaries (Anhert, 1960; Dalrymple, et al., 1992), is affected by the
Q,/Qf ratio. Morphodynamics of north Australian macrotidal systems with
different Q,/Qf values differ, accordingly. Funnelled estuaries in this region
include one case with very active sinuous meanders and scroll plains (the
Daly), several with long reaches of stable, cuspate estuarine meanders (e.g.,
South and East Alligator Rivers), and the extinguished estuarine system of the
Mary River. The smaller the fluvial influence, the weaker the seaward
transport of sediment during wet-season floods and the stronger the tendency
for upstream-moving sediment to remain within the estuary. Seaward-
migrating sinuous meanders in the Daly estuary, which develop rapidly from
cusps after meander cutoff, reflect net downstream sediment transport forced
by high wet-season discharges from the relatively large catchment of this river.
The South Alligator, with its smaller catchment, has stable cuspate meanders
and probably maintains a neutral sediment budget. The yet smaller Mary
estuaries have become extinguished through sedimentary infill from seaward.
The Adelaide system, with a catchment similar to that of the Mary, formerly
had a funnelled estuary which probably infilled in the same manner,
redirecting the tidal river back to its former, fluvially defined channel
(Fig. 5.3b)
Sedimentary infill of the lower Mary and the former Adelaide estuary may
partly reflect an increase of Q,/Qf either through coastal progradation, as
suggested earlier, or from diminution of rainfall since mid Holocene times.
Alternatively, the former estuarine funnels may have developed under an ebb-
dominated tidal regime, within extensive mangrove swamp; in this scenario,
 Macrotidal estuaries 207

original channel

a channel widens, releasing sediment

b mud from nearshore shoals moves upriver

c fluvial sediment spills to floodplain

Figure 5.9. Schematic diagram of macrotidal funnel showing ( a ) overbank flooding and
sediment deposition associated with estuary widening, ( b )upstream sediment transport
in flood-tide dominated regime, and ( c )downstream transport and overbank deposition
due to fluvial flooding. Intertidal shoals are shaded.

subsequent transition from mangrove to floodplain, through continued

sedimentation, would induce a flood-tide dominated regime, which eventually
led to extinction of the former estuaries. These and other aspects of Holocene
evolution of macrotidal systems are explored in the next section.
208 J. Chappell & C.D. Woodroffe
Postglacial sea-level changes and estuary evolution
Rising sea level invaded the rather shallow, broad valleys of far northern
Australia towards the end of the postglacial transgression (Chappell & Polach,
199 1). The following summary of ensuing sedimentation and estuary develop-
ment is based on extensive drillhole studies reported by Woodroffe et al.
(1986, 1989, 1993), Woodroffe & Mulrennan (1993), and Chappell (1990,
1993). The prior valley floor is indicated by either quartzose alluvial sands, or
lateritic gravel, or highly oxidised sandy clay. This prior surface is shown by
thermoluminescence dating to be of Late Pleistocene age (unpublished data: D.
Price, pers. commun., 1992), and lies about 14m below the estuarine plains of
the South Alligator and Daly systems and at 8-10m depth in the Adelaide and
Mary systems. Mangrove facies composed of organic muds containing
mangrove pollen and wood, with local mangrove peat, commonly overlies the
prior valley floors, recording the inundation of the valley by the rising sea level
around 7000-8000 years BP. Transgressive basal sediments pass upwards into
mangrove facies showing that .these macrotidal estuaries became areas of
vertical accretion under widespread mangrove forest, which has been termed
the ‘big swamp’, throughout the last 8 m or more of sea-level rise (Woodroffe,
Chappell & Thom, 1985; Woodroffe et al., 1987; Chappell, 1993). The presence
of tidal channels within these swamps is shown by buried sediment bodies of
sands and muds containing foraminifera with estuarine affinites. Most of these
lie close to the present tidal rivers but some are up to 2-3 km away (Wang &
Chappell, in press). Facies relationships are illustrated in Fig. 5Sb, using the
example of the South Alligator system.
These vertically accreting sedimentary facies underlie a thin cover
(0.5-2 m) of freshwater floodplain sediments which accumulated after sea
level stabilised. The transition from mangrove-dominated ‘big swamp’
sedimentation to freshwater sedge and grassland plains is identified by pollen
analysis of drillcores, which typically show succession, at shallow depth, from
Rhizophoraceous communities to mixed mangroves, especially Avicennia,
followed by a sharp passage to freshwater sedges and grasses (Woodroffe
et al., 1985; Russell-Smith, 1985; Chappell & Grindrod, 1985; Grindrod,
1988; Clark & Guppy, 1988). Transition from mangrove ‘big swamps’ to
freshwater estuarine plains occurred soon after sea level stabilised, around
5000-6000 years BP, in the Daly and South Alligator systems, and somewhat
later in the Mary and other smaller systems (see Fig. 1.5). Sediments of the
rising-sea-level phase are interrupted near the axis of each valley by deposits
of the meandering tidal rivers; the extent of the meander tract reflects the
balance between fluvial input and tidal energy, being most extensive at the
Daly River, narrower and less active at the South Alligator, and least on the
Adelaide plains. Changes of positions and geometries of the estuarine channels
 Macrotidal estuaries 209

during the last 6000 years are shown by surface traces of palaeochannels and
subsurface facies of channel sands and muds which often show lamination and
point-bar structures. Progradation of the coastal plains seawards from the
innermost fossil shoreline occurred in the last 6000 years and decreased in rate
in the last 2000-3000 years in most systems other than the Daly.
Extinguished estuarine funnels of the Mary and Adelaide rivers imply that
these formed under fluvio-tidal regimes different from those of today, as noted
earlier. Radiocarbon-dated estuarine sediments within lower reaches of the
Daly and South Alligator indicate that estuarine funnels existed during the ‘big
swamp’ phase of rising sea level (Woodroffe et al., 1986; Chappell, 1993).
These are likely to have been ebb-dominated due to tidal flooding within the
extensive mangroves; the transition from mangrove ‘big swamp’ to freshwater
plains would have been accompanied by a change to flood-dominated regimes
and enhancement of upstream sediment transport leading to infilling of smaller
systems.

Comparison of Australian macrotidal rivers with other systems

Estuaries are a sink for sediment, and gain sediment both from upstream by
direct riverine input and from seaward by tidal and salt-wedge circulations
(Guilcher, 1967; Nichols and Biggs, 1985; Dronkers, 1986). North Australian
macrotidal rivers lie within Holocene sediment bodies dominantly derived
from seaward; in terms of the classification proposed above (Fig. 5.4), systems
such as the South Alligator and Daly lie in the region defined by high Q,/Qf
and low Vf/Ve. Other, comparable north Australian macrotidal systems have
been described, including the Ord in Western Australia (Wright et al., 1973;
Wright, Coleman & Thom, 1975; Thom, Wright & Coleman, 1975; Coleman
and Wright, 1978) and the Fitzroy River estuary (Jennings, 1975; Semeniuk,
1980, 1982). Stratigraphic relationships and chronology suggest a similar
pattern of development to that of the South Alligator and neighbouring rivers
in the Northern Territory (Woodroffe, 1988). It is useful to compare the north
Australian systems with others which either are macrotidal with relatively
higher fluvial sediment yields from their catchments (higher VfIV,), or are
mesotidal with similar fluvial sediment yields (lower Q,/Qf). There are
systematic morphodynamic differences in both cases which relate to Vf/Ve and
Q,/Qf values, rather than to other boundary conditions.
Rivers entering the Gulf of Carpentaria drain catchments with geomorphic
and climatic regimes which are similar to those of the Daly-Alligators
macrotidal suite but their estuaries have substantially smaller values of Q,/Qf
as tides are dominantly diurnal and have maximum tidal range of about 3 m.
Wave climate varies with exposure and affects estuarine morphology; the
210 J. Chappell & C.D. Woodroffe
deltaic McArthur estuary, for example, is protected by an offshore group of
islands and has two major and several minor distributary channels (Woodroffe
& Chappell, 1993), while the more exposed Norman estuary has a single
channel traversing a very broad chenier plain (Rhodes, 1982). The Gilbert
River, on the eastern margin of the Gulf of Carpentaria, similarly is dominated
by riverine processes and morphology throughout much of its tidal reach
(Jones, Martin & Senapati, 1993). All have relatively low funnelling
coefficients (k), reflecting lesser tidal influence than the macrotidal cases.
These systems lack the extensive mid Holocene ‘big swamp’ mangrove
sediments that characterise the north Australian macrotidal systems, described
above, and prograded late Holocene coastal sediments overlie mid Holocene
nearshore and estuarine sands and muds.
The Bay of Fundy, with perhaps the world’s largest tidal range, includes the
Cobequid Bay macrotidal estuarine system which shares many morphological
similarities with north Australian systems (Dalrymple et al., 1990). Dalrymple
et al. (1 992) identify three estuarine segments with different channel morphology.
A relatively straight seaward funnel, dominated by tidal and marine processes,
has sediment moving in a net landward direction; a central reach is a zone of
sediment convergence and has a sinuous, highly meandering channel; and the
upstream reach is relatively straight, least influenced by marine processes and
has net seaward sediment movement. Similarly, an estuarine funnel and upper
estuarine meandering channels are recognised in the Gironde estuary in
southwestern France but, in this case, coastal wave energy is high and the outer
region is being transgressed by a sandy barrier (G. Allen, 1991; Allen &
Posamentier, 1993). In this system all the fluvially derived sand, and about
75% of the mud, is deposited in the estuary with only a minor output of mud to
seaward (Allen et al., 1980; Castaing & Allen, 1981). The macrotidal Severn
estuary in southwestern Great Britain has similar features. Studies, which
commenced last century, suggest landward movement of sand (Sollas, 1883;
Murray & Hawkins, 1977; Wang & Murray, 1983). The exact pathways of
landward sand transport are still not clear (Harris & Collins, 1985, 1988,
199 1 ; Stride & Belderson, 1990), whereas it appears that the dominant source
of mud in the Severn at present is from rivers which feed the estuary (J. Allen,
1991), there may be a net export of mud to the continental shelf through cliff
erosion of channel margins related to relative sea-level rise (Allen, 1987,
1990).
Despite similarities between these systems and those of northern Australia,
there are differences. Longitudinal sand bars andahoals are more prominent in
the outer funnels and estuary-embayments of,dhe Severn and Bay of Fundy
/
than in most north Australian macrotidal systems (with the partial exception of
 Macrotidal estuaries 21 1
the Daly: Fig. 5.3a). Perhaps the most significant difference is represented by
mid-Holocene development, in north Australian systems, of mangrove ‘big
swamps’ in which sedimentation kept pace with rising sea level. We note that
extensive Holocene swamp facies occur in cool temperate systems such as the
English Fenlands (Godwin, 1978), but these appear to be time-trangressive
horizons associated with sea-level changes. Our tropical, vertical sequences
may have no exact counterpart in cool temperate regions, not only by virtue of
their mangrove vegatation. Earlier, we suggested that vertical sedimentation of
mangrove-facies muds in north Australian palaeo-estuaries was sustained by
landward ‘pumping’ of muds in this highly seasonal, highly evaporative
environment, even in ebb-dominated regimes. We note that extensive mid
Holocene mangrove sediments overlain by freshwater sediments occur in other
tropical, western Pacific estuaries such as the macrotidal Klang River
(Coleman, Gagliano & Smith, 1970) and other sites in southeast Asia
(Anderson & Muller, 1975; Haseldonckx, 1977; Bosch, 1988; Kamaludin,
1993; Woodroffe, 1993). Not all of these are macrotidal and, in the light of
tropical, mesotidal cases which lack a mid Holocene mangrove ‘big swamp’
phase, such as those outlined above in the Gulf of Carpentaria, it appears that
further work is required before the boundary conditions for vertically accreting
‘big swamps’ are identified.
Several recent facies models of macrotidal estuaries have been linked to the
procedures of sequence stratigraphy (Cobequid Bay, Dalrymple et al., 1992;
Daly River, Chappell, 1993; Gironde estuary, Allen & Posamentier, 1993).
These adopt similar approaches, identifying a lowstand system tract
(equivalent to the pre-Holocene alluvial valley floor in the north Australian
examples), a transgressive tract, and various highstand system tracts.
Differentiation of transgressive and highstand system tracts is not always easy;
in the case of the Daly, Chappell(l993) considered that the maximum flooding
surface separates the transgressive tract from a vertical sedimentation tract,
within which there is an estuarine-freshwater transition which appears
synchronous across the estuarine plains. Within the highstand tract, different
rivers in north Australia demonstrate different degrees of reworking of their
floodplains with different extents of meander tract. A regressive tract is
represented by progradation of the coastal plain.

Likely effects of future sea-level rise

Freshwater ecosystems of the estuarine plains and backwater swamps of
northern Australia are endangered if sea level rises in future, which is a widely
anticipated consequence of global warming. Tops of channel banks are close to
P
 Macrotidal estuaries 213

the level of high spring tide throughout lower reaches of most macrotidal rivers
of the region and, away from main tidal-river channels, floodplain surfaces and
backwater swamps lie at or below the elevation of highest spring tide levels at
the coast. Saline waters today intrude into these areas where small, high-tide
creeks breach the low, broad levees. Networks of small creeks which have
extended rapidly in the lower Mary system, causing extensive salinisation of
former freshwater floodplain over the last 40 years, are shown in Fig. 5.8; this
example dramatically illustrates a process which has occurred in parts of many
estuarine plains in northern Australia (see Fig. 5.106). Although this historical
process arguably is attributed to compaction of sediments and the effects of
feral buffalo rather than sea-level rise, as outlined earlier (see Woodroffe &
Mulrennan, 1993), it indicates the most likely first stage of floodplain
inundation which would occur if sea level were to rise in future.
Substantial overbank flooding would occur at high tide if highest tide level
were increased by only 30-50 cm, overtopping the levees, whether as the result
of global eustatic sea-level rise or a local increase of tidal amplitude. The extra
volume of water, flooding overbank, would increase the total tidal prism and
flow velocities in the channel would increase if all the additional flow was
carried by the channel and its dimensions remained constant. Evidence cited
earlier indicates that channel cross-section throughout estuarine funnels of the
north Australian macrotidal type is adjusted to the maximum flow velocity.
When the tidal prism is contained within the tidal river, maximum velocities
occur at low- to mid-flood tide. Overbank flooding with rising sea level will
enlarge the tidal prism but channel response is not necessarily immediate,
because a modest degree of overbank flooding close to high water will not
affect maximum flow velocities. Ebb velocities will be enhanced by return
flow from the floodplain, however, and rising sea level will cause a gradual
shift from flood- to ebb-dominance. A threshold is crossed when enhanced ebb
velocities exceed the peak velocity which prevailed before sea-level rise
commenced, and it is likely that channel adjustment to rising sea level will be
slight until this occurs. Once the threshold is exceeded, channel widening is
likely to proceed more rapidly. Alteration of the floodplain ecology will
commence before the threshold is reached, however, by salt-water invasion
similar to the historical processes in the lower Mary system (Fig. 5.10d).
Once initiated by rising sea level, channel widening may contribute
sediment to the estuarine plains. Sediment tends to remain in the estuarine
system of long macrotidal, tropical rivers even when ebb dominated, and
sediment released by channel enlargement should find its way overbank onto
the plains. This will offset the effect of flooding and may lead to steady vertical
sedimentary accretion. In this respect, the past is a partial key to the future:
214 J. Chappell & C.D. Woodrofle
patterns of development of macrotidal estuaries under rising sea level in earlier
Holocene times may be revisited in future. This review has outlined some of
the morphodynamic factors which need more precise study before the
consequences of any future scenario can be forecast accurately.

Acknowledgements
We particularly wish to thank Eric Wolanski and Rob Vertessy for valuable
discussions in the course of this work, and all those who participated with us in
research in northern Australia in the past decade, especially Monica
Mulrennan, Eugene Wallensky and Colin Campbell.

References
Ahnert, F. (1960). Estuarine meanders in the Chesapeake Bay area. Geographical
Review, 50,390-401.
Allen, G.P. (1991). Sedimentary processes and facies in the Gironde estuary: a recent
model for macrotidal estuarine systems. In Clastic tidal sedimentology, ed.
D.G. Smith, G.E. Reinson, B.A. Zaitlin & R.A. Rahmani, pp. 2 9 4 0 . Canadian
Society of Petroleum Geologists Memoir, 16.
Allen, G.P. & Posamentier, H.W. (1993). Sequence stratigraphy and facies model of
an incised valley fill: the Gironde estuary, France. Journal of Sedimentary
Petrology, 63,378-9 1.
Allen, G.P., Salomon, J.C., Bassoullet, P., Du Penhoat, Y. & De Grandpre, C. (1980).
Effects of tides on mixing and suspended transport in macrotidal estuaries.
Sedimentary Geology, 26,69-90.
Allen, J.R.L. (1 987). Reworking of muddy intertidal sediments in the Severn Estuary,
southwestern U.K.: a preliminary survey. Sedimentary Geology, 50, 1-23.
Allen, J.R.L. (1990). Salt-marsh growth and stratification: a numerical model with
special reference to the Severn Estuary, southwest Britain. Marine Geology, 95,
77-96.
Allen, J.R.L. (1991). Fine sediment and its sources, Severn Estuary and inner Bristol
Channel, southwest Britain. Sedimentary Geology, 75, 57-65.
Anderson, J.A.R. & Miiller, J. (1975). Palynological study on a Holocene peat and a
Miocene coal deposit from NW Borneo. Review of Palaeobotany and
Palynology, 19,291-35 1.
Bagnold, R.A. (1966). An approach to the sediment transport problem from general
physics. US Geological Survey Professional Paper, 422( I).
Bosch, J.H.A. (1988). The Quaternary deposits in the coastal plains of Peninsular
Malaysia. Geological Survey of Malaysia, Quaternary Geology Section, Report
QG/1 of 1988.
Boyd, R., Dalrymple, R. & Zaitlin, B.A. (1992). Classification of clastic coastal
depositional environments. Sedimentary Geology, 80, 139-50.
Byrne, R.J., Gammisch, R.A. & Thomas, G.R. (1981). Tidal prism - inlet areas
relationships for small tidal inlets. In 17th International Conference on Coastal
Engineering, Sydney, pp. 244-5. New York: American Society of Civil
Engineers.
 Macrotidal estuaries 215
Castaing, P & Allen, G.P. (1981). Mechanisms controlling seaward escape of
suspended sediment from the Gironde: a macrotidal estuary in France. Marine
Geology, 40, 101-18.
Chappell, J. (1982). Evidence for smoothly falling sea levels relative to north
Queensland, Australia, during the past 6000 years. Nature, 302, 406-8.
Chappell, J. ( 1 987). Late Quaternary sea-level changes in the Australian region. In
Sea-level changes, ed. M.J. Tooley & I. Shennan, pp. 296-331. New York:
Blackwells.
Chappell, J. (1990). Some effects of sea level rise on riverine and coastal lowlands. In
Lessons f o r human survival: nature’s record from the Quaternary. pp. 37-49.
ed. P. Bishop Geological Society of Australia Symposium Proceedings, I .
Chappell, J. (1993). Contrasting Holocene sedimentary geologies of lower Daly River,
northern Australia, and lower Sepik-Ramu, Papua New Guinea. Sedimentary
Geology, 83, 339-58.
Chappell, J., Chivas, A., Wallensky, E., Polach, H. A. & Aharon, P. (1983). Holocene
palaeo-environment changes, central to north Great Barrier Reef inner zone.
BMR Journal of Australian Geology and Geophysics, 8, 223-35.
Chappell, J. & Grindrod, J. (1985). Pollen anlysis: key to past mangrove communities
and successional changes in North Australian coastal environments. In Coastal
and tidal wetlands of the Australian monsoon region, ed. K.N. Bardsley, J.D.S.
Davie & C.D. Woodroffe, pp. 225-36, North Australia Research Unit
(Monograph), Darwin: ANU press.
Chappell, J & Polach, H. (1991). Postglacial sea-level rise from a coral record at
Huon Peninsula, Papua New Guinea. Nature, 349, 147-9.
Chappell J., Rhodes E.G., Thom B.G. & Wallensky E. (1982). Hydro-isostasy and the
sea-level isobase of 5500 B.P. in north Queensland, Australia. Marine Geology,
49,8 1-90.
Chappell, J. & Thom, B.G. (1986). Coastal morphodynamics in north Australia:
review and prospect. Australian Geographical Studies, 24, 1 10-27.
Chappell, J. & Woodroffe, C.D. (1985). Morphodynamics of Northern Territory tidal
rivers and floodplains. In Coastal and tidal wetlands of the Australian monsoon
region, ed. K.N. Bardsley, J.D.S. Davie & C.D. Woodroffe, pp. 85-96, North
Australia Research Unit (Monograph), Darwin: ANU press.
Clark, R.L. & Guppy, J.C. (1988). A transition from mangrove forest to freshwater
wetland in the monsoon tropics of Australia. Journal of Biogeography, 15,
665-84.
Coleman, J.M., Gagliano, S.M. & Smith, W.G. (1970). Sedimentation in a Malaysian
high tide tropical delta. In Deltaic sedimentation: modern and ancient, ed. J.P.
Morgan, pp. 185-97. Society of Economic Palaeontologists and Mineralogists,
Special Publication, 15.
Coleman, J.M. & Wright, L.D. (1978). Sedimentation in an arid macro-tidal alluvial
river system: Ord River, Western Australia. Journal of Geology, 86, 621-42.
Dalrymple, R.W., Knight, R.J., Zaitlin, B.A. & Middleton, G.V. (1990). Dynamics
and facies model of a macrotidal sand-bar complex, Cobequid Bay-Salmon
River estuary (Bay of Fundy). Sedimentology, 37, 577-612.
Dalrymple, R.W., Zaitlin, B.A. & Boyd, R. (1992). Estuarine facies models:
conceptual basis and stratigraphic implications. Journal of Sedimentary
Petrology, 62, 1030-43.
Dronkers, J. ( 1 986). Tide-induced residual transport of fine sediment. In Physics of
shallow estuaries and bays, Lecture notes on coastal and estuarine studies, vol.
16, ed. J. van de Kreeke, pp. 228-44, Berlin: Springer-Verlag.
216 J. Chappell & C.D. Woodroffe
Godwin, H. ( 1 978). Fenland; its ancient past and uncertain future. Cambridge
University Press.
Grindrod, J. (1988). The palynology of Holocene mangrove and saltmarsh sediments,
particularly in northern Australia. Review of Palaeobotany and Palynology, 55,
2294.
Guilcher, A. (1967). Origins of sediments in estuaries. In Estuaries, ed. G.H. Lauff,
pp. 149-57. American Association for the Advancement of Science Publication,
83.
Hacker, J.L.F. (1988). Rapid accumulation of fluvially derived sands and gravels in a
tropical macrotidal estuary: the Pioneer River at Mackay, North Queensland,
Australia. Sedimentary Geology, 57,299-3 15.
Harris, P.T. & Collins, M.B. (1985). Bedform distributions and sediment transport
paths in the Bristol Channel and Severn Estuary, U.K. Marine Geology, 62,
153-66.
Harris, P.T. & Collins, M.B. (1988). Estimation of annual bedload flux in a macrotidal
estuary: Bristol Channel, U.K. Marine Geology, 83, 237-52.
Harris, P.T. & Collins, M.B. (1991). Sand transport in the Bristol Channel: bedload
parting zone or mutually evasive transport pathways. Marine Geology, 101,
209-16.
Haseldonckx, P. (1977). The palynology of a Holocene marginal peat swamp
environment in Johore, Malaysia. Review of Palaeobotany and Palynology, 24,
227-38.
Jennings, J.N. (1975). Desert dunes and estuarine fill in the Fitzroy estuary,
north-western Australia. Catena, 2 , 2 15-62.
Jones, B.G., Martin, G.R. & Senapati, N. (1993). Riverine-tidal interactions in the
monsoonal Gilbert River fandelta, northern Australia. Sedimentary Geology,
8 3 , 3 19-37.
Kamaludin, b.H. (1993). The changing mangrove shorelines in Kuala Kurau,
Peninsular Malaysia. Sedimentary Geology, 83, 187-97.
Knighton, A.D., Mills, K. & Woodroffe, C.D. (1991). Tidal creek extension and saline
intrusion in northern Australia. Geology, 19, 83 1-4.
Knighton, A.D., Woodroffe, C.D. & Mills, K. (1992). The evolution of tidal creek
networks, Mary River, northern Australia. Earth Surface Processes and
Landforms, 17, 167-90.
Lambeck, K. & Nakada, M. (1990). Late Pleistocene and Holocene sea-level change
along the Australian coast. Palaeogeography Palaeoclimatology and
Palaeoecology (Global and Planetary Change Section), 89, 143-76.
Murray, J.W. & Hawkins, A.B. (1976). Sediment transport in the Severn Estuary
during the past 8000-9000 years. Journal of the Geological Society of London,
132,385-98.
Nakada, M. & Lambeck, K. (1989). Late Pleistocene and Holocene sea-level change
in the Australian region and mantle rheology. Geophysical Journal, 96,
497-5 17.
Nichols, M.M. & Biggs, R.B. (1985). Estuaries In Coastal sedimentary environments,
2nd edn, ed. R.A. Davis, Jr. pp. 77-186, New York: Springer-Verlag.
Nielsen, A.F. & Gordon, A.D. (1981). Tide1 inlet behavioural analysis. In
Proceedings 17th International Coastal Engineering Conference, Sydney,
pp. 2517-33. New York: American Society of Civil Engineers.
Pirazzoli, P.A. (1991). World atlas of Holocene sea-level changes. Amsterdam:
Elsevier.
Rhodes, E.G. (1982). Depositional model for a chenier plain, Gulf of Carpentaria,
Australia. Sedimentology, 29, 201-21.
 Macrotidal estuaries 217
Roy, P. S. (1984). New South Wales estuaries: their origin and evolution. In Coastal
geomorphology in Australia, ed. B.G. Thom, pp. 99-122, Sydney: Academic
Press.
Russell-Smith, J.J. (1985). A record of change: studies of Holocene vegetation history
in the South Alligator River region, Northern Territory. Proceedings of the
Ecological Society of Australia, 13, 191-202.
Semeniuk, V. (1980). Quaternary stratigraphy of the tidal flats, King Sound, Western
Australia. Journal of the Royal Society of Western Australia, 63,65-78.
Semeniuk, V. (1982). Geomorphology and Holocene history of the tidal flats, King
Sound, north-western Australia. Journal of the Royal Society of Western
Australia, 65,47-68.
Semeniuk, V. (1985a). Development of mangrove habitats along ria coasts in north
and northwestern Australia. Vegetatio, 60, 3-23.
Semeniuk, V. (1985b). Mangrove environments of Port Darwin, Northern Territory:
the physical framework and habitats. Journal of the Royal Society of Western
Australia, 67, 81-97.
Sollas, W.J. (1883). The estuaries of the Severn and its tributaries: an enquiry into the
nature and origin of their tidal sediment and alluvial flats. Quarterly Journal of
the Geological Society of London, 38.61 1-28.
Stride, A.H. & Belderson, R.H. (1990). A reassessment of sand transport paths in the
Bristol Channel and their regional significance. Marine Geology, 92, 227-35.
TAPSM (Task Committee for Preparation of Sediment Manual) (1971). Sediment
transportation mechanics. H. Sediment discharge formulae. Journal of the
Hydraulics Division, American Society of Civil Engineers, 97,523-67.
Thom, B.G. & Roy, P.S. (1985). Relative sea levels and coastal sedimentation in
southeast Australia in the Holocene. Journal of Sedimentary Petrology, 55,
257-64.
Thom, B.G., Wright, L.D. & Coleman, J.M. (1975). Mangrove ecology and
deltaic-estuarine geomorphology, Cambridge Gulf-Ord River, Western
Australia. Journal of Ecology, 63,203-22.
Vertessy, R. (1990). Morphodynamics of macrotidal rivers in far northern Australia.
Unpublished PhD thesis, Australian National University.
Wang, P. & Chappell, J. (in press). Foraminifera as Holocene environmental
indicators in the South Alligator River, Australia. Marine Geology.
Wang, P. & Murray, J.W. (1983). The use of foraminifera as indicators of tidal effects
in estuarine deposits. Marine Geology, 51,239-50.
Wolanski, E. ( 1 986). An evaporation-driven salinity maximum zone in Australian
tropical estuaries. Estuarine Coastal and Shelf Science, 22,415-24.
Wolanski, E., Chappell, J., Ridd, P. & Vertessy, R. (1988). Fluidisation of mud in
estuaries. Journal of Geophysical Research, 93,235 1-61.
Woodroffe, C.D. (1988). Changing mangrove and wetland habitats over the last 8000
years, northern Australia and Southeast Asia. In Northern Australia: progress
and prospects, vol. 2, Floodplains research, ed. D. Wade-Marshall &
P. Loveday, pp. 1-33, North Australia Research Unit, ANU Press.
Woodroffe, C.D. (1993). Late Quaternary evolution of coastal and lowland riverine
plains of Southeast Asia and northern Australia: an overview. Sedimentary
Geology, 83, 163-75.
Woodroffe, C.D. & Chappell, J. (1993). Holocene emergence and evolution of the
McArthur River Delta, southwestern Gulf of Carpentaria, Australia. Sedimentary
Geology, 83, 303-17.
Woodroffe, C.D., Chappell, J., Thom, B.G. & Wallensky, E. (1986).
Geomorphological dynamics and evolution of the South Alligator River and
plains, N.T. North Australia Research Unit Monograph, ANU Press,
218 J. Chappell C? C.D. Woodrofe
Woodroffe, C.D., Chappell, J., Thom, B.G. & Wallensky, E. (1989). Depositional
model of a macrotidal estuary and floodplain, South Alligator River, Northern
Australia. Sedimentology, 36, 737-56.
Woodroffe, C.D. & Mulrennan, M.E. (1993). Geomorphology ofthe Lower Mary
River Plains, Northern Territory. Darwin: North Australia Research Unit.
152 pp.
Woodroffe, C.D., Mulrennan, M.E. & Chappell, J. (1993). Estuarine infill and coastal
progradation, southern van Diemen Gulf, northern Australia. Sedimentary
Geology, 83,257-75.
Woodroffe, C.D., Thom, B.G. & Chappell, J. (1985). Development of widespread
mangrove swamps in mid-Holocene times in northern Australia. Nature, 317,
71 1-13.
Woodroffe, C.D., Thom, B.G., Chappell, J., Wallensky, E., Grindrod, J. & Head, J.
( 1 987). Relative sea level in the South Alligator River region, North Australia,
during the Holocene. Search, 18, 198-200.
Wright, L.D. (1985). River deltas. In Coastal sedimentary environments, 2nd edn,
ed. R.A. Davis, pp. 1-76. New York: Springer-Verlag.
Wright, L.D. & Coleman, J.M. (1973). Variations in morphology of major river deltas
as functions of ocean wave and river discharge regimes. American Association
of Petroleum Geologists Bulletin, 57, 370-98.
Wright, L.D., Coleman, J.M. & Thom, B. G. (1973). Processes of channel
development of a high-tide range environment: Cambridge Gulf-Ord River
Delta, Western Australia. Journal of Geology, 81, 15-4 I .
Wright, L.D., Coleman, J.M. & Thom, B.G. ( 1 975). Sediment transport and
deposition in a macrotidal river channel: Ord River, Western Australia. In
Estuarine research, vol. 11, Geology and engineering, ed. L.E. Cronin,
pp. 309-21. London: Academic Press.
 6
Lagoons and microtidal coasts
J.A.G. COOPER

Introduction
This chapter deals with the evolution of lagoonal coasts. Despite being widely
distributed around the globe, coastal lagoons have been studied in a
fragmented manner and few co-ordinating studies have been undertaken.
Lagoons exhibit many different morphologies and form under a variety of
environmental conditions which determine not only the principal
morphodynamic processes but also the evolutionary path which a lagoon
follows. This chapter outlines various types of lagoon. It then reviews several
documented examples of lagoonal evolution and assesses the main
evolutionary processes. Variations between documented studies of lagoon
evolution are assessed and the main controls on the evolutionary path are
discussed. Existing models of lagoonal evolution are reviewed and current
research deficiencies are highlighted.

Coastal lagoon distribution

Lagoons are present on many coasts of the world in a variety of environmental
settings; they are most common in microtidal environments, although
examples do occur in mesotidal and even macrotidal environments (Hayes,
1975). A broad range of physical and chemical characteristics consequently
exists. Current estimates (Table 6.1) suggest that lagoons border between
17% (Cromwell, 1971) and 56% (Berryhill, Kendall & Holmes, 1969) of
North America, and 13% of the world’s coastline (Berryhill er al., 1969;
Cromwell, 1971; Barnes, 1980) from the polar to tropical latitudes. The wide
discrepancy in the estimates for North America highlights the difficulty in
definition of a barrier lagoon. Lagoons may be even more widespread as many
occur on bedrock coasts in discrete embayments, rather than linear barrier
island coasts.

219
220 J.A.G. Cooper
Table 6. I . Distribution of barrierhgoon coasts by continent

Barriedlagoon % of continent’s % of world’s

Continent coast (km) coastline lagoonal coast
N. America 10765 17.6 33.6
Asia 7126 13.8 22.2
Africa 5984 17.9 18.7
S. America 3302 12.2 10.3
Europe 2693 5.3 8.4
Australia 2168 11.4 6.8
Total 32038

From Barnes (1980), originally after Cromwell (1971).

Coastal lagoons are generally considered ephemeral features on a

geological time scale as they form, evolve and infill within a short timespan.
This chapter is concerned with the processes by which they form and evolve.
Evolutionary processes vary according to environmental setting, and as
lagoons occur in most latitudes and occupy a transitional location between land
and sea, their morphodynamic evolution varies accordingly. Most studies of
lagoon morphodynamics have centered on one or a few examples (Emery &
Uchupi, 1972) and relatively few studies (Lankford, 1977; Nichols & Allen,
1981; Kjerfve, 1986; Carter et al., 1989; Nichols, 1989) have synthesised
information on a variety of lagoons.
Precise definition of a coastal lagoon is problematic and many definitions
have been proposed. Considerable overlap between lagoons and estuaries has
been identified. As morphodynamic systems, lagoons have been defined as
‘coastal water bodies which are physically separated, to a greater or lesser
extent, from the ocean by a strip of land’ (Ward & Ashley, 1989). The
imprecise definition of coastal lagoons is perhaps the main problem in the lack
of co-ordinated research, as many features variously termed estuaries, blind
estuaries, embayments, coastal ponds, coastal lakes, bays and sounds may
alternatively be regarded as lagoons. Carter (1 988) identified three types of
lagoon according to the flux of water within them. In one group the inflow of
seawater equals the outflow over a tidal period. In the second, evaporation of
lagoon water means that the inflow of water exceeds the outflow. This
produces a sabkha. In the third group there is a net seaward discharge due to
the addition of freshwater discharge. Nichols & Allen (1981) identified four
main lagoon types based on dominant processes: estuarine lagoon; open
lagoon; partly closed lagoon; closed lagoon. Many other classifications of
 Lagoons and microtidal coasts 22 1
lagoons exist, tailored to the needs of the proposing author. Several
morphodynamic realms exist within lagoons and the typical features are
depicted in a generalised model in Fig. 6.1.

Processes of lagoon formation

Lagoons form where coastal embayments or depressions are separated from the
adjacent sea by a barrier. Barriers comprise either clastic material (sand or gravel)
or are created by vegetation, coral growth or tectonics (Lankford, 1977). Lagoons
are best formed on transgressive coasts, particularly where the continental margin
has a low gradient and sea-level rise is slow (Emery, 1967); however, they are
also present on embayed rocky coasts (Roy, 1984; Cooper, 199 1a).
Clastic barriers originate in different ways. They may arise through spit
growth across an embayment through longshore drift (Hoyt 1967; Martin,
Gadel & Barusseau, 1981), behind cuspate barriers produced by wave-
reworked deltaic sediment, or through flooding of coastal lowlands behind
former dune ridges or beaches (Hoyt, 1967). In the latter scenario the original
width and depth of a lagoon depends on the antecedent slope of the land
surface and extent of sea-level rise. In other cases lagoon enclosure results

WASHOVER DUNES
TIDAL CHANNELS,
.BEACH
TIDAL FLATS
FLUVIAL
DELTA

MARSH FLOOD EBB TIDAL

TIDAL DELTA
DELTA
14L€7

LAGOON

BARRIER
BEACH
SHOREFACE

Figure 6. I . Generalised depiction of morphosedimentary features of coastal lagoons

(modified after Reinson, 1980). Rarely will all features or salinity zones be present in a
single lagoon and the relative importance of each may vary with time.
222 J.A.G. Cooper
from landward transport of a barrier across the shoreface during transgression
until it becomes stabilised. If excess transgressive sediment is available the
barrier may then prograde seaward, typically by beach-ridge formation
(Phleger & Ewing, 1962; Thom, 1983; see also Fig. 1.6). If additional sediment
is unavailable the lagoon may be stressed such that it breaches, or the barrier
may degenerate as barrier sediment is eroded (Phleger, 1981). In some cases
clastic barriers completely enclose a lagoon, leaving no connection with the
sea. This typically arises where the sediment supply and transport capacity of
littoral drift or wind exceeds tidal or fluvial currents which act to maintain a
connecting channel.
Early concepts of lagoon evolution are typified by the work of Lucke (1934)
whose conceptual model was termed the ‘Lucke Model’ by Oertel et al.
(1989). According to this model (Fig.6.2) coastal lagoons evolve from a
marine embayment to a relatively deep, partially enclosed, back-barrier
lagoon. Sedimentation in the lagoon occurs through development of flood-
tidal deltas into marsh islands, a process recently documented in detail by
Cleary, Hosser & Wells (1979), and fluvial sedimentation in the form of deltas.
Ultimately, the originally deep lagoon is transformed into a marsh or deltaic
plain through which rivers discharge to the sea. This is the most widely held
concept of lagoon evolution at present (Barnes, 1980) but, as will be shown
below, several alternative views have recently been expounded. In some cases,
however, unquestioning acceptance of this model has led to misconceptions
regarding lagoonal processes and consequent attribution of apparent deviations
to human impacts. Begg (1978, 1984), for example, attributed the shallow
nature of Natal lagoons to increased sediment yield from recently cultivated
catchments, without recognising that naturally high-sediment yields from the
humid, subtropical hinterland throughout the Holocene had rendered most
lagoons in a mature stage of evolution several thousand years before
widespread cultivation (Cooper, 199 1a).

Examples of lagoonal evolution

In this section a variety of examples is used to illustrate the variation in
processes of evolution in different settings and under different conditions. In
the subsequent section they are compared.

South-east African lagoons

The microtidal coastline of Natal and Zululand represents an excellent natural
laboratory for the study of coastal lagoons in a humid (annual rainfall
1000mm) subtropical setting. Mean tidal range at Durban is 1.72 m on spring
 Lagoons and microtidal coasts 223

a INLET BARRIER OCEAN

SHALLOW LAGOON
CHANNEL

SHALLOW LAGOON /TIDAL FLATS

MA1NLAND

MARSH DEVELOPMENT

Figure 6.2. A summary diagram of the ‘Lucke model’ of lagoon evolution (after Oertel
et al., 1989). This represents one of the earliest, and perhaps still most widely held,
concepts of lagoon evolution. An initially deep lagoon ( a ) becomes progressively
infilled by sediment accumulation ( b ) (in this case from marine sources) and is
ultimately transformed into a swamp with fluvial channels ( c ) .
224 J.A.G. Cooper
and 0.5 m at neap tide. The Natal coast, whose northern boundary is the Tugela
River mouth, is a linear, rock-based coastline with shallowly indented
embayments and sandy beaches overlying rock outcrops. The coastal
hinterland rises steeply to over 3000m and the continental shelf ranges in
width from 10 km to 40 km (Fig. 6.3). The 100 m isobath marks the edge of the
shelf. The Zululand coast, north of the Tugela River, has a low-lying sandy

KOSl LAGOON

LAKE
-
T
2 SlBAYA
3
a
v
3

Y)
L

0' LAKE
c1 sr. LUCIA

IUGELA R

DURBAN

Figure 6.3. The coastal hinterland of Natal and Zululand, showing continental shelf
width and hinterland topography. Only the major rivers are shown. Note the steep
gradients in Natal (south of Tugela River), compared with those in Zululand (north of
Tugela River), where an extensive coastal plain occurs. The Natal coast is mainly rocky
and owes its linear nature to its tectonic orgin. The Zululand coastal plain comprises
unconsolidated sands and muds. Aeolianite and beachrock form the main solid
outcrops. 0indicates beach-ridge plain.
 Lagoons and microtidal coasts 225
coastal plain which widens northward to 80 km and the continental shelf is as
narrow as 3 km. There is a prograding beach-ridge plain for the 45 km north of
the Tugela River. On this strongly wave-dominated section of the coast, river
courses are diverted northward in the direction of net littoral drift to produce a
number of shore-parallel coastal lagoons.

Siyai Lagoon
The Siyai Lagoon in southern Zululand (Fig. 6.4) is located on a coastal sector
where historical progradation rates average up to 5 m per year, but are strongly
episodic (Cooper, 1991b). The lagoon drains an 18km2 catchment and is
generally isolated from the sea. Barrier breaches caused by heavy rainfall are
generally sealed within a week (Begg, 1978). Between 1937 and 1977 the
barrier of the Siyai Lagoon prograded by about 125 m through beach-ridge
accretion and subsequent stabilisation by dune vegetation. This coastal
progradation through northward longshore transport of sediment from the
Tugela River caused the northward elongation of the Siyai Lagoon by 800 m
(Fig. 6.4), a 35% increase in total length. Over the same period rapid growth of
reeds in the back-barrier caused loss of open water area as the channel
narrowed and apparently shallowed. Reduction in species diversity of fish and
aquatic birds was also noted (Begg, 1978; Benfield, 1985) and attributed to
decreased frequency of outlet formation and incursion of seawater. The
changes could be attributed to reduced stream gradient and loss of stream
power, and to the use of fertilisers and destruction of riparian vegetation in the
catchment.

Accretion
1937- 1977
Lagoon extension
0 500m 1937-1977
Beachridge
crest-lines

SlYAl LAGOON
Dense Vegetation 1937
mouth

1937 1977
mouth
1977
H WM
Indian Ocean

Figure 6.4. The Siyai Lagoon on the prograding coast of southern Zululand, showing
the seaward edge of stabilised beach ridges in 1937 and 1977. Progradation of the coast
by 125 m (light shading) was accompanied by elongation of the lagoon through 800 m
(dark shading), causing reduced frequency of outlet formation, reed encroachment,
shallowing of the lagoon and decreased species diversity.
 PONTA 00
Tidal flats WRO

H o l o m e beahond dunes

L w m I redimentr

Fleirtaene dune mymlex Inlet

Bathymetry

2 km

Figure 6.5. The Kosi lagoon in northern Zululand. Segmentation has given rise to a unique set of aquatic environments and habitats in each
segment. The fetch and depth of each segment determines morphodynamic processes. Key: I . Amanzimnyama, 2. Nhlange, 3. Mpungwini, 4.
Makawulani, 5. Khalu Inlet, 6. Zilonde. The modern bathymetry (right, in metres) preserves the outline of an incised Pleistocene river valley
in which the lagoon formed.
 Lagoons and microtidal coasts 227
Kosi Lagoon
The Kosi lagoon in northern Zululand is an excellent example of the effects of
segmentation (Zenkovich, 1959) on lagoon evolution (Fig. 6.5). The lagoon is
situated on the sandy Zululand coastal plain (see Fig. 6.3) and is fed by several
small streams with negligible sediment input, other than suspended vegetal
matter. The evolution of the lagoon has not been studied in detail but
preliminary results of seismic profiling and vibracoring suggest that Holocene
sediments are very thin (<1 m), and comprise mainly gyttja. Underlying and
adjacent sediments of probable Late Pleistocene age (Cooper, Kilburn & Kyle,
1989) show that the modern lagoon reoccupied the site of a Pleistocene lagoon
in an inter-dune depression. Present water depths up to 3 5 m in the lagoon
reflect incision during the last glacial maximum (Hill, 1969). Closure of a
former outlet at Bhanga Nek by aeolian deposition isolated the southern part of
the lagoon from the sea. Contact with the sea is maintained through a narrow
northern outlet.
Closure of the southern outlet may have been associated with regression
from a Holocene highstand for which there is ample geomorphological
evidence in the area. Low sediment supply from the narrow shelf and coastal
streams prevented coastal progradation during regression; instead, the
coastline has evolved into a series of log spiral bays in apparent equilibrium
with the ambient wave field, backed by high aeolian dunes.
Segmentation of the Kosi Lagoon was enabled by its uncohesive, sandy
margins from which sediment was reworked by wind-generated lagoon waves
to form cuspate spits which ultimately coalesced, dividing the lagoon into a
series of segments. Preliminary results from seismic profiling suggest that
progradation of the spits into deeper water occurred through avalanching and
turbidite deposition. The course of the original incised channel is still evident
from the deep sections of adjoining segments (Fig. 6.5). Segmentation of the
lagoon was aided by fluctuations in late Holocene sea levels which elevated
formerly subtidal parts of the cuspate spits. Narrow channels connect the
segments. The origin of these channels is uncertain: they may have arisen by
incision associated with falling water levels after floods or sea-level
highstands, or as paths followed by hippopotami.
Segmentation of the lagoon has produced a series of distinctive environ-
ments and habitats. Lake Amanzimnyama (diameter IOOOm) is a 1.5m deep
marginal basin whose water is entirely fresh due to its distance from the tidal
inlet. Lake Nhlange, (diameter 5000m) is dominated by wind waves which
cause mixing throughout its 35 m depth. Long-term (decade-scale) salinity
changes, between 5 and O%O, occur (Kyle, 1986). The lake margins are
dominated by wave-rippled sand while fine organic matter settles in the deeper
n
c
e
 Lagoons and microtidal coasts 229

basins. Lake Mpungwini occupies a central position in the lake system and
receives seawater on each tidal cycle. Its great depth (maximum 21 m) and
small fetch (1 800 m) has produced marked salinity stratification ranging from
over 2 0 % ~at depth to less than 5 % at ~ the surface. Anoxic and sulphidic
conditions occur at depth (Ramm, 1992). During equinoctial high tides
accompanied by strong winds the stability of the water column is broken down
and fish kills occur when sulphidic water reaches the surface. Lake
Makawulani is 8 m deep and displays marked stratification. Salinities range
from 25 %O at the bottom to 10%0at the surface. Between this lake and the tidal
inlet is a 5 km stretch of low-gradient sandy intertidal flats with a narrow
channel. Comparison of historical aerial photographs indicates that intertidal
deposition is minimal at present but former landward progradation of the tidal
flats is indicated by small deltas in Lake Makawulani and the Khalu inlet. The
current lack of tidal sedimentation may be attributed to lack of longshore
supply of sand on the equilibrium planform coastline.
Kosi lagoon has preserved its incised channel almost intact through low
depositional rates from the catchment. Its stability is further enhanced by the
log-spiral coastline, which minimises littoral drift and prevents inlet closure
and flood-tidal delta buildup. Longshore sand supply is further reduced by the
large distance from the closest sediment supplying river ( 175 km to the south)
and interception of shelf-sediment pathways by submarine canyons (Ramsay,
1990).

Lake St Lucia
Lake St Lucia is the largest lagoon in Zululand and covers an area of 380 km2
(Fig. 6.6). Several large rivers form deltas in the lagoon which has an average
depth of 1.5 m (Hobday, 1976). The modern lagoon occupies the same site as
a Late Pleistocene lagoon, deposits of which (Port Durnford Formation) are
preserved locally (Hobday & Orme, 1974). During the last glaciation these
deposits were partly eroded and incised to depths up to 40 m (Sydow, 1987);
seismic profiling reveals the presence of a former river channel under the
lagoon (Fig. 6.6). For much of the Holocene an outlet was maintained through
a northern channel. This was ultimately sealed by aeolian deposition, perhaps
associated with slight regression during the late Holocene (Sydow, 1987).
Subsequently, an outlet formed in the present position in the southern end of
the lagoon.
Cores through the lagoonal fill show various mixtures of clay, silt and fine
sand containing molluscan and foraminifera1 fauna typical of brackish
lagoonal conditions (Hobday, 1976). Lagoonal confinement was evidently
established at an early stage of the transgression behind an earlier Pleistocene
230 J.A.G. Cooper
barrier dune. Sedimentation probably kept pace with transgression as there is
no evidence of particularly deep water sedimentation. In marginal areas of the
lagoon, brackish lagoonal deposits overlie pre-Holocene freshwater swamp
deposits, reflecting progressive drowning of surrounding swamps during
transgression.
Due to its shallow nature and large water area Lake St Lucia has been
reported to become hypersaline when fluvial discharge is low (Orme, 1973;
Begg, 1978). The presence of gypsum in cores from 20 m below the present
sediment surface (Hobday, 1976) provides evidence of similarly shallow
conditions early in the evolution of the lagoon. Segmentation of the lake is
only weakly developed (Orme, 1973) a fact which may be linked to rocky
outcrops on the landward lagoon margin and to the presence of cohesive
muddy sediment whose cohesion limits wave-transport and shoreline
realignment.
The entire lake is now above wave base (Hobday, 1976) and wave-
generated currents and direct wave action are the principal sediment dispersal
agents. The high sediment level may be in part a reflection of sedimentation
during high Holocene levels, since which time fluvial discharge has been
unable to erode the accumulated muddy sediment.
Around the St Lucia Lagoon, Late Pleistocene lagoon and barrier sediments
of the Port Durnford Formation are preserved. Focusing of drainage channels
into low-lying areas during the last glacial maximum caused differential
erosion of Pleistocene deposits from the back-barrier area but preserved them
under the Pleistocene barrier. Holocene transgression drowned the incised
valley and deposited additional sand against the Pleistocene barrier while a
new set of lagoonal deposits were laid down in the incised valley. The nature
of the Holocene transgressive sequence has not been investigated but may well
duplicate that of the Late Pleistocene, documented by Hobday & Orme (1 974)
and Hobday & Jackson (1 979). This sequence comprises a lower lagoonal
facies, overlain by inter-layered washover sands and lagoonal muds and a
barrier with an upper aeolian component (Fig. 6.7).

Small bedrock-confined lagoons in Natal

On the mainly rocky continental margin of Natal south of the Tugela River,
small lagoons occupy fluvial palaeovalleys incised in bedrock during lowered
sea levels. The lateral confinement of these channels has led to complete
erosion of pre-Holocene deposits (Maud, 1968). Most of these lagoons are
relatively narrow ( 4 0 0 m) and coast-normally orientated but some exhibit
limited coast-parallel extension behind sandy barriers. In most cases river
 Lagoons and microtidal coasts 23 1

4m

Aeolian dunes
0 200111

Washover deposits
Tidal channel
Peat
Barrier complex

Lagoonal facies

Older Quaternary Deposits Oyster bed

Figure 6.7. A model of the trangressive barrier lagoon deposits of the Pleistocene Port
Durnford Formation preserved under the Pleistocene barrier of Lake St Lucia (after
Hobday &Jackson, 1979).

discharge is insufficient to maintain an outlet. The sandy barriers are

dominated by overwash lamination and have an upper aeolian component.
Flood-tidal deltas are absent and ephemeral ebb-tidal deltas form opposite
flood-generated breaches (Cooper, 1990b). The ephemeral outlet channels are
shallow and are rapidly reworked after formation. In this respect they are
similar to the outlets described from the Oregon coast (Clifton, Phillips &
Hunter, 1973)
Stable conditions are characterised by high water levels and low-energy
conditions in which suspension settling dominates sedimentation. The main
sediment sources are the catchment, surrounding vegetation and local bank
erosion. Marine sediment input is limited to overwash and deposition is
restricted to the immediate barrier area. Sediment supply from the hinterland is
high and can be correlated directly with catchment size (Cooper, 1991a). The
volume of overwash input depends on the length and height of the barrier. In
 J.A.G. Cooper
uMgababa Lagoon (Grobbler, 1987; Grobbler, Mason & Cooper, 1987)
marine sediment input through overwash is low due to the short (100 m) barrier
while in Mhlanga Lagoon the 900 m-long barrier promotes overwash
sedimentation (Cooper 1989).
Periodic, short-lived, high-energy phases are precipitated by barrier
breaching and outlet formation (Cooper, 1990b). This appears to be caused
mainly by increased fluvial discharge as noted by Carter, Johnson & Orford,
(1 984) in Irish lagoons. In the cell-confined barriers of Natal, where long-term
changes are minimal (Cooper, 199 lc) alternative processes of breaching such
as stretching and separation (Carter er al., 1989) have not been documented.
Overwash may lower the barrier and enable rising water levels to form an
outlet there (Begg, 1984; Cooper, 1989). Rapid outflow of water following
breaching causes erosion of accumulated fine sediment from the bed of the
lagoon and promotes selective retention of sand (Cooper 1989; Grobbler 1987).
Floodplain deposits comprise vegetated wetlands, typically fringed with
either Phragmites reeds or Barringtonia, the ‘freshwater mangrove’. True
mangroves are generally absent. Back-barrier channels which are compara-
tively unconfined are shallow and sandy, while those which are laterally
confined by vegetated or muddy banks are deeper.
Cores from such lagoons reveal impounded brackish water conditions early
in their evolution. Incursions of the sea during late Holocene highstands may
have promoted breaching more frequently, as evidenced by the presence of
brackish water molluscan remains in some lagoonal deposits above present sea
level. At present most lagoons appear to have reached equilibrium (Cooper,
1989, 1991a); fine-grained fluvial sediment resides temporarily in the lagoon
before being scoured during breaching. This preferentially preserves sand in
the sedimentary record. Slow buildup of overwash sands adjacent to some
barriers occurs as outflowing currents are insufficient to remove such grain
sizes.
In Mdloti Lagoon (Grobbler, 1987), the preservation of a widespread
lagoonal mud in the lagoon fill (Fig. 6 . 8 ~reflects
) a period when deeper-water
conditions prevailed and fluvial currents were unable to scour deeply enough
to remove this sediment. Brackish-water conditions are indicated by the
molluscan shell remains. This is overlain by fluvial sands, which now
dominate the lagoon. In contrast, the valley fill of the Mtwalume Lagoon
(Fig. 6.8b), which has similar dimensions to the Mdloti, is almost entirely
sandy. This was interpreted by Cooper (1991a) as a result of higher sediment
supply from the inflowing river which continually kept pace with drowning of
the bedrock valley in a similar manner to estuaries of the southeast United
States (Nichols & Biggs, 1985).
 Lagoons and microtidal coasts 233

a Mdloti N

mud
muddy sand
sand
conglomerate
bedrock

borehole
5 b Mtwalume N
10

-10

-20

200 300 400

iml

Figure 6.8. Cross-sections of sedimentary fills of the bedrock-confined ( a ) Mdloti

Lagoon (after Grobbler, 1987) and ( b ) Mtwalume Lagoon (after Orme, 1974) on the
mainly rocky Natal coast. The presence of a well-defined muddy horizon in the Mdloti
fill indicates deep-water brackish lagoonal conditions. The upper sands are fluvial and
were deposited in shallow water: mud is preferentially eroded during floods. In the
Mtwalume the entire fill is dominated by sand, suggesting that deep-water conditions
were never attained and coarse-grained fluvial sedimentation kept pace with drowning
of the bedrock valley.
234 J.A.G. Cooper
Coastal lagoons of Virginia, USA
The coastal lagoons of Virginia receive relatively low fluvial inflow and are
largely filled with marsh deposits and narrow tidal channels. Inlets in the sandy
barriers are maintained by tidal currents. Their evolution is characterised by
barrier retreat, resulting from rising sea level and a nearshore sediment deficit
(Finkelstein & Ferland, 1987), which has led to lagoon narrowing and
reduction of tidal prisms. This in turn caused inlet constriction and reduction in
tidal current strength. A corresponding increase in sedimentation was noted on
marshes and tidal flats. Stratigraphic cross-sections show changes from a high-
energy lagoon to muddy tidal flat and marsh environments, accompanied by
increasingly quiescent conditions, during a sea-level rise of about 6 m in the
past 5000 years. The resultant ‘regressive’ back-barrier stratigraphy was
produced under transgressive conditions through sedimentation under
progressively lower-energy conditions.
Infilling of these lagoons occurred principally through fluvial inputs, which
were probably increased by deforestation since the Colonial period. Under
slowly rising sea level the lagoon infilled while under more rapid sea-level rise
it was drowned. In the past 40 years an increased rate of sea-level rise has been
matched by increased sedimentation, which has prevented lagoon rejuvenation.

Port Royal Bay, Bermuda

Port Royal Bay is a sheltered lagoon, surrounded by high ground and sheltered
from wind action. A summary of its Holocene evolution is included here as an
example of lagoonal development under low-energy conditions and low
sedimentation rates (Ashmore & Leatherman, 1984).
At approximately 9500 years BP, rising sea level increased the groundwater
table to support a freshwater swamp on top of pre-Holocene soils in a
depression in aeolianite bedrock. Continued rise of sea level outstripped
swamp aggradation and by 8200 BP a freshwater pond was formed in which
2.5 m of gyttja accumulated. This pond became increasingly brackish with
time as seawater percolated through the surrounding aeolianite. The lagoon
was drowned about 7000 BP when the sea overtopped its margins. This enabled
colonisation by marine molluscs and corals. Low sedimentation rates
accompanying continued sea-level rise produced deeper water until reducing
conditions and development of a seasonal thermocline led to the demise of the
molluscan population and deposition of lime mud in the deeper sections. Coral
growth was maintained in shallow marginal areas. The lagoon is over 2 5 m
deep as a result of low sedimentation rates but Ashmore & Leatherman (1984)
predict continued infilling of the lagoon by coral growth under present sea-
level conditions.
 Lagoons and microtidal coasts 235
Warnbro Sound, Australia
Warnbro Sound, a semi-enclosed lagoon in Western Australia, represents a
lagoon with very low terrigenous sedimentation in which wave action affects
the marginal areas (Carrigy, 1956). The lagoon is protected from ocean waves
by a line of aeolianites, which are the remnant of a former barrier. A self-
generating sedimentary system of calcareous sands and silts derived from
pelagic foraminifera1 and benthic molluscan debris has led to a sedimentary
equilibrium in the lagoon. Fine sediment is restricted to the deeper sections
(>14 m) while the shallow, wave-dominated marginal areas are sandy.
Warnbro Sound receives negligible clastic sediment input either from marine
or terrestrial sources and carbonate precipitation and dissolution balance to
preserve the equilibrium. Excess marine sediment is removed by longshore
drift.

Lagoons in New South Wales, Australia

On the embayed bedrock coast of New South Wales (NSW), lagoons occupy
incised fluvial valleys, often protected from direct ocean influences by large
embayments. Roy (1984) presented an evolutionary model of the evolution of
lagoons or ‘saline coastal lakes’ based on widespread drilling records and
observation of modern lagoons at various stages of development (Fig. 6.9).
These evolved from a relatively deep back-barrier basin of marine to marginal
marine salinity. Water area and volume were reduced gradually by fluvial delta
progradation and barrier overwash and at the same time salinity variations
increased, favouring prolonged fresh to brackish conditions. This change was
accompanied by swamp encroachment and deposition of freshwater muds.
Ultimately, the lagoons were infilled by swampy floodplains surrounding a
narrow channel. The possibility of increased barrier breaching as a result of
channelisation in the back-barrier was cited as a possible evolutionary trend.

Coastal lagoons in Rio de Janeiro State, Brazil

The lagoons of the embayed rocky coastline of Rio de Janeiro State, Brazil are
completely isolated from the sea by sandy barriers. They formed during the
Holocene and the earliest dates for their formation are about 7800 years BP
(Ireland, 1987). Subsequently, sea level has oscillated around its present level
in a series of minor transgressions and regressions. In most cases Holocene
lagoonal sediments are only 4 to 5 m thick, and are underlain by similar pre-
Holocene lagoonal facies. The typical sedimentary succession comprises a
basal non-marine sand or peat, passing upward into clay and clayey sand of
brackish to marine affinity in a low-energy environment. Diatoms indicated
236 J.A.G. Cooper

Barrier sands
Overwash sands
Flood-tidal delta
Estuarine mud
Floodplain deposits
Fluvial channel

Inlet
b

Lagoon transformed to floodplain

Inlet
C
Figure 6.9. Model of evolution of saline coastal lakes in New South Wales (after Roy,
1984; see also Fig. 2.6 and Fig. 4.24). This model predicts gradual infilling of an
initially deep basin (a) and subsequent transformation into a swamp ( b ) with confined
fluvial channels ( c ) . It is similar to the Lucke (1934) Model; however, the principal
sediment source is fluvial rather than marine.
 Lagoons and microtidal coasts 237
steadily declining marine influence with time and a concomitant increase in
freshwater influence. The sequence is capped by a layer of freshwater peat and
organic detritus, indicating complete closure of the barrier inlet. Salinity
variation in the back-barrier deduced from diatom assemblages could be linked
to periods of increased or decreased connection with the sea via barrier inlets.
The general evolutionary trend was for barriers to be breached as sea level rose
and to close when it stabilised. Although no explanation was offered for barrier
breaching during transgression, it was probably a result of low sediment
supply to the barrier. Lagoonal fills for each transgressive phase suggest that
sedimentation rates were comparatively rapid.

Lagoons of Maine, USA

The coastal lagoons of Maine occur on an irregular rocky coastline and number
over 200 (Duffy, Belknap & Kelley, 1989). Their coarse-grained barriers are
typically <1000m long and assume a number of forms including barrier spits,
looped barrier, cuspate barriers, double tombolos and pocket barriers.
Sedimentation in these barrier lagoon systems began about 9000 years BP and
reached maximum development when sea level rise slowed after 5000 BP.
Barrier sediments were eroded from glacial till as sea level rose. Limited
sediment supply and rapid sea-level rise caused the barriers to retreat into
embayments where they became sheltered from littoral drift. Holocene
lagoonal facies accumulated on top of glaciomarine sediments and a thin
regressive tidal flat or beach deposit. The 3 to 4 m thick Holocene facies
comprises gyttja and three types of marsh with distinctive flora. Up to 0.8 m of
gyttja accumulated in shallow brackish-water lagoons early in their evolution.
It is typically overlain by swamp deposits whose plant remains indicate
increasingly marine conditions. This transgressive sequence, which is similar
to those identified in adjacent areas of New England and Boston, therefore
reflects increased marine influence during relative sea-level rise and is the
most common stratigraphic sequence found. Fresh- to brackish-water swamps
accumulated as lagoons infilled or as water levels were lowered by outlet
formation.
Some lagoons showed prolonged periods of brackish-water conditions,
without increasing marine influence with time, while a few showed variations
in marine influence with time. In a few lagoons a reverse sequence of
progressively fresher water conditions was indicated by marsh progradation
over back-barrier flats. The latter, regressive, situation is interpreted as a result
of abundant sediment or a slow rate of sea-level rise, which permitted marsh
expansion. Duffy et al. (1989) concluded that the presence of a barrier inlet
238 J.A.G. Cooper
was the most important control on back-barrier stratigraphy in Maine.
Variations in barrier morphology, rate of sea-level rise, tidal range and bedrock
morphology were not reflected in back-barrier stratigraphy.

Discussion and comparison

Comparison of the examples above shows that there is wide variation in the
evolution of coastal lagoons. Lagoon evolution was summarised by Kjerfve &
Magill (1989) as ‘the combined action of marine and fluvial processes ...
[leading to] ... trapping and infilling of semi-enclosed coastal systems,
including coastal lagoons, and the re-shaping of seaward boundaries’. It is,
however, clear that evolution progresses in a variety of ways and with different
results in different lagoons.
The size of coastal lagoons varies greatly. In Zululand, for example, much
larger coastal lagoons are present on the sandy coastal plain than on the steep
rocky coast of Natal. This is a direct result of shallower gradients and soft
surrounding sediment in Zululand.
Where the continental shelf is narrow, erosion during lowered sea levels
incises fluvial valleys to great depths. This produces deep depressions in which
to form lagoons. In contrast, areas with wide continental shelves offer only
shallow depressions in which to form lagoons. In Natal and Zululand,
Holocene lagoonal sequences are up to 40 m thick, while in Maine and Brazil
lagoonal deposits are only 3 to 4 m thick.
The maximum age of Holocene lagoonal sediments in Virginia is 5000
years (Finkelstein & Ferland, 1987) and in Brazil, 7800 years (Ireland, 1987).
The Canet-St Nazaire Lagoon in the Mediterranean formed only during the
past 6000 years (Martin et al., 1981). Young ages and shallow sedimentary
fills are apparently more typical of areas where lagoons occupy coastal
depressions rather than incised river valleys. Holocene lagoons formed in
incised river valleys on steep-gradient high-latitude coasts may have
originated as much as 10000 years BP (Thom, 1983) and have required much
greater sediment volumes to be infilled. This impacts on the age of sediments
and the length of time for which a lagoon may persist.
The role of sediment supply is illustrated by comparison of Lake St Lucia
and Kosi Bay (Cooper, 1990a), which clearly indicates the influence of
sediment availability on lagoon coast evolution. The bedrock valley of St
Lucia Lagoon is filled by 40 m of fluvially derived Holocene sediments which
reduced its mean depth to 1.5 m, while in Kosi Lagoon the incised valley is
largely preserved by the modern bathymetry as no major rivers discharge into
the lagoon.
 Lagoons and microtidal coasts 239
The transgressive back-barrier sequence found in most Maine lagoons
contrasts with the apparently more typical regressive back-barrier fills
described from Brazil and New South Wales. The differences may be
attributed to different balances between sediment supply and rate of sea-level
rise. In Maine, sediment supply was low and sea-level rise rapid, while in
Brazil and New South Wales, sediment supply was comparatively higher and
prevented complete inundation of back-barrier areas. In Natal, ubiquitous
lagoon fills such as that of the the Mtwalume (Fig.6.8b), suggest an
oversupply of sediment, which maintained shallow fluvial conditions through
most of the lagoon’s evolution.
The coastal lagoons of New South Wales infilled from an initially deep
basin through successive shallowing with time to a swamp-surrounded river
channel (Roy, 1984). Lagoons can be found at various stages of development
in an evolutionary sequence. This compares directly with the progressive
infilling of initially deep lagoons in other areas such as Brazil (Ireland, 1987)
and southeast France (Martin et al., 1981) along the lines of the Lucke model.
In subtropical Natal extended periods of chemical weathering since the
Miocene (Partridge & Maud, 1987) have produced large quantities of fluvial
sediment. Consequently, in lagoons such as the Mtwalume, sedimentation
matched the rate of infilling and so deepwater lagoonal conditions never
developed. Furthermore, later Holocene fluctuations of sea level in bedrock-
confined Natal lagoons caused incision of former subtidal deposits and their
emergence as floodplains during minor regressions. This has had the effect of
increasing evolutionary rates: consequently, such phases in infilling could not
be recognised in the Natal lagoons, as all are geomorphologically mature.
The role of fluvial sediment supply on lagoonal evolution processes is
particularly well illustrated by comparison of the Kosi and St Lucia lagoons on
the Zululand coastal plain. In Lake St Lucia, rapid fluvial sedimentation
accompanied the Holocene transgression and even promoted desiccation early
in its Holocene evolution. In contrast, in Kosi Lagoon the Late Pleistocene
channels are preserved as a result of low fluvial influx. The idea that
segmentation reduces sedimentation rates by trapping inflowing sediment (as
in Lake Amanzimnyama) can be rejected from Kosi as seismic profiling
reveals no Holocene sediment accumulation, but this may be an effective
process elsewhere in reducing lagoonal infilling rates.
Sediment supply is in part influenced by climate. Martin et al. (1981), for
example, noted a sixfold increase in sedimentation rates in Mediterranean
lagoons during the Subboreal Period through increased rates of soil erosion
and increased ‘storminess’ in the hinterland. Sediment supply rates in Natal
and Zululand are high in relation to New South Wales, for example, but factors
240 J.A.G. Cooper
such as catchment size also influence sediment supply to particular lagoons. It
is probable that a relationship exists between catchment size, runoff volume
and lagoon volume; if runoff to a particular lagoon were to increase a breach
may form and even persist, whereas if discharge were reduced the breach may
close.
Closure of tidal inlets as a result of decreased tidal prism, documented by
Finkelstein & Ferland (1987) from Virginia, contrasts with the situation in
Natal and New South Wales where normally closed outlets may open more
frequently as a result of increased competence arising from fluvial
channelisation (Roy, 1984; Cooper 1 9 9 1 ~ ) This. is a direct result of the
importance of fluvial discharge in Natal and New South Wales in comparison
with tidal currents, which maintain inlets in Virginia and elsewhere.
The response of a barrier to sea-level rise has a direct bearing on lagoonal
evolution. If a transgressive lagoonal barrier stabilises against a break in slope
or topographic high and accretes upward rather than landward then the volume
of the lagoon may increase if sea-level rise exceeds fluvial sediment supply. If
the barrier continues to move landward as sea level rises then the actual lagoon
volume may remain constant. In certain cases the barrier progrades seaward
and this may reduce contact with the sea, as in the case of the Siyai lagoon.
Barrier grain size apparently limits the influence of washover in the back-
barrier environment (Duffy et al., 1989). The lagoons of Maine contain limited
volumes of overwash-derived sediment, in contrast to lagoons with sandy
barriers where washover fans may extend far into the lagoon (Andrews 1970;
Cooper & Mason, 1986) and become an important component in the back-
barrier stratigraphy (Hobday & Orme 1974) and a major sediment source
(Bartberger, 1976). Carter & Orford (1984) noted that swash percolates rapidly
into porous coarse-grained barriers and deposits entrained material. This limits
washover by building up the barrier. Eventually, the accumulated material
slumps into the barrier as a series of small lobes (Orford & Carter 1982).
Most of the examples in the literature (for example New South Wales,
Brazil, Virginia) depict the evolution of a lagoon from an initial basin, which
is progressively isolated from the sea and is ultimately infilled by a swamp or
floodplain. The source of sediment in some cases (New South Wales) is fluvial
and differs from the Lucke Model of marine-derived sediment accumulation in
this respect. The examples cited from Maine, which show increasing marine
influence, differ from this model because of low sedimentation rates in relation
to sea-level rise while some of those in Natal differ in that sedimentation
matched or exceeded the rate of drowning during sea-level rise.
 Lagoons and microtidal coasts 24 I
Processes of lagoon evolution
Numerous distinct evolutionary processes contribute to the evolution of
lagoons. In some cases a single process dominates, while in others a variety of
processes play roles of differing importance. The relative importance of each
necessarily influences the evolution of a particular lagoon. The major
processes are discussed below under back-barrier and barrier headings. The
division is for ease of discussion only as one may impact on the other.

