978 0883850305

Benthic suspension feeders and flow
Benthic suspension feeders

and flow
DAVID WILDISH
Biological Station, St. Andrews, Canada
DAVID KRISTMANSON
University of New Brunswick
CAMBRIDGE
UNIVERSITY PRESS
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo
Cambridge University Press
The Edinburgh Building, Cambridge CB2 2RU, UK
Published in the United States of America by Cambridge University Press, New York
www. Cambridge. org

Information on this title: www.cambridge.org/9780521445238
Cambridge University Press 1997
This publication 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 1997
This digitally printed first paperback version 2005
A catalogue recordfor this publication is available from the British Library
Library of Congress Cataloguing in Publication data

Wildish, David, 1939-
Benthic suspension feaders and flow / David Wildish, David
Kristmanson.
p. cm.
Includes bibliographical references and index.
ISBN0-521-44523-X
1. Benthos. 2. Marine biology. 3. Hydrodynamics.
I. Kristmanson, D. D. II. Title
QH91.8.B4W55 1997
577.7'4-dc21 97-17118
CIP
ISBN-13 978-0-521-44523-8 hardback
ISBN-10 0-521-44523-X hardback
ISBN-13 978-0-521 -02347-4 paperback

ISBN-10 0-521-02347-5 paperback
for Ann and Bernice
Contents
Preface page ix
Acknowledgments xi
Introduction 1
Asking the right question 3
Scientific methods 7
Scientific communication 9
Aims and structure 12
References 15
Methods of study 17
Flow simulation 18
Velocity measurement 30
Seston sampling 38
Specialized methods 43
References 60
Dispersal and settlement 69
Benthic suspension feeder life cycle 69
Dispersal 73
Larval dispersal 74
Post-larval dispersal 91
Local hydrodynamics 93
Larval settlement/recruitment 95
Summary 117
References 120
Flow and the physiology of filtration 128
Background 128
Passive suspension feeders 132
Active suspension feeders 138
vn
Vlll Contents
Facultative active suspension feeders 155

Combined passive-active suspension feeders 158
Deposit-suspension feeders 161
Summary 163
References 165
Mechanisms of seston capture 170
Nature of seston 170
Fluxes, encounter rates, and collection efficiency 172
Mechanics of capture 174
Seston capture mechanics 200
Summary 207
References 209
Behavioral responses to flow 213
Optimizing growth by orientation to flow 214
Avoiding flow dislodgement 238
Epifauna and aggregation 247
Swimming 254
Summary 260
References 263
Benthic populations and flow 270
Benthic limitation by flow 270
Water movement impoverishment 299
Biological interactions in mixed benthic communities 301
Summary 304
References 306
Ecosystems and flow 313
Background 313
Marine plant canopies: kelp forests 315
Marine plant canopies: seagrass meadows 326
Bivalve reefs 341
Summary 353
References 355
Future directions 362
Selecting pertinent scientific questions 362
Proximate fundamental questions 367
Proximate, applied benthic biology questions 374
Ultimate questions 378
References 387
Glossary 391
Index 405
Preface
Hydrodynamics and benthic biology represent a very broad field encom-

passing many possible levels of hierarchical organization and subjects.
Thus, it is not possible for any author(s) to have a detailed personal
knowledge and experience of all the subjects discussed herein. Nor does
the book seek to cover every subject which could be included; for exam-
ple, there is little on corals or freshwater biology, although we have
occasionally mentioned such examples to illuminate the points we are
trying to make. The subject matter of each chapter is also limited to
matters where hydrodynamics are centrally involved. Hence, in Chapter
3, for example, larval biology is limited to dispersal and settlement where
flow is clearly important, and our overview is not intended to replace the
more complete reviews of this subject which are available.
It became clear to us early in the genesis of this book that our topic, at
most hierarchical levels of benthic biology, was at a very early stage in
the learning curve. Indeed, one of the reviewers suggested that we wait
for a few years to begin the book in the hope that by then the subject
would have reached a firmer theoretical footing. As the evidence in front
of you will suggest, we did not take this advice because of our own
interests and agenda. It is graduate students who are in the best position
to carry on the large amount of work still to be done in the field we have
outlined here. Consequently, our book has been shaped specifically to-
wards their needs.
Our collaboration began in 1970 and became interesting soon after-
wards during joint studies involving pollution of a marine tidal inlet in
the Bay of Fundy. We were both struck by the powerful control that the
tidal flow of the Bay appeared to have in shaping the benthic populations
which were present there. We realized that we had the means of an
interdisciplinary collaboration at our disposal: D. D. Kristmanson's
IX
x Preface
training in chemical engineering and special interest in how flows allow

molecules to come together and react, and D. J. Wildish's training in
marine biology and special interest in benthic animals, inclusive of sus-
pension feeders. Our modus operandi evolved into one in which new
ideas were tested against accepted hydrodynamic theory, before design-
ing specific null and alternate hypotheses for biological experimental
testing.
Until recently, most marine biologists were not required to study fluid
mechanics during undergraduate training. We believe that such training,
as first introduced by Steven Vogel in his book Life in Moving Fluids:
The Physical Biology of Flow, will become mandatory during the train-
ing of undergraduates in marine and freshwater biology.
This book was written in St. Andrews at the Biological Station and in
Fredericton at the University of New Brunswick, between 1992 and 1995.
So that we could interact better, we have also spent shorter periods
together at the Huntsman Marine Science Centre in St. Andrews and
in a small office, without a telephone, in Vancouver. At both of the
latter locations, we were able to visit a public aquarium and saw
some familiar benthic animals. At the Aquarium-Museum of the
Huntsman Marine Science Centre in St. Andrews, we saw the solitary
lampshell Terebratulina septentrionalis and the giant scallop Placopecten
magellanicus, which looks much more attractive in the living state than as
a shell used as an ashtray. At the Vancouver Aquarium in Stanley Park,
we saw the beautiful sea pen Ptilosarcus qurneyi and the polychaete
worm Eudostylia vancouveri. We suggest to readers that, unless they are
accomplished SCUBA divers, this is the best way to see some of the
animals which are the subject of this book.
Acknowledgments
Many people, in various ways, helped in the completion of this book.

Without naming all, we thank them for their efforts on our behalf.
The following research scientists provided frank and detailed critiques
of one or two chapters: Edwin Bourget, Cheryl-Ann Butman, Jim
Eckman, Craig Emerson, Marcel Frechette, Barry Hargrave, C. Barker
J0rgensen, Beth Okamura, and J. Evan Ward. All of this group (part
of the Northwest Atlantic invisible college of the benthos) have, with
the exception of Professor J0rgensen, attended at least one of the
Biennial Benthic Workshops held in St. Andrews in October. These
meetings began in 1983 to help celebrate the St. Andrews Biological
Station's 75th Anniversary. After missing a year in 1987, the Biennial
Benthic Workshops are now even-year events, with the next one planned
for 1998. The larger invisible college, probably in excess of 200 persons
over the years, has been an important source of information exchange,
inspiration, criticism, and friendship. We thank all of our mud-grubbing
colleagues who have helped to make these meetings so friendly and
stimulating to us.
This book could not have been completed without the excellent serv-
ices available at the St. Andrews Biological Station library and the
dedicated help of the library staff, Marilynn Rudi and Joanne Cleghorn.
Brenda Best, also of the St. Andrews Biological Station, was able to
translate our scribbles or unformatted chapters on disk into readable
English efficiently, and Susan Laurie-Borque, of Ottawa, created new or
redrafted published art material into the excellentfiguresincluded in this
book.
The forbearance of the Department of Chemical Engineering of the
University of New Brunswick allowed David Kristmanson to pursue the
research interests included herein, and the Department of Fisheries and
XI
xii Acknowledgments
Oceans, in the persons of Mike Sinclair and Wendy Watson-Wright,

encouraged Dave Wildish in this endeavor.
We thank the following for permission:
Prof. Dr. Klaus Liming, Biologische Anstalt Helgoland
Springer-Verlag GmbH & Co. KG for material from Marine Biology,
vol. 81, p. 259; vol. 110, pp. 94, 101; vol. 97, p. 119; vol. 119, p. 526; vol. 85,
p. 128; vol. 72, p. 128; vol. 105, p. 120; vol. 92, p. 324; from Oecologia
(Biol.), vol. 8, p. 26; and from Bivalve Filter Feeders (1993) by R. F. Dame
(ed.)
The Scottish Association for Marine Science
Dr. Cheryl Ann Butman, Woods Hole Oceanographic Institution
D. O. Duggins
Gary L. Taghon
S. W. Nixon
Richard Emlet, Oregon Institute of Marine Biology, University of Oregon
Charles H. Peterson
Steven Vogel, Department of Zoology, Duke University
Journal of Phycology
Walter de Gruyter & Co.
Annual Reviews Inc. for Figure 5, modified, with permission, from the An-
nual Review of Fluid Mechanics, Volume 9, 1977, by Annual Reviews Inc.
The American Meteorological Society for illustrative material from Jackson,
George A., 1984: Internal wave attenuation by coastal kelp stands. J. Phys.
Oceanogr., 14,1300-1306
Estuaries for illustrative material from Hidu, H., and H. H. Haskin. 1978.
Swimming speeds of oyster larvae Crassostrea virginica in different salinities
and temperatures. Estuaries 1:252-255
NRC Research Press
Material from Elsevier Science: Reprinted from J. Exp. Mar. Biol. Ecol., vol.
65, p. 121; vol. 83, p. 241; vol. 148, p. 217; vol. 174, p. 68; vol. 181, p. 12; vol.
114, p. 2; vol. 173, pp. 64-5; vol. 155, p. 163; vol. 165, p. 253; vol. 174, p. 76; vol.
113; p. 212; vol. 164, p. 149 with kind permission from Elsevier Science - NL,
Sara Burgerhartstraat 25, 1055, KV Amsterdam, The Netherlands
Academic Press Ltd, London, England, for material from The Biology of
Estuaries and Coastal Waters (1974) by E. J. Perkins
Material from J. Fish. Res. Board. Can., vol. 24, pp. 328,332; and Can. J. Fish.
Aquat. Sci., vol. 40 (Supplement 1), pp. 302, 307, reproduced with permission
of the Minister of Supply and Services Canada 1996
Blackwell Science Ltd., Osney Mead, Oxford OX2 0EL England for material
from J. Anion. Ecol., vol. 22, p. 337
Inter-Research, Nordbiinte 23, D-21385 Oldendorf/Luhe, Germany, for ma-
terial from Mar. Ecol. Prog. Ser. vol. 61, p. 165; vol. 42, p. 164; vol. 28, p. 190;
vol. 92, p. 143; vol. 63, p. 213; vol. 34, p. 70
The Bulletin of Marine Science for material from Bull. Mar. Sci., vol. 39, pp.
290-322
Material from Bioscience, vol. 28, p. 639, and vol. 29, p. 349, 1978 American
Institute of Biology and Sciences
Brian Helmuth
H. G. Verhagen (personal communication), Agricultural University
Wageningen, The Netherlands
Acknowledgments xiii
Material modified with permission from D. O. Duggins, C. A. Simenstad, &

J. A. Estes. Magnification of secondary production by kelp detritus in coastal
ecosystems. Science, vol. 245, p. 171. Copyright 1989 American Association
for the Advancement of Science.
Material modified with permission from G. L. Taghon, A.R.M. Nowell, & P.
A. Jumars. Introduction of suspension feeding in spionid polychaetes by high
particulate fluxes. Science, vol. 210, pp. 562-A. Copyright 1980 American
Association for the Advancement of Science.
Material modified with permission from S. Vogel & W. L. Bretz. Interfacial
organisms: passive ventilation in the velocity gradients near surfaces. Science,
vol. 175, pp. 210-11. Copyright 1972 American Association for the Advance-
ment of Science.
Let you who enter here
Cast off all cant
Take on the eyes
Of the wondering child.
Be blessed with fields
Richly carpeted with
Insight and imagination.
Stare in wonder
At this world of beauty
Of which you are a part.
1
Introduction
One objective of benthic ecology is to describe the spatial distribution of

living organisms at or near the sediment-water interface, as well as to
explain how and why the distribution occurs at that particular location on
the seabed. The descriptive part of the job of the benthic ecologist thus
becomes a sampling challenge since a depth gradient is involved from the
intertidal to the hadal zone - a depth range of 0 to >6000m (Parsons et
al. 1977). Another sampling problem is that of substrate variability, thus
hard, e.g. rocks and corals, versus soft substrates, e.g. muds, sands, and
gravel, which means that appropriate samplers for each type must be
quite different. Because it is considerably more difficult to sample at
depth, the bulk of the available sampling results are for the littoral and
sublittoral zones down to SCUBA diving depth. Thus, much of our
knowledge of the spatial distribution of benthic organisms is heavily
biased to these limited depths in the nearshore environment. The benthic
animals which we would expect to be present in a grab or core sample of
soft sediments in the nearshore region would include micro-, meio-, and
macrofauna (Table 1.1). The size ranges shown are arbitrary and based
on sampling convenience; thus, if a 0.5- or 0.8-mm mesh was used in
sieving, then the lower size limit for macrofauna becomes 500 or 800 Jim.
Because of a need to limit the size of this book, we have had to limit its
subject matter to marine or estuarine benthic macrofauna and exclude
micro- and meiofauna from detailed consideration. This is simply for
presentation convenience and it is obvious to us thatflowwill have many
important direct influences on micro- and meiofauna also. Thus, if barna-
cle cyprids can distinguish between microbial films developed on hard
substrates in either low or high shear stress flows (see the section, Hard
substrates, Chap. 3), then flow itself must be an important factor in
shaping the microbial association which develops there.
1
Benthic Suspension Feeders and Flow
Table 1.1. Arbitrary size divisions of benthic fauna.
Size range
Name (urn) Examples
Microfauna 1-100 Some smaller Protozoa

Meiofauna 100-1000 Some Protozoa, e.g. Foraminifera; nematodes;
harpacticoid copepods; gastotrichs; isopods;
turbullarians; small macrofauna
Macrofauna >1000 Amphipods, mysids, echinoderms, sponges,
brachiopods, corals, hemichordates,
echinoderms, molluscs, polychaetes
Megafauna Cannot be adequately sampled with grab or
corer, e.g. the mud star Ctenodiscus
crispatus or benthic fish
Source: Based on Parsons et al. (1977).
The reader should be aware that there is a parallel interest among

freshwater biologists (e.g. Hildrew and Giller 1994) in the effect of flow
on river or stream benthos. For the same reasons of space limitation, we
have not considered this work, except incidentally where research de-
scribed presents a particularly good example of a pertinent interdiscipli-
nary study.
This book is concerned only with those macrofauna characterized
as suspension feeders that live on either soft or hard substrates of the
seabed and feed in a particular way, that is by filtering microscopic
particulates, or seston, transported to them by ambient flows. Our focus
is the interactions which occur between individuals, populations of sus-
pension feeders, or communities and ecosystems containing suspension
feeders and the various types of water movement characterized as flow in
the last paragraph of this section.
Functionally, benthic macrofauna can be classified as suspension feed-
ers, deposit feeders, carnivores, omnivores, and algal scrapers (Chap. 7,
the section, Benthic limitation by flow). The two major trophic types of
macrofauna can be distinguished by their mode of feeding - deposit
feeders are limited to soft sediments and ingest sediment which has
already been deposited after seawater transport, while suspension feed-
ers feed by capturing seston, a mix of microscopic particles which may
contain detritus, bacteria, microalgae, small animals, and sediments, sus-
Introduction 3
pended in and transported by seawater flows. If the seston is carried by

flow directly to the capture surface, the suspension feeder is described as
passive. If, on the other hand, seawater containing seston is pumped by
a ciliary or muscular pump to the capture surface, the suspension feeder
is described as active. Further classification of passive and active suspen-
sion feeders is considered in Chapter 4 (see the section, Background).
Suspension feeders as a group comprise a diverse assortment of taxa
drawn from most of the major invertebrate phyla inclusive of corals,
hydrozoans, bryozoans, brachiopods, some polychaetes and bivalve mol-
luscs, and a few echinoderms and crustaceans. Taxa which are found on
hard substrates, e.g. barnacles, are generally epifaunal, living firmly at-
tached to the rock or coral reef surface, and are said to be sessile. A few
specialized species are able to burrow into some hard substrates, e.g. the
rock-boring bivalve Petricola pholadiformis, which is also a suspension
feeder. Soft sediment macrofaunal taxa may be either epifaunal and
sessile, e.g. tube-living polychaete worms such as many spionids, or
infaunal, that is to say, able to live and burrow within the sediment
matrix, e.g. free-living polychaete burrowers such as Chaetopterus
variopedatus. Again, both taxa are suspension feeders.
Flow is the general term that we have used for all the types of water
movement which can occur in the marine or estuarine environment.
Seawater movements include those caused by the action of winds di-
rectly on the sea surface {wind-wave effects), by the variable gravita-
tional attraction of both moon and sun on seawater {tidal currents), or by
density differences between different water masses due either to in-
equalities of heat exchange across the air-seawater interface or to differ-
ences in salinity (Tait 1968).
Asking the right question

Two general types of question apply in biology: proximate and ultimate
(Alcock 1989). The proximate type equates to the "how does it work"
questions of Table 1.2 and requires an answer based on a reductionist
experimental program of inquiry. By contrast, the ultimate type is a
"why" question involving knowledge of evolutionary mechanisms
and history of evolutionary trends, often over significant geological
periods.
Typical ultimate questions listed in Table 1.2 clearly demonstrate that
most cannot be answered by reductionist experimental methods. Thus,
Table 1.2. Some general questions related to the disciplines listed and
referable to suspension feeders or the communities in which they live.
Discipline No. Proximate Ultimate

Physiology 1 How do suspension Why did active suspension
feeders filter feed? feeders evolve from passive
ones (or vice versa)?
How can energy costs of Why did a ciliary pump evolve?
ciliary pumping be
estimated?
Ethology 3 How do competent larvae Why have some sessile

choose substrates suitable suspension feeders evolved a
for initial colonization? short, and others a long, period
of planktonic life?
How do epifaunal Why do hydrodynamic factors
suspension feeders cause epifaunal suspension
become aggregated? feeders to aggregate?
How do some bivalve What adaptive imperatives
molluscs swim? resulted in swimming bivalves?
Ecology 6 How does larval How do larvae with specific

retention in an estuary responses to environmental
occur? variables, e.g. light, gravity, and
flow, evolve?
How do mixed benthic What causes a mixed benthic
assemblages dominated assemblage dominated by
by suspension feeders suspension feeders to evolve?
occur?
How are materials cycled Why do ecosystems evolve
between pelagic and over geological time?
benthic parts of an
ecosystem?
for the first physiological ultimate question of Table 1.2, we have no

direct knowledge of the ancient environments which may have resulted
in the development of an active suspension feeder, nor any idea of the
evolution of its ciliary pump since they are soft structures which are not
preserved very well in the geological record. Although circumstantial
evidence from the geological record may help, e.g. general climatic con-
ditions deduced from stratigraphic deposition or from the morphology
and physiology of extant ciliary pumps from "primitive" to "advanced"
Introduction 5
forms, such deductive methods often do not provide an adequate answer

for the ultimate questions. For this reason, we have limited the questions
addressed in Chapters 3 to 8 to those that are clearly proximate in nature.
In Chapter 9, where possible future benthic biological questions are
posed, we include consideration of ultimate ones.
As pointed out by Fenchel (1987), biology as a scientific discipline is
organized hierarchically. At the most complex organizational level is the
ecosystem, by which is meant a community of living organisms, together
with the physical and chemical environmental factors present interacting
together to form a recognizable ecological unit, e.g. an estuary. Descend-
ing the list of hierarchical order, we thus arrive at successively lower
levels - ecosystem, community, population, individual, organ, cells,
organelles, and molecules (Fenchel 1987). As pointed out by Fenchel,
these levels of organization correspond to the specialized subdiscip-
lines of biology: ecology, inclusive of ecosystems, communities, or
populations; ethology, inclusive of social groups and individuals; physiol-
ogy, inclusive of individuals and organs; cell biology, inclusive of cells
and organelles; and biochemistry, inclusive of organic molecules. From
this list we have selected the three higher hierarchical levels and thus
subdisciplines of biology - ecology, ethology, and physiology - to inves-
tigate benthic interactions with flow.
Where the focus of concern is the individual animal as in ethology and
physiology, it is easy to appreciate how a reductionist research program
could be developed into an experimental study of any suspension feeder.
In ecology, where the unit of study ranges from a population to an
ecosystem, that is to say, a mussel reef to the whole estuary in which the
mussels live, it is much more difficult to see which questions are appro-
priate. Thus, it becomes necessary to determine exactly what are valid
ecological questions. Fenchel (1987) defines ecology as "the study of the
principles which govern temporal and spatial patterns of assemblages of
organisms." This succinct wording is a good encapsulation but does not
quite convey the depth of the subject matter, and a more complete
snapshot of ecology, at least in the mid-1980s, can be gleaned from an
opinion survey of 645 members of the British Ecological Society
(Cherrett 1989). This group was composed of university trained ecolo-
gists with mixed backgrounds and employment, who suggested 236 dif-
ferent concepts as fundamentally important in ecology. Considerable
overlap in the concepts existed and, using the data provided by Cherrett
(1989), we suggest that they can be reduced to the seven shown in
Table 1.3.
6 Benthic Suspension Feeders and Flow
Table 1.3. Important concepts in ecology derived from a survey

conducted by the British Ecological Society.
No. Description Examples

1 Spatial Studies involving the ecosystem, community, and
niche
2 Temporal Studies involving succession, climax, and diversity
3 Energy flow Energy flow, materials cycling, and food webs
process study
4 Biological Competition, density dependence, and predator-prey
interactions studies
5 Limiting factors Physical limitations in ecology at various hierarchical
levels
6 Evolution Life history strategies, adaptations, and co-evolution
7 Ecosystem Conservation, ecosystem fragility, and maximum
management sustainable yield
Source: Cherrett (1989).
In this book we have excluded the long-term, temporal perspective in

line with our resolve to exclude ultimate questions (number 6, Table 1.3),
and because succession, climax, and diversity concepts (number 2, Table
1.3) have not been studied in relation to flow. This means that our
attention is given entirely to numbers 1, 3, 4, and 5, but omitting 7, in
Table 1.3. This is not as ruthless as it may sound since there is only a small
amount of work published on evolutionary aspects of benthic biology. In
any case, we return briefly to pertinent questions concerning numbers 6
and 7 (Table 1.3) in Chapter 9.
Benthic biology is a relatively young scientific discipline. Its quantita-
tive study dates from the work of the Danish investigator Petersen
(1911), who studied those factors which regulate benthic macrofaunal
density and biomass in order to predict the productivity of benthic feed-
ing and commercially important groundfish. It is only since the early
1970s that the focus of benthic biological research has moved from
observation to inferential experimentation, with a more rigorous effort
at field or laboratory experimentation. This may help explain why
benthic biology has so few widely applicable predictive models to offer
the newcomer to this field.
Introduction
Table 1.4. Major types of scientific method.
Type Description
Observation Simple: involves direct experience, e.g. looking down a
microscope. Includes descriptive taxonomy, natural
history observations.
Complex: sampling or measurement which may involve
bias, e.g. benthic grab sampling involving different sieve
mesh sizes in sorting the animals.
Theory/model Reductionist: a way of looking at a subfield of benthic
biology which can be represented verbally or by a simple
conceptual model, e.g. the benthic limitation by flow
theory (see Chap. 8).
Holistic: as above, except that because of complexity or
the nature of the model, formal mathematical
representation is required, e.g. an ecosystem simulation
of a bivalve reef based on energy flow.
Inferential Laboratory or field experimental tests of formulated null,
experimentation Ho, and alternate, H1? hypotheses.
Criticism Logical refutation of any conclusions or constructs
derived from any of the types of scientific method
described here.
Scientific methods
Science is concerned with the creation and communication of knowl-
edge. Concerning the first of these, scientific knowledge is created by one
of the four major methods shown in Table 1.4.
Observation, in both the field and laboratory, is fundamentally impor-
tant to the scientific enterprise. During this process, realistic questions
and, hence, theories, models, or hypotheses are formulated. Such formu-
lations are possible constructs determined by logical reasoning concern-
ing the relations of benthic animals to the observed external world.
As pointed out by Peters (1991), there is no consensus on how words
like theory, model, and hypothesis are used, and he recommends using
them synonymously. In our account, we define hypothesis as a working
explanation of observations proposed in such a way that an experimental
test can choose between two contrasting constructs - null and alternate.
The hypothesis may apply fully, or only partially, to the theory or model
constructs which are conceived to be at a higher hierarchical level. The
Librarian Research Scientist

Information retrieval Literature review
Bibliographic control Fieldwork observation
Information collection Conception
Information storage Formalization: Ho, H1
Client assistance/training Experimental test
MS preparation
|lnternal review|-- Reject

Pub isher
Science publishing External review - Reject
Figure 1.1 Outline of the process used in science to conceive, experimentally

test, disseminate, and store information.
alternative and null hypotheses are tested by inferential experimentation

with the desired outcome that one or other of the constructs can be
discarded as a result of the test. As pointed out by Sprintall (1990), care
must be taken to distinguish between post facto and experimental re-
search. The former concerns field studies where independent and de-
pendent variables are not assigned before the work begins and, hence,
causation or rejection cannot be established. In experimental research,
the independent and dependent variables are established before the
work begins and, if the results warrant, can lead logically to the rejection
or establishment of cause. If a theory or model can be represented by a
simple conceptual model, we refer to it as a reductionist model, in
contrast to a complex model, which requires mathematical formalism to
express it and which is referred to as an holistic model.
Criticism, either from one's self, using accumulated experience, or
from others and formally expressed as in a written review, informally as
during a workshop, and as part of the peer review process before primary
publication (Fig. 1.1), is important to the well-being of the scientific
knowledge database. It is during the critical process that judgements of
hypotheses are made.
An outline of the process used to conceive, test, and store scientific
information is shown in Fig. 1.1. Peters (1991) has characterized
hypothetico-deductive science as an alternation between creation and
criticism. For the research scientist, it is in the private or synthetic phase
of the process shown in Fig. 1.1 (first three items in the box) where the
Introduction 9
individual creates a new theory and/or new hypothesis referable to an

already established theory. Exactly what is involved in the synthetic
phase of science has proved difficult to define (Peters 1991). The creative
process involves keen observation and intuition, an ability to juxtapose
two apparently unrelated ideas, or use of accepted theories from related
disciplines. It is the latter process which is the main approach used in this
book with benthic biology as the recipient of theory from the well estab-
lished discipline of hydrodynamics.
It is frequently necessary to use multiple hypotheses to solve a scien-
tific question, e.g. in determining the precise mechanism of filtration in
suspension feeders (Chap. 5), and this approach is championed by Platt
(1964) and Chamberlin (1965). Quinn and Dunham (1983) drew atten-
tion to the difficulty of applying Baconian experimental tests to pairs of
mutually exclusive hypotheses in many ecological questions where mul-
tiple alternative hypotheses represent the usual situation. As an exam-
ple, we site the growth or production of a suspension feeder population.
Here, the null hypothesis might be that flow does not affect suspension
feeder growth, but is swamped by multiple alternative environmental
factors, e.g. seston concentration, seston quality, temperature, or
growth-limiting flow at high velocity. According to Simberloff (1983),
multiple alternative hypotheses should be met by increased ingenuity
from researchers in framing unambiguous hypotheses capable of dealing
with multiple alternatives and the potential interactions between them.
Scientific communication
Concerning the communication of science results mentioned earlier, the
active participants in science tend to form into natural groups which are
referred to as an "invisible college" because of shared research interests
and problems. Communication is the most important function of the
invisible college and various ways have been tried to optimize it. These
include personal communication (often electronic), workshops, sympo-
sia, published periodicals, and books.
In regard to periodicals, there are four distinct types: science maga-
zines, which provide interpretive articles for a multidisciplinary audi-
ence, but not original research articles (e.g. New Scientist, Scientific
American); technology transfer periodicals, whose purpose is to transfer
science to a commercially active field such as aquaculture (e.g. World
Aquaculture) but which does not usually carry original research; a few
multidisciplinary journals, notably Science and Nature, which contain
original abbreviated research articles thought to be of wide general

interest and from all fields of science, including physics, chemistry, biol-
ogy, geology and medicine; and single discipline or primary journals,
which contain original research articles from a specific area of research
or subdiscipline and therefore are targeted to a specialized audience. As
a result of recent advances in electronic publishing and the widespread
use by research scientists of personal computers, the future possibility of
online journals is real (Maddox 1992). Nevertheless, in 1997, the printed
word remains the pre-eminent means of communication for the aquatic
sciences.
The creation and communication of scientific knowledge in printed
multidisciplinary or primary journals involve three different types of
professional (Fig. 1.1). The research scientist creates scientific knowl-
edge and describes the work in research articles or evaluates the work of
other research scientists in review articles. The publisher prepares copy
for the journal consisting of a series of research articles and is responsible
for preparing a sufficient number of copies for wide dissemination
among the extended invisible college. Wide dissemination of printed
copy is less important today because of the availability of online elec-
tronic abstracting services with which, and by using key words, relevant
articles can be obtained. The librarian assists the research scientist by
facilitating access to published sources by collecting, storing, and retriev-
ing articles involving bibliographic control inclusive of indexing, ab-
stracting, and classification.
Newcomers to the interdisciplinary field of hydrodynamics/benthic
biology may need assistance in finding reliable literature sources which
explain the physics of flow. Some of the sources which we have found
helpful may be of interest. There are two excellent introductions to
hydrodynamics for biologists: Vogel (1981), Life in Moving Fluids: The
Physical Biology of Flow, recently revised and expanded in a second
edition (Vogel 1994); and Denny (1993), Air and Water: The Biology and
Physics of Life's Media. Both provide an introduction to the fundamen-
tal principles of hydrodynamics and many applications of interest in
benthic biology. Denny (1988) has also provided a similar introduction
and also an account of flow phenomena on the wave-swept shore. These
books interpret the solid body of accepted hydrodynamic theory devel-
oped by mathematicians, physicists, and engineers over a period of more
than 150 years. For the more mathematically adept, there are many
introductory texts in fluid dynamics which might be consulted. We have
found Daugherty et al. (1985), Fluid Mechanics with Engineering Appli-
Introduction 11
cations, to contain clear descriptions of most of the ideas of interest to

benthic biologists. When fluid flows past a solid boundary, it is arrested
at the boundary and a sheared zone develops. Thisflow,which has been
studied in great detail by aerodynamicists, is called a "boundary layer."
In benthic biology, boundary layers influence forces on submerged ob-
jects and the flow around them. They also influence the processes by
which nutrients and other dissolved molecules are transferred between
biological surfaces and the ambientflow.The standard reference for such
flows is Schlichting (1979), Boundary-Layer Theory. Boundary layers
over smooth or uneven marine sediments are described by an extension
of engineering boundary layer theory. A lucid introduction to this sub-
ject is given in Mann and Lazier (1991). Since the terminology will be
unfamiliar to many biologists, we have included a Glossary beginning
after Chapter 9. It contains brief definitions of those words which we
consider to be the most important in the benthic interdisciplinary field
considered in this book. The Glossary includes terms used in hydrody-
namics, benthic biology, and surficial geology.
With regard to the biological aspects of the interdisciplinary field of
hydrodynamics/benthic biology, J0rgensen (1966), Biology of Suspen-
sion Feeding, is concerned with the wide taxonomic range of inverte-
brates which are suspension feeders. This book pioneered the physi-
ological and ethological study of feeding by suspension feeders. A later
book by J0rgensen (1990) presents selected updates on suspension feed-
ing physiology and ecology. This includes the Danish school's elegant
description of the bivalve gill as a ciliary pump and a novel hypothesis
regarding the mechanism of seston capture on the bivalve gill surface. In
addition, we have found the biological dictionary of Lincoln, Boxshall,
and Clark (1982) to be useful and it can be used profitably when reading
this book.
Among the plethora of aquatic, multidisciplinary, and primary jour-
nals, we have selected those which are most frequently used to carry
original research articles in the English language on benthic suspension
feeders and flow. The results of an analysis of the references used in
Chapters 2 through 8 are shown in Table 1.5. It is of interest that the first
four journals in rank order carry nearly 40% of all 799 cited articles.
Some journals, e.g. Limnology and Oceanography and the Canadian
Journal of Fisheries and Aquatic Sciences, clearly stick to ecological
rather than physiological research articles. Table 1.5 also demonstrates
another point: that the relevant research articles are spread over a wide
range of journals, as is evidenced by the "Other journals" entry. The
Table 1.5. Most cited journals in Chapters 2-8.
Chapters
lOldl iVdUK
Journal 2 3 4 5 6 7 8 citations order
J. Exp. Mar. Biol. Ecol. 16 17 21 4 16 13 15 102 1
Mar. Ecol. Prog. Ser. 11 14 13 8 7 13 14 80 2
Mar. Biol. 15 10 14 11 13 7 8 78 3
Limnol. Oceanogr. 24 15 1 2 6 8 7 63 4
Biol. Bull. 7 1 7 6 5 0 1 27 5
Can. J. Fish. Aquat. Sci. 7 3 1 0 5 7 4 27 6
J. Mar. Res. 5 6 1 0 3 5 4 24 7
J. Mar. Biol. Assoc. U.K. 6 4 5 1 2 2 0 20 8
Science 0 4 3 3 4 1 5 20 9
Can. J. Zool. 2 3 0 2 6 0 0 13 10
Helgol. Meersunters. 4 1 1 0 1 3 1 11 11
Ophelia 3 2 2 1 0 1 1 10 12
Oceanogr. Mar. Biol. Ann. Rev. 1 4 0 1 0 3 0 9 13
Nature 3 3 0 0 2 0 1 9 14
Bull. Mar. Sci. 1 7 0 0 0 0 0 8 15
Neth. J. Sea Res. 2 2 0 0 0 2 1 7 16
Other journals 36 31 11 16 36 25 41 196
Books, Proceedings, etc. 18 12 9 11 19 13 13 95
Totals 161 139 89 66 125 103 116 799
total of 196 citations here represents 25% of the overall number of cited
articles.
Aims and structure

Our central aim is to present interdisciplinary studies which use hydrody-
namic theory as an explanation for selected aspects of the benthic biol-
ogy of suspension feeders. We have deliberately considered as wide a
range of hierarchical levels of biology as possible, from organs and
physiology to ecosystems and ecology. We wished to consider the widest
range possible because we feel that it is at the interfaces of these levels
that interesting new insights may be discovered. We also hope to demon-
strate that hydrodynamics is an important factor at all hierarchical levels
of biology.
Some of the earliest writers in modern biology had realized the impor-
tance of water movement in enhancing growth. Thus, Darwin (1842)
observed how coral reefs grew fastest where wave energy was most
intense. The question as to how growth and productivity were enhanced
by water movement, and similar basic questions in other subdisciplines
Introduction 13
Table 1.6. Pioneering modern interdisciplinary studies involving

hydrodynamics and benthic biology.*
Subdiscipline Chapter Authors and date

General All Vogel (1981)
Larval biology 4 Crisp (1955)
Eckman (1979)
Hannan (1984)
Feeding physiology 5 Kirby-Smith (1972)
J0rgensen et al. (1986)
6 Rubenstein and Koehl (1977)
J0rgensen (1983)
Ethology 7 Leversee (1976)
Gruffyd (1976)
LaBarbera (1977)
Ecology 8 Wolff et al. (1976)
Wildish and Kristmanson (1979)
9 Odum (1971)
Odum et al. (1979)
Rowe (1971)
a
Note: Full citations are given at the end of the appropriate chapter.
(Table 1.6), did not begin to be adequately addressed until the 1970s,
except by Crisp (1953). Thus, most of the research that will be discussed
is less than 25 years old.
The decision to limit the questions considered to proximate ones we
believe to be justifiable and liberating. It is liberating because it is then
possible to use the same experimental approach of strong inference
introduced by Francis Bacon in 1620, rounded out by the method of
multiple hypotheses advocated by Platt (1964), for most of the studies
reviewed herein. Purely descriptive or theoretical/model studies, unless
they have been tested by inferential experimentation, have been omitted
or de-emphasized. The exclusion of evolutionary questions is also liber-
ating because it absolves us from the responsibility to devise satisfactory
methods of study. That is not to say that benthic biologists should not be
interested in questions involving neo-Darwinian evolution (see Chap. 9).
Another difficulty we faced during the preparation of the book was,
Whom are we writing it for, and what should be its style? We view our
audience as primarily active research scientists, either already profes-
sionals in the interdisciplinary field of hydrodynamics/benthic biology or

recently graduated students whom we would like to attract to it. As
regards the style, we did not set out to produce a definitive review of each
chapter topic but rather an eclectic overview of the literature up to the
middle of 1995. We are painfully aware that a different author or combi-
nation of authors might produce a quite different account. Specific rea-
sons for taking the overview - more general - approach are that because
of the wide range of subjects we wished to introduce we could hardly
claim to be comprehensive in each of them and because it became clear
to us during the book's preparation that many important questions in
this field remain to be formulated and many formulated questions have
not yet been satisfactorily answered.
We believe that an explanation of how the nine chapters are organized
will help the reader to use this book. The first chapter is a general
introduction to the subject matter considered. In Chapter 2, we present
an overview of currently used methods to studyflowand benthic biologi-
cal questions, in both the field and the flume laboratory, with selected
examples of the special methods involvingflowin larval biology, feeding
physiology, ethology, and ecology. The final chapter, Chapter 9, is a
discussion of the possible future research directions that could be taken
in the field of flow-related aspects of the benthic biology of macrofaunal
suspension feeders.
Each chapter has its own bibliography, which we hope will make
locating references to a particular subject convenient for the reader. One
unfortunate effect of this is that there is some repetition of references
between chapters.
Chapters 3 through 8 bear two similarities in their organization that
distinguish them from the others. The first is that each chapter (except
Chap. 5) is organized around a group of hypotheses which are considered
to be the major ones in the subfields concerned. Much of each chapter
involves a discussion of the experimental tests which have been reported
in connection with each hypothesis. The second similarity is that each of
the six chapters also contains a summary, including a judgement on each
null and alternative hypothesis. Each alternative hypothesis is desig-
nated as accepted, rejected, and status unclear or further work needed.
The subject matter of Chapters 3 though 8 is as follows: In Chapter 3
we deal with those aspects of larval biology most affected byflow- larval
dispersal and larval settlement/recruitment processes. Two chapters con-
cern those parts of the physiology of feeding by suspension feeders which
are most affected by flow: seawater transport to the filtration surface
Introduction 15
(Chap. 4) and capture mechanisms at the filtration surface (Chap. 5). In

Chapter 5 it proved difficult to erect a suitable null hypothesis for the
multiple possibilities of the aerosol theory, and an attempt to organize
this chapter around pertinent hypotheses had to be abandoned. In Chap-
ter 6 we present an introduction to studies involving behavioral re-
sponses to flow, and although a number of different questions are raised,
the dearth of published work means that not enough suitable hypotheses
have yet been proposed. The next two chapters deal with ecology and
flow at different hierarchical levels. The first refers to populations of
suspension feeders and flow (Chap. 7) and the second to selected eco-
systems and flow (Chap. 8).
References
Alcock, J. 1989. Animal behaviour: an evolutionary approach. 4th ed.
Sinnauer Associates, Sunderland, Mass. p. 596.
Chamberlin, T. C. 1965. The method of multiple working hypotheses. Science
148: 754-759.
Cherrett, J. M. 1989. Key concepts: the results of a survey of our members'
opinions, p. 1-16. In J. M. Cherrett (ed.) Ecological concepts: the
contribution of ecology to an understanding of the natural world.
Blackwell Scientific, Oxford.
Crisp, D. J. 1953. Changes in the orientation of barnacles of certain species in
relation to water currents. J. Anim. Ecol. 22: 331-343.
Darwin, C. 1842. The structure and distribution of coral reefs. Smith, Elder,
London.
Daugherty, R. L., J. B. Franzini, and E. J. Finnemore. 1985. Fluid mechanics
with engineering applications. 8th ed. McGraw-Hill, New York.
Denny, M. W. 1988. Biology and the mechanics of the wave swept
environment. Princeton University Press, Princeton, N. J.
1993. Air and water: the biology and physics of life's media. Princeton
University Press, Princeton, N. J.
Fenchel, T. 1987. Ecology - potentials and limitations. Ecology Institute,
Oldendorf/Luhe.
Hildrew, A. G., and P. S. Giller. 1994. Patchiness, species interactions and
disturbance in the stream benthos, p. 21-62. In P. S. Giller, A. G.
Hildrew, and D. F. Raffaelli (eds.) Aquatic ecology, scale, pattern and
process. Blackwell, Oxford.
J0rgensen, C B. 1966. Biology of suspension feeding. Pergamon Press,
London.
1990. Bivalve filter feeding: hydrodynamics, bioenergetics, physiology and
ecology. Olsen and Olsen, Fredensborg, Denmark.
Lincoln, R. J., G. A. Boxshall, and P. F. Clark. 1982. A dictionary of ecology,
evolution and systematics. Cambridge University Press, Cambridge.
Maddox, J. 1992. Electronic journals have a future. Nature 356: 559.
Mann, K. H., and J. R. N. Lazier. 1991. Dynamics of marine ecosystems:
biological-physical interactions in the oceans. Blackwell Scientific,
Boston.
Parsons, T. R., M. Takahashi, and B. Hargrave. 1977. Biological

oceanographic processes. 2nd ed. Pergamon, Oxford.
Peters, R. H. 1991. A critique for ecology. Cambridge University Press,
Cambridge.
Petersen, C. G. J. 1911. Valuation of the sea. 1. Animal life of the sea-bottom,
its food and quantity. Rep. Dan. Biol. Sta. 20: 1-79.
Platt, J. R. 1964. Strong inference. Science 146: 347-353.
Quinn, J. F., and A. E. Dunham. 1983. On hypothesis testing in ecology and
evolution. Am. Nat. 122: 602-617.
Schlichting, H. 1979. Boundary-layer theory. 7th ed. McGraw-Hill, New York.
Simberloff, D. 1983. Competition theory, hypothesis testing, and other
community ecological buzzwords. Am. Nat. 122: 626-635.
Sprintall, R. 1990. Basic statistical analysis. 3rd ed. Prentice Hall, Englewood
Cliffs, N. J.
Tait, R. V. 1968. Elements of marine ecology: an introductory course.
Butterworths, London.
Vogel, S. 1981. Life in moving fluids: the physical biology of flow. Willard
Grant Press, Boston.
1994. Life in moving fluids: the physical biology of flow. 2nd rev. ed.
Princeton University Press, Princeton, N. J.
2
Methods of study
The purpose of this chapter is to present an overview of the methods

available to research scientists involved infieldobservations and inferen-
tial field and laboratory experiments concerning benthic suspension
feeders. As well as hydrodynamics, the disciplines we consider are larval
biology, physiology, ethology, and ecology. Because of space limitations,
the methods described for each discipline could not be dealt with com-
prehensively. Instead, we have selected a few themes for each which we
believe will be most useful for the novice research scientist trying to set
up an interdisciplinary hydrodynamic/benthic biology study.
Previous reviews concerned with some aspects of hydrodynamics/
benthic biological methodology involving suspension feeders have been
presented in Vogel (1981, 1994), Denny (1988), and Jorgensen (1990).
The content of each was briefly considered in Chapter 1, and we suggest
that the appropriate book should be consulted for suitable methods if
they are not included here.
A common difficulty for a research scientist in thisfieldis to determine
whether a laboratory or field experiment, or both, is the best approach to
use in a particular research project. Thus, laboratory flume experiments
are reductionist and yield simplified environments which are question-
ably realistic. On the other hand, field experiments provide realistic
environmental conditions, but many of the variables cannot be con-
trolled or replicated, rendering the results unique. Resolution of this
difficulty, of generality versus special conditions, depends on the nature
of the question. Hence, it is more likely to be field experimentation the
higher in the biological hierarchy that the question is framed. Generality
versus uniqueness is considered by Sulkin (1990) from the point of view
of larval orientation mechanisms.
The subjects dealt with here include how to simulate flow in both field
17
and laboratory flumes. We also describe various ways of measuring

velocity, as well as methods of sampling seston quantitatively. The final
section is concerned with specialized methods associated with the inter-
disciplinary fields of flow/benthic biology studies related to larvae, feed-
ing physiology, behavior, and ecology of suspension feeders.
Flow simulation
It is required to simulateflowso as to perform experiments in which flow
speed (unidirectional flow) orflowspeed and periodicity offlowreversal
(oscillating flows) are controllable variables. The flume requirements for
physiological and behavioral experiments are often less rigorous than the
full dynamic similarity required for realistically simulating sediment-
water interface conditions within flumes.
Although for convenience we have separated methods of simulating
flow into those in the field and those in the laboratory, the distinction is
not always easy to make. This is because there is a gradual change from
one to the other, with the difference in the degree of control of the
important variables: velocity, salinity, and temperature.
Field flows
Flow simulators suitable for ecological studies in the field have been
designed to measure growth rates of small populations of bivalves, or to
determine ecosystem level fluxes of materials, e.g. seston, plant nutri-
ents, carbon, or dissolved oxygen. The former are required to reproduce
natural flows, so that growth rate is a function of independent variables
such as bulk velocity or bottom shear stress ([/*), unhindered by wall
effects or an undeveloped boundary layer. For ecosystem fluxes, the
channels require well-defined flows so that the sampling at inlet and
outlet gives a good estimate of the fluxes of the materials of interest. All
of these are flowthrough devices and use a natural seawater supply. Flow
velocities used are either natural tidal currents or simulated ones using a
submersible pump or head tank and gravity flow.
The Kirby-Smith growth apparatus consists of eight 1.5-m-long acrylic
tubes which receive seawater from a constant level head tank (Fig. 2.1).
In our experiments (Wildish and Kristmanson 1985), bivalves were sup-
ported in the mid-centre line by plastic mesh inserts which minimally
impededflow.Flow rates through the tubes could be regulated by chang-
ing the diameter of the outlet tube fitting. The bulk flow velocities are
Methods of Study 19
A -d
I
I
lU
1 metre
Figure 2.1 Kirby-Smith growth tube apparatus as used by Wildish and

Kristmanson (1985): a, inlet seawater; b, screen; c, growth tube; d, outlet tube; e,
main outlet.
calculated by timing collections of the volumetric flow and then dividing

by the tube cross sectional area:
(2.1)
where
U is the bulk velocity in cm s"1
Q is the volumetric flow in cm3 s"1
A is the cross sectional tube area, cm2
Pipe flows are well known hydrodynamically (Schlichting 1979). Mean
velocity profiles and turbulence characteristics in empty pipes can be
related to the mean pipe velocity and diameter. However, the pipes used
in the Kirby-Smith apparatus are too short for a fully developed profile
and the presence of inserts and bivalves disturbs theflow.Thus, the bulk
velocity measures recorded may not estimate the velocity experienced
by an experimental subject very well and may be apparatus-specific in
determining their responses.
1.5 m 5m
!:
i ; 64"
if r.m ; o
\i i
ii I ; 45.7
ii i ; cm
g
Figure 2.2 Multiple-channel flume of Wildish and Kristmanson (1988): upper,
plan view; lower, elevation view; a, inlet port; b, perforated plywood; c, brass
screens; d, perforated bulkhead; e, adjustable weir; f, acrylic plastic (Plexiglas)
windows; g, drain.
The hydrodynamics of flow in a tube also differs from that in a natural

benthic boundary layer. Test animals which are suspended at the centre
of the tube would not experience the same flow field as those placed on
the tube wall in the boundary layerflow.Flumes offer a better simulation
for measuring the growth of suspension feeders than tubes since within
each channel a benthic boundary layer develops on the bottom and
benthic animals experience shear flows of similar magnitude to that in
ambient flows. In most flumes discussed here, however, the benthic
boundary layer never becomes as fully developed as it does in the sea.
A multiple-channel flume was designed by Wildish and Kristmanson
(1988) to measure the population growth rate of bivalves while varying
the unidirectional velocity. The multiple-channel flume (Fig. 2.2) al-
lowed for four different velocities to be run at the same time, each with
the same initial seston concentration at the inlet end of each channel.
Seawater is pumped to the intake box by a submersible pump where a
constant head is maintained by an adjustable weir. Velocity in each
channel is controlled by rubber stoppers which can be removed from the
perforated bulkhead. Hydrodynamic observations in this flume sug-
gested that the flow had a developing boundary layer of ca. 20 cm thick-
ness at the downstream end, and, therefore, this simulation provides
flows which are more like the turbulent shear of the benthic boundary
layer than can be achieved in tube flow.
Methods of Study 21
Multiple-channel flumes have also been used in field conditions so the

natural substrate is present. The flume used by Grizzle, Langan, and
Howell (1992) consisted of four channels, 3.5 m long by 0.2 x 0.2 m, oper-
ated as three units side by side so that 12 different flow combinations
were possible. The aim of this work was to determine the effect of
velocity on representative siphonate and non-siphonate bivalve growth
rates. Each unit was supplied with natural seawater from a constant head
tank and the flow to each channel controlled by a ball valve. A plastic
meshflowstraightener was included at the inlet and a sediment box near
the outlet to permit experiments with burrowing bivalve infauna. Judge,
Coen, and Heck (1992) describe a multiple-channel flume for use in the
field with natural water movement which was similar to the single-
channelflumesused in ecosystem studies described later. Its purpose was
also to measure bivalve growth rates as a function of velocity. It con-
sisted of nine adjacent channels with walls 7m long by 1.2m high, made
of plywood sheets which were dug 0.15 m into the sediment during de-
ployment. The flume was aligned channel end to the major flow in a
seagrass meadow so that flows were regulated by size of the openings,
which were equal at either end. Channel width was varied from 0.6 to
1.2 m and was not constant along its length. A 0.25-m2 experimental plot
was located mid-way down the channel so that it was beyond any wall
boundary layer flow effects.
The other major purpose of field flumes was to measure fluxes of
materials in subsystems, such as bivalve reefs, whieh are part of larger
estuarine ecosystems. The simple tunnel designed by Dame, Zingmark,
and Haskin (1984) utilized natural tidal currents, but it required precise
orientation and did not include wind-wave forces unless the wind
direction was of the same orientation. The benthic ecosystem tunnel of
Dame et al. (1984) was constructed of 1/4-in-thick acrylic sheet,
supported by a fibreglass strip and neoprene skirt to prevent leaking
(Fig. 2.3).
A much larger, single-channel field flume of 40m x 2m width x 0.66m
was also used in the North Carolina salt marshes to measure material
fluxes from the high to low marsh (Wolaver et al. 1985). The flume walls
of the device could be removed after use and consisted of corrugated
fibreglass panels. A similar apparatus - the Sylt flume (Fig. 2.4) - of 70 m
length was constructed by Asmus and Asmus (1991) with 2-m-high walls
and a channel width of 2 m. Plastic sheet formed the walls and could be
removed from the wooden frame when not in use. To determine the flux
of materials, e.g. oxygen, which occurs as a result of the canalized water
within the flume at inlet (Ca, g O2-nT3) and outlet (Q, g O2-nT3), the
Figure 2.3 The benthic ecosystem tunnel of Dame et al. (1984): a, acrylic sheet;
b, neoprene skirt; c, reinforcement rod.
Figure 2.4 Diagram of the Sylt flume. The plastic sheeting can be removed
when the flume is not in use (Asmus and Asmus 1991).
instantaneous rate of oxygen uptake, /, in units of g O 2 m 2 h 1, is

calculated:
(2.2)
LW
where
Q is the volumetric flow over the bed in m3 h l
L is the distance, in m, between a and b
W is the path length of sediment, in m, over which the canalized
water flows
Methods of Study 23
This equation can only be used if the flow is well mixed with respect to
dissolved oxygen concentrations at the inlet and outlet. Otherwise, the
inputs and outputs, R, must be determined by integration of the form
R = \\vcda (2.3)
where v and c are the velocity and concentration of dissolved oxygen,

respectively, at the infinitesimal element of area normal to the flow, dA,
and A is the cross sectional area at inlet or outlet. Under favorable
circumstances, similar studies can be conducted without the use of a
tunnel, e.g. when the flow is shallow and vertically well mixed. In this
case, the benthic boundary layer is depth limited and an estimate of the
outlet concentration can be made as in equation (2.2) without the need
for vertical positioning in obtaining inlet and outlet samples. This
method was followed by some authors (Nixon et al. 1971; Wright et al.
1982) in determining dissolved oxygen uptake by mussel reefs. In deeper
conditions when the benthic boundary layer does not reach the water
surface, the bottom water may be depleted of dissolved oxygen. In these
conditions, it will be necessary to use equation (2.3), requiring multiple
vertical and coincident velocity sampling.
Surficial geologists (e.g. Scoffin 1968) have used an enclosed channel
firmly pegged to the sediment (3 m x 10 cm x 13 cm) through which water
is drawn by a propeller to determine the initiation of sediment motion by
flow. The propeller was driven by a sealed dynamo motor, and it was
necessary to use SCUBA divers to observe the threshold velocity
for resuspension. The Seaflume consists of a channel 4m x 15cm
high x 61 cm wide (Young 1977) and a propeller pump to create the flows
required. A camera and strobe light allow close-up photography to ob-
serve the erosional behavior of the sediments.
Detailed hydrodynamic studies have not been made for any of the field
flow simulators described. Narrow growth tubes do not provide a flow
environment like the benthic boundary layer in the ocean and flumes do
a better job of simulating natural flows. However, laboratory and field
flumes, as well as tunnels, suffer from the same limitation - that cross-
flows which are not parallel to the flume walls and are created by wind
effects, for example, are excluded. Additionally, the tunnel suffers from
a restriction in vertical mixing which could result in lowerfluxes.A study
by Asmus et al. (1992), comparing fluxes measured in a benthic ecosys-
tem tunnel and the Sylt flume at the same intertidal site, supports this
view, although not conclusively, because no replication and hence statis-
Table 2.1. Recirculating flumes designed for measuring physiological or

behavioral responses of suspension feeders to unidirectional flows.
Volumetric Free stream

Flume type capacity, L *7max, cm-s" 1 Taxa Reference
Water tunnel, 200 30 Marine plants Charters and Anderson

low speed (1980)
Vogel-LaBarbera, 20-800 40-50 "General purpose" Vogel and LaBarbera
low speed (1978)
Vogel (1981)
Raceway 60 50 Gorgonian coral Leversee (1976)
Vogel-LaBarbera 115 50 Gorgonian coral Sponaugle and LaBarbera
(1991)
Vogel-LaBarbera 98 50 Octocoral Patterson (1991)
Vogel-LaB arbera ? 50 Sea pen Best (1988)
Holland 11 4 Crinoid echinoderm Leonard (1989)
Vogel-LaBarbera 276 50 Bryozoa Okamura (1984)
Raceway ? ? Phoronids Emig and Bechernii (1970)
Belle W. Baruch ? 7 Polychaete Luckenbach (1986)
Institute Field Lab
Vogel-LaB arbera 15 7 Brachiopods LaBarbera (1977)
Saunders-Hubbard 200 100 Bivalves Wildish and Saulnier (1993)
Denny, high speed 1500 500 "Wave-exposed" Denny (1988)
taxa
tical analysis was included. A further restriction for field flumes is that
they can only operate intertidally where the tidal prism is less than a few
metres, whereas tunnels could operate subtidally at least to SCUBA
diving depths.
Unidirectional laboratory flows

Attempts have been made with laboratory flow devices to create a con-
tinuous and adjustable unidirectionalflow,an oscillatingflowto simulate
wind-wave effects, and realistic simulation of sediment-water interface
conditions.
Simpleflowsin which physiological or behavioral responses of suspen-
sion feeders can be determined do not require the large flumes found in
civil engineering or naval laboratories. The design considerations of
suitable smaller flumes are discussed by Vogel and LaBarbera (1978),
and this work has resulted in many flumes which have housed a variety
Methods of Study 25
JL
Tl U "b
1 metre
Figure 2.5 Vogel-LaBarbera recirculating flume in side view is constructed of
plywood and wooden beam supports: a, viewing window; b, outlet; c, collimators;
d, propeller shaft (Vogel and LaBarbera 1978).
of taxa (Table 2.1). Vogel (1981) pointed out that round pipe, rather than
the original square type of the Vogel and LaBarbera flume, is preferable,
and it has been used by subsequent authors listed in Table 2.1. All of the
flumes shown in Table 2.1 are recirculating, meaning that afixedvolume
of seawater is propelled through a closed, or partially open, circular
arrangement of conduits. If filtration rates are being measured on the
basis of the temporal change in seston concentration, the necessity is to
limit the volume appropriately to the filtration power of the suspension
feeder subjects. A corollary is that this limits the exposure time available
for experimental purposes with suspension feeders because of a buildup
of excretory substances.
Vogel-LaBarbera flumes are limited to flows ^SOcm-s"1, because at
greater velocities, air bubbles begin to appear as a result of cavitation
effects at the propeller. Because the flow entrance and exit conditions of
the Vogel-LaBarbera type flume (Fig. 2.5) do not reduce the instabilities
introduced by forcing the flow around four corners, it causes wave for-
mation, which limits its usefulness at higher flows. If higher flows are
required, then two flumes, the Saunders and Hubbard and Denny
high speed, are available - the former useful to 100 and the latter to
500 cm s \ The large Denny high speed flume is expensive to build, while
Figure 2.6 Saunders-Hubbard recirculatingflume:upper, side view; lower, plan

view; a, propeller; b,flumeworking section with false floor; c,fixedflowdepth; d,
flap for controlling surface flow; e, throat section; f, air removal; g, collimator; h,
turning vanes (Wildish and Saulnier 1993).
the Saunders and Hubbard (Fig. 2.6) is mid-way in expense between the
Denny and Vogel-LaBarbera flumes. The design improvements of the
Saunders and Hubbard and Denny flumes which allow higher flows free
of waves and aeration were borrowed from naval architects (Saunders
and Hubbard 1944). For both flumes, they include a throat section choke
and turning vanes at each of the corners.
An annular, open channel flume or raceway was used by Emig and
Becherini (1970) and Leversee (1976), and this design appears to have
similar maximum velocity characteristics to the Vogel and LaBarbera
flumes.
The constant head flow tube or water tunnel of Charters and
Methods of Study 27
Table 2.2. Oscillatingflowflumesfor measuring physiological and

behavioral responses of suspension feeders to bidirectional flows.
Wave character
Volumetric Oscillation umax
Flume type capacity (L) period (s) (cm-s-1) Reference
Lofquist >1000 3-25 30-150 Lofquist (1977)
Lofquist >1000 3-30 30-150 Hunter (1989)
Vogel-LaB arber a 15 l->3600 Trager et al. (1990)
Lofquist >1000 3-30 30-150 Turner and Miller
(1991), Miller et al.
(1992)
Anderson (1980) is a specialized apparatus for measuring the effect of

flow on macroalgal physiology.
Oscillating laboratory flow

Svoboda (1970) considered various possibilities of simulating the oscilla-
toryflowspresent near the sediment-water interface which are the result
of wind action at the seawater surface. A consideration of laboratory
methods of creating surface waves, primarily of interest to engineers, and
problems such as beach erosion control are presented by Sorensen
(1993). Oscillatingflowtables or eccentrically attached, swinging aquaria
were thought to be unsatisfactory, and a water tunnel in which a piston
at one end drove seawater through a horizontal tube which acted as the
working section was used (Svoboda 1970). Lofquist (1977) also em-
ployed this principle to simulate wave action for surficial sediment stud-
ies (Table 2.2) characteristic of all but the most exposed natural condi-
tions. Although it is expensive to construct, the Lofquist oscillating flow
flume (Fig. 2.7) has been used by biologists (Table 2.2) to determine
oscillating flow effects on a wide taxonomic range of suspension feeders.
The flume consisted of a water tunnel of U tube design with the vertical
part made of paired cylinders at each end, one of which contained tight-
fitting pistons, the other acting as a reservoir. A 3/4 horse power motor
moved the pistons, either in unison or separately, so that two flow
simulations could be run together. The device produces a sinusoidal flow
1m
Figure 2.7 Lofquist oscillatingflowflume(from Lofquist 1977): a, paired, hori-

zontal cylinders with drive pistons; b, paired horizontal reservoir cylinders.
with periods adjustable between 3 and 25 s and peak velocities of

SOcm-s1. The working section was divided by a removable partition
which allowed either single- or double-channel use.
A flume of much smaller volumetric capacity (Table 2.2) was used by
Trager et al. (1990) in studies with barnacles. The design of this system
required a specialized electronic circuit to control the motor-driven pro-
peller in either direction. With this system, it was possible to produce
linear- and sine-wave accelerating and decelerating seawater flows. The
maximum velocities achievable with the system described by Trager,
Hwang, and Strickler (1990) are probably <50cm-s~1 because of the
Vogel and LaBarbera flume used.
Sediment-water interface simulation

Theoretical considerations needed in designing a flume to simulate
sediment-water interfaces have been considered in reviews by
Muschenheim, Grant, and Mills (1986) and Nowell and Jumars (1987).
They indicate that the flume conditions required to simulate sediment-
Methods of Study 29
Table 2.3. Sediment-water interfaces and unidirectional flow flumes.
Flume type Dimensions Reference

Unidirectional, 3-m-Long working section x Muschenheim et al.
flowthrough 0.35 m wide (1986)
2.0 m Outside diameter x
Annular - lab 0.1m wide Taghon et al. (1984)
Annular - lab 0.1m wide Wainwright (1990)
Annular - field 0.15 m wide Amos et al. (1992)
Recirculating flume 3-m-Long working section x O'Riordan et al. (1993)
0.6 m wide
water interactions realistically are the most demanding of the flume

designer's art. Wall effects can be minimized by placing the walls as wide
apart as possible and preferably with a width/depth ratio >5:1. Nowell
and Jumars (1987) point out that the flume must be capable of simulating
the following flow and flow-sediment characteristics as they occur in the
field:
Reynolds number
Froude number
Strouhal number
Roughness Reynolds number
Shields parameter
Rouse parameter
assuming a uniform bedform and a steady, unidirectional seawater flow.
Further information on some of the listed variables is available in the
reviews cited at the beginning of this section.
A selection of the types of flume which have been used in determining
sediment-water interface interactions is shown in Table 2.3. The flume
described by Muschenheim et al. (1986) was designed to investigate
suspension-deposit feeding of spionid polychaetes in realistic initial sedi-
ment motion conditions as affected by turbulent boundary layer flows.
Annular flumes are more suitable for determining continuous sedimen-
tary transport, such as transport distance, but because of secondary
circulations inherent in annular flow are not ideal for determining sedi-
mentary transport rates. Flow is driven in the annular flumes by a float-
ing ring, with or without paddles, propelled from a small motor operating
in air. A field version of the annular flume - the Sea Carousel - from
Amos et al. (1992) was designed to measure in situ sediment erodibility
of shallow inter- or subtidal locations. For this purpose, it is equipped
with optical backscatter sensors, a video camera, and an electromagnetic
flow meter.
The interaction of siphonal exhalant currents generated by bivalves
with the development of the concentration boundary layer was studied
by introducing rhodamine dye into the exhalant flow by O'Riordan et al.
(1993). The effect of different boundary layer velocities on the concen-
tration field was determined by using a non-intrusive snapshot of the
dye concentration field. The dye was illuminated with a thin sheet of
laser light, and a charged coupled device camera was used to capture
images of the dye concentration. As the fluorescent light intensity is
proportional to the dye concentration, the method indicates rhodamine
concentration.
Velocity measurement
In order to choose a suitable velocity measurement system, answers to
some key questions must be obtained. The questions include, Is the
measurement to be of ambient velocity in the field or in the laboratory
flume?, At what spatial and temporal scales is velocity to be determined?
and At what hierarchical level of biological organization are the velocity
measurements to apply?
If the hierarchical level of biological organization of interest is ecologi-
cal, then the standard techniques of physical oceanography, such as
electromagnetic current meters, are applicable. Such current meters pro-
duce a continuous temporal record of velocity, and often current direc-
tion, stored electronically and retrievable with a laptop computer with-
out moving the sensor (Table 2.4). The interested biologist, by either
luck or design, may also be able to use a predictive physical oceano-
graphic model to input mean and maximum tidal velocities (e.g. Wildish,
Peer, and Greenberg 1986).
A limitation of electromagnetic current meters, and the models they
give rise to, is that they usually do not measure orbital motions due to
wind-wave effects. Studies by Denny (1982) on wave-exposed rocky
shores were based on transducers to measure the wave force received by
models of suspension-feeding organisms, e.g. barnacles. The force trans-
ducer output was converted to a frequency modulated (FM) radio signal
Methods of Study 31
Table 2.4. Partial list of commercial suppliers of equipment required

for flow measurement
Equipment type Supplier

Electromagnetic current Marsh-McBirney, Inc., 8595 Grovemont Circle,
meters Gaithersburg, Maryland, 20760 USA
Inter Ocean Systems Inc., 3540 Aero Court,
San Diego, California, 92123 USA
Acoustic current meters EG&G Neill Brown Instruments Systems, 1140
Route 28A, P.O. Box 498, Cataumet,
Massachusetts, 02534 USA
Miniature flow meters Nixon Instrumentation, Cirencester Road,
Cheltenham, Gloucestershire, UK GL53 8DZ
Omega Engineering Inc., One Omega Drive,
P.O. Box 4047, Stamford, Connecticut, 06907
USA
Thermistors Victory Engineering Corp., 1-T Victory Rd.,
Springfield, New Jersey, 07081 USA
Laser Doppler flow meters DISA Electronics Inc., 779 Susquehanna Ave.,
Franklin Lakes, New Jersey, 07417 USA
TSI Inc., Fluid Mechanics Instrument Div., P.O.
Box 64201, St. Paul, Minnesota, 55164 USA
Flow meters Blue White Industries, 14931-T Chestnut St.,
Westminster, California, 92683 USA
received and stored at a shore-based laboratory (Denny 1982). The

transducers gave information on the magnitude and maximum force
exerted by waves, but further interpretation (Denny 1985) provided
information on drag on the model, flow acceleration near the model, lift
forces to which the model was subject, velocity and acceleration of
waves, as well as the timing of each wave. This system is obviously
expensive to construct, with data difficult to interpret and the instrumen-
tation not very portable. A simpler, cheaper method was designed by
Bell and Denny (1994) to determine maximum velocity during a passing
wave. The equipment (Fig. 2.8) consisted of a spring and a hollow ball
attached by monofilament line which acts as a drag element. As the wave
passes the hinge-anchored instrument, the ball and spring are stretched,
thus moving a rubber tube stretched over the line, indicating its maxi-
mum extension. From the linear relationship between the force and
spring extension distance, the drag force on the ball or maximum wave
Figure 2.8 Maximum force recorder of Bell and Denny (1994): a, spring; b,
rubber sleeve indicator of maximum force; c, upper plug; d, monofilament line; e,
practice golf ball.
velocity can be determined. The spring extension response is non-linear

with respect to wave velocity and maximum velocities measurable, lim-
ited to >2.5 m s"1 as a result of inertia of the spring used.
The gypsum dissolution method can be used as a crude means of
indicating water movement. The method incorporates any water motion
due to tidal as well as wind-wave effects in the same integrated measure,
although only recently have attempts (Thompson and Glenn 1994) been
made to provide an absolute measure in seawater. Further advantages of
the gypsum dissolution method are that it is inexpensive, the results are
simple to interpret, and it is widely applicable for use in nearshore
continental shelf environments at all levels of biological organization.
The rate limiting step for dissolution of gypsum CaSO4 2H2O is trans-
port through the diffusive sublayer (Lui and Nancollas 1971). Thus,
increases of water movement near the diffusive sublayer enhance diffu-
sion because of a decrease in the thickness of the boundary layer and/or
an increase in the concentration gradient across it due to rapid transport
of Ca2+ and SO2^ away from their site of dissolution.
The dissolution rate of gypsum follows a first order kinetics equation
(Lui and Nancollas 1971). Thus, for any shape of gypsum cast, the rate of
loss of mass is given by
v 5 7
dt ~ v^# v
where
m = mass of the cast in kg at time t
K = m-s"1, rate constant for a specified shape
A = m2, area exposed to flow
Cs = kg m3, saturation constant of gypsum
C = kg m3, concentration of dissolved gypsum
Methods of Study 33
Following Petticrew and Kalff (1991), this equation can be reformulated

as a flux, F, as follows:
j ( ) (2.5)
where
F = transfer rate of gypsum per unit area, kg m2 s"1
D = diffusion coefficient for CaSO4 in water, m2 s"1
8 = concentration boundary layer thickness, m
The dissolution rate of gypsum is sensitive to temperature for two rea-
sons: The diffusivity increases with temperature and decreases in vis-
cosity are caused by increases in temperature, which, in turn, reduce
boundary layer thickness, 8.
A number of shapes and materials have been used to measure dissolu-
tion rates - small rings, actually candy (Lifesavers) - sewn to a kelp
surface (Koehl and Alberte 1988) to "clod cards" of gypsum cast in
plastic ice cube trays (Doty 1971; Jokiel and Morrissey 1993; Thompson
and Glenn 1994) to cylinders (Wildish et al. 1997). The ideal shape for
such measurements is a sphere, since flow from any direction should
equally influence the dissolution process. A capped right cylinder will
also be insensitive to the direction of flow as long as its main axis is
oriented perpendicular to the flow. The weight initially (Mo) and after a
measured exposure period, drying, and reweighing (Mx) may be used
in a comparative or absolute measure of water movement. The
dimensionless index proposed by Doty (1971), the diffusion factor, DF,
is calculated as
[ J flow
(2.6)
still
where the weight loss, determined in the field for a standard time, is
divided by the weight loss in still seawater of the same salinity and
temperature.
In order to relate gypsum cylinder weight losses to flow in an absolute
measure of velocity, it is necessary to calibrate flume flows and relate
weight loss to salinity, temperature, and velocity. In the field, replicate
cylinders may be deployed (Fig. 2.9) on a frame which includes a swivel
and vane so that the gypsum cylinders face the current.
Five general techniques have been used to measure velocity in labora-
Figure 2.9 Deployment system for gypsum cylinders: A, frame constructed of

steel wire for four gypsum cylinders held in place by rubber bands; B, buoy and
anchor system for deployment and retrieval from a small boat.
tory flumes: electrochemical, mechanical, thermistor anemometry, laser

Doppler velocimetry, and flow visualization. The method used will de-
pend on the purpose of the velocity measurement, the velocity range
required, as well as the size of the animal and flume size used in the
experiment.
Electrochemical methods are usually not suitable for studies involv-
ing living animals because the chemicals involved may be toxic
(Muschenheim et al. 1986). The hydrogen bubble method developed for
use in freshwater (Schraub et al. 1965) is one exception. It was success-
Methods of Study 35
fully adapted for measuring flume boundary layer velocities above a

mussel reef in seawater by Wildish and Kristmanson (1984), who con-
structed an electronic square wave simulator in which the electrical
frequency and pulse width to the electrode could be adjusted. Velocity
was determined by finding the pulse width of known time which just
filled a gap of known distance with a continuous sheet of hydrogen
bubbles, followed by calculating unit distance/time. Use of a very fine
platinum wire as the cathode is required to encourage the formation of
sufficiently small and hence slowly rising bubbles. However, because the
electrical current densities required to overcome the polarization voltage
in seawater are high, the bubbles tend to be large and, consequently, rise
too rapidly towards the surface.
The standard mechanical method of hydrodynamics is the Pitot tube
(Bradshaw 1964), but applicable only where a steady uniform flow is
available. The method measures the dynamic pressure where the flow is
sampled and velocity is calculated from the Bernoulli equation (Vogel
1994). At 20cm s^-flows, the Pitot tube pressure differential is ca. 0.2cm
freshwater. The precision of the velocity measurement will depend on
the precision of the manometer used, but will not likely give satisfactory
results below 20 cm s"1. Electronic pressure transducers may yield better
results at lower velocities.
The other common mechanical method of measuring flume velocity is
by miniaturized current meters with cup or rotors in special bearings
which minimize frictional torque. We have used a Streamflo model 403
probe, which is 47 cm long and has a rotor diameter of 11mm, protected
by a metal ring (Table 2.4). The rotor measures the integrated velocity in
an area of 95 mm2 and is thus suitable for measuring free-stream flows or
determining velocity related to physiological or behavioral experiments
of the same size, e.g. of bivalves or larger-sized organisms. The square
wave signal generated has a pulse rate proportional to the rotor speed
and hence ambient velocity. The indicator used with the rotor probe has
a three-digit light-emitting diode (LED) display in hertz with continuous
or time-integrated sampling of 1 and 10 s possible. The output can also be
displayed on a recorder or imported to a personal computer (PC)
spreadsheet. The rated velocity range of the model 403 probe is 2.5-
150 cm-s"1 with an accuracy of 1% for most of the range, except at
<15cm-s~1, the accuracy declines to 5% at 2.5 cm-s"1.
Suitable thermistors for constructing a homemade thermistor bead
velocimeter according to the temperature compensated resistor bridge
circuit diagram given in Vogel (1981) are available from the Victory
Engineering Corporation (Table 2.4). Complete thermistor anemometer

systems are available from DISA Electronics Incorporated (Table 2.4).
Because of the small size of the sensing heads or beads, typically 0.1-
2.0 mm in diameter, the thermistors are suitable for resolving com-
pressed flume boundary layer flows in steps of 1 mm. Charriaud (1982)
has used a thermistor probe of 0.1-mm diameter mounted on a micro-
manipulator to measure velocity profiles across the actively pumping
cloacal siphon of an ascidian in steps of 1 mm. Each thermistor probe
used requires separate determination of a probe constant and is tedious
to calibrate. The amplifier output is a curvilinear response of velocity, so
the useful range may be limited. Further details of thermistor
anemometry can be found in Gust (1982).
The Doppler effect can be used to profile flow velocities within the
boundary layer with a laser Doppler velocimeter (Table 2.4). Further
details of these systems can be found in Penner and Jerskey (1973).
Optical receivers measure the scattering of laser light by natural particles
or fluorescent dyes (Koochesfahani and Dimolokis 1985) within the
ambient flows; thus this is a specialized version of the flow visualization
method.
Theflowvisualization method has the distinct advantage of being non-
invasive; therefore, the sensor itself does not influence local flows. An
early example of the flow visualization method was reported by Vogel
and Feder (1966), in which low speed flows seeded with plastic particles
illuminated by pulses of light (12flashes-s"1) from a strobe source were
recorded photographically. Carey (1983) used denatured egg white
particles in the size range 0.25-1.00 mm in a seawater flume. Miron,
Pelletier, and Bourget (1995) used polyvinyl chloride particles of 125-
300|nm in diameter in flume studies of scallop spat collectors. Time-lapse
photography through the acrylic plastic wall of the flume allowed calcu-
lation of particle trajectories and velocity by determining the particle
streak length for the known time-lapse period.
The use of natural particles such as seston as the flow streak line
monitor is a considerable improvement on the previous methods. Trager
et al. (1990) used such a system with the small oscillating flow flume
described in the section, Oscillating laboratory flow. The optical part of
the apparatus consisted of a fiber-optic white lamp and expanded colli-
mated laser beam of 740 nm, both focused onto the barnacle subject in
the 10-cm-wide flume channel. A video camera fitted with a dark field
lens system, capable of magnifications up to 50 times, was used to record
the images. The latter, inclusive of the sestonic particles, appear white on
Methods of Study 37
Figure 2.10 Velocity calibrator: a and b, flow meters of different capacity; c,

velocity controls; d, position for insertion of probe (Streamflo 403); e, flow
formers; f, seawater drain valve.
a black background and, after frame-by-frame analysis, were used to

provide data on particle velocity as well as feeding movements of the
barnacles. This system was originally described by Strickler (1985) and
Leonard (1989).
A general purpose velocity calibrator design is shown in Fig. 2.10. It
consists of a 20-cm-diameter acrylic tube and an overflow tank. Seawater
from a piped supply is passed through one or both adjustable volumetric
flow meters (Table 2.4) and so into the bottom of the acrylic tube. The
collimated flow passes through a 5-cm-diameter orifice, just above which
test probes such as the model 403 Streamflo can be introduced so that the
rotor is normal to the flow. Factory-calibrated Streamflo probes were
Figure 2.11 Thermistor flow calibrator: a, inlet; b, baffle chamber; c, Nytex

screen; d, entrance chamber; e, flared acrylic tube; f, thermistor (Muschenheim et
al. 1986).
used to confirm that velocity at the test point was equal to the volumetric
flow divided by the orifice area. Calibration plots are made of probe
output in hertz as a function of volumetric flow rate.
A similar but miniaturized version of the design was used by
Muschenheim et al. (1986) to calibrate small thermistor bead probes
(Fig. 2.11).
Seston sampling
The purpose of seston sampling is to determine the concentration and
composition of seston particles in seawater. During feeding by suspen-
sion feeders, it is necessary to take grab samples of seawater with small
diameter sampling tubes, because of the localized nature of the concen-
tration gradient of seston within the benthic boundary layer.
Three general types of seston sampling may be used to determine
seston concentration and composition: isokinetic, non-isokinetic, and
active suspension feeder-biased. Non-isokinetic is the default option,
where sampling is arbitrary, not considering flow and concentration
boundary layer conditions. Isokinetic is the appropriate means of sam-
pling for passive suspension feeders, and it, or species-biased sampling to
simulate the live animal, is appropriate for suspension feeders with an
active ciliary or muscular pump.
Isokinetic sampling
A theory of sampling solid particles from flowing fluids has been devel-
oped for aerosols in industrial applications (Hinds 1982). Sampling er-
Methods of Study 39
rors may arise in sampling flowing fluids to measure the solids content as
a result of the following:
A mismatch between ambient and sampling tube velocities
Variations in settling velocities of the sestonic particles
The orientation of the sampling tube with respect to the flow
With regard to the latter, Hinds (1982) showed that as the orientation of
the sampling tube was moved from opposing the flow (0) to an increas-
ing angle until it was normal to the flow (90), so sampling errors due to
mismatch of flows also increased. Estimates of this error are given in
graphical form in Hinds (1982).
The key hydrodynamic parameter for sampling is the dimensionless
Stokes number, Stk
W JJ
Stk = ^ ^ (2.7)
where
Ws = passive terminal settling velocity of seston, cm s"1
Uo = ambient velocity, cm s"1
g = gravitational constant, 981 cm s~2
D = diameter of sampling tube, cm
The Stokes number indicates the likelihood that the particle will follow
the fluid streamlines, with higher values of Stk giving greater sampling
error. Thus, if Stk < 0.01 and the velocity mismatch between ambient
and sampling flow is within 4 times the isokinetic rate, then the sam-
pling error will be <5% (Hinds 1982). The maximum terminal settling
velocity of a sestonic particle at a fixed Stk = 0.01 can be calculated
from
W,<9-f (2.8)
Assuming D = 0.64 cm, we can calculate at Uo = Scm-s"1 that seston with
a settling velocity of -l^Scm-s" 1 will be collected without significant
sampling error. Similarly, at lOcm-s"1, but otherwise the same condi-
tions, the limit for collection bias is particles of settling velocity of
-0.63 cm-s"1. These considerations suggest that sestonic particles such as
typical phytoplankton cells with Ws < 0.01 cm s"1 would be collected
without sampling error (see also Chap. 5). However, if the seston load
contained denser sedimentary particles >100|mn, then sampling errors, if
strict isokinetic sampling were not practised, might occur.
Table 2.5. Matching flume flows, Uo, and sample tube flow at the
iso kinetic poinC
Calculated isokinetic Static head Time to empty