Back-barrier changes
Segmentation
This is the process whereby a single lagoon is converted into a series of
isolated basins. It arises through the reorientation of a lagoon shoreline by
erosion and deposition to assume equilibrium with the ambient wave field.
Zenkovich (1959) proposed a model (Fig. 6.10) for complete segmentation of

cuspate spits

BARRIER
SEA

original paired
shoreline sp1ts

Figure 6.10. A model of lagoon evolution through segmentation (based on Zenkovich,

1959). This process operates best on uncohesive sandy shores in micro- or non-tidal
environments. An open lagoon ( a ) becomes progressively segmented ( b )and ( c ) .It is
well illustrated by the Kosi Lagoon (Fig. 6.5).
 Next Page
242 J.A.G. Cooper
lagoons in which cuspate forelands develop into paired spits, which ultimately
divide the lagoon into a series of roughly circular basins. Rosen (1975)
subsequently noted that while cuspate spits form only in microtidal areas,
complete segmentation only occurs in non-tidal areas because elsewhere tidal
currents limit spit growth. Zenkovich’s (1959) suggestion that segmentation
occurred when the long axis of a lagoon extended parallel to the dominant
winds was disputed by Rosen (1975), who argued that fetch limitation in an
elongate lagoon was itself sufficient to force wave approaches to be
perpendicular to the shoreline regardless of wind approach direction.
Segmentation has been reported from many parts of the world from the Arctic
to the tropics and occurs in lagoons ranging from a few tens of metres in
diameter to the 350 km long Lagoa dos Patos in southern Brazil. One of the
main impacts of segmentation is to create distinct sedimentary cells, which in
the case of Kosi Lagoon exhibit distinctive sedimentary, chemical and
biological characteristics. Segmentation may potentially preserve greater
depths in certain segments by isolating them from fluvial sediment inputs, or
create different evolutionary responses in adjacent segments.

Vertical accretion
Vertical accretion in lagoons is mediated by the rate and nature of sediment
supply. Lagoon sediments may originate from the sea or land or as biogenic
particles generated within the lagoon. Quiet water deposition of silt and clay
occurs in lagoons in a manner analogous to pro-delta environments, where
material is carried in by a river. Deposition from suspension of fluvial
sediment is often enhanced by flocculation in the more saline lagoonal waters
(Hobday, 1976). In certain cases increases in sedimentation rates have been
attributed to human impacts (Finklestein & Ferland, 1987).
Organic or skeletal material is produced within lagoons below water level
and on marginal intertidal flats by lagoonal organisms and vegetation. In some
cases (Warnbro Sound and Port Royal Bay) this sediment is skeletal carbonate,
while in others it is vegetal, originating from fringing vegetation such as
mangroves in the tropics or salt-marsh vegetation in more temperate regions.

Fluvial delta progradation

Growth of fluvial deltas into lagoons is one mechanism by which the volume
of a lagoon may be reduced. The size of a delta and its morphology depend
largely on the volume of sediment carried by the inflowing river and energy
levels within the lagoon. In theory, lagoonal deltas could exhibit most of the
morphologies exhibited by deltas deposited directly in the sea (Galloway,
1975).
 7
Coral atolls
R.F. McLEAN AND C.D. WOODROFFE

Introduction
Coral atolls consist of an annular reef rim surrounding a central lagoon. On the
atoll rim there may be reef islands; either sandy cays or shingle motu (Stoddart
& Steers, 1977). They appear particularly fragile constructions, exposed to a
range of oceanographic, atmospheric and anthropogenic processes, and yet the
prototypes of present day sea-level atolls have endured in a variety of forms for
millions of years. Of the 425 atolls in the world (Stoddart, 1965), most are in
the Indian and Pacific Oceans, though isolated atolls do occur in the Caribbean
(e.g. Stoddart, 1962a)
The evolution of atolls is of particular relevance in the context of coastal
evolution studies, because one of the most central theories of atoll evolution
dates back to Charles Darwin and the voyage of HMS Beagle. This voyage
gave rise to the concept of natural selection and evolution of species, with the
publication of On the Origin of Species in 1859, but Darwin’s view on
evolution of atolls, was published much earlier with his observations on the
geology of reefs (Darwin, 1842).
Darwin had formulated his theory of reef development shortly after leaving
South America, and before seeing the reefs of the Pacific. He refined his ideas
as a consequence of visiting reefs in the Society Islands, and had completed a
draft of his theory before reaching New Zealand (see Stoddart, 1962b). The
only atoll that Darwin landed on was the Cocos (Keeling) Islands in the Indian
Ocean, and at that stage he was keen to verify his intuition. Darwin’s theory
was later tested first by deep-drilling that was undertaken on the atoll of
Funafuti in Tuvalu (previously the Ellice Islands), and later as a consequence
of atomic testing on other Pacific atolls.
The Cocos (Keeling) Islands in the Indian Ocean and Funafuti in the Pacific
Ocean were the best-known atolls at the turn of the century. In this chapter the
history of theories of reef development is examined, with particular emphasis

267
268 R.F. McLean & C.D. Woodroffe
on the Cocos (Keeling) Islands and Funafuti. The long-term development of
atoll structure is considered, based upon the considerations which Darwin
outlined, and atoll evolution is accommodated into the concepts of plate
tectonics. Late Quaternary development of atolls, during the oscillations of sea
level, is examined, and the Holocene morphodynamics of reef growth,
lagoonal infill and reef island formation are considered.
At the time of Darwin’s voyage, the most widely held view of atolls was
that they were submerged volcanic crater rims veneered with a thin cover of
coral (Lyell, 1832). Darwin recognised the improbability that these volcanic
rims would everywhere reach into the shallow depths required for reef growth,
and proposed that there:
is but one alternative; namely the prolonged subsidence of the foundations on which the
atolls were primarily based, together with the upward growth of the reef-constructing
corals. On this view every difficulty vanishes; fringing reefs are thus converted into
barrier-reefs; and barrier-reefs, when encircling islands, are thus converted into atolls,
the instant the last pinnacle of land sinks beneath the surface of the ocean.
(Darwin, 1842: p . 109.)

This view, illustrated by Darwin’s own sketch (see Fig. 7.1), envisaged
fringing reefs, barrier reefs, and atolls as stages in an evolutionary sequence,
and that the process driving the sequence was gradual subsidence of the
volcanic basement on which the reef initially established, in combination with
vertical reef growth.
On Cocos, Darwin’s attention was drawn to erosion of the shoreline, which
had resulted, as it still does, in toppling of coconut palms. He considered this
‘tolerably conclusive evidence’ of subsidence. We now know, as John Clunies
Ross knew at the time, that localised shoreline erosion on Cocos, as on other
atolls, is easily explicable by other means. Indeed, concerned at Darwin’s
suggestion that the islands of which he was in possession were about to subside
beneath the sea, Ross (1 855: p. 9) claimed that a
moderately attentive investigation of the Cocos islets affords ample reasons for
believing that they have stood up to the present time above the level of the ocean during
hundreds if not thousands of years.

Darwin’s theory has proved robust, and is widely supported by drilling

results from other atolls, and by evidence that is described below from both
Cocos and Funafuti. Surface geomorphological evidence from Cocos,
however, indicates that there has been a relative fall of sea level (emergence)
over the last few thousand years, quite the converse of what Darwin had
 Coral atolls 269

A C B B C A

A B .B A
Level of Sea

C .B‘ C
Level of Sea

Figure 7.1. Darwin’s (1 842) original woodcut illustrations redrawn. In the upper
diagram the original volcanic island is shown with afringing-reefA-B. A section of the
reef and island after subsidence is given by the dotted lines which show an encircling
barrier-reefwith A’ the outer edge of the reef, B’ the shore of the encircled island, and
C the lagoon-channel between the reef and island’s shore. In the lower diagram the
newly formed barrier reef is represented by unbroken lines. As the island continues
subsiding the coral reef will continue growing up on its own foundation and after the
sea covers the highest volcanic pinnacle a ‘perfect afoll’ is formed with A“ representing
the outer edge and C’ the lagoon of the newly formed atoll.

concluded. To understand this it is necessary to discriminate between two

contrasting aspects of the evolution of coral atolls: the structure of atoll
foundations, including the relationship between reefal limestone and volcanic
basement, on the one hand; and surface morphology of atolls, including
modern reefs, surficial deposits, lagoonal infill and the origin of reef islands,
on the other (Stoddart, 1973). It is important to realise that ‘the movement of
subsidence, if coral growth was to keep pace with it, would have to be too slow
to be observable’ (Stoddart, 1983: p. 525). The distinction between structure
and surface morphology is of significance because the rates of operation of the
formative processes and the timescales involved are so different.
270 R.F. McLean & C.D. Woodroffe
The Cocos (Keeling) Islands and Funafuti Atoll
The Cocos (Keeling) Islands comprise the South Keeling Islands, a horseshoe
shaped atoll (12" 12' S , 96" 54' E), and North Keeling, an isolated atollon
(almost entirely closed reef-top island) (1 1" 5 0 ' S , 96'49' E), 27 km to the
north. These rise as a single feature from the ocean floor, which is about
5000 m deep, and 60-90 Ma old (Jongsma, 1976). They are dominated by the
southeast tradewinds which blow strongly during most of the year, and
maintain a large swell particularly on the southern and eastern margins of the
atoll.
The South Keeling Islands, henceforth referred to as Cocos, consist of a
series of reef islands on a horseshoe-shaped atoll rim, with two deep passages
southwest and southeast of Horsburgh Island, the northernmost island
(Fig. 7.2). Shallow interisland passages separate the other reef islands, with a
broad reef flat between the southern elongate islands of West Island and South
Island. The reef front is relatively barren of living scleractinian corals; the reef
crest is algal veneered, with surge channels at intervals of 50-250 m. The reef
flat is covered by 1-2 m of water at high tide and much of it dries at low tide.
The atoll is microtidal with a spring tidal range of 1.2 m. The northern part of

COCOS
(KEELING)
ISLANDS FUNAFUTI

5 kilometres
8035'5

PACIFIC

5 kilometres

Figure 7.2. The Cocos (Keeling) Islands and Funafuti Atoll, Tuvalu, showing location
and reef and island outlines.
 Coral atolls 27 1
the lagoon averages around 15 m deep and is covered with sand, or in places
dead coral. The southern part of the lagoon is shallow, but contains a network
of 'blue holes', individual deep holes with coral rims, and a muddy fill, up to
15 m deep in their centres. Extensive sand flats and sand aprons occur around
the margin of the lagoon, especially towards the southern end of the lagoon
(Smithers et al., 1994).
Funafuti Atoll (8" 30'S, 179" 1O'E) is substantially larger than Cocos
(Fig. 7.2). Its long axis, aligned approximately north-south, is about 25 km
long, and the atoll is about 18 km west to east at its widest. The lagoon
comprises a deep central basin of average depth 40-50m, which although
generally of low relief is in places punctuated with deeper depressions and
pinnacles some of which reach to the surface. The basin slopes gently upwards
initially and then much more steeply (slopes of 10-20") towards the atoll rim.
Depths of 30 m are reached within 500 m of the shore in the north and east and
within 2 km of the shore in the west. Much of the narrow southern lagoon is
less than 10m in depth and in the extreme south the lagoon shoals at low tide.
Funafuti's rim is cut by two narrow and deep (40-50m) passages in the
west and several shallower ones in the north and southeast. Compared with
Cocos, the atoll rim is narrow and solid reef flat is exposed at low tide on both
the ocean and lagoon margins of islands in the east and south. In the west the
wider more discontinuous reef flat is occupied by a number of small more-
compact islands. Maximum tidal range is l .7 m and like Cocos, Funafuti is
subject to tradewind-generated seas from the east and southeast.

Atoll structure
Although not the first attempts to drill reefs (Stoddart, 1992), deep drilling was
undertaken on Funafuti in 1896-1898 in order to examine Darwin's
interpretation of atolls that they developed as a result of a gradually subsiding
volcanic basement around which vertical reef growth had been established.
After several attempts, 333 m of coral rock was drilled, with the upper 194 m
in coral limestone, and the rest in dolomite, which Judd (1904) and others
regarded as being of shallow-water origin throughout. While volcanic
basement was not encountered, the Funafuti borehole provided some
substantiation of Darwin's ideas. Volcanic basement was further indicated by
seismic refraction experiments carried out in the Funafuti lagoon during the
H.M.S. Challenger expedition (Gaskell & Swallow, 1953; Locke, 1991).
Gaskell & Swallow concluded that at Funafuti there is about 550 m of coral
limestone overlying the volcanics.
Davis (1928), in an extensive review of the development of coral reefs,
272 R.F. McLean & C.D. Woodroffe
considered that the Darwinian subsidence theory was the most credible of the
competing theories. He discounted the solutional theory of atoll development
put forward by Murray (1 889), and the sedimentation view, propounded by
Wood-Jones (1912), who spent over a year on the Cocos Islands as the medical
doctor.
Although the Funafuti boring did not confirm that basalt underlaid the atoll,
it did show greater thicknesses of shallow-water carbonates than could occur in
the water depths that were drilled, implying subsidence. Similar results were
obtained from the drilling at Kita Daito Zima, southeast of Kyushu, Japan
(Hanzawa, 1938), which penetrated 43 1 m without encountering basalt.
Post-war deep-drilling of atolls, such as that on Bikini and Enewetak in the
Marshall Islands, which revealed more than 1000 m of shallow-water
carbonates overlying a basalt core (Ladd, Tracey & Lill, 1948; Ladd et al.,
1953; Emery, Tracey & Ladd, 19541, has generally substantiated the
subsidence theory of coral atoll development proposed by Darwin (1 842).
On Midway Atoll in the Hawaiian Islands, volcanic basement was
encountered at less than 200 m in the lagoon and around 400 m on the atoll rim
(Ladd, Tracey & Gross, 1970), and in the Tuamtous on Mururoa a series of
drillholes encountered basalt at just below 400 m (Lalou, Labeyrie & Delibrias,
1966; Labeyrie, Lalou & Delibrias, 1969) and on Fangataufa Atoll boreholes
reached the volcanic basement at 360 m under the atoll rim and 230 m under
the lagoon (Guillou et al., 1993). The combined results demonstrate ‘with a
certainty rarely obtained in geomorphology, that these ... atolls were built by
corals and associated organisms during a long-continued subsidence’
(Guilcher, 1988).
That the Cocos Islands are underlain by a volcanic basement can be inferred
from magnetic surveys showing an anomaly (Chamberlain, 1960; Finlayson,
1970), and a basalt and tuff pebble dredged from the western end of the Cocos
Rise (Bezrukov, 1973).
The Darwinian sequence of coral atoll development, in which the driving
forces are gradual subsidence of the volcanic basement and vertical reef
growth (Braithwaite, 1982), has been incorporated into plate tectonic theory,
and much of the subsidence can be explained in association with the aging and
contraction of the ocean floor as the oceanic plate moves (Parsons & Sclater,
1977; Scott & Rotondo, 1983). Atolls in the Pacific are often found in linear
chains, some of which (i.e. the Hawaiian and Society Islands) demonstrate
stages in the Darwinian sequence - fringing reefs, barrier reefs, and coral atolls
- along their length (see Chapter 12). Reef growth, however, has not been
continuous, but has been interrupted by a series of solutional unconformities
(Schlanger, 1963), indicating periods of subaerial erosion resulting from
 Coral atolls 273
fluctuations of sea level during the Tertiary and Quaternary. Atolls have been
exposed above the sea and subjected to weathering and erosion during low
stands of sea level, and submerged with reefs re-establishing over their
previously exposed surface during high stands of sea level (Lincoln &
Schlanger, 1991).

Late Quaternary sea-level fluctuations

Superimposed upon the gradual subsidence of the atoll basement there have
been pronounced fluctuations of sea level of around 120-1 50 m amplitude over
the Quaternary. It is these alternate periods of subaerial exposure during
glacial lowstands and drowning and reef re-establishment over the upper
surface of the platform during interglacial highstands which have given rise to
the solutional unconformities in atoll stratigraphy.
The importance of sea-level fluctuations was realised by several reef
geologists, particularly Reginald Daly (Daly, 1915, 1925, 1934). Daly’s
glacial control theory of coral reef development maintained not only that there
was contraction of reefs from marginal seas during glacial periods, but that reef
platforms were planed off at a lower sea level, explaining the consistency of
lagoon depths, with postglacial reef growth constructing the atoll rim. These
views were perpetuated in the interpretations of atoll development propounded
by Wiens (1959, 1962). In fact the rate of planation of reefs is generally too
slow for such dramatic truncation to occur (Stoddart, 1969). Furthermore,
outcrops of Pleistocene reef limestone which have clearly not been planated
indicate that wave abrasion is not as effective as envisaged by Daly. More
effective than wave abrasion are karst erosion processes, particularly at lower
sea levels, and their importance has been stressed by Purdy in what has become
known as the antecedent karst hypothesis (Purdy, 1974).
In the COCOSIslands shallow boreholes (up to 30m deep), undertaken as
part of water resource investigations and geomorphological studies of the
development of the atoll, have encountered an unconformity at around
10-12m below sea level. The lower limestone is well lithified, porous, and
contains micritic cements indicating subaerial diagenesis. The pattern of
boreholes and the depth of this unconformity are shown in F i g . 7 . 3 ~ ;the
shallowest at which it is encountered is at 6-7 m at the southern and far western
sides of the atoll. Uranium-series disequilibrium dating of coral from below
the unconformity has yielded dates of around 120000 years BP (Woodroffe
et al., 1991a), indicating that it was deposited as part of a Last Interglacial reef.
The level of the sea at that time was roughly comparable to that at present
or even slightly above present sea level. Solution of this thickness of
 WEST IS.

West Island
Northern HIZ(9.4)
Lens

Wll7
(10.5)
w121
(10.4) sourn
WH8 KEELING
(8.5)
w1: ISLANDS
(10.1)
WllO
(10.5) West Is.
Wl22
(6.8) West Island South
Airfield Lent
WIl(10.3)
Quaranfine WlZ(8 6 )
Stallon ~eleoroIo91CSIStallon South Island
Wlll(10.3) WS(8.6) L'?"S*S
wm(s.6)
W7(10.8) I
Airstrip CK15 f
--
CK13C6.7) ('I6' n
Freshwater lens sll(ll.9)
P
Demhr bebv water la& 10 the u r o n l a m l y
Stratigraphic borehole between HObcene an@PC~loccncSaUment.i
Salinity monitoring borehole "Imetres M W " h t .acLL,*

Figure 7.3. ( a )The location of boreholes on islands on the rim of Cocos, showing the depth to the Thurber unconformity (updated from Woodroffe
ef al., 1991a). (b) schematic cross-section of lagoon, showing Last Interglacial reef beneath West and Home Islands (after Woodroffe ef al.,
1991a). HAT = Highest Astronomical Tide; LAT = Lowest Astronomical Tide. ( c ) seismic reflection profile of western margin of Cocos lagoon,
representing about 250 m of trace running southeast from W14. S = Lagoon floor; A, B and C are reflectors; A is taken to represent the Last
Interglacial surface (after Searle, 1994, Fig. 2b).
 Coral atolls 275

limestone over this time appears unlikely, and the depth of this Thurber
discontinuity (cf. Thurber et al., 1965) appears to support subsidence since its
deposition.
Much of the lagoon at the northern end of Cocos is 12-15m deep
(Fig. 7.3b). However, seismic reflection profiling across the lagoon (Fig. 7 . 3 ~ )
indicates a distinct reflector which occurs at a similar depth as the
unconformity in the vicinity of the islands, but which slopes relatively steeply
into the lagoon, and which occurs at depths of 20-22 m throughout much of the
lagoon (Searle, 1994).
On Funafuti, unlike Cocos, the Last Interglacial reef has not been positively
identified through radiometric dating. Nevertheless, extensive seismic
reflection profiling in the Funafuti lagoon has indicated the presence of several
unconformities, the uppermost being interpreted as a ‘coral surface’, the base
of which is coincident with a widespread unconformity, U1 (Gibb Australia,
1985). Overlain by 20-25 m of unconsolidated sediment in the deep lagoon,
the morphology of the ‘coral surface’ is highly variable with depressions and
pinnacles; some of the latter extend to the lagoon floor. We suggest that this
karst surface represents the Last Interglaical solutional unconformity.
From the deep lagoon basin the surface rises towards the atoll rim along
with a decline in thickness of the overlying unconsolidated sediments. In the
east we have attempted to correlate the seismic reflection results with previous
interpretations of data from the 1896-98 drilling program on Fongafale Island
and the adjacent lagoon (Fig. 7.4). Ladd (1948) identified a fossil land snail
Ptychodon sp. A in partially leached coralliferous limestone from a depth of
50.6-5 1.8 m in the main bore, which he suggested was Pleistocene to Recent in
age and represented a time during a glacial epoch when sea level was lower
and when ‘the top of the atoll stood above the sea’. Ginsburg et al. (1963)
identified a ‘probable surface of subaerial exposure’ at about 42 m in the main
borehole and 55 m in the lagoon borehole. The latter coincides with the depth
of the main unconformity (Ul) distinguished in the seismic record. The
relationships between this unconformity and the ‘coral surface’ (Gibb
Australia, 1985), the ‘coral reefs and blocks’ facies between depths of 10-26 m
in Borehole No 1 (Sollas, 1904) and the ‘first calcite cemented skeletal debris’
at 40m in the main borehole (Cullis, 1904; Ginsburg et al., 1963) remain
unclear and need to be resolved through further analysis of the cores and
absolute dating techniques.
During the 125000 years or so that it has taken for the sea to return to its
present level, Cocos and Funafuti appear to have subsided, perhaps partly
through cooling and contraction of the ocean floor, but also as a result of other
factors, so that the former (Last Interglacial) surface is everywhere below
276 R.F. McLean & C.D. Woodroffe

PACIFIC

Funafuti

Seismic line No 56
Lagoon
Borehole "C Halirneda Main
Lagoon Samplks Borehole

Sollas No1
Borehole

OCEAN

FONGAFALE ISLAND

kilometre

Sand with
b some coral blocks (GI

20% Carllr
Sand with 80% Unconiolidated
L m m " flow IOrnB Coral IkeletaI debris (G)
blocks ( 5 )
Coral surface (GA)
Unconrolidated
HaJimeda debris 42m First Calcite cemented
skeletal debris (G)
Probable Surface of 51m FOISII land mail (L)
rubaerial ~ X P O L U I(G)
~
First coral casts &
moulds (GI
Calcite cemented Coral mcteasing 67m Last aragomte ( c )
skeletal debris (GI with depth (G)

kilometre

Figure 7.4. Shallow subsurface structure of Fongafale Island and adjacent lagoon,
Funafuti Atoll. (a) Location of the Main, Sollas No I and lagoon boreholes drilled
during the 1896-98 Royal Society Expedition, and the seismic reflection line number
56, stations 130-154 and the radiocarbon dated sediment (Halirneda) samples taken
from 3 m below the surface at depths of 18 m (location 12) and 22 rn (location 26) by
Gibb Australia (1985). (b)Sedimentary facies and subsurface unconformities based on
several sources: (C) Cullis, 1904; (G) Ginsburg et al, 1963; (L) Ladd, 1948; (S) Sollas,
1904; (GA) Gibb Australia, 1985. See text for details.
 Coral atolls 277
present sea level. To what extent this is true of other atolls is discussed further
below. The sea rose rapidly during the postglacial marine transgression
(Fairbanks, 1989; Chappell & Polach, 1991; Eisenhauser et al., 1993), in each
case drowning the former atoll surface. The upper 10-15 m of the present rim
on these atolls has formed as a result of Holocene reef growth over the last few
thousand years. These late Quaternary sea-level fluctuations ‘were not
envisaged by Darwin though they fit perfectly into later versions of the
subsidence theory, and are easily explainable by almost any version of the
glacial control theory’ (Ladd, 1948: p. 197).

Holocene reef development

Reef response to sea-level change

Reefs are not necessarily able to track sea level exactly. At least three patterns
of reef response to sea level can be envisaged. These have been termed keep-
up, catch-up and give-up (Neumann & Macintyre, 1985). Keep-up reefs are
reefs which do grow close to sea level and which are able to accrete vertically
at rates fast enough to remain at sea level. They are characteristic of gradually
or rapidly uplifting coasts (i.e. Barbados, Bard et al., 1990 and Huon
Peninsula, Chappell & Polach, 1991), but can also be seen on some stable
coasts, even close to the latitudinal limit to reef growth (Collins et al., 1993).
Catch-up reefs are reefs where vertical growth initally lags behind sea level
and then catches up when sea level stabilises or the rate of rise declines. Most
reefs on Australia’s Great Barrier Reef have grown by catch-up mode (see
Chapter 8). Give-up reefs are reefs that founder as accretionary growth falls
behind sea level until growth declines or ceases.
The pattern of sea-level change over the late Quaternary has been driven
primarily by changes in the volume of water in the oceans. However, the
earth’s crust and upper mantle have adjusted to variations in the load of both
polar ice and ocean water, and these glacioisostatic and hydroisostatic
adjustments, together with local tectonic movements, mean that land-sea
changes have varied around the world. These variations have become
especially apparent over the mid and late Holocene, because ocean water
volume has changed little during that period, and therefore subtle isostatic
adjustments are not masked by rapid eustatic change (Pirazzoli, 1991).
Geophysical modelling of the response of a viscoelastic earth to changes in
ice and water loads, calibrated with reconstructed ice-melt histories and key
sea-level data, allows the extrapolation of relative-sea-level curves to regions
278 R. F. McLean & C.D.Woodrofle
from which data are presently scant (Clark, Farrell & Peltier, 1978; Peltier,
1988; Nakada & Lambeck, 1988; Tushingham & Peltier, 1991). This
modelling supports the recognition of a ‘Caribbean sea-level curve’, which has
been rising, but at a decelerating rate over the last 6000 years, and a ‘Indo-
Pacific sea-level curve’, which had reached a level close to present 6000 years
ago, rose above that level, and which has been falling for the last few thousand
years (Fig.7.5, see Adey, 1978; Davies & Montaggioni, 1985), though some
variability also occurs at a regional scale. Coral atolls, together with other

14C ka BP 14Cka BP

CARIBBEAN INDO-PACIFIC
0

12

16 I
kW
20 n

24

28

Acmpora Branching Massive Dead or

palmeta acroporid corals truncated
corals reef

Sea level Keep-up Catch-up Give-up

Figure 7.5. Caribbean and Indo-Pacific sea-level curves (based upon the curves shown
in Davies & Montaggioni, 1985: Caribbean curve after Lighty et al., 1982; Indo-Pacific
curve is a combination of curves for Australia and the Polynesian curve of Pirazzoli
et al., 1988b). Keep-up, catch-up and give-up reef growth strategies in relation to sea
level are shown. Give-up reefs are represented by drowned reefs, or in the case of Indo
Pacific reefs, growth may be truncated by a fall of sea level.
 Coral atolls 279
islands less than 25 km in diameter, are likely to provide one of the least-
complicated records of relative sea-level change, because they should be
relatively immune to local hydroisostatic effects (Nakada, 1986).
The keep-up, catch-up and give-up reef growth strategies are most often
expressed in terms of Caribbean reefs, where they often form distinct facies.
However, similar growth patterns can be identified in the Indo-Pacific reef
province. In the Indo-Pacific, keep-up reefs are generally characterised by
branching framework corals, whereas there are greater proportions of massive
coral heads in catch-up reef, though without any as yet recognised patterning
of genera or species. Give-up reefs can be found as dead reefs in 20-30 m of
water depth, upon which there may be sparse modern coral growth but which
appear to have been drowned and to have accreted slowly if at all. The
drowned banks and atolls, such as the Chagos Bank (Stoddart, 1971a), and
Saya de Malha (Guilcher, 1988), are examples of probable give-up reefs,
though less work has been done on these than on equivalent structures in the
Caribbean. A second type of give-up reef found in the Indo-Pacific region is a
truncated reef which has given up because it has been exposed by a relative fall
in sea level (emergence). Areas of such give-up reef, often with scattered flat-
topped intertidal corals, termed microatolls, which emerge at lowest tide,
characterise many reef flats in the region and are described in more detail
below.

Holocene reef development in the Cocos (Keeling) Islands

Fig. 7.6 shows details of stratigraphy and radiocarbon chronology of three
transects from a series of drillhole transects across reef flat, and reef islands on
Cocos (from Woodroffe et al., 1994). Recovery in drillholes is generally poor,
but is dominated by shingle clasts of branching acroporid corals apparently in
a sandy matrix. Except immediately beneath the reef flat, or cemented
conglomerate platform, this coral shingle is rarely lithified. The oldest dates
obtained are around 7000 years BP at 14m below sea level in boreholes on
Home Island, recording reef establishment over the Pleistocene surface after
flooding by the rapidly rising postglacial sea level. Similar dates (around 6800
years BP) have been obtained at shallower depths (around 9 m) in a drillhole on
an island in the southern passage. At the northern end of the atoll, however,
dates at similar depths in drillholes on Horsburgh Island yielded ages of
4600-5500 years BP, suggesting later establishment of reefs at this end of the
atoll (profile X, Fig. 7.6).
Vertical reef accretion appears to have been rapid. Ages do not necessarily
record in situ reef growth, as some of the reef fabric could have accumulated
 PROFILE X: HORSBURGH

MSL

Reef Flat
Beachrock
Conglomerate Platform
Sand
Shingle
Mud
Pleistocene Limestone
Solid Coral
Shingle Sticks
Algal Cemented Shingle
UNCONFORMITY- -? Radiocarbon Date (coral)
Radiocarbon Date (Tridacna)

CK1
CK19 CKlOA CKT.. cK3
C K 1 8 CK1oB C K 2 0
MSL

PROFILE II:
HOME IS.
UWCONFORMITY

PROFILE VIII:
WEST IS.

Figure 7.6. Transects of boreholes on COCOS,

showing stratigraphy and radiocarbon chronology. Location of transects can be seen in Fig. 7.3~7.
 Coral atolls 28 1
as detrital material. Nevertheless, dates in CK3 and CK12 suggest average
rates of accretion of 4-40 mm a- '. In Fig. 7.7 radiocarbon ages from a series of
drillholes and from surface exposures of conglomerate platform are shown on
an age-depth plot (updated from Woodroffe, McLean & Wallensky, 1990b).
Several dates indicate that reefs were 2-3m below present sea level around
6000 years ago, though the sea appears to have reached present level by this
time elsewhere in the Indo-Pacific region. This disparity might either result
from a different hydro-isostatic response for mid-ocean sites to those in
continental settings (Chappell, 1974), or more probably might reflect reef
growth lagging behind the rapidly rising sea (Neumann & Macintyre, 1985)
and only later catching up when sea level had stabilised.
Data on shallow subsurface stratigraphy and radiometric chronology are
now available from several Pacific atolls. Fig. 7.7 also summarises age-depth

RADiOCARBON YEARS BP RADIOCARBON YEARS BP

8000 6000 4000 2000 aooo 6000 4000 2000

INDIAN OCEAN PACIFIC OCEAN

r
m
rn
<
2.

.
O
m

-3
I
3
v
-
i
3
1

Tarawa, subsurface
Cocos, coral Kiribati and
Cocos, mlcroatoli Tuvalu, surface
Tuvalu, surface

Figure 7.7. Age-depth plot of radiocarbon dates from Cocos, and from Kiribati and
Tuvalu. Dates are on coral (or in some instances on Triducnu). Only those labelled as
microatolls are known to be in growth position. Three phases of Holocene
development, numbered 1-3, can be identified; see text for details.
282 R.F. McLean & C.D. Woodroffe
relationships of radiocarbon dates on coral and Tridacna samples from
Tarawa, Kiribati, (Marshall & Jacobson, 1 9 8 3 , together with surface samples
from Tuvalu and Kiribati (Schofield, 1977a; McLean & Hosking, 1991) which
are discussed later. In Tarawa, drilling undertaken on reef islands on the atoll
rim several hundred metres from the reef crest indicates that Holocene reefs
established over the former atoll surface between 7000 and 8000 years ago,
and there was then a period of rapid vertical growth of reefs, catching up with
the rapidly rising sea.
Surface features of Indian and Pacific Ocean atolls do not appear to record
the attainment of modern sea level at around 6000 years BP, as is generally
observed for sea-level curves from the region (Pirazzoli, 199 1) and at several
sites close to the atolls under consideration (i.e. Fiji and Tonga, Miyata et al.,
1990; Nunn, 199 1 , in the Pacific; Sri Lanka, Katupotha, 1988, in the Indian
Ocean). Instead, throughout Pacific atolls there is evidence to indicate that
reefs on many atolls were at or above present level 4500-4000 years BP
(Hopley, 1987; Pirazzoli & Montaggioni, 1988a, 1988b), with dates from
Makatea, Mataiva and Takapoto indicating that in French Polynesia they may
have grown to above present sea level by 5500 years BP. We believe that the
first phase of Holocene reef development on atolls was characterised by catch-
up reef growth.

Three stage model of development of the Cocos Islands

Based on the data in Fig.7.7, a three-stage model can be proposed for the
Holocene evolution of the Cocos Islands (Woodroffe et al., 1990a). The first
phase from about 8000 to 4500 years BP was a phase of rapid vertical reef
growth as the reefs strived to ‘catch-up’ with a rapidly rising sea level
(Neumann & Macintyre, 1985). The radiocarbon-dated coral samples
portrayed in Fig. 7.7 do not directly indicate the position of sea level, except
for the microatoll samples shown from Cocos, because corals can grow in a
range of water depths (Davies & Montaggioni, 1985). Nevertheless, there is
some consistency of dates from around the atoll rim at Cocos. Dates from a
couple of cores do lag significantly behind the general pattern, but these were
from cores on the lagoonward side of a leeward reef (i.e. those shown in
transect X, Fig. 7.6).
The second phase from about 4500 to 3000 years BP was a phase of reef flat
formation as reefs caught up with sea level and consolidated. This former reef
flat is indicated by the presence of fossil corals, particularly specimens of
massive and branching Porites in growth position, which underlie a
conglomerate platform of cemented coral boulders, which is widespread
 Coral atolls 283

around the island (Fig. 7 . 8 ~ )In. situ Porites microatolls (Fig. 7.8b), radiocarbon-
dated at about 3000 years BP, growing up to 80cm above their modern living
counterparts, indicate that the sea level was higher and that there has been late
Holocene emergence (Woodroffe et al., 1990b). This represents a give-up reef.
The boulder conglomerate appears in many places on Cocos to have developed
as a reef flat deposit under higher sea level, though on other atolls the origin of
similar boulder conglomerates is more controversial and is discussed below.
The third phase, perhaps starting around 3500 years ago and continuing to
the present, was a phase of reef island formation. It is clear that islands are
geologically very young. Radiocarbon dating of shingle from within the sands
of West Island on Cocos indicates that the core of the island was formed
around 3500-3000 years BP overlying the conglomerate platform, and that
there has been gradual accumulation of island sediments since then, with the
extension of a series of recurved spits flanking the larger interisland passages.
Guppy (1 889) speculated on how the reef islands might have formed, putting
particular island shapes into an evolutionary sequence. There are insufficient
dates on other reef islands around the atoll rim, however, to substantiate his
theories. It has been suggested by several researchers that reef islands on other
atolls may have formed as a result of a net fall in sea level in the last few
thousand years (Gardiner, 1903; Kuenen, 1933; Schofield, 1977a). It remains
unclear how important changes in sea level are to patterns of island
accumulation, but reef islands are now recognised as being very dynamic, and
are examined below.
Lagoonal infill occurred throughout the three stages of Holocene reef
development, but it changed in nature. During the first phase, as reefs were
catching up with sea level, we infer that the lagoon was a very open
environment characterised by uninhibited water exchange, and probably
covered with flourishing coral cover. During the second, phase the catch-up of
the atoll rim and the development of reef flats decreased the exchange of water
and appears to have permitted extensive branching coral growth in the
relatively sheltered lagoon. Coral from vibrocores in the southern lagoon
indicate abundant acroporid corals around 3000 years BP. During the third
phase, reef island formation has meant that the input of sediment into the
lagoon from ocean reef and reef flats is restricted to interisland passages, and
quiet water environments and enclosed lagoonlets (termed teloks on Cocos,
and barachois elsewhere) have developed in which finer grained sediments
accumulate (Smithers et al., 1994).
The foregoing three-stage model relating sea-level change to reef growth,
reef flat formation and island formation appears applicable to many atolls. In
Fig. 7.7 the same interpretation may be applied to atolls in Kiribati; the catch-
Figure 7.8. ( a ) An oblique aerial view of Pulu Pandan on the eastern rim of Cocos. The island sits upon a conglomerate platform which can be
seen outcropping along the eastern shore. (6)Microatolls of massive Porites, above the modern upper limit to growth of coral and dating around
3000 years BP. These corals indicate that at that time the sea was around 80cm above present relative to Cocos. ( c )Emergent reef of the blue
coral, Helioporu, along the southern shore of Bairiki, Tarawa, Kiribati. This fossil coral underlies the conglomerate platform (background), and
is evidence of a Holocene highstand of sea level.
 Coral atolls 285
up phase of reef growth is clear from dating boreholes on Tarawa (Marshall &
Jacobson, 1985), reef-flat development may be indicated by the extensive
areas of fossil Heliopora reef exposed on modern reef flats of southern Tarawa
(see Fig. 7.8c), and reef islands would seem to have developed after the fossil
reef and conglomerates had formed.
However, the rates of upward reef growth can also vary between (and
within) atolls, and this variation is likely to have ramifications for the absolute
timing of stages 2 and 3. In fact, lags in all three stages can be expected. For
instance, on Funafuti there is no indisputable evidence of emergence in the
form of in situ emergent corals. Though fossil corals were reported to imply a
higher sea level in the past by David & Sweet (1904), and several Holocene
oscillations in sea level were inferred by Schofield (1977b), it has not been
proven that any of these are beyond the reach of modern processes, (McLean
& Hosking, 1991). Radiometric dates from the modern reef flats both on the
ocean and lagoon sides of the atoll and from the fossil reef flat (at the same
elevation) beneath reef islands indicate that the major phase of reef flat
development took place 2000-3000 years BP on this atoll (McLean & Hosking,
1991). Late catch-up is implied, the reef reaching sea level at a time when
elsewhere sea level may have been descending from its higher position or was
approaching its present level. A corollary of this is that the main phase of
island building on Funafuti also lagged behind, and this is confirmed by the
fact that all of the 21 samples from island deposits around the atoll so far
radiocarbon dated have ages less than 2000 years BP.

Surface morphology
The reef rim of atolls, from seaward reef crest to the first significant break of
slope into the lagoon, frequently comprises three fundamental units: oceanside
reef flat, reef island (cay or motu), and lagoonside reef flat or sand flat. Sollas
(1904) suggested that on Funafuti the ‘belt of land which separates the ocean
from the lagoon presents much the same character as in the case of Keeling
Atoll, described by Darwin’ and he presented two diagrams (Fig. 7.9) as being
‘sufficient to show this’. Our surveys and observations indicate that the
margins of these two atolls have some important differences which were not
highlighted by Sollas but are of significance when considering the recent
evolution of the reefs and particularly the islands.
First, the width of the atoll rim is generally greater on Cocos, at least where
there is a wide reef flat at the southern end. Similarly, reef islands are much
wider on Cocos, where islands frequently are more than 300m wide, than on
Funafuti, where they are rarely more than 100 m wide (see Fig. 7.2).
286 R.F. McLean & C.D. Woodrofle

FUNAFUTI ATOLL
E

LJ-

;igure 7.9. Diagrammatic sections through the rim of Keeling (Cocos) and Funafuti
atolls, redrawn with double vertical exaggeration from Sollas (1904, Fig. 4). The upper
diagram was originally in Darwin ( 1 842) and reproduced by Sollas unmodified.
A. Constantly submerged portions of reef. B. Nullipore (algal) rim. C. Reef flat of coral
rock. D. Ledge of coral rock (conglomerate platform, Cocos; breccia sheet, Funafuti).
E. Seaward or outer ridge (hurricane bank, Funafuti). F. Floor of lagoon. J. Central flat
of island. L. Lagoon mound. R. Growing reefs of lagoon

Second, the ‘ledge of coral rock’ (D on Fig.7.9) has quite different

characteristics. The conglomerate platform on Cocos is a flat surface, with few
coral clasts of more than 50cm diameter, and with in situ corals representing a
former reef flat formed around 3000 years BP. In the case of Funafuti, it is a
narrow breccia sheet of loosely cemented coral rubble which does not appear
to extend far beneath the islands. This breccia is interpreted as the base of an
early island beach formed around 1500 years BP.
Third, on Funafuti reef flats there are large storm-tossed blocks and
extensive boulder tracts on the eastern and western sides of the atoll
respectively. Such coarse storm-generated deposits are virtually absent on
Cocos. A similar difference applies to the islands also, with coral gravel and
boulders the dominant island material on motu in Funafuti (excepting sandy
cays to the north of the atoll), while on Cocos islands are predominantly sandy,
with rubble on the modern beach but rare elsewhere.
Fourth, on Funafuti, the lagoon shore drops steeply from solid reef flat into
the lagoon with a distinct break of slope 3 4 m high. On Cocos there is a broad
sand flat flanking the southern islands, with sand spilling into and filling a
complex of lagoonal blue holes. Beachrock outcrops are found on the steeper
lagoon beaches of islands on Funafuti, but beachrock is rare in Cocos.
 Coral atolls 287
Finally, although the reef rims of Cocos and Funafuti display similar gross
characteristics, in detail their surface morphologies are quite different and such
contrasts are common on atolls elsewhere around the world.

Synthesis
A pattern of atoll development over an interglacial-glacial-interglacial cycle,
the last 125 000 years, is shown in Fig. 7.10. Such a pattern is presumed to have
also occurred in a similar manner over previous sea-level cycles. The sea-level
curve shown for that period is based upon oxygen isotope analysis of
foraminifera from deep-sea cores, and correlation with sequences of raised reef
terraces on uplifted shorelines (Chappell & Shackleton, 1986). The response of
an atoll is shown schematically. When sea level was high, as during the Last
Interglacial (marine oxygen isotope substage 5e), an atoll similar to that
presently found would have existed. The extent to which the lagoon was
infilled and to which the rim contained reef islands probably varied from atoll
to atoll, as it does between atolls today, though as the Last Interglacial appears
to have contained a period of around 12 000 years of relatively unchanging sea
level (Lambeck & Nakada, 1992), many atolls probably filled in to a far greater
extent than they have been able to during the last 6000 years of relatively stable
sea level. During interstadials (such as oxygen isotope substages 5c and 5a)
shallow lagoonal sediments and peripheral reefs were presumably exposed.
This seems to have been associated with phases of aeolianite building in
marginal reef areas such as Bermuda (Hearty, Vacher & Mitterer, 1992), but
has not yet been recognised in the subsurface stratigraphy of atolls.
As the sea became progressively lower, the atoll would have been exposed
as an emergent limestone island, such as present day Niue. During this
emergence, it underwent solution giving rise to a highly eroded karst surface,
with karst and speleothem formation (Li et al., 1989; Lundberg et al., 1990)
being possible at depths of tens of metres below present sea level, at the peak
of the last glaciation (1 8 000 radiocarbon years BP, corresponding to around
21 000 years BP determined by uranium-series dating, Bard et al., 1990).
Solutional features, such as blue holes, were formed, or enlarged during this
emergence, and internal drainage is likely to have concentrated solution into
the lagoon (Purdy, 1974). As the sea rose so the three phases of Holocene reef
development were experienced (indicated by 3, 4 and 5 in Fig.7.10), with
inundation of the last interglacial platform occurring before sea level reached
its present level because the atoll had subsided.
While this model may apply to Cocos, it is not equally applicable to all
atolls, and we consider below some of the factors which may lead to a slightly
 Sand Porous Limestone
Reef Coral and Sand
0
reef islands

unconformify
h

100
3
v

150
100.000
Years BP

Figure 7.10. A summary of the late Quaternary evolution of an atoll such as COCOS.The sea-level curve for the last 140000 years is from
Chappell & Shackleton (1986). Stage I is the Last Interglacial with an atoll like that at present. Stage 2 is the peak of the glacial, the previous
unconsolidated lagoonal and reefal sediments are lithified or eroded, and the island is highly karstified. Stages 3 , 4 and 5 correspond to I , 2 and
3 of Fig. 7.7, and are described in the text.
 Coral atolls 289

different atoll history. On some atolls the Last Interglacial surface is exposed
at the surface on the modern atoll, such as in the drier Line (i.e. Christmas
Island, unpublished results) and Phoenix Islands of eastern Kiribati (Tracey,
1972, and pers. commun., 1992), Anaa atoll in the Tuamotu Archipelago
(Veeh, 1966; Pirazzoli et al., 1988a), and Aldabra Atoll in the western Indian
Ocean (Braithwaite, Taylor & Kennedy, 1973). On these atolls there might be
either no subsidence, or subsidence might be countered by lithospheric flexure
(see Chapter 12). Last Interglacial reefs occur at 7-14m beneath present sea
level on Enewetak (Tracey & Ladd, 1974; Szabo, Tracey & Goter, 1985), at
6-1 1 m beneath Mururoa Atoll in the Tuamotus (Labeyrie, et al., 1969;
Trichet, Repellin & Oustribre, 1984; Perrin, 1990), at 8-17m on Tarawa in
Kiribati (Marshall & Jacobson, 1985), and at 7-22m beneath the lagoon on
Aitutaki, Pukapuka and Rakahanga in the Cook Islands (Gray et al., 1992).
Some lowering of the previous atoll surface can undoubtedly be attributed to
solution, which is likely to have been least on the drier atolls (Menard, 1982).
However, if Last Interglacial reefs were all at sea level (and some may have
lagged behind as catch-up reefs), the occurrence of the Thurber discontinuity
at different depths indicates either flexure of the crust and upper mantle
(Lambeck & Nakada, 1992) or differential subsidence (see Chapter 12).
Differential vertical movements appear particularly likely for the Cook
Islands, where Last Interglacial limestones are emergent on the makatea
islands and range in elevation from 10-20 m above sea level (Woodroffe et al.,
1991b), but occur at 7-22m below sea level on atolls (Gray et al., 1992).
Although there are karst features developed on the surface of the emergent
makatea reefs (Stoddart, Woodroffe & Spencer, 1990), differences in the
elevation of the Last Interglacial surface between islands in a similar climatic
setting (i.e. Aitutaki and Atiu, which are <200 km apart), must result from
vertical movements and cannot be entirely ascribed to solution.
Holocene reef growth rates are known to differ substantially from place to
place (Hopley, 1982; Buddmeier & Smith, 1988). Individual reefs may adopt a
keep-up, catch-up or give-up strategy, and there may be considerable
variations from one part of a reef to another. As a result the time at which a reef
reaches sea level also varies. Whereas in COCOSthat time seems to have been
around 3500 years BP, in other parts of the Pacific it may be around 4500 years
BP (Hopley, 1987), or, as in parts of French Polynesia, around 5500 years BP
(Pirazzoli et al., 1988a, 1988b). On Funafuti, we have suggested that it may
have been as recently as 2000 years BP. In the Maldives there is also some
evidence for reef flat development around modern level 3000 years BP, but
evidence of higher sea level is ambiguous (Woodroffe, in press). It is tempting
to speculate that this late time of catch-up is linked to a deep Last Interglacial
290 R.F. McLean & C.D. Woodroffe
surface from which vertical Holocene reef growth commenced; perhaps this
might be a consequence of more rapid subsidence. However, there would need
to be far more drilling and dating results from Funafuti before this could be
substantiated.
Variations in the timing of reef catch-up to sea level may also explain the
variety of surface rubble conglomerate deposits on atolls. Boulder deposits,
cemented into a conglomerate platform, on tropical islands have been
interpreted either as evidence of a sea level higher than present (Daly, 1934 ;
Newell, 1961), or alternatively as a result of storm-wave action (Shepard et al.,
1967; Newell & Bloom, 1970). In addition, observations of fossil coral reefs at
higher elevations than their modern, living counterparts appeared to indicate
that the sea stood 1-2m above its present level with respect to many of the
coral atolls of the Pacific and Indian Oceans (Stearns, 1945; Cloud, 1952), and
that it has subsequently fallen relative to those islands. Cemented coral
conglomerates on the reef flats and islands of atolls and above the present limit
to coral growth have been radiometrically dated on many islands to about
4000-3000 years BP, e.g. from the Marshall Islands (Tracey & Ladd, 1974 ;
Buddemeier, Smith & Kinzie, 1975), Kiribati and Tuvalu (Tracey, 1972;
Schofield, 1977a, 1977b; Valencia, 1977; Guinther, 1978; McLean &
Hosking, 1991), Fiji (Ash, 1987; Miyata et ul., 1990; Shepherd, 1990; Nunn,
1991), and the Cook Islands (Scoffin et al., 1985; Yonekura et al., 1988;
Woodroffe et al., 1 9 9 0 ~ )The
. evidence has been examined in great detail
through French Polynesia, where the pattern of higher sea level from atolls in
the Tuamotu Archipelago can be demonstrated at around +0.8 t 0.2 m from
prior to 4000 years BP to at least 1500 years BP (Montaggioni & Pirazzoli,
1984; Pirazzoli & Montaggioni, 1986, 1988a, 1988b; Pirazzoli et al., 1987a,
1987b, 1988a). Similar boulder conglomerates have been described from atolls
in the Indian Ocean, and again have been inferred to indicate a sea level higher
than present in the Maldives (Gardiner, 1903; Sewell, 1935, 1936a, 1936b;
Stoddart, Spencer Davies & Keith, 1966), and Diego Garcia (Stoddart, 197 1b).
Where extensive conglomerate platform occurs as a result of fossil reef flat
becoming emergent, then it forms an anchor upon which reef islands may
develop. That combined with the relatively long period of potential
accumulation may explain the larger islands in COCOSthan in atolls such as
Funafuti which have had reef islands forming for a shorter time. In addition,
there are different processes operative.
There have been few studies of when reef islands began to form, or of their
chronology of development (Stoddart, 1969). From the northern Great Barrier
Reef, radiocarbon dates on shingle clasts indicate that similar reef islands
formed a core at least 4000 years ago, with a major phase of island building
 Coral atolls 29 1

forming an upper terrace on shingle islands, sand cays and composite islands
3500-3000 years BP, and some further building 1500 years BP (McLean &
Stoddart, 1978; McLean et al., 1978; Stoddart, McLean & Hopley, 1978). On
the other hand, a study based upon extensive radiocarbon dating on an island
in the southern Great Barrier Reef, has demonstrated that shingle islands
appear to have accreted uniformly over the last few thousand years (Chivas
et al., 1986).
However, it is clear that not all atolls have experienced similar storm and
sea-level histories to islands on the Great Barrier Reef, and that given the
generally larger tidal range, sediment movement processes may be different on
the Great Barrier Reef to those experienced on atolls. Little is known about the
pattern of sediment accretion on atoll reef islands over the -3500 years over
which there has been a substrate upon which islands could form. At least three
different possibilties exist: first, islands may have accumulated uniformly over
that time, and thus be continuing to build up at a similar gradual rate (as
demonstrated in the case of the study of Chivas et al., 1986); second, islands
may have formed initially around 3500 years ago as conditions were first
favourable for supratidal accumulation, and then have received progressively
less sediment with time, or have been subject to less storm activity as a result
of reef crest growth dissipating the energy from the open sea (analogous to the
Holocene high-energy window suggested for the Great Barrier Reef by
Hopley, 1982, 1984; see also Chapter 8); third, islands may have accreted
during one or more episodes within that period (as suggested for the northern
Great Barrier Reef islands, McLean et al., 1978). It would be extremely useful
to know which, if any, of these simplified models of island growth is
appropriate, as it would give an insight into modern and future island
dynamics.
It is important to stress that reef islands are naturally dynamic, and that
sediment production occurs around reef islands, and that erosion, deposition
and cementation can occur concurrently on atolls today (Wiens, 1962). Some
islands may be in a stable equilibrium with neither addition nor loss of
sediment. However, on most islands, sediment is added and lost over time, and
there is more likely to be a dynamic equilibrium between inputs and outputs.
Islands adjust over a range of timescales. They may adjust to seasonal changes;
thus in the Maldives the seasonal reversal of the prevailing monsoon can lead
to sediment accumulation at one end of an island for a part of the year, and its
redistribution to the other end for the rest of the year. In Kiribati, during the
1983 El Niiio, mean monthly sea level rose 40cm. Shoreline erosion was
detected in places on Tarawa, but was not especially rapid or devastating
(Howorth & Radke, 1991).
292 R.F. McLean & C.D. Woodrofle
By contrast, one severe tropical cyclone can totally devastate an entire atoll,
destroying vegetation and reducing (or increasing) island area. Catastrophic
storms play an important role in both construction and destruction in those
areas which experience hurricanes over a longer timescale (Stoddart, 197 1c;
Bourrouilh-Le Jan & Talandier, 1985), though the effect of these episodic but
high-magnitude events is also likely to differ between islands. Rubble is
usually an important constituent of islands in the storm belt, a legacy of past
storms. Sand on the other hand tends to be stripped off islands by storms, but
to be moved back onto islands by the more regular processes that operate
between storms (see Fig. 1.1). Sand and shingle motus, typical of high-energy
settings of atolls which experience storms, are built up in part by rubble-sized
material moved by storms onto the reef flat. The impact of Hurricane Bebe on
the atoll of Funafuti, Tuvalu has been studied in particular detail (Maragos,
Baines & Beveridge, 1973). A rubble rampart, composed of corals ripped from
off the reef front, was thrown onto the reef flat or onto the elongate reef
islands, adding about 10% to the total land area of the atoll. Regular less severe
storms have broken down and redistributed the storm rubble (Baines,
Beveridge & Maragos, 1974; Baines & McLean, 1976a, 1976b).
Not all islands are affected equally by storms and the morphology of islands
differs accordingly (McLean, 1980). Fig. 7.1 1 illustrates schematically some
of the features of the morphology of islands in areas of different storm
frequency and intensity. Strong tradewinds outside hurricane belts can also
serve to modify islands, as in Kiribati or the Maldives where storms of
hurricane force are not frequent. Where storms are frequent, reef flats
generally contain rubble ramparts or degraded rubble deposits on the motus on
the more-exposed side of the atoll. There is often a well-developed algal ridge
(though this may not be related to storm frequency), and conglomerate
platforms are prominent and may extend across much of the reef flat,
underlying entire islands (i.e. northern Cook Islands); sand cays are found in
the less-exposed areas. Where storms are not as frequent or as severe, there are
less-extensive rubble deposits; the algal ridge is less prominent and the
conglomerate platform is not as extensive across the reef flat (i.e. COCOS
Islands; some islands in Kiribati). In storm-free, low-energy areas, rubble is
not a major component in island construction; instead sand cays are found even
on the outer atoll rim (i.e. many islands in Kiribati, Maldives).
Fig. 7.1 1 also depicts a series of schematic responses of island form to
processes (modified from Bayliss-Smith, 1988). Storms result in loss of sand
from cays and motus, but an input of rubble to the motus in the form of ramparts.
The motus which receive an input of rubble adjust with the medium-term
 Coral atolls 293

OCCASIONAL SEVERE STORMS m

I
P
MOW

CAY

STORM-FREE, RARE SURGE

Sand locallsed
b..ohrOck

Figure 7.1 I . The morphology of reefs and reef islands in areas of different storm
occurrence, and the response of form to process. Where there are frequent storms there
is usually abundant coral shingle and boulders, and there may be rubble ramparts and
reef blocks. Boulders are often cemented into a boulder conglomerate. Where there are
only occasional storms of less severity, there are generally less-extensive boulder
deposits on the reef flat, and where there are only rare storms, islands are predominantly
composed of sand (after Woodroffe, 1994). Storm events are indicated by bars in the
process diagram, and storms generally add material to shingle motu, but erode sand
cays (following Bayliss-Smith, 1988). Rapid change during storms is followed by
redistribution of coarse material on motus, and slow recovery of beaches on cays
between storms.

breakdown and redistribution of that material as they return to equilibrium.

Cays lose sand during storms, but are rebuilt towards equilibrium by beach
recovery through normal processes. Relaxation time (time taken to readjust
between high energy events) and recurrence interval (frequency of such
events) are important in controlling reef island morphology. When storms are
very frequent (or very severe) motus and cays may be in disequilibrium for
most of the time. When storms are occasional, complete recovery is possible
between storms and islands may be in dynamic equilibrium. Where there are
no storms cays should reach a stable equilibrium; a rare disturbing event such
as a storm surge can cause devastation on these islands, and such catastrophes
will require a long time for recovery (Fig. 7.1 1).
294 R.F. McLean & C.D. Woodroffe
Conclusion
In this chapter the response of atolls to large-scale movements of sea level
during the Quaternary has been discussed first, followed by an examination of
how atoll reefs adjusted to sea-level rise during the Holocene and how reef
islands were subsequently formed. Atoll structure appears relatively similar
between atolls and can be explained as a function of long-term geological
processes, upon which large-scale sea-level changes ( 100-1 50 m amplitude)
have been superimposed throughout the Quaternary. Surface morphology of
atoll reefs, reef islands, and lagoons, on the other hand, varies considerably
within and between atolls, and reflects finer-scale sea level (0-5 m amplitude)
and other environmental changes as well as contemporary processes.
There is an intimate relationship between coral atolls, reef islands and sea
level, which has been the focus of studies of two types. First, there have been
studies on the determination of sea-level histories based upon data from atolls
which serve as ‘dipsticks’ in mid-oceanic locations (i.e. Bloom, 1967). Such
data contribute to the record of sea-level change itself and have also been used
to test and constrain geophysical models of the response of a viscoelastic earth
to changes in ice and water loads (i.e. Clark et al., 1978; Peltier, 1988).
Second, there have been studies of how atoll environments have responded to
sea-level changes. For coral atolls adjustments (responses) have been
associated with major shifts in sea level operating over long periods of
geological time ( lo5 years), while for reef islands adjustments have been
associated with much more recent and subtle movements of sea level (lo3
years). Although other physical and biological processes are also involved, it
is clear that the development of both coral atolls and reef islands cannot be
understood without an assessment of the role of sea-level change.
Recognition of the sensitivity of atoll environments and reef islands to
variations in sea level has resulted in widespread concern that future sea-level
rise, as a consequence of global warming (Houghton, Jenkins & Ephraums,
1990; Houghton, Callander & Varney, 1992), will have devastating effects.
Coral atolls, including several entire nations (e.g. Kiribati, Tuvalu, Maldives),
appear particularly vulnerable (Commonwealth Secretariat, 1989; Lewis, 1990);
indeed, the total disappearance of atoll reef islands has been foreshadowed under
the more extreme predictions (Roy & Connell, 1989, 1990, 1991), along with the
imminent displacement of the populations of atoll nations (Connell & Roy,
1989). The challenge for atoll geomorphology is to develop robust models of the
natural morphodynamics of reefs, reef flats and particularly reef islands, which
can then be used to assess the impact of changes in boundary conditions, whether
those be associated with sea-level rise, storm incidence, or anthropogenic
modification (usually degradation) of reef ecosystems.
 Coral atolls 295
References
Adey, W.H. (1978). Coral reef morphogenesis: a multidimensional model. Science,
202,83 1-7.
Ash, J. (1987). Holocene sea levels in northern Viti Levu, Fiji. New Zealand Journal
of Geology and Geophysics, 30,43 1-5.
Baines, G.B.K., Beveridge, P.J. & Maragos J.E. (1974). Storms and island building at
Funafuti Atoll, Ellice Islands. In Proceedings Second International Coral Reef
Symposium, Brisbane, vol. 2, pp. 485-96.
Baines, G.B.K. & McLean, R.F. (1976a). Resurveys of 1972 hurricane rampart on
Funafuti atoll. Search, 7, 36-7.
Baines, G.B.K. & McLean, R.F. (1976b) Sequential studies of hurricane bank
evolution at Funafuti atoll. Marine Geology, 21, MI-M8.
Bard, E., Hamelin, B., Fairbanks, R.G. & Zindler, A. (1990). Calibration of the I4C
timescale over the past 30000 years using mass spectrometric U-Th ages from
Barbados corals. Nature, 345,405-1 0.
Bayliss-Smith, T.P. (1988). The role of hurricanes in the development of reef islands,
Ontong Java Atoll, Solomon Islands. Geographical Journal, 154,377-91.
Bezrukov, P.L. (1973). Principal scientific results of the 54th cruise of the R.V. Vitian
in the Indian and Pacific Oceans (Feb-May 1973). Oceanology, 13, 761-6.
Bloom, A.L. (1967). Pleistocene shorelines: a new test of isostasy. Geological Society
of America Bulletin, 78, 1477-94.
Bourrouilh-Le Jan, F. & Talandier, J. (1985). SCdimentation et fracturation de haute
energie en milieu ricifal: tsunamis, ouragans et cyclones, leurs effets sur la
skdimentation et la gCometrie d’un atoll. Marine Geology, 67,263-333.
Braithwaite, C.J.R. (1982). Progress in understanding reef structure. Progress in
Physical Geography, 6,505-23.
Braithwaite, C.J.R., Taylor, J.D. & Kennedy, W.J. (1973). The evolution of an atoll:
the depositional and erosional history of Aldabra. Philosophical Transactions
of the Royal Society of London, Ser. B, 266, 307-40.
Buddemeier, R.W. & Smith, S.V. (1988). Coral reef growth in an era of rapidly rising
sea level: predictions and suggestions for long-term research. Coral Reefs, 7,
5 1-6.
Buddemeier, R.W., Smith, S.V. & Kinzie, R.A. (1975). Holocene windward reef-flat
history, Enewetak Atoll. Geological Socieo of America Bulletin, 86, 1581-4.
Chamberlain, N.G. (1960). Cocos Island magnetic survey, 1946. Bureau of Mineral
Resources, Record, 1960/124.
Chappell, J. ( I 974). Late Quaternary glacio- and hydro-isostasy on a layered earth.
Quaternary Research, 4,429-40.
Chappell, J. & Polach, H. (1991). Post glacial sea level rise from a coral record at
Huon Peninsula, Papua New Guinea. Nature, 349, 147-9.
Chappell, J. & Shackleton, N.J. (1986). Oxygen isotopes and sea level. Nature, 324,
137-40.
Chivas, A., Chappell, J., Polach, H., Pillans, B. & Flood, P. (1986). Radiocarbon
evidence for the timing and rate of island development, beach-rock formation
and phosphatization at Lady Elliot Island, Queensland, Australia. Marine
Geology, 69, 273-87.
Clark, J.A., Farrell, W.E. & Peltier, W.R. (1978). Global change in post glacial sea
level: a numerical calculation. Quaternary Research, 9, 265-87.
Cloud, P.E. (1952). Preliminary report on geology and marine environments of
Onotoa Atoll, Gilbert Islands. Atoll Research Bulletin, 12, 1-73.
Collins, L., Zhu, Z.R., Wyrwoll, K.H., Hatcher, B.G., Playford, P., Chen, J. H.,
Eisenhauser, A. & Wasserburg, G.J. (1993). Late Quaternary facies
296 R.F. McLean & C.D.Woodroffe
characteristics and growth history of a high latitude reef complex: the Abrolhos
carbonate platforms, eastern Indian Ocean. Marine Geology, 111, 203-1 2.
Commonwealth Secretariat ( 1989). Climate change: meeting the challenge. Report of
a Commonwealth group of experts, London, September 1989.
Connell, J. & Roy, P. (1989). The greenhouse effect: the impact of sea level rise on
low coral islands in the South Pacific. In Studies and reviews of greenhouse
related climatic change impacts on the Pacific Islands, ed. J.C. Pernetta & P.J.
Hughes, pp. 106-33. Majuro: Association of South Pacific Environmental
Institutions.
Cullis, C.G. (1904). The mineralogical changes observed in the cores of the Funafuti
borings. In The Atoll of Funafuti: borings into a coral reef and the results,
pp. 392-420. Royal Society, London.
Daly, R.A. (19 15). The glacial-control theory of coral reefs, Proceedings of the
American Academy of the Arts and Science, 51, 155-25 1.
Daly, R.A. ( I 925). Pleistocene changes of level. American Journal of Science, Series
5,10,281-313.
Daly, R.A. (1934). The changing world ofthe Ice Age. New Haven: Yale University
Press. 27 1 pp.
Darwin, C. (1 842). The structure and distribution of coral reefs. London: Smith,
Elder & Co.
Darwin, C. (1 859). On the Origin of Species. London: Smith, Elder & Co.
David, T.W.E. & Sweet, G. (1904). The geology of Funafuti. In The Atoll ofFunafuti:
borings into a coral reef and the results, pp. 61-124. Royal Society, London.
Davies, P.J. & Montaggioni, L.F. (1985). Reef growth and sea-level change: the
environmental signature. In Proceedings Fifth International Coral Reef
Congress, Tahiti, vol. 3, pp. 477-515.
Davis, W.M. ( 1928). The coral reefproblem. American Geographical Society, Special
Publication, 9. 596 pp.
Eisenhauser, A., Wasserburg, G.J., Chen, J.H., Bonani, G., Collins, L.B., Zhu, Z.R. &
Wyrwoll, K.H. (1993). Holocene sea-level determination relative to the
Australian continent: UiTh (TIMS) and I4C (AMS) dating of coral cores from
the Abrolhos Islands. Earth and Planetary Science Letters, 114, 529-47.
Emery, K.O., Tracey, J.I., Jr. & Ladd, H.S. (1954). Geology ofBikini and nearby
atolls. United States Geological Survey, Professional Paper, 260 A. 265 pp.
Fairbanks, R.G. (1989). A 17,000-year glacio-eustatic sea level record: influence of
glacial melting rates on the Younger Dryas event and deep-ocean circulation.
Nature, 342, 637-42.
Finlayson, D.M. (1970). First-order regional magnetic survey at COCOS Island,
Southern Cross and Augusta. Bureau of Mineral Resources, Record, 1970/101.
Gardiner, J.S., ed. ( 1 903). Thefauna and geography of the Maldive and Laccadive
Archipelagoes, being an account of the work carried on and of collections
made by an expedition during years 1899 and 1900. Cambridge University
Press (2 volumes).
Gaskell, T.F. & Swallow, J.C. (1953). Seismic experiments on two Pacific atolls.
Occasional Papers, Challenger Society, 3, 1-8.
Gibb Australia (1985). Tuvalu Lagoon Bed Resources Survey. Report to Australian
Development Assistance Bureau, Canberra. 26 1 pp.
Ginsburg, R.N., Lloyd, R.M., Stockman, K.W. & McCallum, J.S. (1963). Shallow-
water carbonate sediment. In The sea: ideas and observations on progress in
the study ofthe seas, vol. 3, ed. M.N. Hill, pp. 554-82. New York: Wiley
Interscience.
 Coral atolls 297
Gray, S.C., Hein, J.R., Hausmann, R. & Radtke, U. (1992). Geochronology and
subsurface stratigraphy of Pukapuka and Rakahanga atolls, Cook Islands: late
Quaternary reef growth and sea level history. Palaeogeography,
Palaeoclimatology, Palaeoecology, 91,377-94.
Guilcher, A. (1 988). Coral reef geomorphology. Chichester: Wiley.
Guillou, H., Browse, R., Gillot, P.Y. & Guille, G. (1993). Geological reconstruction
of Fangataufa Atoll, South Pacific. Marine Geology, 110,377-9 I .
Guinther, E.B. (1978). Observations on terrestrial surface and subsurface water as
related to island morphology at Canton atoll. Atoll Research Bulletin, 221,
171-83.
Guppy, H.B. (1 889). The Cocos-Keeling Islands. Scottish Geographical Magazine, 5,
281-97,457-74,569-88.
Hanzawa, S. (1938). Studies on the foraminifera fauna found in the bore cores from
the deep well in Kita-Daito-Zima (North Borodino Island). Proceedings of the
Imperial Academy, Tokyo, 14,38490.
Hearty, P.J., Vacher, H.L. & Mitterer, R. (1992). Aminostratigraphy and ages of
Pleistocene limestones of Bermuda. Geological Society of America Bulletin,
104,47 1-80.
Hopley, D. (1984). The Holocene ‘high energy window’ on the central Great Barrier
Reef. In Coastal geomorphology in Australia, ed. B.G. Thom, pp.135-50.
Sydney: Academic Press.
Hopley, D. (1987). Holocene sea-level changes in Australasia and the Southern
Pacific. In Sea surjace studies: a global review, ed. R.J.N. Devoy, pp. 375-408.
London: Croom Helm.
Houghton, J.T., Callander, B.A. & Varney, S.K., eds. (1992) Climate change 1992:
the supplementary report to the IPCC scientific assessment. Cambridge
University Press.
Houghton, J.T., Jenkins, G.J. & Ephraums, J.J., eds. (1990). Climate change: the
IPCC scientific assessment. Cambridge University Press.
Howorth, R. & Radke, B. (1991). Investigation of historical evidence of shoreline
changes: Betio, Tarawa Atoll, Kiribati, and Fongafale, Funafuti Atoll, Tuvalu.
Proceedings workshop on coastal processes in the South Pacific nations, Lae,
1987, CCOP/SOPAC Technical Bulletin, 7 , 9 1-8.
Jongsma, D. (1976). A review of the geology and geophysics of the Cocos Islands and
Cocos Rise. Bureau of Mineral Resources, Record, 1976/38.
Judd, J.W. (1904). General report on the materials sent from Funafuti and the methods
of dealing with them. In The Atoll of Funafuti: boring into a coral reef and the
results, pp. 167-85. Royal Society, London.
Katupotha, J . (1 988). Evidence of high sea level during the mid-Holocene on the
southwest coast of Sri Lanka. Boreas, 17,209-13.
Kuenen, P.H. (1 933). Geology of coral reefs. In Snellius Expedition in the eastern
part of the Netherlands East Indies, 1929-30. Vol. 5. Geological Results, No.2.
Utrecht: Kemink. 125 pp.
Labeyrie, J., Lalou, C. & Delibrias, G. (1969). Etude des transgressions marines sur
l’atoll de Mururoa par la datation des diffkrents niveaux de corail. Cahiers du
Pacifique, 13,59-68.
Ladd, H.S. (1948). Fossil land snails from western Pacific atolls. Journal of
Paleontology, 32, 183-98.
Ladd, H.S., Ingerson, E., Tonsend, R.C., Russell, M. & Stephenson, H.K. (1953).
Drilling on Eniwetok Atoll, Marshall Islands. American Association of
Petroleum Geologists Bulletin, 37,2251-80.
298 R. F. McLean C
? C.D. Woodroffe
Ladd, H.S., Tracey, J.I., Jr. & Gross, M.G. (1970). Deep drilling on Midway Atoll.
US.Geological Survey Professional Paper, 680A, A 1-22.
Ladd, H.S., Tracey, J.I., Jr. & Lill, G.G. (1948). Drilling on Bikini Atoll, Marshall
Islands. Science, 107, 5 1-5.
Lalou, C., Labeyrie, J. & Delibrias, G. ( 1 966). Datation des calcaires coralliens de
I’atoll de Mururoa (Archipel des Tuamotu) de l’epoque actuelle jusqu’a-
500,000 ans. Comptes Rendus Academie des Sciences, Paris, 263, 1946-9.
Lambeck, K. & Nakada, M. ( 1 992). Constraints on the age and duration of the last
interglacial period and on sea-level variations. Nature, 357, 125-8.
Lewis, J. (1990). The vulnerability of small island states to sea level rise: the need for
holistic strategies. Disasters, 14, 241-8.
Li, W-X., Lundberg, J., Dickin, A.P., Ford, D.C., Schwarz, H.P., McNutt, R. &
Williams, D. (1989). High precision mass spectrometric uranium-series dating
of cave deposits and implications for paleoclimate studies. Nature, 339, 534-6.
Lighty, R.G., Macintyre, I.G. & Stuckenrath, R. ( 1 982). Acropora palmata reef
framework: a reliable indicator of sea level in the Western Atlantic for the last
10,000 years, Coral Reefs, 1, 125-30.
Lincoln, J.M. & Schlanger, S.O. ( I 99 I ). Atoll stratigraphy as a record of sea level
change: problems and prospects. Journal of Geophysical Research, 96,6727-52.
Lloyd, J.W., Miles, J.C., Chessman, G.R. & Bugg, S.F. (1980). A ground-water
resources study of a Pacific Ocean atoll - Tarawa, Glbert Islands. Water
Research Bulletin, 16,646-53.
Locke, C.A. (1991). Geophysical constraints on the structure of Funafuti. South
Pacijic Journal of Natural Sciences, 11, 129-40.
Lundberg, J., Ford, D.C., Schwarc, H.P., Dickin, A.P. & Li, W-X. (1990). Dating sea
level in caves. Nature, 343, 2 17-1 8.
Lyell, C. ( 1 832). Principles of geology. London: Murray.
McLean, R.F. (1980). Spatial and temporal variability of external physical controls on
small island ecosystems. In Population-environment relations in tropical
islands: the case of eastern Fiji, ed. H.C. Brookfield, pp. 149-175, Paris:
UNESCO.
McLean, R.F. & Hosking, P.L. (1991). Geomorphology of reef islands and atoll motu
in Tuvalu. South Pacijic Journal of Natural Science, 11, 167-89.
McLean, R.F. & Stoddart, D.R. (1978). Reef island sediments of the northern Great
Barrier Reef. Philosophical Transactions of the Royal Society of London, Ser.
A , 291, 101-17.
McLean, R.F., Stoddart, D.R., Hopley, D. & Polach, H.A. (1978). Sea level change in
the Holocene on the northern Great Barrier Reef. Philosophical Transactions of
the Royal Society of London, Ser. A , 291, 167-86.
Maragos, J.E., Baines, G.B.K. & Beveridge, P.J. ( 1 973). Tropical cyclone creates a
new land formation on Funafuti atoll. Science, 181, 1161-4.
Marshall, J.F. &Jacobson, G. (1985). Holocene growth of a mid-plate atoll: Tarawa,
Kiribati. Coral Reefs, 4, 11-17.
Menard, H.W. (1982). Influence of rainfall upon the morphology and distribution of
atolls. In The Oceanfloor, ed. R.A. Scrutton & M. Talwanii, pp. 305-1 I . New
York: Wiley.
Miyata, T., Maeda, Y., Matsumoto, E., Matsushima, Y., Rodda, P., Sugimura, A &
Kayanne, H. ( 1 990). Evidence for a Holocene high sea-level stand, Vanua
Levu, Fiji. Quaternary Research, 33, 352-9.
Montaggioni, L.F. & Pirazzoli, P.A. (1984). The significance of exposed coral
conglomerates from French Polynesia (Pacific Ocean) as indicator of recent
relative sea-level changes. Coral Reefs, 3, 29-42.
 Coral atolls 299
Murray, J. ( I 889). Structure, origin, and distribution of coral reefs and islands.
Nature, 39,424-8.
Nakada, M. (1986). Holocene sea levels in oceanic islands: implications for the
rheological structure of the Earth’s mantle. Tectonophysics, 121, 263-76.
Nakada, M. & Lambeck, K. (1988). The melting history of the late Pleistocene
Antarctic ice sheet. Nature, 333, 36-40.
Neumann, A.C. & Macintyre, I.G. (1985). Reef response to sea level rise: keep-up,
catch-up or give-up. In Proceedings Fifth Znternational Coral Reef Congress,
Tahiti, vol. 3, pp. 105-1 0.
Newell, N.D. ( 1 96 I ) . Recent terraces of tropical limestone shores. Zeitschrifrfiir
Geomorphologie, N.F., Suppl. Bd., 3, 87-106.
Newell, N.D. & Bloom, A.L. (1970). The reef flat and ‘two-meter eustatic terrace’ of
some Pacific atolls. Geological Society of America Bulletin, 81, I88 1-94.
Nunn, P. (1991). Sea-level changes during the last 6000 years in Fiji, Tonga and
Western Samoa: implications for future coastline development. SOPAC
Technical Bulletin, 7,79-90.
Parsons, B & Sclater, J.G. (1977). An analysis of the variation of ocean floor
bathymetry and heat flow with age. Journal of Geophysical Research, 82,
803-27.
Peltier, W.R. (1988). Lithospheric thickness, Antarctic deglaciation history, and ocean
basin discretization effects in a global model of postglacial sea level change: a
summary of some sources of nonuniqueness. Quaternary Research, 29,
93-1 12.
Perrin, C. (1990). Genkse de la morphologie des atolls: le cas de Mururoa (PolynCsie
FranGaise). Comptes Rendus des Acadernie de Sciences, Paris, 311,11,67 1-8.
Pirazzoli, P.A. (1991). World atlas of Holocene sea-level curves. Amsterdam:
Elsevier.
Pirazzoli, P.A., Delibrias, G., Montaggionin, L.F., Salikge, J.F. & Vergnaud-Grazzini,
C. (1987a). Vitesse de croissance laterale des platiers et Cvolution
morphologique rCcente de I’atoll de Reao, iles Tuamotu, PolynCsie fraqaise.
Annales de l’institut oceanographique, ns, 63, 57-68.
Pirazzoli, P.A., Koba, M., Montaggioni, L.F. & Person, A. (1988a). Anaa (Tuamotu
Islands, central Pacific): an incipient rising atoll? Marine Geology, 82, 261-9.
Pirazzoli, P.A. & Montaggioni, L.F. (1986). Late Holocene sea-level changes in the
northwest Tuamotu Islands, French Polynesia. Quaternary Research, 25,
350-68.
Pirazzoli, P.A. & Montaggioni, L. ( 1 988a). The 7000 yr sea-level curve in French
Polynesia: geodynamic implications for mid-plate volcanic islands. In
Proceedings Sixth International Coral Reef Symposium, Townsville, vol. 3,
pp. 467-72.
Pirazzoli, P.A. & Montaggioni, L.F. (1988b). Holocene sea-level changes in
French Polynesia. Palaeogeography, Palaeocli/natology, Palaeoecology, 68,
153-75.
Pirazzoli, P.A., Montaggioni, L.F., Salvat, B. & Faure, G. (1988b). Late Holocene sea
level indicators from twelve atolls in the central and eastern Tuamotus (Pacific
Ocean). Coral Reefs, 7,57-68.
Pirazzoli, P.A., Montaggioni, L.F., Vergnaud-Grazzini, C. & Saliege, J.F. ( 1987b).
Late Holocene sea levels and coral reef development in Vahitahi Atoll, eastern
Tuamotu Islands, Pacific Ocean. Marine Geology, 76, 105-16.
Purdy, E.G. (1974). Reef configurations, cause and effect. In Reefs in time and space,
ed. L.F. Laporte, pp. 9-76. Society of Economic Palaeontologists and
Mineralogists, Special Publication, 18.
300 R. F. McLean & C.D. Woodroffe
Ross, J.C. (1855). Review of the theory of coral formations set forth by Ch. Darwin in
his book entitled: Researches in Geology and Natural History. Natuurkundig
Tijdschrift voor Nederlandsch Indie, 8, 1-43.
Roy, P. & Connell, J. ( 1 989). Greenhouse: impact of sea level rise on low coral
islands in the South Pacific. Research Institute for Asia and the Pacific
Occasional Paper 6. 55 pp.
Roy, P. & Connell, J. ( 1 990). Greenhouse effects on atoll islands in the South Pacific.
Geological Society of Australia Symposium Proceedings, 1, 57-70.
Roy, P. & Connell, J. (1991). Climatic change and the future of atoll states. Journal of
Coastal Research, 7, 1057-75.
Schlanger, S.O. (1963). Subsurface geology of Eniwetok Atoll. United States
Geological Survey, Professional Paper, 260-BB, 99 1-I 066.
Schofield, J.C. (1977a). Effect of Late Holocene sea-level fall on atoll development.
New Zealand Journal of Geology and Geophysics, 20, 53 1-6.
Schofield, J.C. (l977b). Late Holocene sea-level, Gilbert and Ellice Islands, west
central Pacific Ocean. New Zealand Journal of Geology and Geophysics, 20,
503-29.
Scoffin, T.P., Stoddart, D.R., Tudhope, A.W. & Woodroffe, C.D. ( 1985). Exposed
limestone of Suwarrow Atoll. In Proceedings Fifth International Coral Reef
Congress, Tahiti, vol. 3, pp. 137-40.
Scott, G.A.J. & Rotondo, G.M. (1983). A model to explain the differences between
Pacific plate island atoll types. Coral Reefs, 1, 139-50.
Searle, D.E. (1994) Late Quaternary morphology of the Cocos (Keeling) Islands. Atoll
Research Bulletin, 401, 1-13.
Sewell, R.B.S. (1935). Studies on coral and coral formations in Indian waters.
Memoirs of the Royal Asiatic Society of Bengal, 9,461-540.
Sewell, R.B.S. (1936a). An account of Addu Atoll. John Murray Expedition
1933-1934, Science Reports, 1,63-93.
Sewell, R.B.S. ( 1 936b). An account of Horsburgh or Goifurfehendu Atoll. John
Murray Expedition 1933-1934, Science Reports, 1, 109-25.
Shepard, F.P., Curray, J.R., Newman, W.A., Bloom, A.L., Newell, N.D.,Tracey, J.I.
& Veeh, H.H. (1967). Holocene changes in sea level: evidence in Micronesia.
Science, 1 5 7 , 5 4 2 4
Shepherd, M.J. (1990). The evolution of a moderate energy coast in Holocene time,
Pacific Habour, Viti Levu, Fiji. New Zealand Journal of Geology and
Geophysics, 33, 547-56.
Smithers, S.G., Woodroffe, C.D., McLean, R.F. & Wallensky, E. (1994). Lagoonal
sedimentation in the Cocos (Keeling) Islands. In Proceedings Seventh
International Coral Reef Symposium, Guam.
Sollas, W.J. ( 1 904). Narrative of the expedition in 1896. In The Atoll of Funafuti:
borings into a coral reef and the results, pp 1-28. Royal Society, London.
Steams, H.T. (1945). Decadent coral reef on Eniwetok Island, Marshall Group.
Bulletin of the Geological Society of America, 56, 283-8.
Stoddart, D.R. ( 1962a). Three Caribbean atolls: Turneffe Islands, Lighthouse Reef
and Glover’s Reef, British Honduras. Atoll Research Bulletin, 87, 1-15 I .
Stoddart, D.R. (1962b). Coral islands by Charles Darwin. Atoll Research Bulletin, 88,
1-20.
Stoddart, D.R. (1965). The shape of atolls. Marine Geology, 3, 369-83.
Stoddart, D.R. (1969). Sea-level change and the origin of sand cays: radiometric
evidence. Journal of the Marine Biological Association of India, 11,44-58.
Stoddart, D.R. (197 la). Environment and history in Indian Ocean reef morphology.
Symposium of the Zoological Society of London, 28, 3-38.
 Coral atolls 30 1
Stoddart, D.R. ( 1 97 1b). Geomorphology of Diego Garcia Atoll. Atoll Research
Bulletin, 149, 7-26.
Stoddart, D.R. (1 97 1 c). Coral reefs and islands and catastrophic storms. In Applied
coastal geomorphology, Ed. J.A. Steers, pp. 155-97, London: Macmillan.
Stoddart, D.R. (1973). Coral reefs: the last two million years. Geography, 58,
3 13-23.
Stoddart, D.R. (1983). Grandeur in this view of life: Darwin and the ocean world.
Bulletin of Marine Science, 33, 521-7.
Stoddart, D.R. (1992). The foundations of atolls: first explorations. In Geology and
offshore mineral resources of the central Pacifc basin. ed. B.H. Keating &
B.R. Bolton, pp. 11-19. Circum-Pacific Council for Energy and Mineral
Resources Earth Science Series, vol. 14. New York: Springer-Verlag.
Stoddart, D.R., McLean, R.F. & Hopley, D. (1978). Geomorphology of reef islands,
northern Great Barrier Reef. Philosophical Transactions of the Royal Society of
London, Ser. B, 284,39-6 I .
Stoddart, D.R., Spencer Davies, P. & Keith, A.C. (1966). Geomorphology of Addu
Atoll. Atoll Research Bulletin, 116, 13-41.
Stoddart, D.R. & Steers, J.A. (1977). The nature and origin of coral reef islands. In
Biology and geology of coral reefs, vol. IV, ed. O.A. Jones & R. Endean,
pp. 59-105. Geology 11. New York & London: Academic Press.
Stoddart, D.R., Woodroffe, C.D. & Spencer, T. (1990). Mauke, Mitiaro and Atiu:
geomorphology of Makatea islands in the Southern Cooks, Atoll Research
Bulletin, 341, 1-61.
Szabo, B.J., Tracey, J.I. & Goter, E.R. (1985). Ages of subsurface stratigraphic
intervals in the Quaternary of Eniwetak Atoll, Marshall Islands. Quaternary
Research, 23,54-61.
Thurber, D.L., Broecker, W.S., Blanchard, R.L. & Potratz, H.A. (1965). Uranium-
series ages of Pacific atoll coral. Science, 149, 55-8.
Tracey, J.I. (1972). Holocene emergent reefs in the central Pacific. In Second National
Conference, American Quaternary Association, Abstracts, pp. 5 1-2. Miami:
University of Miami.
Tracey, J.I. & Ladd, H.S. (1974). Quaternary history of Eniwetok and Bikini atolls,
Marshall Islands. In Proceedings Second International Coral Reef Symosium,
Brisbane, vol. 2, pp. 537-50.
Trichet, J., Repellin, P. & Oustriere, P. (1984) Stratigraphy and subsidence of the
Mururoa Atoll (French Polynesia) Marine Geology, 36, 24 1-57.
Tushingham, A.M. & Peltier, W.R. (1991). Ice-3G: a new global model of late
Pleistocene deglaciation based upon geophysical predictions of post-glacial
relative sea level change. Journal of Geophysical Research, 96,4497-523.
Valencia, M.J. (1977). Christmas Island (Pacific Ocean): reconnaissance geologic
observations. Atoll Research Bulletin, 197, 1-14.
Veeh, H.H. (1966). Th23n/U23x and U234/U23x ages of Pleistocene high sea level stand.
Journal of Geophysical Research, 71, 3379-86.
Wiens, H. ( 1 959). Atoll development and morphology. Annals of the Association of
American Geographers, 49, 3 1-54.
Wiens, H. (1962). Atoll environment and ecology. New Haven: Yale University Press.
Wood-Jones, F. (1912). Coral and atolls: a history and description of the Keeling-
Cocos Islands, with an account of their fauna andflora, and a discussion of the
method of development and transformation of coral structures in general.
London: Reeve and Co.
Woodroffe, C.D. (1994). Morphology and evolution of reef islands in the Maldives. In
Proceedings Seventh International Coral Reef Symposium, Guam. pp. 1229-38.
302 R.F. McLean & C.D. Woodroffe
Woodroffe, C.D., McLean, R.F., Polach, H. & Wallensky, E. (1990~)Sea level and
coral atolls: Late Holocene emergence in the Indian Ocean, Geology, 18, 62-6.
Woodroffe, C.D., McLean, R.F. & Wallensky, E. (1990b) Darwin’s coral atoll:
geomorphology and recent development of the Cocos (Keeling) Islands, Indian
Ocean. National Geographic Research, 6,262-75.
Woodroffe, C.D., McLean, R.F. & Wallensky, E. (1994). Geomorphology of the
Cocos (Keeling) Islands. Atoll Research Bulletin, 402, 1-33.
Woodroffe, C.D., Short, S.A., Stoddart, D.R., Spencer, T. & Harmon, R.S. (I991 b).
Stratigraphy and chronology of late Pleistocene reefs in the southern Cook
Islands, South Pacific. Quaternary Research, 35, 246-63.
Woodroffe, C.D., Stoddart, D.R., Spencer, T., Scoffin, T.P. & Tudhope, A.W.
( 1 990a). Holocene emergence in the Cook Islands, South Pacific. Coral Reefs,
9, 3 1-9.
Woodroffe, C.D., Veeh, H. H., Falkland, A., McLean, R.F. & Wallensky, E. (1991a)
Last interglacial reef and subsidence of the Cocos (Keeling) Islands, Indian
Ocean. Marine Geology, 96, 137-43.
Yonekura, N., Ishii, T., Saito, Y., Maeda, Y., Matsushima, Y., Matsumoto, E. &
Kayanne, H. (1988) Holocene fringing reefs and sea-level change in Mangaia
Island, southern Cook Islands. Palaeogeography, Palaeoclimatology,
Palaeoecology, 68, 177-88.
 8
Continental shelf reef systems
D. HOPLEY

Introduction
Major advances in the understanding of landforms occurred when it was
recognised that environment could control, largely through climate, the suite
of processes operating on the landscape and in extreme instances produce
distinctive landforms. Climatic geomorphology was further advanced when it
was realised that climates have not remained stable during the period in which
modern landforms had evolved, and that modern processes may be operating
on relict landscapes (Stoddart, 1968). For example, in the areas peripheral to
the great ice sheets of the major glaciations, present fluvial processes are
superimposed over landscapes which may have been largely determined by
periglacial processes. In the tropics, etchplanation and deep weathering may
have alternated with periods of regolith stripping in response to wetter and
drier climatic phases to produce the characteristic savanna landforms of the
present day (Thomas, 1974).
In the coastal zone, landforms are produced by combination of subaerial
processes (in which climate may play a part) and oceanic processes which have
been presumed to be much more stable through time. Only very broad
morphogenetic regions have been recognised, for example, the humid tropics
(Bird & Hopley, 1969). Coastal landforms as we know them today however,
are young, being produced only during the relatively short period when sea
level has been at, or close to, its present position. Probably because of this,
inheritance within the coastal zone through the re-occupation of early
interglacial coastline positions and the reworking of landforms produced
125 000 years ago or earlier has been recognised as being of great importance
(e.g. Hopley, 1985).
Coral reefs at their broadest level of distribution are the ultimate example of
climatically controlled landforms being limited to largely tropical or
immediately subtropical waters. However, within the tropics zonality is

303
304 D.Hopley
difficult to recognise and it has been advocated that there are greater contrasts
across continental shelves and into the open oceans, than there are in a
latitudinal direction (Hopley, 1989a). Environmental control on the biological
organisms which produce coral reefs is therefore strong, and not surprisingly
the fringing reefs and shelf reefs found along continental margins are in many
ways very different when compared with those of volcanic islands and atolls in
open oceans (see Chapter 7).
This chapter examines the distinctive features of shelf reefs, with particular
reference to the Great Barrier Reef of northeastern Australia. It also suggests
that coral reefs, like landforms, inherit many of their characteristics from
earlier events and that just as changes to climate can pass through important
landscape-forming thresholds, changes within the ocean may have similar
effects on coral reefs.

The special features of fringing and shelf reefs

The proximity of the continental landmass is of far greater importance to coral
reefs than any kind of zonal gradient which can be found within tropical
waters. Continentality produces amplified temperature ranges in nearshore
waters which can extend across a continental shelf, though much reduced
towards the outer margins. Nearshore areas are also influenced by run-off from
the landmass and reefs have a restricted tolerance to reduced salinities,
turbidity, sediment loading and increases in nutrients which may emanate from
the land (for review see Hopley, 1989a).
Sea-level changes also have much greater impact on the shelf reef systems.
At the maximum of each glaciation, with a fall of 100 metres or more in sea
level, entire reef systems are left high and dry. Along the Great Barrier Reef
such a fall in sea level would put the coastline on the top of the continental
slope, most of which is steeply sloping producing a cliffed coastline at glacial
maxima, along which reef formation would have been difficult and very
limited (Hopley, 1982). Recolonisation during the postglacial transgression
may well have been from more distant refuges such as the Queensland Plateau
reefs and have required the establishment of suitable oceanic circulation
patterns for larval transport (Davies, Marshall & Hopley, 1985). Whilst low
sea levels may also have limited coral development to steeply sloping sides of
atolls and open ocean volcanic islands, these were immediately adjacent to the
reef systems which were subsequently colonised during the transgression.
There, the effect of changing sea level has been one of moving a vertical
zonation up and down a slope with little horizontal migration. In contrast, on
continental shelves, horizontal migration is the major effect of changing sea
 Continental shelf reef systems 305
SHELF REEF OPEN OCEAN ATOLL
Living reef zone
Living reef zone - c. 100 km -c. 10 km
Terrestrial influence

SL SL

Modern S.L.

Living reef zone

-
Living reef zone c. 80 km
-c.lOoO m

Terrestrial influence

SL

S.L. -3om

Living reef zone Living reef zone

- c. 1 km c. 1 m

Terrestrial
influence

S.L. -120117

Active reef growth Recent reef growth Pleistocene reef formatioi

Oceanic influence Terrestrial influence Volcanic foundation

Figure 8.1. A schematic comparison of the reef development on a continental shelf and
an atoll during Quaternary sea-level changes.

level and, as it has been established that the most influential gradient for coral
reefs is one which is normal to the continental shoreline, the critical zone of
freshwater influence, sediments and nutrients will have migrated to and fro
across the shelf with each transgression/regression cycle.
306 D.Hopley
Oceanic circulation is important in many ways. It provides the mechanism
for larval dispersal, may produce upwelling and nutrient enhancement, and
determines the residence time of water bodies which may not be ideal for reefs.
Circulation around open ocean reefs will be little different at low sea-level
stages as the obstruction formed by a volcanic island or atoll will have changed
little (presuming that major oceanic circulation has not changed). In contrast,
shelf reefs show major changes, and some circulation may be completely
restricted. For example, at any sea level lower than -20m Torres Strait
between Australia and New Guinea is closed and circulation across northern
Australia between Indian and Pacific Oceans is halted (Jennings, 1972). Major
ocean currents are also brought up hard against the steeply sloping shelf
margin, rather than being able to flood across the outer shelf region as during
an interglacial high. Changes to upwelling patterns may be expected.
Shelf reefs therefore have a steeper environmental gradient to withstand at
the present time and also experience a far greater contrast in ambient
conditions during sea-level change than do oceanic reefs. The history of
evolution of such reefs would be expected to show responses to changes in the
environment. The possibility that these changes pass across important
thresholds is discussed below.

The major ecological controls on coral reef growth

Within the general intertropical area defined by temperature, hermatypic or
reef building corals are limited by a number of ambient environmental
parameters. In many instances these controls are applied indirectly via the
zooxanthellae, the minute algal symbionts which inhabit the endodermal tissue
of the coral polyp. These algae are important in removing waste, providing the
coral with nutrients and most importantly, through controlling a number of
biochemical processes, influence the rate of calcification, or skeletal
deposition, as the coral colony grows. It is important that many of these
ecological controls are applied via plants not animals. Plants require light for
photosynthesis and respond directly to nutrients such as phosphate and nitrate.
Thus, whilst salinity as a control may act directly on the coral polyps
(a tolerance range between 30%0 and 40%0is required), the normal response to
prolonged reduced salinity is expulsion by the coral colonies of their symbiotic
zooxanthellae, resulting in coral bleaching and ultimately death. Similarly,
temperature extremes may also affect the polyps, but the response is also
zooxanthellae expulsion and bleaching. (Veron, 1986)
Other controls may be applied to the corals via the zooxanthellae. Although
heavy sedimentation may smother coral polyps and cause mortality, Great
 Continental shelf reef systems 307
Barrier Reef corals have been shown to be able to withstand sedimentation
rates and levels of turbidity an order of magnitude higher than those published
for Caribbean and Pacific reefs (e.g. Pastorok & Bilyard, 1985), (Table 8.1). At
lower levels of turbidity, the important factor is the more rapid attenuation of
light through the water column and whereas coral may be found at depths of
more than 100 metres in clear open ocean waters with a widely spaced and
intricate vertical zonation, in inshore waters the lower level of coral growth
may be as little as 8 metres with a greatly compressed vertical zonation. The
important factor is the amount of light required for photosynthesis by the
zooxanthellae. On the Great Barrier Reef, corals have been found on
Myrmidon Reef, located on the outermost shelf, at a depth of 100 metres, and

Table 8.1. Measured rates of sedimentation on reefs of the Great Barrier Reef
Province* and elsewhere

Mean or range
of sedimentation
Location (mg cm-2 day-') Source

Great Barrier Reef

I . Cape Tribulation, mainland 145.8 Hoyal(1986)
adjacent to disturbed rainforest
catchments
2. Cape Tribulation, mainland 88.4 Hoyal (1986)
adjacent to partially disturbed
rainforest catchment
3. Cape Tribulation, mainland 26.1 Hoyal ( 1986)
undisturbed rainforest catchment
4. Magnetic Island, high granitic 20-1 I4 Mapstone et al. ( 1 989)
inshore island
5. Low Isles, inner-shelf reef 69.7 (0.6-899.9) Marshall & Orr (193 1)
6. John Brewer Reef, mid-shelf 0.17-2.87 Hoyal (1986)
reef lagoon
Other
Kaneohe Bay, Hawaii, oceanic 3 5 41 Maragos ( 1 972)
volcanic island
Puerto Rico, large continental island 12.8-1 179.9 Cortes & Risk ( 1 985)
Canton Atoll, open ocean atoll 0.13 Smith & Jokiel (1975)
(based on calcification
and metabolism only)

"Rates quoted here are an order of magnitude greater than those quoted in Pastorok &
Bilyard (1985). This frequently quoted reference based on data from Guam does not
appear applicable to the Great Barrier Reef.
308 D. Hopley
almost 100% coral cover of Leptoseris, Pachyseris and Endophyllia spp. was
found between 60-80 metres depth (Fig.8.2). On the inner shelf of the
Cumberland and Northumberland Groups of islands, where tidal ranges in
excess of 6 metres are found and high turbidity as a result of re-suspension of
sediments is constant, the lower limit of coral growth may be as little as 4
metres (van Woesik, 1992).
Nutrients may also be limiting. Whilst the traditional view of oceanic reefs
living in a ‘nutrient desert’ has now been challenged and reefs can grow in a
wide range of naturally occurring nutrient environments (Kinsey, 199 l),
inshore reefs may be growing close to their tolerance limits with respect to
phosphate and nitrogen and are therefore susceptible to anthropogenically
enhanced nutrient levels or spikes in the natural system caused by disturbances
such as tropical cyclones (Furnas & Mitchell, 1986).
Thus, there may be a relatively wide envelope of oceanographic conditions
within which reefs will grow and within which the metabolic performance is
predictable. Reef metabolic performance may be expressed as gross primary
production, but more significantly for geological purposes as net calcification
(Kinsey, 1985). Over shallow areas with 100% coral cover this may reach
10kg m-2 a-I which, taking into account the density of the aragonite being
laid down and the porosity of the reef fabric, converts to a vertical growth
potential of approximately 7 to 8 m m per year (7 to 8 metres per thousand
years). This correlates well with the geological record as determined by dated
coral cores (Davies & Hopley, 1983). In other areas of the reef, net
’
calcification rates are proportionately slower: for example, 4 kg m2a- on algal
pavements and as low as 0.5 kg m2a- I on sandy lagoon floors.
However, reefs under stress may show different metabolic performances.
These stress situations may be produced directly by environmental
perturbations such as tropical cyclones. Shelf reefs, particularly those close
inshore, may be living continuously in conditions which are at the lower end of
their tolerance limits and thus may be under a continuous chronic stress
situation. These reefs are therefore much more susceptible to acute stresses
whether natural or human induced (Kinsey, 1988). Such stresses may be
expressed through expulsion of zooxanthellae and bleaching with the resultant
decreases in net calcification rates.
The response to nutrient increases is particularly interesting and relevant to
the geological history of shelf reef systems. For example, small increases in
phosphate have been shown to result in an increase in coral growth rates.
However, the skeleton laid down may be more porous and the actual rate of
calcification may be reduced. (Rasmussen, Ness & Cuff, 1994a) The coral
 Continental shelf reef systems 309
0

Modern spur and groove

soh corals
Lower spur and groove zone
Stepped
drop-off with
corals
(Montiporal

Convex coral covered buttresses

Pure stands Leptoseris
on buttresses, Pachyseris
and Endophyllia in grooves

Platey corals only

10
E
100- 4 cwo
Halimeda sand dope
110-
Limit of living corals
120 -
and crustosa corallines
130 -
Lower outcrop
140 - Maximum cwe and
pitting development
and scree slope

150 -

160 - Halimeda sand and

coarse debris 45O
170-
310 D. Hopley
skeleton becomes more susceptible to mechanical breakage by waves and tidal
surge. Simultaneously, algae, particularly macroalgae, which are competing
with corals for substrate space, may gain significantly from the fertilisation
effect. Within the normally accepted envelope of ambient nutrient levels algae
may be kept under competitive control by grazing fish. Once beyond a
particular threshold, however, they may grow to the extent that they smother
the corals and in particular limit the amount of light reaching the corals living
in their shade. Competition from macroalgae is as important as water
temperature in limiting the poleward limits of coral reefs (Wiebe et al., 1982).
The concept of thresholds is of particular importance to coral reefs. Beyond
a particular limit, a reef may change from coral dominated to algal dominated.
The concept of catastrophe theory has been applied to such situations in coral
reef areas with the cusp model, in which the point of change from one
condition to another may be different dependent on the direction of change,
being particularly appropriate (Buddemeier & Hopley, 1988). Thus, the
establishment of a coral reef community may require far higher levels of water
quality with respect to nutrients, salinity, turbidity and light than is required to
maintain that reef community, once established. The demise of the coral reef
community will take place in water quality conditions far inferior to those in
which the reef became established.
It is suggested that during a transgressive/regressive sea-level cycle shelf
reef systems may pass through important thresholds several times and that this
can be identified in the present constituent components of shelf communities
and in the vertical column of the Holocene reefal veneer. In recognising such
environmental flips it is important that one of the algal response communities
may also result in the formation of bioherms. The calcium carbonate producing
alga Halimeda under certain circumstances can be extremely prolific with
cover of up to 90%. Growth can be extremely rapid with dense stands having
productivity rates of 4 g C rnw2d-' gross (Hillis-Collinvaux, 1974). On the
Great Barrier Reef dense Halimeda meadows have been shown to have
densities as great as 4637 g m-2 (Drew & Abel, 1988), with the potential for
vertical sedimentation rates of 1 metre in about 500 years. Such a figure has
been confirmed by radiocarbon dating of sediments from Halimeda bioherms
with a rate of 1.75 m per 1000 years quoted (Marshall & Davies, 1988). Littler,
Littler & Lapointe, (1988) showed that shallow species of Halimeda are
adapted to take advantage of episodic nutrient pulses. Although they believed
that light was not a major controlling factor, the limited distribution of
Halimeda on fringing and nearshore reefs of the Great Barrier Reef (Drew &
Abel, 1988) suggests that nutrient enhancement when combined with low
salinity and sediment yield will not produce luxurious Halimeda meadows.
 Continental shelf reef systems 31 1
Origin and evolution of the Great Barrier Reef
The Great Barrier Reef (Fig.8.3) is the world's largest coral reef system,
extending off northeastern Australia for a distance of 2300 km over 14" of
latitude (Hopley, 1982). Even excluding the Torres Strait region, the shelf
waters exceed 230000km2 and of this almost 9% is occupied by reefs or
submerged shoals (Hopley, Parnell & Isdale, 1989).

PAPUA
NEW GUINEA

CORAL SEA

CORAL SEA PLATEAU

TOWNSVILLE

MACKAY

Northumberland
Group
AUSTRALIA

ROCKHAMPTON

BUNDABERG

GLAD STONE

Figure 8.3. The Great Barrier Reef province.

312 D. Hopley

The northern region of the reef has developed on a narrow shelf no more
than 50 km wide. Its most distinctive features are the linear or ribbon reefs
running parallel to the edge of the continental shelf almost as far south as
Cairns. They form a near continuous barrier of individual reefs up to 25 km
long. Inside many of the ribbon reefs are large banks at depths of 20-40
metres. These are formed almost entirely of Hulimedu, which appears to have
built structures equal in size to many of the coral reefs (Fig. 8.4). The middle
shelf is occupied by large platform reefs with extensive reef flats, some up to
25 km in length (Fig. 8 . 5 ~ )Closer
. to shore is a more open area of inner shelf,
where a number of small reefs capped by distinctive low wooded (mangrove)
islands are found. Most high islands of this region, which are not large in
number, have fringing reefs and in some areas fringing reefs also extend along
the mainland, most distinctly in the Cape Tribulation area some 100 km north
of Cairns (Fig. 8 3 ) .
As the continental shelf widens south of Cairns, the Great Barrier Reef
occupies only the outer one-third of the shelf. Reefs are more widely spaced
and generally have less-well-developed reef flats. Most reefs are irregular reef
patches, or crescentic features, aligned across the dominant southeasterly trade
winds. No ribbon reefs are found on the outer shelf but in some areas at least,
notably just south of Cairns and outside the Pompey Complex in the south,
there are linear outer shoals rising from depths of about 70 metres. South of
approximately 20" S the continental shelf widens even further to about 300 km.
With an increase of tidal range (to more than 4 metres on even the outermost
reefs in the Pompey complex) narrow, well-defined tidal channels up to 70
metres deep intersect the reefs. Even the innermost reefs are lOOkm from the
mainland. The outermost reefs of the Pompey Complex stretch for 200 km as a
solid mass of reef and lagoons 15 km wide with narrow, intricate channel
systems (Fig. 8 5 ) . Individual reefs may be more than 100 km2 in area. The
southern extent of the Pompey Complex is a distinctively T-line junction of

BANK CORRIDOR BANK CORRIDOR BANK

Large excavations-
'up to 2 metres d a p
Dense c o w
Caulerpa,
Halimeda etc pnd Vary harp b r r k
betwmvl3#6tated
occaiofml CINanous a d u m m & Low. i r r ~ l r
corals dead coral WtCmp.
outcrop, Plristocm? Pleistocene?

Figure 8.4. Idealised morphology of the Hulimedu banks behind Ribbon 5, northern
Great Barrier Reef. Determined from 1986 Platypus submersible transect.
 Continental shelf reef systems 313
reefs, to the south of which are the contrasting Swain Reefs, smaller flat-top
reefs closely spaced and with numerous sand cays. Halimeda shoals have been
reported in this area (Searle & Flood, 1988).
Although no fringing reefs are found along the mainland coast in the central
and southern section of the Great Barrier Reef, they are exceptionally well
developed around all continental islands north of about 21" S . At this point, the
tidal range increases to such an extent (more than 5 metres) that resuspension
of sediments appears to have depressed major reef development (Kleypas,
1991; van Woesik, 1992).
South of the Capricorn Channel the shelf narrows again to less than 100km.
The Bunker Capricorn Group of reefs, the southernmost of the Great Barrier
Reef, are a series of 22 reefs and 11 shoals of only moderate size but with
numerous vegetated sand cays.
In spite of its size, the Great Barrier Reef is one of the youngest reef systems
in the world, most of it being less than 1 million years old (Davies, 1992). Until
about 75 million years ago, Australia and Antarctica were joined. Most of
Australia lay south of 40" S far from waters warm enough for coral growth.
About 65 million years ago Australia began to split from Antarctica and moved
northwards. Subsequently, northeastern Australia was formed by rifting
between the Australian and Pacific plates and by the time the continental shelf
had formed, northern Australia lay close to 30" S latitude. Uplift, rifting and
vulcanism produced a complex rift basin system that was to control the
location and form of the continental shelf. As Australia continued to move
north, shelf evolution was dominated by fluvial sediment yield. Current annual
sediment input from north Queensland rivers alone is estimated at 28 million
tonnes (Pringle, 1986). Deltaic progradation produced extension of the shelf
margin (Symonds, Davies & Parist, 1983).
Surprisingly, even though Australia was within sufficiently warm waters by
26 million years ago, seismic records show little reef development apart from
on the northernmost area of the Great Barrier Reef. Reef building began less
than 500000 years ago over most of the Great Barrier Reef (Davies, 1992).
The reef sequence is therefore thin, less than 300 metres in thickness. Its entire
development has taken place during periods of rapidly fluctuating sea levels
produced by high-latitude glaciation. Reef growth has occurred during short
periods of high sea level. During the intervening periods of low sea level, the
reefs were subaerially eroded by karstic processes and continual recolonisation
of older foundations has produced reefs that are composite features made up of
a series of remnant reefs separated by unconformities, a process aided by
general shelf subsidence. The Holocene has been the final period of high sea
level, during which the final layer has been added to the reefal column. On
314 D.Hopley

Figure 8.5. ( a ) Mid-shelf reefs, although growing in shallower water than their outer-
shelf counterparts, probably started to grow at around about the same time, i.e. 8000
years BP. However, on the central Great Barrier Reef in particular, their foundations
tended to be deeper ( 1 5 to 30 metres). They therefore trailed sea level by 10 metres or
more as they grew upwards and only reached modern sea level subsequent to 6000
years BP. Nonetheless, extensive reef flats have developed though deep lagoons still
remain. (Darley Reef, central Great Barrier Reef.) ( b )Mainland fringing reefs are rare
and occur only north of Cairns. These reefs near Myall Creek south of Cape Tribulation
commenced to grow shortly before 7000 years BP on foundations provided by gravel
fans laid down by steep coastal streams. By 5400 years BP most reef growth had taken
place. Subsequently, as coastal water conditions deteriorated, these reefs have been in a
 Continental shelf reef systems 315

Caption for fig. 8.5 (cont.).

delicate state of balance typical of many mainland and nearshore island reefs in north
Queensland. (Cape Tribulation reefs.) ( c ) Even the largest reefs of the Great Barrier
Reef accreted principally in the short period between 8000 years BP,and the maximum
of the transgression about 6500 years BP. These reefs are part of the Pompey Complex
in the south central Great Barrier Reef with much of the growth and carbonate
production since 6500 years BP being directed towards lagoon infilling. (d)Flood plume
of the Murray River near Tully, north Queensland, shortly after the passage of Cyclone
Winifred in 1986. Such plumes carrying massive amounts of sediment have increased
during the period of European occupation due to extensive clearance of natural
rainforest catchments for agriculture.
316 D. Hopley
average this is approximately 20 metres in thickness but can vary from zero to
more than 30 metres. The evolution of the reef during the Holocene throws
much light on the overall development of this reef system.

Recolonisation, growth and maturation of a shelf reef system in

response to changing water quality
Hopley (1982) has documented the stages in reef growth following the
inundation of older Pleistocene foundations during the Holocene trans-
gression. Following inundation, growth on elevated portions of the
foundations and in particular on the windward margins enhances the original
relief. With stabilisation of sea level and vertical reef growth catching up, reef
flat development commences and, after extending around the margins of the
platform, enclosing single or multiple lagoons, infilling of the lagoons results
in a final reef form which is planar.
Whilst drilling has largely confirmed this model of reef development, it has
also been recognised that reef growth did not commence immediately the
antecedent platform was inundated. Initially (Davies & Marshall, 1979), this
was attributed to poor water-quality conditions over reef tops as regolith,
developed during the low sea-level phase, was reworked. However, during the
1980s extensive work on the Great Barrier Reef highlighted three important
factors:
Holocene reef growth commenced on the Great Barrier Reef almost
exclusively between 8320 and 7500 years BP regardless of the time of
inundation. A timelag of between 1200 and 2000 years separates
transgression from initial colonisation (Davies et ul., 1985).
0 extensive areas of the Great Barrier Reef, formerly thought to be submerged
coral reefs are now known to be Hulimedu bioherms (Orme, Flood &
Sargent, 1978; Orme, 1985; Orme & Salama, 1988; Davies & Marshall,
1985; Drew & Abel, 1985, 1988; Marshall & Davies, 1988; Phipps, Davies
& Hopley 1985; Searle & Flood, 1988). These algal bioherms commenced
to grow more than 10000 years ago, i.e. more than 2000 years before reef
growth was initiated.
further, recent studies (e.g. Kleypas & Hopley, in press) have also indicated
that the rate of reef growth has not been constant over the last 8000 years.
Various explanations have been put forward to explain these phenomena,
many of them related to water quality. However, no holistic attempt has been
made to explain the relationship between the changes in water quality and reef
growth during the Holocene. Geographical changes produced by rising sea
 Continental shelf reef systems 317
level were also accompanied by changes in climate, water circulation and
water quality. Changes are still taking place at the present time, exacerbated by
anthropogenic influences which ultimately have the potential for greatest
impact.
The major environmental changes which may have affected reef growth
include:

1. Flooding of the continental shelf

Up to about 1 1 000 years BP the postglacial transgression had risen against the
steeply sloping shoulder of the continental shelf and for much of the reef,
particularly in the north, a cliff shoreline would have predominated.
Subsequently, however, as the transgression flooded onto the shelf,
environmental conditions changed rapidly. There is evidence that at least
locally extensive fluvial sediments have been laid down over the shelf and
these were available for reworking. Soils which had formed over these
deposits and over older Pleistocene coastal plain and reefal features would
have stored nutrients which were also released into the nearshore waters.
Initially at least, the strong upwelling which may have taken place against the
steep continental slope would have remained, but especially north of Cairns
the outer shelf would have consisted of a linear range of limestone hills, the
precursors of the present ribbon reefs with only narrow passes through which
the sea flooded. Strong nutrient jetting through these is suggested during the
very early Holocene. Marshall (1988) has suggested that this upwelling of cold
deep water also played a major role in modulating surface temperature along
the margins of the shelf.

2. Flooding above the level of the older reefal foundations

This took place about 8500 years ago as the average depth of reefal
foundations is approximately 20 metres. Although some of the foundations in
the northern Great Barrier Reef, southernmost Great Barrier Reef and some
fringing reefs are shallower, once sea level had reached approximately -20
metres freer movement of water from the Coral Sea onto the shelf was
possible. As the older Pleistocene foundations no longer formed a barrier, this
exchange of water would have almost certainly increased the water quality of
the middle shelf. It also produced the mid-Holocene high-energy window
(Hopley 1984), which produced more turbulent mixing of waters than had
previously been possible and existed until the modern reefs grew up to sea
level, generally 5000-4000 years ago.
Figure 8.6. Seismic reflection profile inshore area northern Great Barrier Reef showing a mud silt wedge of up to 8 metres thickness overlying
Pleistocene surface.
 Continental shelf reef systems 319
3. The commencement of the postglacial stillstand
Modern sea level was achieved between 6000-6500 years BP along most of the
Great Barrier Reef. For reef growth this was important as it allowed reefs
which had formerly lagged behind the rise in sea level to catch up and
commence the lateral development of reef flats. Strong energy gradients
developed from windward to leeward margins of reefs and lagoons became
sinks for not only sediments but also nutrients. On the mainland, the
stabilisation of sea level had a further important effect. In contrast to the
previous 20000 years or more the coastline remained in almost exactly the
same position, apart from progradation in areas of rapid sediment yield.
Initially at least the area of sea floor immediately in front of the coastline had
little sediment. Pleistocene sediments had largely been driven onshore and
seismic surveys have shown that over much of the sea floor, the Pleistocene
surface outcropped (Searle et al., 1982; Johnson & Searle, 1984). However, in
the subsequent 6000 years the coastal streams have been delivering enormous
amounts of sediment to the coastline (Pringle, 1986) and these have built up a
mud and silt wedge (Fig. 8.6) adjacent to the coast up to 20 metres in thickness
and extending as much as 20 km from the mainland, i.e. out to at least the
innermost of the high islands with fringing reefs. This sediment, often in less
than 10 metres of water depth, is resuspended even in moderate southeasterly
conditions of 20-25 knots as can be found for up to 50% of the time during the
winter months. Nearshore waters have therefore changed from relative clarity
to highly turbid over the period of stillstand. Stability of the coastline has also
allowed the accumulation of nutrients not only in lagoons, but also in coastal
estuaries, particularly in mangrove areas.

4. Climatic change
Details of climatic change during the Holocene are available for at least the
Atherton Tableland behind Cairns (Kershaw, 1978) and may be indicative of
much of the central and northern Great Barrier Reef. Prior to 7800 years BP
sclerophyll vegetation existed on the tableland and rainfall totals are estimated
as being less than 1800mm. Between 7800 and 6500 years BP the vegetation
changed to simple notophyll vine forest and rainfall totals rose to about
2500mm. Between 6500-3000 years BP a complex mesophyll vine forest
existed and rainfall totals were as high as 3500mm. Subsequently, rainfall
totals have declined to about 2500 mm annually and a complex notophyll vine
forest has become established. These results suggest that the major growth
phase for coral reefs took place during a period of increasing rainfall when
totals have been less than present. The period of stillstand was one of
320 D.Hopley
maximum rainfall totals during which reefs reached modern sea level, whilst
the period of reef flat development has been one of declining rainfall. These
changes are important for natural sediment yield to the coast.
However, the relationships are not simple. Maung Maung Aye (1 976)
investigated drainage basin parameters related to erosion in north Queensland
catchments. Minimum denudation was related to mean annual rainfall totals of
between 2500-3000mm. An increase to totals above those of present
(>3000 mm) would automatically have resulted in delivery of greater sediment
loads and fresh water by small coastal catchments. The reason is that the
rainforest canopy gives no greater protection than it does between 2500-
3000mm annual rainfall, and the throughput of water is greater. Similarly, a
decrease in rainfall totals below 2500 mm is also demonstrated to be related to
an increase in sediment yield. Such rainfall totals are usually related to
seasonal rainfall and open eucalypt woodland vegetation, which gives far less
protection to the ground surface than rainforest. The implications for changing
natural sediment yield from coastal catchments for reef development is
discussed further below.

5. Human impacts
Although Aborigines occupied Australia more than 40 000 years ago and
almost certainly wandered the coastal plains which were the exposed
continental shelf at low sea-level stages, their impact on water quality, even
allowing for practices which included firing of native vegetation, would have
been minimal. In contrast, 200 years of white settlement has seen major
changes to the environment, resulting in increased sediment yield, more rapid
run-off and higher flood peaks and delivery of nutrients from agricultural
catchments through run-off (Fig. 8 . 5 4 . Other changes may be more visible
(Hopley, 1988, 1989b), but these are the ones which have caused a decline in
water quality, particularly since the middle of the twentieth century.

Phases in reef development in relation to water quality changes

Although not absolutely synchronised, many of these changes have taken place
close enough to the same time to produce important responses in coral reef and
associated biohermal growth, largely because most changes have been
reinforcing, i.e. have been in the same direction in terms of water quality, or
have been large enough on their own to pass through a reefal threshold. The
main phases recognised are:
 Continental shelf reef systems 32 1

1. Commencement of the Holocene to c.8000 years BP

Water quality was low during this period as seas were shallow and probably
highly turbid, as previously developed soils and regolith were reworked.
Mainland river mouths were sufficiently close to what is now the main reef
tract to allow their plumes to reach out to the older Pleistocene foundations of
the modern-day reefs. Nutrients released from the soils were reinforced by
nutrient upwelling on the shelf edge and jets extending through the passes on
the shelf edge. Coral growth during this period appears to have been
negligible. Problems with water quality may have been compounded by
incursions of cold upwelled water on the shelf edge and by larger temperature
extremes on the shelf because of its shallower nature compared with the
present.
It has also been suggested that the absence of larval replenishment centres
locally may also have delayed recolonisation processes with major reef refuges
occurring on the Queensland Plateau. Davies et al. (1985) have suggested that
the likely low sea-level water circulation pattern would have been dominated
by north to south flows across the Queensland Plateau, which would have
inhibited the westward dispersal of planulae and would have persisted until
sea-level rise was advanced sufficient to flood the Torres Strait between
Australia and New Guinea. At such time, a new oceanographic pattern would
have been established comparable to that which operates during the winter
months today, i.e. westward flows from the Coral Sea onto the Queensland
continental shelf. Such a change in circulation pattern would have occurred
around 8000 years BP when the shelf had been largely flooded.
Thus, coral reef growth to about 8000 years BP was negligible. Instead,
carbonate productivity appears to have been concentrated in what are now
recognised as major Halimeda banks. First recognised on the northern Great
Barrier Reef (Orme et al., 1978), extensive Halimeda deposits probably occur
along much of the outer third of the shelf in water depths exceeding 30 metres
and have been recently recognised as far south as the Swain Reef Complex
(Searle & Flood, 1988). The banks are up to 20 metres in thickness, overlying
pre-Holocene substrate, and began to grow about 10000 years BP. Living algae
still cover these banks (not always predominantly Halimeda) and although
accumulation rates could be as high as 2 metres per thousand years, i.e.
sufficient to produce the present thickness of the banks, this is probably at the
upper end of the scale at the present time. The indications are that the bulk of
these banks formed in the first half of the Holocene very rapidly and in
response to water conditions which favoured their growth, rather than that of
coral. Growth may have commenced soon after the immediate terrestrial
322 D. Hopley
influences (which currently appear to restrict Hulimeda) had migrated away
from the outer shelf.
Not surprisingly, several metres of loosely cemented Hulimedu gravels are
also found at the base of some Holocene reefal sequences, mainly on
horizontal surfaces, e.g. Yonge and Raine Island Reefs (Fig. 8.7). Although
there are no ages for the Holocene Hulimeda gravels, the oldest dates obtained
for overlying corals are 6580 2 120 years BP (Beta 40822) on Yonge Reef and
7040 ? 140 years BP (ANU 6621) on Raine Island. Notably amongst all cores
from the Great Barrier Reef to which this author has access, although
Halimedu may occur as individual fragments within the core, Hulimedu
gravels are found nowhere except at the base.
On both Yonge Reef and Raine Island, the top of the Pleistocene is also
composed of Hulimeda grainstones. Radiocarbon ages indicate a Pleistocene
age, but are not absolute and the Halimedu may have accumulated either
during an interstadial high sea level, or during the fall from the Last
Interglacial high sea-level stand. In both instances, the palaeogeography would
have been similar to that during the early Holocene with similar low water
quality and apparently the same ecological response.
Coral growth on fringing reefs of the present mainland and high continental
islands did not take place pre 8000 years BP for more obvious reasons, i.e. these
shallower depths had yet to be inundated (e.g. see Partain & Hopley, 1989)

2. 8000 years to 6500 years BP: The final period of transgression

Sea level was close to its modern position by 6500 years BP along much of the
Great Barrier Reef. In the final 1500 years of the transgression the rate of rise
may have been little more than lOmm per year. Nonetheless, the shoreline
retreated rapidly over the low-gradient inner-coastal shelf, removing terrestrial
influence away from the outer third of the shelf where most reef growth takes
place. With the removal of terrestrial run-off influences, conditions for coral
reef growth became ideal. All regolith by this time had been removed from the
older reefal foundations which, when colonised, may have been as much as 10
metres below the rising sea level. As the Torres Strait became flooded the
present oceanic circulation within the Coral Sea became established and larval
recruitment from the Coral Sea Plateau low sea-level refuges was more easily
achieved. The Atherton Tableland palaeoclimatic record (Kershaw, 1978)
shows increasing rainfall during this period, with the vegetation changing from
sclerophyll woodland to a simple notophyll vine forest. However, this is likely
to have restricted the terrestrial influence as the greater vegetational cover may
 Continental shelf reef systems 323
RAINE ISLAND REEF YONGE REEF

0-
5100f80
- 3each rock and
phosphatic cay sandstone
2- 51902 120
5170f8O
- $7402130 5680f90

4-
5010-+80
-
52 10280
6-
5600-+80
-
8- 58702 1 10
65802 120
-
10 63002 130 57002 lclo
g 10- 7w02 140
5
-
12-

-
14- 63002 130
-
16 - 257202 1540

-
18 -
24600f360
-
20 - Shingle an Halimeda
Coral, in situ? rubble packstone
Cemented shingle sad Cavity
and rubble
Figure 8.7. Dated cores from Yonge and Raine Island Reefs, in the far northern section
of the Great Barrier Reef.
324 D. Hopley
have reduced sediment yield and nutrient recycling by the rainforest system
could well have reduced the natural flow of nutrients to the ocean.
There is a strong possibility that these conditions saw a decline in the
growth of the previously flourishing Halimeda banks. The early Holocene
bathymetry of the shelf edge, formed by linear barriers of Pleistocene reefs and
narrow passes, was now replaced by a far more open developing shelf-edge
reef system. Although there is some evidence that parts of the outer ribbon
reefs may have kept up with sea-level rise, it is highly unlikely that the outer
shelf retained the full barrier morphology. The tidal jet vortex systems which
had previously delivered nutrient-rich water from below the thermocline to the
outermost shelf, as modelled by Thomson & Wolanski (1984) and Wolanski
et al. (1988), may have been substantially reduced. Until the outer reefs grew
up to sea level and re-established a continuous outer barrier, Halimeda growth
was probably retarded.
In contrast, coral reef growth was prolific and analysis of drilling results
suggests that up to 80% of framework construction in the Holocene reefal
veneer took place during this period and immediately afterwards. The reason
for this is clear. Kinsey (1985) has indicated sharp zonality in the production
of calcium carbonate over reefs, with maximum amounts of about 10kg per m2
per year occurring in areas of 100% coral growth. Today, such areas are
limited to the outermost perimeters of the reefs, particularly on windward
sides. Leeward margins may be made up of large amounts of sediment and
talus. In contrast, during this rapid growth period, reefs were still largely below
the rising sea level and 100% coral cover across the entire reef platform may
have been the norm. Much of the more fragile coral growth may have been
removed during cyclonic events but regrowth in such environments has been
shown to be extremely rapid.
The pattern for the inshore fringing reefs shows many similarities (e.g.
Hopley et al., 1983; Hopley & Barnes, 1985; Johnson & Risk, 1987; Partain &
Hopley, 1989; Kleypas, 1991) although growing in shallow water and
therefore inundated much later than the outer shelf foundations. Removal of
regolith does not seem to have retarded reef growth in many areas. Indications
are of almost immediate colonisation once sea level had risen sufficiently. By
definition, these reefs would still have experienced terrestrial influences, but in
contrast to the outer reef, which had been in similar situations 2000 years or
more earlier, larval recruitment for the fringing reef would not have been a
problem. However, common occurrence of alcyonarian spiculite (Konishi,
1982) most probably formed from the soft coral Sinularia in the basal section
of the Holocene in many reefs of the Cumberland and Northumberland Group
and also in the Palm Islands may well have been the result of shallower and
 Continental shelf reef systems 325
more turbid inshore conditions between 8000 and 6500 years BP (e.g. Kleypas,
1991).
Another important environmental factor during this period was the much
lower amount of protection given to the mainland coastline by reefs, many of
which were several metres below sea level. Today, most oceanic swells are
excluded by the reefal barrier along much of the north and central Queensland
coastline and fetch during both normal weather conditions and tropical
cyclones is limited by the reefal barrier. These influences would have been far
less during the final part of the transgression and even during the next phase
the mainland coastline may have experienced far higher energy conditions
than at the present time. This has been referred to as the Holocene 'high energy
window' (Hopley, 1984) and may have also been one of the factors promoting
reef growth. Far better circulation of inshore waters may have resulted and, in
particular, widespread settlement of extensive mud areas close inshore may
have been restricted.

3. 6500 years to 3000 years BP: stillstand and lateral growth

Between 6500 years and 3000 years BP the majority of reef reached modern sea
level. A slowing down of reef growth within the zone affected by stormwave
activity has long been recognised (Davies & Marshall, 1979) and much of the
carbonate productivity of this period has gone into detrital accumulation rather
than main framework construction. Not surprisingly, as Kleypas (1991) has
noted, most I4C date reversals in coral cores are found within this period.
After stillstand the gross morphology of reefs was changing rapidly. As
reefs reached sea level reef flat construction commenced and maximum
carbonate productivity of 10 kg m-2a-' became limited to the reef perimeters.
Outer reef flats now produced at about 4 k g m P 2 a - ' and much of the inner
sandy reef flats and lagoons may have been reduced to 0.5 kgm-2a-'. As a
result, reefs became much more complex ecological systems, with zones and
habitats determined by energy gradients from windward to leeward coming
into existence. For the first time in the Holocene, extensive reef flats were
available for the accumulation of sediments and the coral cays of the Great
Barrier Reef came into existence during this period.
Although reef growth was predominantly on the leeward sides as calcium
carbonate produced on windward margins was transported to the leeside,
leeside framework construction became much more limited. There is
considerable seismic evidence which shows that flourishing reef patches to the
leeward side of reefs became rapidly buried by sediment once the adjacent reef
structure reached sea level and this horizontal transport of carbonate
326 D. Hopley
commenced, a process which Goreau & Land (1974) have termed the
endogenous control of reef growth.
A further factor which aided in the emergence of reef flats and the
production of reef flat sediments, at least on reefs of the inner shelf, was a fall
in sea level of 1.0 to 1.5 metres (Chappell et al., 1983; Hopley, 1983a,b).
Attributed to hydroisostatic processes, this relative emergence has affected the
inshore fringing reefs in particular leaving inner reef flats still within the
modern tidal range, but with numerous dead microatolls. Subsequent reef
growth strategy has relied on framework construction on the reef front only
aided at times by detrital accumulation (e.g. Hopley et al., 1983). Retarded
growth becomes even more prominent close to the mainland and has been
particularly noted on the Cape Tribulation reefs (Fig. 8.8) along the mainland
north of Cairns (Partain & Hopley, 1989). The slight fall in relative sea level is
partly responsible but there are other factors. By 3000 years BP the stability of
sea level had meant that the position of the mainland coastline had been more
or less the same for 3500 years and the closing of the Holocene high energy
window had allowed the accumulation of a significant part of the nearshore
mud-silt wedge. Serious decline in water-quality conditions due to re-
suspension of sediments now took place whenever wind speeds were sufficient
to produce significant wave activity (circa 15-20knots). In the very high tidal
range area of the Cumberland and Northumberland Islands, tidal motion alone
became sufficient to maintain highly turbid waters almost continuously
(Kleypas, 1991).
The climatic record, at least from north Queensland, also suggests that this
was the period of maximum rainfall during the Holocene, with rainfall totals
on the Atherton Tableland being approximately 3500 mm, or more than double
what they had been some 4000 years previously. This is recorded in the change
in the vegetation to a complex mesophyll vine forest, but from a geomorpho-
logical viewpoint this would have given no more protection than the earlier
notophyll vine forest and it is likely therefore that there was greater run-off and
possibly even higher sediment yield during this period.

4.3000 years BP to 200 years BP: a reef system under increasing natural stress
All the trends of declining water quality noted in the previous 3500 years were
intensified during the millennia leading up to the present time. Declining
energy conditions with increasing reef barrier development accompanied the
expansion of nearshore mud wedges and increases in turbidity. Reef flat
continued to expand and reef framework development became even more
restricted to reef margins.
 Continental shelf reef systems 327
Olrtance In metrar
0 100 200 300

EMMAGEN R E E F LEGEND: REEF UNITS

E-2 E-1 MHWS (2.2) Frarnestone/Baundstone Unit

n
-
-coralllna a l g a
insitu coral
-ancNlt*ra/borarr
$
Detrital Unit
539W60 - rand. loosa Nbbla
535Oi50
c Basal Framestone Unit
0
p 628W80 - in situ coral
Fan Gravel Unit
- alluvlal cobblaa

6 8 6 M 60 6 8 0 M 1 10
RYKERS REEF R-4 R-3 R-2 R-’ MHWS

f
d

778W 260 708W 50 722M 50

SOUTH M Y A L L REEF
615W70 683W 70 MHWS
M - 2 M-3
LEGEND OF CORES
0 Coralllne algae-
e dominated Ilmartone
E Algal-camentad
-
C
rubbla
Loore rubbla
f
0
a Sand

in situ Porrtes
and Acroporld
Cavlly

667W 66 73301: 60

Figure 8.8. Age and lithology of the Cape Tribulation fringing reefs, far north
Queensland. MHWS = Mean High Water Springs.

Accompanying the accumulation of fine sediments in the nearshore zone

has been the increasing storage of nutrients in sinks in the natural system.
These include the mangroves of expanding estuarine and deltaic systems on
the mainland and lagoonal deposits within the reefs themselves. Release of
these nutrients is possible during occasional high-energy events. This was
328 D. Hopley
illustrated during cyclone Winifred in 1986 when disturbance of lagoonal
deposits resulted in a short-lived period of algal blooms (Furnas & Mitchell,
1986). Short periods of acute nutrient stress may therefore be superimposed on
a regime in which water quality under natural conditions, though not reaching
chronic stress proportions, is certainly declining.
These conditions are most pertinent to the inshore reefs. However, similar
though less-intensive trends may also have taken place on the outer shelf. The
outer barrier became fully established and morphological conditions returned
to those which initiated the early growth of Halimeda banks during the first
part of the Holocene. Certainly, living Halimeda is found over most of the
present banks and productivity levels appear to be high (Drew & Abel, 1988).
Research has clearly shown that cold nutrient-enriched water enters passages
from the Coral Sea and is transported to the areas in which the banks are
currently found. Although the water is only enriched from the normal depleted
level of 0.07 p M to about 0.25 p M nitrate with very little ammonia present, it is
suggested that even this input on three days of spring tides every two weeks
could still satisfy the entire annual nitrogen requirements of this very productive
algal vegetation. (Thomson & Wolanski, 1984; Wolanski, Drew & Abel, 1987;
Wolanski et al., 1988). Moreover, geochemical and geomechanical signals
which have been correlated with nutrient increases in coral skeletons clearly
show a variability which has been linked to the El Nifio phenomenon, which is
also considered to be a driving force in this nutrient upwelling (Rasmussen et
al., 1994a). Whether or not this nutrient enhancement is sufficient to affect reef
communities is debatable. Reef communities of the middle shelf are the most
prolific and most diverse and although this has been attributed to other factors
such as energy levels, it is the middle-shelf area which is best buffered from
the terrigenous influences to the west and the upwelling phenomena to the east.