(cm-s-1) point (cm^s"1) (cm) 180-cm-long tube (s)
3 0.05 18 65
5 0.08 24 40
8 0.12 35 24
10 0.15 42 20
12 0.19 48 16
15 0.23 60 13
18 0.28 65 11
20 0.31 77 10
25 0.39 95 8
a
Tubing is D = 0.140 cm, A = 0.0154 cm2. PE200 Clay-Adams Intramedic
tubing, Becton and Dickson, Parsipanny, NJ 07054, USA. Static head required
to deliver matched flows is shown with the time required to empty the total
tube volume completely.
The position of the passive or active suspension feeder's sampling

surface or inhalant opening with respect to boundary layerflowis critical
in determining where sampling should take place. A simple method of
sampling from a flume is to deploy a tube near the sampling surface or
inhalant of the experimental animal and then allow the other end to
siphon. The distance from the free flume surface to the outlet end of the
tube is the static head and determines the sampling velocity and volumet-
ric flow rate. Different tube diameters and lengths require experimental
calibration (e.g. Table 2.5), which is conveniently achieved with a volu-
metric cylinder and stopwatch. The static head can then be varied to
obtain a range of different exhalant flow velocities. Isokinetic sampling
requires sampling tubes opposing the flow and matching the sampling
tube flow to the ambient flow at the appropriate boundary layer height.
Alternatively, the isokinetically matched flow can be achieved with an
adjustable clamp and pump.
Non-isokinetic sampling
From the preceding discussion, we conclude that non-isokinetic sampling
will rarely lead to serious sampling errors in collecting seston. Only
where sedimentary particles are resuspended during feeding obser-
vations or experiments, as was the case in a study of a facultative
Methods of Study 41
Figure 2.12 Active suspension feeder biased sampling. Sampling ports at 1, 5,

15 cm above the substrate and with internal diameter of 6.4 mm to simulate a
6-cm Mercenaria mercenaria (Judge et al. 1993).
suspension-deposit feeding polychaete by Muschenheim (1987), for

example, could non-isokinetic sampling lead to serious seston sampling
errors.
Active suspension feeder-biased sampling

In active suspension feeder-biased sampling an attempt is made to simu-
late the localized hydrodynamics of a specific suspension feeder in its
initial feeding responses. Judge et al. (1993) described an original at-
tempt to do this with the hard clam Mercenaria mercenaria. The sampling
device used to simulate hard clam filtration is shown in Fig. 2.12. Simul-
taneous samples were collected at three depths to simulate clam position
on eelgrass stems (1, 5, and 15 cm above the sediment) by manual opera-
tion of syringes so that the intake velocity matched that previously
reported in hard clams (Walne 1972). Sampling tube diameter,
D = 0.64 cm, was chosen to follow closely that of the inhalant siphon
diameter of a 6-cm-valve-length hard clam.
Whether this form of sampling is more widely applicable to other
active suspension feeders will depend on whether it is possible to deter-
mine accurately the following required variables for a wide range of

species:
The inhalant velocity (see the section, Initial feeding responses)
The inhalant siphon area
The siphon height above the sediment surface
The orientation of the inhalant siphon with respect to ambient flow
An integral assumption of active suspension feeder-biased sampling is
that these variables are constants, yet evidence suggests that this is not
the case. Thus, inhalant siphon area and filtration capacity are inversely
linked to ambient flow in scallops (Wildish and Saulnier 1993); it is
difficult to see how this could be simulated.
Seston collection
Before a quantitative estimate of seston concentration can be made, it is
necessary to filter the sample - all or a suitable subsample - so that the
result can be given as dry weight per unit volume of seawater. Other
measures of concentration commonly used include chlorophyll a, deter-
mined either fluorometrically or spectrophotometrically (Strickland and
Parsons 1968); adenosine triphosphate (ATP), as an indication of all
living cells or tissue (Holm-Hansen and Booth 1966); and the acridine
orange epifluorescence direct counting technique of Hobbie, Daley,
and Jaspar (1977) to indicate numbers of live bacteria (Wildish
and Kristmanson 1984). Electronically sizing and counting sestonic par-
ticles (Strickland and Parsons 1968) are other frequently used options
in determining seston concentration. More recently, the wider availa-
bility of the flow cytometer has meant that this technique has been
used (e.g. Shumway et al. 1985; Boucher, Vaulot, and Partensky 1991)
to measure particle numbers and fluorescence simultaneously. Shumway
et al. (1985) showed that even in mixed phytoplankton cultures con-
taining three species, this technique could indicate the proportional
presence of each species. The flow cytometric technique as it applies
to seston in seawater is reviewed by Yentsch et al. (1983); also see
the review of particle measurement in seawater edited by Demers
(1991).
All of the measures of seston concentration described - except the
simplest, dry weight determination - are selective, and none of them is
capable of indicating the quality of the seston as a food source to a
named suspension feeder.
Methods of Study 43
Seston quality
In a study of blue mussel initial feeding responses with a unialgal culture,
Wildish and Miyares (1990) found that two factors influenced filtration
rate - velocity and chlorophyll a content per cell. In these experiments,
the mussels were all fed the same initial cell density of the cultured
microalga, but, as a result of growth/time related differences in different
cultures, the chlorophyll a content varied, with evidence that the mussels
preferred those cells with higher levels of chlorophyll a. This same
method cannot be used with a mixed culture or with wild-caught
phytoplankton because of differences in amounts of chlorophyll a typical
of a species as well as the intraspecific variances already noted. Use of
flow cytometric techniques may allow more species of cultured
microalgae in a mixed culture to be resolved (e.g., see Olson and Zettler
1995), even where preferential clearance of one species over another is
present. However, it too may be of limited use in field samples, where
many phytoplankton species are usually present and only a few of them
are being ingested.
The nutritional quality of seston for suspension feeders could not be
adequately expressed by gross biochemical measures of the seston
(Navarro and Thompson 1995). Navarro and Thompson proposed a
dimensionless food index based on the proportion of protein, carbohy-
drate, and lipid content of the total seston available. They demonstrated
seasonal changes of the food index, but did not relate them to the growth
of any named suspension feeders.
The nutritional value of seston to suspension feeders can only be
determined in standardized feeding growth bioassays. The availability of
methods to create microparticulate diets (Jones, Mumford, and Gabbott
1974; Langdon, Levine, and Jones 1985; Southgate, Lee, and Nell 1992)
with known types and quantities of nutrients suggests that for larvae,
juveniles, and adults it is possible to create a standard diet for genetically
standardized suspension feeders. Such a diet would be of great value in
comparing the quality of the available natural seston as a food for sus-
pension feeders.
Specialized methods
Many different methods have been used in studying the larval biology,
physiology, ethology, and ecology of benthic suspension feeders. Be-
cause there are so many, we have selected a few methods for each of the
four disciplines concerned and present them here.
Larval swimming, feeding, settlement, and recruitment

One advantage of the small size of the ciliated larvae of suspension
feeders is that live specimens can be studied individually in relatively
small apparatuses. This represents a considerable saving, for example, in
the amount of live culture necessary to maintain the experimental stock
and run feeding experiments. In reviewing the literature on swimming
velocity of larvae, Chia, Buckland-Nicks, and Young (1984) described
two general methods in which swimming was observed directly with a
microscope, or indirectly with tethered larvae followed by high speed
photography and frame-by-frame analysis of the exposed film. From the
latter, the larval swimming speed can be calculated as one-half that of the
ciliary tip speed. An important consideration with this method is that at
such low Reynolds numbers, the drag experienced by a swimming larva
is influenced by walls as far away as several hundred body diameters
(Chia, Buckland-Nicks, and Young, 1984).
The use of tethered larvae was introduced in feeding studies by
Strathman, Jahn, and Fontsia (1972); also see Strathman and Leise
(1979) and Gallagher (1988). Tethering involves attaching the larva to a
hair or thin glass rod by cyanoacrylate adhesive or with a fine suction
pipet to immobilize it. A small recirculating flow tank of low volumetric
capacity, inclusive of the head tank (Fig. 2.13), was designed by Emlet
(1990) for examining the effect of tethering in flows up to 4mm-s~1.
Pediveliger larvae of Crassostrea gigas were one species used in these
experiments with film exposed at 200 frames-s"1 (Emlet 1990).
Fluorescent microparticles have been used to label food particles in
experiments on settling with living bivalve larvae (Lindegarth, Jonsson,
and Andre 1991). In this method, the larvae are fed polyvinylchloride
particles with densities similar to those of the larvae so that they do not
influence sinking rate after ingestion. Before the microparticles are fed
to the larvae, they are coated with a fluorescent pigment so that if
illuminated with a suitable light source, the particles fluoresce, and a
camera fitted with an appropriate light filter can detect the fluorescent
label. The label remains visible in the larval gut for a short time - a half-
life -10 h - so that experiments must be short term if a single pulse
feeding is employed.
Laboratory experiments involving larval settlement over soft
Methods of Study 45
1 0
0
o
o
o
b
c
(
71
1 cm
Figure 2.13 Miniature flow tank for observing tethered larvae (after Emlet
1990): a, air lift; b, upstream reservoir; c, downstream reservoir; d, inlet with
control valve; e, collimator; f, larva; g, coverslip; h, tether; i, exit port with control
valve.
sediments require full dynamic similarity of the sediment-water inter-

face in the flume (see the section, Sediment-water interface simulation)
with that in the field. Examples of laboratory experiments reaching this
degree of control in flume setups are those of Eckman (1979), Hannan
(1984), Butman, Grassle, and Webb (1988) and Grassle, Snelgrove, and
Butman (1992).
Field sampling of larvae has been achieved with plankton nets or
sediment traps. It was shown by Hargrave and Burns (1979) that the
trap shape and sampler height/mouth diameter, or aspect ratio, affected
the amount of sediment collected for a given sampling area. For high
aspect ratio sediment traps, the amount of particulate matter collected
varied asymptotically with the amount of turbulent mixing. Similar high
aspect ratio traps are also suitable for sampling larvae if a means of
preventing active larval swimming and advective removal from the trap
is provided. The latter can be achieved by including a formalin layer in

the bottom of the trap which kills and preserves the captured larvae
(Yund, Gaines, and Bertness 1991). Butman (1986) pointed out that trap
sampling biases were common: Trap larval catches are relative and de-
pendent on design, sampling efficiency decreases with particle size, and
different ambient flows influence larval collection efficiency. Thus, it is
not possible to compare the many studies reported because of differ-
ences in shape of sampler and flow conditions sampled. However, within
a given study, a particular trap design should give adequate comparative
data as long as sampling is directed to one larval size class (Yund et al.
1991).
Because of the dependence of particulate catch rates on the velocities
near the trap, Hargrave et al. (1994) designed a sampler which can
respond to different current directions and velocities. The current-
activated sediment trap consists of a steep-walled funnel of aspect ratio
1:5, a motor-driven carousel carrying up to 12 sampling cups, and an
acoustic current meter. A microcomputer uses the current direction and
velocity to switch any sampling cup to the collector position, according to
pre-programmed thresholds of velocity and compass direction. A sam-
pler of this type may be useful in larval trap design in order to remove
velocity/direction biases.
Studies of larval recruitment on hard, rocky substrates can be achieved
by allowing the larvae to settle on a cleared portion of the natural
substrate or on a specially designed flat settlement plate. The type of
plate - thin, faired; bluff; split or angled - was used by MuUineaux and
Butman (1990) and MuUineaux and Garland (1993) to manipulate flows
over the plate as a means of investigating larval settlement caused by
differences in small-scale boundary layer flows.
In the deep sea where it is required to conduct larval settlement
studies, various types of larval mud boxes, lowered by submersible or
wire and retrieved by transponder and a flotation device, have been
designed (Snelgrove, Butman, and Grassle 1995). These devices were
tested in flume experiments to show the hydrodynamic artifacts which
can be associated with them. According to Snelgrove et al. (1995), care-
ful design consideration can limit these flow artifacts.
Larval settlement rates have been shown to be affected by the density
of previously settled larvae. Thus, in field studies where natural rock
surfaces were used, Minchinton and Scheibling (1991) showed how
the timing of removal of larvae in the low intertidal zone influenced
the densities estimated for recruitment. Both recruitment and post-
Methods of Study 47
settlement mortality rates decreased exponentially as the sampling fre-

quency was decreased. This finding suggests that measured recruitment
densities will not be comparable unless both studies incorporate the
same sampling frequency. It is of obvious importance in studying the
supply side ecology of aggregating suspension feeders.
Initial feeding responses

The physiological responses of feeding, growth, and elimination in sus-
pension feeders are defined in Chapter 4. Because we are interested only
in those feeding responses influenced byflow,we limit discussion here to
initial feeding responses. The initial feeding of suspension feeders in-
volves rate processes defined as follows:
Seston uptake rate, S, the mass of seston removed from the water column in
unit time by a standard weight of a suspension feeder
Clearance (=filtration)rate, R, the calculated volume of seawater filtered,
assuming complete removal of sestonic particles, in unit time and for a
standard weight suspension feeder
S is not necessarily equal to the seston consumed, e.g. bivalves may reject
some seston as pseudofaeces. Hence,
Feeding rate, F, the mass of seston ingested in unit time by a standard weight
of a suspension feeder
R cannot always be calculated from the volumetricflowrate through the
trophic fluid transport system because the retention efficiency of the
filtration surface will rarely be 100% or known. It may also temporarily
change, e.g. by gill bypassing in bivalves (Famme and Kofoed 1983). The
efficiency of the filtration surface in retaining sestonic particles is also of
interest for studying mechanisms of particle capture. Thus:
Retention efficiency, RE, is the proportion of seston particles captured at the
filtration surface in relation to the total of those which approach it
The amount of energy-requiring work done in order to pump seston-
containing seawater through the trophic fluid transport system of active
suspension feeders is presumably related to the rate of this process:
Pumping rate, P, which is the volume of seawater that actually passes through
the trophicfluidtransport system of a standard weight animal, irrespective of
retention efficiency, in unit time
The clearance rate should not be confused with the pumping rate, and
these terms are related by the retention efficiency thus:
R = P(RE) (2.9)
Seston particles which differ in size or chemical composition will vary in
retention efficiency in bivalves (Famme and Kofoed 1983), so that the
clearance rate can be said to be particle-type-specific.
Historically, indirect methods for measuring S or R have been most
commonly used. Such measurements have been made in closed tanks
of fixed volume ("static") or in variously shaped flowthrough vessels.
The difference between the initial, Q, and final, C2, concentrations
of seston, or the inlet and outlet concentrations is used to estimate S
or R. The volume of the static system, V, and the mass of animals, M,
must be known as well as the elapsed time of the experiment, T, to
calculate S:
(2.10)
MT
Typical units for this equation might be g-seston-g"1 dry weight -h"1.
The clearance rate for a unit weight of animal can be estimated from
R = SC~1 {111)
with reasonable accuracy if Cx and C2 are not greatly different (say,
C2 > 0.9 C\). Unacceptable inaccuracies may arise because the equation
is based on the assumption that the seston concentration of the filtered
seawater remains at Cx during the experiment. Hildreth and Crisp (1976)
recognized this difficulty and suggested that Q in equation (2.11) be
replaced by the actual seston concentration around the suspension
feeder throughout the experimental test.
If Q divided by C2 is relatively large, the formula of Coughlan (1969)
is preferable:
If a flowthrough system is employed, the Wrterm in equations (2.10)

and (2.12) can be replaced by the volumetric flow rate. Then the terms Q
and C2 will denote concentrations in inlet and outlet flows, respectively.
The same difficulties in specifying the concentration term in equation
(2.11) - that is, the actual concentration of seston present in inhaled
seawater - will also be present. This concentration will be between Q
and C2, its exact value depending on the mixing characteristics of the
flow-through vessel.
Methods of Study 49
The standard indirect methods of determining S or R currently in use

suffer a number of drawbacks:
Static containers or experimental boxes cannot simulateflownear the suspen-
sion feeder, which, in some cases, may directly influence S or R.
Seston concentration depletes during the experimental test and this may
change the suspension feeders' response to it. Winter (1973) used a constant
volume apparatus for measuring R at a constant concentration. Again, the
mixing characteristics are critically important in determining whether the
results are simply apparatus-specific or more universally applicable.
The indirect method is based upon the assumption, which cannot be fully
satisfied, of instantaneous back-mixing of exhalant flow with the rest of the
seawater within the experimental chamber (see Hildreth and Crisp 1976;
Riisgard 1977). The actual extent of back-mixing is dependent on the geom-
etry of the experimental chamber and the flow rates.
Errors arising from incomplete mixing can be reduced by arranging experi-
ments with relatively small changes in Q, but this is a useful strategy only if
Cx and C2 can be measured with high precision. The standard error of the
estimate of R is mostly determined by that associated with Q - C2. If Cx and
C2 are close, large standard errors result in calculating R even with relatively
small errors in C. For example, if Q = 10.0 and C2 = 9.0 with standard errors
for each of 0.1, the greatest fractional error in Cx - C2 is 0.14 (Topping 1957).
A number of different direct methods of determining filtering or

pumping characteristics have been tried. The constant level apparatus
for bivalve molluscs originated with Galstoff (1928) and has been im-
proved by a number of subsequent authors (e.g. Drinnan 1964). The
method involves separating the inhalant and exhalant flows when the
pumping rate can be directly determined from timed collections of
the exhaled water. Filtration retention efficiency (RE) can be deter-
mined by comparing the seston concentrations at the inhalant and exha-
lant. It is necessary to ensure that hydrostatic pressure differences be-
tween the inhalant and exhalant sides do not influence the pumping rate
(see Chap. 4). The method is difficult to set up and bivalves may respond
to the flow divider by reducing pumping, although considerable use has
been made of the constant level apparatus introduced by Famme,
Riisgard, and J0rgensen (1986). The water level in the exhalant chamber
is monitored so that any pressure difference between the two chambers
can be maintained with the shunt closed (Fig. 2.14).
Other direct ways of measuring pumping rates include micromethods
to determine velocities or pressure changes in the exhalant flow. Because
of the small size of the exhalant siphon, it is important to use probes of
small dimensions to prevent interference with the flow. Amouroux,
Revault d'Allonnes, and Rouant (1975) described a hot film probe with
a 0.2- x 0.4-mm tip for measuring local velocity; Foster-Smith (1976) and
Figure 2.14 Direct method of determining the pumping rate of a mussel (after
Famme et al. 1986): a, laser source; b, mirror; q and c2, inhalant and exhalant
sides of a divided chamber; d, membrane; e, shunt; f, pump; g, scale.
Jones and Allen (1986) describe a pressure sensor with a <l-mm tip for
sampling inhalant/exhalant pressures. Because of the complex flow and
pressurefieldswithin the exhalantflow,it is critical to position the sensor
precisely. Because of these exacting requirements, such methods have
not often been successfully used.
Mohlenberg and Riisgard (1978) describe an apparatus which can be
used in non-flowing conditions to sample the exhalant flow of active
suspension feeders isokinetically. The sampling flow is adjusted so that
the seston concentration is minimal, at which point the flows are consid-
ered to be isokinetic. Mohlenberg and Riisgard's (1978) method could
not be used satisfactorily in flume experiments because of the complicat-
ing ambient flow around the bivalve (Wildish and Saulnier 1993); direct
exhalant flow sampling (Wildish and Saulnier 1993) was substituted. The
bivalve gill retention efficiency (RE) can be directly determined from the
concentration of inhalant (Q) compared to exhalant seston (C2):
Methods of Study 51
RE = 1 - ^ - (2.13)
c1
The value of R can be calculated as in equation (2.11) and the pumping
rate, P, can be calculated by dividing R by RE, provided the estimates
are for the same periods.
Suspension feeding bioassays ideally require that seston be main-
tained at a constant concentration during the experiment. Winter (1973)
originally used a gently stirred tank of 10-L capacity containing blue
mussels. Improvements to this technique have been suggested by
Riisgard and Mohlenberg (1979) and a less expensive version proposed
by Haupt (1979). All of these methods use a photometer to monitor the
particle concentration by differences in light transmission, without dis-
tinguishing the preferred microalga from other sestonic particles offered
to the mussel. As the mussels consume the seston, the increasing light,
through a relay, switches on a dosing valve connected to a microalgal
culture for a predetermined period. This allows the culture to be added
to the tank at a rate to maintain a pre-set seston concentration. The rate
of addition of the microalgae is also a direct measure of mussel uptake
or clearance, after a suitable allowance for microalgal settling or growth
and cell division has been made in a control run with mussels absent
but for the same period as the experiment. Photometric determination
is not specific for microalgal particles and any dense particle, such
as faeces or pseudofaeces, reduces the light reaching the photoelectric
sensor.
A more specific method for microalgal cells involves fluorescence
measurements with two suitable filters, providing exciting light at 430 nm
and emitted light at 650 nm, which is suitable for continuous measure-
ment of in vivo chlorophyll concentration (Lorenzen 1966) in a continu-
ous flow fluorometer (Table 2.6; Fig. 2.15). Some limitations of the in
vivo fluorescence method should be kept in mind (Loftus and Seliger
1975), e.g. high concentration quenching or direct absorption of light by
the microalgae, when using this method. Bayne, Widdows, and Newell
(1977) were successful in using flow fluorometry to measure the feeding
rate of blue mussels fed unialgal cultures. A suitable setup for maintain-
ing a constant seston concentration with feeding bivalves present in a
flume flow is shown in Fig. 2.15. One unsatisfactory feature of this
method is that, because of the different fluorescent properties of each
microalgal species, it is not possible to maintain the strict relationship
Table 2.6. Partial list of commercial suppliers of equipment and

supplies required for flow-related biological observation or
experimentation.
Type Supplier
Temperature measurement Hewlett Packard, 6877 Goreway Drive,
Ontario, Canada L4V 1M8
Photoperiod control Suntracker, Paragon Electric Canada Ltd., 221
Evans Ave., Etobicoke, Ontario, Canada M8Z
1J5
Three-dimensional Velmex Inc., East Bloomfield, NY, 14443 USA
positioner made from
Unislide assembly parts
Spaghetti tags for identifying Floy Tag & Manufacturing Inc., 4616 Union
experimental subjects, Bay Place, N.E., Seattle, Washington, 98105
gypsum cylinders, etc. USA
Flow fluorometer, AU-005 Turner Designs, 845 W. Maude Ave.,
Sunnyvale, California, 94086 USA
Melamime formaldehyde Radiant Color, Richmond, California, 94809
particles USA
Endoscope Olympus Corp., Medical Instrument Div., 4-T
Nevada Drive, Lake Success, New York, 11042
USA
Sony video system with PEMTEK Technologies, 201 Brownlow 33B,
miniature camera Dartmouth, Nova Scotia, Canada B3B 1W2
Aerolab, 9580 Washington Blvd., Laurel, MD,
Schlieren systems 20723 USA
Yellow Springs Instrument Co. Inc., Box 279,
Dissolved oxygen probes Yellow Springs, Ohio, 45387 USA
observed between fluorescence and cell density if two or more species

are present (see also Bayne et al. 1977), particularly if the bivalve is
preferentially feeding on a particular microalga, as is likely to be the
case. A review of the in vivo fluorescence technique to measure the
clearance rates of bivalves was presented by McLatchie (1992).
Other means of controlling seston particle concentrations in this same
general way have been tried, e.g. electronic particle sizing and counting
(Klein, Breteler, and Laan 1993), and presumably flow cytometry could
also be used for this purpose. The former method is not readily adapt-
able to continuous measurement as are photometry, flow fluorometry,
Methods of Study 53
b
d
Figure 2.15 Constant seston concentration in flume flow: a, metering pump

controlled by c; b, aerated algal supply; c, personal computer controlled data
logger with switch relays; d, personal computer; e, Model AU-005 field
fluorometer (Turner Designs); f, Little Giant submersible pump; g, flume.
and cytometry (Kolber and Falkowski 1993). Flow cytometry is capable

of resolving the identity of microalgal species on the basis of particle
sizing and resolution of the fluorescence spectra of in vivo cells.
Carey (1989) employed melamine formaldehyde particles of 3-6 jum
and specific gravity = 1.4g-cm3 to study benthic animal-sediment dy-
namics (Table 2.6). The particles are treated with dyes so that they can
be determined fluorometrically. Perhaps such a technique could be
adapted by combining microparticulate, formulated diet particles
treated with dyes and determined byflowfluorometryto estimate seston
uptake or clearance rates by suspension feeding animals.
Behavioral responses and models

A central method of observing behavior in both swimming and sessile
suspension feeders uses various forms of photography. In the field, the
use of SCUBA with hand-held underwater still or video cameras has
furthered our understanding of the swimming behavior of bivalves (e.g.

Hartnoll 1967). Meyer et al. (1984) used a moored video camera in an
underwater housing to take time-lapse photographs at ~l-min intervals
of feather stars on a coral reef at depths of 3-13 m. With this means
of recording frequent observations, short-term crawling episodes,
behavioral responses to currents, and spawning events were detected.
An underwater stereo camera system mounted on a benthic support
frame was used by Wildish and Lobsiger (1987) to take time-lapse pho-
tographs of tube-living polychaetes and amphipods in soft sediments at
80 m depth (but with a rated depth to 3000 m). The camera was fitted
with a microcomputer which allowed the time interval for exposure
to be pre-set. Dolmer, Karlsson, and Svane (1994) used underwater
stereophotography to observe natural and experimentally manipulated
groups of blue mussels with respect to flow in a 9-m-depth man-made
channel.
Photographic methods are also appropriate in laboratory observations
or experiments involving behavior. Film which records the swimming
behavior of scallops can be analyzed at leisure after the experiments by
frame analysis to estimate, for example, the swimming speed (Moore and
Trueman 1971; Dads well and Weihs 1990). Underwater video footage
can be analyzed with a video cassette recorder able to analyze frame by
frame. Specialized equipment for computerized analysis of multiple-
channel events recorded on video tapes is available (Krauss, Morrel-
Samuels, and Hochberg 1988) with suitable software (Noldus 1991) to
analyze the data obtained.
Video viewing through the acrylic plastic walls of a flume during
feeding experiments of scallops showed that the exhalant opening area
was controlled by the ambient velocities (Wildish and Saulnier 1993). A
similar technique was used by Dolmer et al. (1994) to measure the valve
gape of blue mussels exposed toflow/no-flowconditions.
Flow patterns over individual scallops were investigated by using mod-
els consisting of cleaned valves filled with plaster of paris (Thorburn and
Gruffyd 1979). Flow lines were visualized by video analysis offinehydro-
gen gas bubbles or condensed milk diluted with seawater. To improve
contrast, white light illumination was provided against a black back-
ground, including painting the valves black. This dark field illumination
technique was also used by Leonard (1989) to observe tube feet feeding
by crinoids.
A trend in the study of suspension feeding has been to observe at
higher magnification in the hope of seeing the detailed mechanisms of
Methods of Study 55
this process. Strickler (1985) suggested the use of a Schlieren optical

pathway and a collimated light beam focused by a condensing lens which
makes seston particles appear white on a black background. Trager et al.
(1990) used a similar system with magnification of 10-50 times to obtain
video images of particles and barnacle feeding appendages. The
endoscope, developed for medical purposes, has been used to observe
some of the details of the suspension-feeding mechanisms, e.g. by Hunt
and Alexander (1991) in barnacles and by Ward, MacDonald, and
Thompson (1993) in bivalves.
Physical models of suspension feeders can be used in flume experi-
ments where all environmental conditions can be closely controlled and
the experiments repeated without danger of tiring the subjects. One
disadvantage of using models is that any subtle behavioral responses to
flow may be missed, so the method must be used judiciously and in
conjunction with observations of natural behavior. A good example of an
appropriate use of this method is the right cylindrical models of tube-
living epifauna studied by Eckman and No well (1984). These authors
were interested in testing tube height and diameter as they affected flow
patterns and erosion-sedimentation phenomena close to the tubes.
Simple models of brachiopods based upon propped-open empty shells
and with a catheter tube containing rhodamine dye were used by
LaBarbera (1977) to visualize flow patterns in and around the shells. A
similar technique was used by Johnson (1988) to assess the local flow
around simple models of phoronids at different ambient velocities. Bi-
valve siphon models with the siphon apertures at a fixed opening diam-
eter and adjustable height above the sediment-water interface were
prepared by Monismith et al. (1990). Siphon jetting and mixing near the
bivalve were visualized by video recording of a fluorescent dye pumped
continuously into the exhalant siphon. A limitation of this technique, as
a behavioral test, is that the siphons could not be adjusted to open and
close partially or fully as occurs in real life in many bivalves.
In studying the factors involved in the dislodgement of epifaunal sus-
pension feeders by flowing or wind-induced seawater movement, life-
sized models can be used in suitable flume experiments (e.g. Gruffyd
1976; Millward and Whyte 1992; Wildish and Saulnier 1993). For
bivalves, the dried valves can be used to prepare suitable models or
complete new models cast with plaster of paris. Millward and Whyte
(1992) mounted scallop models on an adjustable arm, while Gruffyd
(1976) suspended scallop models directly in flow by means of three
threads. Wildish and Saulnier (1993) first determined the buoyant weight
of live scallops by weighing a suspended scallop in motionless seawater

from the arm of a balance (see Lowndes 1942). Model scallops were then
prepared by the addition of plasticine to cleaned valves so that the model
had the same buoyant weight as the live scallop.
As pointed out by Vogel (1981), it is possible to study hydrodynamic
phenomena in air, rather than in seawater, since at the same tempera-
ture, air is 15 times more kinematically viscous. This means that for an
object of the same dimensions, air flows must be 15 times greater than
seawater velocity to achieve the same Reynolds number. Vogel (1985)
employed wind tunnel experiments to show that flow-induced forces
assisted the elasticity of the hinge ligament of the scallop Argopecten
irradians in reopening the left valve during coordinated swimming
movements.
Reef metabolism and production

Because ecology is such a vast subject area, we have concentrated only
on methods to measure oxygen demand by reefs and to calculate pro-
duction potentials for benthic populations or communities (see Chaps. 7
and 8).
The mechanisms involved in oxygen exchanges across the sediment-
water interface, specifically that of primarily suspension-feeding
populations such as a bivalve reef, include diffusion from seawater into
sediments caused by the physical gradient, increased exchange of
seawater with sediments due to the bioturbation activities of infaunal
burrowers, as well as photosynthetic production of molecular oxygen by
surface-living diatoms. From within the sediment biological oxygen
demands from aerobic microorganisms, meiofaunal and macrofaunal
respiration, and chemical oxygen demands due to reduced chemical
compounds produced by anaerobic microorganisms all result in utiliza-
tion of dissolved oxygen in sediment pore water and overlying seawater.
Thus, in measuring dissolved oxygen fluxes in the seawater above a reef,
we are integrating the resultant of all these processes.
The effect of water movement and bottom surface roughness on the
flux of dissolved oxygen across the sediment-water interface was appre-
ciated by the benthic ecologists whose work is discussed in Chapter 8 in
the section, Reef metabolism. Their results should be considered as
crude measures of reef oxygen demands. A notable problem is that the
sampling points chosen are arbitrary and not based on the particular
hydrodynamic conditions prevailing at a specific site. In the case of flow
Methods of Study 57
canalization caused by the benthic ecosystem tunnel, advective water

motion from above is prevented from mixing with canalized seawater.
Recently, Dade (1993) has introduced suitable theory for estimating the
distribution and flux of dissolved substances such as oxygen within the
benthic boundary layer.
Many of the commonly used methods of measuring sediment oxygen
demand (Bowman and Delfino 1980) are not adaptable for obtaining
accurate field estimates of reef oxygen demands influenced by ambient
water movements. The most commonly used enclosed chamber tech-
nique - a cylinder pushed into the sediment to enclose a known volume
of seawater, <1 - 38 L (Boynton et al. 1981) - cannot readily be adapted
to simulate ambient water movement. It has been suggested that the
passage of waves in nearshore waters creates differential hydrostatic
pressures in the sediment, which, in turn, cause pumping of seawater in
and out of the sediment. This phenomenon has been experimentally
examined by Malan and McLachlan (1991). They provided a small,
unstirred chamber with a flexible membrane (rendered impermeable by
bonding sheets of polyethylene and nylon) sealed on the open top of the
device. This allowed wave-induced interstitial seawater to enter the
chamber through the sediment as a result of compensatory movements
of the membrane. A comparison of unstirred flexible-topped and solid-
topped chambers deployed in the field showed a 73% increase in oxygen
uptake in the former, supporting the view that the passage of waves
increased dissolved oxygen fluxes.
We consider that two sensing devices - commercially available dis-
solved oxygen probes (Table 2.6) and oxygen microelectrodes
(Revsbech et al. 1980; Revsbech and J0rgensen 1983) - could be used to
provide accurate estimates of oxygen fluxes from suspension-feeding
reefs. Advantages of the microelectrode are that the sensing tips are
small, 5- to 10-|um diameter, and that 50- to 100-jim sections of the
sediment profile can be resolved. From dissolved oxygen profiles within
the sediment and application of Fick's law of diffusion, thefluxof oxygen
at the sediment-water interface can be determined (Revsbech and
J0rgensen 1983). These results, coupled with benthic boundary layer
theory and detailed contemporary observations by a physical oceanogra-
pher, could help to determine the whole reef flux. This method is limited
by the necessary assumption that only Fickian diffusion is involved. This
assumption is reasonable in fine sediments of low permeability, but in
sandy sediments of higher permeability it is not. For the latter, the
assumption is invalid because of the passage of waves (Malan and
Table 2.7. Conversions and predictions of potential benthic production.
Conversions
Wet weight, g, to kcal, multiply by:
1.0 (Steele 1974)
0.6 x fresh weight (Mills and Fournier 1979)
0.5 (Crisp 1975)
kcal to kJ, multiply by:
4.186
kJ to g carbon, divide by:
50
Predictions
Production/biomass, P:B, ratios may be determined by multiplying W, wet
body mass at maturity, g, or lifespan in years, based on the following
equations:
Applicability Equation Reference
Macroafauna + meiofauna P:B = 0.971 W^167 Schwinghamer (1981)
Macrofauna P :B = 0.563 W'0302 Schwinghamer (1981)
Invertebrates, 5-20C P:B= 0.640W0310 Banse and Mosher (1980)
Invertebrates, less insects P:B = 0.6UW0390 Banse and Mosher (1980)
Macrofauna P:B = 4.571 lifespan Robertson (1979)
Production of suspension feeders, P(g-rar2'yf1), from mean tidal velocity,

U(m-s~l), in the Bay of Fundy:
log10P = 5.862*7- 1.155 (Wildish and Kristmanson 1979)
McLachlan 1991) in sediments and/or presence of roughness elements of

biological or mechanical origin (Huettel and Gust 1992) at the sediment
surface, which destroys the well defined laminar sublayer.
Production estimates of benthic communities are required in ecosys-
tem analysis, e.g. in pelagic-benthic coupling studies. The units in which
such estimates are made are convertible (Table 2.7). The crudest esti-
mate of benthic production is made from quantitative grab or core
sampling data as biomass (g-m~2). Empirically derived relationships
based on the annual turnover ratio, P:B, from a series of field studies in
temperate climates (Table 2.7) are based on either lifespan (Robertson
1979) or mature biomass (Banse and Mosher 1980; Schwinghamer 1981).
These relationships can be used to compute crude annual production
estimates by multiplying P:B and the observed mean annual biomass for
a species.
Methods of Study 59
It is possible to produce a more accurate measure of benthic produc-

tion by obtaining detailed information for each cohort of each species in
the population. This method requires periodic estimates of population
density and estimation of mean growth rates throughout the life of the
cohort. Cohort production is determined by summing the time period
estimates by one of two computation methods - cohort summation of
losses or the Allen curve method (Crisp 1984) - or, if it is impossible to
distinguish the cohorts, a size-frequency method (Wildish and Peer
1981). Because of extensive work required to complete the measure-
ments and analysis even for one species population, it is usually only
possible to complete it for a few of the most dominant ones. Probably
because of the time costs involved, only a few studies using this method-
ology have been completed (e.g. Buchanan and Warwick 1974). Because
of the often near-monoculture conditions of a bivalve reef, such a
method is well suited to it.
Ecosystem-level estimates of benthic production require a detailed
and accurate map of the predominant benthic community. The develop-
ment and use of digital multibeam bathymetric systems by surficial ge-
ologists to provide detailed and rapidly obtained maps of the seabed (see
Courtney and Fader 1994) may prove to be important in indicating
benthic communities and, hence, lead to more accurate charts of benthic
production than hitherto available. Commercially available systems
which adapt ships' echo sounders to produce sedimentary roughness and
hardness estimates and plot it on a chart have also been used to identify
epibenthic communities, e.g. Magorrian, Service, and Clarke (1995).
Some attempts have been made to determine the limiting factors
involved in controlling suspension feeder-dominated communities in
natural conditions (e.g. Wildish and Kristmanson 1979) or in bivalve
suspension culture (e.g. Emerson et al. 1994). Thus, in the lower Bay of
Fundy, a simple linear relationship (Table 2.7) was found between mean
tidal velocity and production calculated on the basis of the Robertson
lifespan method. Further field studies in the upper Bay of Fundy
(Wildish et al. 1986), where tidal velocities and wind-wave effects were
more energetic, showed that tidal velocities did not provide a good
predictor of production, suggesting that the relationship was not univer-
sal. Emerson et al. (1994) measured monthly growth rates of the local
scallop Placopecten magellanicus and found that up to 68% of growth
variation could be explained by temperature and seston quality meas-
ured at biweekly intervals. This suggests that a crude estimate of produc-
tivity at a specified site might be predicted with just these two environ-
mental variables, although uncertainties remain as to how to measure

seston quality.
References
Amos, C. L., J. Grand, G. R. Daborn, and K. Black. 1992. Sea carousel: a
benthic annular flume. Estuar. Coast. Shelf Sci. 34: 557-578.
Amouroux, J. M., M. Revault d'Allonnes, and C. Rouant. 1975. Sur la mesure
directe du debit de filtration chez les mollusques lamellibranches. Vie
Milieu (Ser. 2B) 25: 339-346.
Asmus, H., R. M. Asmus, T. C. Prins, N. Dankers, G. Frances, B. Maas, and
K. Reise. 1992. Benthic-pelagic flux rates on mussel beds: tunnel and tidal
flume methodology compared. Helgolander Meeresunters. 46: 341-361.
Asmus, R. M., and H. Asmus. 1991. Mussel beds: limiting or promoting
phytoplankton? J. Exp. Mar. Biol. Ecol. 148: 215-232.
Banse, K., and S. Mosher. 1980. Adult body mass and annual production/
biomass relationships of field populations. Ecol. Monogr. 50: 355-379.
Bayne, B. L., J. Widdows, and R. I. E. Newell. 1977. Physiological
measurements on estuarine bivalve molluscs in the field, p. 57-68. In B.F.
Keegan, P. 6. Ceidigh, and P. J. S. Boaden (ed.) Biology of benthic
organisms. Pergamon, Oxford.
Bell, E. C, and M. W. Denny. 1994. Quantifying "wave exposure": a simple
device for recording maximum velocity and results of its use at several
field sites. J. Exp. Mar. Biol. Ecol. 181: 9-29.
Best, B. A. 1988. Passive suspension feeding in a sea pen: effects of ambient
flow on volume flow rate and filtering efficiency. Biol. Bull. 175: 332-342.
Boucher, N., D. Vaulot, and F. Partensky. 1991. Flow cytometric
determination of phytoplankton DNA in cultures and oceanic
populations. Mar. Ecol. Prog. Ser. 71: 75-84.
Bowman, G. T., and J. J. Delfino. 1980. Sediment oxygen demand techniques:
a review and comparison of laboratory and in situ systems. Water Res. 14:
491^99.
Boynton, W. R., W. M. Kemp, C. G. Osborne, K. R. Kaumeyer, and M. C.
Jenkins. 1981. Influence of water circulation rate on in situ measurements
of benthic community respiration. Mar. Biol. 65: 185-190.
Bradshaw, P. 1964. Experimental fluid mechanics. Pergamon Press, Oxford.
Buchanan, J. B., and R. M. Warwick. 1974. An estimate of benthic
macrofaunal production in the offshore mud of the Northumberland
coast. J. Mar. Biol. Assoc. U.K. 54: 197-222.
Butman, C. A. 1986. Sediment trap biases in turbulent flows: results from a
laboratory flume study. J. Mar. Res. 44: 645-693.
Butman, C. A., J. P. Grassle, and C. M. Webb. 1988. Substrate choices made
by marine larvae settling in still water and in aflumeflow.Nature 333:
771-773.
Carey, D. A. 1983. Particle resuspension in the benthic boundary layer by a
tube-building polychaete. Can. J. Fish. Aquat. Sci. 40: 301-308.
1989. Fluorometric detection of tracer particles used to study animal-
particle dynamics. Limnol. Oceanogr. 34: 630-634.
Charriaud, E. 1982. Direct measurements of velocity profiles and fluxes at the
cloacal siphon of the ascidian Ascidiella aspersa. Mar. Biol. 70: 35-40.
Methods of Study 61
Charters, A. C, and S. M. Anderson. 1980. A low velocity water tunnel for