5.200 years BP to the present: an anthropogenic overlay

Although there was an Aboriginal presence in the Great Barrier Reef region
throughout the Holocene and impact on the landscape may have been
significant through seasonal firing of the vegetation, the impact on the Great
Barrier Reef was almost certainly minimal. (However, a recent media report on
an Ocean Drilling Programme core taken just off Cairns suggested that an
increase in charcoal in the sediments, immediately off the Great Barrier Reef,
during the last 100000 years was related to Aboriginal settlement.)
The early exploitation of Great Barrier Reef resources and the settlement by
Europeans over the last 200 years has had a widespread impact, as reviewed by
Hopley (1988, 1989b). These impacts range from direct, such as the mining of
guano from reef islands and damage to reefs by anchor chains, to indirect, such
 Continental shelf reef systems 329
as the enhancement of nutrients close to sewage outfalls. It has been generally
regarded that because of its size, most of these impacts are highly local and that
the general health of the reef, particularly under the protection given since
1975 by the Great Barrier Reef Marine Park Act has left the Great Barrier Reef
in a better state than most other reef systems in the world.
Whilst this is probably so, anthropogenic activities have still tended to
aggravate the natural chronic stresses which have built up over the reef over
the last 3000 years. Land clearance and development of unsealed roads close to
fringing reefs, such as those of Cape Tribulation, have clearly increased
sediment yield. In the example of Cape Tribulation reefs north of Cairns, the
sediment yield increase in the adjacent streams may be as much as 14 fold due
to roadworks (Hopley et al., 1993). However, the increasing stress levels of the
late Holocene may have acclimatised the reefs to naturally occurring high
turbidity. Sedimentation levels which can be tolerated by Great Barrier Reef
corals (Table 8.1) appear far higher than those reported elsewhere in the world
(Pastorok & Bilyard, 1985).
There are increasing signs of an impact on widespread areas of the Great
Barrier Reef. Damage to the reef as the result of at least three major outbreaks
of the coral-eating crown-of-thorns starfish over the last 30 years has been
attributed by some to anthropogenic causes. Whilst this is yet to be proven, it
is still a possibility and research into the water-quality record retained by the
coral skeleton itself in the annual growth rings laid down in the colonies
suggests a drastic decline in water quality over the last 30 years on the inshore
reefs within the reach of river plumes in the Cairns area, an area of major
agricultural activity and fertiliser application. (Rasmussen, Cuff & Hopley,
1994b). Such results appear to confirm anecdotal accounts of reef deterioration
during the twentieth century, as for example, suggested by Sir Maurice Yonge
for Low Isles between his visits in 1928/29 as leader of the Royal Society
Expedition and his return to the area during the 1970s. The effects of the
decline of water quality on the Great Barrier Reef from sources such as
agricultural fertilisers and sewage have been the focus of great debate in
Australia in recent years. One result have been a major $1 million research
programme initiated by the Great Barrier Reef Marine Park Authority in 1991
into water-quality problems.
The Great Barrier Reef over the last 3000 years has clearly been growing in
what may be considered deteriorating water-quality conditions. Just as clearly,
there is a record of degradation of these water-quality conditions over the last
30 years or more as the result of human activities. Thresholds at which reef
systems can change direction very quickly (Buddemeier & Hopley, 1988) may
not yet have been reached, but there are clear signals that they may be
approaching.
330 D. Hopley
Over the next century or so effects of global environmental change as the
result of the Greenhouse Effect may be superimposed over these recent trends.
The effects for coral reefs may not be as drastic as those for other ecosystems
(Hopley & Kinsey, 1988; Kinsey & Hopley, 1991; Buddemeier & Smith,
1988). A slight rise in sea level and increase in storminess as predicted for
Great Barrier Reef waters may lead to renewed coral growth and greater
flushing of lagoon and other nutrient sink areas, producing conditions which
are analogous to those of the latter part of the postglacial marine transgression
7000 years ago. Such changes ultimately may do much to counteract the
negative human impacts of the last century.

Worldwide comparisons
There is no single major shelf reef system which has available as
comprehensive a history as that for the Great Barrier Reef. However, there is
sufficient evidence from the Caribbean to suggest that changing water-quality
conditions during the Holocene transgression have similarly been responsible
for many fossilised features of the continental shelves (e.g. Macintyre, 1967,
1972, 1988; Macintyre & Milliman, 1970; Adey et al., 1977; Lighty 1977;
Lighty, Macintyre & Stuckenrath, 1978; Neumann & Macintyre, 1985;
Fairbanks, 1989, and Macintyre er al., 1991).
Macintyre (1988) summarised four stages of development of eastern
Caribbean reefs as follows:
a) Pleistocene shelves were exposed to subaerial weathering and a soil zone developed
on their surfaces.
b) With rising seas, fringing reefs became established along the upper slopes or shelf
edges. In some areas, fringing reefs developed into barrier reefs with associated
lagoons.
c) The continuing rising seas that flooded broad areas of the shelves caused widespread
erosion of soil and lagoon deposits. The resulting high turbidity and high nutrient
levels terminated reef growth on the outer edges of these shelves. In northerly areas,
such as the east coast of Florida, the intermittent cooling of shallow-shelf waters
could also have led to the demise of shelf-edge reefs.
d) By the time water conditions returned to normal, the advancing seas were too deep
to allow shallow-water communities of Acropora palmafa to establish. The rapid
reef growth associated with these communities was thus transferred to shallow-
water substrates adjacent to coastlines. As a result, ‘modern barrier reefs are notably
absent in this part of the world’ (Macintyre, 1972, p.737). Instead, this area is
characterised by relict ‘give-up reefs’ (Neumann & Macintyre, 1985 [see also
Fig. 7.51) along the upper slopes and shelf edges and by relatively young (less than
7000 years BP) late Holocene reefs fringing most coastlines. The age limit of inner
shelf reef development is related to the time of flooding of these shallow areas and
has nothing to do with regional cooling related to continental glaciation.
 Continental shelf reef systems 33 1
Probably the most distinctive feature of this sequence is the demise of Late
Pleistocene and early Holocene shelf-edge reefs spanning depths of up to 80
metres and dated up to 12000 years BP. The death of these reefs has variously
been referred to ‘being shot in the back by your own lagoon’ (Neumann &
Macintyre, 1985) or ‘killed off by inimical bank waters’ (e.g., Schlager, 1981).
Hallock & Schlager (1986) attribute the demise of these reefs almost entirely
to eutrophication due to release of nutrients from a submerging shelf at a time
when sea level was still rising rapidly.
The eastern Caribbean and Great Barrier Reef appear to be different in two
major ways.
1. the pattern outlined above for the Great Barrier Reef did not incorporate
‘give-up’ shelf edge reefs; and
2 . the eastern Caribbean appears to have no major equivalent of the Halimedu
banks of the Great Barrier Reef.
However, deep shelf-edge reefs have been described along the entire length
of the Great Barrier Reef for over 65 years (Paradice, 1925; Hopley, 1982
pp.271-3; Harris & Davies, 1989). Depth ranges are between 40 and 150
metres (Fig. 8.9). Very little is known about these reefs, not even to the extent
of whether or not they support living deepwater coral communities and nothing

Figure 8.9. Seismic profiles of shelf-edge reefs off the Pompey Reefs, south central
section of the Great Barrier Reef.
332 D. Hopley

is known about their age or sea-level relationship. Nonetheless, their

ubiquitous presence and similar shelf-edge location to features in the eastern
Caribbean suggests that they could also be Late Pleistocene or early Holocene
in age and could have been largely retarded in their growth at the time when
the shelf was first flooded and Hulimeda banks began to accumulate.
The lack of Halimedu banks or equivalent structures in the Caribbean story
may also be more apparent than real. Hulimedu is present in the Caribbean and
some Hulimedu structures have been described, but the depth of the shelf in
relation to the different history of sea-level rise probably meant that conditions
of eutrophication were present for a relatively short period, perhaps less than
2000 years. This was sufficient to kill off shelf reefs, but insufficient for major
bioherm construction. Hine et al. (1988), however, have described Hulimedu
bioherms which form a nearly continuous band bordering the margins of the
Miskito Channel in the southwest Caribbean. Banks with a relief of 20-30
metres and depths of 40-50 metres appear to be built of largely dead Hulimeda
material. Their location and depth suggests that they could have been formed
at a time when bank flooding was first taking place and upwelling was also
present from the adjacent deep water. Similar deep bank edge Hulimedu banks
have also been described from the Java Sea (Phipps & Roberts, 1988; Roberts,
Aharon & Phipps, 1988). Banks have depths of 20-100 metres and upwelling
and nutrient overloading are suggested as explanations for the remarkable algal
growth at the expense of reef-building corals during early platform drowning.

A model for tropical carbonate accumulations during a

transgression-stillstand cycle (Fig. 8.10)
The data from the Great Barrier Reef and worldwide are sufficiently consistent
to suggest that there are predictable changes to water quality during a
transgression and stillstand period. Although the pattern could apply to any
transgression and stillstand period, the Late Pleistocene/Holocene stages are as
follows:
1. Late PleistoceneEarly Holocene
Oceanic waters suitable for coral growth rising against a steeply sloping
continental shelf leading to extensive narrow fringing reef development.
2. Early Holocene
Shelf flooding with rapid decline in water-quality conditions, an increase in
nutrients and turbidity in particular leading to a decline in reef growth
whilst sea level is still rising rapidly and, in suitable locations, the
development of algal bioherms at the expense of reef growth.
 Continental shelf reef systems 333

Anthropogenic input

Natural sediments. nutrients Natural sediments. nutrients

and fresh water and fresh water

SL SL

( V ) MODERN (IV) LATE HOLOCENE

Limited input from land

Upwelling water
Input from land
Reworking of roils,
r-lith, Gc.

SL

Cool nutrient
rich water
(111) MID HOLOCENE (11) EARLY HOLOCENE

Indicative water quality

High LOW
Limited
input from
land into Terrigenour sandlmud
oceanic water
Algal (Halimedal reef
SL Coral reef

Figure 8.10. A model for tropical carbonate accumulations during a transgression-

stillstand cycle. Water-quality conditions are closely linked to the sea-level stage. Note
that pre-existing reefal platforms are not illustrated in this schematic representation.
334 D. Hopley
3. Mid Holocene
A period of sea-level stabilisation, but at a level too high for regeneration of
the former shelf edge reefs. Water quality has improved and the major
phase of coral reef growth commences.
4. Late Holocene
A late Holocene phase in which reef growth on the innermost shelves
becomes restricted as terrigenous sediments build outwards from the
coastline (for a Caribbean example the Galeta Point reef, Panama shows
many features similar to the Cape Tribulation reefs of the Great Barrier
Reef (Macintyre & Glynn, 1976; Macintyre, 1988).
5. Modern
At a global scale, the anthropogenic overprint which has been recognised on
the Great Barrier Reef is even more distinctive. The decline of Caribbean
reefs in particular as a result of several centuries of human impact is well
documented (e.g. Brown, 1987; Lapointe, 1989; Shinn, 1989; Glynn 1990).
Many of these reefs are showing the classic, chronic stress symptoms as
described by Kinsey (1988), resulting in widespread bleaching and death at
times of acute stress such as produced by abnormally high temperatures
during recent El Nifio events.

ConcIusion
It is suggested that this morphogenetic approach to shallow-water marine
ecosystems will have the same advantages as a climatic or morphogenetic
approach in terrestrial geomorphology. Explanation is given to what may seem
anomalous features on continental shelves, whilst simultaneously these
features may provide insight into changing environmental conditions.
Applications are possible to the longer-term geological record. Changing
ecological patterns have been described at geological timescales (e.g., Copper,
1988). Whilst many large-scale changes are related to plate tectonics (for
example the movement of coral reef platforms on the Pacific plate northwards
out of the zone of modern reef growth to produce the drowned guyots of the
Emperor Seamount chain (Grigg & Hey, 1992)), other changes have been
related to changing water-quality conditions (Copper, 1989). These may be
cyclic, as for example the 100000 year ocean nutrient cycle occurring during
glacial and interglacial cycles as suggested by Wilson (1989) resulting in rapid
changes in low-latitude nutrient levels, or individual and catastrophic as the
Miocene phosphate spike (Riggs, 1984). Nonetheless, the sequence of events
described in this chapter appear to be logical and distinctive enough to warrant
research for similar patterns covering transgressive-stillstand periods in the
long-term geological record.
 Continental shelf reef systems 335
Acknowledgements
This paper draws on the work of many of my Honours and Postgraduate
students. In particular, I would like to acknowledge Bruce Partain, Cecily
Rasmussen, Joanie Kleypas and Rob van Woesik.

References
Adey, W.H., Macintyre, I.G., Stuckenrath, R. & Dill, R.F. (1977). Relict barrier reef
system of St. Croix: its implications with respect to late Cenozoic local reef
development in the western Atlantic. In Proceedings Third International Coral
Reef Symposium, Miami, vol. 2, pp. 15-2 I .
Bird, E.C.F. & Hopley, D. (1969). Geomorphological features on a humid tropical
sector of the Australian coast. Australian Geographical Studies, 7, 87-1 08.
Brown, B.E. (1987). Worldwide death of corals: natural cyclical events or man-made
pollution? Marine Pollution Bulletin 18.9-1 3.
Buddemeier, R.W. & Hopley, D. (1989). Turn-ons and turn-offs: causes and
mechanisms of the initiation and termination of coral reef growth. In
Proceedings Sixth International Coral Reef Symposium, Townsville, vol. I ,
pp. 253-62.
Buddemeier, R.W. & Smith, S.V. (1988). Coral reef research in an era of rapidly
rising sea level: predictions and suggestions for long-term research. Coral
Reefs, 7 , 5 1-6.
Chappell, J., Chivas, A., Wolanski, E., Polach, H.A. & Aharon, P. (1983). Holocene
palaeoenvironmental changes, central to north Great Barrier Reef, inner zone.
Bureau of Mineral Resources Journal of Australian Geology and Geophysics,
8,223-36.
Copper, P. (1988). Ecological succession in Phanerozoic reef ecosystems: is it real?
Palaios, 3, 136-52.
Copper, P. (1989.) Enigmas in Phanerozoic reef development. Memoirs of the
Association of American Palaeontologists, 8, 37 1-85.
Cortes, J.N. & Risk, M.J. (1985). A reef under siltation stress, Cahuita, Costa Rica.
Bulletin of Marine Science, 36, 339-56.
Davies, P.J. (1992). Origins of the Great Barrier Reef. Search, 23, 193-6.
Davies, P.J. & Hopley, D. (1983). Growth facies and growth rates of Holocene reefs
in the Great Barrier Reef. Bureau of Mineral Resources Journal of Australian
Geology and Geophysics, 8, 237-5 I .
Davies, P.J. & Marshall, J.F. (1979). Aspects of Holocene reef growth - substrate age
and accretion rate. Search, 10, 276-9.
Davies, P.J. & Marshall J.F. (1985). Halimeda bioherms, low energy reefs, northern
Great Barrier Reef. In Proceedings Fifk International Coral Reef Symposium,
Tahiti, vol. 5, pp. 1-8.
Davies, P.J., Marshall, J.F. & Hopley, D. (1985). Relationships between reef growth
and sea level in the Great Barrier Reef. In Proceedings Fifth International
Coral Reef Symposium, Tahiti, vol. 3, pp. 95-103.
Drew, E.A. & Abel, K.M. (1985). Biology, sedimentology and geography of the vast
inter reefal Halimeda meadows within the Great Barrier Reef province. In
Proceedings Fifth International Coral Reef Symposium, Tahiti, vol. 5 , pp.
15-20.
Drew, E.A. & Abel, K.M. (1988). Studies of Halimeda. I. The distribution and species
composition of Halimeda meadows throughout the Great Barrier Reef
province. Coral Reefs, 6, 195-205.
336 D.Hopley
Fairbanks, R.G. (1989). A 17,000 year glacio-eustatic sea level record: influence of
glacial melting rates on the Younger Dryas event and deep-ocean circulation.
Nature, 342, 637-42.
Furnas, M. & Mitchell, A. (1986). Oceanographic aspects of cyclone Winifred. Great
Barrier Reef Marine Park Authority workshop on offshore effects of cyclone
Winifred, pp. 41-2.
Glynn, P.W. (1990). Coral mortality and disturbances to coral reefs in the tropical
eastern Pacific. In Global ecological consequences of the 1982-83 El NiAo-
Southern Oscillation, ed. P.W. Glynn, pp. 52,55-126, Amsterdam: Elseiver
Oceanography Series.
Goreau, T.F. & Land, L.S. (1974). Fore-reef morphology and depositional processes,
North Jamaica. In Reefs in time and space, ed. L.F. Laporte, pp. 77-89. Society
of Economic Palaeontologists and Mineralogists Special Publication 18.
Grigg, R.W. & Hey, R. (1992). Palaeoceanography of the tropical eastern Pacific
Ocean. Science, 255, 172-8.
Hallock, P. & Schlager, W. (1986). Nutrient excess and the demise of coral reefs and
carbonate platforms. Palaios, 1, 389-98.
Harris, P.T. & Davies, P.J. (1989). Submerged reefs and terraces in the shelf edge of
the Great Barrier Reef, Australia: morphology, occurrence and implications for
reef evolution. Coral Reefs, 8, 87-98.
Hillis-Collinvaux, L. (1974). Productivity of the coral reef alga Halimeda (Order
Siphonales). In Proceedings Second International Coral Reef Symposium,
Brisbane, vol. 1, pp. 3 5 4 2 .
Hine, A.C., Hallock, P., Harris, M.W., Mullins, H.T., Belknap, D.F. & Jaap, W.C.
(1988). Halimeda bioherms along an open seaway: Miskito Channel,
Nicaraguan Rise, Southwest Caribbean Sea. Coral Reefs, 6, 173-8.
Hopley, D. (1982). Geomorphology of the Great Barrier Reef: Quaternary
development of coral reefs. New York: John Wiley-Interscience. 453 pp.
Hopley, D., ed. (1983a). Australian sea levels in the last 15 000 years. James Cook
University of North Queensland, Department of Geography Monograph Series
Occasional Paper, 3. 104 pp.
Hopley, D. (1983b). Deformation of the north Queensland continental shelf in the late
Quaternary. In Shorelines and isostasy, ed. D.E. Smith & A.G. Dawson,
pp. 347-66. London: Academic Press.
Hopley, D. (1984). The Holocene ‘high energy window’ on the central Great Barrier
Reef. In Coastal geomorphology in Australia, ed. B.G. Thom, pp. 135-50.
Sydney: Academic Press.
Hopley, D. (1985). Geomorphological development of modern coastlines: a review. In
Themes in geomorphology, ed. A.F. Pitty, pp. 56-71. London: Croom Helm.
Hopley, D. (1988). Anthropogenic influences on Australia’s Great Barrier Reef.
Australian Geographer, 1 9 , 2 6 4 5 .
Hopley, D. (1989a). Coral reefs: zonation, zonality and gradients. Essener
Geographische Arbeiten, 18,79-123.
Hopley, D. (1989b). The formation, use and management of the Great Barrier Reef.
Melbourne: Longman Cheshire. 54 pp.
Hopley, D. & Barnes, R.G. (1985). Structure and development of a windward fringing
reef, Orpheus Island, Palm Group, Great Barrier Reef. In Proceedings Fifth
International Coral Reef Symposium, Tahiti., vol. 3, pp. 141-6.
Hopley, D. & Kinsey, D.W. (1988). The effects of rapid short term sea level rise on
the Great Barrier Reef. In Greenhouse: planning for climatic change, ed. G.E.
Pearman, ppl89-201. Melbourne: CSIRO Division of Atmospheric Research.
 Continental shelf reef systems 337
Hopley, D., Parnell, K.E. & Isdale, P.J. (1989). The Great Barrier Reef Marine
Park: dimensions and regional patterns. Australian Geographical Studies, 27,
47-66.
Hopley, D., Slocombe, A.M., Muir, F. & Grant, C. (1983). Nearshore fringing reefs in
north Queensland. Coral Reefs, 1, 151-60.
Hopley, D., van Woesik, R., Hoyal, D.C.J.D., Rasmussen, C.E. & Steven A.D.L.
(1993). Sedimentation resulting from road development Cape Tribulation area.
Great Barrier Reef Marine Park Authority Technical Memorandum.
Hoyal, D.C.J.D. (1986). The effect of disturbed rainforest catchments on
sedimentation in an area of nearshore fringing reef Cape Tribulation, North
Queensland. Unpublished honours thesis, Department of Geography, James
Cook University of North Queensland.
Jennings, J.N. (1972). Some attributes of Torres Strait. In Bridge and barrier: the
natural and cultural history of Torres Strait, ed. D. Walker, pp. 29-38. ANU
Canberra: Research School of Pacific Studies Publication BG3.
Johnson, D.P. & Risk, M.J. (1987). Fringing reef growth on a terrigenous mud
foundation: Fantome Island, central Great Barrier Reef, Australia.
Sedimentology, 34,275-87.
Johnson, D.P. & Searle, D.E. (1984). Postglacial seismic stratigraphy, central Great
Barrier Reef province. Sedimentology, 31, 335-52.
Kershaw, A.P. (1978). Record of last interglacial-glacial cycle from northeastern
Queensland. Nature, 272, 159-6 1.
Kinsey, D.W. (1985.) Metabolism, calcification and carbon production. I. Systems
level studies. In Proceedings Fijth International Coral Reef Symposium, Tahiti,
VOI.4, pp. 505-26.
Kinsey, D.W. (1988). Coral reef system response to some natural and anthropogenic
stresses. Galaxea, 7, 1 13-28,
Kinsey, D.W. (1991). The coral reef an owner-built, high-density, fully serviced, self
sufficient housing estate - or is it? Symbiosis, 10, 1-22.
Kinsey, D.W. & Hopley, D. (1991). The significance of coral reefs as global carbon
sinks - response to Greenhouse. Palaeogeography Palaeoclimatology
Palaeoecology, 89, 1-15.
Kleypas, J.A. (1991). Geological development of fringing reefs in the southern Great
Barrier Reef, Australia. Unpublished PhD thesis, James Cook University of
North Queensland. 199 pp. + appendices.
Kleypas, J.A. & Hopley, D. (1994). Reef development across a broad continental
shelf, southern Great Barrier Reef, Australia. In Proceedings Seventh
International Coral Reef Symposium, Guam.
Konishi, K. (1982). Alcyonarian spiculite: limestone of soft corals. In Proceedings
Fourth International Coral Reef Symposium, Manilla, vol. I , pp. 643-9.
Lapointe, B.E. (1989). Caribbean coral reefs: are they becoming algal reefs? Sea
Frontiers, Mar-Apr, 1989, pp. 83-9 I .
Lighty, R.G. (1977). Relict shelf-edge Holocene coral reef southeast coast of Florida.
In Proceedings Third International Coral Reef Symposium, Miami, vol. 2, pp.
215-21.
Lighty, R.G., Macintyre, I.G. & Stuckenrath, R. (1978). Submerged early Holocene
barrier reef southeast Florida shelf. Nature, 276, 59-60.
Littler, M.M., Littler, D.S. & Lapointe, B.E. (1988). A comparison of nutrient and
light-limited photosynthesis in psammophytic versus epilothic forms of
Halimeda (Caulerpales, Halimedaceae) from the Bahamas. Coral Reefs, 6,
219-25.
338 D.Hopley
Macintyre, I.G. (1967). Submerged coral reefs, west coast of Barbados, W.I.
Canadian Journal Earth Sciences, 4,461-74.
Macintyre, I.G. (1972). Submerged reefs of the eastern Caribbean. American
Association of Petroleum Geologists Bulletin, 56, 720-38.
Macintyre, I.G. (1988). Modem coral reefs of the western Atlantic: new geological
perspective. American Association of Petroleum Geologists Bulletin, 72, 1360-9.
Macintyre, E.G. & Glynn, P. (1976). Evolution of modern Caribbean fringing reef,
Galeta Point, Panama. American Association of Petroleum Geologists Bulletin,
60, 1054-72.
Macintyre, I.G. & Milliman, J.D. (1970). Physiographic features on the outer shelf
and upper slope, Atlantic continental margin, southeastern United States.
Geological Society of America Bulletin, 81, 2577-2598.
Macintyre, I.G., Rutzler, K., Norris, J.N., Smith, K.P., Cairns, S.D., Bucher, K.E. &
Stenick, R.S. (1991). An early Holocene reef in the western Atlantic
submersible investigations of a deep relict reef off the west coast of Barbados,
W.I. Coral Reefs, 10, 167-74.
Mapstone, B.D., Choat, J.H., Cumming, R.L. & Oxley, W.G. (1989). The fringing
reefs of Magnetic Island: benthic biota and sedimentation: a baseline survey.
Unpublished report to Great Barrier Reef Marine Park Authority, 88 pp.
Maragos, J.E. (1972). A study of the ecology of Hawaiian reef corals. Unpublished
PhD thesis, University of Hawaii, Honolulu.
Marshall, J.F. (1988). Potential effects of oceanic deep waters on the initiation and
demise of coral reefs. In Proceedings Sixth International Coral Reef
Symposium, Townsville, vol. 3, pp. 509-12.
Marshall, J.F. & Davies, P.J. (1988). Halimeda bioherms of the northern Great Barrier
Reef. Coral Reefs, 314, 13943.
Marshall, S.M. & Orr, A.P. (1931). Sedimentation on Low Isles reef and its relation to
coral growth. Scientific Reports Great Barrier Reef Expedition, 1,94-133.
Maung Maung Aye (1976). Variation in drainage basin morphometry in north
Queensland. Unpublished MA thesis, James Cook University of North
Queensland, Townsville. 282 pp.
Neumann, A.C. & Macintyre, I.G. (1985). Reef response to sea level rise: keep-up,
catch-up or give-up. In Proceedings Fifth International Coral Reef Symposium,
Tahiti, vol. 3, 105-10.
Orme, G.R. (1985). The sedimentological importance of Halimeda in the development
of back reef lithofacies, northern Great Barrier Reef, Australia. In Proceedings
Fijih International Coral Reef Symposium, Tahiti, vol. 5 , pp. 3 1-8.
Orme, G.R., Flood, P.G. & Sargent, G.E.G. (1978). Sedimentation trends in the lee of
outer (ribbon) reefs, northern region of the Great Barrier Reef. Philosophical
Transactions Royal Society London, Ser. A., 291, 85-9.
Orme, G.R. & Salama, M.S. (1988). Form and seismic stratigraphy of Halimeda
banks in part of the northern Great Barrier Reef province. Coral Reefs, 6,
131-7.
Paradice, W.E.J. (1925). The pinnacle or mushroom-shaped coral growths in
connection with the reefs of the outer barrier. Reports of the Great Barrier Reef
Committee, 1,52-9.
Partain, B.R. & Hopley, D. (1989). Morphology and development of the Cape
Tribulation fringing reefs, Great Barrier Reef, Australia. Great Barrier Reef
Marine Park Authority Technical Memorandum 21,45 pp.
Pastorok, R.A. & Bilyard, G.R. (1985). Effects of sewage pollution on coral reef
communities. Marine Ecolology Progress Series, 21, 175-89.
 Continental shelf reef systems 339
Phipps, C.V.G., Davies, P.J. & Hopley, D. (1985). The morphology of Halimeda
banks behind the Great Barrier Reef east of Cooktown, Queensland. In
Proceedings Fifth International Coral Reef Symposium, Tahiti, vol. 5 , pp.
27-30.
Phipps, C.V.G. & Roberts, H.H. (1988). Seismic characteristics and accretion history
of Halimeda bioherms on Kalukalukuang Bank, eastern Java Sea (Indonesia).
Coral Reefs, 6, 149-59.
Pringle, A.W. (1986). Causes and effects of changes influvial sediment yield to the
northeast Queensland coast, Australia. Department of Geography, James Cook
University, Monograph Series Occasional Paper, 4.
Rasmussen, C., Cuff, C. & Hopley, D. (1994b). Evidence of anthropogenic disturbance
retained in the skeleton of massive corals from Australia's Great Barrier Reef. In
Proceedings Seventh International Coral Reef Symposium, Guam.
Rasmussen, C., Ness, S. & Cuff, C. (1994a). A correlation between the El
NiAo-Southern Oscillation and crystallographic variability in corals of the
Great Barrier Reef. In Proceedings Seventh International Coral Reef
Symposium, Guam.
Riggs, S.R. (1 984). Palaeoceanographic model of Neogene phosphorite deposition,
US Atlantic continental margin. Science, 223, 123-3 1.
Roberts, H.H., Aharon, P. & Phipps, C.V.G. (1988). Morphology and sedimentology
of Halimeda bioherms from the eastern Java Sea (Indonesia). Coral Reefs, 6,
161-72.
Schlager, W. (1981). The paradox of drowned reefs and carbonate platforms.
Geological Society American Bulletin, 92, 197-2 I I .
Searle, D.E. & Flood, P.G. (1988). Halimeda bioherms of the Swain Reefs, southern
Great Barrier Reef. In Proceedings Sixth International Coral Reef Symposium,
Townsville, vol. 3, pp. 139-44.
Searle, D.E., Harvey, N., Hopley, D. & Johnson, D.P. (1982). Significance of results
of shallow seismic research in the Great Barrier Reef province between
16'10's and 20'05'S. In Proceedings Fourth International Coral Reef
Symposium, Manilla, vol. 1, pp 532-9.
Shinn, E.A. (1989). What is really killing the corals? Sea Frontiers, Mar-Apr, 1989,
72-8 1.
Smith, S.V. & Jokiel, P.L. (1975). Water composition and biogeochemical gradients
in the Canton Atoll lagoon: budgets of nitrogen, phosphorus, carbon dioxide
and particulate materials. Marine Science Communications, 1, 165-207.
Stoddart, D.R. ( 1968). Climatic geomorphology: review and reassessment. Progress
in Geography, 1, 161-222.
Symonds, P.A., Davies, P.J. & Parisi, A. (1983). Structure and stratigraphy of the
central Great Barrier Reef. Bureau of Mineral Resources Journal of Australian
Geology and Geophysics, 8, 277-9 1.
Thomas, M.F. (1974). Tropical geomorphology: a study of weathering and landform
development in warm climates. London and Basingstoke: MacMillan, 332 pp.
Thomson, R.E. & Wolanski, E. (1984). Tidal period upwelling within Raine Island
entrance, Great Barrier Reef. Journal Marine Research, 42,787-808.
Veron, J.E.N. (1986). Corals of Australia and the Indo-Pacific. Sydney: Angus and
Robertson. 644 pp.
Wiebe, W.J., Crossland, C.J., Johannes, R.E., Rimmer, D.W. & Smith, S.V. (1982).
High latitude (Abrolhos Islands) reef community metabolism: what sets
latitudinal limits on coral reef development? In Proceedings Fourth
International Coral Reef Symposium, Manilla, vol. 1, pp. 721.
340 D. Hopley
Wilson, T.R.S. (1989). Climate change: possible influence of ocean upwelling and
nutrient concentration. Terra Nova, 1, 172-6.
van Woesik, R. (1992). Ecology of coral assemblages on continental islands in the
southern section of the Great Barrier Reef, Australia. Unpublished PhD thesis,
James Cook University of North Queensland. 226 pp.
Wolanski, E., Drew, E.A. & Abel, K.M. (1987). The Barrier Reef Halirneda banks -
why? Australian Marine Science Bulletin, 98, 10-1 1.
Wolanski, E., Drew, E.A. & Abel, K.M. and O’Brien, J. (1989). Tidal jets, nutrient
upwelling and their influence on the productivity of the alga Halirneda in the
Ribbon Reefs, Great Barrier Reef. Estmarine Coastal and Shelf Science, 26,
169-20 1.
 Arctic coastal plain shorelines
P.R. HILL, P.W. BARNES, A. HEQUETTE
AND M-H RUZ

Introduction
In this chapter we discuss shorelines which are loosely grouped under the
name ‘Arctic coastal plain shorelines’. These shorelines are formed in polar
regions where unconsolidated and permafrost-affected sediments form low
bluffs and where marine processes are affected by the seasonal presence of sea
ice. The emphasis will be exclusively on coastlines where relative sea level is
rising. Such coastlines are found along the perimeter of the Arctic Ocean -
along the Alaskan and Canadian Beaufort Sea coast and northeastern Siberia
(Fig. 9.1). Much of the remaining Arctic coastline is characterized by either

120E. 150E. I80 150W. @+

LAP T€V
SEA ARCT/C BANKS
ISLAND
OCEAN
BEAUFORT
SEA

€AS T SIBERAN
S€A
CANADA

CUUCKCU/ ALASKA
SEA
SIBERIA
ARCTIC
CIRCLE
Arctic Coastal Plains

Figure 9.1. The Arctic Ocean margin showing the approximate extent of Arctic coastal
plains in northwestern North America and eastern Siberia. These regions are
characterized by low tundra bluffs and rising relative sea level.

34 1
342 P.R. Hill et al.

bedrock cliffs or falling sea level (Taylor & McCann, 1983), and is excluded
from this chapter.
The broad aims of this chapter are to review the present understanding of
coastal evolution in Arctic coastal plain settings and to evaluate the importance
of truly Arctic (zonal) processes to this evolution. A synthesis of the important
zonal processes, using several examples from the Alaskan and Canadian
Beaufort Sea coasts, will be presented. We emphasise that this does not
represent all polar coasts and that, whereas the factors we have identified here
may be important in other polar regions, the range of possible climatic,
geologic and oceanographic conditions dictates that other factors may be
important elsewhere.
One of the most important considerations when evaluating the information
presented here is the problem of data acquisition in these remote regions.
Although adequate aerial photography is available over most of the Beaufort
Sea coastline, the distribution of ground surveys is relatively patchy, surveys
being conducted from either remote field camps, small boats or aircraft.
Measurement of marine processes, such as waves and currents, is extremely
limited in both duration and geographic distribution, the sea ice making
equipment moorings particularly hazardous. Furthermore, there is a strong bias
towards data collected during the summer season. Although a few winter
studies have been carried out, studies during autumn and spring have been
extremely limited due to logistic difficulties during the freeze-up and break-up
of the sea-ice cover. Unfortunately, these periods may have considerable
significance in terms of potentially important coastal processes and the lack of
information is a major limitation to the understanding of Arctic tundra coasts.

Important zonal factors

The concept of ‘zonality’ as it relates to latitudinal or climatic variation of
coastal geomorphology has been discussed by numerous authors (including
Guilcher, 1953; Davies, 1972; Kelletat, 1989). This concept recognizes that
whereas some coastal processes, such as tides and currents, do not show a
marked variation with latitude or climatic zones, others are strongly related to
distinct latitudinal zones. Coastal processes associated with extremely cold
conditions provide probably the clearest example of zonality, being restricted
to polar latitudes (Kelletat, 1989). Several zonal factors characterize Arctic
coastal plain shorelines and have a potentially important influence on coastal
evolution. These can be grouped under the two major headings: permafrost and
sea-ice phenomena (Fig. 9.2).
 Arctic coastal plain shorelines 343

LANDFAST ICE
GROUNDED PRESSURE
BOTTOMFAST RIDGES OF SHEAR PERMANENT PACK ICE
ICE OR "STAMUKHI" ZONE

20-30m

FROZEN, WELL-BONDED SEDIMENT

FROZEN,POORLY-BONDED SEDIMENT

UNFROZEN SEDIMENT

Figure 9.2. Schematic illustration of sea-ice zonation and permafrost distribution. First
year sea ice in the landfast ice zone reaches up to 2 m in thickness, except in regions
where pressure ridging occurs. Multi-year ice of the permanent pack is thicker,
averaging 3 m. The representation of permafrost is based on studies from the Canadian
Beaufort Sea (Dallimore et af.,1988; Blasco et al., 1990). The coastal zone represents
a transition zone from continuous permafrost to patchy sub-sea permafrost.

Permafrost phenomena
Permafrost is ubiquitous throughout northern Siberia, Canada and Alaska. On
land, permafrost is aggradational, the depth of frozen ground corresponding to
the depth of the 0°C isotherm, which may reach 700 m depth in Wisconsinan
or older sediments (Burgess et al. 1982; Judge, 1986; Lachenbruch et al.,
1987). In terrestrial Holocene sediments such as those of the Mackenzie Delta,
the thickness of permafrost may be considerably less than 100 m (Smith, 1975)
due to the relatively short time period of permafrost aggradation. As summer
temperatures rise above O"C, a seasonally active layer develops in the top 1 to
2 m of unconsolidated sediments.
Permafrost is also present in continental shelf sediments offshore (Hunter &
Hobson, 1974; Neave & Sellman, 1984), but is presently degrading seaward of
the bottomfast ice zone, due to the warming effect of the sea. The 0 "C isotherm
occurs at up to 600 m depth below the seabed on the Canadian Beaufort Shelf,
but sediments containing saline pore water may not be frozen to this level
(Judge, 1986).
344 P.R. Hill et al.
Most of the perennially frozen ground of the Arctic coastal plains of Alaska
and Canada contains both interstitial pore ice and larger ground-ice bodies.
The quantity of both depends on the particular lithology and local processes of
groundwater migration (Mackay, 1971; Rampton, 1988a). Ground ice is
present in several forms, including veins, lenses and massive ice bodies
(Fig. 9 . 3 ~ )The
. melting of pore ice and ground ice in bluff faces results in mass
failure as ‘thaw flow-slides’ (Fig. 9.36; Harry, 1985) and as thermoerosional
falls along ice wedges (Barnes, Rawlinson & Reimnitz, 1988). These
processes result in very rapid local and short-term bluff retreat. Certain sites
monitored along the Canadian Beaufort Sea coast have retreated 23 m in two
years (Hill et al., 1990). Average retreat rates range from 1-2ma-’ to over
15ma-’ (Reimnitz, Graves & Barnes, 1988; Harper, 1990). The products of
these cliff failures are deposited temporarily on the beach and shoreface
(Fig. 9.3). Depending on the efficiency of shoreface erosion and sediment
transport, the bluff may stabilize for several years until erosion again causes
the cliff to become thermally and gravitationally unstable.
Thermally triggered slumping also contributes to the development of
thermokarst topography, a morphology typical of Arctic coastal plains
(Czudek & Demek, 1970; Rampton, 1988b). Caused initially by the presence
of standing water, this morphology is formed by continued ground ice melting
and development of an unfrozen ‘talik’ beneath the standing water. The pond
gradually expands in size due to preferential consolidation of the unfrozen
sediments, assisted by thaw flow-slide activity along the margins. Eventually,
the terrain becomes dominated by numerous lakes, commonly several
kilometres in diameter and a few metres deep. Rampton (1988b) has suggested
that the development of thermokarst topography in northern Canada may have
originated around 9000 to 10 000 years BP during climatic optimum conditions.
Thermokarst topography has a considerable influence on the development of
shoreline morphology in Arctic coastal plain settings, through the
development of narrow headlands and broad embayments (Wiseman et al.,
1973; Ruz, Htquette & Hill, 1992). This provides a particular form of
antecedent morphology to a shoreline retreating under rising relative sea level,
although its importance to coastal evolution has been questioned (Reimnitz,
et al., 1988).
A less-understood phenomenon related to permafrost is the subsidence
associated with degrading permafrost at the coastline. In a general sense, the
shoreline and nearshore zone represent the boundary between regions of
aggrading and degrading permafrost. Where sediments remain frozen, the rate
of consolidation, and therefore subsidence, may be reduced because the pore
 Arctic coastal plain shorelines 345

Figure 9.3. ( a ) Typical fresh exposure of ground ice in low tundra bluff, North Point,
NWT, Canada. Stake on the left is approximately 1 m high. (b) Extensive thermal
erosion of ice-rich cliffs, Hooper Island, NWT, Canada. Thaw flow-slides transport
cliff material onto the beach, often forming an extensive bench on the upper foreshore.
Beach width is approximately 20m.
346 P.R. Hill et al.
space would be filled or partially filled with ice (Sellman, Brown & Lewellen,
1975; Reimnitz et al., 1988). As sediments thaw, consolidation may accelerate
as the pore spaces collapse and water is expelled. In detail, this phenomenon is
complex, depending on the thermal regime, the sediment properties (thermal
conductivity, porosity and permeability), the presence of massive ice,
sedimentation rates and the rate of coastal retreat (Dyke, 1991).
The thermal profile of an Arctic shoreface depends principally on the
overlying air or water temperature. Where sea ice is frozen to the bottom, the
seabed experiences winter temperatures below 0 "C because heat is conducted
to the atmosphere by the sea ice. In the Beaufort Sea, the winter thickness of
sea ice reaches between 1.5 and 2.0m. Thus, much of the shoreface and
nearshore zone may be seasonally frozen. Dallimore, Kurfurst & Hunter
(1988) have shown that the thermal profiles and permafrost conditions of two
sections of coastline within 5 km of each other can be considerably different
(Fig. 9.4). Although both sections were characterized by a gently sloping
shoreface underlain by ice-bonded sediment, the permafrost table dipped
seaward more steeply at the site with the steeper nearshore profile than at the
site with a gently sloping nearshore profile. This suggests that the depth of
permafrost in the nearshore depends on the length of time the seabed is
exposed to low winter temperatures through freezing of the sea ice to the
seabed and the temperature of coastal water in the summer (Dyke, 1991). In the
Canadian Beaufort Sea, summer near-bottom water temperatures range from
0 to 11 "C due to the influence of Mackenzie River water and transgressed
permafrost will tend to thaw. The thaw will be intermittent within the
bottomfast ice zone and continuous seaward of it.
Whereas the thaw of frozen sediment may result in delayed but predictable
subsidence, the thaw of massive ground ice in transgressed sediments may
cause rapid subsidence. Dallimore et al. (1988) calculated that the melting of a
massive ice body discovered in the drilling program described above could
result in the seabed settling locally by as much as 10m. Depressions in the
transgressive unconformity surface may be explained by such a process (Fig. 9.4).

Sea ice phenomena

The Arctic Ocean is characterized by a permanent ice pack, consisting of
aggregates of multi-year ice and rare fragments of glacier ice (ice islands). The
modal thickness of the ice pack is approximately 3 m, with pressure ridge keels
and ice islands reaching 45 m (Untersteiner, 1990; Hobson et al., 1989). The
margins of the ice pack consist of thinner 'first year' ice formed during freeze-
up of a single year. This ice is generally less than 2 m thick, but may be thicker
 Arctic coastal plain shorelines 341
0 200 400 600 800 1000 m

GROUNDING
LINE SEA L E V E L
0

40

-200 0 200 400 600 800 1000 1200 1400m-

GROUNDING
LINE

v
h

-10 g
-20
a
-30
-40

-
UNFROZE N FROZEN FROZEN WELL BONDED MASSIVE ICE
POORLY BONDED WITH EXCESS ICE
--- CONTACT HOLOCENE UNCONFORMITY

Figure 9.4. Thermal conditions of gently sloping (top) and more steeply sloping
(bottom) nearshore zones, off western Richards Island, NWT, Canada. Vertical lines
represent borehole penetration. Note the more steeply dipping permafrost table in the
lower diagram, associated with the narrower bottomfast ice zone. (Modified from
Dallimore et al., 1988.)

locally where ridging occurs. Along most sections of the Arctic Ocean coast,
open water is present during the summer. The extent of open water depends on
both atmospheric forcing and the resultant oceanic circulation (Mysak &
Manak, 1989), as well as on local conditions such as wind stress or river
discharge. As a consequence, it is highly variable from year to year (Fissel,
Birch & Melling, 1990).
The presence of sea ice is an important control on the fetch and consequent
wave regime experienced by Arctic coastal plain shorelines (Hill et al., 1991).
The development of waves in ice-laden waters is poorly understood. The
presence of ice inhibits the generation of waves, but it is not clear under what
348 P.R. Hill et al.
conditions a wave train can be formed. Pinchin, Nairn & Philpott, (1985)
achieved reasonable calibration of their wave hindcast model assuming that
stable wave trains form in water with less than 3/10 ice cover. In the Beaufort
Sea, the fetch distance changes throughout the summer, generally reaching a
maximum in late August. The most pronounced variability in fetch is
alongshore. Along the Alaskan and Yukon (western Canadian Beaufort Sea)
shorelines, the fetch distance rarely exceeds 30-40 km due to the presence
of grounded ice along the stamukhi zone (Reimnitz et al., 1988). However, in
the eastern Beaufort Sea, close to the influence of the Mackenzie River, the
fetch can reach up to several hundred kilometres. Where the fetch direction
coincides with predominant storm winds, relatively high wave conditions can
be generated. For most of the Beaufort Sea, long fetches to the northwest and
strong winds from the same quadrant result in the most extreme wave
conditions. Estimation of return periods for extreme waves, however, requires
evaluation of the joint probability of long fetches and extreme storms. Also,
the shallowness and flat profile of Arctic shelves means that substantial depth
limitation of waves would occur. Hodgins (1985), based on an evaluation of
five different hindcast models, estimated a deep-water significant wave height
of 7 to 9 m for a return period of 100 years in the Canadian Beaufort Sea.
The presence of sea ice in the water column may modify considerably the
processes of erosion, transport and deposition of sediment in the Arctic
nearshore. Ice may be present as large blocks, as small newly formed crystals
(frazil ice) or as a surface layer of slush ice, consisting of both frazil ice and
fragments of older sea ice (Reimnitz & Kempema, 1987; Reimnitz, Kempema
& Barnes, 1987).
Large blocks of sea ice may interact with the nearshore seabed through the
process of ice wallow. Wallowing is the term used to describe the motion of
grounded ice in waves (Reimnitz & Kempema, 1982). The wallowing of
grounded ice floes in 2 to 7 m water depth, within 1 km of the coast, forms
local, crater-like erosion surfaces with relief of 2 to 3 m and spacing of 50 to
100m (Reimnitz & Kempema, 1982). These features form preferentially in
granular sediments and repetitive surveys show that they are unstable over a
period of three years. These observations suggest that coastal processes rapidly
cause infill of wallow depressions. Ice wallow may nevertheless exert a
considerable influence on beach and nearshore dynamics.
Frazil ice forms during the autumn when turbulent coastal waters become
supercooled. The fine frazil crystals are characteristically sticky and are known
to adhere together, to the seabed, and to other objects (Martin, 1981).
Typically, frazil ice is generated during autumn storms, which mix and cool
the coastal water column (Reimnitz et al., 1987). The downward transport of
 Arctic coastal plain shorelines 349
frazil ice by turbulence causes the formation of anchor ice, where the seabed
becomes coated with a mass of frazil particles (Reimnitz et al., 1987). The
presence of anchor ice can have an armouring effect mitigating against
entrainment of sediment, but only during the storm which produces it. Heat
conducted from the seabed leads to the eventual release of the anchor ice which
floats up to form part of the developing ice canopy (Reimnitz et al., 1987).
This process can result in considerable erosion of sediment from the bottom
and its incorporation into the winter ice cover.
Before complete freezing of the ice canopy, there is a period of several
weeks where a layer of slush ice floats at the sea surface. Reimnitz &
Kempema (1987) have documented the characteristics of this layer on coastal
transects off Alaska. The slush ice effectively filters out high-frequency wave
oscillations and inhibits wave breaking. Nevertheless, in the free water below
the slush ice layer, Reimnitz & Kempema (1987) observed a rippled seabed,
with a wavelength similar to that of the observed waves in the slush, and
transport of cobble-sized clasts at 0.95 m depth, indicating that sediment
transport occurs on the shoreface under these conditions. The long-term effects
of freeze-up processes on beach and nearshore equilibrium profiles have not
been evaluated.
Apart from modifying hydrodynamic processes, the ice cover can have a
direct impact on the seabed. The phenomenon of ice scouring (or ice gouging
-terminology varies from country to country) has been well documented in the
Beaufort Sea (Fig. 9.5; Lewis, 1978; Barnes, Rearic & Reimnitz, 1984; Rearic,
Barnes & Reimnitz, 1990; Shapiro & Barnes, 1991). Ice scouring is caused by

Figure 9.5. Sidescan sonar image showing extensive ice-scour erosion of lower
shorefacehner shelf, off the Tuktoyaktuk Peninsula, NWT, Canada.
350 P.R. Hill et al.
the impact of ice ridge keels on the seabed. The highest intensity of seabed
scouring occurs on the seabed below the shear zone between the moving Arctic
pack ice and the more stable landfast ice (Barnes et al., 1987). This region
occurs in 15-40m water depth over most of the Beaufort Sea (Barnes et al.,
1984). Smaller pressure ridges and ice blocks cause scouring inside the main
shear zone, within the landfast ice, in shallow water (Shapiro & Barnes, 1991).
Frequent scouring also occurs in shallow water by the grounding of isolated ice
floes during the summer. HCquette & Barnes (1990) have shown that, at least
locally, ice-scour erosion of the nearshore seabed may change the equilibrium
profile and lead to coastal retreat (Fig.9.6). Ice scouring may bulldoze
sediments in a net onshore direction. Reimnitz, Barnes & Harper (1990)
calculated that a single scouring event moved 1-2 m3 m-' at 8 m water depth.
Two other sea-ice processes which have a potential impact on the coastal
zone are the phenomena of ice pile-up (Fig.9.7~)and ice ride-up (Barnes,
1982; Kovacs, 1983, 1984; Reimnitz et al., 1990). These result from the
thrusting of ice onto the shoreline in winter, ice pile-up occurring where the ice
buckles and fails, forming a pile of broken ice on the beach, and ice ride-up

-5

-E
v
I987
:
a
-10 I971
W
(3

-15

ACCRETION
EROSION
-20
DISTANCE FROM 1987 COASTLINE (km)

Figure 9.6. Nearshore bathymetric profile showing erosion attributed to ice scour in
water depths of 12-15 m, Tuktoyaktuk Peninsula, NWT, Canada (from HCquette &
Barnes, 1990).
 Arctic coastal plain shorelines 35 1

Figure 9.7. ( a ) Ice pile-up ridge formed along the spit at Atkinson Point, Tuktoyaktuk
Peninsula, NWT, Canada. Note 2 m figure on ridge showing scale (photograph by D. F.
Dickins); ( b ) mounds of ice-pushed sediment resulting from an ice pile-up event,
Camden Bay, Alaska, USA.
352 P.R. Hill et al.

where a large sheet of ice slides smoothly up on to the shore, continuing for
hundreds of metres (Reimnitz et al., 1990). In some cases, ice has been
documented to override bluffs of several metres high (Kovacs, 1983, 1984)
and on a regional basis may occur every year. Ice pile-up and ride-up may
occur very quickly, the whole process taking 15-30 minutes (Reimnitz et al.,
1990). Subsequent melt-out of this ice often leaves a substantial deposit of
sediment on the beach (Fig. 9.7b) and/or bluff (Barnes, 1982; Reimnitz et al.,
1990).

Evolution of Arctic coastal plain shorelines

When considering the morphodynamic evolution of Arctic coastal plain
shorelines, it is useful to consider two end-member cases: 1. those shorelines
whose evolution is strongly influenced by zonal (i.e. Arctic) processes; and 2.
those shorelines where zonal processes are subordinate to the normal (azonal)
processes of wave action. Obviously, the importance of both zonal and azonal
processes along any one section of shoreline varies with time, both seasonally
and at longer time scales. It is impossible to define any one region to be
exclusively controlled by one or other type of process. Nevertheless, the two
examples presented below are thought to represent as far as possible these two
end-member cases.

Shorelines strongly influenced by zonal processes: Alaskan Beaufort Sea

Background
The Alaskan Beaufort Sea (Fig. 9.8) is almost completely ice-covered for nine
months of the year (October to June). Even during the summer months, when
some open water exists along the coast, the maximum fetch is rarely less than
30 to 40 km and drifting ice commonly reduces the effective fetch (Barnes
et al., 1988). A mobile ice belt is also commonly stranded on the shoreface for
extended periods. As a result, wave heights are generally less than 0.5 m with
periods less than 3 s throughout the open-water season (Short, Coleman &
Wright, 1974) and the impact of waves on the shoreface is intermittent and
modified by the presence of ice. Higher waves are observed in coastal lagoons
than in the open sea (Reimnitz et al., 1988). Storm waves reach 1 to 3 m in
height, with periods of 5 to lOs, the strongest winds and longest fetches
coming from the northwest (Short et al., 1974; Reimnitz et al., 1988). Extreme
wave heights estimated at 3 m were observed during a one-in-25-year storm in
September 1970, associated with a storm surge which left driftwood at
elevations as high as 3.4m above mean sea level (Reimnitz & Maurer, 1979).
 Arctic coastal plain shorelines 353

BEAUFORT SEA
Pt Barrow

Drew Pt

Gape
Halkett Pingok Island
Smith Bay Oliktok Pt
Cross Island
Floxman
Island
Harrison Bay

Comden Bay
Prudhoe Demarcafioq
Bay Boy
B ~ ~ ~ u S RANGE

Figure 9.8. The Alaskan Beaufort coastline, showing the principal place names
mentioned in the text. Note the presence of offshore barrier islands along much of the
coast. (Modified after Barnes et al., 1988.)

Because the maximum tidal range is less than 0.15 m, water-level fluctuations
at the coast are influenced strongly by meteorological events, with positive and
negative surges causing as much as 1 m of additional amplitude (Short et al.,
1974).
Sea level has risen in the Alaskan Beaufort shelf region from a late-
Wisconsinan minimum of 96 to 116m (Dinter, Carter & Brigham-Grette,
1990). In the absence of reliable tide gauge data, the present rate of relative
sea-level rise is thought to be close to the eustatic rate (1.0 to 1.5 m m a - ' ;
Gornitz, Lebedeff & Hansen, 1982; Reimnitz et al., 1988). Holocene tectonic
movements are not thought to be significant contributors to relative sea-level
changes because older (c. 120000 years BP) shoreline deposits are not
deformed and have a consistent elevation across the region (Reimnitz et al.,
1988).

Geornorphology
Most of the Alaskan coastal plain is made up of unconsolidated and ice-
bearing Quaternary deposits of the Gubik Formation (Carter, Brigham-Grette
& Hopkins, 1986), overlain by late Quaternary aeolian deposits and Holocene
lacustrine sediments (Barnes et al., 1988). The coastal plain is characterized by
thermokarst topography in the form of patterned ground, thaw lakes and
pingoes. Several major rivers, with sources in the Brooks Range, form alluvial
fans that cross the coastal plain creating large deltas (Fig. 9.8). At the coast, all
these deposits outcrop as low bluffs, generally 1 to 3 m high, but locally up to
354 P.R. Hill et al.
10m, fronted by narrow sand and gravel beaches. Along approximately half of
the Alaskan coast, barrier islands of coarse sand and gravel are present several
kilometres offshore, having subaerial relief of 1 to 2 m and separated by inlets
up to 5 m deep. These island chains are in places anchored by low tundra
islands and are separated from the coast by shallow (1 to 3 m ) lagoons
(Fig. 9.8).
Nearshore morphology varies considerably along the coast. Seaward of the
barrier islands well-developed bar systems with crests 3 to 4 m below the sea
surface are developed in water depths up to 7 m (Fig. 9 . 9 ~Short
; et al., 1974;
Short, 1975). Three types of bar were identified by Short (1975): multiple
parallel bars, long parallel bars attached to the shore and short en echelon
(transverse) bars, also attached to the shore. The grain size characteristics of
these bars are not described by Short (1975). However, the shore-attached bars
join with sand and gravel beaches and Barnes et al. (1988) describe some
transverse bars as ‘clean sands’.
Bars are also present off certain sections of tundra bluffs (Short, 1975), but
in other regions where bars are not present the seabed slopes gradually to the
2 m isobath, forming a shallow-water bench (Fig. 9.9b). Reimnitz, Graves &
Barnes (1988) interpret this to be related to the presence of bottomfast ice

m ’
w 2
a NEARSHORE
L o
=z -2 BAR

I -4
-6
U
n -8
100 200 300 400 500 600
METRES

SHALLOW-WATER BENCH

I
I- -10
x -15
Q

5 10 15 20
K I LOMETRES
Figure 9.9. Typical seabed profiles from the Alaskan Beaufort Sea: ( a ) barred nearshore
off Pingok Island (from Short, 1975); ( b )non-barred coast showing the presence of the
2 m bench (from Reimnitz et al., 1988).
 Arctic coastal plain shorelines 355
during the winter. Superimposed on this bench is an irregular series of closed
depressions and mounds related to ice-wallow processes. These features have
relief of 2 to 3 m and are known to be rapidly formed and obliterated (Reimnitz
& Kempema, 1982). Off both barred and non-barred coasts, the inner shelf
shows a broadly concave profile to the 20 m isobath (Fig. 9.9). Just inside the
20 m isobath, a series of linear stamukhi shoals is present (Barnes et al., 1987).
Although the origin of these shoals is unclear, they are commonly the site of
sea-ice grounding, which forms the main constraint on fetch for wave
generation (Reimnitz & Kempema, 1984).

Coastal evolution
A geomorphologic model of coastal evolution for the Alaskan region was
proposed by Wiseman et al. (1973). In this model, the present distribution of
broad embayments, offshore islands and barrier islands develops from the
thermal collapse, breaching and drowning of thaw lakes (Fig. 9.10). Continued
transgression results in coalescing of breached lakes into lagoons, while
erosion transforms the intermediate headlands and islands into spit and barrier
island systems that eventually become isolated many kilometres from shore.
Reimnitz et al. (1988) have raised some objections to the Wiseman et al.
(1973) model, namely that coastal thaw lake depths are insufficient, many lake
bottoms are perched several metres above sea level and lake and valley
orientations are not always parallel to the shore as implied by Wiseman et al.
(1973). Ruz et al. (1992) have proposed a model similar to that of Wiseman
et al. (1973) for the geomorphologic evolution of the Canadian Beaufort Sea
coast, where lake and valley orientation is, if anything, predominantly
perpendicular to the shoreline. The orientation of coastal lagoons would be
controlled not only by the original size of the lake, but also by the extent of
erosion of the headlands formed after lake breaching. Thus, it seems that the
Wiseman et al. (1973) model provides a reasonable starting point for
understanding coastal evolution in the area. However, the model is only
applicable at a relatively large scale and says little about the dynamic processes
involved.
The bars present along much of the Alaskan coast attest to the importance
of wave action on the shoreline, despite the limited fetch and relatively low-
energy conditions (Short, 1975). The bar systems do not fit easily into classical
beach and nearshore categories (Greenwood & Davidson-Arnott, 1979;
Wright & Short, 1983) and there have been few measurements of wave
parameters to help classification. Wiseman et al. (1974) measured waves at
Pingok Island (Fig. 9.8) over a single summer season, including a storm with
wave heights of 1.5 to 2.5m and periods 9 to 10s. Spectral analysis showed
356 P.R. Hill et al.

Beuuforf Seo
Drew Point Lone Iy

Beuuforf Sea

Kogru River

Pingok Island
Beuuforf Seu
Simpson Lagoon
Oliktok Point

Barrier Island

0 5 10km

Figure 9.10. Selected coastal morphologies, representing the evolution of lagoons and
barrier islands, Alaskan Beaufort Sea coast (after Wiseman et al. 1973). The basis of
this model is the development of lagoons and barrier islands from thermokarst lakes and
headlands. Reimnitz ef al. (1988) have challenged this hypothesis.

significant infragravity wave energy and spectral peaks compatible with

standing wave frequencies predicted from nearshore bar spacing.
These shorelines appear to have the dual characteristics of an apparently
dissipative nearshore fronting a relatively coarse-grained, more reflective
beach. This differentiation, perhaps most often associated with macrotidal
coasts, may be here related to ‘meteorological’ tides, with the gravel beach
being active only during storm surges. The importance of wave motions during
freeze-up conditions and of sub-ice current flow during winter is not known.
The coarse sediment may be supplied by zonal processes such as ice push. This
 Arctic coastal plain shorelines 357

process has been observed to transport intermittently large volumes of

sediment onto barrier islands (Fig. 9.7; Reimnitz et al., 1988).
Although the antecedent topography in the Wiseman et al. (1973) model is
zonal in nature, being formed by degradation of permafrost, neither this model
nor the formation of nearshore bars described by Short (1975) and Wiseman et
al. (1974) require the short-term dynamic processes acting on this topography
to be uniquely Arctic. By contrast, Reimnitz et al. (1988), using a sediment
budget and equilibrium profile approach, have suggested that the long-term
evolution of these shorelines is strongly zonal in nature. These authors
calculated the sediment yield from bluff erosion, based on observed retreat
rates, bluff composition and ice content. They then assumed that the present
nearshore and inner-shelf profile represents an equilibrium profile and
calculated the sediment yield from erosion of the inner shelf. Comparison of
bathymetric profiles from 1952 to 1980 suggested erosion to approximately
5 km offshore (approximately 8 m water depth; Fig. 9.9).
The result of this analysis was that the sediment yield appears to be many
times larger than the available sediment sinks in either coastal lagoons or on
the shelf. Lagoons cannot act as permanent sediment sinks because they are
shallow and being transgressed. Seismic profiles suggest that the shelf is
predominantly erosional, except for a thin veneer, known as the ‘ice-keel
turbate’ layer, formed by the scouring action of ice (Barnes et al., 1984).
Reimnitz et al. (1988) concluded that this turbate layer may act as a temporary
storage but that, in order to maintain the equilibrium profile, sediment must be
transported off the shelf and eventually into the Arctic Ocean basin.
Reimnitz et al. (1 988) also point out that the rates of coastal erosion in the
Alaskan Beaufort Sea are high compared with those of other coastlines of
North America; comparable, for example, with those along the Gulf of Mexico
coast, despite the very limited wave energy available. This requires a very
efficient removal of sediment from the coastal zone and these authors conclude
that zonal processes must be important because wave energy is insufficient.
There is evidence that much of the erosion and transport of sediment occurs
during autumn storms, when frazil and anchor ice is forming. When these
events are intense, ice scours are completely obliterated on the inner shelf to a
distance of 15 km (13 m water depth; Barnes & Reimnitz, 1979), implying
extensive bottom transport of sediment. Much sediment is also incorporated
into the developing ice canopy as it becomes attached to frazil particles or is
lifted to the surface in anchor ice (Reimnitz et al., 1987). ThiS provides a
potential mechanism for transport of sediment across the shelf and into the
Arctic Ocean basin. Ice scouring may also contribute to sediment transport on
the inner shelf, although the predominant transport directions appear to be
alongshore or onshore (Rearic et al., 1990).
358 P.R. Hill et al.
Shorelines strongly influenced by azonal processes: Canadian Beaufort Sea
Background
The period of complete sea-ice cover in the Canadian Beaufort Sea is similar
to that of the Alaskan Beaufort Sea, lasting from October to June. However,
the extent of open water is commonly much greater in summer, the edge of the
polar ice pack ranging from a few kilometres to more than 300 km offshore.
Break-up of the ice cover is controlled largely by the outflow of the Mackenzie
River (Fig. 9.1 l), although the distribution of ice during the summer months is
controlled by more complex variables, including air temperature and wind
patterns. As a result, wave energy at the coast is determined largely by the
available fetch across open water and the wind direction. The wave regime is
characterized by wave heights and periods of mostly less than 4 m and 8 s
respectively (Pinchin et al., 1985). The strongest winds generally blow from
the northwest and fetches are commonly also longer in this direction, so that
wave energy is concentrated from this orientation. Easterly winds tend to be
lighter, with the result that the Tuktoyaktuk Peninsula (Fig. 9.11) and
northwest-facing coastlines experience the most severe wave attack.
Water-level elevations resulting from positive storm surges are extremely
important and have been studied using numerical models (Henry, 1975, 1984)
and by field observations (Harper, Henry & Stewart, 1988). Mapping of drift
log shoreline deposits in the Kugmallit Bay region (Fig. 9.1 1) suggests a
maximum storm surge level of 2.4 m above mean sea level.

Land over 200m elevation

Extent of Holocene
Delta deposits
Mojor spits and barrier islands

WOLFE
SPIT

KUGMALLIT
BAY

BRITISH RICHARDS
MTS IS1 ANO

KING
POINT
MAC KENZIE
- , A , , * - - - - . ,
DELTA
MlLilAMU3VIY
140° MTS

Figure 9.1 1. The Canadian Beaufort Sea coast, showing principal place names
mentioned in text. Barrier islands are absent along the coast west of the Mackenzie
Delta, but present elsewhere, particularly along the Tuktoyaktuk Peninsula.
 Arctic coastal plain shorelines 359
Under open water conditions, current circulation is primarily wind-driven
(Hill et al., 1991), currents correlating strongly with winds in coastal waters
(Fissel & Birch, 1984; Davidson, de Margerie & Lank, 1988). During periods
of northwesterly winds, there is a net easterly drift of surface water (MacNeil
& Garrett, 1975; Fissel & Birch, 1984). This pattern is reversed during periods
of easterly winds. However, the flow is complex and, particularly in coastal
waters, influenced by topography and water levels in coastal embayments
(Hill et al., 1990; HCquette & Hill, pers. commun.).
Relative sea level in the Canadian Beaufort Sea rose during the Holocene
from a late-Wisconsinan minimum of -70m (Forbes, 1980; Hill et al., 1985).
The present rate of relative rise is thought to be greater than the eustatic rate of
1.0 to l S m m a - ' , but recent work suggests that the rate does not exceed a
maximum of 2.5 mma-' (Hill, HCquette & Ruz, 1993).

Geomorphology
The coastal plain of the Canadian Beaufort Sea is also made up of Pleistocene
deposits, within which the Holocene Mackenzie Delta occupies a broad
erosional trough (Fig. 9.1 1; Rampton, 1982; 1988b). As on the Alaskan coastal
plain, thermokarst topography predominates, resulting in an irregular shoreline
of embayments and narrow headlands.
To the west of the Mackenzie Delta, the coastal plain is less than 50 km
wide. The deposits are very similar to those of the Alaskan coastal plain
(Carter et al., 1986), but the coastal morphology is very different. Bluffs are
relatively high (up to 50m) and often near-vertical, reflecting erosion by a
combination of thermal and wave processes. They are fringed with narrow
sand and gravel beaches and small embayments may be partially or completely
enclosed by gravel spits and barriers. Offshore barrier islands are absent. The
barrier beach at King Point appears to have developed within the last 200
years, beginning as a spit and now accreting seaward at its distal end (Hill,
1990; Fig. 9.12). The nearshore profile off King Point is relatively steep and
reflective, compared with profiles to either the west or the east (Hill et al.,
1986).
To the east of the Mackenzie Delta, the coastal plain widens and sandy
lithologies dominate (Rampton, 1988b). Most of the bluffs east of the delta are
low (4 m) and consist of well-sorted fine to medium sand, although units of
gravel-rich sediments do occur locally. Cliff-top coastal dunes are common
along much of this shoreline. The abundance of thermokarst lakes in this
region produces a highly indented coastline with narrow headlands separated
by broad embayments (Fig. 9.1 1). Several small tundra islands lie offshore of
Richards Island (Fig. 9.1 1). Headlands and islands are typically fringed by
sand spits, with crestal elevations less than 1.5 m above mean sea level and
360 P.R. Hill et al.

KING POINT
BEAUFORT SEA
9'

LAGOON

6 9'06"
0 500 m

Figure 9.12. Evolution of the King Point barrier beach, based on aerial photographs
from 1954 to 1984. Inset shows typical beach profiles across the barrier beach. The
barrier-beach developed from a southeastward prograding spit between 1954 and 1970.
Since the closing of the lagoon, accretion of the beach has occurred at the eastern end.
No inlets have been observed along this gravel-rich barrier in recent years. MSL = mean
sea level. (After Hill, 1990; Ruz et al., 1992.)

composed predominantly of fine to medium sand (Fig. 9.13). At Wolfe Spit,

where the adjacent headland supplies a proportion of gravel, a storm ridge with
a somewhat higher elevation of 2 m has developed (Fig. 9.13). Along the
Tuktoyaktuk Peninsula coast (Fig. 9.1 1 ), several long barrier island systems,
with crestal elevations less than 1 m are present and are migrating landward at
rates of 1 to 3 ma-' (Hill et al., 1990; Htquette & Ruz, 1991).
 Arctic coastal plain shorelines 36 1

520000 525000
WASHOVER CHANNEL
ACCRETED BERM (1974-1985)

SAND+GRAVEL
STORM RIDGE SAND BEACH
38 39
7735000-
40 41

INTERTIDAL
PLAT FORM

Figure 9.13. Wolfe Spit, Canadian Beaufort Sea, based on aerial photograph analysis.
Grid line spacing is 5 km. Profiles 38 to 41 are shown in Fig. 9.14. Note that the spit is
differentiated into a sand and gravel storm beach, a sand beach and an intertidal barred
platform, suggesting partition of sediment transport between storm surge and
fairweather conditions. (After Hill & Frobel 1991.)

Most of the sandy shorelines east of the Mackenzie Delta are characterized
by nearshore bars (Fig. 9.14). Parallel bars are the most common form visible
from aerial surveys (Harper 1990; Htquette & Ruz, 1991). Transverse bars are
found commonly on the foreshore, particularly along relatively sheltered low-
energy shorelines. Some low-lying spits and barrier islands, where overwash is
common, are made up of accreted transverse foreshore bars (Fig. 9.13).

Coastal evolution
At the geomorphic level, Ruz et al. (1 992) have developed a conceptual model
(Fig. 9.15) for evolution of the Canadian Beaufort coast. The model is similar
to that of Wiseman et al. (1973) in that the antecedent morphology created by
thermokarst processes favours the development of headland and spit
morphology. Continued erosion leaves isolated islands with flanking spits,
which are transformed eventually into true barrier islands when the tundra
island remnants are totally eroded. Deprived of a continued sediment supply
from alongshore, the barrier islands migrate landward as a result of continued
washover. Evidence from seismic records on the inner shelf suggests that at
least some barrier islands are eventually inundated and partially preserved on
the inner shelf within thermokarst depressions (Ruz et al., 1992; HCquette &
362 P.R. Hill et al.
2.0 Sand and Gravel
I .5
90-38
Accreted Sand
I .o Beach HHW
0.5 LLW
0.0
-0.5
-1 .o
2.0
90-39 Sand Beach N
I .5 Accreted Sond
I .o Beach
0.5
-- -0.5
E
0.0

-I .o V.E.172
2
I- 2.0
1.5 90-40 N
lnteri idal Platform
y 1.0 HHW
W 0.5 LLW
0.0
-0.5
-I .o V.E.=72
2.0
I .5 90-41 N
In v t i d a l PIC form
I .o Shore-Parallel Bars HHW
0.5 LLW
0.0
-0.5
-1.0 V. E .= 72
0 100 200 300 400 500 600 700 800 900 1000
DISTANCE (in)

Figure 9.14. Beach profiles from Wolfe Spit, Canadian Beaufort Sea. Note the presence
of several bars on the lower foreshore and the broad intertidal platform. Location of
profiles shown in Fig. 9.13. V.E. = vertical exaggeration; HHW=high high water;
LLW = low low water. (After Hill & Frobel I99 I .)

Hill, 1989). The basal parts of some thermokarst lake basins appear to be
preserved on both gravel-rich, steep-profile shorelines, such as King Point
(Hill, 1990), and on the sandy, low-angle shorelines of Richards Island and the
Tuktoyaktuk Peninsula (Htquette & Hill, 1989; Hill & Frobel, 1991). Thus, in
contrast to the Alaskan Beaufort Sea, a relatively clear ravinement surface
(Nummedal & Swift, 1987) is preserved in the shelf stratigraphic record.
The geomorphologic evolution described above is clearly zonal in nature,
being strongly controlled by the presence of permafrost. However, within the
above framework, the development of individual beach or spit morphology
reflects local conditions of sediment supply, grain size and exposure to wave
 Arctic coastal plain shorelines 363
TYERMOKARST
TOPOORAPHV

Coastal accumulation features

Barrier Island drowning
Pleistocene sediments

-EAOdNOAVD EMBAYMENI (21 SPI-OEVELOPYE\T

:ORMATION

(4) BARRIER RETREAT (3) BARRIER ISLAUO FORMATION

Figure 9.15. Model for coastal evolution of thermokarst terrain, Canadian Beaufort Sea
coast. Note similarities to the model shown in Fig. 9.10. (From Ruz et al., 1992.)

energy rather than zonal sea-ice-related processes. The barrier beach at King
Point developed through southeasterly longshore transport of sand and gravel
eroded from the northwest (Fig. 9.12). Sediment budget calculations suggest
that the source of sediment for construction of the initial spit was the former
headland, now mainly eroded, but that rapid closure of the barrier and later
beach accretion at the southeastern end were related to straightening of the
coast and the resultant increased sediment supply (Hill, 1990). Calculations of
potential sediment transport based on a hindcast model of wave conditions at
King Point have been used to suggest that this morphological evolution is
limited principally by the sediment transport potential (Hill, 1990). Thermal
erosion of bluffs by thaw-flow slide provides a potentially large local sediment
supply, but the bluffs remain stabilized until wave energy has removed all the
364 P.R. Hill et al.
flow-slide material deposited on the beach. Thus, thermal erosion, although
maximizing bluff retreat, does not necessarily accelerate coastal evolution in
regions of low wave energy.
Mixed sand and gravel beaches east of the Mackenzie Delta show a
differentiation between gravel-rich storm ridges and barred sandy foreshores.
Wolfe Spit, for example, is characterized by a recurved 2 m high sand and
gravel ridge that extends over 1 km eastward of the headland (Fig. 9.13). A
sandy beach has accreted in front of this ridge and extends eastward almost
1 km beyond it, increasing the overall length of the spit. This part of the spit has
a maximum elevation of 1.2 m and was breached by a broad washover channel
between 1974 and 1985. In front of these two elevated sections of spit, a series
of oblique, northeastward-facing bars form a broad platform, with a maximum
elevation of 0.6 m, that presently extends over 4 km eastward. Between 1974
and 1985, a narrow sandy berm, with a maximum elevation of 1.Om accreted
seaward of the proximal end of the spit and now extends more than 2 km from
the headland. Aerial photographs indicate the the presence of at least two shore-
parallel nearshore bars seaward of Wolfe Spit and the headland to the west.
Analysis of this morphology suggests a complex evolution characterized by
a partition of sediment transport between storm surge conditions and
‘fairweather’ conditions when water levels are lower. During northwesterly
storms, most of the spit, with the exception of the storm ridge, is submerged.
During moderate storms, waves break directly on the nearshore bars and, under
extreme surge conditions, it is likely that reformed waves would break on the
gravel spit, reactivating sediment transport. Sediment transport during such
storms would be predominantly eastward along the nearshore bars, the sandy
berm and the upper beach foreshore. The oblique foreshore bars appear to be
in equilibrium with easterly wind (‘fairweather’) conditions and suggest that
accretion of these bars reflects a return westward transport of sand.
The importance of sea-ice processes to coastal evolution in the Canadian
Beaufort Sea appears to be less crucial than in the Alaskan region, but it is
important to note that there have been few observations of coastal processes
during the autumn and winter periods. Ice pile-up and ride-up are relatively
rare at most locations, although pile-up events have been observed along the
Tuktoyaktuk Peninsula (Dickins, 1987; Fig. 9.7). Ice-scouring is common in
water depths less than 10m, and undoubtedly occurs inshore, but sidescan
sonar studies of the outer shoreface during the summer shows that reworking
of sediments by wave and current processes obliterates ice scours (HCquette &
Hill, pers. commun.; M. Desrosiers, pers. commun.). There have been no
observations of ice-wallow relief on beaches, but detailed surveys of the
foreshore have been by no means exhaustive.
 Arctic coastal plain shorelines 365
Discussion
Both Canadian and Alaskan sectors of the Beaufort Sea coastline have clearly
zonal influences, particularly at the morphologic level (Wiseman et al., 1973;
Ruz et al., 1992). The principal difference between the two regions lies in the
relative importance of zonal and azonal processes on the shoreface. These
differences are observable in two characteristics: 1. the equilibrium between
the rate of shoreface retreat and the littoral sediment budget; and 2. the
preservation of the ravinement surface on the inner shelf. On the Alaskan
coast, the fact that sediment supply from bluff retreat greatly exceeds the
sediment that is stored in the coastal zone (Reimnitz et al., 1988) suggests
strongly that sediment transport on the shoreface is not in equilibrium with
energy supplied by the relatively modest waves and currents of the region.
Zonal processes, such as entrainment of sediment in the ice canopy during
autumn storms (Reimnitz & Kempema, 1987) and transport of sediment by ice
scouring (Rearic et al., 1990) or ice push (Barnes & Reimnitz, 1988), are
important over the longer time scale of coastal evolution. This is reflected in
the absence of coastal deposits or ravinement surfaces on the shelf (Barnes &
Reimnitz, 1974). The sedimentary record of transgression is reduced to a
highly deformed ice-keel turbate layer.
In contrast, the Canadian Beaufort coast appears to be more in equilibrium
over the long term with the energy supplied by wave and current processes.
The sediment budget at King Point over a 30 year period, for example,
indicates that the sediment supplied by bluff retreat can be accounted for in the
sediments stored within the barrier beach (Hill, 1990). Although ice scouring
causes considerable deformation of post-transgression sediments, a clear
ravinement surface can be recognized across the shelf (Blasco et al., 1990). In
addition, a coastal sand sheet is present on the inner shelf and possible
drowned barrier islands and thermokarst lake basins are preserved below the
ravinement surface (Htquette & Hill, 1989). Zonal processes may be locally
important. HCquette & Barnes (1990) have suggested, for example, that ice-
scour erosion on the lower shoreface may accelerate shoreline retreat by
deepening the submarine profile and requiring further erosion on the upper
shoreface to maintain dynamic equilibrium. It is not clear, however, that this
process is as efficient as wave and current transport, which could rapidly
negate the effect of ice-scour erosion.
At shorter time scales, both regions show beach and nearshore bar
morphodynamics that are not particularly zonal in nature, but rather show a
morphological response to the relatively moderate wave regime, dominated by
storms and associated surges. This is in the form of highly differentiated
beaches and nearshore bars. Although the moderate wave regime can be
366 P.R. Hill et al.
considered a zonal response to fetch limitation by sea ice, the cause of this
differentiation is more probably the response of a microtidal shoreline to storm
surge conditions.
The conclusion from the above comparison would appear to be that the
relative importance of zonal and azonal processes depends principally on the
degree of fetch limitation imposed on the shoreline by the seasonal ice cover.
More precisely, it is the joint probability of long fetches and extreme storms
that controls the long-term evolution of Arctic shorelines. Because of the
unconsolidated nature of the coastal plain deposits being eroded and the
importance of thermal degradation of the exposed bluffs, sediment supply is
generally high. When the fetch is very restricted, ice-related processes are
certainly dominant and appear to be efficient enough agents of sediment
transport to remove the sediment supplied by a rapidly eroding coast. Where
the fetch is less limited, wave energy is higher and transgression proceeds as
on any other wave-dominated coastline by progressive erosion of the
shoreface, with the added influence of ice-related processes. The importance
of these zonal processes is not clear, due largely to the lack of observations at
critical times of year.
Many zonal processes, such as ice scour and ice wallow, are very
intermittent in their action and therefore difficult to evaluate in terms of their
efficiency. Observations, such as the obliteration of ice scours on the inner
shelf during large storms (Barnes & Reimnitz, 1979) suggest that over the
short term, their action is considerably less efficient than waves and currents.
However, over the longer term the cumulative effect appears to be significant
(Reimnitz et al., 1988; Rearic et al., 1990).
Other zonal processes, particularly those related to freeze-up conditions, are
very poorly understood in terms of their efficiency as agents of sediment
transport. The work of Reimnitz and others (Reimnitz & Kempema, 1987;
Reimnitz et al., 1987; Reimnitz et al., 1992) has recognized the potential
importance of autumn storms for removing sediment from the shoreline and
that of sea ice as a transport agent. Hydrodynamic conditions may be greatly
modified by the presence of frazil ice in the water column and anchor ice at the
seabed. It is therefore probable that the morphology of beaches and nearshore
bars along the Alaskan coast is in dynamic equilibrium with these autumn
storm events. It is not known how important this would be on the more wave-
influenced Canadian Beaufort Sea coast. Hydrodynamic studies have been
carried out in summer, but not during this potentially critical time of the year,
nor during the winter.
For a more complete understanding of beach and nearshore
 Arctic coastal plain shorelines 3 67
morphodynamics and of the relative importance of zonal versus azonal
processes, field research needs to be carried out in the autumn and the winter.
Standard equipment, moored through the ice, could be used potentially to
investigate winter circulation on the lower shoreface and inner shelf during the
winter, after a solid ice cover has been established. However, measurements
conducted in the critical autumn period are problematic due to instrument
icing, hazards from drifting ice and simply access to the foreshore under
potentially dangerous conditions. Nevertheless, research into the dynamic
processes of this period should be a future priority.

Acknowledgements
Funding for this paper was provided by operating grants from the Natural
Science and Engineering Research Council, Canada, and Research
Agreements with Energy, Mines and Resources, Canada to Hill and HCquette.
We would like to acknowledge the continued support and encouragement
provided by colleagues from both sides of the AlaskdYukon border,
particularly Steve Blasco, Don Forbes, Erk Reimnitz, and Steve Solomon.

References
Barnes, P.W. (1982). Marine ice-pushed boulder ridge, Beaufort Sea, Alaska. Arctic,
35,312-16.
Barnes, P.W., Asbury, J.L., Rearic, D.M. & Ross, C.R. (1987). Ice erosion of a sea-
floor knickpoint at the inner edge of the stamuckhi zone, Beaufort Sea, Alaska.
Marine Geology, 76,207-22.
Barnes, P.W., Rawlinson, S.E. & Reimnitz, E. (1988). Coastal geomorphology of
Arctic Alaska. In Arctic coastal processes and slope protection design, ed. A.T.
Chen & C.P. Leidersdorf, pp. 3-30. New York: American Society of Civil
Engineers.
Barnes, P.W., Rearic, D.M. & Reimnitz, E. (1984). Ice gouging characteristics and
processes. In The Alaskan Beaufort Sea: ecosystems and environments, ed.
P.W. Barnes, D.M. Schell & E. Reimnitz, pp. 185-212. Orlando, FL: Academic
Press.
Barnes, P.W. & Reimnitz, E. (1974). Sedimentary processes on Arctic shelves off the
northern coast of Alaska. In The coast and shelfof the Beaufort Sea, ed. J.C.
Reed & J.E. Sater, pp. 439-76. Arlington, VA: Arctic Institute of North
America.
Barnes, P.W. & Reimnitz, E. (1979). Ice gouge obliteration and sediment
redistribution event; 1977-1978, Beaufort Sea, Alaska. US Geological Survey
Open File Report 79-848,22 pp.
Barnes, P.W. & Reimnitz, E. (1988). Construction of an Arctic barrier island by
alternating sea-ice pileup and overwash. In Geologic studies in Alaska by the
US.Geological Survey during 1987, ed. J.P. Galloway, & T.D. Hamilton, pp.
180-2. US.Geological Survey Circular 1016.
 P.R. Hill et al.
Blasco, S.M., Fortin, G., Hill, P.R., O’Connor, M.J. & Brigham-Grette, J.K. (1990).
The Late Neogene and Quaternary stratigraphy of the Canadian Beaufort
Continental Shelf. In The Arctic Ocean region, The geology of North America
series, Vol. L, ed. A. Grantz, L. Johnson & J.F. Sweeney, pp. 491-502.
Boulder, CO: Geological Society of America.
Burgess, M., Judge, A., Taylor, A. & Allen, V. (1982). Ground temperature studies of
permafrost growth at a drained lake site, Mackenzie Delta. In Proceedings of
Fourth Canadian Permafrost Conference (Calgary, Alberta), ed. H.M. French,
pp. 3-1 I . National Research Council of Canada, Associate Committee on
Geotechnical Research.
Carter, L.D., Brigham-Grette, J. & Hopkins, D.M. (1986). Late Cenozoic marine
transgressions of the Alaskan Arctic coastal plain. In Correlations of
Quaternary deposits and events around the margin of the Beaufort Sea, ed. J.A.
Heginbottom & J-S. Vincent, pp. 21-6. Geological Survey of Canada Open
File Report 1237.
Czudek, T. & Demek, J. (1970). Thermokarst in Siberia and its influence on the
development of lowland relief. Quaternary Research, 1, 103-20.
Dallimore, S.R., Kurfurst, P.J. & Hunter, J.M. (1988). Geotechnical and geothermal
conditions of nearshore sediments, southern Beaufort Sea, Northwest
Territories, Canada. In Proceedings, V International Conference on
Permafrost, Trondheim, Norway, pp. 127-3 I . Trondheim, Norway: Tapir
Publishers.
Davidson, S., de Margerie, S. & Lank, K. (1988). Sediment transport in the
Mackenzie River plume. Geological Survey of Canada Open File Report 2303.
123 pp.
Davies, J.L. ( 1972). Geographical variation in coastal development. London:
Longman. 204 pp.
Dickins, D.F. (1987). Aerial reconnaissance survey of ice break-up processes in the
Canadian Beaufort Sea coastal zone. Geological Survey of Canada Open File
Report 1687. 16 pp.
Dinter, D.A., Carter, L.D. & Brigham-Grette, J. (1990). Late Cenozoic geologic
evolution of the Alaskan North Slope and adjacent continental shelves. In The
Arctic Ocean region, The Geology of North America series, Vol. L, ed.
A. Grantz, L. Johnson & J.F. Sweeney, pp. 459-90. Boulder, CO: Geological
Society of America.
Dyke, L.D. (1991). Temperature changes and thaw of permafrost adjacent to Richards
Island, Mackenzie Delta, N.W.T. Canadian Journal of Earth Sciences, 28,
1834-42.
Fissel, D.B. & Birch, J.R. (1984). Sediment transport in the Canadian Beaufort Sea.
Unpublished report by Arctic Sciences Ltd., Victoria, B.C. for Geological
Survey of Canada, Dartmouth, N.S. 165 pp.
Fissel, D.B., Birch, J.R. & Melling, H. (1990) Interannual variabiliry of
oceanographic conditions in the southeastern Beaufort Sea. Canadian
Contractor Report on Hydrography and Ocean Sciences No. 35, Fisheries and
Oceans Canada. 102 pp.
Forbes, D.L. (1980). Late-Quaternary sea levels in the southern Beaufort Sea. Current
Research, Part B. Geological Survey of Canada, Paper 80- I B, 75-87.
Gornitz, V., Lebedeff, S. & Hansen, J. (1982). Global sea-level trends in the past
century. Science, 215, 161 1-14.
Greenwood, B. & Davidson-Arnott, R.G.D. ( 1 979). Sedimentation and equilibrium in
wave-formed bars: a review and case study. Canadian Journal ofEarth
Sciences, 16, 3 12-32.
 Arctic coastal plain shorelines 369
Guilcher, A. (1953). Essai sur la zonation et distribution des formes littorales de
dissolution de calcaire. Annals de GPographie, 62, I6 1-79,
Harper, J.R. (1990). Morphology of the Canadian Beaufort Sea coast. Marine
Geology, 91,75-9 I .
Harper, J.R., Henry, R.F. & Stewart, G.G. (1988). Maximum storm surge elevations
in the Tuktoyaktuk region of the Canadian Beaufort Sea. Arctic, 41,48-52.
Harry, D.G. (1985). Ground ice slumps, Beaufort Sea coast, Yukon Territory. In
Fourteenth Arctic workshop, Arctic land-sea interactions, Abstracts,
pp. 1 15-17. Dartmouth, N.S., Canada: Bedford Institute of Oceanography.
Henry, R.F. (1975). Storm surges. Beaufort Sea Project technical report no. 19,
Victoria, B.C., Canada: Department of the Environment. 41 pp.
Henry, R.F. (1984). Flood hazard delineation at Tuktoyaktuk. Department of Fisheries
and Oceans, Canadian contractor report of hydrography and ocean sciences No.
19. Sidney, B.C., Canada: Institute of Ocean Sciences. 28 pp.
HCquette, A. & Barnes, P.W. (1990). Coastal retreat and shoreface profile variations
in the Canadian Beaufort Sea. Marine Geology, 91, 1 13-32.
HCquette, A. & Hill, P.R. ( 1 989). Late Quaternary seismic stratigraphy of the inner
shelf seaward of the Tuktoyaktuk Peninsula, Canadian Beaufort Sea. Canadian
Journal of Earth Sciences, 26, 1990-2002.
HCquette, A. & Ruz, M-H. (1991). Spit and barrier island migration in the
southeastern Canadian Beaufort Sea. Journal of Coastal Research, 7,677-98.
Hill, P.R. (1990). Coastal geology of the King Point area, Yukon Territory, Canada.
Marine Geology, 91,93-11 I .
Hill, P.R., Blasco, S.M., Harper, J.R. & Fissel, D.B. (1991). Sedimentation on the
Canadian Beaufort Shelf. Continental ShelfResearch, 11, 8 2 1 4 2 .
Hill. P.R., Forbes, D.L., Dallimore, S.R. & Morgan, P. (1986). Shoreface
development in the Canadian Beaufort Sea. In Proceedings, symposium on
cohesive shores, ed. M.G. Skafel, pp. 428-48. Ottawa: National Research
Council of Canada, Publication No. 26134.
Hill, P.R. & Frobel, D. (199 I). Documentation of summer NOGAP activities,
July 22- August 25, 1990. Geological Survey of Canada Open File Report
245 1 . 8 4 ~ ~ .
Hill, P.R., HCquette, A. & Ruz, M-H. (1993). Holocene sea-level history of the
Canadian Beaufort shelf. Canadian Journal of Earth Sciences, 30, 103-8.
Hill, P.R., HCquette, A., Ruz, M-H. & Jenner, K.A. (1990). Geological investigations
of the Canadian Beaufort Sea coast. Geological Survey of Canada Open File
Report 2387.348 pp.
Hill, P.R., Mudie, P.J., Moran, K. & Blasco, S.M. (1985). A sea-level curve for the
Canadian Beaufort shelf. Canadian Journal of Earth Sciences, 22, 1383-93.
Hobson, G.D. & the Canadian Ice Island Scientific Party (1989). Ice Island field
station: new features of Canadian Polar Margin. Eos, 70, 833, 835, 838-9.
Hodgins, D.O. (1985). A review of extreme wave conditions in the Beaufort Sea.
Department of Fisheries and Oceans, Canadian contractor report of
hydrography and ocean sciences no. 12. Sidney, B.C., Canada: Institute of
Sciences. 160 pp.
Hunter, J.A. & Hobson, G.D. (1974). Seismic refraction method of detecting subsea
bottom permafrost. In The c o a ~and
t shelfof the Beaufort Sea, ed. J.C. Reed &
J.E. Sater, pp. 401-16. Arlington, VA: Arctic Institute of North America.
Judge, A. (1986). Permafrost distribution and the Quaternary history of the Mackenzie-
Beaufort region: a geothermal perspective. In Correlation of Quaternary
deposits and events around the margin of the Beaufort Sea, ed. J.A. Heginbottom
& J-S. Vincent, pp. 41-5. Geological Survey of Canada Open File Report 1237.
370 P.R. Hill et al.
Kelletat, D. (1989). The question of ‘zonality’ in coastal geomorphology -with
tentative application along the east coast of the USA. Journal of Coastal
Research, 5,329-44.
Kovacs, A. ( 1 983). Shore ice ride-up and pile-up features. Part 1. Alaska’s Beaufort
Sea coast. Hanover, NH: US.Army Cold Regions Research and Engineering
Laboratory, Report 83-9. 5 1 pp.
Kovacs, A. (1984). Shore ice ride-up and pile-up features. Part 2. Alaska’s Beaufort
Sea coast. Hanover, NH: US.Army Cold Regions Research and Engineering
Laboratory, Report 84-26. 29 pp.
Lachenbruch, A.H., Sass, J.H., Lawver, L.A., Brewer, M.C., Marshall, B.V., Munroe,
R.J., Kennelly, J.P., Jr. & Moses, T.H., Jr. (1987). Temperature and depth of
permafrost on the Alaskan North Slope. In Alaskan north slope geology, ed.
I.L. Tailleur & P. Weimer, pp. 544-58. Bakersfield, CA: Pacific Section,
SEPM and the Alaska Geological Society.
Lewis, C.F.M. (1978). The frequency and magnitude of drift ice groundings from ice
scour tracks in the Canadian Beaufort Sea. In Proceedings 4th International
Conference on Port and Ocean Engineering Under Arctic Conditions, pp.
568-79. St. John’s, Newfoundland: Memorial University of Newfoundland.
Mackay, J.R. (197 I). The origin of massive icy beds in permafrost, western Arctic
coast. Canadian Journal of Earth Sciences, 8, 397-422.
MacNeil, M.R. & Garrett, J.R. (1975). Open water surface currents. Beaufort Sea
Project, Technical Report No. 17, Environment Canada.
Martin, S. (1981). Frazil ice in rivers and oceans. Annual Review of Fluid Mechanics,
13,379-97.
Mysak, L.A. & Manak. D.K. (1989). Arctic sea ice extent and anomalies, 1953-1984.
Atmosphere-Ocean, 27, 376405.
Neave, K.G. & Sellmann, P.V. (1984). Determining distribution patterns of ice-
bonded permafrost in the U.S. Beaufort Sea from seismic data. In The Alaskan
Beaufort Sea: ecosystems and environments, ed. P.W. Barnes, D.M. Schell &
E. Reimnitz, pp. 237-58. Orlando, FL: Academic Press.
Nummedal, D. & Swift, D.J.P. (1987). Transgressive stratigraphy at sequence-
bounding unconformities: some principles derived from Holocene and
Cretaceous examples. In Sea levelfluctuations and coastal evolution, ed.
D. Nummedal, O.H. Pilkey & J.D. Howard, pp. 241-60. SEPM Special
Publication 41.
Pinchin, B.M., Nairn, R.B. & Philpott, K.L. (1985). Beaufort Sea coastal sediment
study: numerical estimation of sediment transport and nearshore profile
adjustment at coastal sites in the Canadian Beaufort Sea. Geological Survey of
Canada, Open File Report 1259.7 12 pp.
Rampton, V.N. (.1982). Quaternary geology of the Yukon coastal plain. Geological
Survey of Canada Bulletin 317. 49 pp.
Rampton, V.N. (1988a). Origin of massive ground ice on the Tuktoyaktuk
Peninsula, Northwest Territories, Canada: a review of stratigraphic and
morphologic evidence. In Proceedings of the V International Permafrost
Conference, Trondheim, Norway, pp. 850-5. Trondheim: Norway: Tapir
Publishers.
Rampton, V.N. ( I988b). Quaternary geology of the Tuktoyaktuk Coastlands,
Northwest Territories. Geological Survey of Canada, Memoir 423. 98 pp.
Rearic, D.M., Barnes, P.W. & Reimnitz, E. (1990). Bulldozing and resuspension of
shallow-shelf sediment by ice keels: implications for Arctic sediment transport
trajectories. Marine Geology, 91, 1 3 3 4 7 .
 Arctic coastal plain shorelines 37 1
Reimnitz, E., Barnes, P.W. & Harper, J.R. (1990). A review of beach nourishment
from ice transport of shoreface materials, Beaufort Sea, Alaska. Journal of
Coastal Research, 6,439-70.
Reimnitz, E., Graves, S.M. & Barnes, P.W. (1988). Beaufort Sea coastal erosion,
sediment flux, shoreline evolution, and the erosional shelf profile. U S .
Geological Survey Miscellaneous Investigations Series, accompaniment to Map
I- 1 18243 ( I :82,000).
Reimnitz, E. & Kempema, E. (1982). Dynamic ice-wallow relief of northern Alaska’s
nearshore. Journal ofsedimentary Petrology, 52,45 1-61.
Reimnitz, E. & Kempema, E. (1984). Pack ice interaction with Stamukhi Shoal, Beaufort
Sea, Alaska. In The Alaskan Beaufort Sea: ecosystems and environments, ed. P.W.
Barnes, D.M. Schell & E. Reimnitz, pp. 159-83. Orlando, FL: Academic Press.
Reimnitz, E. & Kempema, E.W. (1987). Field observations of slush ice generated
during freeze-up in Arctic coastal waters. Marine Geology, 77,219-31.
Reimnitz, E., Kempema, E.W. & Barnes, P.W. (1987). Anchor ice, seabed freezing
and sediment dynamics in shallow Arctic seas. Journal ofGeophysica1
Research, 92, C13, 14671-8.
Reimnitz, E., Marincovich, L., Jr., McCormick, M. & Briggs, W.M. (1992).
Suspension freezing of bottom sediment and biota in the Northwest Passage
and implications for Arctic Ocean sedimentation. Canadian Journal ofEarth
Sciences, 29, 693-703.
Reimnitz, E. & Maurer, D.K. (1979). Effects of storm surges on the Beaufort Sea
coast, northern Alaska. Arctic, 32, 3 2 9 4 4 .
Ruz, M-H., HCquette, A. & Hill, P.R. (1992). A model of coastal evolution in a
transgressed thermokarst topography, Canadian Beaufort Sea. Marine Geology,
106,25 1-78.
Sellmann, P.V., Brown, J. & Lewellen, R.I. (1975) The class(fication ofthaw lakes on
the Arctic coastal plain, Alaska. U S . Army Corps of Engineers, Cold Regions
Research and Engineering Laboratory, Research Report 344. 2 1 pp.
Shapiro, L.H. & Barnes, P.W. (1991). Correlation of nearshore ice movement with
seabed ice gouges near Barrow, Alaska. Journal ofGeophysica1 Research, 96,
16 979-89.
Short, A.D. (1975). Offshore bars along the Alaskan Arctic coast. Journal of Geology,
83,209-2 1.
Short, A.D., Coleman, J.M. & Wright, L.D. (1974). Beach dynamics and nearshore
morphology of the Beaufort Sea coast, Alaska. In The coast and shelfofthe
Beaufort Sea, ed. J.C. Reed & J.E. Sater, pp. 477-88. Arlington, VA: Arctic
Institute of North America.
Smith, M.W. (1975). Permafrost in the Mackenzie Delta, N. W.T. Geological Survey
of Canada Paper 75-28. 34 pp.
Taylor, R.B. & McCann, S.B. (1983). Coastal depositional landforms in northern
Canada. In Shorelines and isostasy, ed. D.E. Smith & A.G. Dawson, pp. 53-75.
London: Academic Press.
Untersteiner, N. (1990). Structure and dynamics of the Arctic Ocean ice cover. In
The Arctic Ocean region, The geology of North America series, Vol. L. ed.
A. Grantz, L. Johnson & J.F. Sweeney, pp. 37-5 I . Boulder, CO: Geological
Society of America.
Wiseman, W.J., Coleman, J.M., Gregory, A,, Hsu, S.A., Short, A.D., Suhayda, J.N.,
Walters, C.D., Jr. & Wright, L.D. (1973). Alaskan Arctic coastal processes and
morphology. Coastal Studies Institute, Technical Report 149, Baton Rouge,
LA: Louisiana State University. 171 pp.
372 P.R. Hill et al.
Wiseman, W.J., Jr., Suhayda, J.N., Hsu, S.A. & Walters, C.D., Jr. (1974).
Characteristics of nearshore oceanographic environment of Arctic Alaska. In
The coast and sheyof the Beaufort Sea, ed. J.C. Reed & J.E. Sater, pp. 49-64.
Arlington, VA: Arctic Institute of North America.
Wright, L.D. & Short, A.D. ( 1 983). Morphodynamics of beach and surf zones in
Australia. In Handbook of coasfal processes and erosion, ed. P.D. Komar, pp.
35-64. Boca Raton: CRC Press.
 10
Paraglacial coasts
D.L. FORBES AND J.P.M. SYVITSKI
The initial shoreline, like that of the present, must have been largely dominated by
drumlin outlines ... a picture the sheer beauty of which will live long in the observer’s
memory.
Douglas Johnson, 1925, The New England - Acadian Shoreline, pp. 118-19

Introduction
Marine transgression across glaciated terrain can produce striking coastal
landscapes. Drumlin archipelagos such as Boston Harbour (in the northeastern
USA), Mahone Bay (in southeastern Canada), or Clew Bay (in the west of
Ireland) are among the most impressive such landscapes, Boston Harbour and
Mahone Bay providing the inspiration for Douglas Johnson’s words of reverie.
Still more spectacular in many cases are the flooded glacial valleys forming
fjords (Fig. 10.1), common along the margins of coastal highlands throughout
the Northern and Southern Hemisphere fjord belts (Fig. 10.2). While
underscoring our central assertion, that glaciation may exert a profound
influence on the later evolution of a coastline, fjords and drumlin archipelagos
are but two examples of the wide range of paraglacial phenomena encountered
on mid- to high-latitude coasts.
Large areas of the globe have been glaciated repeatedly during late
Cenozoic time and intermittently on earlier occasions in the geological record
(Harland & Herod, 1975; Fulton, 1989). Major continental ice sheets have
experienced cyclic growth and decay during the last two million years or more,
while the record of glaciation along the Gulf of Alaska continental margin goes
back to the late Miocene (Eyles, Eyles & Lagoe, 1991). Marine cores from the
Norwegian shelf have yielded evidence of mountain glaciation extending back
some 35 Ma (Jansen & Sjprholm, 1991) and the Antarctic has a similar history.
Although ice sheets of continental scale are still present over Greenland and
Antarctica, and numerous smaller ice masses persist elsewhere, most areas
covered by Late Quaternary glaciation have been ice-free for many thousands
of years (typically 8 to 22 ka, depending on the distance from former ice centres).
The environmental impact of glaciation in formerly ice-covered areas
remains pervasive despite the passage of time (see for comparison the Arctic
coastal plain shorelines described in Chapter 9). This legacy includes: distinctive

373
374 D.L. Forbes & J.P.M. Syvitski

Figure 10.1. North Arm, a fjord on the northeast coast of Baffin Island in the eastern
Canadian Arctic. Looking up-fjord with side-entry glacier and valley-wall talus sheets
and cones. Photo courtesy R.B. Taylor.

landforms such as moraines, drumlins, fjords and other overdeepened

embayments; widespread deposits of glacial or proglacial origin, including
glacial diamicts, outwash sands and gravels, and glaciomarine or
glaciolacustrine silts and clays; and ongoing vertical movements of land and
sea that continue to influence the rates and processes of coastal change.
Natural systems thus affected by glaciation are described as paraglacial.
This term was introduced by Ryder (1971) and Church & Ryder (1972) to
describe the initial postglacial excess and subsequent relaxation of sediment
supply on alluvial fans, and in river basins generally, in glaciated terrain.
Sediment transport in paraglacial river systems of modest scale (Fig. 10.3) has
 Paraglacial coasts 375
NORTHERN FJORD BELT

SOUTHERN FJORD

Figure 10.2. Generalized global distribution of fjords and other paraglacial coastal
environments (reproduced from Syvitski et al., 1987).

4U SMALL VALLEYS
a
c

MAJOR VALLEYS

PRIMARY SUBAERIAL DENUDATION RATE

TIME
Figure 10.3. Sediment supply in paraglacial river systems (Church & Ryder, 1972;
redrawn after Church & Slaymaker, 1989).

been shown to decline rapidly from late-glacial or early postglacial maxima to

much lower values in the mid- to late Holocene (Church & Ryder, 1972). In
larger basins (Fig. 10.3), intermediate storage in valley deposits can induce a
lag in the peak sediment delivery and prolong the relaxation process (Church
376 D.L. Forbes C? J.P.M. Syvitski
& Slaymaker, 1989). This latter pattern may also occur in relatively small
basins in areas of glacioisostatic rebound, where high-level deltas and terraces
are later reworked as base level falls. Cold-climate effects, including winter ice
cover and periglacial conditions such as permafrost and ground ice, may be
present incidentally in paraglacial systems and play an important role in
coastal development (see e.g. Taylor, 1978; Rosen, 1979; McLaren, 1982;
Taylor & Forbes, 1987; Reimnitz et al., 1991; Ruz, Htquette & Hill, 1992;
Forbes & Taylor, 1994; see also Chapter 9), but are not essential elements.
Although glacioeustatic, glacioisostatic, and related hydroisostatic changes
in relative sea level are effective on a global scale (Walcott, 1972), such that
any coast in the world could be considered quasi-paraglacial in the broadest
sense, we adopt a more restrictive definition in this review. Here we define
paraglacial coasts to be those on or adjacent to formerly ice-covered terrain,
where glacially excavated landforms or glaciogenic sediments have a
recognizable influence on the character and evolution of the coast and
nearshore deposits. In other words, the distinguishing features relate to
sediment type, sediment supply, or the presence of glacial landforms.
Proglacial settings in which ice is still present occupy one end of the
paraglacial spectrum. Other glaciogenic effects such as changes in relative sea
level are crucial to an understanding of coastal change, but are not part of the
definition adopted here. Indeed, sea-level changes are known to have an almost
ubiquitous influence on coastal development and related stratigraphic
signatures throughout the Holocene (e.g. volume edited by Nummedal, Pilkey
& Howard, 1987b) and more generally (e.g. Posamentier, Jervey & Vail, 1988;
Cant, 1991).
Even under the restricted definition given above, paraglacial effects are
present and frequently dominate the character and development of coasts in
many parts of the world. Major areas of paraglacial influence (Fig. 10.2)
include ice-free portions of Antarctica, islands in the Southern Ocean, parts of
Tasmania, the South Island of New Zealand, southern South America, large
areas of northern North America, northern Europe and Asia. In this review we
examine a variety of paraglacial coasts, drawing our examples primarily from
the North Atlantic region (eastern Canada and Iceland). We focus on cases
from this part of the world because of the wide range of conditions represented
here and our greater familiarity with these examples.
It is only within the past few years that paraglacial effects have been
explicitly identified as such in the context of coastal development and
sedimentation (Forbes, 1984; Forbes & Taylor, 1987; Syvitski, Burrell & Skei,
1987) and research activity in glaciated regions has begun to focus attention on
paraglacial systems (e.g. volume edited by FitzGerald & Rosen, 1987). This is
 Paraglacial coasts 377
not to say that the effects of glaciation on coastal landforms had not been noted
much earlier. Fjords had been recognized as coastal features of glacial origin
as early as the 1820s (Esmark, 1827), although opposing theories of fluvial or
tectonic origin fuelled discussion of the so-called ‘fjord problem’ until at least
as late as 1913 (Johnson, 1915, 1919). Proglacial outwash deposits, including
deltaic sinks in lakes or the sea, have received considerable attention over the
years (e.g. Davis, 1890, 1896; Stone, 1899; more recently Church, 1972, 1978;
papers in Jopling & McDonald, 1975; Boothroyd & Nummedal, 1978; among
others), as have features associated with shoreline development in coastal
drumlin fields (e.g. Johnson & Reed, 1910; Taylor et al., 1986; Piper et al.,
1986; Wang & Piper, 1982; Boyd, Bowen & Hall, 1987; Carter & Orford,
1988; Carter et al., 1990b; among others; see also Fig. 1.4). Distinctive aspects
of coastal development in glaciated terrain were noted by Johnson (1919),
while his 1925 monograph included chapters on ‘Initial shorelines determined
by glacial forms’, on ‘Erosional forms bordering unconsolidated [glaciogenic]
deposits’ (with a section on the ‘Rapidity of marine erosion of glacial
deposits’), and on ‘Wave-built forms bordering unconsolidated [glaciogenic]
deposits’.
Temporal patterns of sediment delivery or availability in paraglacial coastal
systems are determined in large part by the disposition of glacial deposits
relative to the coastline and by relative sea-level changes that control access to
these sediments. In fjord-head and other settings where river supply plays a
major role, fluvial patterns of sediment supply with relaxation scales of a few
thousand years, recognized previously by Church & Ryder (1972), may
prevail. Sea level plays a secondary role at the river outlet, where an
isostatically induced relative drop in base level may initiate downcutting into
earlier, high-level, delta deposits (e.g. Church, 1972, 1978), augmenting
sediment discharge to the sea. On coasts where river inputs are unimportant,
sea level may exert a major control on sediment supply, shortening the
paraglacial cycle where water levels are stable and extending it where rising
sea levels give access to deposits at progressively higher levels (Forbes, Taylor
& Shaw, 1989). Periods of growth and decline in sediment supply play a
critical role in the evolution of the coast and this aspect of paraglacial systems
is considered further in a later section.

Coastal systems in the postglacial context

Coastal development in paraglacial environments occurs within a variety of
distinctive physiographic frameworks (Figs. 10.2 and 10.4). These include:
fjords (Fig. 10.l), fjord-head deltas, and associated shallow marine settings;
 e

Figure 10.4. Diversity of coastal forms in paraglacial settings: (a) Irregular rocky coastline with pocket beach and coastal barrens, near Grand
Bruit, southwest Newfoundland. (b) Story Head, a drumlin headland on the Eastern Shore of Nova Scotia, showing trailing drift-aligned spit on
the left (Fisherman’s Beach) and mailing ridge feeding sediment to swash-aligned barrier on the right. Incipient breaching of the barrier at the
point of connection (arrow) may lead to substantial changes in this system over the coming few years. Note well-developed longshore cell
structure on Fisherman’s Beach. Chezzetcook Inlet in background. (c) Oblique view of bayhead barrier at Portugal Cove South, southeast
Newfoundland, showing low, relict, progradational beach-ridge complex being ovemdden by high, transgressive, type-2, gravel storm ridge. (d)
Raised gravel beach ridges on south shore of Inglis Bay, northwest Devon Island, eastern Canadian Arctic (photo. by S.B. McCann). (e) Aerial
oblique view of boulder barricade in front of a thin gravel beach, west coast of Newfoundland north of Port au Choix. (f) Ground view of
boulder-strewn tidal flat and boulder barricade at head of Freshwater Bay, Bonavista Bay, northeast Newfoundland.
380 D.L. Forbes & J.P.M. Syvitski
outwash plains on exposed coasts; areas of extensive ice-contact or outwash
deposits in eroding bluffs; areas with thin or localized sediment sources on
coastlines of variable complexity (Fig. 1 0 . 4 ~and c); coastal drumlin fields in
which glaciogenic sediments form eroding headlands (Fig. 10.4b); and
regions in which coastal development is or has been strongly influenced by
rapid uplift (Fig. 10.4d). Johnson (1925) distinguished erosional coastal
features of glacial origin (fjords) and depositional shorelines associated with
glaciogenic deposits (end moraines, ground moraine, drumlins, outwash
deltas, eskers and kames).
Coastal evolution involves changes in shore-zone morphology and
associated facies geometry through time. The process can be reversed or
altered by changes in one or more parameters of the coastal system or when
thresholds are exceeded. In other cases, subtle changes in water levels, storm
frequency, sediment supply, or combinations of these factors may tip the
balance between progradation and transgression in susceptible systems (Shaw
& Forbes, 1987; Forbes et al., 1989) or otherwise alter the response of the
system to external forcing (cf. Allen, 1990). Subaerial and subaqueous parts of
a coastal system are intimately related and cannot be viewed in isolation
(e.g. Bruun, 1962; Shaw & Forbes, 1992).
Rates of coastal change and the form and facies characteristics of resulting
deposits depend on the geological and geographical setting, including the
following four major factors:

- antecedent morphology of the coastline and adjacent seabed;

- sources, physical properties, and rates of sediment supply;
- climatic and oceanographic environment; and
- rate and direction of relative sea-level change.

Changes in relative sea level (resulting from ocean-volume changes and

glacioisostatic uplift or subsidence) may determine access to sediment sources
and rates of sediment supply. Changes in relative sea level over the past 10 ka
(Fig. 10.5) exceed 200 m at some locations (Dyke, Morris & Green, 1991). The
direction and rate of relative sea-level change vary widely, from +10 mm a- '
or more on the inner Scotian Shelf off eastern Canada during part of the
Holocene (Forbes, Boyd & Shaw, 1991a; Shaw, Taylor & Forbes, 1993) to
-40mma-' in Glacier Bay, Alaska, as ice ablation and retreat succeed the
Neoglacial ice advance (Matthews, 1981).
 Paraglacial coasts 38 1

240 A BATHURST INLET, NWT

A B GASPESIE,QUEBEC
C ANTICOSTI, Q U h E C
220 D NORTHEAST NEW BRUNSWICK
E SOUTHWEST NEWFOUNDLAND
F INNER SHELF OFF HALIFAX, NOVA SCOTIA
200 G OUTER SHELF NEAR SABLE ISLAND, NOVA SCOTIA

180

160

d
>
y 120
B

U
w
C
D

5 40
E
5y 20
W

-20

-60 G

-'qS 14 12 10 8 6 4 2 0
TIME (ka BP)

Figure 10.5. Relative sea-level curves from various sites in eastern Canada (after Grant,
1989; Scott et al., 1989; Dyke et al., 1991; Forbes et al., 1991a), showing a range of
conditions from emergence to submergence and temporal-spatial variability in rates of
relative sea-level change.
 Paraglacial sediments
One of the primary distinguishing features of paraglacial coasts is the
predominance of glaciogenic sediment sources, often providing the only
significant inputs of material to the shore-zone. This material can include a
very wide range of grain sizes and sorting coefficients (Fig. 10.6). Lithologies
range from varved clays and well-sorted silts to bouldery rubble. Glaciogenic
source facies include tills, other ice-contact diamicts, glaciofluvial outwash,
and glaciolacustrine and glaciomarine deposits, which themselves can range
from diamicts through sandy turbidites to well-sorted silts and muds (e.g.
Eyles, Eyles & Day, 1983; Eyles et al., 1991; Miall, 1983). The grain size of
ice-contact diamicts can range over at least six orders of magnitude (more than
24 4 units), often without distinct gaps or modes (Fig. 10.6). Flow competence
is rarely an issue in glacial entrainment; subglacial erosion processes favour
the production of all grain sizes from very large blocks to so-called glacial
flour.
Competence does become important in proglacial and paraglacial contexts,
however. As the glacial melt component of a river's discharge subsides,
coarser sediment is left behind, largely immobile. Over-sized lag gravel

GRAIN SIZE (mm)

0.001 0.01 0.1 1.o 10

99
98
95
90
80
60
40
20
10

CLAY SILT -
SAND GRAVEL
10 9 8 7 6 5 4 3 2 1 0 -1 - 2 - 3 - 4 -5
GRAIN SIZE ( 0 UNITS)
Figure 10.6. Typical grain-size distributions of glacial till, outwash gravel, and
glaciomarine muds (redrawn after Boulton, 1976).
 Paraglacial coasts 383
constituents can armour the transport surface, reducing and possibly
eliminating the availability (or accessibility) of an underlying sediment source.
Where the paraglacial sediment input to the coastal zone is below the sea
surface, the same punctuation in sediment supply can develop. The winnowing
action of waves or tidal currents on shallow-water deposits may remove the
fines to deep or more-protected waters (e.g. Piper et al., 1986), leaving behind
an armouring lag of coarser particles that can ultimately reduce the availability
of finer sediment. In some paraglacial environments, flow competence has
increased over time. The tidal range in the Bay of Fundy has increased with
rising sea level and changing basin geometry during the Holocene (Wightman,
1976; Scott & Greenberg, 1983; Amos & Zaitlin, 1985). The bay has
undergone a transformation from microtidal (tidal range less than 1.5 m) in the
early Holocene to the present macrotidal condition (spring tidal range up to
16m), with concomitant high-energy currents (Amos & Long, 1980) and
distinctive sedimentary environments (Dalrymple et al., 1990).
Because of the wide spectrum of particle size entrained and deposited by
glacial processes (Fig. 10.6), sedimentary deposits derived from glaciogenic
sources often incorporate large volumes of gravel. Proglacial outwash
complexes are commonly gravel-dominated, at least in their proximal reaches
(e.g. Church, 1972; Church & Gilbert, 1975; Boothroyd & Nummedal, 1978;
Stravers, Syvitski & Praeg, 1991). Gravel is a common constituent of beach,
shoreface, and continental shelf sequences in glaciated terrain (e.g. Shepard,
1963; Stanley, 1968; Swift, Stanley & Curray, 1971; Forbes & Boyd, 1987,
1989; Forbes & Taylor, 1987; Kelley, 1987; Shipp, Staples & Ward, 1987;
Orford, Carter & Jennings, 1991b; Carter & Orford, 1993). With the exception
of tectonically active regions (e.g. Emery, 1955; Kirk, 1980; Massari & Parea,
1988; Moseley, Wagner & Richardson, 1992) and exceptional areas such as
the south coast of England (e.g. Carr, 1969; Jennings & Smyth, 1987), coastal
gravel is otherwise relatively rare and usually found only in immediate
proximity to cliff sources.
On the other hand, some glacial deposits are dominated by relatively fine-
grained facies and supply little gravel. Thus, sandy beaches predominate in the
southern Gulf of St. Lawrence, a glaciated area of low-lying Permian and
Carboniferous sandstones. The tills in the area are sandy and the rocks (being
easily erodible) represent important coastal sand sources in their own right
(Owens, 1974; McCann, 1979; Armon & McCann, 1977,1979). In parts of the
North American Great Lakes region, tills (Krumbein, 1933; Dreimanis, 1976;
Scott, 1976) and lake sediments (Eyles et al., 1983) exposed in coastal bluffs
are deficient in both gravel and sand, yielding relatively little material for
beach development (Davidson-Arnott, 1986; Philpott, 1986).
384 D.L. Forbes & J.P.M. Syvitski
Another result of high sediment discharge in early paraglacial time is the
extensive distribution on parts of the inner continental shelf of sands derived
from wasting of major ice masses such as the Laurentide Ice Sheet (e.g. Boyd,
Scott & Douma, 1988). Darby (1990) has argued that the elemental
composition of Fe-Ti oxide minerals (primarily ilmenite) in sands of the US
Atlantic Shelf points to the Hudson River as the dominant source of sands as
far south as Cape Hatteras (North Carolina). This implies very large discharge
and sediment supply from the Hudson River, which ‘may have rivalled ...
much larger rivers like the Mississippi or Amazon ..., at least during brief
glacial melting events’. In a partly analogous context along the QuCbec North
Shore of the Gulf of St. Lawrence, Syvitski & Praeg (1989) recognized an
acoustic stratigraphic unit interpreted as early paraglacial, representing the
melting of land-based ice caps and the rapid progradation of outwash deltas
under conditions of abrupt glacial-isostatic rebound (falling relative sea level).
In other areas, such as the former Champlain Sea basin, in the St. Lawrence
Lowland of eastern Canada, large marine outwash deposits now lie well above
present sea level (e.g. Rust & Romanelli, 1975; Sharpe, 1988).

Paraglacial settings and coastal outcomes

Paraglacial settings are arguably the most diversified set of coastal
environments treated in this volume. It is a challenge to select sufficient and
appropriate examples to represent the range of paraglacial coasts. This
treatment is necessarily selective. Following a review of glacial erosion
features (fjords and other overdeepened basins) and coastal sedimentation
within them, we proceed to a consideration of coastal development related to
various types of glaciogenic sediment source and their spatial and temporal
distribution. Numerical models are applied in two cases in order to clarify the
effects of major sediment supply and delivery processes.

Coasts dominated by glacial erosion: overdeepened basins and coastal barrens

Fjords are glacially scoured valleys partially submerged below sea level
(Fig. 10.1). They may include one or more overdeepened basins, which are
commonly confined at their down-valley limits by shallower sills. Glacially
overdeepened basins also occur in other forms, including: fjord-like lake
basins (cf. Eyles, Mullins & Hine, 1990), converted from fjords to lakes by
isostatic uplift and emergence of the fjord sill; submarine valley complexes
(Fig. 10.7), wherein subglacial meltwater is sometimes invoked as the erosional
agent (e.g. Boyd et al., 1988); and broader basins formed by differential
 Paraglacial coasts 385
50' 45' 40' 35' 58030'W 25
35' 35'
PORTAU PORT BAY RWAINES
BROOK

ISTHMUS
STEPHENVILLE
PORT
HARMON
STEPHENVILLE
CROSSING

480 480
30' 30'

LIMITOF
DETAILED
SURVEY
25 25'

BARRIER PLATFORM
AND SPILLOVER
FLAT BAY SEDIMENTS
BROOK
BARRIER DEPOSITS
BEDROCK VALLEY

BANK ERODING COASTAL

HEAD BLUFFS

50 45' 40 35' 58030'W 25'

Figure 10.7. Overdeepened basins in St. George's Bay, southwest Newfoundland

(redrawn after Shaw & Forbes, 1990b, 1992). Basin depths are greatest (>90 m off Flat
Island and 260 m off Stephenville) over partially filled submarine valleys more than
180 m deep that extend beneath the barriers at Stephenville and Stephenville Crossing
(cf. Fig. 5 of Shaw & Forbes, 1990b). The 12 km long Flat Island spit and the 3.5 km
beach-ridge barrier complex at Stephenville are fed from eroding coastal bluffs along
the north and south sides of the bay and rest on thick subaqueous barrier platforms
developed along and across the basins (cf. Shaw & Forbes, 1992).

excavation of weaker lithologies and subsequently flooded to form arms of the

sea (Miller, Mudie & Scott, 1982; Piper et al., 1983). Two distinct forms of
coastal sedimentation in glacial basins (fjord-head delta deposits and basin-
margin barrier platforms) are treated in separate sections below.
Some coasts lie within areas of extensive glacial or glaciofluvial erosion,
the result of which is a rock-bound coast of moderate relief, almost completely
devoid of sediment (Fig. 10.4a). Coastal barrens of this type are known, for
example, from parts of the Scandinavian and Greenland coasts and from a
number of areas in eastern Canada. The latter include areas of scoured granites
and metasediments on the Atlantic coast of Nova Scotia (Owens & Bowen,
1977) and Newfoundland (Forbes, 1984).
386 D.L. Forbes & J.P.M. Syvitski
Sedimentation in fjord-head and proximal basin settings
Fjords by their very nature, being deep and elongated, create a huge storage
capacity for sediments delivered to them. Syvitski et al. (1987) estimated that
24% of all the sediment transported from the land to the world ocean over the
last lOOka now resides within fjord basins. This estimate reflects both the
storage capacity of silled basins and the paraglacial delivery of first- and
second-cycle glaciogenic sediments. The enormous volumes of sediment that
lie within a fjord basin are not obvious. Saguenay Fjord, a north-shore tributary
to the St. Lawrence Estuary, contains 120 km3 of fill. The rate of sediment
delivery at the present time can account for only 7% of sediment deposition
during the Holocene. This points to an earlier paraglacial episode of high
sediment delivery (Syvitski & Praeg, 1989).
Fjord deposits reflect these high rates of sediment accumulation and
provide a highly resolvable stratigraphic record of the processes affecting
sediment delivery. Their interbasin variability reflects the wide range of
coastal and hinterland parameters affecting both sediment supply and
mechanisms of dispersal within the basin. These include glacial history (styles
and rates of glacier movements and time since deglaciation), fluvial conditions
(including seasonal distribution of runoff and partitioning of sediment between
bedload and suspended load), geography, oceanography, climate and ecology
(including basin shape, fetch, tidal characteristics, flushing dynamics,
temperature, wind, sea ice, biomass), and tectonic and geotechnical conditions
(including sea-level changes, earthquake frequency, and slope failures).
Many fjords are dominated by a large hinterland drainage basin delivering
runoff to the head of the fjord. Fluvial water and sediment discharge produce a
fjord-head delta (Fig. 10.8). The surface morphology of the delta is controlled
by discharge characteristics, tides, and wave energy, among other factors. The
land-sea interface may take the form of a braided channel complex, a single
outlet channel, several distributaries, or tidal channels dissecting mud flats.
Delta-front features can include sand and gravel barriers and river-mouth bars.
These topset deposits seldom represent more than a small fraction of the
sediment within the delta and seaward in the fjord itself, where sediment
thickness can exceed 500 m.
The geomorphology of fjord-head sandurs, such as those described from
Baffin Island (Church, 1972, 1978; Syvitski & Hein, 1991), reflects both the
abrupt decrease in sediment delivery during the early to mid Holocene and the
rapid emergence of the land resulting from isostatic rebound. As the ice sheet
melted out of the local drainage basins, both sediment availability and the
glacial meltwater component of fluvial discharge were reduced rapidly,
leaving a runoff pattern reflecting the cold, snowmelt-dominated, semi-arid
 Paraglacial coasts 387

RECENT DELTA
PROGRADATION
INTERCHANNEL HIGH
TOPSET
INSTRUMENT PACKAGE
RECENT FAILURE FORESETS POOR DATA RESOLUTION
CHUTES

0 SIDE.ENTRY FAN DELTA

20 ROCK WALLS
PARTIALLY
40 BURIED
SCARPS
60 ACTIVE
CHANNEL
ARCUATE + LlSTRlC
0 RIPPLED SAND FAULT SCARPS
100

DELTA FRONT
SANDS 300 CHANNEL WALLS
2 lo 5 m HIGH
400
PRODELTA 5w
HEMIPELAGIC VARVES
INTERLAYERED WITH GRAIN FLOW SANDS 6w
BRITTLE STAR
7w POLYCHAETE
COMMUNITY ESTABLISHED
GRAIN 800 ON PRODELTA HIGHS
FLOW SAND

Figure 10.8. ( a ) Vertical air photograph of fjord-head Klinaklini River delta and side-
entry Franklin delta in Knight Inlet, British Columbia, on the Pacific coast of Canada.
Note distributary network with bars, alluvial flats, and intertidal flats on subaerial delta.
Note also amalgamated fan morphology of Franklin delta and steep valley sidewall
channels delivering sediment directly into the fjord. Arrows denote submarine channels
that direct sand down the proximal prodelta slope. Reproduced from Syvitski & Farrow
(1989). ( b )Schematic drawing of bedload-dominated prodelta environments at head of
Itirbilung Fiord, Baffin Island. Reproduced from Syvitski & Farrow (1989).
388 D.L. Forbes & J.P.M. Syvitski

100
Ekolugod Fiord
Ilirbilung Fiord
HoI rood Pond St Mory's Boy)
60 Flor Island (St ieorge's Boy)

n
F q T ISLAND BARRIER
EKALUGAD SANDUR
E
w 20 HOLYROOD POND
BARRIER
z sill 1
0 -20
F
Q
ITIRBILUNG
2,
1 -60 DELTA
W sill 2

-100
sill 3

-140
EKALUGAD FIORD
W A L I K ARM
ice

60 SKEIMRAR
SANDUR
n

w
E 20
ATUWTIC OCEAN
Z
0 -20

W
1 -60
W

-100
V.E.= 125 bedrock

Figure 10.9. Representative topographic and bathymetric profiles for selected large-
scale systems building into deep basins (ocean to the right in all cases; V.E. = vertical
exaggeration): ( a ) Fjords and other over-deepened basins: Ekalugad Fiord, Baffin
Island, eastern Canadian Arctic: sandur and fjord-head delta building into Tasiujaq
Cove behind breached moraine (sill I), with two, more distal, sill-confined basins to
seaward in north arm of fjord (after Knight & Church, 1970; Church, 1972); Itirbilung
Fiord, Baffin Island: fjord-head delta and prodelta slope (cf. Fig. 102%); not shown are
the four sills at depths of 200, 350, 400, and 270m located 7.5, 16, 24, and 46km,
respectively, down-fjord (see Fig. 8 of Syvitski & Hein, 1991); Holyrood Pond, St.
Mary's Bay, southeast Newfoundland: large, high, sand-gravel barrier resting on fjord
 Paraglacial coasts 389

conditions prevailing at the present time (Church, 1974). The extent of active
delta progradation beyond raised early Holocene terraces reflects in large part
the size of the hinterland draining to the fjord. Trapping of sediment in lake
basins and the depth of the marine receiving basin both play a role, among
other factors, in determining the size of the modern sandur surface, but these
are of a secondary nature.
In some basin settings, wave-built ridges can be well preserved on emerging
deltas. On an actively prograding fjord-head delta, however, channel migration
and rapid progradation combine to reduce the effectiveness of wave processes
in the fjord (Syvitski & Farrow, 1983). Furthermore, the depth of the receiving
basin puts much of the fluvial sediment out of reach of wave action or tidal
currents. Thus, once the sediment reaches the seafloor, it tends to remain in
place unless removed into deeper water by submarine slides, slumps, or gravity
currents (Prior et al., 1982; Prior, Bornhold & Coleman, 1983; Syvitski &
Farrow, 1989). The suspended load is dispersed along the surface of the fjord
within a buoyant plume. The latter can extend over 100 km from the river
mouth within the restricted confines of a fjord, with corresponding effects on
the extent of hemipelagic deposits. The process can be likened to a perforated
conveyor belt, with the highest rates of sedimentation near the river mouth,
decreasing exponentially down the fjord (Syvitski et al., 1985, 1987). The
bedload is dumped along the length of the river valley with seaward decrease
in river gradient (Fig. 1 0 . 9 ~ ;Church, 1972), after which any remaining
bedload is deposited at the river mouth (Kostaschuk, 1985). Because of the
laterally confined nature of the delta front between the fjord walls (Fig. 10.8a),
the bedload accumulates as steep delta foresets within the first few hundred
metres of the river mouth. Alongshore currents are absent, a factor that
contributes (along with the high rates of bedload deposition) to both
oversteepening and overloading of the sand and gravel sediment pile. Delta-
front failures typically occur in a series of chutes, commonly 10 to 30 m wide,
that wander down the delta front (Fig. 10.8b). The chutes can provide an
indication of the failed volume. Where they are large, with significant relief

Caption for fig. 10.9 (cont.).

sill (Forbes, 1984; Forbes & Taylor, 1987; Shaw et al., 1990); fjord basin (Holyrood
Pond) to left, St. Mary’s Bay to right; Flat Island spit, St. George’s Bay, southwest
Newfoundland (Shaw & Forbes, 1987, 1990b, 1992): sand and gravel beach-ridge spit
complex and sandy barrier platform developed along the side of a deep coastal basin
(see Fig. 10.7). ( b ) Two parallel profiles about lOkm apart across Skeidarfirsandur
(Fig. 10.1I), southeast Iceland (after Boothroyd & Nummedal, 1978; Hine &
Boothroyd, 1978; Nummedal et a/., 1987a). Ice of Skeidarirjokull (Fig. 10.1 I), a
piedmont outlet glacier of Vatnajokull, is shown at upper left.
390 D.L. Forbes & J.P.M. Syvitski
(10 to 30 m), they attest to the existence of turbidity currents, which can travel
many tens of kilometres. The proximal part of the prodelta (the first few
kilometres) is often a zone of bypassing, where the gravity flows do not deposit
their sediment load but often erode the seafloor, enlarging the channels that
confine the flow (Fig. 10.8). Typically, the channel form disappears as the
seafloor slope falls below 1'. The sediment load is then deposited as the flow
decelerates and spreads across the seafloor.
Submerged fjord-head outwash deltas and shoreface platforms, such as
those found along the south coast of Newfoundland (Flint, 1940; Forbes &
Shaw, 1989), also attest to a rapid decrease in sediment supply in early
postglacial time. These features reside in an area of mid to late Holocene
submergence (Forbes, Shaw & Eddy, 1993). In the absence of significant
modern sediment discharge from hinterland basins, they now form narrow
estuaries with shallow, steep-fronted, relict-delta platforms. The lips of these
platforms are found in present water depths down to about 30 m.
Numerical models can help to elucidate the relationships between sedimen-
tation processes, environmental forcing and resulting depositional architecture.
DELTA2 is a process-response model that simulates delta growth in a fjord-
head setting (Syvitski et al., 1988). In this model, river sediment is discharged
to the coast as bedload or suspended load and is spread into the sea by a
combination of up to four processes: 1. delta-front progradation through
bedload dumping at the river mouth; 2. hemipelagic sedimentation of
suspended particles from the river plume; 3. proximal delta-front bypassing by
turbidity currents and cohesionless debris flows; and 4. downslope diffusive
processes (wave or tidal current reworking or small slides) that work to smear
previously deposited sediment into deeper water (Syvitski et al., 1988;
Syvitski, 1989). The model computes the seafloor surface at different time
intervals and the accumulation rates produced by the various depositional
mechanisms. Input data include the seasonal river discharge, suspended
sediment and bedload, the shape of the river channel, and the dimensions of the
basin to be filled with sediment. The model can be used to predict sediment fill
under complex sea-level fluctuations and over irregularly shaped basins.
Fig. 10.10 presents a DELTA2 simulation of sedimentation in Itirbilung
Fiord, Baffin Island (cf. Figs. 10.8b 10.9a), a bedload-dominated basin. Input
parameters were taken from Syvitski & Hein (1991) and include two periods
of sediment delivery: 1. the Hypsithermal period between 7.0 and 5.6ka BP, a
time of warm dry summers, when valley glaciers from the Foxe Ice Sheet
ablated and receded rapidly (Church, 1978); and 2. the Neoglacial period from
5.6 to 0.1 ka BP, a moister time but with much cooler summers. The
Hypsithermal condition was represented using an annual bedload transport
 Paraglacial coasts 39 1

600 SLUMP EVENT

500

SEAFLOOR
400 EROSlON
h SEAFLOOR
YE AFTER
I- 400 YEARS
I 300
(3
W
I
200
ORIGINAL
BASIN SHAPE
100

0
0 2 4 6 8 10 12
DISTANCE FROM ORIGINAL RIVER MOUTH (km)
Figure 10.10. Simulation of delta growth and fjord-basin sedimentation, showing the
Holocene record of successive seafloor profiles every 400 years in Itirbilung Fiord,
Baffin Island (for modelling details see Syvitski & Daughney, 1992). Two distinct
climatic intervals, involving different fluvial discharge and oceanographic conditions,
are reflected in the simulated sediment record: 1. an early paraglacial episode of high
sediment supply lasting 2400 years; 2. a 'postglacial' interval of lower sediment input
continuing for 5600 years (see text for further details).

(sand) amounting to OSMt, primarily in the spring freshet, and annual

suspended load (mud) of 0.17 Mt, introduced from the spring through to the
autumn. The early paraglacial influx was associated with an exponentially
decreasing drop of 5 0 m in relative sea level (primarily resulting from
glacioisostatic rebound). The Neoglacial period had much lower rates of
sediment input (bedload and suspended load estimated at 0.016Mt a-' and
0.013 Mt a-', respectively), occurring largely in the late spring and early
summer. During this time, relative sea level fell in a roughly linear fashion
over a range of 10m, and the delta front prograded slowly into a much more
energetic basin (more open water, more wave activity, and stronger currents).
The resulting simulation (Fig. 10.10) shows the computed profile of the
seafloor along an axial line normal to the delta front at 400-year intervals. The
model output shows the rapid delta-front progradation during the Hypsithermal
interval and the concomitant generation of sediment gravity flows into an
392 D.L. Forbes & J.P.M. Syvitski
otherwise quiet basin. The transition from Hypsithermal to Neoglacial is
associated with a major adjustment of the delta front in the form of a large
slump. This event temporarily slows the progradation of the delta and causes
the formation of a terrace. Increased deepwater currents during the Neoglacial
period cause subsequent erosion of hemipelagic deposits that mantle some of
the bedrock highs in the outer reaches of the fjord. High-resolution seismic
reflection profiles collected along the axis of Itirbilung Fiord show
stratigraphy similar to that produced by the model (Syvitski & Hein, 1991).