biological research. J. Hydronaut. 14: 3-4.
Chia, F., J. Buckland-Nicks, and C. M. Young. 1984. Locomotion of
invertebrate larvae: a review. Can. J. Zool. 62: 1205-1222.
Coughlan, J. 1969. The estimation of filtering rate from the clearance of
suspensions. Mar. Biol. 2: 356-358.
Courtney, R. C, and G. B. J. Fader. 1994. A new understanding of the ocean
floor through multibeam mapping, p. 9-14. In A. Fiander (ed.) Science
review 1992 and 1993. Department of Fisheries and Oceans, Bedford
Institute of Oceanography, Dartmouth, Nova Scotia, Canada.
Crisp, D. J. 1975. Secondary productivity in the sea, p. 71-89. Productivity of
world ecosystems. National Academy of Sciences, Washington, D.C.
1984. Energy flow measurements, p. 284-372. In N. A. Holmes, and A. D.
Mclntyre (ed.) Methods for the study of marine benthos. IBP Handbook
16. 2nd ed. Blackwell, Oxford.
Dade, W. B. 1993. Near-bed turbulence and hydrodynamic control of
diffusional mass transfer at the sea floor. Limnol. Oceanogr. 38: 52-69.
Dadswell, M. J., and D. Weihs. 1990. Size-related hydrodynamic characteristics
of the giant scallop, Placopecten magellanicus (Bivalvia: Pectinidae). Can.
J. Zool. 68: 778-785.
Dame, R. F., R. G. Zingmark, and E. Haskin. 1984. Oyster reefs as processors
of estuarine materials. J. Exp. Mar. Biol. Ecol. 83: 239-267.
Demers, S. (Editor). 1991. Particle Analysis in Oceanography. NATO ASI
Series G. Vol. 27. Springer-Verlag, Berlin.
Denny, M. W. 1982. Forces on intertidal organisms due to breaking ocean
waves: design and application of a telemetry system. Limnol. Oceanogr.
27: 178-183.
1985. Wave forces on intertidal organisms: a case study. Limnol. Oceanogr.
30: 1171-1187.
1988. Biology and the mechanisms of the wave-swept environment.
Dolmer, P., M. Karlsson, and I. Svane. 1994. A test of the rheotactic
behaviour of the blue mussel Mytilus edulis L. Phuket. Mar. Biol. Cent.
Spec. Publ. 13: 177-184.
Doty, M. S. 1971. Measurements of water movement in reference to benthic
algal growth. Bot. Mar. 14: 32-35.
Drinnan, R. E. 1964. An apparatus for recording the water-pumping
behaviour of lamellibranchs. Neth. J. Sea Res. 2: 223-232.
Eckman, J. E. 1979. Small-scale patterns and processes in a soft-substratum,
intertidal community. J. Mar. Res. 37: 437-457.
Eckman, J. E., and A. R. M. Nowell. 1984. Boundary skin friction and
sediment transport about an animal-tube mimic. Sedimentology 31: 851-
862.
Emerson, C. W., J. Grant, A. Mallet, and C. Carver. 1994. Growth and
survival of sea scallops Placopecten magellanicus: effect of culture depth.
Mar. Ecol. Prog. Ser. 108: 119-132.
Emig, C. C, and F. Becherini. 1970. Influence des courants sur l'ethologie
alimentaire des phoronidiens. Etude par series de photographies
cycliques. Mar. Biol. 5: 239-244.
Emlet, R. B. 1990. Flow fields around ciliated larvae: effects of natural and
artificial tethers. Mar. Ecol. Prog. Ser. 63: 211-225.
Famme, P., and L. H. Kofoed. 1983. Shunt water flow through the mantle
cavity in Mytilus edulis (L.) and its influence on particle retention. Mar.
Biol. Lett. 4: 207-218.
Famme, P., R. U. Riisgard, and C. B. J0rgensen. 1986. On direct
measurements of pumping rates in the mussel, Mytilus edulis. Mar. Biol.
92: 323-327.
Foster-Smith, R. L. 1976. Pressures generated by the pumping mechanism of
some ciliary filter feeders. J. Exp. Mar. Biol. Ecol. 25: 199-206.
Gallagher, S. M. 1988. Visual observations of particle manipulation during
feeding in larvae of a bivalve mollusc. Bull. Mar. Sci. 43: 344-365.
Galstoff, P. S. 1928. Experimental study of the function of the oyster gills and
its bearing on the problems of oyster culture and sanitary control of the
oyster industry. Bull. U.S. Bur. Fish. 44, Doc. 1035: 1-39.
Grassle, J. P., P. V. R. Snelgrove, and C. A. Butman. 1992. Larval habitat
choice in still water and flume flows by the opportunistic bivalve Mulinia
lateralis. Neth. J. Sea Res. 30: 33-44.
Grizzle, R. E., R. Langan, and W. H. Howell. 1992. Growth responses of
suspension-feeding bivalve molluscs to changes in water flow: differences
between siphonate and nonsiphonate taxa. J. Exp. Mar. Biol. Ecol. 162:
213-228.
Gruffyd, L. D. 1976. Swimming in Chlamys islandica in relation to current
speed and an investigation of hydrodynamic lift in this and other scallops.
Nor. J. Zool. 24: 365-378.
Gust, G. 1982. Tools for oceanic small-scale high-frequency flows: metal clad
hot wires. J. Geophys. Res. 87: 445-447.
Hannan, C. A. 1984. Planktonic larvae may act like passive particles in
turbulent near-bottom flows. Limnol. Oceanogr. 29: 1108-1115.
Hargrave, B. T., and W. M. Burns. 1979. Assessment of sediment trap
collection efficiency. Limnol. Oceanogr. 24: 1124-1136.
Hargrave, B., G. Siddal, G. Steeves, and G. Await. 1994. A current-activated
sediment trap. Limnol. Oceanogr. 39: 383-390.
Hartnoll, R. G. 1967. An investigation of the movement of the scallop. Helg.
Wiss. Meeresunters. 15: 523-533.
Haupt, K. 1979. A simple device to control concentrations of food algae in
feeding experiments with filter feeding organisms. Veroff. Inst.
Meeresforsch, Bremerhaven 17: 241-244.
Hildreth, D. I., and D. J. Crisp. 1976. A corrected formula for calculation of
filtration rate of bivalve molluscs in an experimental flowing system. J.
Mar. Biol. Assoc. U.K. 56: 111-120.
Hinds, W. C. 1982. Aerosol technology. John Wiley & Sons, New York.
Hobbie, J. E., R. Daley, and S. Jasper. 1977. Use of nucleopore filters for
counting bacteria by fluorescence microscopy. Appl. Environ. Microbiol.
33: 1225-1228.
Holm-Hansen, O., and C. R. Booth. 1966. The measurement of adenosine
triphosphate in the ocean and its ecological significance. Limnol.
Oceanogr. 11: 510-519.
Huettel, M., and G. Gust. 1992. Impact of bioroughness on interfacial solute
exchange in permeable sediments. Mar. Ecol. Prog. Ser. 89: 253-267.
Hunt, M. J., and C. G. Alexander. 1991. Feeding mechanisms in the barnacle
Tetraclita squamosa (Brugiere). J. Exp. Mar. Biol. Ecol. 154: 1-28.
Hunter, T. 1989. Suspension feeding in oscillating flow: the effect of colony
morphology and flow regime on plankton capture by the hydroid Obelia
longissima. Biol. Bull. 176: 41-49.
Methods of Study 63
Johnson, A. S. 1988. Hydrodynamic study of the functional morphology of the

benthic suspension feeder Phoronopsis viridis (Phoronida). Mar. Biol.
100: 117-126.
Jokiel, P. L., and J. I. Morrissey. 1993. Water motion on coral reefs: evaluation
of the "clod card" technique. Mar. Ecol. Prog. Ser. 93: 175-181.
Jones, D. A., J. G. Mumford, and P. A. Gabbott. 1974. Micro capsules as
artificial food particles for aquatic filter feeders. Nature 247: 233-235.
Jones, H. D., and J. R. Allen. 1986. Inhalant and exhalant pressures in Mytilus
edulis L. and Cerastoderma edule (L.). J. Exp. Mar. Biol. Ecol. 98: 231-
240.
J0rgensen, C. B. 1990. Bivalve filter feeding: hydrodynamics, bioenergetics,
physiology and ecology. Olsen and Olsen, Fredensborg, Denmark.
Judge, M. L., L. D. Coen, and K. L. Heck, Jr. 1992. The effect of long-term
alteration of in situ currents on the growth of Mercenaria mercenaria in
the northern Gulf of Mexico. Limnol. Oceanogr. 37: 1550-1559.
1993. Does Mercenaria mercenaria encounter elevated food levels in
seagrass beds? Results from a novel technique to collect suspended food
resources. Mar. Ecol. Prog. Ser. 92: 141-150.
Klein Breteler, W. C. M., and M. Laan. 1993. An apparatus for automatic
counting and controlling density of pelagic food particles in culture of
marine organisms. Mar. Biol. 116: 169-11 A.
Koehl, A. R. M., and R. S. Alberte. 1988. Flow, flapping and photosynthesis of
Nereocystis luetkeana: a functional comparison of undulate and flat blade
morphologies. Mar. Biol. 99: 435-444.
Kolber, Z., and P. G. Falkowski. 1993. Use of active fluorescence to estimate
phytoplankton photosynthesis in situ. Limnol. Oceanogr. 38: 1646-
1665.
Koochesfahani, M. M., and P. E. Dimolakis. 1985. Laser-induced fluorescence
measurements of mixed fluid concentration in a liquid plane shear layer.
J. Am. Inst. Aeronaut. Astrophys. 23: 1700-1707.
Krauss, R. M, P. Morrel-Samuels, and J. Hochberg. 1988. VIDEOLOGGER:
a computerized multichannel event recorder for analyzing video tapes.
Behav. Res. Methods Instrum. Comput. 20: 37-40.
LaBarbera, M. 1977. Brachiopod orientation to water movement. I. Theory,
laboratory behaviour and field orientations. Paleobiology 3: 270-287.
Langdon, C. J., D. M. Levine, and D. A. Jones. 1985. Microparticulate feeds
for marine suspension feeders. J. Microencapsulation 2: 1-11.
Leonard, A. B. 1989. Functional response in Antedon mediterranea (Lamarck)
(Echinodermata: Crinoidea): the interaction of prey concentration and
current velocity on a passive suspension feeder. J. Exp. Mar. Biol. Ecol.
127: 81-103.
Leversee, G. J. 1976. Flow and feeding in fan-shaped colonies of the gorgonian
coral, Leptogorgia. Biol. Bull. 151: 344-356.
Lindegarth, M., P. R. Jonsson, and C. Andre. 1991. Fluorescent microparticles:
a new way of visualizing sedimentation and larval settlement. Limnol.
Oceanogr. 36: 1471-1475.
Lofquist, K. E. B. 1977. A positive displacement oscillatory water tunnel. Misc.
Rep. 77-1, U.S. Army Corps of Engineers, CERC, Fort Belvoir, Va.
Loftus, M. E., and H. H. Seliger. 1975. Some limitations of the in vivo
fluorescence technique. Chesapeake Sci. 16: 79-92.
Lorenzen, C. J. 1966. A method for the continuous measurement of in vivo
chlorophyll concentration. Deep Sea Res. 13: 223-227.
Lowndes, A. G. 1942. The displacement method of weighing living aquatic

organisms. J. Mar. Biol. Assoc. U.K. 25: 555-574.
Luckenbach, M. W. 1986. Sediment stability around animal tubes: the roles of
hydrodynamic processes and biotic activity. Limnol. Oceanogr. 31: 779-
787.
Lui, S. T., and G. H. Nancollas. 1971. The kinetics of dissolution of calcium
sulphate dihydrate. J. Inorg. Nucl. Chem. 33: 2311-2316.
Magorrian, B. H., M. Service, and W. Clarke. 1995. An acoustic bottom
classification survey of Strangford Lough, Northern Ireland. J. Mar. Biol.
Assoc. U.K. 75: 987-992.
Malan, D. E., and A. Mclachlan. 1991. In situ benthic oxygen fluxes in a
nearshore coastal marine system: a new approach to quantify the effect of
wave action. Mar. Ecol. Prog. Ser. 73: 69-81.
McLatchie, S. 1992. Time series measurement of grazing rates of zooplankton
and bivalves. J. Plankton Res. 14: 183-200.
Meyer, D. L., C. A. Lattaye, N. D. Holland, A. C. Arneson, and J. R.
Strickland. 1984. Time-lapse cinematography of feather stars
(Echinodermata: Crinoidea) on the Great Barrier Reef, Australia:
demonstrations of posture changes, locomotion, spawning and possible
predation by fish. Mar. Biol. 78: 179-184.
Miller, D. C, M. J. Bock, and E. J. Turner. 1992. Deposit and suspension
feeding in oscillatory flows and sediment fluxes. J. Mar. Res. 50: 489-520.
Mills, E. L., and R. O. Fournier. 1979. Fish production and the marine
ecosystems of the Scotian Shelf, eastern Canada. Mar. Biol. 54: 101-108.
Millward, A., and M. A. Whyte. 1992. The hydrodynamic characteristics of six
scallops of the super family Pectinacea, class Bivalvia. J. Zool. London
227: 547-566.
Minchinton, T. E., and R. E. Scheibling. 1991. The influence of larval supply
and settlement on the population structure of barnacles. Ecology 72:
1867-1879.
Miron, G., P. Pelletier, and E. Bourget. 1995. Optimizing the design of giant
scallop (Placopecten magellanicus) spat collectors: flume experiments.
Mar. Biol. 123: 285-292.
M0hlenberg, F., and H. U. Riisgard. 1978. Efficiency of particle retention in 13
species of suspension feeding bivalves. Ophelia 17: 239-246.
Monismith, S. G., J. R. Koseff, J. K. Thompson, C. A. O'Riordan, and H. M.
Nepf. 1990. A study of model bivalve siphonal currents. Limnol.
Oceanogr. 35: 680-696.
Moore, J. D., and E. R. Trueman. 1971. Swimming of the scallop Chlamys
opercularis (L.). J. Exp. Mar. Biol. Ecol. 6: 179-185.
Mullineaux, L. S., and C. A. Butman. 1990. Recruitment of encrusting benthic
invertebrates in boundary-layer flows: a deep-water experiment on Cross
Seamount. Limnol. Oceanogr. 35: 409-423.
Mullineaux, L. S., and E. D. Garland. 1993. Larval recruitment in response to
manipulated field flows. Mar. Biol. 116: 667-684.
Muschenheim, D. K. 1987. The dynamics of near-bed seston flux and
suspension-feeding benthos. J. Mar. Res. 45: 473-496.
Muschenheim, D. K., J. Grant, and E. L. Mills. 1986. Flumes for benthic
ecologists: theory, construction and practice. Mar. Ecol. Prog. Ser. 28:
185-196.
Navarro, J. M., and R. J. Thompson. 1995. Seasonal fluctuations in the size
spectra, biochemical composition and nutritive value of the seston
Methods of Study 65
available to a suspension-feeding bivalve in a subarctic environment. Mar.

Ecol. Prog. Ser. 125: 95-106.
Nixon, S. W., C. A. Oviatt, C. Rogers, and K. Taylor. 1971. Mass and
metabolism of a mussel bed. Oecologia 8: 21-30.
Noldus, L. P. J. J. 1991. The Observer: a software system for collection and
analysis of observational data. Behav. Res. Methods Instrum. Comput. 23:
415-429.
No well, A. R. M., and P. A. Jumars. 1987. Flumes: theoretical and
experimental considerations for simulation of benthic environments.
Oceanogr. Mar. Biol. Annu. Rev. 25: 91-112.
Okamura, B. 1984. The effects of ambient flow velocity, colony size, and
upstream colonies on the feeding success of Bryozoa. I. Bugula stolonifera
Ryland, an arborescent species. J. Exp. Mar. Biol. Ecol. 83: 179-193.
Olson, R. J., and E. R. Zettler. 1995. Potential of flow cytometry for "pump
and probe" fluorescence measurements of phytoplankton photosynthetic
characteristics. Limnol. Oceanogr. 40: 816-820.
O'Riordan, C. A., S. G. Monismith, and J. R. Koseff. 1993. A study of
concentration boundary layer formation over a bed of model bivalves.
Limnol. Oceanogr. 38: 1712-1729.
Patterson, M. R. 1991. The effects of flow on polyp-level prey capture in an
octocoral, Alcyonium siderium. Biol. Bull. 180: 93-102.
Penner, S. S., and T. Jerskey. 1973. Use of lasers for local measurement of
velocity components, species densities and temperatures. Annu. Rev.
Fluid Mech. 5: 9-30.
Petticrew, E. L., and J. Kalff. 1991. Calibration of a gypsum source for
freshwater flow measurements. Can. J. Fish. Aquat. Sci. 48: 1244-1249.
Revsbech, N. P., and B. B. J0rgensen. 1983. Photosynthesis of benthic
microflora measured with high spatial resolution by the oxygen
microprofile method: capabilities and limitations of the method. Limnol.
Oceanogr. 28: 749-756.
Revsbech, N. P., J. Sorensen, T. H. Blackburn, and J. P. Lomholt. 1980.
Distribution of oxygen in marine sediments measured with
microelectrodes. Limnol. Oceanogr. 25: 403-411.
Riisgard, H. U. 1977. On measurements of the filtration rates of suspension
feeding bivalves in a flow system. Ophelia 16: 167-173.
Riisgard, H. U., and F. M0hlenberg. 1979. An improved automatic recording
apparatus for determining the filtration rate of Mytilus edulis as a function
of size and algal concentration. Mar. Biol. 52: 61-67.
Robertson, A. I. 1979. The relationship between annual production: biomass
ratios and lifespans for marine macrobenthos. Oecologia 38: 193-202.
Saunders, H. E., and C. W. Hubbard. 1944. The circulating water channel of
the David W. Taylor model basin. Soc. Naval Architects Mar. Engin.
Trans. 52: 325-364.
Schlichting, H. 1979. Boundary-layer theory. 7th ed. McGraw-Hill, New York.
Schraub, F. A., S. J. Kline, J. Henry, P. W. Runstadler, and A. Littell. 1965.
Use of hydrogen bubbles for quantitative determination of time-
dependent velocity fields in low speed water flows. J. Basic Eng. 87: 429-
444.
Schwinghamer, P. 1981. Characteristic size distributions of integral benthic
communities. Can. J. Fish. Aquat. Sci. 38: 1255-1263.
Scoffin, T. P. 1968. An underwater flume. J. Sediment. Petrol. 38: 244-247.
Shumway, S. E., T. L. Cucci, R. C. Newell, and C. M. Yentsch. 1985. Particle
selection, ingestion, and absorption in filter-feeding bivalves. J. Exp. Mar.

Biol. Ecol. 91: 77-92.
Snelgrove, P. V. R., C. A. Butman, and J. F. Grassle. 1995. Potential flow
artifacts associated with benthic experimental gear: deep-sea mud box
examples. J. Mar. Res. 53: 821-845.
Sorensen, R. M. 1993. Basic wave mechanisms for coastal and ocean
engineers. John Wiley & Sons, New York.
Southgate, P. C, P. S. Lee, and J. A. Nell. 1992. Preliminary assessment of a
microencapsulated diet for larval culture of the Sydney rock oyster
Saccostrea commercialis (Iredale and Roughly). Aquaculture 105: 345-
352.
Sponaugle, S., and LaBarbera, M. 1991. Drag-induced deformation: a
functional feeding strategy in two species of gorgonians. J. Exp. Mar. Biol.
Ecol. 148: 121-134.
Steele, J. H. 1974. The structure of marine ecosystems. Harvard University
Press, Cambridge, Mass.
Strathman, R. R., T. L. Jahn, and J. R. C. Fontsia. 1972. Suspension feeding by
marine invertebrate larvae: clearance of particles by ciliated bands of a
rotifer, pluteus, and trocophore. Biol. Bull. 142: 508-519.
Strathman, R. R., and E. Leise. 1979. On feeding mechanisms and clearance
rates of molluscan veligers. Biol. Bull. 157: 524-535.
Strickland, J. D. H., and T. R. Parsons. 1968. A practical handbook of
seawater analysis. Bull. Fish. Res. Board Can. 167: 311.
Strickler, J. R. 1985. Feeding currents in calanoid copepods: two new
hypotheses. Symp. Soc. Exp. Biol. 39: 459-485.
Sulkin, S. D. 1990. Larval orientation mechanisms: the power of controlled
experiments. Ophelia 32: 49-62.
Svoboda, A. 1970. Simulation of oscillating water movement in the laboratory
for cultivation of shallow water sedentary organisms. Helgol. Wiss.
Meeresunters. 20: 676-684.
Taghon, G. L., A. R. M. Nowell, and P. A. Jumars. 1984. Transport and
breakdown of fecal pellets: biological and sedimentological consequences.
Thompson, T. L., and E. P. Glenn. 1994. Plaster standards to measure water
motion. Limnol. Oceanogr. 39: 1768-1779.
Thorburn, I. W., and L. D. Gruffyd. 1979. Studies of the behaviour of the
scallop Chlamys opercularis (L.) and its shell inflowingseawater. J. Mar.
Biol. Assoc. U.K. 59: 1003-1023.
Topping, J. 1957. Errors of observation and their treatment. Chapman & Hall,
London.
Trager, G. C, J.-S. Hwang, and J. R. Strickler. 1990. Barnacle suspension-
feeding in variable flow. Mar. Biol. 105: 117-129.
Turner, E. J., and D. C. Miller. 1991. Behaviour of a passive suspension feeder
{Spiochaetopterus oculatus) (Webster) under oscillatory flow. J. Exp. Mar.
Biol. Ecol. 149: 123-137.
Vogel, S. 1981. Life in moving fluids: the physical biology of flow. Willard
1985. Flow-assisted shell reopening in swimming scallops. Biol. Bull. 169:
624-630.
1994. Life in moving fluids: the physical biology of flow. 2nd rev. ed.
Methods of Study 67
Vogel, S., and N. Feder. 1966. Visualization of low-speed flow using suspended
particles. Nature 209: 186-187.
Vogel, S., and M. LaBarbera. 1978. Simple flow tanks for research and
teaching. Bioscience 28: 638-643.
Wainwright, S. C. 1990. Sediment-to-water fluxes of particulate material and
microbes by resuspension and their contribution to the planktonic food
web. Mar. Ecol. Prog. Ser. 62: 271-281.
Walne, P. R. 1972. The influence of current speed, body size and temperature
on the filtration rate of five species of bivalves. J. Mar. Biol. Assoc. U.K.
52: 345-374.
Ward, J. E., B. A. MacDonald, and R. J. Thompson. 1993. Mechanisms of
suspension feeding in bivalves: resolution of current controversies by
means of endoscopy. Limnol. Oceanogr. 38: 265-272.
Wildish, D. J., and D. D. Kristmanson. 1979. Tidal energy and sublittoral
macrobenthic animals in estuaries. J. Fish. Res. Board Can. 36: 1197-
1206.
1984. Importance to mussels of the benthic boundary layer. Can. J. Fish.
Aquat. Sci. 41: 1618-1625.
1985. Control of suspension-feeding bivalve production by current speed.
Helgol. Wiss. Meeresunters. 39: 237-243.
1988. Growth response of giant scallops to periodicity of flow. Mar. Ecol.
Prog. Ser. 42: 163-169.
Wildish, D. J., and U. Lobsiger. 1987. Three dimensional photography of soft
sediment benthos, S.W. Bay of Fundy. Biol. Oceanogr. 4: 227-241.
Wildish, D. J., and M. P. Miyares. 1990. Filtration rate of blue mussels as a
function of flow velocity: preliminary experiments. J. Exp. Mar. Biol.
Ecol. 142: 213-219.
Wildish, D. J., and D. Peer. 1981. Method for estimating secondary production
in marine Amphipoda. Can. J. Fish. Aquat. Sci. 38: 1019-1026.
Wildish, D. J., D. L. Peer, and D. A. Greenberg. 1986. Benthic macrofaunal
production in the Bay of Fundy and the possible effects of a tidal power
barrage at Economy Point-Cape Tenny. Can. J. Fish. Aquat. Sci. 43:
2410-2417.
Wildish, D. J., and A. M. Saulnier. 1993. Hydrodynamic control of filtration in
Placopecten magellanicus. J. Exp. Mar. Biol. Ecol. 174: 65-82.
Wildish, D. J., J. D. Trynor, and H. M. Akagi. 1997. Measurement of water
movement by gypsum dissolution method in the Bay of Fundy. Can. Tech.
Rep. Fish. Aquat. Sci. (in press).
Winter, J. E. 1973. The filtration rate of Mytilus edulis and its dependence on
algal concentration, measured by a continuous automatic recording
apparatus. Mar. Biol. 22: 317-328.
Wolaver, T., G. Whiting, B. Kjerfve, J. Spurrier, H. McKellar, R. Dame, T.
Chrzanowski, R. Zinmark, and T. Williams. 1985. The flume design: a
methodology for evaluating material fluxes between a vegetated salt
marsh and the adjacent tidal creek. J. Exp. Mar. Biol. Ecol. 91: 281-291.
Wright, R. T., R. B. Coffin, C. P. Ersing, and D. Pearson. 1982. Field and
laboratory measurements of bivalve filtration of natural marine
bacterioplankton. Limnol. Oceanogr. 27: 91-98.
Yentsch, C. M., P. K. Horan, K. Muirhead, Q. Dortch, E. Haugen, L.
Legendre, L. S. Murphy, M. J. Perry, D. A. Phinney, S. A. Pomponi, R.
W. Spinrad, M. Wood, C. S. Yentsch, and B. J. Zahuranic. 1983. Flow
cytometry and cell sorting: a technique for analysis and sorting of aquatic
particles. Limnol. Oceanogr. 28: 1275-1280.
Young, R. A. 1977. Seaflume: a device for in situ studies of threshold erosion
velocity and erosional behaviour of undisturbed marine muds. Mar. Geol.
23: M11-M18.
Yund, P. O., S. D. Gaines, and M. D. Bertness. 1991. Cylindrical tube traps for
larval sampling. Limnol. Oceanogr. 36: 1167-1177.
3
Dispersal and settlement
Study of larval biology has been under way for -150 years and a large
literature is available, causing Young (1990) to suggest that ideas are
often rediscovered by each new generation of scientists interested in this
field. This literature has been repeatedly reviewed; the most useful and
recent include Crisp (1974), Chia, Buckland-Nicks, and Young (1984),
Scheltema (1986), Butman (1987), Pawlik (1992), Sammarco and Heron
(1994), and McEdwards (1995). The overview presented here relies
heavily on these publications and on selected references which have
appeared in the decade subsequent to 1985. The basis for selection of
references in this chapter is that only those with flow-related content are
included.
Consistent with the aims outlined in Chapter 1, we emphasize hydro-
dynamic mechanisms in larval biology where these are known. Conse-
quently, our concern here is limited to larval and post-larval dispersal at
spatial scales >0.1km and on the environmental factors, notably hydro-
dynamic ones, which influence larval settlement/recruitment at spatial
scales <0.1km on both soft and hard substrates. We begin with back-
ground coverage of the typical benthic suspension feeder life cycle and
dispersal in general.
Benthic suspension feeder life cycle

The life cycle of many sessile benthic suspension-feeding animals in-
cludes a pelagic larval stage important in dispersal (Fig. 3.1). In the
barnacle Semibalanus balanoides, for example, a number of larval stages
can be recognized morphologically, culminating in the cyprid, which is
the competent stage for settlement on a suitable hard substrate. Once
settled on such a surface, the cyprid undergoes a metamorphosis to
69
Adult
Juvenile
Figure 3.1 Life cycle diagram of the acorn barnacle, Semibalanus balanoides
(based on Pyefinch 1948).
produce the post-larval stage or juvenile and, eventually, the adult barna-
cle. Barnacles are hermaphroditic and cross fertilization, by means of an
intromittent organ, involves internal fertilization, followed by release of
the first larval stage to begin the pelagic phase of life. There is a great
range in shapes and sizes of larvae among suspension feeding inverte-
brates, although not all, e.g. polychaete larvae, have recognizably distinct
stages as, for example, do barnacles. Some polychaetes simply add seg-
ments to the body as larval life in the plankton proceeds.
The duration of pelagic life for larval suspension feeders is dependent
on both genetic factors and environmental variables. A characterization
Dispersal and Settlement 71
Table 3.1. Type of larvae.
Type Feeding behavior Examples

1. Lecithotropic Do not feed Spirorbinidae
Many Bryozoa
Ascidiacea
2. Planktotrophic Feed only in the Cirripede barnacles,
pre-competent stage where the cyprid is the
non-feeding competent
stage
3. Planktotrophic Feed in both pre- and Some gastropod molluscs
post-competent stages,
but do not grow further
in the latter period
4. Planktotrophic Feed in both pre- and Sand dollars (Highsmith
post-competent stages and Emlet 1986)
and continue to grow Some bivalve molluscs,
in the latter period e.g. blue mussels
(Bayne 1965)
Source: Scheltema (1986).
based on whether they feed during planktonic life may indicate the
approximate duration of larval life (Table 3.1). Thus, lecithotrophic lar-
vae are short-lived (0.1-2 d) and do not feed, whereas planktotrophic
larvae have the potential for a longer life (e.g. some teleplanic larvae ~2
years) and may feed at various periods during the larval stage (Sheltema
1986). The length of larval life in the plankton generally increases from
type 1 through type 4 in Table 3.1.
Competent larvae are those with the ability to settle on a soft or hard
substrate in the general way shown in Fig. 3.2. According to this scheme,
the competent larva may make several contacts with the substrate, by
either passive sinking or active swimming, before temporarily or perma-
nently burrowing or attaching to it. Whether attachment or burrowing is
involved after initial contact depends on the competent larva receiving
ecologically relevant cues at sensory receptors. If appropriate environ-
mental cues are present, the larvae undergo a hormonally coordinated
physiological change, or metamorphosis, to produce the post-larval stage,
which may precede, coincide with, or follow settlement (Scheltema 1974;
Butman 1987). Settlement simply refers to the change from planktonic to
benthic life of the competent larva; it may involve repeated explorations
Competent larva
Initial surface contacst
.nervous, .sensory^ .Environmental

system receptor cue
Temporary attachment
crawling v
activity >v -
\ nervous sensory Environmental
/ system ~ "receptor" cue
Permanent attachment catalytic factor

metamorphic factor
masked
metamorphic factor
unmasked
Post-larva
Figure 3.2 General model of settlement and metamorphosis of a plank-

totrophic larva, e.g. a barnacle cyprid (based on Chia 1978; Rittschof et al. 1984).
of the substrata, nervous coordination, and appropriate environmental

cues (Scheltema 1974). As pointed out by Butman (1987), this definition
is functionally inadequate in some cases as all benthic stages of soft
sediment benthos, inclusive of post-larva, juvenile, and adult, may under-
take pelagic excursions. Thus, if the post-larva cannot be morphologi-
cally distinguished from the competent larva, as in some polychaetes,
confusion can occur. In those other cases where metamorphosis is coin-
cident with settlement, as in barnacles, there will be no difficulty in
distinguishing between competent and post-larval stages. The term re-

cruitment is considered to be independent of the life history of the
suspension feeder and is defined for investigator sampling convenience.
Hence, the number of juveniles recruited to a settlement plate represents
those retained on the sieve after removing them at some period after
settlement (Keough and Downes 1982). This is not necessarily the same
as the initial settlement of cyprid larvae (Butman 1987), since it excludes
those that may have died after settlement, but before capture, or are too
small to be retained by the sieve.
Dispersal
Dispersal is the general process by which organisms become distributed
spatially, at scales generally >0.1km, to potentially hospitable niches.
Scheltema (1986) outlined four main types of dispersal common among
benthic invertebrates:
Rafting dispersal, generally involving passive transport of post-larval stages

on a floating or submerged raft
Synanthropic dispersal, involving passive transport by human intervention
which includes all life history stages, particularly larvae
Larval dispersal
Post-larval dispersal
Most of this chapter deals with larval and post-larval dispersal, although
we first briefly consider rafting and synanthropic dispersal. The purpose
of this is to ensure for the reader that these forms of dispersal are not
forgotten when pertinent questions regarding it are considered and dis-
persal hypotheses framed.
Dispersal by rafting occurs on a wide variety of living and dead floating
substrates, which may carry all life history stages of suspension feeders
wherever the surface currents dictate. Descriptions of the associated
fauna with surface drifting macroalgal rafts stress the importance of a
resident community, e.g. off the coast of Florida (Stoner and Livingston
1980). DeVantier (1992) described rafting of tropical marine animals,
inclusive of reef and pearl oysters, as well as species of Bryozoa, on the
buoyant skeleton of the reef coral Symphylla agarica, in which chambers
in the coral were air-filled, giving lift during rafting. Other species of
macroalgae may be submerged and moved by benthic currents along the
seabed, rather like marine tumbleweeds. Holmquist (1994) reported that
Laurencia poiteau does this and can therefore carry rafted fauna to new
benthic niches.
The importance of synanthropic dispersal of a wide range of marine
organisms is becoming better appreciated; Carlton and Geller (1993), for

example, report on one important route for this. Since the 1880s, ships
have pumped seawater into their holds as ballast when running empty
between ports. In a study of 159 Asian ships which used Pacific coast
ports of Oregon, Carlton and Geller (1993) found at least 367 taxa in
ballast seawater - mostly larvae - inclusive of many suspension feeders.
One was the larva of the Asian clam Potamocorbula amurensis. This
clam, probably after discharge of ship ballast water (Carlton et al. 1990),
rapidly spread through San Francisco Bay after 1986 and replaced the
previously introduced exotic, the soft shell clam Mya arenaria, which in
turn had replaced an indigenous bivalve Macoma balthica (Nichols and
Thompson 1985). Another well studied example of synanthropic disper-
sal from ballast water is that of the zebra mussels Dreissena polymorpha
and a second species of this genus D. bugensis introduced from Europe
to the North American Great Lakes (Hebert, Mancaster, and Mackie
1989; May and Marsden 1992; Ricciardi, Serrouya, and Whoriskey 1995),
where it has become a problem in blocking freshwater intake pipes.
These dramatic examples of the introduction of exotics and their
sometimes rapid spread in the new location suggest the importance of
this and similar processes in modifying nearshore habitats. Many other
species introduced synanthropically may have replaced the indigenous
fauna and flora without being recognized. It is often difficult to study
rafting and synanthropic dispersal in the field because of the stochastic
nature of these processes.
Larval dispersal
Planktotrophic larvae are generally <1000|im in length and swim/sink at
Reynolds numbers in the creeping flow range. Such larvae swim by
means of beating cilia, although larvae which are longer than 1000 Jim
are unable to swim in this way as a result of inefficiencies linked to size,
and are forced to use muscular contractions to swim (Chia et al. 1984).
Examples include the later stages of some polychaete larvae or the
lecithotrophic tadpole larvae of ascidians. Chia et al. (1984) showed that
larvae using muscular power to swim increased their speed proportional
to an increase in body size, although this was not the general case with
ciliary swimming larvae.
Although data are readily available for larval size in the literature (for
example, see the review of larval stages of bivalves presented by
Ackerman et al. 1994), concomitant and reliable data on swimming
speed and sinking rate are often lacking, explaining the few entries in
Table 3.2.
The key factors which influence the final distances achieved by larvae
during dispersal and before settlement of the competent stage are dura-
tion in the plankton, larval sinking rates, larval swimming speed, and
direction and velocity of ambient currents. Environmental factors which
influence the length of larval life, and therefore the potential dispersal
distance, are physical, e.g. salinity and temperature, or biological, e.g. the
availability and quality of sestonic food or predation.
The length of time that larvae are planktonic, both between species
and within species, is highly variable. Hence, lecithotrophic larvae may
spend a few hours or days in the plankton, with the outcome that they
either settle or die because of predation or limiting energy resources
(Scheltema 1986). Longer-lived planktotrophic larvae are affected by
physical factors in various ways. As an example, the dispersal of the
larval stages of the blue mussel, following the release of sperm and egg,
is depicted in Fig. 3.3. After fertilization, the trocophore is followed by
the veliger, the velichoncha, the eyed veliger, and the pediveliger, which
is the competent settling stage, and this sequence can take from 26 to 60 d
(Bayne 1964a). After settling on seaweed, the pediveliger metamor-
phoses to the first post-larval or spat stage, termed the plantigrade (not
shown in Fig. 3.3). The plantigrade can actively migrate to a more suit-
able hard substrate nearby to grow into the juvenile and adult forms of
the blue mussel. Bayne (1965) conducted growth experiments with blue
mussel pediveligers in which optimum salinities of 30-33 parts per thou-
sand (%o) and food concentrations of 1 x 105 cells L"1 were offered while
varying the temperature. It was found that mixed cultures of microalgae
led to better growth rates than if a unialgal culture was offered as food
and that the delay of metamorphosis achieved by pediveligers ranged
from ~2 to 46 d, in inverse relation to temperature (Bayne 1965). Salinity
also affected the time spent as a pediveliger, so that the delay was greater
at the lower salinities tested of 12.5%o (Bayne 1965).
From a knowledge of the duration of planktonic life it is possible to
predict crudely the distances the larvae could disperse on the basis of the
dispersion diagrams prepared by Okubo (1971). A three-dimensional
mathematical model of ocean circulation over Georges Bank was con-
structed by Tremblay and Sinclair (1994) to predict the passive transport
of larvae of the giant scallop Placopecten magellanicus. The model was
reasonably successful in predicting the flow field and, consequently, in
simulating the scallop larval drift over Georges Bank. Important vari-
Table 3.2. Larval size, swimming speed, sinking rates, and duration of stay in the plankton for some benthic
suspension feeders.
Size, urn
Sinking Swimming speed, cm s
Competent rate, Duration,
Taxon Egg larva cm s"1 Up Down Horizontal days Reference
Cnidaria, Octocorallia
Eunicella stricta 0.38 0.2 Theodor (1967)
Annelida, Polychaeta
Phragmatopoma 0.07-0.16 0.20-0.50 Pawlik et al.
lapidosa californica (1991)
Streblospio benedicti 300-1400 0.01-0.30 Butman (1989)
Mollusca, Bivalvia
Crassostrea virginica 240-320 -0.83 0.13-0.27 0.13-0.27 0.01-0.1 Hidu and Haskins
(1978)
Pecten maximus 66-70 240-260 0.10 0.12 32-41 Cragg (1980)
C. virginica >250 0.037-0.102 >12-24 Mann (1988)
Spisula solidissima 196.1 0.22 0.03-0.04 24 Mann et al. (1991)
Mulina lateralis 159.7 0.13 0.03 0.02 19 Mann et al. (1991)
Rangia cuneata 168.5 0.17 0.02-0.05 0.06 12 Mann et al. (1991)
Crustacea, Cirripedia
Balanus crenatus 0.17-0.55 12-20 DeWolf (1973)
L3
straight hinge veliger
L2
young veliger L4
veliconcha
L1
trochophore
fertilized L5
zygote eyed
veliger
Figure 3.3 Diagram of larval dispersal and settlement of the blue mussel
Mytilus edulis (based on photomicrographs andfiguresin Bayne 1976). The post-
larval plantigrade is not shown.
ables in determining larval drift were thought to be larval growth rates,