Exposed outwash coasts

The difference between deeply embayed basins receiving sediment from active
or relict glacial outwash sources (the fjord-head deltas described above) and
partially embayed or open coasts fronting outwash plains is superficially
striking, reflecting the influence of wave processes in unprotected settings.
Under these circumstances, beach and barrier facies develop along the sandur
front (Htquette & Ruz, 1990; Shaw, Taylor & Forbes, 1990). Distributary
migration and longshore dispersal of sediment under waves may contribute
jointly to the maintenance of a relatively simple shoreline form. Although
minor sinuosity may initiate some longshore cell structure, this is inhibited by
the large sediment supply in active proglacial environments.
The sandur coast of southeast Iceland (King, 1956; Hine & Boothroyd,
1978; Nummedal, Hine & Boothroyd, 1987a) is an eminent example, perhaps
the largest active feature of its kind. It represents proglacial conditions of very
high sediment supply from piedmont glaciers at the southern margin of
Vatnajokull (Fig. 10.11). The sediment flux (typically about 0.4 X 106tons per
year) is enhanced by volcanogenic glacier-burst (iokulhlaup) floods that
generate discharges as high as 100000m3s-', total flow volumes ranging up
to 15 km3, and sediment transport estimated at about 29 X 106tons for a burst
discharge of 3.5 km3 in 1954 (Rist, 1957; Nummedal et al., 1987a). These
conditions have produced extensive sand and gravel outwash (sandur) plains
filling broad depressions in the underlying Tertiary and Quaternary basaltic
rocks. The SkeiiSarksandur (Fig. 10.9b) has built seaward to a broad front 20
to 30km south of the present ice margin and covers an area in excess of
1000km2. Braided channel networks cross the sandur (Boothroyd &
Nummedal, 1978), attaining widths as great as lOkm (SkeiiSarB) at the
shoreline, where they break through the barrier beach at a number of
distributary outlets. The subaerial slope of the sandur, 0.010 (0.57') or more
near the ice margin, decreases to less than 0.003 distally. The much steeper
shoreface slope, up to 0.026 (-5') off the SkeiiSarB outlet (Fig. 10.9b),
 Paraglacial coasts 393
24' 22' 20' 18' 16' 14'

660
17°00s
65'

640

kin 100
64"
00'

SKEIDARAR
SANDUR

ING6LFS-
HOFDI

ATLANTIC OCEAN
BEDROCK ICE FIELD OUTWASH PLAIN BARRIER SPITS/ISLANDS

Figure 10.11. Outwash plain of SkeiBarBrsandur, southeast Iceland, and barriers

developed along its seaward margin (redrawn after Nummedal et al., 1987a).

diminishes westward to 0.020 (-1') and decreases seaward to 0.006 at the

100 m isobath. Maximum clast size (A-axis) decreases down the SkeiaarA
channel system from about 320 mm near the ice front to approximately 10 mm
near the coast (Fig. 6 of Boothroyd & Nummedal, 1978); the dominant
sediment type over the distal sandur (below about 5 0 m elevation) is sand.
Seismic refraction indicates that depth to bedrock at the shoreline is about
lOOm at the mouth of SkeiBarA, rising to sea level lOkm to the east at
Ing6lfshofai (Figs. 10.9b and 10.11; Nummedal et al., 1987a).
This coast is subjected to a highly energetic wave climate resulting from
northeastward movement and intensification of winter storms, which generate
southwesterly swell followed by higher waves (significant wave height, up to
16 m) from the southwest through southeast (Nummedal, 1975; Hine &
Boothroyd, 1978; Vigg6sson et al., 1994). The waves generate substantial
longshore drift during the winter months, reworking the seaward margin of the
outwash plain to produce a set of broad sandy barrier islands. These are
delimited by distributary outlets and impound wide wind-tidal flats. The
barriers increase in width from 200m to about 750m to the west (Hine &
394 D.L. Forbes & J.P.M. Syvitski
Boothroyd, 1978). They are characterized by a broad storm ridge or berm
rising to approximately 5 m, supporting widespread aeolian bedforms
(transverse dunes), fronted by a lower berm during the summer months, and by
nearshore bars off the major distributaries (Fig. 10.9b; Hine & Boothroyd,
1978; Nummedal et al., 1987a). Despite the great quantity of sediment
delivered to the coast by meltwater runoff and glacier bursts, short-lived
progradation of about 200 m is the maximum observed in association with any
one event off Skeibarti; over the past century, the mean position of the
shoreline has been essentially stable in that area (Nummedal et al., 1987a).
This implies rapid reworking of littoral sediment alongshore and downslope,
consistent with the widening barriers and broader submarine slope to the west
(Fig. 10.9b). In terms of the large-scale architecture of the system, however,
wave-formed structures are of minor importance. The sandur deposits
constitute a very large, seaward-prograding, coarsening-upward clastic wedge,
prograding seaward across a broad front. The large barrier-beach complex
form$ only a minor surface decoration.
The Skeibartirsandur has formed in a laterally unconfined, high-wave-
energy, and recently stable sea-level setting. The outwash coast of the Gulf of
Alaska is a roughly analogous case (Boothroyd & Nummedal, 1978). Where
relative sea levels are rising and topography provides some lateral
confinement, the resulting morphology can include transgressive barrier
structures behind which tidal estuaries may develop (Fig. 3 of Shaw et al.,
1990). Where relative sea level has been falling throughout most of the
Holocene, the coastal morphology may be dominated by superficial reworking
of the outwash surface into a succession of quasi-parallel beach ridges
(Fig. 10.4d; Blake, 1975), locally amalgamated or separated by shallow
depressions, the morphology of which records short-lived details of longshore
transport, coastal cell development, spit extension and recurve development,
and other common features of sediment-limited gravel beaches (Fig. 10.4b;
HCquette & Ruz, 1990).

Basin-margin coasts with large sediment supply

The coast of St. George’s Bay, southwest Newfoundland, provides a case
study of coastal evolution in the presence of a large paraglacial sediment
supply close to a former ice margin and glacially overdeepened coastal basins.
The shores of the bay are marked by extensive coastal bluffs, typically 20 to
3 0 m high, partially stabilized in some parts and actively eroding in others.
These eroding sections reveal thick sequences of ice-contact diamicts,
proximal outwash sands and gravels, and glaciomarine muds of Late
 Paraglacial coasts 395
Wisconsinan age (MacClintock & Twenhofel, 1940; Brookes, 1974, 1987;
Grant, 1987, 1991). They represent the most-extensive exposure of
glaciogenic sediments to be found on the Newfoundland coast. Erosional
bluffs extend for lOkm along the north shore of the bay and 40 km to the
southwest along the south shore (Fig. 10.7). Along with a depositional
complex at least 90m thick (probably thicker over buried valleys), forming a
broad sill across the middle of the bay, and even thicker sequences within the
valleys, these deposits record a major late-glacial depocentre in the area (Shaw
& Forbes, 1990b). Two basins landward of the sill, the deepest extending to
more than 90 m water depth under present conditions, are underlain by
partially filled bedrock depressions with the appearance of subglacial tunnel
valleys. These contain thick stratified sequences overlying acoustically
unstratified sediments. The top of the unstratified unit occurs as deep as 190 m
below present sea level, beneath the present-day beaches at Stephenville and
Stephenville Crossing (Fig. 10.7).
Relative sea level stood at 44m above present at the time of initial ice
recession about 13.7 ka BP (Brookes, Scott & McAndrews, 1985; Grant, 1991),
but dropped to a minimum at about 25 m below present c. 9.5 ka BP (Forbes
et al., 1993). St. George’s Bay is open to the prevailing wind from the
southwest, with potential fetches in excess of 700 km across the Gulf of St.
Lawrence. Wave-formed ripples with wavelengths up to 3 m developed in
pebble gravels (modal grain sizes of 2 to 29mm) over a large area of the sill
(Fig. 15 of Shaw & Forbes, 1990b), attest to wave conditions at least
equivalent to Hs>3.8m at peak periods of 11 s (cf. Forbes & Boyd, 1987).
Significant wave heights of this magnitude have a return interval of less than
1 year (Ouellet & Llamas, 1979), although wave energy is significantly limited
by sea ice from late December to early May (Forbes & Taylor, 1994; see also
Chapter 9).
Along the southeast side of St. George’s Bay, the 12 km long Flat Island
barrier (Fig. 10.7) forms a large, drift-aligned, sand and gravel spit complex
(Shaw & Forbes, 1987, 1992). It rests on a sandy subaqueous barrier platform,
partially filling the adjacent basin and providing a foundation for subaerial
barrier growth. The platform has a pronounced break in slope (Figs. 1 0 . 9 ~and
10.12) that rises to 5 m water depth at the distal end of the spit, converging
distally on the beach from an initial width of 2 km and depth of 25 m at the
proximal (southwest) end. This platform is analogous to spit platforms
described elsewhere, in particular one from a raised paraglacial shoreline
sequence in Denmark (Nielsen, Johannessen & Surlyk, 1988), but is larger and
thicker than other reported examples. It consists of a clinoform prism of sand,
up to 50m thick, draped along the south flank of the southern basin in St.
396 D.L. Forbes & J.P.M. Syvitski
DISTANCE (kilometres)
5 1.0 1.5 2 .o 2.5
O0

.02
20

.04 -v)

-
v)
40
.06 o
U
C

0
al m
L

60 .08 -
W
lil

J
w
21 22 o z
20 I-
.02 >
a
B
B A R R I E R PLATFORM .046
3
!-
G LAC IOMA R I N E .06

19
60 .06

.I 0
80
GAS

20
FLAT ISLAND
FIAT BAY
0 MWL

3 20
a
t;
5 4c
km
60

80

Figure 10.12. Seismic reflection (sparker) profile through outer part of Flat Island
barrier platform (Figure 10.7) and underlying glaciomarine sediments in an area of
subaqueous slope failure on the seaward face of the platform. Note seaward-dipping
clinoform structures underlying platform surface at right and slump structures on the
seaward slope. Numbers on profile refer to sediment samples. Grain sizes range from
muddy sand with a mode of about 3 I$ (mean 4 I$) at sample 19 to well-sorted medium
sand (modal size about 1.6 I$)on the upper shoreface at sample 22. Bottom panel shows
a cross-section through the Flat Island barrier and subaqueous platform at a much-
reduced vertical exaggeration relative to the same profile in Figure 1 0 . 9 ~Reproduced
.
from Shaw & Forbes (1992).
 Paraglacial coasts 397
George's Bay. The seaward-dipping beds within the platform have typical dip
angles of about 6". The seaward face in an area showing evidence of recent
failure (Fig. 10.12) has a slope of about 14" at the top, running out to a gentler
slope of 3" at the base and the sediment size ranges from medium sand on the
upper shoreface to muddy fine sand on the lower slope (Shaw & Forbes, 1992).
Gravel occurs on the beach and in the trough of the nearshore bar system. The
subaerial barrier (Fig. 10.13a) consists of three distinct segments (Shaw &
Forbes, 1987):
- a narrow proximal section that widens downdrift to include a number of
partially truncated beach ridges aligned at an angle to the present shoreline,
with thin salt marsh accumulating along the landward shore and in
depressions between the gravel beach ridges;
- a central section that was breached by extensive washover during the early
1950s, developing a wide flood-tidal shoal on the landward side and
interrupting the littoral drift;
- a broad distal section, more than 1 km wide, consisting of several discordant
sequences of gravel beach ridges capped with aeolian sand.
Another platform structure is present along the north shore of St. George's
Bay, extending across the northern valley at Stephenville, around a rocky
promontory (Indian Head, Fig. 10.7), and feeding sediment to the barrier at
Stephenville Crossing. This barrier rests on a submerged early Holocene delta
sequence graded to a sea level about 25 m below present (Fig. 6 of Shaw &
Forbes, 1990b). In the Stephenville embayment north of Indian Head, the
barrier platform has provided the foundation for progradation of a gravel
beach-ridge complex almost 4 km in length and 1 km wide (Grant, 1975; Shaw
& Forbes, 1992). Yet another large clinoform sediment body has developed on
the landward side of the sill, building landward into the inner basins. This
spillover unit results from reworking of sediments primarily at lower sea levels
(most of the sill was subaerial for a time in the early Holocene), but continues
to grow under present conditions of active reworking on the sill. The spillover
wedge has a steep basin face with a slope of 10" to 20°, is 20 to 50 m thick,
consists of landward-dipping units of sand and gravel overlying glaciomarine
muds in the basin, and extends for more than 15 km across the bay on the inner
side of the sill (Shaw & Forbes, 1992).
Coastal sediment complexes can develop on shallow or emergent fjord sills,
impounding lakes or estuaries (cf. Pickrill, Irwin & Shakespeare, 1981). The
barrier at Holyrood Pond in southeast Newfoundland (Fig. 10.9a; Forbes,
1984, 1985; Shaw et al., 1990) is an example of beach development on a
shallow fjord sill. This sand and gravel barrier, more than 200m wide and over
 Next Page
398 D.L. Forbes & J.P.M. Syvitski

Nearshore L q Shoal (CobMo-BwMer)

k t N e Beach (sand6 OrOV94
Vegetated dunes (Send)
Bosch-FlidpeComplex (sand6 Gravel)
san Msnh
FM-Tidal Dels (Sand6 Wnvel)
Active CM
Relict CIM

Figure 10.13. Beach-ridge morphology showing evidence of progradation (drawn from

air photographs at various scales): (a) Proximal and distal sections of Flat Island barrier
in St. George’s Bay (after Shaw et al., 1990; see Fig. 10.7). (b) Prograded beach-ridge
complex being overrun by transgressive storm ridge, Shoal Bay, southeast Newfoundland
(see Forbes, 1984). ( c ) Beach-ridge pattern at Lawrencetown Beach (Fig. 10.15) on the
Eastern Shore of Nova Scotia, showing early progradation and evidence of trailing spits
from former drumlin anchor off the middle of the present beach (see Boyd et al., 1987;
Forbes et al., 1990).
 11
Coastal cliffs and platforms
G.B. GRIGGS AND A.S. TRENHAILE

Introduction
Our individual perceptions of the coastline are clearly related to our local
coastal geomorphology. On global and regional scales coastal morphology
often correlates closely with tectonic setting (Inman & Nordstrom, 1971).
Frequently, major geomorphic contrasts exist between the subdued coastlines
along passive continental margins and the rugged coastlines along convergent
plate boundaries. Along most coastlines, modern (active) coastal landforms are
similar to their Pleistocene counterparts, which suggests that current tectonic
and coastal processes have been fairly uniform over extended periods of recent
geological time (Lajoie, 1986; see also Chapter 12).
Many exposed coastlines along passive continental margins, such as the
southeast and Gulf coasts of the United States or parts of the northeast coast of
Australia, for example the Gold Coast, consist of a low-relief coastal plain,
bordered offshore by a wide continental shelf (see Chapters 2 and 4). These
relatively stable coasts are also characterized primarily by meso- and
macrotidal conditions, energy dissipation and depositional landforms, such as
wide sandy beaches and dunes (see Chapter 4), as well as offshore barrier
islands or bars (or a barrier reef in the case of northeastern Australia; see
Chapter 8).
In marked contrast to the Atlantic coast of the United States and Canada, the
Pacific Coasts of North and South America are active margins, and are either
present or former collision coasts, where two plates have collided. Even a
casual glance at almost any individual segment of this coastline will
immediately reveal striking differences in coastal landforms and geological
history. In addition to seismicity and vulcanism, collision coasts or active
margins are characterized by deep-sea trenches offshore, narrow continental
shelves, coastal mountains and often uplifted marine terraces. Erosional
landforms such as steep sea cliffs, rocky headlands, sea stacks and islands

425
426 G.B. Griggs & A.S. Trenhaile
produce a very different coastline than many trailing-edge coasts with their
depositional landforms. Thus, coastal cliffs and platforms, the focus of this
chapter, commonly have primary origins which are related directly to their
large-scale tectonic setting.
Having made these generalizations, and recognizing the importance of
tectonic setting (which was first recognized by Suess (1888) over a hundred
years ago), it is important to point out that rugged coasts with rocky cliffs are
not restricted to collision coasts. There are high cliffs, for example, around the
British Isles, and in northern France, southeastern Australia and eastern
Canada. All of these coastlines would be appropriately termed passive margins
or trailing edges. The presence of rocky cliffs, sea stacks and arches is very
common along these shorelines in part because the structural grain of the land
is often at a high angle to the coast, which leads to exposure of rocks of
variable resistance alongshore. Additionally, however, factors such as limited
alluvial sediment contributions or loss of material into estuaries or
embayments may lead to the development of rocky cliff coasts. Although the
structural grain of collision or Pacific-type coasts, in contrast, is parallel to the
shore, the tectonic activity and uplift which characterizes these coastlines
typically produces the sea cliffs, terraces, headlands, and other characteristic
features enumerated above.
Coastal cliffs, as a result of their tectonic setting, structural and lithologic
framework and recent geologic history, vary widely in their height and
morphology. Although it is common to think of cliffs as being solely a product
of marine erosion, terrestrial or subaerial processes can be equally important in
affecting the landforms exposed along the shoreline at any particular location.
In most such locations it is the combination of marine and terrestrial processes
which creates the distinctive coastal slopes.

Rock coasts
A large proportion of the coasts of the world is rocky, and even many sandy
and cobble beaches are backed by rock cliffs or underlain by shore platforms.
The prominence of such features is not reflected, however, in the modern
process-orientated coastal literature, where most emphasis has been placed on
beaches and other systems that respond rapidly to changing environmental
conditions. While chemical and physical analyses, geochronometric dating,
physical and mathematical modelling, and careful measurement of processes
and erosion rates are providing valuable insights, we can still often only
speculate on the mode of development of rock coasts.
 Coastal cliffs and platforms 427
Processes
Rock coasts are the legacy of marine and subaerial processes that have been
operating for thousands of years (Trenhaile, 1987). The type, intensity, and
focus of these processes have varied with shifts in relative sea level, and with
temporal and spatial changes in climate, exposure, and rock type. Although
some attempt has been made to identify and measure the results of the
processes operating on rock coasts, scientists are still largely ignorant of their
precise nature and relative importance. This is unfortunate, as coastal cliffs are
being populated at an increasing rate as more people in the world’s developed
countries move to the coast. The study of processes on rock coasts, especially
the acquisition of quantitative data, has been hindered by the very slow rates of
change, the role of high-intensity - low-frequency events, exposed and often
dangerous environments for wave measurement and submarine exploration,
the lack of access to precipitous or heavily vegetated cliffs, poor research
funding, and the small body of active researchers in this area. In any case, even
if we completely understood the nature of contemporary erosive processes, it
would still be difficult to explain the morphology of coasts that often retain the
vestiges of environmental conditions that were quite different from today.

Wave action
Mechanical wave action is the primary erosional agent in most storm-wave and
more vigorous swell-wave environments. In polar, tropical and other relatively
low-energy environments, where waves are generally much weaker, they still
have an important role in removing the products of physical and chemical
weathering, simply because they are the dominant source of energy for
transporting material in the coastal zone.
Mechanical wave erosion is accomplished by a number of processes, but
we usually have to gauge their relative importance from ambiguous
morphological evidence. The presence of large, angular debris and fresh rock
scars has convinced many workers that wave quarrying is usually the dominant
erosive mechanism in vigorous wave environments. Wave quarrying usually
requires the alternate presence of air and water. Water hammer, high shock
pressures generated against structures by breaking waves (up to 690 kPa has
been recorded), and probably most importantly, the compression of pockets of
air in rock crevices, are most effective in a narrow zone extending from just
below the still water level up to the wave crest (Sanders, 1968). Although
abrasive processes are not as closely associated with the water level, their
efficacy rapidly decreases with depth below the water surface.
428 G.B. Griggs & A.S. Trenhaile
Abrasion or corrasion takes place as sand or other coarse granular material
is washed, dragged or rolled across a rock surface or when this material is
thrown against a coastal cliff or bluff. The importance of this process at any
particular site is related to the presence of tools (sand, gravel or pebbles) and
the wave energy available to move these tools around. Abrasion tends to
produce much smoother surfaces than the quarrying processes described
above, although where lithologic inhomogeneities or structural weaknesses
occur, abrasion can produce local scouring or grooving of the bedrock. An
extreme example of this focused abrasion is in the formation of potholes where
large pebbles or cobbles abrade or carve circular depressions on intertidal rock
platforms. The abrasional process is limited in its shoreward extent by the
elevation to which waves can throw or wash material up against the flanking
seacliff. Abrasion is most effective where the sediment cover in the littoral
zone is sufficiently thin (up to perhaps 1Ocm) such that the available wave
energy is able to agitate or dislodge particles down to the underlying bedrock
(Robinson, 1977).
Most mathematical models suggest that the pressures exerted on vertical
structures by standing, breaking, and broken waves are highest at, or slightly
above, the still water level (Trenhaile, 1987). The zones of maximum wave
pressure and most effective erosional processes are therefore essentially
coincident, occurring at or near to the water surface. Consequently, the long-
term level of greatest erosion must be related to the neap high and low tidal
levels, the elevations most often occupied by the still water level. Wave energy
is probably greater during the high tidal periods because of greater water
depths and reduced friction. The highest and most effective waves occur
during storms, however, when the water surface is raised above its still water
or tidal level. It is at these times that wave impact and runup reach farthest
inland and the uppermost reaches of the shoreline or seacliff are attacked most
vigorously. The zone of greatest wave erosion is therefore probably above the
neap high tidal level, particularly in microtidal environments and where hard
rocks resist all but the most vigorous storm waves that operate at elevated,
supratidal levels.

Chemical and salt weathering

Cliffs and intertidal shore platforms subjected to alternate wetting and drying
by ocean spray, splash, and tides are suitable environments for many chemical
and physical weathering processes. They are probably only major erosive
mechanisms in sheltered areas and on particularly susceptible rocks in cooler
regions, but they assume an important and sometimes dominant role in warmer
climates.
 Coastal cliffs and platforms 429

Chemical and salt weathering tend to operate together and it is usually very
difficult to distinguish the results in the field. They contribute to the general
weathering of cliffs and shore platforms, the development of tafoni and
honeycombs, and to the suite of processes collectively referred to as water
layer levelling (So, 1987; Matsukura & Matsuoka, 1991). There is no
evidence, however, of permanent saturation above the low tidal level, or of its
corollary, an abrupt transition in the intertidal zone from less-resistant,
weathered rock above the saturation level, to lower, more-resistant,
unweathered rock below; concepts that have dominated the Australasian
literature on shore platforms for almost a century (Bartrum, 1916).
The nature and efficacy of the destructional processes operating on
limestone coasts are contentious issues. Although surface sea water is
normally saturated or supersaturated with calcium carbonate, marine karren in
the intertidal and splash-spray zones of limestone coasts is similar to the
variety of sharp pinnacles, ridges, grooves and circular basins etched on land
by meteoric water. Carbonate solution may occur in pools at night, however,
when faunal respiration, which is not taken up by algae during the hours of
darkness, reduces the pH of the water and causes calcium carbonate to be
transformed into more soluble bicarbonate. Nevertheless, other biochemical
processes may inhibit or prevent solution through the coating of rock surfaces
with dissolved organic substances, and the building, by organic substances, of
complexes with calcium ions. It has been estimated that solution is responsible
for about 10% of the total erosion of the coast of Aldabra Atoll in the Indian
Ocean (Trudgill, 1976), but alternative bioerosional mechanisms are more
important in western Ireland (Trudgill, 1987) and on Grand Cayman Island
(Spencer, 1988), and chemical solution does not even occur in the rock pools
of the northern Adriatic (Schneider, 1976).

Bioerosion
Many workers have argued that coastal ‘solutional’ features are primarily the
result of bioerosional activities by a variety of marine organisms. The removal
of the substrate by direct organic activity is probably most important in tropical
regions, where there are an enormously varied marine biota and calcareous
substrates that are particularly susceptible to biochemical and biomechanical
processes (Spencer, 1988; Fischer, 1990).
Algae, fungi, and lichen are pioneer colonizers in the inter- and supratidal
zones. In addition to their role as rock borers and their effect on water
chemistry, microflora allow subsequent occupation by gastropods, chitons,
echinoids and other grazing organisms. Grazers abrade rock surfaces by
mechanical rasping as they feed on epilithic and endolithic microflora. In the
430 G.B. Griggs & A S . Trenhaile
mid-tidal zone of Aldabra Atoll, for example, grazing organisms are
responsible for about one-third of the surface erosion where sand is available
for abrasion, and as much as two-thirds where sand is absent (Trudgill, 1976).
Other erosion is accomplished by barnacles, sipunculoid and polychaete
worms, gastropods, echinoids, bivalve molluscs, Clionid sponges and
numerous other rock borers, particularly in the lower portions of the intertidal
zone. Borers directly remove rock material, but also, by enlarging crevasses
and creating weaknesses in the rock structure, they render the residual rock
more susceptible to wave action and, by increasing the surface area, to
weathering.
There is a fairly large body of information on bioerosional rates, but its
reliability and applicability vary enormously (Trenhaile, 1987; Spencer, 1988).
Nevertheless, the overall erosion rate on both vertical and horizontal limestone
surfaces in a variety of environments, has often been found to be between
about 0.5 and 1 mma-’, which may reflect the maximum boring rate of
endolithic microflora (Schneider & Torunski, 1983).

Expansion-contraction mechanisms
The expansion and contraction of certain clay minerals such as the
montmorillonitehmectite group is probably an important erosive mechanism
in shales and other argillaceous rocks subjected to tidally induced cycles of
wetting and drying. Clay-rich rocks can also be weakened by temperature-
dependent wetting and drying resulting from the attraction of the positively
charged ends of water molecules to the negatively charged surfaces of clay
particles contained within small rock capillaries.
However, we neither know the extent to which the traditional field evidence
for freeze-thaw action can be explained by the temperature-dependent
adsorption of water, nor do we understand the precise nature of such ‘frost’
action. The two mechanisms may act together, the suction generated by
unfreezable adsorbed water, trapped in rock pores by expanding ice, causing
free or bulk water to migrate to the freezing front. The pressures generated by
this mechanism may be a more-important cause of frost shattering than the
approximately 9% expansion of water upon freezing (Matsuoka, 1988).
Although we are still unsure of the responsible processes, we do have a
general sense of the conditions most suitable for their operation. Cool coastal
regions appear to be almost optimum environments. Rocks in the intertidal and
spray and splash zones can attain high levels of saturation (Trenhaile &
Mercan, 1984), and they experience many more frost cycles than those further
inland. More than 200 freeze-thaw alternations can occur in a year in the
 Coastal cliffs and platforms 43 1

intertidal zone in coastal Maine, for example, compared with 3 0 4 0 inland

(Kennedy & Mather, 1953). Tidally induced frost cycles occur when rocks are
alternately frozen when exposed to freezing air temperatures, and thawed
when submerged in water above freezing. Rocks thaw very rapidly when
inundated by rising tides, but they may need several hours to freeze once they
have been exposed by falling tides (Robinson & Jerwood, 1987). Whether
critical saturation levels can be maintained in the rocks over this period
remains to be determined.

Mass movement
Fresh rock faces and accumulations of debris attest to the importance of rock
falls on many coasts (Fig. 11.1). They are more common than deep-seated
slides, but most are fairly small. Falls occur in well-fractured rocks, especially
over notches cut into the cliff foot by waves, chemical solution, or bioerosional
agencies. Large coastal landslides require suitable geological conditions
(Grainger & Kalaugher, 1987). Argillaceous and other easily sheared rocks
with low bearing strength are particularly susceptible to translational sliding,
as are alternations of permeable and impermeable strata, and massive rocks
overlying incompetent materials (Pitts, 1986). Rotational slumps usually occur

Figure 1 1.1. Rockfall from coastal bluffs in central California along shore-parallel and
shore-normal joint sets in sedimentary rocks (note person at lower right for scale).
432 G.B. Griggs & A S . Trenhaile
in thick, fairly homogeneous deposits of clay, shale or marl (Flageollet &
Helluin, 1987). Most deep-seated mass movements may be attributed to
undercutting at the base of the slope and the build-up of groundwater. They
therefore tend to take place during or shortly after periods of prolonged and/or
intense precipitation, or as a result of the accumulation of water from human
activities (including septic systems, runoff disruption, and landscape
irrigation). Coastal cliff failures may therefore be rather instantaneous events
in the case of rock falls due to wave undercutting, or seismic shaking, or may
take place slowly in the case of deep-seated mass movements initiated by
excess groundwater.

Shoreline erosion and retreat

Factors affecting the form and rate of shoreline erosion

The rate of coastal cliff erosion and the resulting landforms are related to both
the cliff-forming materials and the physical processes to which they are
exposed, as well as the time over which these processes have operated. The
tectonic and climatic histories of any particular coastline are additional
variables which have major effects on the relief, the rock types, stratigraphy
and structure present, as well as weathering processes and vegetation
development.

Rock type
The type of rocks exposed in coastal cliffs or platforms is an extremely
important factor in determining their rate of erosion. In general, crystalline
rocks (igneous and metamorphic) are more resistant to marine and subaerial
erosion than sedimentary rocks, and consolidated or cemented sedimentary
rocks are more resistant than those which are unconsolidated. Where rocks of
variable resistance are interbedded (sandstones and shales, for example) and
exposed on the shoreline, we can expect the erosion rates and resulting
landforms to express these differences. Although it is difficult to separate the
importance of lithologic variables from other factors controlling shoreline
erosion, such studies are needed in order to quantify the importance of these
individual variables. Within a distance of 60 km along the shoreline of central
California, a variety of materials including unconsolidated Pleistocene dune
sands, Pliocene sandstones, Miocene mudstones, and Mesozoic granites are
exposed in the coastal bluffs. Erosion rates in these materials range from
2.5 m a - ' in the dune sands to imperceptible rates in the granites.
 Coastal cliffs and platforms 433
Structural weaknesses
Internal structural weaknesses, typically joints, fractures and faults, exert a
primary control on both the morphology and rate of cliff erosion in many areas.
These weaknesses provide zones of accelerated wave erosion which often
account for the presence of arches and caves in these cliffs (Fig. 11.2).
Although it appears that this process is often underappreciated, it is abundantly
clear that the overall planform of many coastal cliffs, as well as the smaller
scale irregularities, are often intimately tied to joint orientation and spacing.
While a cliff may appear to consist of a massive or thick-bedded mudstone or
siltstone, closer inspection reveals that it is the presence of joints, which may
owe their origin and orientation to the recent tectonic history of the area, which
ultimately determines the size and geometry of the blocks which fail or are
dislodged from the cliff. Even the weakest, but mechanically intact and
unweathered, rock could theoretically withstand pressures generated by cliffs
over a thousand metres in height. These heights are never attained, however,
because the critical height is limited by the presence of joints, faults, and other
mechanical defects, rather than the integral strength of the rock itself
(Terzaghi, 1962).

Figure 11.2. Caves eroded at base of cliffs by wave action along structural weaknesses
in sedimentary rocks along the central California coast.
434 G.B. Griggs & A S . Trenhaile
Along the shoreline of northern Monterey Bay, California, the young
sedimentary rocks (dominantly mudstones and siltstones) are jointed in both
shore-parallel and shore-normal directions. Failure typically occurs as large
(up to 3 m thick) joint-bounded blocks fall to the beach below following
undercutting of the base of the bluff by wave action (Fig. 11.1). In this
instance, the strength or the degree of cementation or consolidation of the
bedrock itself (as determined by typical soil or rock engineering methods) is of
considerably less significance in affecting the rates of shoreline erosion and the
resulting cliff configuration, than is the distribution and orientation of the joint
sets and the degree to which they are developed.

Exposure to wave action

A combination of the regional wave climate or severity and frequency of wave
attack, and the degree of exposure of the sea cliff to waves, is of fundamental
importance in determining the rate of cliff retreat. Because of the irregular
nature of storm-wave attack, we typically witness cliff retreat as an episodic
and largely unpredictable process. Little erosion or retreat may occur for years
and then, under the right combination of cliff weakening or weathering
(perhaps due to excess groundwater or pore pressures) and wave attack and
tidal elevation, major failure and significant erosion may take place virtually
overnight.
Local variations in nearshore and offshore topography often serve to focus
or disperse wave energy such that adjacent sections of coast may experience
widely varying degrees of wave attack and therefore erosion rates. Within
central Monterey Bay on the California coast, for example, the head of
Monterey Submarine Canyon extends nearly to the shoreline. Wave energy is
refracted away from the canyon head, due to its greater depths, such that wave
heights are typically several times greater immediately north or south of the
canyon than in the area immediately landward. Similarly, the distribution and
relief of nearshore bedrock platforms leads to a complex pattern of wave attack
at the shoreline.
The presence or absence of a protective beach, whether permanent or
seasonal, may provide the most important buffer to a cliff from direct wave
attack. There are those areas where no beach ever forms, due to a combination
of offshore topography, relief of the coastline, resistant nature of the rocks, and
lack of sufficient sand. In these instances, wave attack of the cliffs is a year-
round process. In other locations, beaches exist seasonally, and cliff exposure
to wave attack is only intermittent. There are also those sites where, either due
to recent uplift or isostatic rebound or the presence of a very wide equilibrium
beach, cliff attack is relatively infrequent, despite the wave conditions.
 Coastal clifls and platforms 435
Relative sea-level rise
The relationship between sea-level rise and the uplift or subsidence of the
coastline and the resulting landforms will be covered more thoroughly in the
next section, on marine platforms and terraces. It is important to note here that
the relative rise or fall in sea level can exert a major impact on the nature of the
coastal cliffs. For example, where tectonic uplift or isostatic rebound is
occurring at a more rapid rate than eustatic sea-level rise, then the coastal cliffs
can be isolated from wave attack and preserved as relict features. At this stage
subaerial processes may act to degrade the slope. Conversely, where the
relative sea-level rise is positive, the cliffs should be reached more frequently
by waves and therefore subject to greater erosion rates.
Although many factors can influence sea level, leading to a noisy record,
various studies utilizing tide-gauge data find an average global rate of sea-level
rise of 1-2mma-', over the last 100 years (Gornitz, 1991). Along the coast of
southeast Alaska in the Juneau and Skagway areas, however, due to postglacial
rebound, the sea level is dropping relative to the coastline at rates of 13 mm a- '
to over 17mma-'. Coastal cliffs are thus being uplifted and preserved at a
relatively rapid rate. The northern British Isles were covered with ice during the
glacial episodes of the Pleistocene, and were correspondingly depressed, while
southern England was ice free. Postglacial rebound has produced a regional
tilting, with Scotland rising in the north while the south coast of England is
sinking, leading to slow inundation. Relative sea-level change thus varies
significantly from north to south and has produced differing effects on the
coastal cliffs.

Tidal variation
Global differences in daily tidal ranges vary from nearly imperceptible to over
12 metres. The range in elevation between high and low tides determines the
extent of the coastal cliffs or bluffs which are alternately exposed and covered
with seawater and impacted by waves. Through the processes which create
intertidal platforms and terraces, this factor is potentially very significant, as
discussed in the next section. Water table weathering to the zone of permanent
saturation, which includes the softening and weakening of the rocks through
weathering and solution, allows for the erosional processes (including
hydraulic impact) to be effective over an increasingly broader zone as the
vertical range between high and low tide increases.

Human activity
As coastal populations worldwide have migrated to the coastline, either
seasonally or permanently, the impacts of those humans and their development
436 G.B. Griggs & A.S. Trenhaile
have begun to alter both coastal landforms and cliff stability. Buildings,
utilities and coastal protection structures have been built on cliff tops, on the
faces of the cliffs themselves, as well as on the fronting beach. There are many
densely populated coastal areas where little of the natural cliffs can any longer
be seen, as they have been completely armoured with protective materials (Fig.
1 1.3).
Heavy construction on bluff tops, as well as landscaping and the required
irrigation or watering required to support that vegetation, has added several
times the normal average annual precipitation along the coastline of southern
California (Griggs & Savoy, 1985). The net result is to increase the pore
pressures in the cliff materials, thereby decreasing their strength and
accelerating the cliff failure process. In addition, the runoff from the
impervious surfaces accompanying urbanization of coastal bluff top areas has
typically been directed into culverts or drains which have focused runoff on the
bluff face, thereby increasing local cliff erosion or stability (Fig. 11.4).
Coastal engineering structures have had several effects on the development
and stability of seacliffs. Large structures, such as breakwaters and jetties,
typically induce upcoast impoundment of littoral drift, thereby widening the
beach and protecting the cliffs from direct wave attack. As a consequence,

Figure 1 1.3. Intensively developed and protected coastal bluffs along the coastline of
southern California, where little natural bluff is exposed.
 Coastal cliffs and platforms 437

Figure 11.4. These 22 metre high cliffs continue to fail due to a combination of rock
weakening by surface runoff and groundwater seepage, undercutting by wave action,
and mass movement along joint sets. The debris at the base of the cliff was due to
failure caused by the 1989 Loma Prieta earthquake. The apartment units were
subsequently demolished due to cracking of the bluff.

however, the downcoast beaches are initially starved and this beach loss leads
to an increase in exposure to wave attack and accelerated cliff erosion can be
expected (Griggs & Johnson, 1976). Local protection structures such as
seawalls and revetments are constructed to control wave-induced cliff retreat,
and as such can temporarily halt cliff erosion and stabilize coastal cliffs or
bluffs. The type of structure utilized and its height, depth, lateral extent and
durability are important factors in determining the effectiveness of these
shoreline erosion control devices (Fulton-Bennett & Griggs, 1986).
43 8 G.B. Griggs & A.S. Trenhaile
Documenting rates of cliff erosion
It has been recognized by coastal reseachers working with sea cliff evolution
and erosion that cliff or bluff retreat is usually an episodic process. Most of the
major episodes of cliff erosion occur during the simultaneous occurrence of
high tides and large storm waves. At these times waves can reach high enough
up on the shoreline to attack those areas which are less frequently inundated
and the material which has been weakened progressively through weathering
can be dislodged and removed. The sequence of processes may include beach
scour followed by direct cliff or bluff attack, undercutting of the base of the
cliff followed by collapse of the overlying unsupported material, or simply
hydraulic quarrying of blocks or rock which were stable during conditions of
lower wave energy. Terrestrial processes, such as landsliding, slumping, or
rock falls triggered independently of wave attack, may also be extremely
important, and even dominant in areas where the base of the bluff is protected
by a coastal engineering structure or seawall (Fig. 1 1S).
Although we often use or publicize average annual cliff retreat or erosion
rates, in reality we are simply comparing the position of the cliff edge at
different points in time (whether from historic maps or aerial photographs, or
actual field measurements), and dividing by the total number of years between
these data points, to derive an ‘average’ erosion rate. Cliffs or bluffs may
remain superficially unchanged for years, and then due to the right
combination of bluff saturation, or tidal level and wave attack, several metres
may fail instantaneously. Averaging this loss over the time interval between
major storms produces an average rate which may vary from millimetres per
year along resistant granitic coasts to metres per year in unconsolidated
sediments.
Most studies of sea cliff erosion have relied on measurements from historic
aerial photographs or maps, as these typically provide the longest data base.
Vertical stereoscopic aerial photographs are common tools used by coastal
geologists and geomorphologists. In populated coastal areas of the United
States, for example, the aerial photographic record may extend back 50 or 60
years. A span this long typically includes a representative interval of storm and
calmer weather conditions. Thus, the range in actual erosion rates derived from
sequential measurements of the location of the cliff edge, relative to some
baseline or benchmark (a road or structure, for example) over the time span of
the photos, can produce a reasonably accurate picture of the pattern of cliff
retreat.
The extent of the data base utilized (aerial photographs, maps, or ground
measurements), the resolution of this data base (e.g. aerial photo scale and
clarity), the skill or experience of the investigator, as well as the historical
 Coastal cliffs and platforms 439

Figure 11.5.These 25 metre high coastal bluffs continued to fail through mass movements
during heavy rainfall despite their isolation from wave action.

representativeness of the time interval spanned by the data base, are all
important factors which affect the reliability of calculated cliff erosion rates.
Wide variations can result depending upon the length of the historical record
utilized, or the particular segment of coast analyzed, for example, due to long-
term climatic or storm-frequency variations, or due to the alongshore
differences in the geological materials and their resistance or susceptibility to
erosion. The long-term hazards of constructing on geologically active coastal
cliffs where long-term erosion rates have not been carefully evaluated, or
where average erosion rates from some nearby areas have been extrapolated to
the site, can be very costly (Figs. 11.4 and 11.5).
One of the few attempts to quantify both the erosional processes and the
rates of cliff retreat in order to develop a quantitative model is the work of
440 G.B. Griggs & A.S. Trenhaile
Sunamura (1975, 1977, 1982). Sunamura used a wave tank which included a
model cliff made of sand and cement to obtain a relationship between process
and response in cliff form which was then compared with field data. Because
Sunamura was primarily interested in the direct erosion of the cliff, he
concentrated his attention on the formation of the notch which formed at the
base of the cliff. The distribution of wave pressure applied to the cliff face by
the waves was calculated theoretically and measured empirically and the
results were compared with the depth of actual notch erosion. Sunamura was
able to show that for waves breaking directly on the cliff face there is a clear
relationship between wave pressure and erosion, but that for broken waves,
erosion occurs at a much slower rate.

Variations in cliff/platform morphology

Cliff profiles
Steep or undercut cliffs are typical of wave-dominated environments, whereas
convex slopes develop where the climate and wave regime are more conducive
to subaerial weathering and erosion. Geological factors, including structural
weaknesses, stratigraphic variations, and the attitudes or orientations of the
bedding, also exert a strong influence on the shape and gradient of cliffs
(Emery & Kuhn, 1980; Trenhaile, 1987). Although there is a complex
relationship between the shape of cliff profiles and the dip of bedding and/or
joint planes in the rock (Terzaghi, 1962), cliffs tend to be steepest in rocks
either horizontally or vertically bedded, especially if, as in the chalk of
southern England and northern France, they are also lithologically very
homogeneous (Fig. 11.6). Slopes are generally more moderate in rocks which
dip seawards or landwards. Seaward dips often produce dip slopes that are
typically quite smooth (Fig. 11.7), but when dips are landward, many different
beds can be exposed and differential erosion may produce a very irregular cliff
face. The resistance and thickness of the sedimentary strata and the position of
the weakest members are also of great importance. A weak stratigraphic unit
exposed to wave action at the base of a cliff, for example, often results in
formation of a notch and eventually collapse of the overlying, unsupported
material.
Composite cliffs have more than one major slope element. They include
bevelled (hog’s back, slope-over-wall) cliffs with convex or straight
seaward-facing slopes above steep, wave-cut faces, and multi-storied cliffs
with two or more steep surfaces separated by more gentle slopes. Some
composite cliffs reflect geological influences or the combined effects of
 Coastal cl@s and platjiorms 44 1

Figure 1 1.6. Steep cliffs cut into the horizontally bedded chalks along the Normandy
coast of northern France.

Figure 1 1.7. Failure along dip slopes in shale along the northern California coast has
produced a smooth cliff profile (arrow shows dip direction).
 442 G.B. Griggs & A.S. Trenhaile
subaerial and marine processes. Others, however, are the result of Pleistocene
changes in climate and sea level. It has been proposed that wave-cut cliffs,
abandoned during the last glacial stage, were gradually replaced by the upward
growth of convex slopes developing beneath the accumulating talus. With
removal of the debris since the sea rose to its present position, marine erosion
has subsequently either trimmed the base of the convex slopes to form
composite cliffs, or, where erosion has been more rapid, completely removed
it to form steep, wave-cut cliffs (Trenhaile, 1987).

A case study
The development of coastal cliffs has been modelled numerically over two
glacial-interglacial cycles, using a range of values to represent cliff height (25
to loom), rates of marine erosion (0.01 to 0.001 ma-’) and subaerial
weathering (0.01 to 0.0001 m a - ’ ) in cool coastal regions. The mathematical
form of the model is broadly similar to Fisher’s (1 866) model, which has been
adopted and extended by a number of other workers. It was found that
(Fig. 11.8):

last interglacial stage last glacial stage today

(high sea level) (cold, low sea levell (high sea level)
penultimate
glacial stage
[cold, low
..I i.r.11

penultimate
interglacial stage
(high sea k r d

* 1.1115.1
b beveled
c two-storied
d thre.-storl.d

1.1”s

Figure 1 1.8. Model results showing cliff development through two glacial-interglacial
cycles, rates of marine and subaerial erosion differ.
 Coastal cliffs and platforms 443
1. Vertical cliffs formed at the end of the first and second glacial-interglacial
cycles, when low subaerial weathering rates were combined with fast wave
erosion.
2. Bevelled profiles developed at the end of the first cycle in almost 80% of
the 72 model runs. This form generally reappeared at the end of the second
cycle, although it was replaced by two-storied cliffs in four runs. Bevelled
profiles were least likely to develop when weathering rates were low.
3. Two-storied profiles developed at the end of the first cycle in runs with slow
subaerial weathering and wave erosion. These profiles always added
another story at the end of the second cycle, and further cycles would
continue to add others, until an equilibrium form had been attained.
4. The probability of formation of bevelled and multi-storied cliffs increases
with the height of the cliff.
The prolonged period of subaerial degradation required for composite cliff
development, according to traditional concepts, could only have occurred in
areas that lay beyond the ice margins and above the marine limit in the last
glacial stage. Much of the coast of northern North America and Europe,
however, was under ice or below sea level at that time. To study the effect of
these factors on cliff development, a survey was made of the coast of Atlantic
Canada, based upon maps, aerial photographs and field work. Steep coastal
slopes were classified as dominantly wave-cut, composite (bevelled and
multi-storied), or essentially unmodified glacial surfaces.
Palaeozoic sedimentary and volcanic rocks underlie most of the area, but
there are also extensive outcrops of Archaean and especially Palaeozoic
intrusions. Almost all of Atlantic Canada experiences fairly vigorous wave
action, although sea ice protects most of the coast in winter. The coast also lies
within the optimum climatic zone for atmospherically and tidally induced frost
cycles (Trenhaile, 1987).
Modelling suggests that wave erosion rates increase with the rate that sea
level rises and decrease with the rate that it falls (Trenhaile, 1989). The rate
and direction of postglacial changes in relative sea level varied in Atlantic
Canada according to local rates of isostatic recovery. Relative sea level has
therefore risen in southerly areas that were near the ice margins, and fallen in
more northerly areas that were under fairly thick ice.
Most cliffs in Atlantic Canada are between 3 and 30 m in height, although
they occasionally range up to 100 m or more in the resistant metamorphic and
igneous rocks of upland areas. There are wave-cut cliffs in areas of low relief.
They occur in sheltered environments in fairly weak Triassic and Palaeozoic
sedimentary rocks, but only in exposed areas in older, more resistant rocks.
444 G.B. Griggs & A.S. Trenhaile
Steep glacial slopes, little modified by wave action, are generally in resistant
Cambrian and Precambrian rocks, especially where glacial ice flowed parallel
to the coast.
Most of the region was buried beneath ice in the late Wisconsin, and many
of the small unglaciated areas, most notably in western Newfoundland, were
depressed below sea level. Therefore, despite the widespread occurrence of
high cliffs, rapid weathering, and other suitable conditions for composite cliffs,
they are largely restricted to northern Cape Breton Island and a few other areas
in the southern portion of the region that lay beyond the ice margins and above
the marine limit (Fig. 1 1.9). Nevertheless, the fact that they are found in a few
areas that were under ice or sea water in the late Wisconsin suggests that they
can be formed in other ways, although one can only speculate at present on
some possible origins:
1. Postglacial development could have occurred in ice marginal areas above
the marine limit where, as in the upper Bay of Fundy, there was early
deglaciation of frost susceptible rocks (Fig. 11.9).

13 i c e margin with a g e in
thousands o f years b e f o r e p r e s e n t

d e g l a c i a t e d b e f o r e 18,000
years a g o

d e g l e c l a t e d b e t w e e n 18,000
and 14,000 years a g o

l a t e Wlsconsin marlne Ilmlt

b e v e l e d cliffs multl-storied cllffs l s o b a s s s (ml

Figure I 1.9. Composite cliffs in Atlantic Canada and late-Wisconsin glacial and
marine limits.
 Coastal cliffs and platforms 445
2. In unglaciated areas below the marine limit where emergence was gradually
succeeded by submergence, debris could have accumulated for a few
thousand years when relative sea level was low, and then been removed as
it rose to its present level. As it is unlikely that talus could have reached the
top of high, resistant cliffs in the short time available, however, this
explanation could only account for the formation of multi-storied profiles.

Although wave erosion at the base of steep glaciated slopes could produce
composite profiles, the general lack of these profiles in Atlantic Canada
suggests that the rocks have been too resistant for significant marine erosion,
especially where relative sea level has been falling.

Shore platforms
The erosion of the coastline and the subsequent removal of the rock and
sediment debris by nearshore currents cause the progressive retreat or
recession of the shoreline. This leaves behind the remnant of the old slope or
shoreline marked by the lowest level to which the erosion reached, the shore
platform. These platforms have been recognized globally for decades and
several differing arguments have been put forward regarding their shape and
evolution, and the relationship of these to the causal factors or physical
processes.
Shore platforms are usually nearly flat or slope gently seaward, and they
range up to about a kilometre in width. Cross-shore profiles vary from linear to
concave to more complex or composite profiles with steeper inner edges and
flatter outer edges (Bradley & Griggs, 1976). These platforms may extend
from approximately the high tide level, at the base of the receding cliff, to an
elevation below and beyond the low tide level in the nearshore zone. In recent
years the genetically neutral term ‘shore platform’ has for the most part
replaced the term wave-cut platform, thereby avoiding the implication that
these flat features owe their origins exclusively to wave abrasion. The most
widely held view is that a variety of different processes operate on these
platform surfaces and that the resulting morphology reflects these differences.
Nevertheless, it has been shown that there is a strong relationship between the
gradient and other aspects of the morphology of shore platforms, and tidal
range, in a wide variety of environments and rock types (Trenhaile, 1980,
1987).
A high-water or structural storm ledge frequently occurs at the junction
between the inner edge of the shore platform and the base of the sea cliff along
446 G.B. Griggs & A.S. Trenhaile
the tectonically active west coast of California. Similar benches, which include
the ‘2 metre’ bench of the Pacific, have been described in many other parts of
the world (Trenhaile, 1971, 1987). This storm ledge or platform is a
geographically restricted feature, at most only 25 to 50 metres in width, and
best developed on a cliffed coast consisting of well-bedded sedimentary rocks
which are nearly flat lying or dip gently seaward (Fig. 1 1.10). Typically, but
not always, the platform forms along a single slightly more resistant bedding
plane. This feature appears to owe its origin to a combination of storm-wave
attack at high tide and perhaps water table weathering.
Although arguments have been put forward in the past that this two metre
bench provided evidence for a slightly higher eustatic sea level in the very
recent past (Fairbridge, 1961), careful analysis and process observations along
the shoreline where these features exist indicate that they are modern features
forming under present storm wave and sea-level conditions. Debris from the
flanking sea cliff commonly accumulates on the back edge of the terrace, but
storm-wave attack under high tide conditions will sweep this material away
and clear the platform surface. Undercutting of the sea cliff may also take place
during these events, and when combined with terrestrial processes, lead to

Figure 11.10. A storm ledge or platform along the central coast of California which has
been eroded into thin-bedded mudstones which dip gently seaward.
 Coastal cliffs and platforms 447
ongoing sea cliff retreat. The outer edge of the bench or platform is constantly
under wave attack, and in order for an equilibrium profile to be preserved, the
sea cliff and the platform must be, over the long term, eroded at the same rate.
If either horizontal or vertical platform erosion were to occur at a greater rate
than sea cliff retreat, then at a constant sea level the platform should soon
disappear completely. On the other hand, if platform erosion were slower than
cliff retreat, the cliff would soon be isolated from wave attack and would
become a relict feature.

Uplifted marine terraces

Elevated marine terraces are common along collision coasts where uplift is
taking place. The coastlines of New Guinea, New Zealand, as well as the west
coasts of Central and North America, are examples of locations where uplifted
marine terraces are particularly well-developed. These terraces typically
resemble a flight of stairs, commonly less than a kilometre in width, which
ascend to elevations of several hundred metres above present sea level. Each
terrace consists of a nearly horizontal or gently seaward dipping erosional or
depositional platform backed by a steep or degraded, relict sea cliff along its
landward margin (Fig. 1 1.1 1). Based on modern nearshore process
observations, the shoreline angle, or the intersection of the relict platform and

Figure 1 1.1 1. A sequence of four uplifted marine terraces along the California coast,
with a storm platform along the base of the sea cliff.
448 G.B. Griggs & A.S. Trenhaile
the relict sea cliff, provides a good approximation of the location and elevation
of the abandoned or former shoreline, and hence a relative-sea-level highstand
(Lajoie, 1986).
Over the past 25 years, a general consensus has developed that a sequence
of uplifted Pleistocene marine terraces is the geologic and geomorphic record
of repeated glacio-eustatic sea-level highstands superimposed on a rising
coastline. Thus, a rising coastline is a continuous strip chart on which
relatively brief sea-level highstands were successively recorded as erosional or
depositional landforms (Lajoie, 1986). While earlier studies of these uplifted
marine terraces focused on surface morphology and the sedimentary deposits
overlying these abrasional platforms, more recent work has concentrated on
the significance of these terrace sequences as tools to help unravel the recent
tectonic history of the associated coastlines. Using a combination of
radiometric dates on the lowermost terraces (where fossils may still be
preserved), the sea-level rise curve developed from oxygen isotope
stratigraphy from deep-sea cores, and the present elevation of each terrace, it
has been possible to determine the uplift rates for particular segments of
coastline. Dividing the absolute amount of uplift experienced by each
successive terrace since it was formed at some palaeo sea level, by the age of
that terrace, allows us to calculate an average uplift rate (Veeh & Chappell,
1970; Bloom et al. 1974; Chappell, 1983; Lajoie, 1986).

Conclusions
Rock coasts are dynamic landscape elements that are adjusting to the
contemporary morphogenic environment. As rates of change are generally
very slow, however, many coasts retain vestiges of former environmental
conditions that have been inherited, with little contemporary modification,
from interglacial stages when the sea level was similar to that of today. The
degree to which coasts are contemporary or inherited features depends upon
rates of erosion, which are determined by the intensity of the erosional
processes and the resistance of the rock, and the amount of time that these
processes have operated. We are still unable, however, to identify and quantify
the erosive processes operating on rock coasts, and the geological
characteristics that determine the resistance of rocks to these processes.
Mathematical modelling can be used to simulate the long-term development of
rock coasts, but reliable models must also be based upon reliable field data.
Despite a dramatic increase in the number of coastal workers in the last few
decades, however, there has probably been an absolute, as well as a relative,
 Coastal clijjs and platforms 449
decline in the number concerned with rock coasts. We are therefore still
dependent, in many areas, on theories developed in the early part of this
century, and unfortunately there appears to be little prospect of eliminating
many of the deficiencies in our knowledge in the near future.

References
Bartrum, J.A. (1916). High water rock platforms: a phase of shoreline erosion.
Transactions of the New Zealand Institute, 48, 1 3 2 4 .
Bloom, A.L., Broecker, W.S., Chappell, J.M., Matthews, R.K., & Mesolella, K.J.
(1974). Quaternary sea-level fluctuations on a tectonic coast:
new 230Th/234U dates from the Huon Peninsula, New Guinea. Quaternary
Research, 4,185-205.
Bradley, W.C. & Griggs, G.B. (1 976). Form, genesis, and deformation of some
central California wave-cut platforms. Bulletin of the Geological Society of
America, 87,433-49.
Chappell, J.M. (1983). A revised sea-level record for the last 300,000 years from
Papua, New Guinea. Journal of Geophysical Research, 14,99-101.
Emery, K. 0. & Kuhn, G. G. (1980). Erosion of rock coasts at La Jolla, California.
Marine Geology, 37, 197-208.
Fairbridge, R. (1961). Eustatic changes in sea-level. Physics and Chemistry of the
Earth, 4, 99-1 85.
Fischer, R. (1 990). Biogenetic and nonbiogenetically determined morphologies of the
Costa Rican Pacific coast. Zeitschrift fur Geomorphologie, 34, 3 13-2 1.
Fisher, 0. (1866). On the disintegration of a chalk cliff. Geological Magazine, 3,
354-6.
Flageollet, J. C. & Helluin, E. (1987). Morphological investigations of the sliding
areas along the coast of Pays d’Auge, near Villerville, Normandy, France. In
International geomorphology, vol. 1, ed. V. Gardiner, pp. 477-86. Chichester,
John Wiley.
Fulton-Bennett, K.W. and Griggs, G.B. (1986). Coastal protection structures and
their effectiveness. Joint publication of California State Dept. Boating and
Waterways and University of California, Santa Cruz, Institute of Marine
Sciences. 48 pp.
Gornitz, V. (1 991). Global coastal hazards from future sea-level rise. Palaeogeography,
Palaeoclimatology, Palaeoecology, 89, 379-98.
Grainger, P. & Kalaugher, P. G. (1987). Intermittent surging movements of a coastal
landslide. Earth Surface Processes and Landforms, 12, 597-603.
Griggs, G.B. & Johnson, R.E. (1976). The effects of the Santa Cruz Small Craft
Harbor on coastal processes in northern Monterey Bay, California.
Environmental Geology, 1,229-3 12.
Griggs, G.B., & Savoy, L.E. (1985). Living with the California coast. Durham, NC:
Duke University Press. 393 pp.
Inman, D.L. & Nordstrom, C.E. (1971). On the tectonic and morphologic classification
of coasts. Journal of Geology, 79, 1-21.
Kennedy, T.B. & Mather, K. (1953). Correlation between laboratory freezing and
thawing and weathering at Treat Island, Maine. Proceedings of the American
Concrete Institute, 50, 141-72.
450 G.B. Griggs & A.S. Trenhaile
Lajoie, K.R. (1986). Coastal tectonics. In Active tectonics, pp. 95-124. Washington,
DC: National Academy Press.
Matsukura, Y. & Matsuoka, N. (1991). Rates of tafoni weathering on uplifted shore
platforms in Nojima-Zaki, Boso Peninsula, Japan. Earth Surface Processes and
Landforms, 16, 5 1-6.
Matsuoka, N. (1988). Laboratory experiments on frost shattering of rocks. Science
Reports of the Institute of Geosciences, University of Tsukuba (Japan), 9A,
1-36.
Pitts, J. (1986). The form and stability of a double undercliff an example from
south-west England. Engineering Geology, 22, 209-1 6.
Robinson, L.A. (1977). Erosive processes on the shore platform of northeast
Yorkshire, England. Marine Geology, 23, 339-61.
Robinson, L. A. & Jerwood, L. C. (1987). Sub-aerial weathering of chalk shore
platforms during harsh winters in southeast England. Marine Geology, 77, 1-14.
Sanders, N.K. (1968). The development of Tasmanian shore platforms. PhD thesis.
University of Tasmania.
Schneider, J. (1976). Biological and inorganic factors in the destruction of limestone
coasts. Contributions to Sedimentology, 6, 1-1 12.
Schneider, J. & Torunski, H. (1983). Biokarst on limestone coasts, morphogenesis and
sediment production. Marine Ecology, 4, 45-63.
So, C. L. (1987). Coastal forms in granite, Hong Kong. In International
geomorphology, ed. V. Gardiner, pp. 1213-29. Chichester: John Wiley.
Spencer, T. (1988). Limestone coastal morphology. Progress in Physical Geography,
12,66-101.
Suess, E. (1888). The Face of the Earth !I (English Translation by H.B.C. Sollas,
1906, Oxford: Oxford University Press).
Sunamura, T. (1975). A laboratory model for wave-cut platform formation. Journal of
Geology, 83, 389-97.
Sunamura, T. (1977). A relationship between wave-induced cliff erosion and erosive
force of wave. Journal of Geology, 85,613-18.
Sunamura, T. (1982). A predictive model for wave-induced cliff erosion, with
application to Pacific coasts of Japan. Journal of Geology, 90, 167-78.
Terzaghi, K. (1962). Stability of steep slopes on hard unweathered bedrock.
Geotechnique, 12, 25 1-70.
Trenhaile, A S . (1971). Lithological control of high-water rock ledges in the Vale of
Glamorgan, Wales. Geografiska Annaler, 53A, 59-69.
Trenhaile, A.S. (1980). Shore platforms: a neglected coastal feature. Progress in
Physical Geography, 4, 1-23.
Trenhaile, A. S . (1987): The geomorphology of rock coasts . Oxford:Clarendon/
Oxford University Press. 384 pp.
Trenhaile, A. S . (1989). Sea-level oscillations and the development of rock coasts. In
Applications in coastal modeling, eds. V. C. Lakhan & A. S. Trenhaile,
Amsterdam: Elsevier. pp. 27 1-95.
Trenhaile, A. S . & Mercan, D. W. (1984). Frost weathering and the saturation of
coastal rocks. Earth Surface Processes and Landforms, 9, 321-3 I .
Trudgill, S . T. (1976). The marine erosion of limestone on Aldabra Atoll, Indian
Ocean. Zeitschrqt fur Geomorphologie, Suppl. Bd., 26, 164-200.
Trudgill, S. T. (1987). Bioerosion of intertidal limestone, Co. Clare, Eire. 3. Zonation,
process and form. Marine Geology, 74, 1 11-21,
Veeh, H.H. & Chappell, J. (1970). Astronomical theory of climatic change: support
from New Guinea. Science, 167, 862-5.
 12
Tectonic shorelines
P. A. PIRAZZOLI
Sea level is the common and unifying element of coastal tectonics.
K.R. Lajoie, 1986

Introduction
Various classifications of coasts have been attempted in the literature, in some
cases inspired by the plate tectonics model (e.g. Inmam & Nordstrom, 1971).
According to Rice (1941) and to the American Geological Institute (1960), the
term ‘tectonic’ is defined in a very wide sense, ‘designating the rock structure
and external forms resulting from the deformation of the earth’s crust’. This
definition implies that certain processes, which are not always considered as
‘tectonic’, must also be taken into account. It is now widely accepted that
phenomena like glacioisostasy and hydroisostasy have produced a vertical
deformation of the earth’s crust in virtually all coastal areas over the last 20 ka
(Clark, Farrell & Peltier, 1978). Moreover, sediment deposition near coasts
and especially in delta areas contributes, together with erosion processes and
volcanic eruptions, to the modification of loads exerted on the earth’s crust,
causing additional vertical deformation (see Chapter 3). Consequently, all the
coasts of the world can be considered as more or less tectonic and none are
vertically stable. Instead of proposing a new classification, the present chapter
aims therefore to identify various kinds of vertical movement which may affect
any of the types of coast considered elsewhere in this volume, with, however,
more emphasis on seismo-tectonic processes. Certain effects of vertical
movements on coastal evolution are also considered.

Main causes of vertical displacements in coastal areas

As sea level is an equipotential surface of the gravity field, it can be influenced
by any factor affecting gravity: from space (astronomical effects), from the
atmosphere (meteorological changes), in the oceans (hydroisostasy, steric
changes) at the earth’s surface (glacioisostasy, volcano-isostasy, sedimento-
isostasy) and in the earth’s interior (thermo-isostasy, density changes).

45 1
452 P.A. Pirazzoli
Geoid changes
Over the long term (several million years), important vertical displacements at
the surface of the earth can be produced by density changes in the interior.
Satellite ranging has revealed that sea-surface topography does not correspond
exactly to that of a rotational ellipsoid, and that there are many bumps and
depressions with a relief of up to 200 m (Gaposchkin, 1973; Marsh & Martin,
1982). This is probably the consequence of an uneven mass distribution inside
the earth. As the earth’s interior is not solid, relative movements between
masses with different densities have certainly occurred in the past and will
probably continue to do so as long as the earth has not completely cooled. Each
density change modifies the earth’s gravity field. The water surface, being
liquid, will adjust to the new gravity field almost instantaneously. The
lithosphere on the other hand, being denser than sea water and liable to be
deformed and warped, although being solid, will also adjust, but with a less
absolute deformation occurring over a much longer time period. The
difference in vertical displacement and the time lag between liquid and solid
adjustment may cause transgression or regression phenomena on a regional or
even continental scale (Morner, 1976). Changes in the earth’s rate of rotation
and in the tilt of the rotation axis may also produce transgressions and
regressions across wide areas (Fairbridge, 1961; Barnett, 1983; Morner, 1988;
Sabadini, Doglioni & Yuen, 1990; Peltier, 1990). Very little information is
available, however, on case studies of such changes in the past, though vertical
displacements deduced from fossil shorelines have in some cases been
ascribed to geoidal changes (e.g. Martin et al., 1985; Nunn, 1986).

Thermo-isostatic and volcano-isostatic deformation

It is now generally accepted that as the ocean crust spreads away from its point
of origin along submarine ridges, it cools and thickens, thereby increasing in
density. As a result the seafloor subsides isostatically, gradually submerging
oceanic islands as they are carried laterally. Over several million years thermo-
isostatic submergence will be of the order of kilometres (Fig. 1 2 . 1 ~ ) .In
tropical waters subsidence is increased by the load of coral reefs, which have
to grow vertically to maintain their sea-level position. The normal evolution of
an oceanic island, in the absence of eustatic changes in sea level, would
therefore be a gradual, continuous submergence and, at the same time, the
rapid erosion at the surface of all emerged parts caused by subaerial
weathering and wave action. In tropical waters such processes would support
the classical sequence ranging from a high island with fringing coral reefs to an
island with a barrier reef, and finally to an atoll (Darwin, 1842; see also
 Tectonic shorelines 453

3
h
E
Y
-
5n 4
$ 5

a 0 20 40 60 80 100 I20 140

Age (Ma)

Ocean

5n Lithosphere
$

Asthenosphere
b
H

Figure 12.1. ( a ) The relation between depth and age of normal oceanic crust (adapted
from Menard, 1986). ( b )Schematic section of the lithosphere along a hot spot trace, the
heat flow H causes temporary thinning and isostatic uplift of the lithosphere; after the
hot spot zone has been left behind, lithospheric subsidence, caused by cooling and
thickening, will occur again. (Adapted from Detrick & Crough, 1978; no scale.)