larval mortality rates, and depth maintained by the larvae during
drifting.
Earlier attempts at measuring sinking rates and swimming speeds of
dispersing larvae involved a measured distance and stopwatch, but were
often made in a too small volume of seawater, e.g. on a microscope slide
(Chia et al. 1984). As a consequence, "wall effects" (a sheared zone near
the slide surface) may have influenced the results obtained in such a way
as to underestimate the capability of the larva tested. Some examples
thought to be free of this effect are shown in Table 3.2. DeWolf (1973)
measured sinking rates of barnacle cyprids in a glass tube of 20 cm
diameter and 1.25 m length. Hidu and Haskins (1978) used a sealed glass
tube of 1.4 cm diameter and 15 cm length containing seawater to investi-
gate both sinking rate and swimming speeds of oyster larvae. Cragg
(1980) used various methods from a slide well up to a 4-L jar of 25 cm
10-
E
o 5-
Q
LU
LU
CL
CO 2-
O
1-
CO
0.5 -
60 100 140 180 220 260 300
MEAN LENGTH, pm
Figure 3.4 Effect of larval length of Crassostrea virginica on the log of swim-
ming speed at S = 25 o/oo and three different temperatures (Hidu and Haskins
1978). To obtain the size range shown, trocophores, straight hinge, and eyed
veligers were used. (From Hidu and Haskins 1978, reused by permission, Copy-
right Estuarine Research Federation.)
depth to determine larval swimming speeds of scallop. Butman (1989)
and Pawlik, Butman, and Starczak (1991) measured sinking rates with
narcotized or dead larvae and swimming speeds on live, suspension-
feeding polychaete larvae in settling chambers (9.5 cm diameter x 122-
cm length or 15 x 15 x 40 cm long) which were enclosed by a tempera-
ture-controlled water bath. Sinking rates shown in Table 3.2 vary from
0.01 to 0.83 cm -s"1, equivalent to the settling rates of inorganic sedimen-
tary particles of <1 to 100|iim equivalent diameter in size. Swimming
speeds in Table 3.2 vary from 0.01 to 0.50cm s"1, and if an average speed
of 0.2cm-s"1 was maintained throughout Id, the larva could swim a
distance of 173 m. Larval ciliary swimming speeds are strongly influenced
by environmental variables such as salinity and temperature because the
efficiency of ciliary propulsion is influenced by seawater viscosity; thus,
Hidu and Haskins (1978) were able to show that oyster larval swimming
speed was inversely related to temperature (Fig. 3.4).
Finally, the dispersion of larvae is critically dependent on ambient
currents, which are often of a larger magnitude than their swimming or
sinking velocities. The scales of relevance, in both time and space, are
very broad and may include the effects of molecular and turbulent diffu-
sion, tides, storm-mixing events and wind-driven currents, Langmuir
circulation, internal waves, mesoscale eddies, and large-scale general
circulation (Okubo 1994).
A number of questions might be asked about the dispersion of larvae.
For instance, how are they distributed downstream of a point or area
from which they are released, either instantaneously or as a steady
stream, from the bottom of a benthic boundary layer? There is advection
of larvae along paths defined by stream lines, but they will also be subject
to turbulent diffusion, which will move them from the mean streamlines
in any direction. This problem can be analyzed in terms of the advection-
diffusion equation, an application of which will be discussed in Chapter
7. Its solution will give the concentration of larvae at any point down-
stream. This equation can be made to include terms for larval mortality
rate and for deposition on the bottom, if these are known.
In many circumstances, larval dispersion is dominated by more com-
plex flows than in the benthic boundary layer. For example, it is well
known that shoreward advection by currents driven by wind or waves
may be an important transport mechanism (LeFevre and Bourget 1992).
Two-layer flows in salt wedge estuaries and partially mixed estuaries
(Dyer 1979) will allow either landward or seaward transport of eggs or
larvae, depending on their depth. Other important flows which may
influence larval transport are tidally forced internal waves and bores, and
coastal up-welling, as will be subsequently discussed in more detail.
Topographically determined flows, such as observed around reefs and
embayments (Fischer et al. 1979), are not easily analyzed by use of the
advection-diffusion equation. Some of these topics are considered by
Okubo (1994). The physical oceanography of larval dispersal has also
been considered in conference proceedings edited by Sammarco and
Heron (1994).
An interaction involving one of the physical oceanographic processes
discussed and the swimming behavioral responses by the larvae of some
suspension feeders is important in determining the outcome of the dis-
persal process. We have collected the most prominent hypotheses con-
cerning such matters in Table 3.3.
Estuarine retention
In estuaries freshwater from a river is mixed with seawater from the
ocean by the action of tides, wind effects at the surface, and river dis-
Table 3.3. Physical oceanographic processes and larval behavior

control the dispersal of larval benthic suspension feeders.
No. Ho/Hi Description Reference
Ho Swimming by larvae in estuaries is Carriker (1951)

not related to positioning in different Bousfield (1955)
water masses and dispersal is passive Mann (1988)
Hj Swimming by larvae causes Thiebaut et al. (1992)
appropriate positioning in two layer
flows which ensures estuarine
retention
Ho Larvae are randomly dispersed in Hawkins and Hartnoll
nearshore seawater (1982)
Hx Shoreward transport of benthic larvae Shanks (1983,1985,
is caused by onshore winds 1986, 1988)
Shanks and Wright
Ho Larvae are randomly dispersed in (1987)
nearshore seawater
Hx Shoreward transport of benthic larvae
occurs after behavioral concentration
in the surface over tidally forced
internal waves
Ho Larvae are randomly dispersed in Pineda (1991)
nearshore seawater
Hx Shoreward transport of benthic larvae
occurs in bottom waters where
internal tidal bores are present
Ho Larvae are randomly dispersed in Farrel et al. (1991)
nearshore seawater
Hj Shoreward transport of benthic
larvae results from shoreward
advection which occurs if wind-
induced upwelling is absent
Ho Plant canopies have no influence on Eckman (1983, 1987)
suspension feeder recruitment and Peterson (1986)
equal numbers of recruits occur on Wilson (1990)
non-vegetated areas
Hx Plant canopies of sufficient density
entrain significantly more benthic
recruits than non-vegetated areas
H2 In plant canopies of sufficient density,
suspension feeder recruitment
depends linearly on volumetric flow
rate
charge moving seaward. Density differences between seawater and river

water may also result in horizontal pressure gradients which affect flow
patterns (Dyer 1979). On the basis of mixing processes and circulation,
estuaries of temperate climates may be classified (Dyer 1979) as fjords,
well mixed, partially mixed, or salt wedge. Only the two latter categories
are of concern to us here. The salt wedge estuary is one in which there is
a marked separation, at the halocline, between the bottom-more saline
seawater and the surface, less saline river water. Seawater is mixed
upwards by a process of entrainment across the halocline. In the partially
mixed estuary, tidal action is dominant and tidal turbulence is the major
mixing mechanism, causing seawater to be mixed upwards and river
water downwards. Typically in the partially mixed estuary, the surface
and bottom water are fresh toward the landward end and salt toward the
seaward end, with the vertical salinity gradient slight. By contrast, in the
salt wedge estuary, the seawater extends far landwards with the vertical
salinity gradient sharp at the halocline.
Concerning larval dispersal in estuaries, the most important physical
feature of both salt wedge and partially mixed estuaries is that the
bottom water has a net movement landwards and surface water seaward
when averaged over a number of tides. The basic estuarine retention
hypothesis (number 1 in Table 3.3) is that by vertical swimming guided
by appropriate environmental cues such as light, salinity, or gravity, the
larva is able to position itself to be transported to a suitable site for
settlement. This may involve a change in sign of the response to an
environmental cue during larval ontogeny with larvae positively
phototaxic early in life so that they are passively transported seawards.
Later in larval life they become negatively phototaxic so that a compe-
tent larva is returned landwards by passive transport in bottom water,
enabling it to settle near the adults producing it. An alternative mecha-
nism, that greater larval size and/or sinking rate is achieved during
ontogeny, accounts for depth position within the water column.
The estuarine retention hypothesis has been the subject of many field
studies by fish and decapod larval biologists, but relatively few concern
suspension feeders (Stancyk and Feller 1986). The field studies listed in
Table 3.4 are of plankton sampling in estuaries where the authors con-
sider that the temporal and spatial distributions of dispersing larvae of
suspension feeders are consistent with the estuarine retention hypoth-
esis. The field study in the Miramichi estuary, New Brunswick, Canada,
by Bousfield (1955) of various barnacle larvae is probably the most
comprehensive in support of an estuarine retention mechanism. This
Table 3.4. Field studies involving plankton net sampling of larvae of

suspension feeders within estuaries.
Major taxa Species Reference

Cirripede barnacles Semibalanus balanoides Bousfield (1955)
Balanus improvisus,
B. crenatus
Bivalve molluscs Crassostrea virginica Carriker (1951)
C. virginica Seliger et al. (1982),
Boicourt (1982)
C virginica Mann (1988)
Tubicolous polychaete Owenia fusiformis Thiebaut et al. (1992)
author showed that larval swimming depth increased with age so that
competent cyprids were found at the greatest depths. Position of larvae
within the water column varied with the tidal stage and current velocity.
Bousfield (1955) calculated that approximately 18 d later, -10% of those
initially spawned remained as competent larvae within the estuary, and
these were sufficient to maintain adult populations. By contrast, if the
larvae of Balanus improvisus maintained a uniform distribution in the
water column, his calculations showed that, as a result of seaward
advection, insufficient numbers would remain to recruit the historic
numbers of adult barnacles within this estuary.
In the study by Mann (1988) of the James River estuary, Chesapeake
Bay, in the United States, oyster larvae were reported to migrate through
a salinity discontinuity front present in the estuary and were then trans-
ported downwards, where they became entrained in the more saline
bottom water which moved landwards. Laboratory behavioral experi-
ments showed that intermediate stage oyster larvae could actively swim
through a salinity gradient (Mann 1988) involving a salinity difference of
3%o.
As pointed out by Stancyk and Feller (1986), some authors using
similar field observations consider that the observed larval distributions
support the null hypothesis of number 1, Table 3.3. In all cases, though,
the physical oceanographic conditions in the estuaries examined do not
conform to salt wedge or partially mixed conditions, rendering the estua-
rine retention hypothesis invalid in these conditions. Thus, DeWolf
(1973), in a study of barnacle cyprids in the western Wadden Sea, the

Netherlands, suggested that the larvae were transported passively by
tidal currents, periodically sinking to the sediment at slack tide and being
resuspended when currents were greater than 35-67 cm-s"1. Likewise,
Korringa (1952) found that the larvae of Ostrea edulis in the well mixed
Oosterschelde estuary, the Netherlands, were passively distributed in
this tidally energetic system, where no salinity stratification was present.
The larvae of the intertidal spionid polychaete Pseudopolydora
paucibranchiata in the San Diego River estuary, California, were found
by Levin (1983) to be mostly passively retained, with no evidence of
adjustment of the vertical position of the larvae within the water column.
This estuary was much impacted by human modification, with no evi-
dence of a salt wedge or partially mixed conditions.
Support for the estuarine retention hypothesis is also to be found in
laboratory measures of larval sinking or swimming rates, as shown in
Table 3.2. The observed movements are well within the capabilities
required to traverse the vertical distances for estuarine retention. Other
support is to be found in the laboratory demonstration of positive or
negative swimming responses directed to the source of the environmen-
tal cue, referred to as taxes, such as light and gravity. As an example, the
earliest veliger larvae of blue mussels Mytilus edulis are briefly
photonegative, but this condition rapidly changes so that no response to
light is present, although during most of larval dispersion they are nega-
tively geotaxic (Bayne 1964b). In the later stages of life, blue mussel
pediveligers are negatively phototaxic and positively geotactic (Bayne
1964b). Other responses to light by larvae may be non-directional so that
swimming continues until the larva reaches shade, a process termed
photokinesis (Crisp 1974). In the review by Crisp (1974), only light,
gravity, and flow direction were environmental cues considered to pro-
vide directional information for larvae. Other physical variables inclu-
sive of light intensity, pressure, temperature, and salinity were scalar
quantities lacking any directional information content.
Although not an estuarine study, but a study of blind ending coastal
bays in the vicinity of Narragansett Bay, Rhode Island, United States,
Gaines and Bertness (1992) showed that over a 10-year period, barnacle
settling rates depended linearly on the average flushing time of each bay
investigated. For Semibalanus balanoides and their particular habitat,
passive dispersal based on the physical oceanographic conditions domi-
nant at the site lead to good prediction of spatfall.
Shoreward advection
For open coastal environments such as wave pounded vertical rock faces,
the question as to how planktotrophic larvae manage to recruit in a
narrow zone on the shore has intrigued many investigators. Barnacle
larvae have been a favorite in such studies because they are
planktotrophic and feed only in the pre-competent stage. Balanus
crenatus, for example, may spend from 12 to 20 d in the seawater of the
Wadden Sea (DeWolf 1973) before settlement occurs. The question
addressed here is, How does the competent cyprid regain the intertidal
rock face required for the completion of the barnacle life cycle?
Working on the North Wales, United Kingdom, coast, Bennell (1981)
suggested that Semibalanus balanoides settled on scraped natural rock
surfaces which were cleaned daily when the wind direction was offshore
and the sea state calm. In contrast, Hawkins and Hartnoll (1982), using
similar field techniques but on the Isle of Man, United Kingdom, found
a positive relationship between S. balanoides settlement density and
timing with onshore winds (wind direction and velocity for the hour near
high water) during May 1979 (see number 2, Table 3.3). In a California
study by Shanks (1986), daily settlement, mainly of Chthalamus sp., was
not correlated with wind direction for a 4-mo period in 1983 or wind
direction at high water. Settlement rates of cyprids here were found not
to be significantly different on days with onshore versus those with
offshore winds.
In the study by Shanks (1986), three possible mechanisms were consid-
ered to explain how barnacle larvae reached suitable rock settlement
sites in the intertidal:
Cyprids swam ashore.
Cyprids were randomly dispersed and deposited on the shore.
Cyprids utilized onshore currents to transport them to shore.
The larval swimming capabilities were considered to be insufficient for
the purpose, since cyprids are commonly found in abundance to ~10km
away from the coast (Shanks 1986) and up to 100 km offshore on the
French Atlantic coast (LeFevre and Bourget 1991). The second of the
possibilities has been used as the null hypothesis for numbers 2-5 in
Table 3.3, although field evidence, e.g. the neustonic concentration of
barnacle cyprids of Verruca stroemia discovered by LeFevre and Bourget
(1991), suggests that, at least for some species, this is not true. The initial
hypothesis of shoreward transport of larvae by internal waves was devel-
oped for crab megalopae (Shanks 1983), and an integral part of the
Rough Slick
^VVW ^^ ^ ^r- w ^ ^r- w v w w w v
/ t \
7
Figure 3.5 Diagram of currents over linear internal waves as postulated by

Shanks and Wright (1987) for the Southern California coast. Larvae are trans-
ported shorewards in surface water by the internal wave.
mechanism involved a behavioral response by crab megalopae (Shanks

1985). Crab megalopae showed positive phototaxis, negative geotaxis,
and kinetic responses to light and pressure in laboratory experiments
and, infieldexperiments involving release of crab megalopae, a tendency
to swim in surface waters (Shanks 1985). Whether the competent stages
of suspension feeding larvae, such as barnacle cyprids, possess simple
behavioral responses which keep them in surface seawater, as is required
by the internal wave transport hypothesis, was not determined in
Shanks's studies, although Crisp (1974) reports that many species of
barnacle cyprids are positively phototaxic in laboratory conditions, and
LeFevre and Bourget (1991) demonstrated that cyprids of V. stroemia
are concentrated in surface seawater.
The physical events caused by the passage of a linear internal wave are
shown in Fig. 3.5. They are caused when the ebb tide slackens over a
large-scale bottom feature such as the continental shelf break, causing a
lee wave to be released, resulting in a series of large amplitude internal
waves (Shanks and Wright 1987). During the internal waves passage, an
up-welling occurs ahead and a down-welling behind the internal wave.

The convergent down-welling is frequently marked by a surface slick
formed by buoyant particles - flotsam - which does not follow the down-
welling flow.
Studies by Shanks (1986) on the California coast, involving daily meas-
urement of cyprid settling rates, showed that the rates were correlated
tidally with peaks of settlement occurring 1-4 d before the spring tide.
This is consistent with the tidal forcing which causes the internal waves.
Further field tests of hypothesis number 3 (Table 3.3) were made in the
San Juan Archipelago on the U.S. West Coast (Shanks and Wright 1987).
These studies involved plankton sampling in surface seawater to com-
pare the cyprid concentrations within and between slicks. As predicted
by the internal wave shoreward transport hypothesis, the abundance of
Balanus glandula and Semibalanus cariosus was significantly higher
within the slicks on three of the four occasions when this was tested.
Surface drifters deployed by Shanks and Wright (1987) also indicated
that surface shoreward transport occurred and that cyprids settled on the
rocks near where the drifters were carried. In a further field study of this
type, Shanks (1988) showed that internal wave shoreward transport oc-
curred, although few suspension-feeding larvae were included. This
study was conducted near Beaufort Inlet, North Carolina, in the Atlantic
Ocean, where the tidal range was lower (<2m) and the continental shelf
wider (>80km) than in previous studies.
Two other shoreward advection hypotheses have been proposed
(numbers 4 and 5, Table 3.3). Thus, Pineda (1991), working at La Jolla,
California, found that invertebrate larvae, inclusive of barnacle cyprids,
were concentrated in bottom water, contrary to the predictions of the
internal wave transport hypothesis. To explain this he proposed that the
shoreward advection of these larvae resulted from internal tidal bores
associated with submarine canyons, which, in this part of the California
coast, are very close to shore. One of the predictions from this hypothesis
- that cyprid settlement peaks should correspond to low temperature
events due to internal tidal bores - was confirmed by Pineda (1991).
Thus, daily cyprid settlement rates of barnacles (Chthamalus sp. and
Pollicipes polymerus) showed a statistically significant negative correla-
tion with seawater temperatures. The shoreward advection of cold sub-
surface seawater by internal tidal bores is predictable because they are
tidally related.
Another mechanism was proposed (number 5, Table 3.3) on the basis
of field studies of the mid-California coast by Farrel, Bracher, and
0.60 n
a:
UJ 0.20 -\
a:
0.00
APRIL I MAY I JUNE JULY AUG SEPT
Figure 3.6 Recruitment density as mean number of recruits per settling plate
(n = 18) against time for local barnacle settlement of B. glandula and
Chthalamalus sp. on the mid-California coast (Farrel et al. 1991). Wind speed is
shown (north > 0, south < 0). Bars at the top show when internal waves occurred.
Roughgarden (1991). Barnacle cyprid recruitment of Balanus glandula

and Chthamalus sp. was monitored every other day and various physical
oceanographic features coincidentally monitored. During the summer
period of 1988, four major barnacle recruitment pulses (shaded area in
Fig. 3.6) which correlated with the advection of warm, low salinity
seawater shorewards were recorded. The pulses were not correlated with
internal waves, onshore winds, or pulses of release of larvae by adults.
Shoreward advection in this case appeared to be controlled by winds
which caused up-welling at the narrow continental shelf edge. Coastal
up-welling in association with along-shore winds caused net offshore
transport. With the prolonged relaxation of winds, the transport reversed
and became shorewards.
Plant canopies
Physical oceanographic interactions with seagrass meadows are dis-
cussed in Chapter 8 in the section, Marine plant canopies: seagrass
meadows, and interactions with kelp forests in the section, Hydrodynam-
ics. In general, they involve the development of a skimming flow over the
plant canopy, dependent on a sufficiently high flow speed and canopy
density, and within the canopy a reduction in speed, dependent on stem
density, which results in reduced shear stress and, therefore, causes
increased sedimentation. The original reason for studying seagrass
meadows was to find an explanation of the higher species diversity and
animal abundance commonly found there. Thus, passive larval entrain-
ment within the canopy could be an explanation of the increased abun-
dance, which is more fully documented later in this section, where other
hypotheses to explain it are also considered.
The hypothesis tested which relates to plant canopies (number 6,
Table 3.3) originated with Eckman (1983). In this study, a field experi-
ment was set up on the Washington coast of the United States. It in-
volved removing bulrush Scirpus americanus marsh plants in small plots
and then inserting simulated marsh plant stalks consisting of drinking
straws into the sediment. Recruitment of facultative deposit/suspension
feeders such as a spionid and sabellarid polychaete, but mainly of
meiofauna, was shown to have a significant positive correlation with the
numerical density of straws, in agreement with a hydrodynamic mecha-
nism of larval dispersal. Field observations on the North Carolina coast
in a seagrass meadow consisting of 80% Zostera marina and 20%
Halodule wrightii were conducted by Peterson (1986) to determine
whether larval supply rates explained the higher density and biomass of
hard clams within the meadow. The hard clam Mercenaria mercenaria
was sampled in 0.25-m2 cores with a sieving procedure that obtained only
clams of >5mm valve height. The size range of 5-25 mm sampled at
annual intervals was considered to be a good indication of larval recruit-
ment. The results, shown in Table 3.5, suggest that significantly more
larval (0-year) recruits do appear within the seagrass beds than on the
unvegetated sandflat. However, Peterson thought that passive hydrody-
namic trapping of larvae within the seagrass meadow was only one part
of the story since the older hard clams were present at a significantly
greater density within seagrass than on the unvegetated sandflat. The
numbers of older hard clams shown in Table 3.5 are higher than numbers
of 0-year recruits, in part because several years are accumulated in
calculating this mean. The larger differential within older hard clams
between seagrass meadow and unvegetated sandflat was thought by
Peterson (1986) to be due to differential survival within the two habitats,
with more efficient post-settlement predation likely to be the cause of the
Table 3.5. Mercenaria mercenaria densities per 0.25m2 in a North

Carolina lagoon based on the mean standard error of 36 subsamples.
1980 1981
Plotl Plot 2 Plotl Plot 2
0-year recruits <25-mm

valve height
Seagrass 0.94 (0.21) 0.72 (0.15) 0.61 (0.13) 0.27 (0.09)
Sandflat 0.33 (0.11) 0.36 (0.11) 0.17 (0.06) 0.05 (0.04)
Older animals >25-mm
valve height
Seagrass 3.11 (0.48) 1.78 (0.33) 3.52 (0.53) 2.20 (0.24)
Sandflat 0.20 (0.11) 0.36 (0.11) 0.11 (0.07) 0.33 (0.09)
Source: Peterson (1986).
greater losses on the sandflat. That predation was more efficient on the
sandflat and inhibited by seagrass roots and rhizomes had been demon-
strated by Peterson (1982) and others.
In considering the second alternative hypothesis of number 6 in Table
3.3, Eckman (1987) mounted a field study in a Zostera marina meadow
on Long Island, United States, in which the recruitment of two
suspension-feeding bivalves - the bay scallop Argopecten irradians and
the common jingle Anomia simplex - were determined at four sites
within the local area. Both larval bivalves settle first on eelgrass blades
by means of byssus threads. By increasing the numbers of available
eelgrass blades with plastic mimics, recruitment as measured at approxi-
mately weekly intervals was dependent on the velocity of seawater
through the meadow. Thus, high density eelgrass meadows had fewer
recruits than those with fewer blades impeding the flow, as a result of
differences of volumetric flow which passed through the meadow.
In all the plant canopy studies so far reviewed, recruitment was meas-
ured and the likelihood of post-settlement mortality at each site not
independently assessed, and hence excluded as a possible confounding
factor in determining whether the initial settlement rate was indicated by
the recruitment rate. One such study in which settlement rates were
determined directly was reported by Wilson (1990). Larval settlement
rates were determined in settlement traps which were embedded in the
sediment and the larval catch counted daily. Thefieldwork,conducted in
Bogue Sound, North Carolina, showed that mean settlement of bivalve

larvae within traps in a Zostera marina-Halo dule wrightii meadow was
1.6 times greater than in nearby sandy areas which lacked seagrass. The
between-habitat differences were reported by Wilson (1990) to be statis-
tically significant and thus support Hx of hypothesis number 6 (Table
3.3).
Duggins, Eckman, and Sewell (1990) reasoned that understorey kelps
which share the same kind of interaction with flow (Chap. 8, the section,
Marine plant canopies) as do seagrass meadows will also act to entrain
larvae (number 6, Table 3.3). The field methods used in the San Juan
Archipelago area of the U.S. West Coast involved acrylic plastic
(Plexiglas) plates which were bolted to the rock bottom. Deployment of
the plates was for 3- to 5-wk periods, although Duggins et al. (1990) claim
that for the four major species which settled on the plates, they could
measure initial settlement because each produced a calcareous secretion
within hours of settling, which persisted and was counted even if the
larva died soon after initial settlement. A complex field experiment was
designed in which understorey kelp densities and substratum orientation
- upward or downward facing plates - were manipulated. This allowed
the following effects on settlement to be analyzed: kelp/no kelp, low
flow/high flow, dark/light effect on microalgal growth, and sedimenta-
tion/no sedimentation. The results suggested that of the four species, all
except Membranipora membranacea had higher recruitment within the
kelp treatment, although for the spirorbid polychaetes which settled, the
early season response changed to a preference for the no-kelp treatment
later in the season. The bryozoan Membranipora preferred high-flow,
low-sedimentation environments, suggesting that it is intolerant of depo-
sition environments. These results therefore provide support for Hx of
hypothesis number 6 (Table 3.3) for some species but not others, and
clearly demonstrate the importance of other factors such as flow, sedi-
mentation, or growth of microalgal mats as involved in larval settlement
within understorey kelps. As with seagrass meadows, kelp forests may
also be subject to a key predator which affects larval settlement rates.
For example, Roughgarden, Gaines, and Pacala (1987) showed that in
central Californian kelp forests, the juvenile rock fish Sebastes sp. could
significantly reduce any zooplankters passing through the kelp. Thus,
barnacle cyprids within the kelp forest were 70 times less abundant than
those outside as a result of rockfishpredation, and when the rock fish left
the kelp in the fall, the numbers of barnacle cyprids rapidly built up
again.
Post-larval dispersal
Post-larval stages of many invertebrates frequently reappear in the water
column after their settlement and metamorphosis. This was documented
in a review by Butman (1987), which lists 33 field observational studies
involving plankton tows, traps, or seawater pumping in which post-larvae
of bivalves, polychaetes, plus a few gastropods and meiofauna were
found in the samples.
A dispersal function for post-larval blue mussels Mytilus edulis in the
Wadden Sea was proposed by Maas Geesteranus (1942). The young
mussels were passively transported by currents until they detected
a suitable substratum for their growth. Byssal detachment and
reattachment of the young mussels could occur many times until a suit-
able substrate was found. Life cycle studies undertaken by Bayne
(1964a) of blue mussels in the Menai Straits, North Wales, referred to
earlier, suggest that the pediveliger larva first settles on afilamentousred
alga. After settlement and metamorphosis to the post-larval or planti-
grade stage, the young mussel may detach and be passively transported
by seawater currents. Bayne (1964a) referred to primary and secondary
settlement processes to describe this part of the blue mussel life history.
However, McGrath, King, and Gosling (1988), in studies on the rocky
shores of the west coast of Ireland, found that pediveligers of M. edulis
settled directly onto the adult mussel reef without first settling on a
filamentous red alga. Sigurdsson, Titman, and Davies (1976) presented
field and laboratory observations designed to elucidate the mechanisms
involved in post-larval dispersal in bivalves. They showed that slow
currents induced post-larval bivalves to produce a long byssus thread
which was functional in drifting dispersal. For example, juvenile tellinids
Abra alba of lmm valve height could be induced to produce a
mucopolysaccharide byssus thread which was 3 cm long and 2-4|iim in
diameter. The byssus thread increases the surface area of the juvenile
bivalve significantly. Because the settling velocity in creeping flow re-
gimes is proportional to surface area, we can expect the settling velocity
of the juvenile bivalve to be significantly less, thus enhancing dispersal.
The post-larval bivalve was able to detach from its byssus thread, upon
which it passively sank to the bottom. Sigurdsson et al. (1976) thought
that byssus drifting was of importance for many bivalve species of up to
2.5 mm in valve height, inclusive of protobranchs, e.g. Nucula tennis, and
many lamellibranchs, e.g. Mytilus edulis and Mya arenaria.
Most of the studies described are observational and designed to aid in
Table 3.6. Hydro dynamic hypotheses related to post-larval life of

suspension feeders.
No. H Q / H ! Description Reference
1 Ho Post-larval dispersal plays no part Maas Geesteranus (1942)

in determining density and Bayne (1964a)
production Sigurdsson et al. (1976)
Hx Post-larval dispersal is important Emerson and Grant (1991)
in determining density and Roegner et al. (1995)
production
2 Ho Post-settlement mortality is not Luckenbach (1984)

related to hydrodynamic- Sammarco (1991)
sediment conditions
Hi Post-settlement mortality is
related to hydrodynamic-
sediment conditions
understanding the possible mechanisms of post-larval dispersal. But find-

ing post-larvae within the water column does not necessarily mean that
the dispersing individuals are successful in recolonizing or indicate the
relative importance of this method of dispersal. In a field study based on
the passive bedload transport of juvenile softshell clams My a arenaria,
Emerson and Grant (1991) were able to frame their hypothesis with an
ecological emphasis, as shown in hypothesis 1 of Table 3.6. Specially
designed subsediment bedload sampling traps, consisting of a 3-cm-
diameter acrylic tube which could be removed from a supporting pipe,
were placed so that the opening was flush with the sediment surface.
Sedimentary bedload and macrofaunal trap catches were recorded on a
daily basis from sandflats in Halifax Harbour, Canada, and compared
to independent estimates of clam population density and spat settlement.
The transport of juvenile clams >6mm was commonly observed within
Halifax Harbour, reaching maximum rates of 790 clams m"1 d"1 at a
sheltered site and 2600 clams m"1 d"1 at a more exposed site. At both
sites, clam transport was correlated with bedload sediment transport,
with the maximum associated with storms when it reached 35 kg m"1 d"1.
At the sheltered site, clam transport was negatively correlated with clam
density, while at the exposed site, it accounted for a large jump in density
in September. This suggested that the juvenile clams were able to
reburrow and recolonize after transport. Clam transport at the exposed

site caused a precipitous decline 2 mo later and the complete removal of
any recently settled spat. These results clearly support Hx of number 1 in
Table 3.6. Roegner et al. (1995) have determined in flume experiments
that 2-wk-old juvenile recruits of the softshell clam behave passively,
essentially like sedimentary particles, since they avoid erosion until ambi-
ent velocities reach those necessary for the initiation of sediment trans-
port. Comparing living and dead clams shows that burrowing behavior of
young clams is one way of avoiding transport away from the site by flows
sufficient to cause sediment erosion.
Early post-settlement mortality, as is implied in the wind-wave ener-
getic sandflat in Halifax Harbour, is recognized to be a potentially impor-
tant variable for soft sediment benthic community structure and func-
tion. The effect of environmental variables, particularly predation, on
early recruit mortality patterns was studied, for example, by Luckenbach
(1984) on coot clams, and Sammarco (1991) on coral reef recruitment
(number 2, Table 3.6). Their results support Ho of number 2, but do not
consider flow as an alternate hypothesis. We could find no experimental
tests which identify flow as a perturbing factor and assess the importance
of early post-larval stage mortality caused by it.
Local hydrodynamics
The focus for the rest of this chapter shifts to a much smaller spatial scale
- <0.1km. Of most concern in benthic studies are flows near solid
boundaries, whether relatively flat bottoms or more irregular shapes
such as boulders and rocky surfaces. Except at very low velocities and at
very small distances from the surface, theseflowsare turbulent. Creeping
flows are seen very close to the solid boundaries, and, although these
regions are thin, they may be large by comparison to typical sizes of
larvae or eggs.
The hydrodynamically smooth benthic boundary layer is typically
found over soft sediments and other fine-grained bottom materials. This
flow is characterized by the following:
The viscous sublayer, which is a viscosity-dominated layer near the bottom
where the flow is mostly laminar
The logarithmic layer, which is turbulent and where the mean velocity varies
as the logarithm of the distance above the bottom
The outer layer, which is also turbulent but with turbulent energy decreasing
with height
The free stream flow, which drives the flow below it (Soulsby 1983)
100:
0.01
0 2 4 6 8 10 12 14 16
HORIZONTAL FLOW SPEED, cm s"1
Figure 3.7 Turbulent velocity profiles at different shear stresses over a soft
sediment: a, smooth, /* = 0.60cm-s"1; b, rough, U* = 0.82cm-s"1; c, rough,
i7* = 0.98 cm -s"1. The curved region of line a represents the viscous sublayer. A
240-fxm pediveliger larva of the blue mussel is shown to scale (based on Butman
1986).
The viscous sublayer is typically of the order of millimetres thick, while

the logarithmic layer is a few metres deep.
A larva which has to move between the bulk flow and the bottom must
cross turbulent zones as well as the viscous sublayer, a flow dominated by
viscosity. A typical velocity profile for a smooth boundary (from Butman
1986) is shown in Fig. 3.7. This profile (curve a) is based on a velocity of
15 cm s"1 at a depth of 50 cm and a shear velocity of 0.60 cm s"1. It is clear
that the viscous sublayer is less than 2 mm thick under these circum-
stances. This dimension might be compared to the dimensions of some
larvae and their sinking or swimming speeds (Table 3.2). A competent
larva of the blue mussel pediveliger is about 0.024 cm in size and can be
expected to have a typical bivalve swimming speed of -0.2 cm -s"1. It is

only in the viscous sublayer, however, that the larva can be said to
control its own trajectory. In the logarithmic layer, the velocity
fluctuations about the mean are typically about 10% of the mean value
(Butman 1986), about lcm-s"1 in the case under discussion. These
fluctuating velocities will dominate the movement of most larvae, al-
though a high settling velocity or swimming in a preferred direction
should result in a net movement across the stream lines.
The hydrodynamically rough boundary layer is characterized by very
thin viscous sublayers on the upstream faces of roughness elements and
a wake flow in their lee. The roughness elements are higher than the
depth of the viscous sublayer which would be expected in their absence.
Movement of larvae will be determined by the random motion in the
flow a short distance above the tops of the roughness elements and by
processes specific to the geometry of the roughness near the boundaries.
Similarly, wave-swept rocky shores will be characterized by thin, time-
dependent boundary layers and wakes which are difficult to generalize in
terms of larval transport and diffusion.
Larval settlement/recruitment
Before we begin an overview of the rapidly expanding body of work
which deals with the detailed hydrodynamics of settlement of competent
suspension-feeding larvae, we should recall the general ecological fea-
tures that are involved.
The life history of suspension feeders begins with spawning, and field
studies have suggested that the timing of this event is seasonally adjusted
for each species (Thorson 1950; Lacalli 1981; Hawkins and Hartnoll
1982). As an example, Lacalli (1981) showed that spawning periods in
Passamaquoddy Bay, within the Bay of Fundy, Canada, e.g. of polynoid
polychaetes (Fig. 3.8), are correlated with the timing of the spring diatom
bloom in Passamaquoddy Bay. The mechanism which allows this linkage
to be made was first suggested in the experimental work of Starr,
Himmelman, and Therriault (1990). In this work, two species were exam-
ined, one of which was a suspension feeder, Mytilus edulis. Adult blue
mussels were triggered to spawn by release of a chemically unidentified
ectocrine produced by phytoplankton cells found in the Gulf of
St. Lawrence, Canada. In the laboratory, phytoplankton such as
Skelatonema costatum, Phaeodactylum tricornutum, and Thallassosira
nordenskioldii produced sufficient ectocrine to induce mussel spawning.
J FMAMJ J A S O N D
Polychaeta
Polydora quadrilobata
Scolelepis squamata
Harmothoe imbricata
Heteromastus filiformis
Capitella capitata
Nephtys incisa
Lepidonotus squamatus
Eteone longa
Prionospio steenstrupi
Nephtys caeca
Myriochele heeri
Pectinaria granulata
Phyllodoce mucosa
Phyllodoce maculata
Pholoe minuta
Nereis pelagica
Lumbrineris fragilis
Dorvillea caeca
Laonice cirrata
Other invertebrates
Boltenia ovifera
Barnacle nauplii
Strongylocentrotus droebachiensis
Cucumaria frondosa
Tonicella marmorea
Acmaea testudinalis
Asteroid larvae (Asterias sp.j
Tubellaria (Aiaurina sp.j
Echnarachinus parma
P i l i d i a (Cerebratulus lacteus)
Figure 3.8 Seasonal presence/absence of planktotrophic larvae for some com-

mon benthic invertebrates in Passamaquoddy Bay, Bay of Fundy, Canada: solid
lines, maximum abundance; dotted lines, lesser abundance in plankton tows
(Lacalli 1981).
In M. edulis, the response was concentration dependent, requiring at

least 40 x 107 cells-L"1 to induce spawning (Starr et al. 1990). As pointed
out by the authors, such high phytoplankton densities would rarely occur
in field conditions; however, continuous suspension feeding by the mus-
sel from lower cell densities might be sufficient to concentrate enough
phytoplankton in the mantle cavity to trigger spawning.
Because of the seasonal nature of spawning in cold temperate condi-
tions, it is likely that larval recruitment will also show a strong seasonal
signal. As far as we are aware, there is limited information (e.g. Menge

1975; Strathman 1985) regarding larval survivorship during dispersal, but
before settlement. Larval settlement has been studied for a variety of
reasons, one of which concerns the economic importance of fouling on
man-made structures in the sea, such as saltwater intake pipes, ship hulls,
oil rigs, and floating mariculture pens. Suspension feeding larvae in the
period when they are near to, and just after recruitment to, the benthos
are subject to a variety of variables - physical, chemical, and biological -
which can influence the outcome of this process. In the following ac-
count, we are concerned primarily with how hydrodynamics at a scale of
<0.1km influences larval settlement.
A central question of benthic ecology is to determine ways to predict
which characteristic macrofaunal assemblage will occur in a particular
de-faunated patch of marine sediment (Gray 1974; Wildish 1977;
Underwood and Fairweather 1989). One obvious and potentially im-
portant variable in determining the spatial pattern of soft sediment
macrofaunal assemblages, as discussed in Chapter 7 (and see Chap. 6 for
a discussion of aggregation), is the supply of colonizing larvae. Larval
supply is affected by the proximity to the patch of adults producing
fertilized zygotes, the length of planktonic life, the species composition
of adults producing larvae, larval production rates, hydrodynamic effects
which affect dispersal and settlement, as well as survival of larvae during
their planktonic life. These considerations may be boiled down to the
following four ecological parameters which can indicate recruitment
success (see also Keough and Downes 1982; Luckenbach 1984):
Survivorship of the larval stage from zygote to final settlement
Success rate of larval settlement
Losses associated with post-settlement dispersal (see this chapter, the section,
Post-larval dispersal)
Early post-settlement mortality (see Chap. 3, the section, Hard substrates).
Because there is limited information available for mortality during the
planktonic phase of the life history of most suspension feeders, most of
this section will be concerned with the factors which influence larval
settlement.
Soft sediments
In a comprehensive review of invertebrate larval settlement in soft
sediments, Butman (1987) considered two major hypotheses - passive
deposition of competent larvae (number 1, Table 3.7) and active habitat
Table 3.7. Hypotheses relating hydrodynamic factors which may

influence larval settlement/recruitment of soft sediment macrofauna.
No. Description Reference