Chapter 7). Average subsidence rates have been estimated at 0.2 mm a- since '
60 Ma in the Marshall Islands (Menard & Ladd, 1963) and 0.12 mm a- since '
the Pliocene in Mururoa Atoll (Labeyrie, Lalou & Delibrias, 1969) (for a
review of results of deep drilling in atolls, see Guilcher, 1988, pp. 68-75). The
normal evolutionary trend of a oceanic island implies, therefore, gradual
subsidence and relatively rapid erosion of exposed rocks. That is why non-
carbonate rocks older than a few million years are so rare in mid-plate oceanic
areas. If an oceanic island is not subsiding, it is anomalous and requires
explanation.
454 P.A. Pirazzoli
Where the oceanic crust approaches a hot spot, which according to Menard
( 1973) corresponds to a fixed asthenospheric bump, the normal cooling
process is reversed, so that the crust is heated, becomes less dense and thinner,
and thus rises. During transport away from the hot spot there is renewed
cooling and subsidence (Menard, 1973; Detrick & Crough, 1978) (Fig. 12.lb).
Near hot spots, extrusion of lava often occurs. The resulting load produces on
the lithosphere an isostatic depression under the volcanic pile, over a distance
generally less than 150 km from the load barycentre, and a peripheral raised
rim at a distance of between some 150 and 300 to 330 km (McNutt & Menard,
1978) (Fig. 12.2), i.e. volcano-isostatic effects similar to the glacioisostatic
effects of an ice sheet with equivalent mass (see below). When the translation
movement of an oceanic plate over a hot spot has produced a line of islands,
and/or two or more hot spots form an alignment in the direction of plate
movement, interaction between isostatically depressed areas and uplifted rims
of nearby islands is possible and may produce complicated sequences of
vertical deformation, with repeated phases of upheaval and sinking (Coudray
& Montaggioni, 1982; Scott & Rotondo, 1983a,b; Montaggioni, 1989).
A good example of these processes is given by the interaction between the
Society Islands chain, its active hot spot and certain nearby Tuamotu atolls, in
the south Pacific. Anaa is a closed atoll located on the western margin of the
Tuamotu Archipelago, presently about 200 km east of the hot spot area of the
Society Islands (Fig. 12.3). In Anaa, exposed ancient reefs dating from the Last
Interglacial period (Veeh, 1966; Pirazzoli et al., 1988), which are usually
submerged in other Tuamotu atolls, are well preserved in the northeastern part
of the reef rim, where they reach an elevation of about 4 m above present sea
level (Fig. 12.4). There is no evidence of older exposed reefs. During the
Holocene, the sea reached here a level about 1.35 5 0.1 m higher than at
present, i.e. 0.3-0.5 m higher than in other Tuamotu atolls. These data indicate
that Anaa started uplifting recently (probably less than 300 ka ago) and that its
relative uplift rate, in relation to other Tuamotu atolls remote from active hot
spot areas, can be estimated to have been at most 0.1 mma-' since the Last

Figure 12.2. A point load P causes a flexure on a floating elastic plate. In the case of the
oceanic lithosphere loaded by a volcano, the flexure will consist of a subsidence zone,
which reaches its maximum under the load barycentre, and an arching at some distance.
 Tectonic shorelines 455
'
Interglacial period, and about 0.1 mm a- during the late Holocene. This slight
uplifting trend can easily be explained by thermal rejuvenation phenomena
(thermo-isostasy) while Anaa is approaching the uplifted side of the
asthenospheric bulge of the Society Islands hot spot (Pirazzoli et al., 1988).On
the other hand, as the distance of Anaa from the barycentre of the most recent
important volcanic load on the lithosphere (which corresponds to the Tahiti
and Moorea volcanoes, see Fig. 12.3a) is over 350 km, volcano-isostasy has
probably not yet contributed to the recent uplift trend.
The Pacific plate translation carries Anaa in a northwest direction, at a
speed of about 110 mm a-'. Following this trajectory, Anaa is approaching the

ZOO(

o m

200c

400C

Figure 12.3. (a) Location map of the Society and northwest Tuamotu Islands. The hot
spot of the Society Islands is now active near Mehetia (ME). Elevated reefs in the
Tuamotus are found only between the f-f curve and Tahiti; their present-day emergence
(in metres) is indicated near the names of the atolls: A = Anaa; MK = Makatea; N =
Niau; K = Kaukura; RN = Rangiroa; TI = Tikehau; MT = Mataiva. The maximum
elevation is observed in Makatea: 20-25 m for mid-Pleistocene reefs, and about 110 m
for the summit of the island. MO = Moorea; TE = Tetiaroa; MI = Maiao;H = Huahine;
RI = Raiatea; TA = Tahaa; BB = Bora Bora; TU = Tupai; MU = Maupiti; MP =
Mopelia; S = Scilly; BE = Bellinghausen. (b)Topographic profile across the Society
Islands; continuous line: profile along X-X (see a ) ; broken line: projections of nearby
volcanic complexes on X-X; A: average WAr ages of the volcanic islands; B: linear
age variation corresponding to a uniform velocity of the Pacific plate of l l o m m a - '
(adapted from Pirazzoli & Montaggioni, 1985).
456 P.A. Pirazzoli

Figure 12.4. Reefs dating from the Last Interglacial reach about +4m at Anaa Atoll
(Tuamotu Islands, Pacific Ocean) (from Pirazzoli & Montaggioni, 1985).

hot spot area at a rate of 45 mma-' and will continue to do so for the next
1.3 Ma, when it will reach the minimum distance of about 140 km from that
'
area. If the 0.1 mm a- uplift rate does not change and vertical erosion can be
ignored, Anaa will then rise by thermo-isostasy to about 135 m in altitude.
Meanwhile, however, new important volcanic loads may be constructed
above the hot spot in a position nearer to Anaa than the Tahiti-Moorea load,
thus creating a new volcano-isostatic arch, which might increase the uplift of
Anaa.
Makatea Island, which was located about 3 Ma ago in the present position
of Anaa in relation to the Society Islands hot spot area, displays the pattern that
Anaa is likely to show in the future. Makatea is a raised atoll (Fig. 12.5)
reaching an altitude of 113 m (Montaggioni, 1985). If a possible first thermo-
isostatic uplift interrupting the normal subsidence trend of the island is
disregarded (about 14 Ma ago, moving near the Pitcairn hot spot, as suggested
by Montaggioni, 1989), Makatea entered the uplifting side of the Society
Islands hot spot swell, just as Anaa is doing now, slightly earlier than 3 Ma
ago. During the Pleistocene, when the huge volcanic masses of Moorea and
Tahiti erupted (mainly between 2.0 and 0.3 Ma ago), Makatea was located near
the axis of their isostatic arch. This caused an additional uplift of the island.
Moving away from the hot spot area, Makatea remained on the isostatic arch
 Tectonic shorelines 457

Figure 12.5. Raised limestone cliffs in Makatea Island (Tuamotu Islands, Pacific
Ocean). Arrows indicate two former shorelines: at +25 m (dated more than 200 ka), and
at +5/+8 m (dated 100 to 140ka) (Montaggioni, 1985)(photo L. F. Montaggioni, adapted).

of Tahiti and Moorea (McNutt & Menard, 1978; Lambeck, 1981), which is still
viscoelastically active: subsidence due to incipient lithospheric cooling is
therefore still compensated in Makatea by volcano-isostatic uplift, resulting in
',
upheaval rates of the order of 0.05 mm a- both for the Late Pleistocene and
for the late Holocene (Pirazzoli & Montaggioni, 1985). Viscoelastic effects are
expected to decrease with time, however, while lithospheric cooling will
continue. The uplift phase of Makatea, which has predominated during the last
3 Ma, is therefore coming to an end, whereas a new uplift trend is beginning in
Anaa Atoll.
Similar thermo- and volcano-isostatic phenomena have been studied in the
chain of the Cook-Austral Islands (Lambeck, 1981; Calmant & Cazenave,
1986; Pirazzoli & Veeh, 1987; Woodroffe et al., 1990; Pirazzoli & Salvat,
1992), Henderson and Pitcairn (Spencer, 1989; Spencer & Paulay, 1989), the
Hawaii Islands (Hamilton, 1957; Walcott, 1970; Scott & Rotondo, 1983a,b;
Campbell, 1986) and can be applied to most raised oceanic islands remote
from plate boundaries, and even to seamounts (Cazenave et al., 1980).
45 8 P.A. Pirazzoli
Glacioisostatic and hydroisostatic deformation
One of the best known causes of lithospheric deformation is glacial isostasy.
Global in its many side effects, according to geophysical models (Clark,
Farrell & Peltier, 1978; Nakada & Lambeck, 1987; Peltier, 1990, 1991), it may
exceed the growth of a mountain range or the filling of a sedimentary basin in
terms of the mass involved (Pirazzoli & Grant, 1987). As shown by Daly
(1934) (Fig. 12.6), the load of an ice sheet deforms the earth’s crust. The
resulting subsidence beneath the ice makes deeper material flow away, and
raises an uplifted rim at a certain distance. When the ice sheet melts, an
unloading occurs, resulting in uplift beneath the melted ice; the marginal rim
will consequently tend to subside and move towards the centre of the vanishing
load. In the oceans the load of melted water will make the ocean floor subside
(hydroisostasy) (Daly, 1934; Bloom, 1971). This will cause a flow of deep
material from beneath the oceans to beneath the continents. The latter will tend
therefore to rise, with reactivation of seaward flexuring at the continental edge
(Bourcart, 1949; Clark et al., 1978). From a tectonic point of view, coasts are
therefore very unstable zones, which may follow the isostatic trend prevailing

Figure 12.6. Crustal deformations caused by the melting of an ice cap. Stage 1 shows
undeformed land surface; stage 2 shows ice sheet at maximum extent and crustal
response; stages 3 and 4 represent melting of ice sheet; and stage 5 indicates return to
initial state. (Adapted from Daly, 1934.)
 Tectonic shorelines 459
in the nearby ocean, or that in the nearby continent, depending on the local
topography and coastal configuration, the width of the continental shelf, the
water depth, etc.
Part of the above isostatic movements are elastic, i.e. contemporaneous to
loading and unloading. However, because of the viscosity of the earth’s
material, part of the movements may continue for several thousand years after
loading or unloading has stopped. Although most of the continental ice sheets
had disappeared by 8000 years BP in Canada and Fennoscandia, uplift and
subsiding movements have continued to be active there ever since and will
probably continue to be active for several more millennia, although at
decreasing rates (Fig. 12.7).
Using reconstructions of the Late Pleistocene-early Holocene melting
history of polar ice sheets and consequent glacio- and hydroisostatic effects
around the globe, geophysical models may predict, with assumptions, trends of
vertical movements caused by the last deglaciation in any coastal sector of the
world (e.g. Clark et al., 1978; Nakada & Lambeck, 1987; Peltier, 1991;
Nakada, Yonekura & Lambeck, 1991). These models are therefore very useful
for the reconstruction of possible postglacial coastal evolution in areas with
insufficient field data.

Vertical deformation near plate boundaries

Approaching tectonic plate boundaries, which are generally areas of high
seismicity, vertical movements become more complex, with various local or
regional geodynamic factors superimposed, often with predominant effects, on
glacio- and hydroisostatic factors.

Subduction zones
Underthrusting side: Near a subduction zone, the behaviour of the under-
thrusting plate will differ notably from that of the overthrusting one. On the
underthrusting side, the oceanic plate is most likely subjected to arching
phenomena, in order to make possible the change from a horizontal translation
movement to a subduction beneath the overriding plate. This arching implies a
wave-like flexuring of the lithosphere, with first a slow gradual uplift, as in the
case of an oceanic island approaching a hot spot, then a gradual subsidence at
accelerating rates. When approaching subduction trenches, however, very few
islands are at the right distance from an oceanic trench to verify the existence
of a wave-like flexuring pattern. In the Loyalty Islands, east of New Caledonia,
the occurrence of marine terraces of the same age at different altitudes in
various islands, depending on the distance of these islands from the New
460 P.A. Pirazzoli
300

270

240

210
Curve Refunsm
A Andrews, 1987
B Allard L T m l a y , 1983
C,D H i l l a i r e - H a r a l , 1960
100 E,E’ Hillaire-Ha-1, 1976
E P P e l t i e r L Ardrau8, 1983
G Andrew8 L Faloorrr, 1969
H P e l t i e r C Mdrows, 1983

90

60

30

0
10 5 0
Age (ka BP)

Figure 12.7. Relative sea-level curves obtained from the eastern shores of Hudson Bay
(Canada) indicate very rapid uplift (of glacioisostatic origin) in the early Holocene, at
rates decreasing gradually during the late Holocene. For comments, see Pirazzoli
(199 1).
 Tectonic shorelines 46 1

Hebrides Trench, was interpreted by Dubois, Launay & Recy (1974) as being
due to pre-subduction arching, probably the first report of this phenomenon in
the literature. More recently, elevated geomorphological features in Christmas
Island, Indian Ocean, 200km southwest of the axis of the Sunda Trench
(Nunn, 1988; Woodroffe, 1988) and in the Daito (Borodino) Islands, 150km
east of the Ryukyu Trench (Ota & Omura, 1992), have been ascribed to similar
arching phenomena.

Overthrusting side: On the overthrusting side, rapid rates of vertical

displacement are frequent. Uplift in particular may be caused by (i) piling up
above the oceanic plate, on the inner side of the trench, of sediments too light
to be subducted, which will raise the overthrusting edge isostatically; (ii)
elastic rebound phenomena linked to the subduction; (iii) tilting of lithosphere
blocks or other tectonic processes; and (iv) volcanic activities.
Tectonic trends often appear to be incremental and gradual over the long
term, although in reality they frequently consist of sudden coseismic vertical
movements, occurring at the time of earthquakes of great magnitude, separated
by more or less long periods of quiescence or even of slow interseismic
movement contrary to the coseismic movement. Fast aseismic displacements
have also been reported. Active folding is frequent and it is detected generally
on the basis of geomorphological and geodetic methods. In coastal areas, short
wavelength folding is clearly recognizable in folded marine terraces. Among
many possible examples, the southern tip of Taiwan displays remarkable
undulations of the middle Holocene coral-reef terrace along a distance of about
24 km, with elevations ranging from 5 to 36 m above present sea level (Liew &
Lin, 1987). Repetitive up-and-down movements are also possible, as well as
accelerations in the rates of vertical displacement.
Examples of coastal coseismic uplift in subduction areas deduced from
ancient shorelines are reported frequently in the literature. Along the western
coast of the Muroto Peninsula, Shikoku, Japan, there are three main
superimposed coastal terraces, all of which decrease in height to the north. The
middle (Fig. 12.8) and lower terraces were formed in the Last Interglacial age
and in the middle Holocene, respectively. These terraces are mainly abrasional
in origin, but in some places they are depositional, showing that they were built
under transgressional sea levels (Yoshikawa, Kaizuka & Ota, 1964). Tectonic
movements in this area are characterized by acute uplift tilting landward,
associated with great earthquakes and chronic subsidence tilting seaward
during interseismic periods. In Fig. 12.9, changes in height obtained by
repeated levellings are shown for the periods 1895 to 1929 and 1929 to 1947.
462 P.A. Pirazzoli

Figure 12.8. Remnants of Last Interglacial marine terraces at about +150 m near Hane
RGer, approx. 17 km northwest of Muroto Cape, Shikoku, Japan (photo P.A. Pirazzoli
1974).

1-5-I920

0
200
b 12w

::j
IW
IW

2
0
I0 20 30 40 M we"'

Figure 12.9. ( a )Vertical change of bench marks surveyed by precise levelling along the
Muroto Peninsula (Shikoku, Japan) (upper graph) and (6) vertical distribution of height
of former shorelines on terraces ascribed to the Last Interglacial (solid circles) and to
the penultimate interglacial (open circles) (lower graph). The amounts of vertical
change of bench marks during the preseismic period from 1895 to 1929 are shown in
relation to the present sea level (scale on right) and the vertical changes of bench marks
during the period 1929 to 1947 (which include the effects of the 1946 Nankai
Earthquake) are shown relative to bench mark 5176 on the left side of the graph (scale
on left) (adapted from Ota, 1975).
 Tectonic shorelines 463
The former period falls between two large earthquakes; the amount of
subsidence during an interseismic period - on average 120 years here - was
estimated at 0.95 m from geodetic and historical records. The latter levelling
period includes the 1946 Nankai Earthquake, when the tip of the peninsula was
uplifted 1.20m. The mean uplift rate is estimated to be 2mma-I at the
southern tip of the peninsula. At this rate the Last Interglacial shoreline would
be expected to be at 240m in altitude and the Holocene terrace at 12m,
reasonably comparable to the observed heights of 190 m and 13 m, respectively
(Yoshikawa, Kaizuka & Ota, 1981).
Various examples of coastal coseismic subsidence have been reported from
the Pacific coast of North America (Gulf of Alaska, Washington and Oregon).
In the Gulf of Alaska, where the Pacific plate is being subducted beneath the
American plate, the great earthquake of 27 March 1964 (magnitude 2 8.4),
which had its epicentre in the Prince William Sound area, was characterized by
vertical crustal movements over a region of at least 200000 km2 (some 700 to
800 km long and 150 to 300 km wide) with trends roughly parallel to the Gulf
of Alaska coast and the Aleutian volcanic arc and trench. Average uplift within
the seaward zone is 2 to 3m, except along a narrow zone where combined
crustal warping and faulting have resulted in an uplift of 11.5 m on land and
possibly more than 15 m on the sea floor. Subsidence in the adjacent zone to
the north averages about 1 m and reaches a maximum of 2.2 m. There is no
abrupt change of level between the two zones, but rather a northward and
northwestward tilting around the zero isobase, with a maximum slope of about
0.11 mkm-l (Plafker & Rubin, 1967).
In the area affected by the 1964 Alaska earthquake, postseismic uplift
occurs where coseismic subsidence was observed, and postseismic subsidence
occurs where coseismic uplift was observed (Savage & Plafker, 1991). At
Anchorage in particular, where the 1964 earthquake caused subsidence of
0.71 m, repeated levellings and sea-level measurements indicate as much as
0.55 m of land uplift in the decade following the earthquake. This deformation
is explained by Brown et al. (1977) as due to creep along the down dip
extension of the fault which ruptured in 1964. As noted by Savage & Plafier
(1991), however, the immediate postseimic response is damped out within the
first decade, and the subsequent vertical displacement rates appear to be fairly
steady.
Geological evidence along the coast of the Gulf of Alaska reveals a
complex history of Holocene tectonic displacements, in which areas of
postglacial net emergence (as much as 55m since 7650 years BP) or
submergence (at least 90 m) largely coincide with areas of significant
earthquake-related uplift and subsidence. In the uplifted zone, for example,
464 P.A.Pirazzoli
Middleton Island displays a series of marine terraces indicating that 40m of
relative emergence has occurred in at least five major upward pulses during the
last 4470 ? 250 years BP. A sixth terrace was formed on the island as a result
of 3.3 m of uplift during the 1964 earthquake (Plafker & Rubin, 1967).
Along the outer coast of Washington State, near the boundary between the
Juan de Fuca plate and the North American plate in the Cascadia subduction
zone, coastal subsidence commonly accompanies a great subduction
earthquake, the coseismic subsidence occurring mainly in an onshore belt,
flanked by a zone of coseimic uplift mostly offshore (Atwater, 1987). In
northern Oregon, Darienzo & Peterson (1990) found buried marsh deposits
indicating a series of episodic, abrupt subsidences of the marsh surfaces to low
intertidal levels. The stratigraphy shows evidence of sudden subsidence
displacements of 1.O to 1.5 m alternating with gradual uplift displacements of
the order of 0.5 to 1.0m. This was interpreted as reflecting coseismic strain
release (abrupt subsidence) following interseismic strain accumulation
(gradual uplift). Recurrence intervals between subsidence events in this area
range from possibly less than 300 years to at least 1000 years, with the last
dated event likely to have taken place in 300 to 400 years BP. Overlaps of
radiocarbon ages for at least four subsidence events suggest that events may
have been synchronous over at least 200 km of the central part of the Cascadia
subduction zone.
Similar events are known to have happened in many seismically active
coastal areas, especially in the Circum-Pacific region (Ota, 1991), in central
Chile (Kaizuka et al., 1973), Japan (Yonekura, 1972, 1975; Nakata et al.,
1979; Ota, 1985), Vanuatu (Taylor et al., 1980; Jouannic, Taylor & Bloom,
1982), and New Zealand (Wellman, 1967; Berryman, Ota & Hull, 1989; Ota,
Miyauchi & Hull, 1990; Ota, Hull & Berryman, 1991a) (Fig. 12.10).
Although uplift or subsidence rates are commonly reported as linear in the
long term, case studies of Late Quaternary accelerations in uplift movements
also exist. In Kikai Island (Ryukyus, Japan), which is entirely capped by a
sequence of marine terraces, the uppermost terrace, at over 200 m in altitude,
has been dated by in situ corals at about 125 ka. This and other dates from
younger, lower terraces have enabled Konishi, Omura & Nakamichi (1 974) to
infer an uplift rate of 1.5-2.0 mm a-' since the Last Interglacial. However,
reef-proper facies of the same upper terrace formation were also dated to
200 ka or more by 230Th/234U,and 400 to 600 ka by electron spin resonance
(ESR) (Omura et al., 1985; Koba et al., 1985). These new results suggest that
the tectonic trend has changed, between 200 and 125 ka, to a faster uplift
movement.
Other examples from the Mediterranean are even more drastic, with a long
 Tectonic shorelines 465

Figure 12.10. View west over the rapidly emerging coastline at Turakirae Head along
the southern coast of the North Island of New Zealand (41.4OS, 174.9"E). In the
foreground: four of the five large storm beach ridges preserved along the coast (paths
follow the upper surface of three of them) can be seen dipping westward (towards
Wellington City). The second lowest ridge represents the storm beach that formed prior
to more than 3 metres of uplift that accompanied the M 2 8.0 earthquake on January 23,
1855. The lowest ridge has formed during the last 137 years. The age of higher and
older storm beach ridges is poorly known, but the highest ridge probably formed about
6500 years ago, when rising Holocene sea level reached its highest and present position.
Each of the four uplifted ridges is inferred to represent a large earthquake similar in
magnitude to that experienced in 1855 (Wellman, 1967; Berryman et al., 1992). In the
background: the age of higher level terraces farther west, tilted in a northwesterly
direction, is also unknown; marine terraces and gravels have been identified more than
300 metres above present sea level, and all may have been deposited during the last
c. 130000 years (Ota, Williams & Berryman, 1981). (Photo by Lloyd Homer, Institute
of Geological and Nuclear Sciences, Lower Hutt, New Zealand.)

subsidence period producing a series of always higher transgressions towards

the end of the Pliocene, which caused the development of extensive marine
platforms in Calabria (Carobene & Dai Pra, 1990), followed by subsequent
rapid Pleistocene uplift, bringing marine sediments as high as 1360m since the
Calabrian (Dumas et al., 1980) or 157 m since the Last Interglacial (Dumas et al.,
1988). The uplift movements would be produced by periodic earthquakes of
the type which occurred near Messina in 1908, with an average return time of
466 P.A. Pirazzoli
1000 to 1500 years (Valensise & Pantosti; 1992). In the Aegean, a similar
change of trend happened slightly later, with transgression (subsidence)
predominant until the lower Pleistocene and regression (uplift) since that time
(Keraudren, 1975); the amplitude of this fluctuation was about 500m in the
island of Karpathos (Keraudren & Sorel, 1984).

Collision zones
Where a continental lithosphere exists on both sides of a plate boundary,
subduction will generally be prevented by the low density of lithospheric
material and collision or transform processes will occur with vertical
movements which can became highly irregular. The island of Rhodes, Greece,
where various distorting movements took place after the late Pliocene, may be
a case in point. Though these predominant movements were ‘a general tilting
movement with the raising up of the northeast part and the sinking of the
southwest’, ‘the island did not react to this stress as a sole rigid body, but was
disrupted in several blocks separated by normal faults’ (Mutti, Orombelli &
Pozzi, 1970, p. 71). A recent study of emerged Holocene shorelines (Pirazzoli
et al., 1989) has shown that the east coast of the island consists of at least eight
small crustal blocks (up to twelve kilometres in length) each exhibiting a
different tectonic behaviour. The number of emerged Holocene shorelines in
each block varies from one to seven and a slightly submerged shoreline can
also be found in the north part of the island (as well as various indications of
submergence along the west coast). Relative sea-level changes in the
northernmost crustal block of the island are shown in Fig. 12.1 1. The upper-
most shoreline (A,), now at about +3.4m (Fig. 12.12), was reached by the sea
around 6000 years BP. Shortly before 4000 years BP, a 1.3m coseismic
subsidence movement brought the sea surface to A3, where typical sea-level
erosion notches were carved into the limestone cliffs. A few centuries later,
however, a reversed coseismic movement brought the shoreline back to the
previous position at +3.4m, where it remained almost stable for about one
thousand years. Soon after 2280 ? 110 years BP (probably at the time of the
222 BC earthquake which destroyed the statue of Colossus in the harbour of
Rhodes), a coseismic uplift suddenly displaced the relative sea level from
about +3.4m to about -0.4m (A8). On the limestone layers which emerged at
that time, several coastal quarries were worked in Roman times, which are now
slightly submerged. Subsidence movements took place during the next fifteen
centuries, leaving marks of three shorelines (A6, A5 and A4), which are now
emerged, followed by renewed uplift during the last 400 years. Other examples
of coseismic vertical movements deduced from fossil shorelines in the eastern
 Tectonic shorelines 467

? 6 I a 3 2 1 0

TIME (ka BP) 4

z
0
p *1 *1
4
W
-4
w A7
0 Present M I L 0
Submerged
Ae quarrloi

-1 -1

Figure 12.1 1. Late Holocene relative sea-level changes in the northernmost part of the
east coast of Rhodes Island (Greece), from the city of Rhodes to Kalitea. Numbered
filled circles with horizontal/vertical bars correspond to radiocarbon-dated samples
with uncertainty ranges. See text for details. (From Pirazzoli er al., 1989.)

Figure 12.12.The notch with a hammer corresponds to shoreline A,, at about +3.4m at
Kalitea (northern part of Rhodes Island). It marks the sea-level position at the time the
Colossus was erected in Rhodes Harbour. In 222 BC a great earthquake is known to have
ruined the Colossus. The same earthquake, according to geomorphological and
radiometric evidence, is likely to have caused a coseismic uplift, displacing the relative
sea level from +3.4 m to about -0.4 m in this part of the island (photo P.A. Pirazzoli).
468 P.A. Pirazzoli
Mediterranean have been reported from Crete (Thommeret et al., 1981),
Turkey (Kelletat & Kayan, 1983; Pirazzoli et al., 1991) and Euboea Island
(Stiros et al., 1992).

Transform zones
The best known example of a transform fault zone crossing a coastal area is
probably that of the San Andreas Fault, which extends south to north over a
distance of about 1300km between the Gulf of California and the Mendocino
Fracture Zone. In the Santa Cruz area, six marine terraces indent a 60 km stretch
of coastal topography. Their raised shell beds were first described by Darwin
(1891). According to Valensise & Ward (1991), these terraces have been raised
by repeated, dominantly horizontal slip earthquakes on the San Andreas Fault,
which could recur every three to six centuries, uplifting the terraces at average
rates between 0.13mma-l and 0.35mma-l. In several areas along the San
Andreas Fault evidence of uplift is missing, however, and in southern California,
a recent aseismic uplift has been reported (Castle, Church & Elliott, 1976).

Volcanic zones
In volcanic areas, very rapid rates of vertical displacement can be reached.
Kaizuka, Miyauchi & Nagaoka (1983) recognized over 20 steps of marine
terraces, up to 120m above sea level; in Iwo (Sulphur) Island, a volcano
situated 1200km south of Tokyo, the average uplift rate was more than
' '
100 mm a- over several centuries, reaching 200 mm a- in some parts of the
island (Kaizuka, 1992). However, the best-known case study of irregular and
fast vertical movements is that of the Phlegraean Fields caldera near Naples,
Italy. Here, burrows of Lithophaga in the columns of the so-called Temple of
Serapis (a Roman market probably built in the second century BC near the
Pozzuoli Harbour), are today found at a level of several metres above present
sea level (Fig. 12.13). Geomorphological, archaeological and historical data
suggest a complex relative sea-level history, produced by alternating
subsidence and uplift. The subsidence rate seems to average 10 mm a-' in the
long term; after the Temple was constructed, subsidence reached a maximum
of 12 m in the tenth century AD, followed by an uplift of 7 m during the Middle
Ages, culminating (at a rate of 350 to 500mma-') in the Monte Nuovo
eruption in 1538 AD, and further subsidence (Suess, 1900; Gunther, 1903;
Flemming, 1969; Yokoyama, 197 1 ; Grindley, 1974). Tide gauge data from
Pozzuoli Harbour show two brief periods (1970 to 1973 and 1982 to 1984) of
rapid (up to 800mma-I) uplift, reaching a total of 3.2m between 1968 and
1984 (Berrino etal., 1984; Lajoie, 1986; Luongo & Scandone, 1991), followed
by renewed subsidence, which was still ongoing in October 1991.
 Tectonic shorelines 469

Figure 12.13. The temple of Serapis at Pozzuoli (Naples) in October 1991. Note the
dark band on the columns produced by molluscan borings (the upper limit of which
varies, according to Suess (l900), between 5.68 m and 5.98 m above the paving of the
temple) and the fact that the floor of the temple is almost completely dry (photo P. A.
Pirazzoli). The same floor still appears submerged in photographs taken in 1982, 1983
and 1984 (Vita-Finzi, 1986, pp. 10-1 1).

Effects of tectonics on coastal evolution

As shown above, rates and duration of tectonic movements can be very
variable and it is obvious that the morphodynamic evolution of a coastal area
which would be uplifted several metres in a sudden jerk, with a recurrence
period of several thousand years, will be different from that of the same area
uplifting gradually and continuously at the rate of 1 mm a- I . In particular,
dated tectonic events (e.g. coseismic uplift movements) can be used as a basis
for assessing development rates of constructional and destructional forms and
shore-forming processes (Kelletat, 1991).
The evolution of a shoreline depends mainly on the sea-level position, the
type of rock exposed, the shore profile and the local climatic, biological,
sedimentologic and hydrologic processes which can produce erosion or
accretion (Carter, 1988). Any relative sea-level change will not only displace
the shoreline kinematically at a new level, but also modify the zones exposed
to wave action and consequently longshore sediment transport and deposition.
Tectonics (in the broad sense) is one of the two main components of relative
470 P.A. Pirazzoli
sea-level changes, the other one being eustasy. During the Quaternary period,
eustatic changes consisted of sea-level oscillations at rates of the order of
10 mm a-' during periods of about 10 ka, interspaced with stillstands during
periods of a few thousand years. This means that on a tectonically stable coast
(assuming that such a coast exists), each main sea-level oscillation would
always affect the same vertical coastal zone. Each new shoreline would
consequently destroy the preceding one at the same level, or bury it inland if
accretion predominates so that such coasts would consist of broad coastal
plains bordered offshore by wide continental shelves and gentle continental
slopes. These coastlines would be characterized, according to Lajoie (1986),
by low-scale depositional landforms such as broad sandy beaches, offshore
barrier bars and low to moderate topographic relief.
The effect of a tectonic trend of uplift or subsidence, on the other hand, will
be to displace vertically the coastal range and continually create new
landforms. There is, however, a considerable difference between transitory
isostatic rebound (such as that induced by glacioisostasy or hydroisostasy) and
sustained tectonic deformation, since the former produces impacts very similar
to those of eustatic changes, whereas the displacements caused by the latter are
generally more durable. Nevertheless, near ice sheet areas, glacio-isostatic
changes can be much greater than eustatic changes (Fig. 12.7) and the morainic
material supplied by each new glaciation is so abundant that an almost
completely new situation will generally be found by the sea after each
oscillation, as in the case of sustained tectonic deformation.
More generally, uplift will tend to produce rugged coastlines and increased
erosion rates on the new relief. Deepening of valleys and stream beds by river
courses will bring more sediments to the coast and feed beaches where
regressional sedimentary sequences will produce progradational shorelines
with the development of shore-parallel sets of beach ridges (Carter, 1988). The
superimposition of long-term uplift trends on eustatic fluctuations will favour
the formation of stepped marine and river terraces and the development of
underground streams.
Subsidence on the other hand will tend to bury valleys and accumulate
sediments in coastal plains, where the reduced river slopes will favour the
development of peaty and marshy areas. Inland migration of shorelines will
result in land loss by submergence. In tropical areas, coral reefs will tend to
move away from the shore and construct barrier reefs. Outside tropical areas,
oceanic islands will gradually disappear, becoming submerged seamounts. The
evolution of shorelines is therefore closely dependent on the evolution of local
tectonic processes.
 Tectonic shorelines 47 1

References
Allard, M. & Tremblay, G. (1983). La dynamique littorale des iles Manitounuk durant
I’Holoche. Zeitschrift fur Geomorphologie, Suppl. Bd., 47,61-95.
American Geological Institute (1960). Glossary of geology and related sciences, 2nd
edn. Washington, DC.
Andrews, J.T. (1987). Glaciation and sea level: a case study. In Sea suvface studies: a
global view, ed. R.J.N. Devoy, pp. 95-126. London: Croom Helm.
Andrews, J.T. & Falconer, G. (1969). Late glacial and post-glacial history and
emergence of the Ottawa Islands, Hudson Bay, Northwest Territories: evidence
on the deglaciation of Hudson Bay. Canadian Journal of Earth Science, 6,
1263-76.
Atwater, B.F. (1987). Evidence for great Holocene earthquakes along the outer coast
of Washington State. Science, 236, 942-4.
Barnett, T.P. (1983). Recent changes in sea level and their possible causes. Climatic
Change, 5, 15-38.
Berrino, G., Corrado, G., Luongo, G. & Toro, B. (1984). Ground deformation and
gravity changes accompanying the 1982 Pozzuoli uplift. Bulletin
Volcanologique, 47-2, 187-200.
Berryman, K.R., Hull, A.G., Beanland, S., Pillans, B. & Deely, J. (1992). Excursion to
Cape Turakirae. In IGCP 274 Programme and Abstracts, Annual Meeting,
Wellington, New Zealand 1992, pp. 64-9. Wellington: Geological Society of
New Zealand miscellaneous publication 65A.
Berryman, K.R., Ota, Y. & Hull, A.G. (1989). Holocene paleoseismicity in the fold
and thrust belt of the Hikurangi subduction zone, eastern North Island, New
Zealand. Tectonophysics, 163, 185-95.
Bloom, A.L. (1971). Glacial-eustatic and isostatic controls of sea level since last
glaciation. In The Late Cenozoic glacial ages, ed. K.K. Turekian, pp. 355-79.
New Haven, CT: Yale University Press.
Bourcart, J. (1949). Gkographie du fond des mers. Paris: Payoy.
Brown, L.D., Reilinger, R.E., Holdahl, S.R. & Balazs, E.I. (1977). Postseismic crustal
uplift near Anchorage, Alaska. Journal ofGeophysica1 Research, 82, 3369-78.
Calmant, S. & Cazenave, A. (1986). The effective elastic lithosphere under the Cook-
Austral and Society islands. Earth and Planetary Science Letters, 77, 187-202.
Campbell, J.F. (1986). Subsidence rates for the southeastern Hawaiian Islands
determined from submerged terraces. Geo-Marine Letters, 6, 139-46.
Carobene, L. & Dai Pra, G. (1990). Genesis, chronology and tectonics of the
Quaternary marine terraces of the Tyrrhenian coast of northern Calabria (Italy):
their correlation with climatic variations. I1 Quaternario, 3, 75-94.
Carter, R.W.G. (1988). Coastal environments. London: Academic Press.
Castle, R.O., Church, J.P. & Elliott, M. (1976). Aseismic uplift in southern California.
Science, 192, 251-3.
Cazenave, A,, Lago, B., Dominh, K. & Lambeck, K. (1980). On the response of the
ocean lithosphere to sea-mount loads from Geos 3 satellite radar altimeter
observations. Geophysical Journal of the Royal Astronomical Society, 63,
233-52.
Clark, J.A., Farrell, W.E. & Peltier, W.R. (1978). Global changes in postglacial sea
level: a numerical calculation. Quaternary Research, 9,265-78.
Coudray, J. & Montaggioni, L. (1982). Coraux et rkcifs coralliens de la province indo-
pacifique: rkpartition gkographique et altitudinale en relation avec la tectonique
globale. Bulletin de la Sociktk Gkologique de France, 24, 98 1-93.
472 P.A. Pirazzoli
Daly, R.A. (1934). The changing world of the ice age. New Haven: Yale University
Press.
Darienzo, M.E. & Peterson, C.D. (1990). Episodic tectonic subsidence of late
Holocene salt marshes, northern Oregon central Cascadia margin. Tectonics, 9,
1-22.
Darwin, C. (1842). The structure and distribution of coral reefs. London: Smith, Elder
& co.
Darwin, C. (1891). Geological observations. London: Smith, Elder and Co.
Detrick, R.S. & Crough, S.T. (1978). Island subsidence, hot spots and lithospheric
thinning. Journal of Geophysical Research, 83, 1236-44.
Dubois, J., Launay, J. & Recy, J. (1974). Uplift movements in New Caledonia-Loyalty
Islands area and their plate tectonics interpretation. Tectonophysics, 24,
133-50.
Dumas, B., GuCrCmy, P., Hearty, P.J., LhCnaff, R. & Raffy, J. (1988). Morphometric
analysis and amino acid geochronology of uplifted shorelines in a tectonic
region near Reggio Calabria, south Italy. Palaeogeography, Palaeoclimatology,
Palaeoecology, 68,273-89.
Dumas, B., GuCrCmy, P., LhCnaff, R. & Raffy J. (1980). Terrasses quaternaires
soulevkes sur la faGade calabraise du DCtroit de Messine (Italie). Comptes
Rendus de 1’Acadimie des Sciences, Paris, D, 290,739-42.
Fairbridge, R.W. (1961). Eustatic changes in sea level. Physics and Chemistry of the
Earth, 4,99-185.
Flemming, N.C. (1969). Archaeological evidence for eustatic change of sea level and
earth movements in the western Mediterranean during the last 2000 years.
Geological Society of America, Special Paper, 109, 1-1 25.
Gaposchkin, E.M. (1973). Smithsonian standard Earth. SAO Special Report, 353.
Washington, DC: Smithsonian Institution, Astrophysical Observatory.
Grindley, G.W. (1974) Relation of volcanism to earth movements, Bay of Naples,
Italy. In Proceedings of the Symposium on ‘Andean and Antarctic Volcanology
Problems ’, Santiago, September 1974, pp. 598-6 12. Naples: Giannini.
Guilcher, A. (1988). Coral reef geomorphology. Chichester: Wiley.
Gunther, R.T. (1903). Earth-movements in the Bay of Naples. Geographical Journal,
22, 121-49 + 269-89.
Hamilton, E.L. (1957). Marine geology of the southern Hawaiian ridge. Bulletin of the
Geological Society of America, 68, 1011-26.
Hillaire-Marcel, C. (1976). La dkglaciation et le relkvement isostatique sur la cBte Est
de la baie d’Hudson. Cahiers de Gkographie du Quibec, 20, 185-220.
Hillaire-Marcel, C. (1 980). Multiple component postglacial emergence, Eastern
Hudson Bay, Canada. In Earth rheology, isostasy and eustasy, ed. N.A.
Momer, pp. 215-30. New York: Wiley.
Inman D.L. & Nordstrom, C.E. (1971). On the tectonic and morphologic classification
of coasts. Journal of Geology, 79, 1-21.
Jouannic, C., Taylor, F.W. & Bloom, A.L. (1982). Sur la surrection et la dCformation
d’un arc jeune: l’arc des Nouvelles-HCbrides. In Contribution d I’Etude
Giodynamique du Sud-Ouest Pacifique, Travaux et Documents ORSTOM,
147, pp. 2 2 3 4 6 . Paris: ORSTOM.
Kaizuka, S. (1992). Coastal evolution at a rapidly uplifting volcanic island: Iwa-Jima,
western Pacific Ocean. Quaternary International, 15/16, 7-1 6.
Kaizuka, S., Matsuda, T., Nogami, M. & Yonekura, N. (1973). Quaternary tectonic
and recent seismic crustal movements in the Arauco Peninsula and its environs,
central Chile. Geographical Reports of the Tokyo Metropolitan University, 8,
1-49.
 Tectonic shorelines 473
Kaizuka, S., Miyauchi, T. & Nagaoka, S. (1983). Marine terraces, active faults and
tectonic history of Iwo-jima (in Japanese, with English abstract). Ogasawara
Research, 9, 13-45.
Kelletat, D. (1991). The 1550 BP tectonic event in the eastern Mediterranean as a basis
for assessing the intensity of shore processes. Zeitschrift f u r Geomorphologie,
Supplement, 81, 181-94.
Kelletat, D. & Kayan, I. (1 983). First CI4 datings and Late Holocene tectonic events
on the Mediterranean coastline, west of Alanya, southern Turkey (in Turkish,
with English abstract). Tiirkiye Jeoloji Kurumu Bulteni, C26, 83-7.
Keraudren, B. (1 975). Essai de stratigraphie et de palCogCographie du Plio-PICistockne
Cgten. Bulletin de la Socie‘tk Gkologique de France, 17, 1 1 10-20.
Keraudren, B. & Sorel, D. (1984). Relations entre skdimentation, tectonique et
morphologie dans le Plio-PICistockne de Karpathos (Grtce): mouvements
verticaux et datations radiomktriques. L’ Antropologie (Paris), 88, 49-6 I .
Koba, M., Ikeya, T., Miki T. & Nakata, T. (1985). ESR ages of the Pleistocene
coral reef limestones in the Ryukyu Islands, Japan. In First International
Symposium on ESR Dating, Ube City, Japan, Tokyo: Ionics Publishing Co.
12 PP.
Konishi, K., Omura, A. & Nakamichi, 0. (1974). Radiometric coral ages and sea-level
records from the late Quaternary reef complexes of the Ryukyu Islands. In
Proceedings Second International Coral Reef Symposium, Brisbane,
pp. 595-613.
Labeyrie, J., Lalou, C. & Delibrias, G. (1969). Etude des transgressions marines sur
I’atoll de Mururoa par la datation des differents niveaux de corail. Cahiers du
Pacifique, 13, 59-68.
Lajoie, K.R. (1986). Coastal tectonics. In Active Tectonics, Studies in Geophysics,
pp. 95-124. Washington DC: National Academy Press.
Lambeck, K. (1981). Lithospheric response to volcanic loading in the Southern Cook
Islands. Earth and Planetary Science Letters, 55, 482-96.
Liew, P.M. & Lin, C.F. (1987). Holocene tectonic activity of the Hengchun Peninsula
as evidenced by the deformation of marine terraces. Memoirs of the Geological
Society of China, 9, 241-59.
Luongo, G. & Scandone, R., eds. (1991). Campi Flegrei. Journal of Volcanology and
Geothermal Research, 48, 1-227.
Marsh, J.G. & Martin, T.V. (1982). The Seasat altimeter mean sea surface model.
Journal of Geophysical Research, 87, 3269-80.
Martin, L., Flexor, J.M., Blitzkow, D. & Suguio, K. (1985). Geoid change indications
along the Brazilian coast during the last 7000 years. In Proceedings Fifth
International Coral Reef Congress, Tahiti, vol. 3, pp. 85-90.
McNutt, M. & Menard, H.W. (1978). Lithospheric flexure and uplifted atolls. Journal
of Geophysical Research, 83, 1206-12.
Menard, H.W. (1973). Depth anomalies and the bobbing motion of drifting islands.
Journal of Geophysical Research, 78,5 128-37.
Menard, H.W. (1986). Islands. Scientific American Library. New York: Freeman & Co.
Menard, H.W. & Ladd, H.P. (1963). Oceanic islands, sea-mounts, guyots and atolls.
In The sea, ed. M.N. Hill, pp. 365-87. New York: Interscience Publishers.
Montaggioni, L.F. ( 1985). Makatea Island, Tuamotu Archipelago. In Proceedings of
the Fifth International Coral Reef Congress, Tahiti, vol. 1, pp. 103-58.
Montaggioni, L.F. ( I 989). Le soul&vementpolyphas6 d’origine volcano-isostasique:
clef de I’Cvolution post-oligoche des atolls du Nord-Ouest des Tuamotus
(Pacifique central). Comptes Rendus de 1 ’Acade‘miedes Sciences, Paris, 11,309,
1591-8.
474 P.A. Pirazzoli
Morner, N.A. (1976). Eustasy and geoid changes. Journal of Geology, 84, 123-5 1.
Morner, N.A. (1988). Terrestrial variations within given energy, mass and momentum
budgets: paleoclimate, sea level, paleomagnetism, differential rotation and
geodynamics. In Secular solar and geomagnetic variations in the last 10,000
years, ed. F.R. Stephenson & A.W. Wolfendale, pp. 455-78. Dordrecht:
Kluwer Academic Publishers.
Mutti, E., Orombelli, G. & Pozzi, R. (1970). Geological studies on the Dodecanes
Islands (Aegean Sea). IX. Geological map of the island of Rhodes (Greece),
explanatory notes. Annales Gtologiques des Pays Helltniques (Athens),22,
77-226.
Nakada, M. & Lambeck, K. (1987). Glacial rebound and relative sea-level variations:
a new appraisal. Geophysical Journal ofthe Royal Astronomical Society, 90,
171-224.
Nakada, M., Yonekura, N. & Lambeck, K. (1991). Late Pleistocene and Holocene
sea-level changes in Japan: implications for tectonic histories and mantle
rheology. Palaeogeography, Palaeoclirnatology, Palaeoecology, 95, 107-22.
Nakata, T., Koba, M., Jo, W., Imaizumi, T., Matsumoto, H. & Suganuma, T. (1979).
Holocene marine terraces and seismic crustal movement. Science Reports of the
Tohoku University., 7th Series (Geography), 29, 195-204.
Nunn, P.D. (1986). Implications of migrating geoid anomalies for the interpretation of
high-level fossil coral reefs. Geological Society of America Bulletin, 97,
946-52.
Nunn, P.D. (1988). Plate boundary tectonics and oceanic island geomorphology.
Zeitsckrift fur Geomorphologie, Suppl. Bd., 69, 39-53.
Omura, A., Tsuji, Y., Ohmura, K. & Sakuramoto, Y. (1985). New data on Uranium-
series ages of Hermatypic corals from the Pleistocene limestone on Kikai,
Ryukyu Islands. Transactions and Proceedings of the Palaeontological Society
of Japan, NS, 139, 196-205.
Ota, Y. (1975). Late Quaternary vertical movement in Japan estimated from deformed
shorelines. Royal Society of New Zealand, Bulletin, 13, 23 1-9.
Ota, Y. (1985). Marine terraces and active faults in Japan with special reference to
coseismic events. In Tectonic geomorphology, ed. M. Morisawa & J.T. Hack,
pp. 345-66. Boston: Allen & Unwin.
Ota, Y. (1991). Coseismic uplift in coastal zones of the western Pacific rim and its
implications for coastal evolution. Zeitschriftfiir Geomorphologie, Suppl. Bd.,
81, 163-79.
Ota, Y., Hull, A.G. & Berryman, R. (1991a). Coseismic uplift of Holocene marine
terraces in the Pakarae River area, eastern North Island, New Zealand.
Quaternary Research, 35, 33 1-46.
Ota, Y., Miyauchi, T. & Hull, A.G. (1990). Holocene marine terraces at Aramoana
and Pourerere, eastern North Island, New Zealand. New Zealand Journal of
Geology and Geophysics, 33, 541-6.
Ota, Y. & Omura A. (1992). Contrasting styles and rates of tectonic uplift of coral
reef terraces in the Ryukyu and Daito Islands, southwestern Japan. Quaternary
International, 15/16, 17-29.
Ota, Y., Williams, D.N. & Berryman, K.R. (1981). Late Quaternary tectonic map of
New Zealand I : 50 000. Parts Sheets Q27, R27 and R28, Wellington.
Wellington: New Zealand Geological Survey.
Peltier, W.R. (1990). Glacial isostatic adjustment and relative sea-level change. In
Sea-level change, Geophysics Study Committee, National Research Council,
pp. 73-87. Washington, DC: National Academy Press.
 Tectonic shorelines 475
Peltier, W.R. (1991). The ICE-3G model of late Pleistocene deglaciation:
construction, verification, and applications. In Glacial isostasy, sea-level and
mantle rheology, NATO Advanced Science Institutes Series, C, 334, ed. R.
Sabadini, K. Lambeck & E. Boschi, pp. 95-1 19. Dordrecht: Kluwer Academic
Publishers.
Peltier, W.R. & Andrews, J.T. (1983). Glacial geology and glacial isostasy of the
Hudson Bay region. In Shorelines and isostasy, ed. D.E. Smith & A.G.
Dawson, pp. 285-3 19. London: Academic Press.
Pirazzoli, P.A. (1991). World atlas of Holocene sea-level changes. Elsevier
Oceanography Series, 58. Amsterdam: Elsevier.
Pirazzoli, P.A. & Grant, D.R. (1987). Lithospheric deformation deduced from ancient
shorelines. In Recent plate movement and deformation, ed. K. Kasahara,
pp. 67-72. American Geophysical Union & Geological Society of America,
Geodynamics Series, 20.
Pirazzoli, P.A., Koba, M., Montaggioni, L.F. & Person, A. (1988). Anaa (Tuamotu
Islands, central Pacific): an incipient rising atoll? Marine Geology, 82,
261-9.
Pirazzoli, P.A., Laborel, J., Salibge, J.F., Erol, O., Kaian, I. & Person, A. (1991).
Holocene raised shorelines on the Hatay coasts (Turkey): paleoecological and
tectonic implications. Marine Geology, 96, 295-3 1 1.
Pirazzoli, P.A. & Montaggioni, L.F. (1985). Lithospheric deformation in French
Polynesia (Pacific Ocean) as deduced from Quaternary shorelines. In
Proceedings Fifth International Coral Reef Congress, Tahiti, vol. 3,
pp. 195-200.
Pirazzoli, P.A., Montaggioni, L.F., Salikge, J.F., Segonzac, G., Thommeret, Y. &
Vergnaud-Grazzini C. (1989). Crustal block movement from Holocene
shorelines: Rhodes Island (Greece). Tectonophysics, 170,89-114.
Pirazzoli, P.A. & Salvat, B. (1992). Ancient shorelines and Quaternary vertical
movements on Rurutu and Tubuai (Austral Isles, French Polynesia). Zeitschrift
fur Geomorphologie, 36,43 1-5 1.
Pirazzoli, P.A. & Veeh, H.H. (1987). Age 230Th/234U d'une encoche CmergCe et
vitesses de soulkvement quaternaire h Rurutu, iles Australes. Comptes Rendus
de 1 'Acaddmie des Sciences, Paris, 11, 305, 9 19-23.
Plafker, G. & Rubin, M. (1967). Vertical tectonic displacements in south-central
Alaska during and prior to the great 1964 earthquake. Journal of Geosciences,
Osaka City University, 10, 53-66.
Rice, C.M. (1941). Dictionary of geological terms. Ann Arbor: Edwards.
Sabadini, R., Doglioni, C. & Yuen, D.A. (1990). Eustatic sea level fluctuations
induced by polar wander. Nature, 345, 708-10.
Savage, J.C. & Plafker, G. (1991). Tide gage measurements of uplift along the south
coast of Alaska. Journal of Geophysical Research, 96,4325-35.
Scott, G.A.J. & Rotondo, G.M. (1983a). A model for the development of types of
atolls and volcanic islands on the Pacific lithospheric plate. Atoll Research
Bulletin, 260, 1-33.
Scott, G.A.J. & Rotondo, G.M. (1983b). A model to explain the differences between
Pacific plate island atoll types. Coral Reefs, 1, 139-50.
Spencer, T. (1989). Tectonic and environmental histories in the Pitcairn Group,
Paleogene to Present: reconstructions and speculations. Atoll Research Bulletin,
322, 1-22.
Spencer, T. & Paulay, G. (1 989). Geology and geomorphology of Henderson Island.
Atoll Research Bulletin, 323, 1-18.
476 P.A. Pirazzoli
Stiros, S.C., Arnold, M., Pirazzoli, P.A., Laborel, J., Laborel, F. & Papageorgiou S.
(1992). Historical coseismic uplift in Euboea Island (Greece). Earth and
Planetary Science Letters, 108, 109-1 7.
Suess, E. (1900). La Face de la Terre (French translation of Das Antlitz der Erde),
t. 11, pp. 598-638. Paris: Colin.
Taylor, F.W., Isacks, B.L., Jouannic, C., Bloom, A.L. & Dubois, J. (1980). Coseismic
and Quaternary vertical tectonic movements, Santo and Malekula Islands, New
Hebrides Island arc. Journal of Geophysical Research, 85, 5367-8 1.
Thommeret, Y., Thommeret, J., Laborel, J., Montaggioni, L.F. & Pirazzoli, P.A.
(1981 ). Late Holocene shoreline changes and seismo-tectonic displacements in
western Crete (Greece). Zeitschrift f u r Geomorphologie, Suppl. Bd., 40,
127-49.
Valensise, G. & Pantosti, D. ( 1 992). A 125 kyr-long geological record of seismic
source repeatability: the Messina Straits (southern Italy) and the 1908
Earthquake (Ms 74). Terra Nova, 4,472-83.
Valensise, G. & Ward, S.N. (1991). Long-term uplift of the Santa Cruz coastline in
response to repeated earthquakes along the San Andreas Fault. Bulletin of the
Seismological Society of America, 81, 1-1 I .
Veeh, H.H. (1966). Th23n/U238 and U234/U238 ages of Pleistocene high sea level stand.
Journal of Geophysical Research, 71, 3379-86.
Vita-Finzi, C. (1986). Recenr earth movements. London: Academic Press.
Walcott, R.I. (1970). Flexure of the lithosphere at Hawaii. Tectonophysics, 9 , 4 3 5 4 6 .
Wellman, H.W. (1967). Tilted marine beach ridges at Cape Turakirae, New Zealand.
Journal of Geosciences, Osaka City University, 10, 123-9.
Woodroffe, C.D. (1988). Vertical movement of isolated oceanic islands at plate
margins: evidence from emergent reefs in Tonga (Pacific Ocean), Cayman
Islands (Caribbean Sea) and Christmas Island (Indian Ocean). Zeitschr$ f i r
Geomorphologie, Suppl. Bd., 69, 17-37.
Woodroffe, C.D., Stoddart, D.R., Spencer, T., Scoffin, T.P. & Tudhope, A.W. (1990).
Holocene emergence in the Cook Islands, South Pacific. Coral Reefs, 9, 31-9.
Yokoyama, I(1971). Pozzuoli event in 1970. Nature, 229, 532-3.
Yonekura, N. (1972). A review on seismic crustal deformations in and near Japan.
Bulletin of the Department of Geography, University of Tokyo, 4, 17-50.
Yonekura, N. (1975). Quaternary tectonic movements in the outer arc of southwest
Japan with special reference to seismic crustal deformations. Bulletin of the
Department of Geography, University of Tokyo, 7, 19-7 1.
Yoshikawa, T., Kaizuka, S. & Ota, Y. (1964). Mode of crustal movement in the late
Quaternary on the southeast coast of Shikoku, southwestern Japan (in Japanese
with English abstract). Geographical Review of Japan, 37, 27-46.
Yoshikawa, T., Kaizuka, S. & Ota, Y. (1981). The landforms of Japan. Tokyo:
University of Tokyo Press.
 13
Developed coasts
K.F. NORDSTROM

Introduction
Modification of the coast by humans has occurred over countless centuries
(Walker, 1984; 1988), but direct alteration of exposed ocean coasts on a
massive scale is a relatively recent phenomenon. Large-scale alterations began
in the nineteenth century, when there was a need to accommodate vessels of
large burden and deep water drafts and there was steam power to enable large
modifications (Marsh, 1885; De Moor & Bloome, 1988; Terwindt, Kohsiek &
Visser, 1988). Development of the coastlines of the world has accelerated in
the last few decades (Wong, 1985, 1988; Koike, 1988; Moutzouris &
Marouikian 1988; Cencini & Varani, 1988). Marco Island, on the west coast of
Florida, progressed from a wilderness to a fully developed shoreline in fewer
than 20 years (Reynolds, 1987). Many coastal communities in the USA
experienced dramatic growth in this period, with construction of high-rise
condominiums (Carter, 1982; Leatherman, 1987; Schmahl & Conklin, 1991).
By the mid-1970s 1687 km (or 37%) of the ocean frontage of the coastal
barriers on the US Atlantic and Gulf coasts were occupied by buildings, roads
and related features (Lins, 1980). Strip development near the ocean shoreline
has historically dominated the land conversion process in coastal communities
(Mitchell, 1987), placing the location of much of the development where it is
readily affected by wave and wind processes. Shore protection structures are
emplaced seaward of these developments where they can have the greatest
impact on shoreline processes and the most dynamic coastal landforms.
The proportion of coastline that is now protected against flooding and
erosion or that is significantly affected by human development is estimated at
12.9% in Italy (Cencini & Varani, 1988), 20.9% in Korea (Park, 1988), 38.5%
in England (Carr, 1988), 40.0% in Japan (Koike, 1988), and nearly the entire
coast of Belgium is protected by seawalls (De Moor & Bloome, 1988). Many
communities that are only partially stabilized are en route to total stabilization

477
478 K. F. Nordstrom
(Pilkey & Wright, 1988), and undeveloped shorelines that are adjacent to
developed and protected shorelines are often profoundly affected by human-
induced sediment starvation (Leatherman, 1984; Nordstrom, 1987a).
Communities that have been severely damaged by coastal storms are often
rebuilt to larger proportions (Fischer, 1989) (Fig. 13.1). Some areas have been
redesigned and rebuilt as human artifacts and bear litttle resemblance to the
coast that formerly existed (Nagao, 1991; Nagao & Fujii, 1991).
Many of the interactions between human activities and coastal processes
have been studied intensively. The engineering literature on the relationship
between structures, coastal processes and beach change is vast, although much
of this research is concerned with design of structures rather than the effects of
structures after they are emplaced. There is a large body of literature on the
effects of jetties on adjacent beaches (Kieslich, 1981; Dean, 1988; Hansen &
Knowles, 1988) and a growing body of literature of field-orientated studies
examining the effects of shore protection strategies at small temporal and
spatial scales, including the effects of groynes (Sherman et al., 1990; Bauer et
al., 1991), seawalls (Kraus & Pilkey, 1988; Plant & Griggs, 1992) and beach
nourishment (Schwartz & Bird, 1990). Many studies have been conducted into
recreational impacts on coastal dunes (Vogt, 1979). Study of the effects of
buildings or activities of shorefront residents on the morphology of the coast is

Figure 13.1. Gulf Shores, Alabama, I98 1, showing reconstruction after Hurricane
Frederick. The slogan during reconstruction ‘Gulf Shores: Bigger and Better’ reflects
the common practice of increasing the level of development after storm damage.
 Developed coasts 479
also represented, with most geomorphic studies on effects on coastal dunes
(e.g. Gares, 1983, 1990; Nordstrom & McCluskey 1985; Nordstrom, McCluskey
& Rosen, 1986; Nordstrom, 1988a). Insight into coastal changes at larger
spatial scales is provided in case studies of effects of a specific type of
protection strategy over several kilometres of shoreline or in several different
locations, including assessments of groyne fields (Everts, 1979), seawalls and
bulkheads (Fulton-Bennett & Griggs, no date) and beach nourishment
(Leonard, Clayton & Pilkey, 1990; Houston, 1991a). There are also case
studies of regional effects of several different kinds of protection structure
(Morton, 1979; Carter, Benson & Guy, 1981; Carter, Monroe & Guy,1986;
Zabawa, Kerhin & Bayley, 1981; Nordstrom, 1988b; Hall & Pilkey, 1991),
and the effect of storms on human structures (Griggs & Johnson, 1983; Fink1
& Pilkey, 1991).
The understanding of the relationship between human activities and coastal
evolution is far from adequate despite these previous studies. Most
investigations have a purpose other than to determine geomorphological
evolution; studies on beach nourishment, for example, usually focus on rates of
sediment loss or economic considerations, rather than on landform
assemblages associated with the new sediment (Houston, 1991b; Schmahl &
Conklin, 1991). The lack of scientific interest in the evolution of coasts
parallels the lack of interest in the evolution of many other developed physical
systems (Gregory, 1985). Many geomorphologists may not want to study
human developed systems because: they are accustomed to examining
landscape evolution at greater time scales than evolve under developed
conditions; they are deterred by the magnitude of the problem of isolating
cause and effect; or they think that geomorphological principles are
inappropriate where the dominant agent of landform change is earth-moving
machinery. In such cases, human alterations are viewed as an aberration, rather
than an integral component of landscape evolution.
Fundamental studies of the evolution of natural coasts are of considerable
scientific interest, but managerial interests require baseline information to
assess the numerous physical, chemical and environmental problems along our
shorelines, and it would be a mistake to neglect areas that have already been
developed. Coasts that are subject to rapid and largely uncontrolled occupation
are often the most at risk from environmental change in the near future, yet
paradoxically, it is these coasts that we often know least about. Although many
people visit developed shorelines, the scientist often rebuffs them, preferring
to seek out sites that are, superficially, more ‘natural’. This chapter attempts, in
a small way, to redress the balance of the rest of this volume, by exploring the
geomorphological evolution of developed coasts.
480 K. F. Nordstrom
The role of humans in the evolution of a developed coast is examined here
by comparing landform characteristics on an open-ocean barrier island coast
prior to extensive development with landform characteristics and evolutionary
trends under developed conditions. Specific examples are provided of the
evolution of the New Jersey coast (Fig. 13.2). A barrier island setting was
selected because of the strong dependence of geomorphic features on coastal
processes. New Jersey was selected because this location has the longest
history of development and stabilization of any barrier island coast in the USA
(Mitchell, 1987; Pilkey &Wright, 1988; Hall & Pilkey, 1991). Information on
the development process is provided for the entire exposed ocean coast, but
evolutionary trends are examined in detail on the barrier spits and islands
(termed barriers). Human adjustment dominates the landscape on some of the
barriers (Figs. 13.3 and 13.4) or has profoundly altered processes, sediment
budgets or landforms on other barriers. The barriers in New Jersey share
common characteristics with other coastal barriers in the USA, including

mkm

Philadelphia

Jersey lslpndEeachslptePark
4- @
- Atlantic
(Long Bwck la..) Ocean

Atlantic
Oceen

Figure 13.2. New Jersey shoreline study area. Dots represent locations of towns; arrows
identify locations of inlets; names in bold print in parenthesis identify barriers for which
data are presented in Table 13.1.
 Developed coasts 48 1

Figure 13.3. Atlantic City, NJ, on Absecon Island, showing pronounced human
influence on the coastal landcape of an intensively developed barrier island.

Figure 13.4. The barrier south of Sea Bright, NJ, showing the seawall that prevents an
inlet from forming at a location where inlets have formed and closed frequently in the
past.
 Table 13.1. Characteristics of the New Jersey shoreline in 1885-86 and on the most recent maps andphotos
~ ~ ~~ ~ ~ ~~

Dune characteristics (m) Width of upland (m)

Location and Minimum width of
year of map Highest elevation Maximum inland extent Maximum Minimum contiguous upland/marsh (m)

North spit
1886 7.6 375 1120 30 30
1981 ND ND 390 (1 193) 60 60
Headlands
NA NA NA NA NA
Island Beach
1886 4.6 610 920 130 140
1989 7.6 (10.7) 120 (580) 1 180 (570) 240 (97) 240 (I 40)
Long Beach Island
1886 9.1 600 720 140 140
1972 10.7 (4.0) 270 (280)) I310 (730) 180 (270) 180 (360)
Pullen Island
1886 3.0 260 740 280 280
1989 6.1 260 760 0 234
Brigantine Island
1885 4.6 470 530 100 260
1989 6.1 (3.0) 370 ( 130) 2070 (320) 340 (105) 340 (315)
Absecon Island
I885 8.5 560 1270 130 230
1989 None None 2980 190 190
Pecks Beach
1886 6.1 370 920 150 550
1972, 1989 4.6 (4.0) 190 (480) 1350 (490) 290 (280) 710 (280)
Ludlam Beach
1886 5.5 600 720 60 340
1972 8.2 200 940 150 290
7 Mile Beach
1886 13.1 680 960 270 500
1972 14.6 280 980 (210*) 390 (20*) 450 (loo*)
5 Mile Beach
1886 4.6 690 770 320 (50) 870
1972 None None 2120 790 790
2 Mile Beach
1886 4.6 190 480 160 420
1972 4.6 390 1120 390 1340
Cape May
1886 None None 280 60 I60
1972 None None 1034 720 720
~ ~ ~ ~ ~

All numbers from 18854 are taken from undeveloped locations. Numbers from most recent maps and photos are for developed portions of the
barriers; numbers in parenthesis are taken in undeveloped portions of developed barriers. All distances have been rounded to nearest 10m.
Elevations have been converted from feet to metres.
Most recent widths of upland (all land above marsh elevation) and uplandmarsh were taken from 1986 air photos at sites where the most recent
map is 1972.
*The undeveloped part of the barrier here at Stone Harbor Point has been eliminated since 1986 due to erosion (see text).
484 K. F. Nordstrom
Rockaway, New York, Ocean City, Maryland, Miami Beach, Florida and
Galveston, Texas, that have been developed for amenity housing, beach-
orientated recreation and related services. Most contain little agriculture or
woodland, and most lack industrial plants or commercial ports for use other
than fishing (Mitchell, 1987).

Methodology
The effects of development are assessed by comparing the dimensions,
configurations, topography, surface cover and mobility of the New Jersey
barriers prior to major human occupance with conditions represented on the
most recent maps and air photos. The first accurate maps of topography along
the New Jersey coast were drawn in 1885-86 using data from surveys
conducted the previous decade. The maps are at a scale of 1 :63 360; marsh
areas are delineated and the contour interval of 5 feet (1.52m) permits
evaluation of dune characteristics. Data on characteristics of the coastline
between 1885-86 and 1934 were also taken from 1 :63 360 scale maps. Data for
conditions after 1934 were determined using 1:9600 and 1:24000 scale
vertical air photos and 1:24 000 scale topographic maps. Information on dates
of construction of shore protection structures was taken from federal
documents (US Army Corps of Engineers (USACOE), 1954, 1957, 1990; US
Congress, 1976) and historical studies (Gares, 1983). The effect of a major
storm on the developed coastal barriers was determined by identifying changes
caused by the 6-7 March 1962 northeaster, which was the most damaging
storm on record, and comparing the effects of the storm with subsequent
human modifications to reconstruct the coastal landscape. Information on the
likelihood of implementation of future protection strategies was derived from
recent planning documents (New Jersey Department of Environmental
Protection (NJDEP), 1981; USACOE, 1989a, b, c) supplemented by interviews
with key personnel in the State and the US Army Corps of Engineers.

Characteristics of the New Jersey coast

The New Jersey ocean coast (Fig. 13.2) is approximately 205 km long and
consists of a northern barrier spit; a headland 4.5 to 7.5 m high, composed of
unconsolidated sediments; a southern barrier spit and barrier island complex;
and a short southern headland near Cape May that is composed of Pleistocene
sand and gravel. Mean tidal range is about 1.3 m along the open coast (National
Oceanic and Atmospheric Administration (NOAA), 1991). The rates of sea-
level rise at Sandy Hook and Atlantic City, based on trends between 1940 and
 Developed coasts 485
1970, are 4.6mma-' and 2.8mma-' (NJDEP, 1981). Dominant winds blow
from the northwest, although storm waves generated by northeasterly winds
account for the dominance of southerly drift south of Barnegat Inlet (Fig. 13.2).
The northern spit and headlands are sheltered from the full effects of these
waves by Long Island, northeast of the study area (Fig. 13.2, inset), and net
drift is to the north. Littoral drift rates increase to the north and south of a nodal
zone located between Manasquan and Barnegat Inlets (Ashley, Halsey &
Buteux, 1986). The drift rate at Sandy Hook is estimated at 382000m3a-'
to the north (Caldwell, 1967); the drift rate at Cape May is estimated at
190000m3a-l to the south (US Congress, 1976). Annual average significant
wave height is 0.8 m with an average wave period of 8.3 s (Thompson, 1977).
Grain sizes of native beach sediments decrease from coarse sand in the
northern headlands to fine sand in the south (McMaster, 1954).

Human modifications on the New Jersey coast

Conditions prior to extensive human development

Native Americans used the shoreline to fish and gather shells during the
warmer months, but they do not appear to have occupied the barriers
(Koedel, 1983). Only a few hunting cabins, homesteads and boarding houses
existed on the islands before construction of railroads in the mid and late
nineteenth century; the barrier islands were used for livestock grazing, timber,
whaling, farming and shell-fishing (Sea Isle City, 1982; Koedel, 1983).
Atlantic City was the first major barrier island resort, being built up rapidly
after a rail line was completed in July 1854.
Most of the barriers were low and narrow prior to development and they
were backed by marsh deposits colonized principally by Spartina alternifloru
and Spurtinu patens. Some of the marshes appear to have formed on locations
of overwash of sediment from the ocean to the bay or former flood deltas. The
upland portions of the barriers were at locations of dunes, former spit recurves
and zones of overwash that covered former marsh surfaces. The widths of the
upland portions of the barriers (Table 13.1, columns 4 and 5) varied from a
maximum of over 1000m at former spit recurves to minimums of <100 m
where wave processes had recently closed inlets through longshore sediment
transport. Multiple dune ridges were common on portions of several of the
barriers. The foredune appeared to take the form of a broad-based ridge along
much of Absecon Island and on Seven Mile Beach (Fig. 13.2) where dunes
were highest (Table 13.1, column 2), but large portions of most barriers were
characterized by isolated hummocky dunes and 28.3% of the barriers appeared
486 K .F. Nordstrom
to have no dunes at all. Dunes were least well-developed near active and
recently active inlets that had closed naturally. Vegetation on the upland
portion of the barriers was characterized by lush growth of cedar, holly and
other trees and a variety of grasses (Sea Isle City, 1982). The coastal barriers
appeared to be highly mobile prior to human development. State Geologist
G.H. Cook, cited in USACOE (1957), noted in 1882 that stumps of trees felled
with a metal axe and hoof prints of shod horses in peat that were found exposed
on the beach at Long Beach Island after a storm indicated that the island
migrated across its width in no less than 200 years. Recent dating of peat
buried offshore of Ocean City indicates a 2000 m inland shift of that barrier in
the past 5730 years (Psuty, 1986).

Construction of buildings and support infrastructure

Growth of coastal resorts was rapid following construction of the railroads.
Atlantic City had a permanent population of nearly 5500 in 1880 (Funnel,
1975), and there were 650 hotels and boarding houses by 1891 (Reynolds,
1981). Railroad lines extended along 57.8% of the New Jersey shoreline in
1885-86, although roads extended along only 19.1% of the barriers. Four of
the barriers lacked access by either road or rail at that time. The barrier island
resort industry developed rapidly in the early twentieth century, due primarily
to increased use of private automobiles (Koedel, 1983).
The first buildings were constructed on the upland portions of the barriers at
inlets and at locations where railway lines from the mainland first made
contact with the barriers. New isolated communities appeared, and growth
then extended outward from these locations, both alongshore on the upland
portion of the barriers and bayward onto the marsh surface. Dunes were graded
to a flatter form to facilitate construction of buildings and roads in the very first
stages of development. Most of the natural vegetation was destroyed in the
process of this conversion.
Inspection of maps reveals that local filling of the marsh behind the upland
occurred on several barriers between 1886 and 1902. The marsh at Ocean City
was filled all the way to the edge of the backbay by 1907. The marsh behind
the barrier north of Cape May was completely covered to a landward distance
of over 700m by 1913. Dredging of channels into the backbarrier marsh to
accommodate boats was first accomplished in the larger settlements south of
Ocean City and occurred between 1905 and 1913. Filling of the bay landward
of the barriers to accommodate new buildings was never widespread, but fill
was used to facilitate road access to bridges to the mainland.
 Developed coasts 487
The sequence of development observed at the locations that developed early
was followed at later time periods on other barriers. Conspicuous filling of the
marsh did not take place near Mantoloking (Fig. 13.2) until the period between
1914 and 1930, when there is the first evidence of lagoon development. Large-
scale conversion of the marsh to lagoon developments did not occur on many
of the barriers until after World War 11, when changes became rapid and
widespread (Fig. 13.5). Most of these projects used the existing marsh as
substrate and placed materials dredged from the new waterways on the marsh
surface. The shapes of the lagoon developments usually mimic the shapes of
the marsh in plan view (Fig. 13.5), but the function of the backbarrier changed
from a marsh to a residential area. Construction of lagoon housing is now
severely restricted by regulations governing use of wetlands, but marshland is
now found in only isolated enclaves on the back sides of the developed
portions of barriers in New Jersey.
Native plant species are prevented from recolonizing on most of the barriers
because of the human preference for using lawn grass and exotic shrubs and
trees for landscaping. Many landowners have replaced all vegetation with
gravel, using weed killer to maintain an unvegetated surface, and these
locations are now barren (Fig. 13.6). Natural vegetation other than foredune
communities only remains in a few natural zones that are maintained as
preserves.

Figure 13.5. South Mantoloking, NJ, showing the construction of lagoon housing on
marsh that was common practice prior to environmental regulations in the 1970s.
488 K. F. Nordstrom

Figure 13.6. Developed barrier near Barnegat Inlet, NJ, showing the level of
development characteristic of most of the barrier islands in New Jersey, the
omnipresent groynes and the small dune used as protection. Photo. was taken just after
the northeasterly storm March 29, 1984.

Activities at inlets
All but eleven of the inlets that existed in 1885-86 and all inlets that formed
since that time were closed artificially by the US Army Corps of Engineers or
kept from re-opening after natural closure to eliminate undesirable shoreline
fluctuations, facilitate land transportation along the coast, or increase the
hydraulic efficiency of nearby controlled inlets. Five of the natural inlets are
now stabilized by jetties. Jetty construction confines flow between structures,
resulting in abandonment of natural ebb and marginal flood channels, causing
wave dominance of adjacent relict ebb tidal delta areas, and resulting in
landward migration of bars and elimination of typical ebb tidal deltas
(Hubbard, 1975; Hansen & Knowles, 1988). Jetties reduce the rate of shoreline
change by preventing inlets from migrating. The average absolute rate of both
erosion and accretion decreased dramatically at the inlets after jetties were
constructed. Shore-parallel change in the throats of the inlets decreased from
an average of 5.6 m a-I to 0.0 m a - ' at the three inlets where jetties are closely
spaced and from 10.7ma-' to 4.0ma-l at Barnegat and Absecon Inlets
(Fig. 13.2), where the jetties are wide enough apart for beaches to form
 Developed coasts 489

between them (Nordstrom, 1987b). The mobility of beaches on the ocean

shoreline updrift and downdrift of jetties decreased from an average of
3.8ma-' to 2.4ma-' at the five inlets. The net change is unidirectional
(accretion updrift, erosion downdrift) rather than bidirectional (erosion-accretion
cycles) as occurred before construction of the jetties (Nordstrom, 1987b).
The ebb tidal delta of a jettied inlet can develop to a new equilibrium size
(Dean & Walton, 1975; Marino & Mehta, 1988), allowing sediment to pass to
the downdrift shoreline. Bypassing is not achieved on most New Jersey inlets
with jetties because of the prominence of the structures (which often act as
terminal groynes) and frequent channel dredging. Disposal of material dredged
from inlets at sea represents a substantial loss of sediment from the longshore
transport system (Dean 1988; Marino & Mehta, 1988). Much of the sediment
dredged from Manasquan, Absecon and Cape May Inlets has been dumped at
sea, although some of this sediment has been placed on adjacent beaches in the
past. The volumes dredged from these three inlets following jetty construction
have exceeded 5.8 X 106m3(Nordstrom, 1987b).
Shorelines are developed adjacent to four of the inlets that are not stabilized
by jetties, and two of these unjettied inlets are maintained by dredging.
Dredging at these inlets has changed the amount of sediment transferred across
inlets and has influenced the location of accretion and erosion on adjacent
shorelines by either changing the location of tidal channels or maintaining
them in place, depending on human preference. Maintenance dredging of an
existing channel configuration is the more common. This practice keeps the
channel from fluctuating as widely as it would under natural conditions, and it
reduces the periodicity of, or virtually eliminates, erosion/deposition cycles
associated with breaching of the ebb tidal delta. The mobility of developed
inlet shorelines in New Jersey can be less than occurred under natural
conditions, even in the absence of jetties or dredging, because bulkheads
adjacent to inlets truncate shoreline displacement during erosional phases of
cycles of accretion and erosion or prevent breaches in the barriers updrift of
inlets (Nordstrom, 1988b).

Shore protection projects

Efforts were made to control erosion as early as 1847 at Cape May and 1857 at
Atlantic City (USACOE, 1957). The first documented case of groyne
construction on the New Jersey shoreline was at Sandy Hook in the 1860s
(Bearss, 1976). Structures resembling seawalls (but termed breakwaters)
existed at Sandy Hook in 1863 and bulkheads were employed there as early as
1878 (Gares, 1983). Bulkheads existed at Sea Bright in 1880 (Psuty, 1988).
490 K. F. Nordstrom
Fig. 13.7 reveals trends in the implementation of shore protection projects
since 1900. The data are based on inventories by the Corps of Engineers
(USACOE, 1954, 1957, 1990) and include local and state projects as well as
federal projects. The totals represent a conservative estimate of the structures
emplaced, because other narratives make reference to structures that are not
reported in the Corps inventories, and many structures identified in the
inventories that were repaired are not counted in the total. The trends reveal a

Groynes
loo 600

80 500
400
300
200
20 100

Bulkheads, seawalls, revetments

8 50

6
40

85 4 130

c
" 2
20

10

0 0
loo0 1920 1ou) 1960 1980 2Ooo 1900 1920 1940 1960 1980 2Ooo
Yau Year

Beach fill

n
E
--.-s
~

f s
4E
5 a.-g
4 "
d
B
1wo la0 1wo 1960 ls80 2000 3 1900 1920 1940 1960 1980 2000
Year Year

Figure 13.7. Implementation of shore protection projects since 1900.

 Developed coasts 49 1
temporal shift in preference for each of the three strategies, from extensive use
of groynes, through a period of extensive construction of shore-parallel
structures (bulkheads, seawalls, revetments), to the present emphasis on beach
nourishment. A considerable number of projects using all three methods were
employed following the storm of 6-7 March 1962 (Fig. 13.7), but groynes and
seawalls have been employed sparingly since the 1960s. Beach nourishment is
now the preferred alternative, but the groynes and shore-parallel structures
remain in place and still have a pronounced effect on shoreline change. Dune
building programmes have been implemented in many communities, and there
have been demonstration projects for small-scale, non-traditional techniques
of erosion control. The effects of the demonstration projects on coastal
evolution have been negligible, but dune building programmes have had a
significant impact on coastal change and are discussed below.

Shore-parallel structures
Over 43 km of New Jersey shoreline are protected by shore-parallel structures
(USACOE, 1990). The majority of them are timber bulkheads, usually 3.0 to
4.0m above mean low water (MLW), but stone and concrete seawalls and
revetments are common. The seawall at Sea Bright (Fig. 13.4) is the largest
structure, with a top elevation of 5.8 m above MLW.
Bulkheads are also common on the bay shoreline (Fig. 13.5) because the
unconsolidated sand used as fill material to provide substrate for house
construction on marshes is mobilized readily by bay waves and currents. These
structures are not represented in Fig. 13.7. They are poorly documented
because design criteria and details of emplacement are lacking and public
funds are rarely required for their construction; the cost of construction is
within the financial capability of homeowners, and permission to construct
bulkheads was easy to obtain.

Groynes
The groynes are constructed of timber, quarrystone or a combination of timber
and stone. They vary greatly in length and height. Some are extremely
effective at trapping sediment, such as the 240 m terminal groyne downdrift of
the Cape May groyne field and the 245m terminal groynes just north of
Hereford Inlet. Many groynes are over 3 m above MLW at the outer end, but
more recent groynes, especially in the southern part of the state, have been
built or rebuilt to allow some sand to bypass them and are as low as 0.6 m at the
outer end.
Groynes reduce the seasonal fluctuations in beach profile (Everts, 1979)
and lower the rate of beach retreat within the groyne field, but they also cause
492 K.F. Nordstrom
accelerated erosion downdrift. The degree of sand starvation downdrift of
groynes, like that downdrift of jetties, is dependent on the efficiency of the
groyne system in trapping sand entering from updrift and the amount of
sediment deflected seaward as well as the rate of net transport to gross
transport and the length and height of the terminal structure (Everts, 1979).

Downdrift efects of structures

Significant regional erosion effects have occurred downdrift of jetties at
Manasquan Inlet and Cape May Inlet where the average rate of shoreline
retreat has been about 3 m a-I (Everts, 1979). Erosion downdrift of Barnegat
and Absecon Inlets has been offset by beach nourishment operations, and
bulkheads and groynes have been constructed downdrift of all jettied inlets to
mitigate adverse erosion. Accelerated erosion has occurred downcoast of the
terminal groynes at Sandy Hook, the developed southern end of Long Beach
Island, Stone Harbor Point (just north of Hereford Inlet), and Cape May City.
The starvation downdrift of the Cape May groyne field is about 2500 m long
with a retreat rate of about 6 m a-I (Everts, 1979). The erosion rate downdrift
of groynes and bulkheads constructed at Stone Harbor Point was 36 m a - I over
the 12 year period following their construction (Nordstrom, 1988b) and the
1.8 km long spit that existed south of the terminal groyne in 1972 is now gone.

Beach nourishment
Data on past beach nourishment operations funded by the State and local
communities are difficult to find and information about privately funded
projects is often not available at all in the public domain (Pilkey & Clayton,
1989), yet existing reports indicate that at least 101 nourishment operations
have been conducted in New Jersey, involving over 29.8 X lo6 m3 of sediment
(USACOE, 1990). Much of the fill emplaced before the 1970s was obtained
from the backbays. More-recent operations have used sediment from offshore
and from inlets. Some of the fill materials have been placed offshore of the
beach foreshore in an attempt to induce natural feeding but have had little
impact on the sediment budget of the beach (US Congress, 1976).
Nourished areas are often readily distinguishable from adjacent areas
because of their great width, particularly during the first few years of
emplacement, and the presence of coarse sediments on the surface of the
backbeach. Accelerated deflation occurs, but dunes are rarely allowed to form
in intensively developed areas; drift accumulations on the beach are removed
to retain wide, flat recreation platforms; and drift accumulations behind the
beach are removed because sand inundates boardwalks and buildings. The fill
is quickly reworked to a steeper upper beach, but eventually, the erosion rate
 Developed coasts 493
slows, and the beach returns to a gentler profile similar to pre-fill conditions.
Nourishment provides a temporary alteration of the process of deposition, the
rate of change, and the location, size, shape, and internal structure of the
landform. The periodic implementation of beach fill projects introduces a
cycle of beach evolution that has a time scale determined by human activities.

Dunes
Dunes are now well established in more than half of the shorefront
communities. Most of the communities that have dunes have a budget for dune
building, and many communities have implemented regulations to prevent
construction of houses in these locations. The State of New Jersey requires
communities to conduct dune-building programmes as a condition for aid for
hazard mitigation (Mauriello & Halsey, 1987), and dunes may be conspicuous
in locations where they would not occur under natural processes nor under
human influence, except for legal requirements (Fig. 13.3). These management
factors increase the likelihood that a dune will form and persist (Nordstrom &
Gares, 1990), but cycles of erosion and accretion as well as the form,
composition and position of the dune may reflect the specific management
practice employed locally. Wave erosion of the seaward face of the dune may
be more rapid than it was when the dune was farther back from the beach under
undeveloped conditions, but restoration of the dune is rapid, because it is aided
by human efforts.
In some locations, the building-up and levelling of the dune follows a
seasonal cycle but solely under the influence of human processes; dunes are
created in the autumn to provide a barrier against storm waves and are
eliminated in the summer to accommodate bathers. In some cases, dunes are a
by-product of beach cleaning operations; the beach surface is scraped and piled
at the landward margin of the beach. Paradoxically, the actions of beach
cleaning (removal of vegetation litter, with its shoots and seeds) eliminates one
of the conditions needed to initiate natural development of foredunes. Dunes
created by earth-moving equipment from beach sediments include size
fractions that could not be moved by eolian processes. The coarse sediments
may form a surface lag, preventing further deflation by eolian processes, which
gives the landform an unnatural appearance.
Dunes that are built with the aid of sand fences and vegetation plantings are
larger than those that would form behind narrow beaches under natural
conditions. Conversely, dunes are absent in several resort communities, where
beaches are hundreds of metres wide, because dunes are graded. Relative to
natural dunes, dunes in human-modified environments may be smaller or
larger for a given source width or grain size because of active or passive human
494 K .F. Nordstrom
modifications. Once dunes in developed areas are shaped according to human
needs, attempts are made to protect them in place to retain their utility. Thus,
initial mobility is high, but the dunes become less mobile than their
undeveloped counterparts (Nordstrom, 1990). The location of the dune on the
beach profile may be different from the location under natural conditions
because dune position is dictated by human preference rather than the interplay
between vegetation growth, sediment supply and wave erosion.

Small-scale effects
Small-scale human-altered landforms may not affect long-term coastal
evolution but they are of interest to geomorphologists, engineers, planners and
managers. Buildings alter wind flow and locally change the location of
accretion and scour on beaches and dunes (Nordstrom & McCluskey, 1985;
Nordstrom et al., 1986). The large buildings on some portions of the New
Jersey shoreline cause a reversal of the regional wind flow and result in
onshore transport during offshore winds; the buildings can create a deflation
surface and accelerate scour of the backbeach at considerable distances from
the buildings (Nordstrom, 1987a; Gundlach & Siah, 1987). Wind-blown sand
accumulates against groynes, bulkheads, boardwalks and buildings, creating a
diverse dune landscape, characterized by shapes and locations distinct from
those that would occur under natural conditions.
Human alterations may be as subtle as the introduction of new heavy
mineral assemblages in beach fill. This change may be trivial in terms of
dynamic coastal evolution, but alteration of the character of the previous
natural fill may eventually prevent interpretation of provenance and ancient
clastics (Galvin, 1991) and obscure interpretation of natural coastal evolution.
The introduction of exotic species has had a pronounced impact on the
evolution of the coastal landscape in some portions of the USA. The most
widespread alteration of dune characteristics attributed to the introduction of
exotic species has occurred along most of the Pacific coast. Here, European
beach grass (Ammophila arenaria), artificially planted to stabilize inlet
shorelines, spread rapidly, creating a higher, more linear and better vegetated
foredune than existed previously. The new dune is a more complete trap to
sand blown landward of the beach, and it has profoundly altered the
morphology and biota of the littoral zone (Cooper, 1958; Wiedemann, 1984).
The impact of exotic species is less pervasive in New Jersey. Developed areas
are frequently barren of vegetation, both native and exotic. The introduced
reedgrass (Phragmites australis) has colonized portions of the backbarrier
above mean high water level. Japanese sedge (Carex cobomugi) is found in
 Developed coasts 495
several areas in the dunes. These species appear conspicuously different from
the natural vegetation, but data are insufficient to determine whether there is a
morphologic response associated with them that would affect the evolution of
landforms.

Effects of storms and post-storm construction

The effects of the northeasterly storm of 6-7 March 1962 and subsequent
human activities demonstrate the relative roles of natural processes and human
agency in determining the evolution of the New Jersey barriers. This storm was
classified as an unusually severe extratropical cyclone. Wind speeds were
relatively low, with maximum gusts of 25.5 m s-', but the fetch distance for
wave generation was about 1600km. Visual estimates of the height of breaking
waves range from 6.1 to 9.1 m (USACOE, 1962). The increase in ocean water
level due to storm surge was 1.2 m. The major reason for widespread erosion
of beaches and dunes and damage to buildings was the duration of the storm
over five high tides. Tidal flooding was common on all the barriers and
accounted for most of the economic losses. Waves and overwash had very
different effects on different portions of the barriers. The greatest
geomorphological changes and damages to buildings were on Long Beach
Island (Fig. 13.2), where beaches were narrow and dunes were low prior to the
storm. A veneer of fresh sand existed everywhere on the surface of this barrier
following the storm, but the major overwash fans appeared where street ends
were located. Overwash penetrated into the bay, even where there were
bulkheads on the bayside (Fig. 13.8). The barrier was breached in five places.
Overwash also reached the bay on Ludlam Beach north of Sea Isle City, but
did not reach the bay on the other barriers. Many oceanside bulkheads on other
barriers failed during the storm, but there was little wave effect landward of
locations where bulkheads remained intact (USACOE, 1962).
The storm caused considerable property damage, but little lasting
geomorphological effect. All breaches that occurred in developed areas as a
result of the storm were closed artificially, starting 9 March, after the storm
abated (USACOE, 1962). There was little new substrate for salt marsh and
little effect on island widening by natural processes. Dredging of lagoons to
facilitate boating resulted in loss of potential for new salt marsh substrate to
form. Overwash fans on the subaerial portions of the island increased ground
elevation, but these locations were not colonized by natural vegetation. There
was greater inland penetration of overwash across the surfaces of shore-
perpendicular streets, but this sand was removed to facilitate transportation.
The loss of sand from the beach and dunes on the oceanside was compensated
for by the creation of a new dune using sediments dredged from borrow areas
496 K.F. Nordstrom

Figure 13.8. Effects of the March 1962 storm on Long Beach Island. The overwash
channels and deposits in the backbay are similar to those that would occur under natural
conditions. Photo. courtesy of the US Army Corps of Engineers.

in the bay. The new dune was linear and shore-parallel, and bore little
resemblance to the hummocky dune that characterized many areas prior to
development and that would have characterized natural dune growth adjacent
to the overwash channels that were created during the storm (Fig. 13.8).
Construction of new buildings and facilities occurred at a rapid pace after the
storm; many of the new structures were more elaborate than those constructed
prior to 1962 (US Congress, 1976), and the new seaward-most construction
line was at the same location it was prior to the storm. The net effect of the
storm in altering the location of the shoreline, re-creating natural landforms or
re-initiating new cycles of natural landform evolution was negligible. The
large number of shore protection projects implemented after the storm
(Fig. 13.7) reduced the likelihood of pronounced geomorphological effects of
future storms.
The normal sequence of events following storm alteration of natural
systems is creation of a post-storm landscape with subsequent evolution of
landforms according to natural processes. In human-altered systems, however,
the post-storm landscape is quickly restored to characteristics suitable to
human perception of the value of the resource. Severe damage to beachfront
homes by storms in 1878 and 1884 at Atlantic City did not prevent post-storm
 Developed coasts 497
reconstruction (USACOE, 1990). Even in the nineteenth century, the scale of
alterations resulting from storms was small enough that human action could
restore buildings and infrastructure before natural processes could re-establish
natural landforms that could dominate the landscape or control coastal
evolution. As a result of past precedent, and taking into account the
development pressure and current level of investment in many barrier islands,
it is likely that future structures damaged by storms will be replaced
(USACOE, 1989a) or there will be an increase in the level of development and
protection.

Human alterations and the coastal landscape

Evolution through past modifications

Offshore depth contours of 1.8, 3.7 and 5.5m below MLW have generally
moved landward on the New Jersey coast since the mid nineteenth century
(USACOE, 1957), but the upper foreshores have undergone less erosion since
development, and several barriers have an accretional trend (Everts &
Czerniak 1977; Dolan, Hayden & Heywood, 1978; Galvin, 1983). As a result,
there has been little onshore migration of the intertidal shoreline over the past
century.
Beach berm widths now vary from <15 m in many locations to a maximum
of about 400m near Hereford Inlet (USACOE, 1990). The maximum
elevations of isolated dunes in developed areas and in presently undeveloped
areas are generally higher than they were in 1886, but they do not extend as far
inland (Table 13.1). Dunes do not exist in many areas, but these locations are
now protected by bulkheads and seawalls. There are numerous locations where
dunes exist but are no higher than 1.5 m above the elevation of the backbeach
and no wider than 15m (USACOE, 1990). These locations would be
eliminated or washed over in a major storm like the one that occurred in March
1962. Despite this limitation to their human utility value, the dunes are a more
continuous barrier than they were prior to development in 1886, and they are
more likely to restrict barrier island mobility.
The greatest change in the dimensions of the barriers due to development is
the dramatic increase in the width of upland (both maximum and minimum)
through filling of the marsh. The minimum widths in Table 13.1 were
measured at the narrowest portion of each barrier in each of the two time
periods, although these locations are usually not at the same place. The
minimum width of contiguous upland and marsh is a measure of the potential
for breaching of the barrier. This potential has decreased at all barriers except
498 K .F. Nordstrom
Absecon Island and Ludlam Beach, although the narrowest portion of Absecon
Island is now protected by bulkheads on both sides of the island. The lower
width at Five Mile Beach under developed conditions (Table 13.1) does not
represent a narrowing of the island through time. The narrowest location is
now at the location of the former inlet that separated Five Mile Beach from
Two Mile Beach and thus reflects a dramatic increase in barrier dimensions
rather than a loss. The minimum widths of the barriers in presently
undeveloped enclaves (in parenthesis in Table 13.1) may be greater or lesser
than in 1886, depending on the nature of events at the adjacent inlets. The
dramatic decrease in minimum width on the undeveloped portion of Seven
Mile Beach is due to erosion resulting from human-induced sediment
starvation at Stone Harbor Point.
Phases in the evolution of barrier islands through time are presented in
idealized form in Figs. 13.9 and 13.10. Absolute dates are not specified
because different barriers reach a given phase at different times because of

Phasa 1: Railroad bulldlna and lnltlal devel0Dment

Phase 2: Devdo~nwnlof upland. lmglnlng of marsh filling and groyne emplacement

Pham 3: Laaaan dnvalooment. bulkheadlnaand nourishment

Marsh Dunes Railroad Bulkhead

0Upland Roads Groynes Jew

Figure 13.9. Plan view of representative stages in development and protection of barrier
islands through time. Transects A and B are locations represented in profiles in Fig. 13.10.
 Developed coasts 499
Tranmcl A Tranmcl B
Beforedevelopmenl
n-d"
upvnd

ocu, w OM
0V-h
10 4 &Y

Phasa 2 Phase 2
Ur-vln W

Phase 3 Phase 3
"?

Figure 13.10. Profile view of representative stages in development and protection of

barrier islands through time. Transects A and B are locations represented in plan view
in Fig. 13.9.

differences in proximity to major population centres and ease of access. The

activities include: elimination of dunes to accommodate initial construction;
fill of marsh to increase developable land and cutting of lagoons to
accommodate boats; protection using groynes and bulkheads; construction of
jetties and implementation of dredging projects to maintain a static shoreline
and a predictable navigation channel at inlets; and beach nourishment to
provide protection and maintain a recreation platform. These activities have
resulted in an increase in the width of the islands, a reduction in the rate of
erosion, and a reduction in the role that natural processes play in altering the
configuration of the islands or the likelihood that they will migrate.
Differences through time represent differences in intensity of development,
which result in selection of different protection alternatives. Several
communities in New Jersey are now protected by a combination of beach
nourishment, groynes, bulkheads and dunes. Most communities employ at
least two of these methods of protection. The level of development at Transect
A in Figs. 13.9 and 13.10 has been sufficiently intensive to justify use of large-
scale nourishment operations to protect buildings and infrastructure. Human
actions along Transect B indicate a reliance on less costly public expenditures
and include dune building and flood proofing (accomplished by raising
buildings) with a bulkhead providing back-up protection.
500 K.F. Nordstrom

Planned future actions

Groynes are not likely to be employed extensively in the future, although
protection projects for potential implementation in New Jersey identify both
construction of new groynes and maintenance of existing groynes (USACOE,
1989a, b, c). Large structures that are planned for future implementation
include a weir-type breakwater and deposition basin at Absecon Inlet
(Fig. 13.2) and a new north jetty and weir at Townsends Inlet along with
extension of the south jetty. Sand bypass at these inlets would reduce the rate
of net change on both sides of the inlet. The progression of human alterations
at these inlets would take them from a condition of cyclic change, with high
rates of mobility, through a phase of unidirectional change with slightly lower
rates of shoreline mobility, to a phase of limited or no cyclic change with low
rates of mobility.
Most of the planned expenditures for shore protection projects are for beach
nourishment operations. A new project began recently at Ocean City, and
projects are planned for the vicinity of Barnegat Inlet, Sea Isle City and Sea
Bright. The Sea Bright project would restore a beach to a location that has not
had one in decades. The creation of new dunes is not always specified in
design plans. Dunes are included in plans for Brigantine Island. These dunes
would have a top design elevation of 4.6 m (15 feet) and the lowest elevation
of the entire protected proportion of the island would be as high as the highest
elevation in 1885 (Table 13.1).

Scenarios of future change

The Seabright and Cape May seawalls are considered the final stage of the
progression of human alterations to maintain a fixed position on a migrating
shoreline (Pilkey, 1981). In this stage, beaches and dunes are replaced by
bulkheads and seawalls; shoreline position is static; sediment availability is
minimal; and the waves are incapable of creating an equilibrium feature.
Nordstrom (1988b) presents a scenario of change for unjettied inlets that has
similarities to Pilkey’s, in that it is based on a reliance on structural protection
and is driven by limitations in sediment availability. The two scenarios are
similar to the no-action scenario used by the Corps of Engineers to determine
the benefit of future protection projects. The assumptions include: 1. the
present bulkheads or groynes will be maintained or improved; 2. sediment
budgets will be diminished in locations where the structures interfere with
sediment transport; 3. shoreline mobility will be reduced in areas protected by
structures; 4. inlets that are now dredged will continue to be dredged, and
existing channels will be maintained in place. Scenarios like these describe a
 Developed coasts 501
closed system of coastal evolution. They are grounded, at a basic level, in the
physical processes operative on developed coastal barriers, but they are
simplistic in evaluating future human responses. Rejuvenation of features
characteristic of a natural landscape can occur with massive inputs of beach
nourishment or changes in regulations that prohibit or limit construction or
reconstruction of protection structures. Activities in New Jersey indicate that
rejuvenation is occurring at a massive scale through implementation of large-
scale beach nourishment projects. These operations re-establish the potential
for an open system (cyclic) model of beach change. The re-establishment of
the potential for cyclic change is the result of conscious human decisions (and
the availability of massive funding). Whether natural geomorphic features will
form on or adjacent to new beaches is dependent on human values for the
resource. Although restored beaches are likely to evolve under natural cycles
of wave action in the short term (at least between nourishment operations),
transfers between the beach and locations farther inland may be prevented by
human actions. A lack of linkage between the beach and the remainder of the
barrier can occur under both the beach nourishment scenario and the seawall
scenario.
Titus (1990) combines physical processes and human action, including
beach nourishment, in four potential scenarios of change for developed
barriers in response to an increased rate of sea-level rise. These may be
compared to the two responses postulated by Psuty (1986) for the New Jersey
barriers (without regard to human action) that include drowning in place and
migration through overwash, wind transport and inlet sedimentation. The
scenarios for developed barriers (Titus, 1990) include: 1. no protection,
leading to eventual abandonment (although development compatible with a
dynamic barrier is still possible); 2. engineered retreat that mimics natural
retreat by artificially filling the bay sides of barriers while the oceanside
erodes; 3. raising the barrier in place by placing sand on the beach and
concomitantly raising buildings and support infrastructure; and 4. constructing
seawalls and flood protection structures around the barrier, creating a ring-
levte enclave. These options were examined in regard to Long Beach Island,
NJ (Fig. 13.2) because its level of development (similar to that portrayed in
Fig. 13.6) placed it near the middle of the spectrum in regard to development
density. The first two scenarios would mimic the migration model of Psuty
(1986); the second two follow the model for drowning in place, although
drowning would be prevented by human efforts.
Titus uses a greater number of assumptions and more explicit assumptions
about human actions than are usually employed in models of evolution for
developed coasts, considering factors such as preference and precedent, rights
502 K.F. Nordstrom
of property owners, legal implications, economic motivation, environmental
interests and constraints of government programmes. The most telling finding
of the economic evaluation is that the rental value for Long Beach Island
exceeds the protection cost for the most expensive option (raising the barrier),
leaving the relevant question ‘how’, not ‘whether’ to protect the island. Given
the economics of protecting Long Beach Island, many developed barriers
could justify protection for almost any conceivable rate of sea-level rise or unit
cost of sand for nourishment (Titus, 1990). The response of the barriers to
accelerated sea-level rise may mimic the natural response by migrating or
remaining in place as water level rises, but the role of human agency in the
process could be greater than at present.
The likelihood that future shore protection projects will be implemented
increases with increasing levels of development. The Corps of Engineers
found that protection of Brigantine Island was not economically justified in
1957, but protection was justified in 1989 (USACOE, 1989a). Federal funding
increases the likelihood of human alterations and their scale. The sequence of
human alterations on the New Jersey coast indicates that the occupation of the
coastal fringe is widespread, inevitable, and incontrovertible under present
management practice (Nordstrom, 1990).

Discussion
Human action may be considered as an external factor that perturbs or changes
the natural system or as part of the system (Phillips, 1991). Human action can
be considered an exogenous factor if the temporal or spatial scale under
consideration is an order of magnitude greater than the temporal or spatial
scale of human influence, or an endogenous factor if the scales under study and
the scales of human agency are similar in magnitude (Phillips, 1991). Human
alterations on the New Jersey shoreline are endogenous in that they are are
persistent, recurrent, and are more frequent than changes made by natural
processes or equal or exceed the spatial scale of natural changes.
There are three basic approaches used to investigate the effect of human
activities on coastal systems. These may be described as the overlay method,
the no-action method and the active human-input method. The overlay method
compares and contrasts a developed area with an undeveloped area that is
assumed to have the same process controls. The undeveloped area may be the
same location prior to development (the approach used here), or it may be a
different location. Use of the same location prior to development may result in
databases that differ in methods of collection, level of detail, or quality of
measurement. Use of an undeveloped shoreline segment that is distant from
 Developed coasts 503
the developed shoreline is problematic because of differences in setting, such
as wave climate or tidal regime. An undeveloped shoreline segment that is
adjacent to a developed segment may have a similar wave climate and tidal
regime but may be affected by the developed segment through sand starvation
(as at Stone Harbor Point). Care must be taken to ensure that a segment that
lacks buildings or infrastructure actually represents a natural condition. Use of
sand fences, vegetation plantings and introduction of exotic species by humans
has converted thousands of kilometres of dune landscape in the USA (Cooper,
1958; Wiedemann, 1984; Godfrey & Godfrey, 1973). Many hundreds of
kilometres of these coastlines have no human structures to provide clues to
human influence. The degree of naturalness of the undeveloped coast should
be specified in all comparative studies.
The no-action method assumes that the kind of shoreline change that
occurred in the recent past will continue unabated by local actions. This
method is useful to evaluate potential changes in the absence of future
adjustments so that these potential changes can be used as the basis for
selecting the optimum solution for protection (Nordstrom, 1988b), but it is
unrealistic for long-term prediction because of the likelihood of future human
action.
Pre-requisites for successful experimental designs in geomorphological
investigations include understanding of appropriate processes and responses
and selection of appropriate means of measuring and analyzing the pertinent
parameters (Sherman, 1989). The active human input approach is the most
realistic to use for investigation of shorelines like New Jersey, where future
changes depend on human decisions to implement numerous or large-scale
structural or non-structural solutions. Predictions must be based on
probabilities of human action, calibrated with knowledge of physical processes
in a multiple scenario format, requiring that coastal scientists know planning
and management policies (Nordstrom, 1987c) and the likelihood that they will
be implemented or altered during the time span represented in models of
coastal evolution.
Human influence is sufficiently important to change both the space and
time frameworks within which geomorphological processes should be
examined (Trofimov, 1987). Human processes alter the previous rate of
shoreline change and the periodicities of landform evolution, and they change
the location, size, shape, and internal structure of the landforms and the
tendency to undergo cyclic development in both the short and long term.
Prediction according to existing geomorphological principles of natural
processes and coastal evolution in New Jersey is restricted to locations
seaward of the berm crest on nourished shorelines and seaward of the seaward-
504 K. F. Nordstrom
most line of pioneer vegetation on most of the non-nourished shorelines. The
history of landform changes in developed systems is determined by human
action, and the linkage between natural process and form is obscured
(Nordstrom, 1990). Although human alterations obscure the effects of natural
processes, they still provide the best clues to the future evolution of coastal
systems where human influence is an endogenous factor.

Conclusions
The incompatibility of many human alterations with the way coastal barriers
change through natural processes is obvious. What is not obvious is the
relevance of natural models of coastal evolution to coastal barriers that have
passed the critical threshold beyond which return to a naturally functioning
system is not an acceptable management option. Traditional models of coastal
evolution based on natural processes may be appropriate to anticipate future
effects on undeveloped barriers and barriers where the scale of human
activities is too small to overcome natural changes, but the precedent
established on the barriers in New Jersey and many other developed barriers in
the USA indicate that there may be fewer of these locations in the future.

References
Ashley, G.M., Halsey, S.D. & Buteux, C.B. (1986). New Jersey’s longshore current
pattern. Journal of Coastal Research, 2,453-63.
Bauer, B.O., Allen, J.R., Nordstrom, K.F. &Sherman, D.J. (1991). Sediment
redistribution in a groin embayment under shore-normal wave approach.
Zeitschrift fur Geomorphologie, Suppl. Bd., 81, 135-48.
Bearss, E. (1976). Historic resources study: Sandy Hook defenses 1857-1948. Draft
report prepared for the Denver Service Center, National Park Service. Denver,
CO: National Park Service.
Caldwell, J.M. (1967). Coastal processes and beach erosion. CERC Reprint 1/67. Ft.
Belvoir, VA: US Army Corps of Engineers Coastal Engineering Research
Center.
Carr, A.P. (1988). UK - England. In Artificial structures and shorelines, ed. H.J.
Walker, pp. 137-44. Dordrecht: Kluwer Academic Publishers.
Carter, C.H., Benson, D.J. & Guy, D.E., Jr. (1981). Shore protection structures:
effects on recession rates and beaches from the 1870’s to the 1970’s along the
Ohio shore of Lake Erie. Environmental Geology, 3, 353-62.
Carter, C.H., Monroe, C.B. & Guy, D.E., Jr. (1986). Lake Erie shore erosion: the
effect of beach width and shore protection structures. Journal of Coastal
Research, 2, 17-23.
Carter, R.W.G. (1982). Condominiums in Florida. Geography, 62,41-3.
Cencini, C. & Varani, L. (1988). Italy. In Artificial structures and shorelines, ed. H.J.
Walker, pp. 193-206. Dordrecht: Kluwer Academic Publishers.
Cooper, W.S. (1958). The coastal sand dunes of Oregon and Washington. Geological
Society of America Memoir, 72.
 Developed coasts 505
Dean, R.G. (1988). Sediment interaction at modified coastal inlets: processes and
policies. In Hydrodynamics and sediment dynamics of tidal inlets, ed. D.G.
Aubrey & L. Weishar, pp. 412-39. New York: Springer-Verlag.
Dean, R.G. & Walton, T.L,. Jr. (1975). Sediment transport processes in the vicinity of
inlets with special reference to sand trapping. In Estuarine research, vol. 2, ed.
L.E. Cronin, pp. 129-50. New York: Academic Press.
De Moor, G. & Bloome, E. (1988). Belgium. In Artificial structures and shorelines,
ed. H.J. Walker, pp. 115-26. Dordrecht: Kluwer Academic Publishers.
Dolan, R., Hayden, B. & Heywood, J. (1978). Analysis of coastal erosion and storm
surge hazards. Coastal Engineering, 2,41-54.
Everts, C.H. (1979). Beach behavior in the vicinity of groins: two New Jersey field
examples. In Coastal Structures 79, pp. 853-67. New York: American Society
of Civil Engineers.
Everts, C.H. & Czerniak, M.T. (1977). Spatial and temporal changes in New Jersey
beaches. In Coastal Sediments 77, pp. 444-59. New York: American Society of
Civil Engineers.
Finkl, C. & Pilkey, O.H., eds. (1991). The impacts of Hurricane Hugo: September
10-22, 1989. Journal of Coastal Research, SI 8.
Fischer, D.L. (1989). Response to coastal storm hazard: short-term recovery versus
long-term planning. Ocean and Shoreline Management, 12, 295-308.
Fulton-Bennett, K. & Griggs, G.B. (no date). Coastal protection structures and their
eflectiveness. Sacramento, CA: California Department of Boating and
Waterways.
Funnell, C.E. (1975). By the beautiful sea. New York: Alfred Knopf.
Galvin, C. (1983). Sea level rise and shoreline recession. In Coastal Zone 83,
pp. 2684-705. New York: American Society of Civil Engineers.
Calvin, C. (1991). Native sand beaches: a disappearing research resource. SEPM
News, 3 (3), 3.
Cares, P.A. (1983). Historical analysis of shoreline changes at Sandy Hook spit. In
Applied coastal geomorphology at Sandy Hook, New Jersey, ed. K.F.
Nordstrom, J.R. Allen, D.J. Sherman, N.P. Psuty, L.D. Nakashima & P.A.
Cares, pp. 15-50. New Brunswick, NJ: Rutgers University Center for Coastal
and Environmental Studies Technical Report CX 1600-6-0017.
Cares, P.A. (1990). Eolian processes and dune changes at developed and undeveloped
sites, Island Beach, New Jersey. In Coastal dunes: form andprocess, ed. K.F.
Nordstrom, N.P. Psuty & R.W.G. Carter, pp. 361-80. Chichester: John Wiley.
Godfrey, P.J. & Godfrey, M.M. (1973). Comparison of ecological and geomorphic
interactions between altered and unaltered barrier island systems in North
Carolina. In Coastal geomorphology, ed. D.R. Coates, pp. 239-58.
Binghamton, NY: State University of New York.
Gregory, K.J. (1 985). The nature of physical geography. London: Edward Arnold.
Griggs, G.B. &Johnson, R.E. (1983). Impact of 1983 storms on the Coastline.
California Geology, 36, 163-74.
Gundlach, E.R. & Siah, S.J. (1987). Cause and elimination of the deflation zones
along the Atlantic City (New Jersey) shoreline. In Coastal Zone 87, pp.
1367-69. New York: American Society of Civil Engineers.
Hall, M.J. & Pilkey, O.H. (1991). Effects of hard stabilization on dry beach width for
New Jersey. Journal of Coastal Research, 7,771-85.
Hansen, M. & Knowles, S.C. (1988). Ebb-tidal delta response to jetty construction at
three South Carolina inlets, In Hydrodynamics and sediment dynamics of tidal
inlets, ed. D.G. Aubrey& L. Weishar, pp. 364-8 1. New York: Springer-Verlag.
Houston, J.R. (l991a). Beachfill performance. Shore and Beach, 59, 15-24.
506 K. F. Nordstrom
Houston, J.R. (1991b). Rejoinder to: discussion of Pilkey and Leonard (1990)
[Journal of Coastal Research, 6(4) 1023 et seq.] and Houston (1990) [Journal
of Coastal Research, 6(4) 1047 et seq.]. Journal of Coastal Research, 7,
565-77.
Hubbard, D.K. (1975). Morphology and hydrodynamics of the Merrimack River ebb-
tidal delta. In Estuarine research, vol. 2, ed. L.E. Cronin, pp. 253-66. New
York: Academic Press.
Kieslich, J.M. (1981). Tidal inlet response to jetty construction. GITI Report 19.
Vicksburg, MI: US Army Engineer Waterway Experiment Station.
Koedel, C.R. (1983). History of New Jersey’s barrier islands. In New Jersey’s barrier
islands: an ever-changing public resource, pp. 17-28. New Brunswick, NJ:
Rutgers University Center for Coastal and Environmental Studies.
Koike, K. (1988). Japan. In Artificial structures and shorelines, ed. H.J. Walker, pp.
317-30. Dordrecht: Kluwer Academic Publishers.
Kraus, N.C. & O.H. Pilkey, eds. (1988). Effects of seawalls on the beach. Journal of
Coastal Research, SI 4.
Leatherman, S.P. (1984). Shoreline evolution of North Assateague Island, Maryland.
Shore and Beach, 52,3-10.
Leatherman, S.P. (1987). Approaches to coastal hazard analysis: Ocean City,
Maryland. In Cities on the beach, ed. R.H. Platt, S.G. Pelczarski & B.K.R.
Burbank, pp. 143-54. Chicago: University of Chicago Department of
Geography Research Paper 224.
Leonard, L., Clayton, T. & Pilkey, O.H. (1990). An analysis of replenished beach
design parameters on U S . east coast barrier islands. Journal of Coastal
Research, 6 , 15-36.
Lins, H.F. (1980). Patterns and trends of land use and land cover in Atlantic and Gulf
coast barrier islands. Reston, VA: US Geological Survey Professional Paper
1156.
Marino, J.N. & Mehta, A.J. (1988). Sediment trapping at Florida’s east coast inlets. In
Hydrodynamics and sediment dynamics of tidal inlets, ed. D.G. Aubrey &
L. Weishar, pp. 284-96. New York: Springer-Verlag.
Marsh, G.P. (1885). Earth as modijied by human action. New York: Charles
Scribner’s Sons.
Mauriello, M.N. & Halsey, S.D. (1987). Dune building on a developed coast. In
Coastal Zone 87, pp. 762-79. New York: American Society of Civil Engineers.
McMaster, R.L. (1954). Petrography and genesis of the New Jersey beach sands.
Trenton, NJ: New Jersey Geological Survey Bulletin 63.
Mitchell, J.K. (1987). A management-oriented, regional classification of developed
coastal barriers. In Cities on the beach, ed. R.H. Platt, S.G. Pelczarski & B.K.R.
Burbank, pp. 31-41. Chicago: University of Chicago Department of Geography
Research Paper 224.
Morton, R.A. (1979). Temporal and spatial variations in shoreline changes and their
implications, examples from the Texas gulf coast. Journal of Sedimentary
Petrology, 49, 1101-12
Moutzouris C.I. & Maroukian, H. (1988). Greece. In Artificial structures and
shorelines, ed. H.J. Walker, pp. 207-15. Dordrecht: Kluwer Academic
Publishers.
Nagao, Y., ed. (1991). Coastlines of Japan. New York: American Society of Civil
Engineers.
Nagao, Y. & Fujii, T. (1991). Construction of man-made island and preservation of
coastal zone. In Coastlines of Japan, ed. Y. Nagao, pp. 212-26. New York:
American Society of Civil Engineers.
 Developed coasts 507
National Oceanic and Atmospheric Administration (NOAA) ( 1 99 I). Tide Tables
1991: East Coast of North and South America. Washington, DC: US
Department of Commerce, NOAA.
New Jersey Department of Environmental Protection (NJDEP) (1981). New Jersey
Shore Protection Master Plan. Trenton, NJ: New Jersey Department of
Environmental Protection.
Nordstrom, K.F. (1987a). Shoreline changes on developed coastal barriers. In Cities on
the beach, ed. R.H. Platt, S.G. Pelczarski & B.K.R. Burbank, pp. 65-79.
Chicago: University of Chicago Department of Geography Research Paper 224.
Nordstrom, K.F. (1987b). Management of tidal inlets on barrier island shorelines.
Journal of Shoreline Management, 3, 169-90.
Nordstrom, K.F. (1987~).Predicting shoreline changes at tidal inlets on a developed
coast. Professional Geographer, 39,457-65.
Nordstrom, K.F. (1988a). Dune grading along the Oregon coast, USA: a changing
environmental policy. Applied Geography, 8, 101-16.
Nordstrom, K.F. (1988b). Effects of shore protection and dredging projects on beach
configuration near tidal inlets in New Jersey. In Hydrodynamics and sediment
dynamics of tidal inlets, ed. D.G. Aubrey& L. Weishar, pp. 440-54. New York:
Springer-Verlag.
Nordstrom, K.F. (1990). The intrinsic value of depositional coastal landforms.
Geograpical Review, 80, 68-8 1.
Nordstrom, K.F. & Gares, P.A. (1990). Changes in the volume of coastal dunes in
New Jersey, USA. Ocean and Shoreline Management, 13, 1-10.
Nordstrom, K.F. & McCluskey, J.M. (1985). The effects of houses and sand fences on
the eolian sediment budget at Fire Island, New York. Journal of Coastal
Research, 1, 39-46.
Nordstrom, K.F., McCluskey, J.M. & Rosen, P.S. (1986). Aeolian processes and dune
characteristics of a developed shoreline. In Aeolian geomorphology, ed. W.G.
Nickling, pp. 131-47. Boston: Allen and Unwin.
Park, D.W. (1988). Korea - South. In ArtiJicial structures and shorelines, ed. H.J.
Walker, pp. 3 1 1-1 6. Dordrecht: Kluwer Academic Publishers.
Phillips, J.D. (1991). The human role in earth surface systems: some theoretical
considerations. Geographical Analysis, 23, 3 16-3 1.
Pilkey, O.H. (1981). Geologists, engineers, and a rising sea level. Northeastern
Geology, 314, 150-8.
Pilkey, O.H. & Clayton, T.D. (1989). Summary of beach replenishment experience on
U.S. east coast barrier islands. Journal of Coastal Research, 5, 147-59.
Pilkey, O.H. & Wright, H.L., 111 (1988). Seawalls versus beaches. Journal of Coastal
Research, SI 4,41-64.
Plant, N.G. & Griggs, G.B. (1992). Interactions between nearshore processes and
beach morphology near a seawall. Journal of Coastal Research, 8, 183-200.
Psuty, N.P. (1986). Impacts of impending sea-level rise scenarios: the New Jersey
barrier island responses. Bulletin of the New Jersey Academy of Sciences, 31,
29-36.
Psuty, N.P. (1988). USA - New Jersey and New York. In Artificial structures and
shorelines, ed. H.J. Walker, pp. 573-80. Dordrecht: Kluwer Academic
Publishers.
Reynolds, W.J. (1981). The social and geomorphic forces that shaped New Jersey’s
shore. Unpublished seminar paper. New Brunswick, NJ: Department of
Geography, Rutgers University.
Reynolds, W.J. (1987). Coastal structures and long term shore migration. In Coastal
Zone 87, pp. 414-26. New York: American Society of Civil Engineers.
508 K.F. Nordstrom
Schmahl, G.P. & Conklin, E.J. (1991). Beach erosion in Florida: a challenge for
planning and management. In Coastal Zone 91, pp. 261-71. New York:
American Society of Civil Engineers,
Schwartz, M.L. & Bird, E.C.F. (1990). Artificial beaches. Journal of Coastal
Research. SI 6.
Sea Isle City (1982). Sea Isle City Centennial 1882-1982. Sea Isle City, NJ: City Hall.
Sherman, D.J. (1989). Geomorphology: praxis and theory. In Applied geography:
issues, questions, and concerns, ed. M.S. Kenzer, pp. 1 15-3 1. Dordrecht:
Kluwer Academic Publishers.
Sherman, D.J., Bauer, B.O., Nordstrom, K.F. & Allen, J.R. (1990). A tracer study of
sediment transport in the vicinity of a groin. Journal of Coastal Research, 6,
427-38.
Terwindt, J.H.J., Kohsiek, L.H.M. & Visser, J. (1988). The Netherlands. In Artificial
structures and shorelines, ed. H.J. Walker, pp. 103-14. Dordrecht: Kluwer
Academic Publishers.
Thompson, E.F. (1977). Wave climate at selected locations along US.coasts.
Technical Report 77-1. Ft. Belvoir, VA: US Army Corps of Engineers Coastal
Engineering Research Center.
Titus, J.G. (1990). Greenhouse effect, sea level rise, and barrier islands: case study of
Long Beach Island, New Jersey. Coastal Management, 18, 65-90.
Trofimov, A.M. (1987). On the problem of geomorphological prediction. Catena
Supplement, 10, 193-97.
US Army Corps of Engineers (1954). Beach erosion control report on cooperative
study (survey) Atlantic coast of New Jersey, Sandy Hook to Barnegut Inlet.
Philadelphia: US Army Corps of Engineers, Philadelphia District.
US Army Corps of Engineers (1957). Beach erosion control report on cooperative
study (survey) of the New Jersey coast Barnegat Inlet to the Delaware Bay
entrance to the Cape May canal. Philadelphia: US Army Corps of Engineers,
Philadelphia District.
US Army Corps of Engineers ( I 962). Coastal storm of 6-7 March 1962: postflood
report. Philadelphia: US Army Corps of Engineers, Philadelphia District.
US Army Corps of Engineers (1 989a). Benefits reevaluation study: Brigantine Island,
New Jersey. Philadelphia: US Army Corps of Engineers, Philadelphia District.
US Army Corps of Engineers (1989b). Benefits reevaluation study: Absecon Island,
New Jersey. Philadelphia: US Army Corps of Engineers, Philadelphia District.
US Army Corps of Engineers (1989~).Benefits reevaluation study: Townsends
InletBeven Mile Island, New Jersey. Philadelphia: US Army Corps of
Engineers, Philadelphia District.
US Army Corps of Engineers (1990). New Jersey shore protection study: report of
limited reconnaissance study. Philadelphia: US Army Corps of Engineers,
Philadelphia District.
US Congress (1976). New Jersey coastal inlets and beaches Hereford inlet to
Delaware Bay entrance to Cape May canal. House Document 94-641.
Washington, DC: US Government Printing Office.
Vogt, G. (1979). Adverse effects of recreation on sand dunes: a problem for coastal
zone management. Coastal Zone Management Journal, 6,37-68.
Walker, H.J. (1984). Man’s impact on shorelines and nearshore environments: a
geomorphological perspective. Geoforum, 15, 395-4 17.
Walker, H.J. (1988). Artificial structures and shorelines: an introduction. In Artificial
structures and shorelines, ed. H.J. Walker, pp. 1-8. Dordrecht: Kluwer
Academic Publishers.
 Developed coasts 509
Wiedemann, A.M. (1984). The ecology of Pacific northwest coastal sand dunes: a
community profile. Washington, DC: US Department of the Interior Fish and
Wildlife Service.
Wong, P.P. (1 985). Artificial coastlines: the example of Singapore. Zeitschrift f u r
Geornorphologie, Suppl. Bd., 57, 115-92.
Wong, P.P. (1988). Singapore. In Artificial structures and shorelines, ed. H.J. Walker,
pp. 383-92. Dordrecht: Kluwer Academic Publishers.
Zabawa, C.F., Kerhin, R.T. & Bayley, S. (1981). Effects of erosion control structures
along a portion of the northern Chesapeake Bay shoreline. Environmental
Geology, 3, 201-1 1.
INDEX

Index Terms Links

accommodation space 42 52 98 107

126 138 158 173
Adelaide River 193 202 212
aeolianite 234 235 244 252
287
Alaska 344 348 352 373
394 435 463
Aldabra Atoll 289 429
allocyclicity 97
alluvial fan 353 374
Amanzimnajama Lake 227
Anaa Tuamotus 289 454
anabranch 202
antecedent karst 273 287 313
antecedent topography 14 34 42 122
127 133 187 192
204 208 227 246
249 344 357 361
380 399
anthropogenic influence 21 320 477

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

on cliffs 435
on deltas 112
on reefs 320 328
Antrim Cape 405
Arctic coastal plain 341
armouring 11 383
Atchafalaya Delta 98 110 113
Atlantic City 481 484 489 496
atoll 5 267 304 452
Aldabra 289 429
Cocos (Keeling) Islands 267
Funafuti 267
Midway 272
Mururoa 272 453
structure 268 304 452
Tuamotu 272 282 289 454
autocyclicity 97
avalanching 227
Avalon Peninsula 406
Avulsion 97 113 203

back barrier 225 230 235 400

channels 232
flats 237
marsh 486

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

backwater swamp 188 191 213

barrier 16 20 48 52
54 70 123 189
191 221 244 397
399
bayhead 378 405
beach 362 392 394
breach 47 55 225 232
244 495 497
cell-confined 232
clastic 221
cuspate 221 237
erosion 244
fringing 404 412
gravel 10 14 16 159
378 397
looped 237
mainland-attached 250
mainland beach 159
overstepping 48 54 133 244
250
overwash 158 231 243 251
361 485 495
Pleistocene 230 247
pocket 237
prograding 58 131 140 152

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

254
receded 159
regressive 140 234 253
retreat 234 497
stationary 156
transgressive 61 127 146 149
165 168 250
translation 48 132 245
barrier island 124 127 140 155
158 250 353 361
393 480 484 497
drowned 127 365
Bay of Funday 105 210 383 444
beach morphodynamics 9 47 68 124
beach nourishment 478 491 499
beach ridge 97 104 253 394
397 405 407 470
beachridge plain 11 99 152 192
204 225 398 402
beachrock 252 286
Beaufort Sea 341 346 352
bedload 196 201
Bermuda 234 244
big swamp phase 18 46 208
bioerosion 429

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

bluff 341 350 353 359

380 383 394 431
437 439
retreat 344 357 362 365
380
boulder barricade 379 410
boulder pavement 410
boundary conditions 37 188
process 38
spatial 37
breakwater 500
Breton Island Cape 443
Bruun rule 4 73 133 250
bulkhead 479 489 491 495
498
Byron Cape 166

Calabria 464
calcification 306
California 431 433 441 445
Cambridge Fiord 411
Cape Antrim 405
Cape Breton Island 443
Cape Byron 166
Cape Tribulation 312 326 329

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

carbonate production 40 246 321 324

carbonate solution 267 272 287 313
429 454
Caribbean 330
Carpentaria Gulf of 209
catastrophe theory 14 310
cay 5 285 325
channelisation 235 240
chaos 13 33
chenier ridge 20 48 121 192
204 210
Chezzetcook Inlet 404
Clarence estuary 189 195
cliff 14 127 129 317
425
abrasion 428
anthropogenic influence 435
bevelled 443
chemical weathering 428
composite 440
drumlin 402 413
equilibrium profile 447
failure 14 344 431 434
mass movement 431 439
multi-storied 443

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

cliff (cont.)
profiles 440
processes 427
rate of retreat 432
relict 447
salt weathering 428
tectonic setting 425
two-storied 443
vertical 443
wave action 427 434
coastal lake saline 167 192 220 235
see also estuary
Cocoas (Keeling) Islands 267
collision coasts 425 466
continentality 304
continuity equation 304
Cook Islands 289
coral reef see reef
coseismicity 461
crevassing 99
cyclone see hurricane

Daly River 192

dams 23 112
delta 87 222 353 376

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

delta (cont.)
387 391 412
anthropogenic influence 112
Atchafalaya 98 110 113
bayhead 108 110
birdfoot 93 103 191
classification 103
complex 97 105 107
cycle 97
deformation 97 108
DELTA2 390
density contrasts 90
ebb tidal 11 58 124 167
174 221 229 488
facies model 103
flood tidal 11 58 167 221
229 243 404 485
fjord-head 377 386
front 88 99 101
Ganges-Brahmaputra 98 105
Gilbert-type 90
Huanghe 92 112
lacustrine 108
lagoon 229 242
lobe 97 105 107
Mackenzie 346 358

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

delta (cont.)
Mahakam 102 105
Mississippi 93 97 103 105
109
Niger 95
Nile 23 87 112
outwash 380 390 412
plain 88 99 222
processes 390
Rhone 98 104
river-dominated 88 103
São Francisco 94 96 104 125
shelf-phase 105
shoal-water 105
succession 99
switching 47 97
tide-dominated 95 103
wave-dominated 94 96 103
developed coasts 477
dissipative beach 47 402
distributary-mouth bars 92 101
dredging 489 495 499
drowned river valley 125 187 192 256
drumlin 16 373 377 399
archipelago 373
island 401

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

drumlin (cont.)
headland 402 413
dune 13 124 132 221
227 229 244 478
482 485
cliff-top 127 149 164
hummocky 485
multiple ridges 485
Pleistocene barrier 229
rebuilding 491
transgressive 149
vegetation 149 152 225 486
494
dynamic equilibrium 5 45 62 131
291 365

earthquake 14 463 466

El Niñom 39 328
equilibrium 42 44
chaotic 45
dynamic 5 45 62 131
291 365
metastable 45
potential 62
steady-state 5 45 291

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

equilibrium profile 4 44 134 350

357 447
erratics 410
esker 380
estuarine funnel 188 198 200 204
estuarine plain 18 187 218 203
estuary 18 49 138 140
145 167 187 218
235
back-barrier 138 400
barrier 49 167 189 235
Bay of Funday 167 189 235
Bay of Funday 105 210 383 444
blind 220
Clarence 189 195
drowned-valley 167 174
Gironde 210
macrotidal 18 105 187
St Lawrence 386 410
saline coastal lake 167 192 220 235
Severn 210
tide dominated 18 105 187 218
etchplanation 303
evaporite 243

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

feedback 20 33 42 139
247
negative 33 42
positive 33 46
fish kill 229
Fitzroy River 229
fjord 373 377 384 388
Flat Island 385 388 395 398
flocculation 242
floodplain 188 203 213
fluid dynamics 33 43 53
Fly River 194
foredune 132 149 152 155
Funafuti Atoll 267
funnel see estuarine funnel
funnelling coefficient 198 202

Galveston Island 253

Ganges-Brahmaputra delta 98 105
geoid changes 452
geomorphological convergence 57
Gilbert River 210
Gironde estuary 210

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

glacial diamict 374 382 394 399

glaciogenic sediment 382 386 394 397
411
glacioisostasy .see isostasy
glaciomarine sediments 237 374 382 394
Grand Cayman Island 429
Great Barrier Reef 290 303
Great Lakes 383
ground-penetrating radar 154
groyne 478 489 491 499
Gulf of Carpentaria 209
gypsum 230
gyttja 227 234 237

Halimeda 310 312 316 321

328 332
Holyrood Pond 397 399
homeostasis 20
honeycombs 429
hot spot 454
Huanghe Delta 92 112
human influence see anthropogenic
influence
hurricane 13 15 244 291
hysteresis 42

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

ice 341
anchor 357 366
block 350
bottomfast 346 354
first year 343 346
floe 348 350
frazil 348 357 366
glacier 346
gouging 349
ground 344 376
ice-keel turbate 357 365
ice-rafted deposits 410
island 346
landfast 350
pack 346 350
pile-up 350
pore 344
push 355 365 410
ride-up 350 364
scouring 349 357 364
sea ice 341 346 410
sea-ice break-up 342
sea-ice freeze up 342 346 355 366
slush 348

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

ice (cont.)
wallow 348 355 366
wedge 344
Iceland 392
incised valley 227 230 238
inheritance 21 34 38 42
54 122 124 128
168 178 198 202
212 227 245 303
319 344 448
inlet closure 229 237 240 244
artificial 488 495
isostasy 451
glacioisostasy 21 25 39 187
277 451 458 470
hydroisostasy 39 277 279 451
458 470
rebound 470
thermo-isostasy 452
volcano-isostasy 452
Itirbilung Fiord 390

jetty 478 488 499

jokulhlaup 392

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

kame 380
karren 429
karst 272 287 313 429
454
thermokarst 344 353 359 361
363
King Point 359 360 362
Kiribati 281 289
Kosi lagoon 226 238
Kugmallit Bay 358

Lagoa dos Patos brazil 242

lag gravel 162 382 405
lagoon 138 145 219 270
283 404
atoll 270 283
back-barrier 124 127 138 222.
404
barrier reef 319
chemical precipitation 243
closed 220
deficit 256
definition 220

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

lagoon (cont.)
Delaware 254
equilibrium 256
estuarine 167 192 220 235
evolution 219
facies 230 235 237
isolation 244
Kosi 226 238
Late Pleistocene 229
Maine 237
Mdoti 232
Mhlanga 232
open 220
partly-closed 220
processes 241
segmentation 227 230 241 248
Siyai 225 240 243 255
south-east African 222
surplus 256
uMgababa 231
landslides 431
large Scale Coastal Behaviour 13 34 64 123
170 174 178
Lawrencetown beach 398
Lawrencetown Lake 404

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

levce 92 99 113 204

lithospheric flexure 456
littoral drift.see longshore drift
Long beach island 486 492 495 501
Long Island 485
longshore drift 104 123 134 137
158 160 165 178
192 221 225 229
235 244 254 362
393 485
low wooded island 312
Loyalty Islands 459 461
Lucke model 222 239 256
lutocline 196

McArthur River 195

Mackenzie Delta 358 359 361
Mackenzie River 346 358
macrotidal estuary 187
Mahakam Delta 102 105
Makatea Island 456
Makawulani Lake 229
Maldives 289
mangrove 208 232 243 249
Marco Island 477

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

Markovian inheritance 34 38 42 54
marsh 222 234 237 243
248 397 485 497
499
Marshall Islunds 272 289 453
Mary River 192 195 203 212
Mdoti lagoon 232
meander 188 196 212
cuspate 192
cutoff 201
estuarine 192 199 212
inherited 202 212
sinuous 194 198 212
tract 208
Mediterranean 3 464
megaripples 162
Mhlanga lagoon 232
microatoll 283 284 326
Middleton Island 463
Midway Atoll 272
Mississippi Delta 93 97 103 105
109
model 16 33
computer 16 122 133
conceptual 16 68
deterministic 66

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

model (cont.)
facies 103 194 230 250
fold catastrophe 14
forward 35 73
inverse 35 73
Lucke 222 239 256
mathematical 68 448
Nichols 256
numerical 69
parametric morphological
behaviour 70
process-response 103 390
simulation 70 123 133
Monterey Bay 434
moraine 373 380 406
morphodynamics 3 8 10 33
86 124 168 196
207
Moruya Beach 63 153 156
motu 5 285
Mpungwini Lake 229
Mtwalume lagoon 232
Muroto peninsula 461
Mururoa Atoll 272 453
Myall I.akes 150 157
Myrmidon Reef 307

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

Natal 222 238 248 254

257
New Jersey coast 480 504
New South Wales 11 50 53 58
64 121 189 235
239 252 254
Newfoundland 163 385 394 414
443
Niger Dlta 95
Nile Delta 23 87 112
non-linear behaviour 13 18 33 43
52
nonstationarity 43 48 57
Nova Scotia 16 18 48 137
385 399 414
nutrients 306 327

Ocean City 480 486

Ord River 18 209
outwash 374 377 382 406
plain 377 380 392 413
overthrusting plate 461
overwash .see washover

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

overstepping 17 48 133 244

250

palaeochannel 188 201 209

paradigm shift 1
paraglacial coasts 373 424
definition 373
parasequence 107
patterned ground 353
permafrost 341 376
pingo 353
Pingok Island 354
plate boundaries 459
plate tectonics 451
point bar 199 201 209
polar coasts 341
Pompey Complex 312
prodelta 88 99 102 387

Raine Island 322

ravinernent surface 111 133 137 162
362 365
reaction time 9
recurrence interval 64 293

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

reef 267 303

algal-dominated 310
anthropogenic influence 320 328
atoll see atoll
barrier reel 268 303 452
catch-up 277
continental shell 303
fringing reef 268 304 313 314
319 322 326 330
452
give-up 277 330
Great barrier Reef 290 303
keep-up 277 324 330
Last interglacial 273
linear 312
mid-shelf 314
mining 328
Myrmidon 307
platform 312
remnant 313
ribbon 312 317 324
shelf 304
reef flat 312 319 325
recf islands 5 283 285 325
reflective beach 47 355 359
regolith stripping 303 324

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

rclaxation time 58 64 293 374

responsc time 58 65
return interval .see recurrence
interval
revetment 491
Rhodes Island 466
Rhone Delta 98 104
ria 187
Richards Islands 347 362
rock fall 431
rubble rampart 5 292
Ryukyus Islands 464

sabkha 220
St George’s Bay 385 394 397 399
St I.awrence Estuary 386 410
St Lawrence Gulf of 383 395
St Lucia I.ake 228 238 243
Suguenay fjord 386
salinisation 204 212
salinity stratification 229
salt wedge 93 101 196
San Andreas Fault 468
sandur 386 392
São Francisco Delta 94 96 104 125

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

scroll bar 199 201

scroll plain 192 201
sea level 1 8 16 39
47 59 63 70
105 124 167 176
187 208 211 240
249 250 268 273
287 294 304 310
316 332 377 380
395 399 406 435
442 451 458 460
469 484 501
sea stacks 426
seawall 477 491 500
sediment budget 11 41 123 134
137 169 357 362
492
sediment supply 22 33 127 128
145 147 194 238
240 246
seismic profiling 71 107 122 141
154 227 229 274
318 357 361 392
396 405
self organisation 49
self regulation see feedback

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

Sepik River 193 195

sequence stratigraphy 67 107 211
Serapis Temple of 468
Severn Estuary 210
shore platform 425 445
relict 447
shore protection project 489 496 500
Sibaya Lake 245
Siberia 341
sidescan sonar 122 349 364
Siyai lagoon 225 240 243 255
Skeidararsandur 392
Society Islands 267
South Alligator River 18 20 46 187
spit 355 359 361 394
412
barrier 237 480
barrier island 158
cuspate 227 241
drift-aligned 378 395 402 412
growth 221 243
headland 128 158
paired 241
platform 395
recurved 140 155
swash-aligned 378 395 402 412

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

spit (cont.)
trailing 402 411
stamukhi shoal 355
stationarity 57 65
step-response function 62
Stephenville 385 395 399
storm 5 6 9 10
121 131 138 158
160 162 170 239
244 286 290 352
357 366 412 479
484 495
storm block 286
storm surge 15 16 131 352
358 364 366
Story head 403 405 411
strandplain 104
subduction zone 459
subsidence 268 271 452 470
atoll 268 271 452
beneath ice sheet 458
coseismic 463
interseismic 463
oceanic islands 268 271 453
permafrost 344
postseismic 463

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

subsidence (cont.)
thermal 178

tafoni 429
Taiwan 461
talik 344
tectonic shorelines 39 173 451
thaw flow-slides 344 362
thermokarst 344 353 356 359
361
threshold 14 18 33 42
46 51 304 310
329
intransitive 9
tidal asymmetry 196 198
tidal channels 208 234 248 312
489
tidal currents 234 240
tidal flats 229 234 237 242
248
tidal prism 198 204 206 213
234 240
till 382 400 405 410
time scale 5 35 58 64
124
reaction time 9

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

time scale (cont.)

relaxation time 58 64 293 374
response time 58 64 65
lag 61
tombolo 237
Top End Northern Territory 18 20 188 209
Torres Strait 306
transform zones 468
transgressive barriers 127 146 165 168
250
tsunami 15 16
Tuamotu Atolls 272 282 289 454
Tugela River 224 230 254 257
Tuktoyaktuk Peninsula 349 350 358 360
362 364
Tuncurry 141 152 161 174
tundra 342 354 361
turbidite 227 382
turbidity 306
turbidity cuurrents 92 176 389

uMgababa lagoon 231

underthrusting plate 459
uplift 454
coseismic 463
postseismic 463

This page has been reformatted by Knovel to provide easier

navigation.
Index Terms Links

uplifted marine terraces 447 463 468

van Diemen Gulf 18 187 218

washover 124 131 158 231

240 243 252 361
397 399 411 485
495
Warnbro Sound 235 244
water hammer 427
water layer levelling 429 435
wave-dominated coasts 121
wave quality 316
wave quarrying 427
West Marsh 404
Wolfe Spit 360

Yonge Reef 322

Yukon 345 348

zonality 303 342 357 362

365
Zululand 222 238 248 252

This page has been reformatted by Knovel to provide easier

navigation.