Ho Competent larval suspension and Hannan (1984), Butman
deposit feeders are not passively (1989), Pawlik et al.
co-deposited like similarly sized (1991), Pawlik and
sediment particles Butman (1993)
Hx Competent larval suspension and Snelgrove (1994),
deposit feeders are passively Harvey et al. (1995)
co-deposited like similarly sized
sediment particles
Ho Larval swimming near the sediment Butman et al. (1988),
surface has no influence on Bachelet et al. (1992)
selection of a settlement site Butman and Grassle
Hj Larval swimming near the sediment (1992), Grassle et al.
surface is guided by chemical cues (1992a,b), Turner et al.
and results in larval choice of a (1994)
settlement site
Hx Recruitment enhancement does not Eckman (1979,1983,
occur near small-scale 1985), Gallagher et al.
protuberances from the sediment (1983)
H2 Recruitment enhancement occurs Ertman and Jumars
near small-scale protuberances, e.g. (1988)
animal tubes, seagrass shoots
selection by competent larvae (number 2, Table 3.7). In the former,

purely hydrodynamic forces are considered to deliver larvae which act
like passive sedimentary particles with similar fall velocities. In the active
selection hypothesis, competent larvae are considered to be able to swim
(Table 3.2) to a preferred ecological niche. A third hypothesis - passive
deposition near sediment protuberances - is really a special case of the
passive deposition hypothesis (number 3, Table 3.7). Considering all
possible types of initial settlement pattern - random; irregular, e.g. after
incorporation of larvae in marine snow (Shanks and Edmondson 1990)
or other rapidly sinking masses; that caused by active swimming; or that
due to passive deposition - we believe that it is unnecessary to use active
habitat selection as Ho and passive deposition as H1? or vice versa, and we
have phrased the null and alternate hypotheses non-exclusively in Table
3.7. This table also lists those studies, both field and flume experimental,
which test the three hypotheses and which, for the latter, pay adequate
attention to scale and hydrodynamic similarity in simulating conditions

in the field. Butman (1987) concluded that passive deposition and active
habitat selection were not exclusive concepts because they often oper-
ated at very different scales of space and time. With regard to space,
competent larvae settle at millimetre to metre scales, with passive depo-
sition often operating at the higher end of the scale. Temporal scales of
importance are tidal (hours) and the stochasticity associated with wind-
wave motion (days to months).
As is frequently the case in marine biology, the origin of useful con-
cepts such as the passive deposition hypothesis resulted from keen obser-
vation during fieldwork. Tyler and Banner (1977), for example, followed
larval transport of the brittle star Ophiura alba, considering the larvae to
be physically sorted by local currents and passively deposited along with
fine sand in the Bristol Channel, United Kingdom. Even earlier field
observations such as those by Pratt (1953), who studied bivalve distribu-
tion in Narragansett Bay, and Baggerman (1953), who studied oyster
spatfall in the Wadden Sea, suggest the importance of passive larval
transport and deposition.
Rigorous experimental testing of the passive deposition hypothesis for
infauna originated with Hannan (1984) and Butman (1989). In these field
experiments, sediment traps of various designs were set 0.4 m above the
soft sediments of Buzzards Bay, Massachusetts, United States. The traps
had previously been calibrated in a flume, which met dynamic similarity
criteria with boundary layer-sediment conditions in the field, and in
which the relative particle collection efficiencies of larval mimics (poly-
styrene or glass spheres of the same fall velocity) had been determined.
The H : that planktotrophic larvae sink in seawater as passive particles in
turbulent flow can be tested by determining the rank order of collection
in traps compared with the a priori determined rank order of larval
mimics. Differences in trap design (see Chap. 2) caused differences in
local disturbances offlowwhich resulted in the differences observed. The
results from field deployment of the traps 1 m above bottom showed that
most larvae were collected according to a priori expectations as passive
particles, including polychaetes, bivalve post-larvae, enteropneust, and
gastropod larvae (Butman 1989). A few exceptions were found; thus the
polychaete larva of Pectinaria gouldii and the larvae and post-larvae of a
seastar (? Asterias sp.) did not settle passively. The results provide some
support for the larval passive deposition hypothesis, although they do
not provide the benthic boundary layer conditions that a larva would
experience in reaching the sediment. Further field experiments in Buz-
zards Bay designed by Snelgrove (1994) overcame this problem. Two

types of settlement tray were deployed at 20 m depth directly in a mud
habitat. One tray had an 11.3-cm-diameter by 10-cm-deep depression
filled with a 2-cm-deep layer of de-faunated muddy sediment, and the
other had a 2-cm-deep cup of the same diameter to which de-faunated
muddy sediment was added so that it was flush with the tray walls. Most
of the naturally occurring settling larvae of bivalves, gastropods, juvenile
Mediomastus ambiseta, and, on one occasion, spionid polychaetes and
juvenile nemerteans preferred the depression trays, thus supporting Hx
of the passive deposition hypothesis (number 1, Table 3.7). Only the
polychaete Capitella sp. did not prefer the depression trays. These ex-
periments were adequately replicated and, since the trays were deployed
for 3-4 d, the results suggest that initial larval settlement was involved.
A laboratory test of the passive deposition hypothesis comes from
flume experiments with the aggregating tube worm Phragatopoma
lapidosa californica. This sabellarid polychaete constructs tubes of sand
grains which can form into extensive reefs (see Chap. 6, Fig. 6.2) on the
hard substrates of the Southern California coast (Pawlik et al. 1991;
Pawlik and Butman 1993). Because the larvae are easily cultured and
delay settlement until they find sand originating from adult worm tubes
(Pawlik 1986), they are ideal for larval settlement choice experiments in
which unidirectional flume flows over sand are offered. The laminar
sublayer and logarithmic layer can be characterized by the shear or
friction velocity, /*, which is related to the shear force exerted by the
bottom. Six different flume flows of 5-30 cm s"1 were tested, correspond-
ing to U* of 0.26-1.22 cm-s"1 calculated from the "full profile" method
(Pawlik and Butman 1993). A long flume 17 m by 0.6m wide which
contained temperature-controlled seawater at 20C was used. In prelimi-
nary tests, larval mimics were used to determine bedload shear velocities
at which sediment transport was initiated. This was determined to be at
U* = 0.47-0.86cm-s1 in the conditions of this experiment. The flume
was filled to a 12-cm depth with filtered seawater and 2000 larvae added
by plastic tubing and funnel at a point just upstream of a flow straight-
ener. At U* > 1.03 cm-s"1, the larval mimics began to make brief excur-
sions into the benthic boundary layer, and, as the shear velocities in-
creased, so the number of spheres present there and distance travelled by
each mimic also increased. Only one pass by live larvae of the acrylic
settlement plate which contained four wells, each with a sand treatment,
was permitted in the later experiments (Pawlik and Butman 1993). Two
treatments were offered - either "filmed" sand, that is, aged sand coated
with a microbial film but not processed by worms, or "tube" sand, pre-
pared from the broken off tubes that had been cemented by older
worms. Results showed that there was a unimodal response between
larval settlement rates determined in 3-h-long experiments and flume
bulk velocity flows. Peak settlement occurred at bulk velocities of 15-
25cm-s~1, and at these velocities, larvae tumbled end over end, some-
times briefly adhering to the floor by means of their tentacles. At
<10cm-s"1, larvae swam near the release point, eventually swimming to
the flume surface, where they were advected downstream in the bulk
layer, so that they did not come into contact with the settlement plate. At
>30cm-s~1, larval settlement was poor, either because the shear velocity
was too high and/or because the increased turbulence meant that fewer
larvae actually came into contact with the sand treatments. In field
conditions, Phragmatopoma is much more likely to experience oscilla-
tory wind-wave flows so it is not possible to place these results in a field
context.
Observations on the free-swimming larvae of the suspension-feeding
bivalve Ceras toderma edule in still water and a Vogel-LaBarbera flume,
3.5 m long x 0.5 wide x 0.4m high, were made by Jonsson, Andre, and
Lindegarth (1991). In still water, upward helical swimming of the compe-
tent pediveliger was observed and interpreted as geotaxis although
statocysts were not involved. In unidirectional flow flume observations,
four bulk velocities were offered: 2, 5,10, and IScm-s"1. At 5-10cm-s"1,
the larvae became confined to the viscous sublayer, where they drifted in
a streamwise direction at 0.45-1.60 mm-s"1, periodically touching the
flume floor. At 15 cm s"1, bedload transport of tumbling larvae occurred
with a much higher probability of larval resuspension. Jonsson et al.
(1991) suggested that boundary-shear forces created by t/>15cm-s" 1
prevented the larvae from swimming properly. An hypothesis involving
viscous drag torque on the larval body by benthic boundary layer forces,
which can help explain settlement by C. edule, is proposed as a specific
mechanism of viscous sublayer larval confinement by Jonsson et al.
(1991).
Spat of the Iceland scallop Chlamys islandica preferentially settles on
variousfilamentousred algae and on the perisarcs of dead hydroids in the
Gulf of St. Lawrence (Harvey, Bourget, and Miron 1993). Model studies
with three-dimensional branching structures made of plastic were de-
signed by Harvey et al. (1995) to determine the factors influencing larval
recruitment. The effect of branching patterns and branch diameter of the
red alga models on scallop larval settlement was determined in field
experiments. Concurrent flume laboratory experiments using similar

algal models and simulated larvae, actually inert polyvinylchloride
microparticles of the right size, became trapped on the branches in
silicone grease. Flume experiments gave results similar to the field ones,
suggesting that passive settlement was the mechanism at a scale of
~10cm, although active processes at lower scales could not be excluded
by this study (Harvey et al. 1995).
Butman (1987) pointed out that experimental tests of the active habi-
tat hypothesis prior to her review were nearly all larval choice experi-
ments in still water. Most experiments were conducted at inappropriate
spatial scales of <10cm so that they were not simulating field conditions.
Nevertheless, this was the "ruling hypothesis" of this subject (number 2,
Table 3.7) prior to 1987, with much compelling evidence that chemical
cues were utilized during the settlement of competent larvae (Pawlik
1992).
The first experimental test in which realistic hydrodynamic conditions
were simulated during larval choices to set on different sediment types
was presented by Butman, Grassle, and Webb (1988). This work in-
volved comparisons of a still water test and unidirectional flow in a
recirculating flume. The annular flume had a 6-m-long working section
x 0.5 x 0.3 m deep, total volumetric capacity of 1000 L, and flow induced
by paddle wheel. Competent larvae introduced to the flume were given
settlement choices in 4 x 4 x 1-cm-deep compartments filled with sedi-
ment. Two choices - a natural mud and an abiotic glass bead mixture
similar in grain size (-70% < 63|im) to the mud - were offered. Compe-
tent larvae tested included Capitella sp. 1, a subsurface deposit feeder
associated with organic-rich mud, and the suspension-feeding hard clam
Mercenaria mercenaria. Bachelet et al. (1992) discovered a confounding
factor in those experiments with the hard clam that was due to decompo-
sition of the larval shell on preservation in acidic conditions, which led to
an underestimate of larval numbers in all mud treatments. These results
will not be discussed further, although Bachelet et al. (1992) showed that
the hard clam larvae showed no preference over mud and clean sand in
still water conditions. For Capitella sp. 1, Butman et al. (1988) found that
for the one flow tested, equivalent to U* = 0.30 cm -s"1, the worm larvae
were able to choose the ecologically appropriate muddy sediment. In a
continuation of these studies with Capitella, Butman and Grassle (1992)
and Grassle, Butman, and Mills (1992a), in choice experiments at flows
of 0,5, and 15 cm s"1 (equivalent to boundary shear velocities of 0.26 and
0.64 cm-s"1), provided further evidence that the larvae of this worm can
select highly organic sediments at up to the highest shear velocity tested.

It is of interest that at U* = 0.64 cm-s"1, greater numbers of larvae settle
in the 2-h flume test than in still water. This is at least in part due to the
greater flux at higher flows, which allows more larval contacts with
sediments in a limited test time.
Using the same annular paddle wheel flume, settlement experiments
were conducted with pediveliger larvae of the coot clam Mulina lateralis
by Grassle et al. (1992b). In 24-h experiments, the clam larvae chose an
organically rich natural mud over abiotic glass beads of approximately
the same particle size in a free-stream flow of ~5cm-s~l and
[/* = 0.30 cm s"1. At zeroflow,the clams made similar significant choices
to those in flow because of their helical swimming abilities. Grassle,
Snelgrove, and Butman (1992b) conclude that the coot clam pediveliger
can exercise habitat choice by swimming, but that older pediveligers
which are less efficient swimmers require ambient flows to transport
them between sediment patches.
Further experiments in the same annular flume with larvae of M.
lateralis and Capitella sp. were designed by Snelgrove, Butman, and
Grassle (1993) to test the relative importance of active selection versus
passive deposition mechanisms. In these laboratory experiments, differ-
ent size depressions in an acrylic plastic (Plexiglas) settlement plate were
variably filled with organic-rich mud or glass beads of the same particle
size. Both species generally chose mud over glass beads, although in M.
lateralis in some treatments, larval settlement was greatest among glass
beads. This "wrong" behavior was probably due to physical trapping in
the depression, i.e. physically controlled or passive dispersal. Because M.
lateralis pediveligers are poor swimmers, they were unable to swim out of
the depressions and settle in the appropriate mud-filled depression.
These results were shown to be scale dependent, and, hence, the rel-
evance of the results in field conditions was not made clear.
One question of importance in regard to hypothesis 2 (Table 3.7) is
whether the chemical cue is adsorbed on a suitable surface, as suggested
in the work on Capitella and Mercenaria by Butman et al. (1988), or is
waterborne, as proposed by Zimmer-Faust and Tamburri (1994) in work
on oyster larvae of Crassostrea virginica. Other chemical inducers of
larval settlement are reviewed in Pawlik (1992) and all are lipophilic and
substrate bound. A laboratory test of the water-soluble chemical hypoth-
esis was conducted by Turner et al. (1994) with C. virginica larvae in
flows of 2 and 6cm-s~1 at U* < 0.25 cm -s"1. The experiments were set up
in a manner similar to those previously described: tests in both still water
and an annular, recirculating flume of 3516-L capacity. The settlement

choice area was an acrylic plastic (Plexiglas) plate in which circular wells
of 6.98 cm diameter 1.27 cm deep were made with a small hole in the
bottom so that solutions were continuously injected at 2ml-s~1 into each
well. The wells were filled with crushed oyster shell which had been
cleaned and dried before use. A low molecular weight peptide at concen-
trations of 10~6 to KT10M glycl-glycl-L-arginine was the chemical tested
because Zimmer-Faust and Tamburri (1994) had found in still water
tests that it induced oyster larval settlement. The oyster larvae clearly
preferred to settle in wells with water-soluble cues over those without the
chemical. The chemical inducer caused rapid downward swimming or
sinking (>0.3 cm s"1) of the larvae, resulting in concentration of the lar-
vae in near-bottom water, and hence enhanced larval contact and settle-
ment. Thus, the larval response does not appear to depend on a chemical
gradient but to be triggered at low threshold concentrations of <10~10M
glycl-gylcl-L-arginine, which caused more rapid swimming. The applica-
bility of this result in the field, of course, depends on whether the chemi-
cal concentrations simulated are realistic and what happens in more
energetic flows.
The study of how small-scale protuberances from the sediment surface
affect larval recruitment was initiated by Eckman (1979). These studies
utilized field experimentation and observation to investigate the distri-
bution of recently settled recruits to an intertidal sandflat in Puget
Sound, United States. Eckman showed that increased abundances of
recently settled spat occurred within 1 cm of the tubes of polychaetes,
amphipods, and tanaids. Recruitment enhancement experiments involv-
ing model tubes in natural sediments suggested that the resultant en-
hancement was controlled by localized flow-tube-sediment interactions.
These relationships were investigated with small (in relation to benthic
boundary layer thickness) tube mimics in flume studies by Eckman and
Nowell (1984). Whether resuspension or deposition near the tube occurs
depends on the tube height and the boundary shear stress imposed by the
flow. Characteristic horseshoe patterns of sediment erosion occur at high
boundary shear stress on the upstream side of the tube, with discrete
depositional areas a few tube diameters downstream of the tube. Re-
cruitment enhancement occurred in depositional areas downstream of
the tube, although it was not possible with the experimental design to
determine whether local transport of larvae, spat mortality, or move-
ment subsequent to settlement was the cause. Eckman (1983) set up field
experiments in which de-faunated sand cores were placed at control or
treatment sites on a Puget Sound sandy beach. Each site had a core at the
centre, with either a bare sand control or a treatment site surrounded by
different densities of simulated marsh grass stalks. The latter were actu-
ally 0.6-cm-diameter plastic straws placed at three densities of 56-2500
straws m2. Coincident experiments in a recirculating flume where three
velocities, equivalent to U* = 0.52-1.67 cm-s"1, and similar straw densi-
ties to the field manipulations were set up. Enhanced recruitment in the
field was found for harpacticoids, amphipods, facultative deposit-sus-
pension feeding, tube-living polychaetes, and meiofauna. The results
closely follow those predicted from flume observations of boundary
shear stresses around rigid cylinder mimics and thus support Hx of hy-
pothesis 3 (Table 3.7). A similar type of field experiment with tube
mimics at densities less than that causing skimming flow showed that
enhancement of bacterial colonization occurred in the downstream
wakes (Eckman 1985). Bacterial colonization may be important for lar-
val recruitment since a bacterial film may stimulate larval settlement
(e.g., Crisp 1974).
Although the field experiments of Gallagher, Jumars, and Trueblood
(1983) were designed to test the well known conceptual model of
Connell and Slatyer (1977) of mechanisms of succession (inhibition,
tolerance, or facilitation), the results provide support for the passive
deposition hypothesis (number 3, Table 3.7). The experiments involved
placing cores containing clean sand into 10-cm2 patches to which live or
simulated tubes of the common tube builders in the Puget Sound, soft-
sediment intertidal area were added. Either live or simulated tubes en-
hanced the recruitment of other taxa to the de-faunated cores in com-
parison to that of controls without tubes, suggesting the physical nature
of the mechanism involved.
The hydrodynamic characteristics of the exhalant jet of a cockle from
Puget Sound, Clinocardium nutalli, were studied by Ertman and Jumars
(1988) in a 2.5-m-long-working-section recirculating flume. Exhalant
velocity in Clinocardium was estimated as 9-llcm-s 1 and patterns of
flow around it varied with ambient velocity in much the same way as
around the animal tube mimics studied by Eckman and No well (1984).
Polystyrene beads with the same settling velocity as that of narcotized
larvae were injected into the exhalant flow of an actively pumping cockle
and the sediment near it sampled with 0.6-cm-diameter plastic drinking
straws as corers. The free stream velocities tested ranged from 3.3 to
16.0cm-s"1, which was equivalent to [/* = 0.19-0.78cm-s"1. The results
were consistent with the passive deposition hypothesis (number 3, Table
3.7) because the downstream distribution of polystyrene beads showed

localized patterns in the downstream wake of the flow, as indicated by
the high variance of the measured densities of recruits. Similar results
were found by field sampling downstream of the exhalant of resident
soft-shell clams Mya arenaria in Puget Sound.
Hard substrates
Just as for soft sediments, for benthic ecologists interested in hard
substrates an important question is the prediction of the type of
macrofaunal assemblage which will develop on a particular de-faunated
area of rock or coral. Like those of larvae settling on soft sediments, the
ecological parameters of primary concern for hard substrate recruitment
are planktonic survivorship, larval settlement rate, and post-settlement
mortality. One way that larval settlement of suspension feeders of hard
substrates appears to differ from that of soft sediment is the absence of
post-settlement dispersal in the former. Hypotheses which have con-
cerned researchers in this field are similar to passive deposition (number
1, Table 3.8) and active habitat selection (numbers 2 and 3, Table 3.8)
of soft sediments. An additional hypothesis, which is more ecological
in outlook, has focused on larval supply and settlement rates, which
are proposed to control the final densities achieved (number 4, Table
3.8). This approach has been referred to as "supply side ecology"
(Roughgarden et al. 1987; Underwood and Fairweather 1989), although
many of the studies in support of it have not adequately separated larval
settlement rates from early post-settlement mortality rates.
Much of the field and laboratory experimental work on hard substrate
settlement has been done with the competent settling stage, or cyprid, of
cirripede barnacles. Crisp (1955) conducted settling experiments with
cyprids in 1.5-m-long, smooth-walled, glass tubes of from 2- to 10-mm
internal diameter. Water flow rates were varied by adjusting a constant-
head device and recording the volumetricflowrate. All of theflowswere
laminar. The velocity gradient at the wall of the tube can be related to
the average velocity and the tube diameter. The friction velocity, /*, is
related to the velocity gradient at the wall by
= E/*V1
dr
)o
Friction velocities have been calculated for the data of Crisp (1955); they
are shown in Table 3.9. The experiments involved introducing individual
Table 3.8. Hypotheses relating hydrodynamic factors which may

control larval recruitment of hard substrate suspension feeders.
No. HQ/H! Description Reference
1 Ho Larval settlement is independent Crisp (1955), Mullineaux

of benthic boundary layer flow and Butman (1990, 1991)
Hx Larval settlement is passively Mullineaux and Garland
controlled by benthic boundary (1993), Abelson et al. (1994)
layer flow
2 Ho Competent larvae settle Wethey (1986), LeTourneux
indiscriminately and Bourget (1988)
Hj Competent larvae behaviorally Walters (1992)
choose particular hard substrates
on the basis of surface roughness
3 Ho Competent larvae settle Rittschoff et al. (1984)
indiscriminately Neal and Yule (1994)
Hx Competent larvae behaviorally
choose particular hard substrates
on the basis of chemical cues
which are either waterborne or
bound to surfaces
4 Ho Larval supply rates are Keough and Downes
independent of the subsequent (1982), Bushek (1988)
density of rocky shore recruits Bertness et al. (1992),
Hj Larval supply rates control the Miron et al. (1995)
subsequent density of rocky
shore recruits
cyprids into the tube in a pipette and then observing their passage in a
1-m section of the tube. Results were expressed as a percentage of those
tested which made contact with the tube walls as a function of flow
velocity. Larval initial contacts were unimodally related to a range of
tube flow velocities tried, equivalent to velocity gradients up to 1000 s"1
for two species of barnacle, as shown in Table 3.9. Maximum attachment
was at velocity gradients up to, but less than, that in which the swimming
cyprid could maintain its position by swimming against the current. Peak
settling, b, occurred at -75 s"1 for Semibalanus balanoides and -40 s"1 for
Elminius modestus. In stage a response, the larvae did not swim and were
moved passively. Flows greater than 50 s"1 caused the cyprids to swim and
attach. Presumably in the c stage, increasing shear made it harder for the
cyprid to attach, until at >400cm"1 for S. balanoides and >700cm~1 for E.
Table 3.9. Summary of initial settling responses of barnacles in laminar

tube flow experiments.
Response Velocity gradient U*

Species stage (cm-s"1)
Semibalanus balanoides a <75 <0.87
b 75 0.87
c 75-400 0.87-2.00
d >400 >2.00
Elminius modestus a <40 <0.63
b 40 0.63
c 40-700 0.63-2.65
d >700 >2.65
Source: Crisp (1955).
modestus no further larval attachment was possible. The hydrodynamic

conditions in the experimental tube in these experiments do not simulate
natural benthic boundary layer conditions very well, although they do
show the importance of the velocity gradient to initial contact in line with
Hi of hypothesis 1 (Table 3.8). Crisp's experiments are among the few
discussed for larvae that adequately consider a wide enough velocity
range.
The technique of using thin and bluff plates to create particular hydro-
dynamic conditions on surfaces without changing the plate surface rough-
ness was introduced by Mullineaux and Butman (1990) in field recruit-
ment experiments at a deepwater seamount of 410 m depth in the central
North Pacific. Circular plates were used in this work, and the recruitment
results suggested that settlement had been influenced by small-scale
boundary layer flows over the plates. Unidirectional flume flows at free-
stream velocities of 5 or 10 cm s"1 were used for visualization of the flow
over rectangular settlement plates and for behavioral studies of settling
by barnacle cyprids of Balanus amphitrite by Mullineaux and Butman
(1991). The objectives of this study were to determine whether and how
larval settlement was influenced by passive benthic boundary layer phe-
nomena, and whether active behavioral processes were involved after
initial larval contact. Three types of polycarbonate plate, 10 x 26 cm long,
were oriented so that the flat surface was parallel to theflumeflow.Each
type of plate produced characteristically different streamlines dependent
on its shape, as shown in Fig. 3.9. About 20,000 competent cyprids were
Faired Plate
Bluff Plate
Split Plate
c.
Figure 3.9 Streamlines over three types of settlement plate placed horizontal
with respect toflumeflow(Mullineaux and Butman 1991).
placed in the ~1000-L-capacity flume and allowed to recirculate past the

plates which had been pre-soaked with adult barnacle extract, known to
induce settlement chemically (Crisp and Meadows 1963). Cyprid settle-
ment was recorded by video viewing of an area 10 x 10 cm during a 48-h
experimental period. Typical results for the bluff plate are shown in Fig.
3.10. Initial cyprid contacts were found to be negatively correlated with
shear stress and vertical advection in support of Hx (hypothesis 1, Table
3.8). After initial contacts, many cyprids began crawling with the aid of
their antennules, and videophotography showed that downstream of the
re-attachment point they moved downstream, whereas above this point
they moved upstream with respect to the bulk flow. As Fig. 3.9 indicates,
near-bed streamlines are in the same direction as larval movement, and
this behavior results in two peaks of settlement (Fig. 3.10B), although
cyprids could, and did, actively swim off the plate. This was not due to
vertical advection
CO
< ."* "V
Q
first contact
settlement
DISTANCE FROM LEADING EDGE

Figure 3.10 Diagram of boundary layer flow conditions over a bluff plate: A,
vertical advection (dashed line) and shear stress (solid line); B, initial contact
points (solid line) and final settlement location (dashed line) of barnacle cyprid
larvae; arrow, reattachment point of the eddy; see also Fig. 3.9 (redrawn from
Mullineaux and Butman 1991).
erosion of the cyprids from the substrate because Eckman et al. (1990)
have shown that the critical drag force required to dislodge cyprids is not
reached at bulk flows of lOcm-s"1. The flume studies of Eckman et al.
(1990) showed that the force required to detach B. amphitrite cyprids was
positively correlated with the duration of attachment in periods of up to
3h.
Mullineaux and Garland (1993) reported on flume and field work with
rectangular settling plates of similar design to that used by Mullineaux

and Butman (1991), but with the addition of a 30 angled plate type.
Hydrodynamic characteristics of each type of settlement plate were de-
termined in flume flows followed by deployment on polyvinylchloride
(PVC) poles so that the plates were raised 1.2 m above the sediment at
sites near Woods Hole, Massachusetts, United States. The settlement
plates were placed into a swivel by SCUBA divers. Flows of ca.
>2 cm s"1 were enough to cause the swivel to turn so the leading edge of
each plate opposed the flow. A hydrodynamic model of passive settling
by planktonic larvae onto hard substrates presented by Eckman (1990)
could be used to analyze the field results obtained by Mullineaux and
Garland (1993). The model considers a depth limited boundary layer,
boundary shear velocities, larval fall velocities, as well as roughness
features of the substrate. Settlement responses of the locally abundant
larvae were characteristic for each species. Thus, the hydroid Tubularia
croeca settled in regions of high turbulence and strong shear stress,
whereas the bryozoan Schizoporella unicornis was associated with re-
gions of high shear only. Another bryozoan, Bugula turrita, settled in a
region of reduced shear. The results obtained by Mullineaux and
Garland (1993) are consistent with Hx of hypotheses 1 and 2 of Table 3.8.
Abelson, Weihs, and Loya (1994) reported some interesting work on
the settlement choices made by four species of coral larvae found at
Eilat, Israel. Experiments were conducted in a special recirculating
flume, termed a nozzle-diffuser flow tank, in which an expansion section
caused deceleration and acceleration flow patterns. Two of the species,
Seriatopora caliendrum and Alveopora daedalea, settled on the flume
walls, in either non-accelerating or decelerating zones, e.g. in the down-
stream wakes of tubes, and two species, Litophyton arboreum and
Dendronephthya hemprichi, preferentially settled only in the accelerat-
ing part of the flume expansion section or on protruding tubes. The latter
species are found in natural conditions on protruding body substrates so
the flume results are consistent with field observations. During the ex-
periments, all four of the larval species were passively deposited by
decelerating flows, although the two species found on protruding tubes
actively swam back into the flow and delayed settling until an appropri-
ate settlement site was found. The mechanism of settlement on protrud-
ing bodies involved secretion of aboral, adhesive mucous threads up to
100 times the larval body length, which allowed nearly instantaneous
attachment to the substrate.
Shown in Table 3.10 are some examples of settlement or recruitment
Table 3.10. Selected larval settlement or recruitment experiments on hard substrate where flow is simulated (S) or
ambient flow (F) is available.^
Species Hypotheses tested Flow conditions Reference

BARNACLES
Semibalanus balanoides, Elminius modestus 1 S Crisp (1955)
Balanus amphitrite 1,3 S Rittschof et al. (1984)
S. balanoides 1,2,3 F Wethey (1984)
S. balanoides, Chthamalus fragilis 1,2 S,F Wethey (1986)
S. balanoides 1,2,3 F LeTournex and Bourget (1988)
S. balanoides 1,2,3 F Chabot and Bourget (1988)
B. amphitrite 1 S Eckman et al. (1990)
B. amphitrite 1,2 S Mullineaux and Butman (1991)
B. amphitrite (Bugula neritana) 1,2,3 F Walters (1992)
B. perforatus, E. modestus 1,3 S Neal and Yule (1994)
Balanus sp. 1,2 F, S Lemire and Bourget (1996)
B. crenatus 1,2,3 F, S Miron et al. (1996)
CORALS
Local coral spat 1,2,3 F Carleton and Sammarco (1987)
Local coral spat 1, 2, 3 + light F Maida et al. (1994)
Senatopora caliendrum 1 S Ableson et al. (1994)
Alveopora daedalea 1 s Ableson et al. (1994)
Litophyton arboreum 1 s Ableson et al. (1994)
Dendronephthya hemprichi 1 s Ableson et al. (1994)
a
Refer to Table 3.8 for hypotheses tested.
experiments where flow was adequately simulated in flume flows and

field experiments where ambient water movement was utilized. The
rugophilic behavior of barnacle cyprids was demonstrated by Crisp and
Barnes (1954) in still seawater experiments. Some tests of responses to
surface roughness have involved field experiments in ambient flow. A
difficulty inherent in field experiments where most factors influencing
the outcome are not controlled is in identifying the causal factor(s). To
overcome this, Wethey (1986) used plastic models made from casts of
adult and juvenile barnacles on rock surfaces which were cemented to
natural rock surfaces on Long Island. The models were examined daily
by photography and -30% of S. balanoides and Chthamalus fragilis
cyprids settled at the same microsites, low in shear stress, on three
replicated plates, and - at least initially - ecologically relevant chemical
cues were absent. Because of the nature of these experiments, it was
impossible to determine whether passive or active processes in settle-
ment were involved (hypotheses 1 and 2, Table 3.8). The experiments
designed by Walters (1992) in a low energy environment off the coast of
North Carolina were an attempt to overcome the difficulty of distin-
guishing passive and active settlement processes. The plastic settlement
plates were 8x8cm with 3-mm-high, 4.5-mm-diameter bumps which
were separated by 3-mm spaces. Half of the plates were treated with a
thin film of silicone vacuum grease, thereby trapping settling larvae at
the place of initial passive contact; half were untreated, thereby indicat-
ing subsequent active movement during the 4-d experimental period.
The metamorphosis of two species selected for study, Balanus amphitrite
and Bugula neritina, occurred most frequently around the bases of
each bump on untreated plates and was significantly different from
the passive deposition observed on greased plates. Active larval crawling
on a millimetre to centimetre scale after settlement explained these
results since differential post-settlement mortality or passive erosion
could be ruled out as explanations. These results were confirmed by
using similar methods for Balanus sp. cyprids and by comparing field
experiments on the North Carolina coast and flume experiments with
larval mimics to separate active and passive processes (Lemire and
Bourget 1996).
The field settling study of LeTourneux and Bourget (1988) involved
plastic settling plates of 100 x 30 cm and observations at metre, millime-
tre, and micrometre scales. The settling of S. balanoides was observed at
two locations: in the Bay of Fundy and in the St. Lawrence estuary. For
the smallest scale observed, surface microheterogeneity was significantly
greater at the site occupied by the cyprid in comparison with vacant sites.
The results suggest that larvae can discriminate microheterogeneity
down to a length of 35|im, which is approximately the size of the
anntenular disc used by the cyprid both to explore and to attach to the
substrate. Again, it is not possible to decide from these field observations
whether the results are due to a passive or active process.
In common with earlier writers who used non-flowing larval bioassays
(e.g. Crisp and Meadows 1963), Rittschof, Sansford Branscomb, and
Costlow (1984) assumed that the chemical cues associated with settling
were adsorbed onto surfaces. The method they used to demonstrate
chemical attractiveness was similar to the glass tube bioassay for cyprid
settling in flow of Crisp (1955). B. amphitrite larvae were used in these
tests with a constant flow rate of 6.8ml-min"1 in a 3.25-mm-internal-
diameter tube, which gave a velocity gradient 0.5 mm from the wall of
39.4 s"1. The experiments carried out by Rittschof et al. (1984) showed
that rinsing the glass tube with 2.5 jug ml"1 of settlement factor, consisting
of ground-up, whole B. amphitrite and adjusted to 50|ig protein-ml"1,
significantly increased settlement in young cyprids compared to a non-
rinsed or control tube rinsed with bovine serum of the same protein
content. Besides attractant chemicals from conspecifics of poorly defined
identity, biofilms on surfaces have been identified as a settlement cue.
Thus, enhancement of settlement of B. amphitrite cyprids (Maki et al.
1990) or inhibition (Maki et al. 1988) by biofilms may occur.
Neal and Yule (1994) designed an experiment in which B. perforatus
and E. modestus cyprids, after initial contact on two contrasting biofilms
- high (83 s"1) and low (15 s"1) shear stress - were tested for the tenacity
of their attachment to each substrate. Thefilmswere developed on cover
slips influmeflowsover a 2-mo. period. The high shear stress biofilm was
thinner and denser than the other and resulted in increased cyprid tenac-
ity. The low shear stress substrate was relatively thicker and looser and
caused cyprids to attach with lower tenacity. In these experiments, it was
not clear how the benthic boundary layer flow would influence the re-
sults, although both cyprid species tested can obviously choose between
the substrates offered.
In combined flume and field experiments in the St. Lawrence estuary,
Miron et al. (1996) were able to examine both passive and active controls
of initial settlement of barnacle cyprids. To distinguish between hydro-
dynamic and chemical settlement cues, responses by Balanus crenatus
cyprids to either live or ceramic model adult barnacles were compared. It
was found that spat were absent on control model barnacle panels but
present on panels with live barnacles, suggesting the importance of a

chemical cue to final settlement location. Flume tests suggested that
initial settlement was passive, but the cyprids could swim away if appro-
priate cues were absent. Live cyprids actively moved after settlement,
using chemical and roughness cues to accept or reject a settlement site
(Miron et al. 1996).
An important question, to which we will return later in Chapter 6, is,
How do groups of suspension feeders become aggregated? One mecha-
nism - gregarious settlement - will be considered here in more detail. We
use the term gregarious in a general sense to apply to aggregated larval
settlement that results from any active or passive aggregating mecha-
nism. It is thus used here differently than by Havenhand and Svane
(1991), who consider gregarious larval settlement to imply active move-
ment and a specific chemical cue. Reviews by Meadows and Campbell
(1972) and Burke (1986) also present the view that gregarious larval
settlement must involve chemicals as pheromones from conspecifics.
A further field study in the Bay of Fundy and Gulf of St. Lawrence
concerning settling of B. balanoides was presented by Chabot and
Bourget (1988). One aim of this study was to determine the effect of
post-settlement conspecifics on the settlement rate. In the Bay of Fundy,
settling S. balanoides larvae preferred open vertical rock faces, the re-
verse of that found in the Gulf of St. Lawrence and elsewhere. At smaller
spatial scales, of cracks <1.5cm deep, settlement increased as adult den-
sity increased up to a maximum when 22% to 30% of the rock was
covered with adult barnacles; at greater densities, settlement rates de-
creased. Crisp (1974) considered that secretions containing arthropodin
were the chemical cues for this behavior after first contact and subse-
quent crawling.
It was suggested by Keough (1984) that larvae of the bryozoan Bugula
neritina settled gregariously. He devised settlement experiments to de-
termine the mechanism in static glass dishes and used aggregation statis-
tics to show that sibling larvae settled in an aggregated pattern, while
unrelated larvae settled randomly. These tests showed that settling lar-
vae responded only to conspecific larvae and not to conspecific juveniles.
Many colonial suspension feeders and some sponges, cnidarians,
bryozoans, and tunicates are known to have self-/non-self-recognition
mechanisms associated with local competition and the ability to fuse if
the adjacent colonies are genetically related (Keough 1984), so a kin
recognition system has already been identified in these species.
In field experiments, Havenhand and Svane (1991) were the first to
show that an hydrodynamic cue could be involved in gregarious settle-

ment, leading to the broadening of the meaning of the term that we have
adopted. In laboratory static bioassays, these authors showed that larvae
of the solitary ascidian Ciona intestinalis settled randomly on a smooth
plastic surface and that extracts of adult body tissues could not induce
gregarious settlement. The field recruitment experiments were con-
ducted in the Gullmarsfjorden, Sweden, with 16 x 16-cm plastic settle-
ment plates to which were attached real and mimic adult C. intestinalis at
one or five adult Ciona per plate. Recruitment around either real or
mimic Ciona at five per plate was significantly higher than that on a bare
plate. The results of Havenhand and Svane (1991) are consistent with the
hypothesis that hydrodynamic rather than chemical cues, operating ei-
ther at the time of initial settlement or at some point following this event
and involving post-settlement movement, are the primary determinants
of gregarious settlement in Ciona intestinalis.
The final hypothesis, number 4 in Table 3.8, we consider quite differ-
ent from the mechanistic nature of the first three because it is ecological
in character and provides a link to some later chapters (Chaps. 6, 7, 8).
The ecological parameters of importance during early life history stages
of suspension feeders are larval supply, settlement, and early post-
settlement mortality rates. Methods for determining these rates for a few
available species are presented in Chapter 2. As suggested by Bertness et
al. (1992), proper consideration of "supply side ecology," a concept
originating with Thorson (see Olafsson, Petersen, and Ambrose 1994),
requires independent and coincident measurement of all three param-
eters for a given species and at a number of different sites.
Keough and Downes (1982) were among the first to distinguish the
possible importance of larval supply and settlement and post-settlement
mortality rates in establishing the spatial structure of the macrofaunal
association of a hard substratum. Settlement plates were employed on
the southern California coast, coupled with predator exclusion cages
over half the panels, to indicate post-settlement mortality, but their work
did not coincidentally monitor larval supply of locally dominant
bryozoans and polychaetes. The study of Gaines and Roughgarden
(1985) on central California coasts included settlement and post-settle-
ment mortality rates of Balanus glandula, but not a measure of larval
supply rates in the coastal waters.
At various sites in Galveston Bay, Texas, United States, Bushek
(1988) measured settlement on ceramic tiles and larval supply by quan-
titative plankton sampling. The species involved were the oyster
Crassostrea virginica and the barnacle Balanus eburneus. The pattern of

zonation on pilings in Galveston Bay was barnacles high and oysters
lower in the intertidal zone. The zonation was suggested by Bushek
(1988) to be controlled by larval supply rates and preferences for larval
settlement in high and low water motion, respectively. This author also
considered juvenile growth with respect to water motion measured by
gypsum dissolution, showing that there is a positive linear relationship
for the barnacle, but no correlation with oyster growth. Two populations
of Semibalanus balanoides found on the rocky shore of Rhode Island
were compared by Bertness et al. (1992). Larval sampling was by espe-
cially designed sediment traps in which larvae collected were preserved
in formalin. Larval settlement was determined on a 5 x 5-cm cleared rock
quadrat, some cleaned daily and others allowed to accumulate recruits,
thereby giving an estimate of post-settlement mortality rates. This study
provides a clear appreciation of the interplay of all of the three poten-
tially important early life history parameters for these two Rhode Island
locations. Thus, at the high settlement rate site, saturation of settlement
sites meant that cleared sites (such as experimentally cleared quadrats)
received 15 times more settlers than those at the low settlement site.
However, recruitment of juveniles to the high settlement site was compa-
rable to that at the other site because of a significantly greater density-
dependent, post-settlement mortality rate at the former. Both field re-
sults provide support for Hx in hypothesis 4 of Table 3.8.
The correlation between supply of competent larvae of Semibalanus
balanoides and daily settlement of cyprids was tested daily on a wooden
substrate fixed vertically to a wharf at St. Andrews, New Brunswick,
Canada, by Miron, Boudreau, and Bourget (1995). The pattern of settle-
ment was assessed daily, followed by removal of the spat. Water samples
at five different depths, including the neustonic surface, were pumped at
400L-min~1 so that 2000 L was filtered through a 64-jnm plankton net for
each depth. The best correlation was found to be with bottom-collected
larvae in the plankton and settled larvae in the low intertidal, supporting
Hx of number 4 in Table 3.8. Increased variability was found by Miron et
al. (1995) if integrated larval abundance at all depths was used to corre-
late with settlement numbers.
Summary
From the foregoing overview of larval research of benthic suspension
feeders, we conclude that larvae, rather than post-larvae, are the most
important means of dispersal and recruitment. We identify two other

passive dispersal methods, rafting and synantrophic transport, which
may also be important and complementary in dispersal of some species
or at particular times in their ontogeny.
During dispersal, a larva a few hundred micrometres in size may be
transported distances >1000km. Yet, during settlement, larval crawling
over distances of a few millimetres may be crucial in allowing the larva to
recruit successfully to the benthos, rather than being swept away by the
flow. Because of these considerations, temporal and spatial scale are of
primary importance in defining a field or laboratory study concerning
larval biology.
The importance of physical oceanographic processes at scales >0.1km
during larval dispersal has been demonstrated by the work discussed
herein. We can conclude that each of the alternate hypotheses in Table
3.3 is supported by some evidence, although the applicability of the result
may be limited and localized in extent. The evidence for estuarine reten-
tion and shoreward advection - both of which presuppose appropriate
directed behavioral responses by swimming larvae - is circumstantial. To
render the evidence more direct, it is necessary to set up a field experi-
ment with live larvae to determine their swimming patterns. Shanks
(1995) conducted one such field experiment by releasing large decapod
megalopa larvae, which were observed directly by SCUBA divers. He
found that some species swam in surface seawater in the direction of the
sun's bearing, while others swam near the surface and parallel to the
prevailing flow direction. These swimming patterns of the decapod lar-
vae helped them to remain in surface seawater, where they were trans-
ported shorewards by surface currents. It should be possible to extend
this type of field experiment to smaller suspension-feeding larvae using
live or dead individuals identified by vital staining or other means, so that
they could act as models during dispersion studies.
Because of limitations imposed by the methods used, most larval
studies reported from within plant canopies measured recruitment
rather than settlement rates. In such studies, it is not possible to deter-
mine the cause of recruitment as settlement or early post-settlement
mortality. Only the study by Wilson (1990), who measured larval catches
on a daily basis, could be used to indicate larval settlement rates. The
results suggest that it was indeed that larval settlement occurred at a
higher rate within the eelgrass meadow than on nearby bare sand.
The possibility that small benthic juvenile post-larval stages are impor-
tant in further dispersal has been confirmed in some bivalves but has not
been adequately investigated in other species. Its importance is indicated

by the degree to which it affects the productivity of populations, e.g. in
mussels and clams (number 1, Table 3.6). Whether flow-sediment in-
teraction influences juvenile mortality rate during dispersal (number 2,
Table 3.6) does not appear to have been adequately considered, and this
possibility remains to be tested.
The central question of both fundamental and applied research in
larval biology is focussed on the environmental factors which influence
initial settlement and recruitment at spatial scales of <0.1 km. The impor-
tance of both passive deposition and active habitat selection by soft-
sediment benthic animals (Table 3.7) has been confirmed by both field
and laboratory experiments. These experiments have mainly been done
with bivalve and polychaete larvae. Both mechanisms are operating
simultaneously but at different spatial scales (Butman 1987), with active
habitat selection common at the lower end of the scale (micrometres to
centimetres). The response of competent larvae to chemical inducers -
either adsorbed on surfaces or free in seawater - has also been confirmed
in laboratory experiments which include flow among the variables.
With regard to small-scale protuberances, field and laboratory experi-
ments strongly suggest that settlement is passively controlled over
millimetre distances downstream from either tubes or bivalve exhalant
jets. Thus, passive deposition mechanisms may also include very small-
scale phenomena within the benthic boundary layer.
For hard substrate suspension feeders, the equivalent hypotheses
(Table 3.8) to passive deposition (1) and active habitat selection include
choice on the basis of surface roughness (2) and on the basis of chemical
cues (3). Both field and laboratory experiments to test the passive depo-
sition and active habitat selection hypotheses have been undertaken
mainly with barnacle cyprids, although in some cases, coral reef larvae
were the experimental subjects. Evidence supporting the alternate in the
first three hypotheses of Table 3.8 is presented, and the importance of
chemical and surface roughness cues confirmed in the presence of natu-
ral or simulated flows.
Evidence for a kin recognition mechanism operating among some
bryozoan larvae at settlement, given by Keough (1984), explained aggre-
gation for this species. There is further evidence from species which use
chemicals or pheromones from conspecific larvae to aid in gregarious
larval settlement. For solitary ascidians, Havenhand and Svane (1991)
showed that hydrodynamic cues were important at settlement. The rela-
tive importance of these active and passive mechanisms among a wider
range of suspension feeders which aggregate remains to be assessed. The

subject of aggregation is further considered in Chapter 6, where Table
6.8 lists some hypotheses in which the mechanisms, enhanced settlement
or differential post-settlement mortality rate, are formally proposed.
Considering the currently popular question concerning supply side
ecology (number 4, Table 3.8), a number of authors have confirmed the
alternative hypothesis that absolute larval supply rates control the subse-
quent density of barnacles, although we may expect species-specific re-
sponses here. The coincident measurement of larval supply, settlement,
and early post-settlement mortality rates perhaps heralds a new and
welcome quantitative approach to the ecology of early life history stages
of suspension feeders.
References
Abelson, A., D. Weihs, and Y. Loya. 1994. Hydrodynamic impediments to
settlement of marine propagules, and adhesive filament solutions. Limnol.
Oceanogr. 39: 164-169.
Ackerman, J. D., B. Sim, S. J. Nichols, and R. Claud. 1994. A review of the
early life history of zebra mussels (Dreissena polymorpha): comparisons
with marine bivalves. Can. J. Zool. 72: 1169-1179.
Bachelet, G., C. A. Butman, C. M. Webb, V. R. Starczak, and P. V. R.
Snelgrove. 1992. Non-selective settlement of Mercenaria mercenaria (L.)
larvae in short-term, still-water laboratory experiments. J. Exp. Mar. Biol.
Ecol. 161: 241-280.
Baggerman, B. 1953. Spatfall and transport of Cardium edule L. Arch.
Neerland. Zool. 10: 315-342.
Bayne, B. L. 1964a. Primary and secondary settlement in Mytilus edulis L.
(Mollusca). J. Anim. Ecol. 33: 513-523.
1964b. The responses of the larvae of Mytilus edulis L. to light and gravity.
Oikos 15: 162-174.
1965. Growth and delay of metamorphosis of the larvae of Mytilus edulis
(L.). Ophelia 2: 1-47.
1976. The biology of mussel larvae, p. 81-120. In B. L. Bayne (ed.) Marine
mussels: their ecology and physiology. Cambridge University Press,
Cambridge.
Bennell, S. J. 1981. Some observations on the littoral barnacle populations of
North Wales. Mar. Environ. Res. 5: 227-240.
Bertness, M. D., S. D. Gaines, E. G. Stephens, and P. O. Yund. 1992.
Components of recruitment in populations of the acorn barnacle
Semibalanus balanoides (Linnaeus). J. Exp. Mar. Biol. Ecol. 156: 199-215.
Boicourt, W. C. 1982. Estuarine larval retention mechanisms on two scales, p.
445-457. In V. S. Kennedy (ed.) Estuarine comparisons. Academic Press,
New York.
Bousfield, E. L. 1955. Ecological control of the occurrence of barnacles in the
Miramichi estuary. Bull. Nat. Mus. Can. 137: 69.
Burke, R. D. 1986. Pheromones and the gregarious settlement of marine
invertebrate larvae. Bull. Mar. Sci. 39: 323-331.
Bushek, D. 1988. Settlement as a major determinant of intertidal oyster and

barnacle distributions along a horizontal gradient. J. Exp. Mar. Biol. Ecol.
122: 1-18.
Butman, C. A. 1986. Larval settlement of soft-sediment invertebrates: some
predictions based on an analysis of near bottom velocity profiles, p. 487-
513. In J. C. J. Nihoul (ed.) Marine interfaces ecohydrodynamics.
Elsevier, Amsterdam.
1987. Larval settlement of soft-sediment invertebrates: the spatial scales of
pattern explained by active habitat selection and the emerging role of
hydrodynamical processes. Oceanogr. Mar. Biol. Annu. Rev. 25: 113-165.
1989. Sediment-trap experiments on the importance of hydrodynamical
processes in distributing settling invertebrate larvae in near-bottom
waters. J. Exp. Mar. Biol. Ecol. 134: 37-88.
Butman, C. A., and J. P. Grassle. 1992. Active habitat selection by Capitella
sp. 1 larvae. I. Two-choice experiments in still water and flume flows. J.
Mar. Res. 50: 669-715.
Butman, C. A., J. P. Grassle, and C. M. Webb. 1988. Substrate choices made
by marine larvae settling in still water and in a flume flow. Nature 333:
771-773.
Carleton, J. C, and P. W. Sammarco. 1987. Effects of substratum irregularity
on success of coral settlement: quantification of comparative
geomorphological techniques. Bull. Mar. Sci. 40: 85-98.
Carlton, J. T., and J. B. Geller. 1993. Ecological roulette: the global transport
of non-indigenous marine organisms. Science 261: 78-82.
Carlton, J. T., J. K. Thompson, L. E. Schemel, and F. H. Nichols. 1990.
Remarkable invasion of San Francisco Bay (California, USA) by the
Asian clam. I. Introduction and dispersal. Mar. Ecol. Prog. Ser. 66: 81-94.
Carriker, M. R. 1951. Ecological observations on the distribution of oyster
larvae in New Jersey estuaries. Ecol. Monogr. 21: 19-38.
Chabot, R., and E. Bourget. 1988. Influence of substratum heterogeneity and
settled barnacle density on the settlement of cypris larvae. Mar. Biol. 97:
45-56.
Chia, F-S. 1978. Perspectives: settlement and metamorphosis of marine
invertebrate larvae, p. 283-285. In F-S. Chia and M. E. Rice (eds.)
Settlement and metamorphosis of marine invertebrate larvae. Elsevier,
New York.
Chia, F.-S., J. Buckland-Nicks, and C. M. Young. 1984. Locomotion of
invertebrate larvae: a review. Can. J. Zool. 62: 1205-1222.
Connell, J. H., and R. O. Slatyer. 1977. Mechanisms of succession in natural
communities and their role in community stability and organization. Am.
Nat. I l l : 1119-1144.
Cragg, S. M. 1980. Swimming behaviour of the larvae of Pecten maximus (L.)
(Bivalvia). J. Mar. Biol. Assoc. U. K. 60: 551-564.
Crisp, D. J. 1955. The behaviour of barnacle cyprids in relation to water
movement over a surface. J. Exp. Biol. 32: 569-590.
1974. Factors influencing the settlement of marine invertebrate larvae, p.
177-265. In P. T. Grant and A. M. Mackie (ed.) Chemoreception in
marine organisms. Academic Press, London.
Crisp, D. J., and H. Barnes. 1954. The orientation and distribution of
barnacles at settlement with particular reference to surface contour. J.
Anim. Ecol. 23: 142-162.
Crisp, D. J., and P. S. Meadows. 1963. Adsorbed layers: the stimulus to
settlement in barnacles. Proc. Roy. Soc. Ser. B. 156: 500-520.
DeVantier, L. M. 1992. Rafting of tropical marine organisms on buoyant

coralla. Mar. Ecol. Prog. Ser. 86: 301-302.
DeWolf, P. 1973. Ecological observations on the mechanisms of dispersal of
barnacle larvae during planktonic life and settling. Neth. J. Sea Res. 6: 1-
129.
Duggins, D. O., J. E. Eckman, and A. T. Sewell. 1990. Ecology of understorey
kelp environments. II. Effects of kelp on recruitment of benthic
invertebrates. J. Exp. Mar. Biol. Ecol. 143: 27-45.
Dyer, K. R. (Editor). 1979. Estuarine hydrography and sedimentation: a
handbook. Cambridge University Press, Cambridge.
Eckman, J. E. 1979. Small-scale patterns and processes in a soft-substratum,
intertidal community. J. Mar. Res. 37: 437-457.
1983. Hydrodynamic processes affecting benthic recruitment. Limnol.
Oceanogr. 28: 241-257.
1985. Flow disruption by an animal-tube mimic affects sediment bacterial
colonization. J. Mar. Res. 43: 419-435.
1987. The role of hydrodynamics in recruitment, growth, and survival of
Argopecten irradians (L.) and Anomia simplex (D'Orbigny) within
eelgrass meadows. J. Exp. Mar. Biol. Ecol. 106: 165-191.
1990. A model of passive settlement by planktonic larvae onto bottoms of
differing roughness. Limnol. Oceanogr. 35: 887-901.
Eckman, J. E., and A. R. M. Nowell. 1984. Boundary skin friction and
862.
Eckman, J. E., W. B. Savidge, and T. F. Gross. 1990. Relationships between
duration of cyprid attachment and drag forces associated with detachment
of Balanus amphitrite cyprids. Mar. Biol. 107: 111-118.
Emerson, C. W., and J. Grant. 1991. The control of soft-shell clam (Mya
arenaria) recruitment on intertidal sandflats by bedload sediment
transport. Limnol. Oceanogr. 36: 1288-1300.
Ertman, S. C, and P. A. Jumars. 1988. Effects of bivalve siphon currents on
the settlement of inert particles and larvae. J. Mar. Res. 46: 797-813.
Farrell, T. M., D. Bracher, and J. Roughgarden. 1991. Cross-shelf transport
causes recruitment to intertidal populations in central California. Limnol.
Oceanogr. 36: 279-288.
Fischer, H. B., J. Imberger, E. J. List, R. C. Y. Koh, and N. H. Brooks. 1979.
Mixing in inland and coastal waters. Academic Press, New York.
Gaines, S. D., and M. D. Bertness. 1992. Dispersal of juveniles and variable
recruitment in sessile marine species. Nature 360: 579-580.
Gaines, S., and J. Roughgarden. 1985. Larval settlement rate: a leading
determinant of structure in an ecological community of the marine
intertidal zone. Proc. Nat. Acad. Sci. U.S.A. 82: 3707-3711.
Gallagher, E. D., P. A. Jumars, and D. D. Trueblood. 1983. Facilitation of
soft-bottom benthic succession by tube builders. Ecology 64: 1200-1216.
Grassle, J. P., C. A. Butman, and S. W. Mills. 1992a. Active habitat selection
by Capitella sp. 1 larvae. II. Multiple-choice experiments in still water and
flumeflows.J. Mar. Res. 50: 717-743.
Grassle, J. P., P. V. R. Snelgrove, and C. A. Butman. 1992b. Larval habitat
choice in still water and flume flows by the opportunistic bivalve Mulina
lateralis. Neth. J. Sea Res. 30: 33-44.
Gray, J. S. 1974. Animal-sediment relationships. Oceanogr. Mar. Biol. Annu.
Rev. 12: 223-261.
Hannan, C. A. 1984. Planktonic larvae may act like passive particles in

turbulent near-bottom flows. Limnol. Oceanogr. 29: 1108-1116.
Harvey, M., E. Bourget, and R. G. Ingram. 1995. Experimental evidence of
passive accumulation of marine bivalve larvae on filamentous epibenthic
structures. Limnol. Oceanogr. 40: 94-104.
Harvey, M., E. Bourget, and G. Miron. 1993. Settlement of Iceland scallop
spat (Chlamys islandica) in response to hydroids and filamentous red
algae: field observations and laboratory experiments. Mar. Ecol. Prog.
Ser. 99: 283-292.
Havenhand, J. N., and I. Svane. 1991. Roles of hydrodynamics and larval
behaviour in determining spatial aggregation in the tunicate Ciona
intestinalis. Mar. Ecol. Prog. Ser. 68: 271-276.
Hawkins, S. J., and R. G. Hartnoll. 1982. Settlement patterns of Semibalanus
balanoides (L.) in the Isle of Man (1977-1981). J. Exp. Mar. Biol. Ecol.
62: 271-283.
Hebert, P. D. N., B. W. Mancaster, and G. L. Mackie. 1989. Ecological and
genetic studies on Dreissena polymorpha (Pallas) a new mollusc in the
Great Lakes. Can. J. Fish. Aquat. Sci. 46: 1587-1591.
Hidu, H., and H. H. Haskins. 1978. Swimming speeds of oyster larvae of
Crassostrea virginica in different salinities and temperatures. Estuaries 1:
252-255.
Highsmith, R. C, and R. Emlet. 1986. Delayed metamorphosis: effect on
growth and survival of juvenile sand dollars (Echinoidea: Clypeasteroida).
Bull. Mar. Sci. 39: 347-361.
Holmquist, J. G. 1994. Benthic macroalgae as a dispersal mechanism for fauna:
influence of a marine tumbleweed. J. Exp. Mar. Biol. Ecol. 180: 235-251.
Jonsson, P. R., C. Andre, and M. Lindegarth. 1991. Swimming behaviour of
marine bivalve larvae in a flume boundary-layer flow: evidence for near-
bottom confinement. Mar. Ecol. Prog. Ser. 79: 67-76.
Keough, M. J. 1984. Kin-recognition and the spatial distribution of larvae of
the bryozoan Bugula neritina (L.). Evolution 38: 142-147.
Keough, M. J., and B. J. Downes. 1982. Recruitment of marine invertebrates:
the role of active larval choices and early mortality. Oecologia 54: 348-
352.
Korringa, P. 1952. Recent advances in oyster biology. Quart. Rev. Biol. 27:
266-308, 339-369.
Lacalli. T. 1981. Annual spawning cycles and planktonic larvae of benthic
invertebrates from Passamaquoddy Bay, New Brunswick. Can. J. Zool.
59: 433-440.
LeFevre, J., and E. Bourget. 1991. Neustonic niche for cirripede larvae as a
possible adaptation to long-range dispersal. Mar. Ecol. Prog. Ser. 74: 185-
194.
1992. Hydrodynamics and behaviour: transport processes in marine
invertebrate larvae. Trends Ecol. Evol. 7: 288-289.
Lemire, M., and E. Bourget. 1996. Substratum heterogeneity and complexity
influence micro-habitat selection of Balanus sp. and Tubularia croeca
larvae. Mar. Ecol. Prog. Ser. 135: 77-87.
LeTourneux, F., and E. Bourget. 1988. Importance of physical and biological
settlement cues used at different spatial scales by the larvae of
Semibalanus balanoides. Mar. Biol. 97: 57-66.
Levin, L. A. 1983. Drift tube studies of bay-ocean water exchange and
implications for larval dispersal. Estuaries 6: 364-371.
Luckenbach, M. 1984. Settlement and early post-settlement survival in the

recruitment of Mulinia lateralis (Bivalvia). Mar. Ecol. Prog. Ser. 17: 245-
250.
Maas Geesteranus, R. A. 1942. On the formation of banks by Mytilus edulis L.
Arch, neerl. Zool. 6: 283-325.
Maida, M., J. C. Coll, and P. W. Sammarco. 1994. Shedding new light on
scleractinian coral recruitment. J. Exp. Mar. Biol. Ecol. 180: 189-202.
Maki, J. S., D. Rittschof, J. D. Costlow, and R. Mitchell. 1988. Inhibition of
attachment of larval barnacles, Balanus amphitrite, by bacterial surface
films. Mar. Biol. 97: 199-206.
Maki, J. S., D. Rittschof, M.-O. Samuelsson, U. Szewzyk, A. B. Yule, S.
Kjeuebert, J. D. Costlow, and R. Mitchell. 1990. Effects of marine
bacteria and their exopolymers on the attachment of barnacle cypris
larvae. Bull. Mar. Sci. 46: 499-511.
Mann, R. 1988. Distribution of bivalve larvae at a frontal system in the James
River, Virginia. Mar. Ecol. Prog. Ser. 50: 29-44.
May, B., and J. E. Marsden. 1992. Genetic identification and implications of
another invasive species of dressenid mussel in the Great Lakes. Can. J.
Fish. Aquat. Sci. 49: 1501-1506.
McEdwards, L. (Editor). 1995. Ecology of marine invertebrate larvae. CRC
Press, Boca Raton, Florida.
McGrath, D., P. A. King, and E. M. Gosling. 1988. Evidence for the direct
settlement of Mytilus edulis larvae on adult mussel beds. Mar. Ecol. Prog.
Ser. 47: 103-106.
Meadows, P. S., and J. I. Campbell. 1972. Habitat selection by aquatic
invertebrates. Adv. Mar. Biol. 10: 271-382.
Menge, B. 1975. Brood or broadcast? The adaptive significance of different
reproductive strategies in the intertidal seastars Lepasterias hexactis and
Piaster ochraceus. Mar. Biol. 31: 87-100.
Miron, G., B. Boudreau, and E. Bourget. 1995. Use of larval supply in benthic
ecology: testing correlations between larval supply and larval settlement.
Miron, G., E. Bourget, and P. Archambault. 1996. Scale of observation and
distribution of adult conspecifics: their influence in assessing passive and
active settlement mechanisms in the barnacle Balanus crenatus (Brugiere).
J. Exp. Mar. Biol. Ecol. 201: 137-158.
Mullineaux, L. S., and C. A. Butman. 1990. Recruitment of encrusting benthic
invertebrates in boundary layer flows: a deep-water experiment on Cross
Seamount. Limnol. Oceanogr. 35: 409-423.
1991. Initial contact, exploration and attachment of barnacle {Balanus
amphitrite) cyprids settling in flow. Mar. Biol. 110: 93-103.
Mullineaux, L. S., and E. D. Garland. 1993. Larval recruitment in response to
manipulated field flows. Mar. Biol. 116: 667-683.
Neal, A. L., and A. B. Yule. 1994. The tenacity of Elminius modestus and
Balanus perforatus cyprids to bacterial films grown under different shear
regimes. J. Mar. Biol. Assoc. U.K. 74: 251-257.
Nicholls, F. H., and J. K. Thompson. 1985. Time scales of change in the San
Francisco Bay benthos. Hydrobiologia 129: 121-138.
Okubo, A. 1971. Oceanic diffusion diagrams. Deep Sea Res. 18: 789-
802.
1994. The role of diffusion and related physical processes in dispersal and
recruitment of marine populations, p. 5-32. In P. W. Sammarco and M. L.
Heron (ed.) Coastal and estuarine studies. Vol. 45. The bio-physics of
marine larval dispersal. American Geophysical Union, Washington, D.C.
Olafsson, E. B., C. H. Petersen, and W. G. Ambrose. 1994. Does recruitment
limitation structure populations and communities of macro-invertebrates
in marine soft sediments: the relative significance of pre- and post-
settlement processes. Oceanogr. Mar. Biol. Annu. Rev. 32: 65-109.
Pawlik, J. R. 1986. Chemical induction of larval settlement and metamorphosis
in the reef building tube worm Phragmatopoma californica (Polychaeta:
Sabellaridae). Mar. Biol. 91: 59-68.
1992. Chemical ecology of the settlement of benthic marine invertebrates.
Oceanogr. Mar. Biol. Annu. Rev. 30: 273-335.
Pawlik, J. R., and C. A. Butman. 1993. Settlement of a marine tube worm as a
function of current velocity: Interacting effects of hydrodynamics and
behaviour. Limnol. Oceanogr. 38: 1730-1740.
Pawlik, J. R., C. A. Butman, and V. R. Starczak. 1991. Hydrodynamic
facilitation of gregarious settlement of a reef-building tube worm. Science
251: 421-424.
Peterson, C. H. 1982. Clam predation by whelks {Busycon spp.): experimental
tests of the importance of prey sizes, prey density, and seagrass cover.
Mar. Biol. 66: 159-170.
1986. Enhancement of Mercenaria mercenaria densities in seagrass beds: is
pattern fixed during settlement season or altered by subsequent
differential survival? Limnol. Oceanogr. 31: 200-205.
Pineda, J. 1991. Predictable upwelling and the shoreward transport of
planktonic larvae by internal tidal bores. Science 253: 548-551.
Pratt, D. M. 1953. Abundance and growth of Venus mercenaria and
Callocardia morrhiana in relation to the character of bottom sediments. J.
Mar. Res. 12: 60-74.
Pyefinch, K. A. 1948. Methods of identification of the larvae of Balanus
balanoides (L.), B. crenatus Brug. and Verruca stoemia O. F. Muller. J.
Mar. Biol. Assoc. U.K. 27: 451-463.
Ricciardi, A., R. Serrouya, and F. G. Whoriskey. 1995. Aerial exposure
tolerance of zebra and quagga mussels (Bilvalvia: Dreissenidae):
implications for overland dispersal. Can. J. Fish. Aquat. Sci. 52: 470-
477.
Rittschof, D., E. Sansford Branscomb, and J. D. Costlow. 1984. Settlement and
behavior in relation to flow and surface in larval barnacles, Balanus
amphitrite Darwin. J. Exp. Mar. Biol. Ecol. 82: 131-146.
Roegner, C, C. Andre, M. Lindegarth, J. E. Eckman, and J. Grant. 1995.
Transport of recently settled soft-shell clams {Mya arenaria L.) in
laboratory flume flow. J. Exp. Mar. Biol. Ecol. 187: 13-26.
Roughgarden, J., S. D. Gaines, and S. W. Pacala. 1987. Supply side ecology:
the role of physical transport processes, p. 491-518. In J. H. R. Gee and
P. S. Giller (ed.) Organization of communities past and present. Blackwell
Scientific Press, Oxford.
Sammarco, P. W. 1991. Geographically specific recruitment and post-
settlement mortality as influences on coral communities: the cross
continental shelf transplant experiment. Limnol. Oceanogr. 36: 496-514.
Sammarco, P. W., and M. L. Heron (Editors). 1994. The bio-physics of marine
larval dispersal. American Geophysical Union, Washington, D.C.
Scheltema, R. S. 1974. Biological interactions determining larval settlement of
marine invertebrates. Thai. Jugosl. 10: 263-296.
1986. On dispersal and planktonic larvae of benthic invertebrates: an

eclectic overview and summary of problems. Bull. Mar. Sci. 39: 290-322.
Seliger, H. H., J. A. Boggs, R. B. Rivkin, W. H. Biggley, and K. R. H. Aspden.
1982. The transport of oyster larvae in an estuary. Mar. Biol. 71: 57-72.
Shanks, A. L. 1983. Surface slicks associated with tidally forced internal waves
may transport pelagic larvae of benthic invertebrates and fishes
shoreward. Mar. Ecol. Prog. Ser. 13: 311-315.
1985. The behavioural basis of internal wave-induced shoreward transport
of the megalopae of Pachygrapsus crassipes. Mar. Ecol. Prog. Ser. 24:
289-295.
1986. Tidal periodicity in the daily settlement of intertidal barnacle larvae
and an hypothesized mechanism for the cross shelf transport of cyprids.
Biol. Bull. 170: 429-440.
1988. Further support for the hypothesis that internal waves can cause
shoreward transport of larval invertebrates and fish. Fish. Bull. 86: 703-
714.
1995. Orientated swimming by megalopae of several eastern North Pacific
crab species and its potential role in their onshore migration. J. Exp. Mar.
Biol. Ecol. 186: 1-16.
Shanks, A. L., and E. W. Edmondson. 1990. The vertical flux of metazoans
(holoplankton, meiofauna, and larval invertebrates) due to their
association with marine snow. Limnol. Oceanogr. 35: 455-463.
Shanks, A. L., and W. G. Wright. 1987. Internal-wave mediated shoreward
transport of cyprids, megalopae, gammarids and correlated longshore
differences in the settling rate of intertidal barnacles. J. Exp. Mar. Biol.
Ecol. 114: 1-13.
Sigurdsson, J. B., C. W. Titman, and P. A. Davies. 1976. The dispersal of
young post-larval bivalve molluscs by byssus threads. Nature 262: 386-
387.
Snelgrove, P. V. R. 1994. Hydrodynamic enhancement of invertebrate larval
settlement in microdeposition environments: colonization tray
experiments in a muddy habitat. J. Exp. Mar. Biol. Ecol. 176: 149-166.
Snelgrove, P. V. R., C. A. Butman, and J. P. Grassle. 1993. Hydrodynamic
enhancement of larval settlement in the bivalve Mulina lateralis (Say) and
the polychaete Capitella sp. 1 in microdepositional environments. J. Exp.
Mar. Biol. Ecol. 168: 71-109.
Stancyk, S. E., and R. J. Feller. 1986. Transport of non-decapod invertebrate
larvae in estuaries: an overview. Bull. Mar. Sci. 39: 257-268.
Starr, M., J. H. Himmelman, and J.-C. Therriault. 1990. Marine invertebrate
spawning induced by phytoplankton. Science 247: 1071-1074.
Stoner, A. W., and R. J. Livingston. 1989. Distributional ecology and food
habits of the banded blenny Parclinus fasciatus (Clinidae): a resident in a
mobile habitat. Mar. Ecol. 56: 239-246.
Strathmann, R. R. 1985. Feeding and non-feeding larval development and life
history evolution in marine invertebrates. Annu. Rev. Evol. Syst. 16: 339-
361.
Theodor, J. 1967. Contributions a l'etude des gorones (VII): ecologie et
comportament de la planula. Vie Milieu Ser. A 18: 291-301.
Thiebaut, E., J. C. Dauvin, and Y. Lagadeuc. 1992. Transport of Owenia
fusiformis larvae (Annelida: Polychaeta) in the Bay of Seine. I. Vertical
distribution in relation to water column stratification and ontogenic
vertical migration. Mar. Ecol. Prog. Ser. 80: 29-39.
Thorson, G. 1950. Reproduction and larval ecology of marine bottom

invertebrates. Biol. Rev. Cambridge Philos. Soc. 25: 1-45.
Tremblay, M. J., and M. Sinclair. 1994. Drift of sea scallop larvae Placopecten
magellanicus on Georges Bank: a model study of the roles of mean
advection, larval behaviour and larval origin. Deep Sea Res. Part II 41: 7-
50.
Turner, E. J., R. K. Zimmer-Faust, M. A. Palmer, M. Luckenbach, and N. D.
Peutcheff. 1994. Settlement of oyster {Crassostrea virginica) larvae: effects
of water flow and a water soluble chemical cue. Limnol. Oceanogr. 39:
1579-1593.
Tyler, P. A., and F. T. Banner. 1977. The effect of coastal hydrodynamics on
the echinoderm distribution in the sublittoral of Oxwich Bay, Bristol
Channel. Est. Coastal Mar. Sci. 5: 293-308.
Underwood, A. J., and P. C. Fairweather. 1989. Supply side ecology and
benthic marine assemblages. Trends Ecol. Evol. 4: 16-20.
Walters, L. J. 1992. Field settlement locations on sub tidal marine hard
substrata: Is active larval exploration involved? Limnol. Oceanogr. 37:
1101-1107.
Wethey, D. S. 1984. Spatial patterns in barnacle settlement: day to day
changes during the settlement season. J. Mar. Biol. Assoc. U.K. 64: 687-
698.
1986. Ranking of settlement cues by barnacle larvae: influence of surface
contour. Bull. Mar. Sci. 39: 393-400.
Wildish, D. J. 1977. Factors controlling marine and estuarine sublittoral
macrofauna. Helgolander. Wiss. Meeresunters. 30: 445-454.
Wilson, F. S. 1990. Temporal and spatial patterns of settlement: a field study
of molluscs in Boque Sound, North Carolina. J. Exp. Mar. Biol. Ecol. 139:
201-220.
Young, C. M. 1990. Larval ecology of marine invertebrates: a sesquicentennial
history. Ophelia 32: 1-48.
Zimmer-Faust, R. K., and M. N. Tamburri. 1994. Chemical identity and
ecological implications of a waterborne, larval settlement cue. Limnol.
Oceanogr. 35: 1075-1087.
4
Flow and the physiology of filtration
In this chapter, our concern is primarily with the first filtration stage of
feeding as defined later. In the next chapter, we deal with the second
stage of suspension feeding, which involves the mechanisms of particle
capture on the appropriate surface. Because these two stages of feeding
are the only ones directly affected byflow,they are the only ones consid-
ered in physiological detail in this book.
Background
In suspension-feeding animals, feeding and growth involves a series of
coordinated steps. Each of these steps is dependent on the preceding one
(Table 4.1). We will refer to suspension feeding as inclusive of the first
four steps shown and reserve the term filtration for the first two only.
Filtration or clearance rates (see Chap. 2, the section, Initial feeding
responses) need not, therefore, equal feeding rates if particle sorting
leads to rejection as pseudofaeces. Growth encompasses the remaining
catabolic and anabolic steps. Growth rates need not be directly related to
the feeding rate because of differential digestion of sestonic particles.
For example, in the giant scallop, some microalgae pass through the gut
undigested (Shumway et al. 1985b). Residence times of food in the gut
may be increased or decreased (Bricelj, Bass, and Lopez 1984) and thus
change the contact time between ingested food, and digestive and ab-
sorptive surfaces. Thus, filtration rate is not necessarily equal to feeding
rate and feeding rate need not always be directly proportional to the
growth rate.
Two major categories of suspension feeding benthic animals can be
recognized by the extent to which the filtration process is dependent on
external flow (Vogel 1981; LaBarbera 1984):
128
Flow and Physiology of Filtration 129
Table 4.1. Stages in the physiology of feeding, assimilation, growth, and

elimination in suspension feeding benthic animals.
Stage
Stage number Description
Feeding 1 - Transport of seawater past the filtration surface
2 - Capture of seston at the filtration surface
3 - Transport of particles to the mouth involving
sorting and rejection, e.g. as pseudofaeces in
bivalves
4 - Ingestion
Assimilation 5 - Transport through the gut
6 - Uptake across specific gut surfaces
Growth 7 - Anabolic metabolism
8 - Catabolic metabolism
Elimination 9 - Waste solids voided as true faeces
10 - Metabolic wastes voided as urine
Passive suspension feeders - the individual is solely dependent on the external

ambient flow to bring seston close enough to the filtration surface for its
capture
Active suspension feeders - the individual must supply its own energy in the
form of ciliary or muscular pump power to transport seawater with its load of
seston across the filtration surface where particle capture is effected
In addition, there are facultative active suspension feeders, where indi-
viduals are able to switch from passive to active feeding and back again,
depending on the ambient flow conditions. These individuals behave as
passive suspension feeders at high velocities and as active suspension
feeders at low velocities. In another category, the combined passive-
active suspension feeders, the individual has two parallel mechanisms,
one passive and one active, for creating a flow through the body for
feeding purposes. The latter group differs from the facultative active
suspension feeder in that there is no switching at critical velocities and
both active and passive filtration may occur simultaneously. The passive
mechanism involves ambient velocity induced flow, and the active
mechanism involves a ciliary or flagellar pump. This is a new category,
not explicitly considered by Vogel and LaBarbera, and tentatively in-
cludes some ascidians, sponges, and, possibly, brachiopods. Finally, there
are the deposit-suspension feeders, individuals that ingest already depos-
ited sedimentary particles at low velocities, but switch to suspension
feeding at a higher threshold velocity. Examples include spionid

polychaetes and some infaunal bivalve molluscs.
We view these generalizations as helpful in analyzing the wide diver-
sity of methods of suspension feeding to be found among benthic ani-
mals. As further detailed knowledge of suspension-feeding mechanisms
accumulates from a wider range of benthic animals, we can expect fur-
ther modification of the categories presented. We feel that creating too
many categories is counter-productive because, eventually, the process
will be competing with taxonomy. It is not the artificial categories that
developed new suspension-feeding mechanisms, but populations from
individual species (Chap. 9).
Expected filtration responses of suspension feeders to increases in
unidirectional flows are shown in Fig. 4.1. In this representation, it is
assumed that the suspension feeder is optimally positioned with respect
to flow direction. The filtration responses appear to be some form of
unimodal function of velocity. In earlier studies this response was de-
scribed as a "reverse ramp" for part of b and c (Wildish and Saulnier
1993). Since this engineering term is unfamiliar to biologists, we have
replaced it with the more general term continuous unimodal function. Its
use does not imply a precise mathematical relationship between filtration
and velocity. For filtration responses and velocity, parabolic, orthogonal,
sharp peaked, or ramp functions may be obtained, depending on the
species and environmental conditions considered. Given constant seston
concentrations, increments in velocity enhance filtration (Fig. 4.1a) at
very low velocities up to a maximum (Fig. 4.1b) beyond which velocity
increase has no further effect on filtration because other processes (e.g.
stage 3 or 4; see Table 4.1) limit it. An example of suspension feeding in
Protozoa is given by Fenchel (1980); in it particle transport after capture
and the rate of ingestion may be the rate-limiting factors. Bivalves have
a pseudofaecal rejection mechanism that avoids overloading ingestion
(stage 4) and assimilation (stages 5 and 6). At higher velocities (Fig.
4.1c), flow conditions begin to have a negative effect on the filtration
process and finally filtration (Fig. 4.1d) is completely suppressed, e.g. by
mantle edge closure or withdrawal of siphons in infaunal bivalve mol-
luscs. The precise mechanisms by which the filtration responses are
governed depend on the species of suspension feeder, as will be dis-
cussed in some examples in later sections of this chapter. Thus, in active
suspension feeders such as siphonate bivalves, in which the inhalant is
normal to theflow,flow-inducedinhibition may be absent, while in non-
siphonate bivalves, ambient flow inhibits seston capture (Fig. 4.1). In
PASSIVE
a ! b i c id
ACTIVE-INHALANT ACTIVE-EXHALANT
NORMAL TO FLOW DOWNSTREAM IN FLOW
U
Figure 4.1 Effects of velocity, U, on filtration rate, R: a, increasing velocity
enhances filtration; b, optimum filtration rates; c, velocity inhibition of filtration;
d, filtration suppressed by velocity.
general, in stage a, filtration is limited by seston supply as a result of

seston dilution and limited transport; in b, filtration is not limited by
seston dilution and transport; and in c, filtration is inhibited by excessive
ambient flow forces that interfere with seston capture processes.
Seston concentration and quality may also have an influence on the
filtration rate, which may be either independent of or interactively linked
to velocity; this is the case, for example, for bivalve molluscs (Wildish,
Kristmanson, and Saulnier 1992). These responses permit active suspen-
Table 4.2. Representative list of macrobenthic passive

suspension feeders.
Common name Scientific name Major taxa Reference
Sea pen Ptilosarcus guerneyi Cnidaria: Pennatulacea Best (1988)

Hydroid Plumilaria setacea Cnidaria: Hydrozoa Warner (1977)
Sea whip Leptogorgia virgulata Cnidaria: Anthozoa Leversee (1976)
Black coral Cirrhipathes lutkeni Cnidaria: Antipatharia Warner (1977)
Feather star Oligometra serripinnia Echinodermata: Leonard et al.
Crinoidea (1988)
Brittle star Ophiothrix fragilis Echinodermata: Warner and
Ophiuroidea Woodley (1975)
Sea cucumber Cucumaria curata Echinodermata: J0rgensen (1966)
Holothuroidea
sion feeders rapidly to increase both the time spent filtering and the
volumetric throughput for filtration to take advantage of temporally
discrete pulses of good quality sestonic food.
For active suspension feeders, the filtration response also depends
critically on whether the exhalant siphon flow follows or is opposed to
the principal direction of ambientflow.Where the exhalant flow opposes
the ambientflow,the adverse pressure field causes a reduction in filtered
flow, as extraciliary pump work cannot be done to overcome the applied
external flow force.
Passive suspension feeders

Passive suspension feeders are found in many of the phyla which have
benthic representatives. Table 4.2 gives an idea of the diversity of the
taxonomic forms represented. A detailed hydrodynamic study of filtra-
tion responses to velocity and the mechanism of seston capture is avail-
able for only a few passive suspension feeders.
Cnidarians
The sea pen Ptilosarcus gurneyi is a common cnidarian of soft, sandy-
mud sediments in Puget Sound and the San Juan Archipelago of the U.S.
West Coast (Best 1988). Its body consists of a central supporting pedun-
cle inserted in the sediment, a joint, and a rachis. The sea pen is oriented
so that its flat surface faces the major prevailingflow.A parallel series of
semi-circular leaves carries the filtering elements. These filtering ele-
ments are polyps with radiating tentacles bearing pinnules which are
located on the downstream part of each leaf. The rachis is flexible and
bends downstream as ambient velocities increase.
Volume flow rates through P. gurneyi have been determined by esti-
mating the rate of dye movement to determine velocity and a measure of
the rachis area from the dye-stained part of the sea pen (Best 1988).
Volumetric flow rates were related to the square of the height since the
filter area increased linearly with size. The maximum volumetric flow
through the rachis occurred at an ambient velocity which was a function
of sea pen size: thus 6.5-8.5 cm -s"1 for small, 12-14cm-s"1 for medium,
and 14-18 cm-s 1 for large sea pens.
Velocity measurements between the leaves and near the polyps made
with a thermistor flow probe showed that near the polyps velocities were
much reduced at the highest free-stream flows tested (25cm-s1) as a
result of drag near the filtration surfaces.
Filtration efficiency (RE; see Chap. 2) was examined in flume experi-
ments by taking seawater samples upstream and downstream of the
sea pen (with a suction sampler so that sampling was possibly not
isokinetic). The sample was examined in a particle counter and the
efficiency of capture of each particle size determined. Smaller sea
pens were significantly less efficient than medium or large individuals.
Free stream velocities ranging from 1.5 to 6.0 cm s"1 significantly reduced
RE.
Filtration and seston uptake rates, determined from independent esti-
mates of volume flow rate and filtering efficiency as outlined, were equal
in P. gurneyi and unimodal functions of velocity (Fig. 4.2). For the sea
pen, increasing velocity in low flow conditions enhances the filtration
seston uptake rate because of the increase in volumetricflowthrough the
filter. The plateau reached around 7 cm s"1 (Fig. 4.2) occurs as ambient
velocity just begins to deform the sea pen. This process eventually causes
both volume flows and filtering efficiency to decline in the c stage be-
cause pinnules and tentacles are swept back as the flow increases and,
consequently, less filtering surface is presented to theflow.For small sea
pens, the c stage begins at 13-16 cm-s"1 and, in medium and larger ones,
at 18-21 cm s"1. Filtration/seston uptake rate inhibition in P. gurneyi can
be seen to begin in Fig. 4.2 but, because the highest velocities tested were
only 25cm-s~1, the full unimodal function responses were not obtained.
/ ^ ^ " ^ a Medium
105
Li x
C
< i
i
i
LU i Small
IO4: i
O
DC
<
CL-
10
0 2 4 6 8 1012 141618 2022
VELOCITY, cm -S" 1
Figure 4.2 Seston uptake rate in Ptilosarcus gurneyi as a function of ambient
velocity: open circles, retention efficiency at 30%; closed circles, at 20%. The
unimodal response limits are suggested (modified from Best 1988).
No test was made of the effect of seston concentration on capture rates

during this study (Best 1988).
The gorgonian corals exhibit whole colony flexibility where water
movement forces reach moderate to high levels. Two species,
Pseudopterogorgia acerosa and P. americana, found off Jamaica, were
studied by Sponaugle and LaBarbera (1991). Coral feeding was studied
in a recirculating flume in unidirectional flows measured within 2 mm
of the polyp using a thermistor flow probe with a sensing head of
0.5 mm. Artemia nauplli were used as food at concentrations of 2.5-4
nauplii L"1. Captured rhodamine-stained nauplii were easily observed in
the individual polyp gastrovascular cavity. Results showed that the cap-
ture rate of Artemia was a unimodal function of velocity with a peak at
intermediate velocities (-IScm-s"1).
Model studies of the individual polyp of the tropical gorgonian
Pseudopterogorgia acerosa were made by Sponaugle (1991). She noticed
that, in addition to colony flexion, which keeps flow velocities near the
polyps low, individual polyps are deformed by increased velocities. Two

effects were noticed: the first that polyp flexion helped maintain reduced
and constant velocities near the polyp with the effect most marked at
intermediate flows; the second that flow patterns downstream of the
polyp at intermediate ambient flows suggested an upward circulation,
similar to that in tube-living suspension feeders discussed in Chapter 6
(see the section, Tube-living epifauna) which enhanced particle capture.
Both mechanisms contribute to the unimodal feeding response of this
gorgonian to flow described by Sponaugle and LaBarbera (1991).
Studies of particle capture by individual polyps of the octocoral
Alcyonium siderium collected from either the Massachusetts or the Cali-
fornia coast were made in a cooled, recirculating flume at threeflows:2.7,
12.2, or 19.8 cm-s"1 (Patterson 1991). These flows were measured with a
thermistor flowmeter probe positioned 1 cm above the coral polyp. It was
shown at the lowest flow tested that Artemia cysts were captured on the
upstream tentacles of the Alcyonium polyp, but that at flows of
12.2 cm-s"1 or greater, capture was transferred to the downstream tenta-
cles. At the highest flow tested, the number of cysts captured per polyp
per unit time was significantly less than at the lower flows. These obser-
vations are consistent with the unimodal feeding response of other sus-
pension feeders to flow, although there are insufficient data for A.
siderium to determine the precise velocity limits. At the 19.8-cm-s"1 flow
the polyps are bent downstream and eddies form over the tentacular
surfaces. At these higher flows, the cysts are captured all over the tenta-
cles in a radially symmetrical pattern and move towards the tentacle tip,
although capture efficiency decreases with increasing Reynolds number.
Filtration efficiency of three species of gorgonian corals were shown
by Dai and Lin (1993) in recirculating flume experiments to be a
unimodal response to velocity with a maximum efficiency near 8cm-s-1
(Fig. 4.3). A wide range of velocities was tested during these experiments
with Subergorgia suberosa feeding limited to the range 7-9 cm-s"1,
Melthaea ochracea to 4-40 cm-s"1, and Acanthogorgia vegae to 2-
22cm-s~1. Dai and Lin (1993) found that the typical feeding range for a
species was related to the polyp size and its deformability in flow.
Echinoderms
The stalkless crinoid Oligometra serripinnia from the Great Barrier Reef
was used in experimental flume tests by Leonard, Strickler, and Holland
(1988). This crinoid lives at 10-15 cm above the sediment-water interface
80 -
A
12 16 20 24 28 36 40
U, cm-s"1
B
8 12 16 20 24 28 32 36 40
(J, cm -s"1
Figure 4.3 Normalized feeding effectiveness with respect to ambient velocity:
, Acanthogorgia vegae; , Melithaea ochracea; and A, Subergorgia suberosa; A,
polyp feeding rates; B, feeding rate per unit surface area (Dai and Lin 1993).
as an ectocommensal on sea fans, where it occupies the downstream side

of its host. The adult animals used in the experiments described here
were star-shaped with 10 radiating arms, each ~3 cm in length. Each arm
had 55-80 pinnules with 40-60 tube feet per pinnule for feeding. There
was a relatively wide gap between arms but smaller distance (up to
0.69 mm) between pinnules. At higher flows, the distance between
pinnules was a function of velocity due to flow deformation.
Velocity was determined by Leonard et al. (1988) to affect filtration in
the following ways:
By changing the rate of encounters with seston particles
By affecting the size and quality of available seston
By changing thefine-scaleflowregime near the filter
By changing the shape of the filter
By changing the efficiency of capture by tube feet and pinnules
Capture efficiency rates were determined independently in this study
by tracking individual particles by dark field photography so that seston
approach, capture, and escape rates as a function of velocity (Fig. 4.4)
could be determined. The results suggest that for the adult size crinoids
used (n = 4), the following unimodal responses occur:
a. 0.9^.8 cm -s'1: rate of encounter with seston low but capture efficiency high
b. 4.8-6.4 cm-s"1: optimumflowrate with rate of encounter with seston higher
and capture efficiencies still high
c. >6.4cm-s~1: highest rate of encounter with seston but high flows causing
deformation of thefilterand the capture efficiency lower
The maximum velocity tested by Leonard et al. (1988) was 13.3 cm s"1, at
which the absolute approach rate of seston particles for a whole crinoid
had reached an asymptotic level.
Studies by Leonard (1989) on another stalkless crinoid, Antedon
mediterranean found in the Mediterranean Sea off Monaco, included
measurement of both velocity and seston concentration to determine the
effect on feeding responses. Feeding activity was recorded by the dark
field illumination technique (see Chap. 2), focused on the pinnules bear-
ing approximately 70 tube feet. The feeding rate was assumed to be
directly proportional to the number of single or multiple tube foot flicks
made when a sestonic particle was captured by the filter. The crinoids in
a small capacity Vogel-LaBarbera recirculating flume were subjected to
a cycle of increasing step changes of velocity (0.4-3.7 cm s"1) followed by
decreasing step changes of velocity (3.7-0.4cm-s"1) for a wide range of
seston concentrations (25-10,000 cells of Hymenomonas elongata/ml).
This microalga is a spherical coccolithophore of -lljim diameter. The
50 - 50- B
LU 40 - 40-
O ai b i ,c
f
r/v
30 - 30
RCEI
20 - a |
; i
20
b i c
LU
Q_
10 - 10"
^ ^
2 4 6 8 10 12 14 2 4 6 8 10 12 14
50 - 50
aj b |

i
LU40 40-
| 30- 30-
O 20 20-
E
"10
2 4 6 8 10 12 14
VELOCITY C c m - s " 1 )
10-
(/. \
2 4 6 8 10 12 14
1
VELOCITY Ccm -s" 3
Figure 4.4 Oligometra serripinnia capturing pollen grains at different ambient

velocities in four individuals (A-D): A, passes; o, captures; , escapes as a
percentage of those possible. The unimodal function limits are suggested (modi-
fied from Leonard et al. 1988).
optimum concentration for filtration was 2000 cells-ml \ beyond which

further increases caused a progressive decrease in feeding activity (flick
responses). For each concentration tested, there was a maximum flick
activity at 2.3 cm-s"1 (a). Beyond 2.3 cm-s"1, feeding rate declined (c),
suggesting a limited b region. Statistical analysis of the data by analysis of
variance suggested that flick activity depends on velocity and particle
concentration, but not on the rate at which particles approach the filter.
Thus, one cannot assume that filtration rate is the product of velocity and
seston concentration.
Active suspension feeders

Bivalve molluscs are perhaps best described as active suspension feeders,
although some are deposit feeders and others are facultative suspension-
Table 4.3. Representative macrobenthic active suspension feeders.
Common name Scientific name Major taxa Reference
Scallop Placopecten magellanicus Bivalvia Wildish et al. (1987)

Mussel Mytilus edulis Bivalvia Walne (1972)
Winkle Serpulorbis natalensis Gastropoda Hughes (1978)
Bryozoan Flustrellidra hispida Bryozoa Best and Thorpe (1983)
Sea squirt Ciona intestinalis Ascidiacea Flood and Fialo-Medioni
(1981)
Worm Chaetopterus Polychaeta Flood and Fialo-Medioni
variopedatus (1982)
Sand dollar Dendrastus excentricus Echinodermata Timko (1976)
deposit feeders. Examples of taxa which are thought to be active suspen-

sion feeders are listed in Table 4.3. Besides bivalves, they include
Bryozoa and sea squirts, plus a few species of echinoderms, polychaetes,
and gastropods.
In most cases, the pump which moves seawater in the trophic fluid
transport system is a ciliary one. Only in a few polychaete species, such
as Chaetopterus variopedatus, is a muscular piston pump employed.
Bivalves
Although many early bivalve biologists (e.g. Kerswill 1949) recognized
the importance of flow in bivalve production, the precise mechanisms by
which this was achieved were not addressed until relatively recently.
Kirby-Smith (1972) was the first to study the effect of flow on bay
scallops (Argopecten irradians) in a growth tube apparatus (Chap. 2).
Flow was measured as the bulk velocity flow rate. Kirby-Smith (1972)
was also the first to demonstrate that higher velocities could inhibit bay
scallop growth. Wildish et al. (1987) have reinterpreted his growth re-
sults as a unimodal function of velocity, in general agreement with later
studies made in a flume with this bay scallop (Cahalan, Siddall, and
Luckenbach 1989; Eckman, Peterson, and Cahalan 1989). Growth re-
sults with the giant scallop {Placopecten magellanicus) in flume experi-
ments also show that the growth of this species has a unimodal function
response to velocity. Flume studies (Wildish and Saulnier 1993) where a
wide range of velocities were tested confirm that filtration by the giant
scallop is a unimodal function of velocity, suggesting that it is the rela-

tionship underlying the growth responses.
Considering the possible mechanisms involved in the giant scallop
filtration response to velocity and using the labelling of Fig. 4.1, the
following observations can be made:
a. U = 0 to ~3 cm s"1. At zero velocity, the filtration rate is time dependent and,

because of mixing limitations, usually affected by the geometry of the experi-
mental apparatus. The exhalant jet itself, consisting as it does of seawater
cleared of seston, and in otherwise static conditions will cause some mixing
and result in seston dilution near the inhalant. As the seston concentration
declines, inhibition of filtration may result. The effect of small increments in
velocity is to help remove the localized buildup of seston-diluted seawater
around the scallop.
b. Beyond the point where the ambient flow is sufficient to remove completely
any localized seston dilution around the scallop,flowhas no further influence
on filtration. Theflat-toppedunimodal response seen and lack of a continuing
increase in filtration rate with increasing velocity argue against the latter
being used for enhancement of filtration rates, as is suggested by LaBarbera
(1977) in brachiopods.
c. As ambient velocity further increases, the pressure field around the animal
will be altered. The difference between ambient pressure at the inhalant (PA)
and the exhalant (PB) may become significant with respect to the pressure
difference generated by the ciliary pump. If PA PB, seawater will tend to be
forced through the system at a rate greater than can be handled by the ciliary
pump filter, as described by J0rgensen et al. (1986a). This causes a behavioral
response involving valve and mantle wall partial closure, thereby reducing the
volume that can be processed. The internal homeorheostat (Wildish and
Saulnier 1993) can to some extent overcome the closing response when seston
concentration is sufficiently high (Fig. 4.5).
d. At high ambient velocities, the siphons retract and close because they cannot
filter. This velocity is not fixed but variable for the scallop because seston
concentration and other environmental variables, such as temperature, can
influence the endpoint.
Behavioral responses of a bivalve to external pressure fields which

oppose the internal pressure field created by the ciliary pump can be
expected in phases c and d. The ciliary pump in some bivalve molluscs
can generate pressure differences of the order of 0.1-2.0 mm of water
(Foster-Smith 1976). If the external pressure differences are of the same
order of magnitude, it can be predicted that the flow through the pump,
in the absence of behavioral response, will be either much reduced
or considerably enhanced, depending on the sign of the pressure
differential.
If the scallop is oriented with the inhalant pointing upstream, the
pressure difference between the inhalant and the exhalant is given ap-
proximately by the stagnation pressure. The streamlined shape of the
0.2-
R
0.1-
Figure 4.5 The seston uptake rate, R, as arbitrary units, is a function of velocity,
U, in cm-s 4 , and seston concentration, C, as cell number-ml"1 of Chroomonas
salinus (Wildish et al. 1992).
valves would suggest that the pressure defect would be relatively small in
the vicinity of the exhalant, which will lie in the turbulent wake of the
scallop. The bulk stream velocity required to give a stagnation pressure
of 1.0 mm of water is 14cm-s"1. Because the flow through the trophic
fluid transport system is laminar, the volumetric flow is linearly related to
the external pressure difference (see J0rgensen et al. 1986a). If the
scallop's ciliary pump can generate a pressure difference of 1.0 mm of
water, the volumetric flow through the animal can be doubled by an
external pressure difference of the same magnitude. How the ciliary
pumping system of other bivalve molluscs would respond to extreme
pressure differences caused by ambient flow is not clear. As has been
demonstrated by Wildish and Saulnier (1993), by direct video viewing of
the exhalant siphon, the live scallop reacts behaviorally by partially
closing the mantle and valves in order to resist the external pressure
Table 4.4. Scallop seston uptake rates as affected by flux, seston

concentration, and velocity?
Seston Seston
concentration Velocity Flux rate uptake rate
cells -mr 1 cm s"1 cells cm2 s"1 jag Chi a-g wt-h"1
10 50 500 0
100 5 500 0.018
50 5,000 0
1,000 5 5,000 0.09
50 50,000 0
10,000 5 50,000 0.30
50 500,000 0.24
100,000 5 500,000 15.0
Source'. Extrapolated from Wildish et al. (1992).

a
Note: Seston is a unialgal culture of Chroomonas salinus.
difference. This results in a reduced volumetric flow rate of inhaled

seawater for filtration and hence feeding rate at increased velocities.
If the scallop is reversed with the inhalant pointing downstream, the
velocity at which the ambient pressure differential is just balanced by a
ciliary pump delivering a 1.0-mm head is, as noted previously, 14cm-s"1.
For a head of 2.0 mm, the required velocity is 20 cm s"1. The reversed live
scallop must maintain its ciliary pumping work rate but, in the
unfavorable pressure conditions, cannot maintain its filtration through-
put, as suggested by J0rgensen (1990) in mussels. Evidence that growth
rates of live scallops forced to oppose unimodal flume velocities
>12cm-s~1 is reduced is presented by Wildish and Saulnier (1992).
In addition to velocity, seston concentration itself has been shown by
Wildish and Saulnier (1993) to interact to influence the degree of valve
and mantle edge opening, as indicated by seston uptake (Table 4.4). For
a unialgal culture of Chroomonas salinus, filtration rates by individual,
starved scallops were maximized at a cell concentration of ~1 x 103
cells-ml"1 and at a constant ambient velocity of Scm-s'1. Flume experi-
ments in which velocities and seston concentrations were widely varied
(Wildish et al. 1992) and giant scallop filtration rates determined show
that the maximum uptake rate occurs at -25 cm s"1 when the optimum
seston concentration (~103 cells-ml"1 of C. salinus) is offered. Presum-
ably, at this optimum point, giant scallops arefilteringnear the maximum
intrinsic capacity of their pump. The ambient velocity at which scallops
Flow
Seawater velocity
direction
Mantle. Gills and Mouth Palps
photoreception
mechanoreception
chemosensory
Sensors 1
tactile
pressure
Integrated control
r
Nervous System
Hormones
i
Mantle Gills Mouth Palps Diaestive Svstem Growth
Responses opening pump on seston selection feeding rate positive
closing pump off seston rejection satiation negative
gill bypass to tissue
seston selection to gonad
to shell
Figure 4.6 Environmental/physiological model of giant scallop feeding and

growth (Wildish and Saulnier 1993).
stop filtering is directly proportional to seston concentration (Fig. 4.5).

At seston concentrations greater than the optimum, sufficient ration can
be obtained at lower clearance rates, and at both 104 and 105 cells ml"1 of
C. salinus, filtration was still occurring at up to 45cm-s~1. This was the
maximum velocity achievable with the recirculating Vogel-LaBarbera
type flume used in these experiments (Wildish et al. 1992).
The demonstration that giant scallop filtration rates are controlled by
at least two independent environmental variables, velocity and seston
concentration (Fig. 4.5), suggests an integrated environmental-
physiological model for feeding and growth (Wildish and Saulnier 1993).
This model (Fig. 4.6) depends on a constant output ciliary pump whose
filtration rate can be controlled by closing/opening valves and mantle
edge according to changes in these environmental variables. The evi-
dence also indicates that velocity and seston concentration are interac-
tive so that pumping and filtration only continue if it is nutritionally
advantageous for the scallop. This implies central physiological control,
perhaps by the nervous system, in which the bioenergetic profit and loss
can be optimized.
The model incorporates evidence that ectocrines (substances released
from cells which elicit chemosensory responses in target organisms)
associated with seston alter the filtration response by causing closing/
opening of the valves, which automatically control volume throughput.

Ward, Cassel, and MacDonald (1992) showed that filtration of the giant
scallop is stimulated by preferred species of microalgae. When toxic
dinoflagellates were fed to giant scallops, they caused valve clapping and
closure (Shumway et al. 1985a). Little is known about the structure of
sensors involved in the physiological control of feeding in the giant
scallop (Beninger 1991).
Surprisingly few authors have investigated the effect of velocity on
mussel filtration/feeding or growth rates. Walne (1972) claimed that
growth rate in blue mussels Mytilus edulis was positively linked to in-
creasing volumetric flow in the range tested (50-700 ml -min"1 in 2.4-L-
capacity boxes). On the other hand, Hildreth (1976) claimed that the
pumping rate of blue mussels was unaffected by volumetricflowsfrom 33
to 700 ml-mill"1. These studies have led to some confusion (e.g. Vogel
1981), but, in both cases, the mixing characteristics within the experimen-
tal chambers, the degree of refiltration by the mussels, or velocities near
the mussel inhalant were not determined, although the latter are likely to
have been <lcm-s"1, and the results remain apparatus specific and not
referable to the much wider range of velocities experienced in the field.
Walne's (1972) results could be explained by small velocity increases
helping to remove seston-depleted water near the mussel.
Preliminary measurements of blue mussel filtration rates as a function
of velocity in the range 6-38 cm-s"1 have been made by Wildish and
Miyares (1990) in a recirculating Vogel-LaBarbera flume of 90-L capac-
ity. The velocity measurements were made with the aid of a small rotor of
11 mm diameter (see Chap. 2) placed in front of the mussel inhalant, with
the mussel firmly attached by the valve to the flume floor so that the
siphons opposed the flow. At velocities in the range 6-25 cm s"1 and at a
constant seston concentration of 104 cells of Chroomonas salinus-ml"1,
filtration rates were inversely related to velocity. At velocities >25 cm s"1
and at the same seston concentration, the filtration rate was zero or a
small percentage of the maximum possible at 6cm-s"1. Although these
results could be interpreted as part of a unimodal function of velocity
where only the c and d stages were observed, the velocity range tested
needs to be widened and seston concentrations tested need to be varied
before a conclusion can be reached.
Growth studies in which velocity is varied while keeping seston con-
centration constant are not available for blue mussels. The flow range
examined by Wildish and Kristmanson (1985) in growth experiments
with blue mussels of 0.1-3.89 cm-s"1 is not considered wide enough to
detect the unimodal response of feeding/growth to velocity. Bayne,

Thompson, and Widdows (1976) suggested that the relationships among
velocity, filtration rate, seston concentration, and respiration rate in
mussels need to be elucidated, but as far as we are aware, this has not
been done.
Seston concentration also affects the filtration rate of Mytilus edulis.
An optimum rate at 75-200 x 103 cells of Phaeodactylum tricor-
nutum - ml"1 was obtained in static seawater test conditions (Foster-Smith
1975). An indirect method was used to determine this with animals of 45-
to 52-mm shell height and 2.56- to 4.00-g wet weight. The mean filtration
rate was 1.47 0.20 L-h"1 per individual or 0.39 +0.04 L-h^-g"1 wet
weight. Widdows, Fieth, and Worrall (1979), working with the same
species and a similar indirect technique, showed that filtration rate was a
unimodal function of the natural seston concentration (expressed as dry
weight per litre, inclusive of sedimentary particles, and therefore not
relatable to Foster-Smith's results). The optimum seston concentration
increased as animal size became larger. Riisgard and Randlov (1981) also
studied the filtration rate of Mytilus edulis over a wide range of concen-
trations of Phaeodactylum tricornutum. They found that filtration was
independent of seston concentration between 1.5 x 103 and 30 x 103
cells ml"1, but at higher concentrations there was a decrease in filtration.
At seston concentrations lower than 1.5 x 103 cells of P. tricornutum, a
decrease in filtration rate resulted from inhalant and valve closing.
Seston quality is a factor which has been extensively investigated in
relation to filtration and feeding in the blue mussel. Mytilus edulis is
considered to be an indiscriminate active suspension feeder whose filtra-
tion rate is not usually stimulated by ectocrines (Ward and Targett 1989),
although seston quality, e.g. exudates from toxic dinoflagellates, may
directly inhibit filtration or cause valve closures (see review by Shumway
and Gainey 1992; Ward and Targett 1989). Some bacteria have also been
reported to be inhibitory to filtration in the blue mussel (Birbeck
and McHenery 1982; McHenery and Birbeck 1986; Birbeck, McHenery,
and Nottage 1987). Ward and Targett (1989) have suggested that
the response mechanism involves epicellular ectocrines which are cues
for particle selection after capture but before ingestion. This response
would be integrated by the nervous system, which then reduced, or
shut, the valves, thus limiting or stopping filtration. Such a response
would be consistent with the model shown in Fig. 4.6. One criticism of
the work done on the effect of seston quality on bivalve filtration is that
ambient flows around the feeding bivalve have not been simulated, and
10
me ifc sbc
-Vmn **&-
Figure 4.7 Diagrammatic representation of the blue mussel pump with typical
static pressures (J0rgensen et al. 1986a) at points along the trophicfluidtransport
system: A, relative pressure heads; B, diagram of pump; hx and h2 hydrostatic
pressures at inlet and outlet; A/Z12, difference between hx and h2 or back pressure;
in, and ex, inlet and exit losses of pressure; io, inhalant opening; me, mantle
cavity; If, laterofrontal cirri; p, pump; lc, lateral cilia; ifc, interfilament canal; f,
filament; sbc, suprabranchial cavity; es, exhalant siphon; 1, interfilament canal
width; vx and v2, velocities at inhalant and exhalant.
it remains to be seen whether this would change the conclusions already

reached.
J0rgensen et al. (1986a) have provided a hydrodynamic model of the
blue mussel pump useful in estimating inhalant and exhalant pressures.
Pumping power is created by the metachronal beating of lateral cilia
which are located at the upstream ends of the interfilament canals. There
are resistances associated with various parts of the trophic fluid transport
system, and these were estimated from their dimensions by J0rgensen et
al. (1986a) to be of the order shown in Fig. 4.7. The pump characteristic
was calculated from estimated resistances, pressure changes of the whole
system, and volume flow rates to give the relationships among pump
power, pressure heads, and volumetric flow rate (see J0rgensen et al.
1986a).
The measurement of bivalve pump characteristics, e.g. velocities at
inhalant and exhalant, is rendered difficult by the rapidity of opening/

closing of the mantle walls. According to J0rgensen and Ockelmann
(1991), when this occurs, ciliary beating (in young, transparent-valved
individuals) abruptly stops when the exhalant closes and immediately
begins beating once the exhalant opens again. Apart from this, velocity
distribution in the exhalant is that of laminar flow in a tube with mid-
point velocities twice that of the bulk velocity. The difficulties involved in
direct measurements of velocity or pressure at exhalant openings were
emphasized in Chapter 2.
Pumping rate varies with the degree of valve gaping or, more precisely,
with the degree of openness of the inhalant and exhalant siphons (Fos-
ter-Smith 1976). Maximally open siphons produce the maximum pump-
ing rate. Retraction of the siphons causes a shortening of the gill axes and
thus of the demibranchs, which results in a reduced lumen diameter of
the interfilament canals (J0rgensen 1990). Complete siphon retraction
and closing reduce the distance between the extended tips of opposing
lateral cilia from ~10|im to zero; this is suggested to be a general prop-
erty of suspension feeding bivalve gills (J0rgensen and Riisgard 1988).
J0rgensen (1990) concludes that pump power and volumetric capacity
are probably direct functions of interfilament canal width, which is
ultimately controlled by valve gaping and the linked degree of siphon
retraction.
The relationship between seawater viscosity and the pumping rate of
blue mussels (J0rgensen, Larsen, and Riisgard 1990) clearly shows the
dependence of pumping rate on kinematic viscosity at different ambient
temperatures. This effect was considered to be due to the resistance to
water flow in the interfilament canals of the trophic fluid transport sys-
tem. J0rgensen and Ockelmann (1991) showed that in young mussels
lateral cilia beat with a mean frequency of -10 Hz at 14C to -15 Hz at
20-21C, corresponding to a Q10 2. Rapid changes of temperature in
intact mussels are reported by J0rgensen (1990) not to lead to obvious
effects on the pumping rate, despite the increased ciliary beating.
Physiologists interested in the energetics of filter feeding in bivalves
such as mussels (e.g. Bayne and Newell 1983) have hypothesized that
filtration is internally controlled, e.g. by gut fullness, to maintain a con-
stant feeding rate over a wide range of environmental conditions. Al-
though not supported by direct evidence, it is commonly held (Winter
1978) that ciliary pumping by bivalves is energetically expensive. Inter-
nal physiological control of filtration would thus be homeorheostatic by
allowing the conservation of energy required for ciliary beating. An
alternative view formulated in J0rgensen (1990) is that ciliary pumping

(see also J0rgensen et al. 1986b) is inexpensive and, anyway, maintained
at a constant rate irrespective of the pump output (see also Chap. 9, the
section, Ultimate questions). Differences in pump output, and hence
filtration rate, were a simple function of the degree of openness of the
valves and siphons (J0rgensen 1990). Some additional evidence is avail-
able for J0rgensen's view; for example, Wildish and Saulnier (1993)
found that the scallop exhalant opening area was a function of ambient
velocity consistent with regulation of valve closure to control seston
uptake and growth rates by velocity. Stenton-Dozey and Brown (1994)
also found that in the South African clam Venerupis corrugatus the
degree of siphon openness correlated positively with clearance rates and
that increasing clearances were achieved without an increase in respira-
tion rate. Both competing hypotheses outlined are compatible with the
environmental-physiological model of feeding proposed by Wildish and
Saulnier (1993).
Bryozoans
The phylum Bryozoa consists of colonial forms, usually attached to the
substratum as an encrusting, stolonate, or bushy growth. The individual
zooid of the colony is protected by a non-living chitinous skeleton, some-
times strengthened by calcareous secretions. The characteristic tentacle-
bearing lophophore is the structure concerned with suspension feeding.
Observation of feeding Bryozoa in a static experimental system, where
the seawater was slowly mixed by a peristaltic pump, was undertaken by
Best and Thorpe (1983) in Flustrellidra hispida. This species is an epi-
phyte of the brown macroalga Fucus serratus, which is found in the lower
intertidal zone of the Isle of Man. The experimental method involved
viewing the lophophore at lOOOx magnification with a stereomicroscope
to observe the rate of movement of the seston particles (= Tetraselmis
suecica of ~10-|Lim diameter at all densities up to 2xlO 5 cells-ml"1).
Because the particles were moving quickly across the field of view, it was
necessary to tune an oscilloscope trace to the same speed so that from
the distance travelled (=fieldof view diameter), a particle speed could be
calculated. Results showed that starved Flustrellidra could increase the
seston particle speed if increased concentrations were supplied by up to
130%. Best and Thorpe (1983) noted that speeds were variable within
the lophophore and further investigation revealed that particles located
at the centre of the lophophore had the greatest speed. Rates of particle
movement depended on ciliary currents generated by cilia present on the

lophophore tentacles and ciliary speed varied in time, related to satia-
tion. Best and Thorpe (1986) extended their investigations to five other
species locally available around the Isle of Man. They found that starved
zooids also responded to increased seston concentration by increasing
the ciliary flow and that this resulted in a higher capture rate of a
preferred food particle. They also found that feeding current velocities
were positively linked to the lophophore height because increased tenta-
cle length increases the number of cilia available to drive seawater
through the lophophore.
Okamura (1984, 1985) has studied the effects of ambient flows on
feeding of Bugula stolonifera, an arborescent species, and Conopeum
reticulum, an encrusting species. Bugula and Conopeum were present on
floating docks in the Pacific Ocean off California. Feeding success was
determined by using latex beads (mean diameter of 11.9 |im 1.9 |im
standard deviation [SD]) offered at 1120-1190 beads-ml 4 . Feeding ex-
periments were of 20-min duration during which there was no settling
out of the beads in the flume. The beads captured by individual zooids
were counted after treatment with sodium hypochlorite, which left the
exoskeleton and beads intact. Small Bugula colonies showed inhibition
of feeding success for individual zooids at the higher flow velocity tested
(10-12cm-s"1). Okamura (1984) suggests that this is because the sestonic
particles are moving at a greater speed past the lophophore, making it
more difficult to divert the streamlines into the lophophore. In addition,
it is suggested that once they are in the lophophore, it may be more
difficult to retain the particles for a sufficient time to effect transfer to the
mouth as a result of the drag force caused by the current. Larger, more
bushy Bugula did not show this effect of velocity on feeding because
flows are decelerated as they pass through the branches, the colony
changes its shape, and central or downstream, flow-protected zooids are
the ones feeding at the greatest rate. In addition to flow protection,
Okamura (1984) thought that the lack of flow inhibition in bushy forms
could be due to a greater availability of seston because of the poor
feeding effectiveness of upstream zooids and perhaps greater ciliary
pumping by the flow-protected zooids.
In the encrusting bryozoan Conopeum reticulum, which were grown
on glass microscope slides in the field, feeding experiments were con-
ducted in the laboratory by placing slides normal to the flow in a small
recirculating flow tank. Feeding success was determined as for Bugula
and tests were made at three velocities: slow (1-2 cm-s"1), intermediate
(4-6 cm s"1) and fast (10-12 cm s"1). The results show an inverse relation
of feeding with velocity, and both large and small colonies were unable
to accumulate beads at higher rates. At low and moderate velocities,
zooids from larger colonies were more effective in feeding than small
colonies, perhaps because the larger number of zooids could better di-
vert seawater to the colony.
Subsequently, Okamura (1987, 1990) tested the interactive effects of
particle size and velocity on the feeding success of Californian Bryozoa.
She was able to show in Bugula stolonifera that medium and large
particles (14.6 1.0 and 19.1 1.1 Jim) were ingested at rates greater
than small particles (9.60.5jim). Flow influenced the particle size in-
gested by B. stolenifera so that this bryozoan fed preferentially on larger
particles at the highest flow tested (10-12cm-s~1). Particle size alone
affected the feeding success of Bugula neritana; the medium and larger
particles were preferred. The mechanism of selective feeding may in-
volve size-dependent particle behavior, active selection, or rejection in-
volving alternative feeding mechanisms such as tentacle flicking.
Sieving suspension feeders

Mucous filter nets have been found in a wide range of taxa, including
ascidians, appendicularians (Flood and Fiala-Medioni 1982), some
vermetid gastropods (Hughes 1978), and polychaetes such as Nereis
diversicolor and Chaetopterus variopedatus (Wells and Dales 1951;
Riisgard 1989, 1991).
Ascidians, or sea squirts, are attached benthic animals which may be
solitary or colonial and are characteristically suspension feeders. A sin-
gle feeding zooid of the colonial ascidian Clavelina species (Fig. 4.8) has
two openings - the buccal siphon for inhalant flow and the atrial siphon
for exhalant flow. The pharyngeal side walls are punctuated by gill slits
or stigmata. The inhalant flow is generated by cilia lining the stigmata
which cause inhaled seawater to pass around the pharynx, through the
stigmata to the atrium, then out via the atrial siphon. Sestonic particles
are captured on a mucous net which is secreted by the endostyle and the
whole thing periodically ingested. When feeding conditions are suitable,
the endostyle continuously produces the mucous net.
Randl0v and Riisgard (1979) have studied the effect of sestonic parti-
cle size (in the range 1- to 7-jLim equivalent spherical diameter) on the
fractions retained by Ascidiella aspersa, Molgula manhattensis, Clavelina
lepadiformis, and Ciona intestinalis. The small particles used originated
buccal siphon
A B atrium
atrial siphon
pharynx
stigmata
mucous strand
mucous rope
endostyle
Figure 4.8 A, Single feeding zooid of Clavelina; B, transverse section of

Clavelina (Carlisle 1979).
from those already present in the seawater system. Cultures of

Monochriysis lutheri and Dunaliella marina were added at a concentra-
tion of 1 x 104 cells-ml"1 to increase the seston size range. Samples were
collected from the inhalant and exhalant areas as in M0hlenberg and
Riisgard (1978) and the concentration determined in an electronic parti-
cle counter/sizer. Results showed that seston of equivalent diameter in
the range 2-3 to 7 |im were completely retained, while particles of <1 |um
diameter had a particle retention efficiency of -70%. Within the experi-
mental error likely in measuring mesh size and retention efficiency, the
results provide support for the hypothesis that food capture in these
ascidians is by sieving. Fiala-Medioni (1978), working with at least one of
the same species, C. intestinalis, suggested that up to one-third of
Monochrysis lutheri (6-10 x 2-3 |im in size), offered as food at 20 x 106
cells I"1 and 15 1C,passed through the filter. As Randl0v and Riisgard
(1979) point out, this result was not obtained by direct observation of
retention efficiency but by calculation from direct observation of pump-

ing rate and indirect observations based on particle clearance rates, and
so may be inaccurate.
Robbins (1984) studied the effect of particle concentration on feeding
of Ciona and Ascidiella spp. and found that ingestion was reduced at
higher concentrations of fuller's earth (mean particle diameter of 25-
30jjm). This was a result of a reduction in pumping rate and an increase
in the frequency of squirting, but there was no change in retention
efficiency, which remained at 100%. Squirting is a sudden muscular
contraction of the zooid which causes excess sestonic food to be rejected
from the pharynx.
The ascidian pump has been analyzed in a manner similar to that of
the mussel by Riisgard (1988). The pump in Styela clava, by comparison
with that in the mussel, is a much lower pressure/energy system. Flow
through the filter is calculated to be 0.03 mm s"1 and the pressure drop
across it to be 0.069mm H2O (using a net dimension of 1.4 x 5.4|um, this
value has been recalculated; see Riisgard 1989). The maximum pressure
rise in the trophic canal system is 1.2 mm H2O. The pumping rate is
reduced by closing the atrial siphon, but this does not automatically
reduce pressure, as it does in bivalves, because the stigmata cilia may
continue to beat. Factors which result in the cessation of ciliary beating
include a lack of sestonic particles or direct physical disturbance of the
ascidian.
Most of the ascidians which have been studied are found in the
nearshore continental shelf area. Carlisle (1979) reports that some con-
tinental slope stalked ascidians such as Bolteniopsis are specialized for
opposing a steady external current and passively receive seawater and
seston in a much enlarged buccal siphon. Concomitantly, there is a great
reduction in the branchial basket, which may no longer be needed as a
pump. Another example is provided by a stalked ascidian, Styela
monteryensis, which is adapted to oscillating wave surge channels of
exposed shores of the Pacific Ocean on the U.S. West Coast (Young and
Braithwaite 1980). Characteristic of this species of Styela is a spar buoy
system (see Chap. 6) in which wave forces passively reorient the stalked
ascidian (Fig. 4.9) which is articulated near its base so that the buccal
siphon faces the flow either on the onshore wave or on its back surge. It
was shown by field experiments that the time required for seawater to
pass through the trophic fluid transport system depends on ambient
velocity conditions. It was also found that seawater was pushed through
by dynamic force from waves, perhaps assisted by viscous entrainment. It
Figure 4.9 Styela monteryensis: responses to oscillating flows (Young and

Braithwaite 1980).
is thus possible that these specialized ascidians are combined passive-

active suspension feeders. Manahan et al. (1982) have shown that this
species can remove free amino acids from seawater, although the impor-
tance of dissolved organic matter in its nutrition remains unclear.
Other species of considerable interest are polychaetes such as
Chaetopterus variopedatus and Nereis diversicolor which are both capa-
ble of producing mucous nets and capturing microscopic food, probably
by sieving. Chaetopterus lives in a parchmentlike tube secreted by the
epidermis and attached to stones, or in soft sediment in a U-shaped
burrow (Fig. 4.10). The parapodia of segments 14, 15, and 16 are disc
shaped and beat to produce a flow of water through the tube. Flow
direction is always from the head to the tail. The mucous net used in
sieving food particles is produced by the modified parapodia of the
mouth ;.;'./.;..;.;:..:;.
parapodia of
'^segments 14-16
Figure 4.10 Life form of Chaetopterus variopedatus in soft sediments (Flood

and Fiala-Medioni 1982).
twelfth segment. The ciliated dorsal groove and cuplike organ of the
thirteenth segment gather the mucus and roll it into a food ball. When
the ball is sufficiently large, the parapodia stops pumping and the food,
plus mucus, is transported anteriorly by cilia in the dorsal groove to the
mouth for ingestion.
The Chaetopterus variopedatus pump is a muscular one. Riisgard
(1989) has used his constant-level chamber to investigate these worms
from the Swedish west coast. The methods employed were similar to
those used with mussels (J0rgensen et al. 1986a) and ascidians (Riisgard
1988). Video recordings of Chaetopterus confirmed that the mechanism
involved a positive displacement pump and that parapod beating de-
creased with increasing back pressures. Imposed back pressures set up in
the constant-level chamber caused the worm to reverse itself in the tube.
The maximum pressure rise in the tube and the total head loss are
approximately twice those in bivalves and five times those in ascidians,
consistent with the filter area's being relatively smaller than in those
species. The pressure drop across the mucous net as estimated by the
Tamada-Fujikawa equation requires knowledge of mesh opening size
and diameters of the mesh fibres. Estimates made by J0rgensen et al.

(1984) of mesh size based on observations of particle retention efficiency
at different particle sizes suggest that the C. variopedatus mucous net
meshes are three times larger than suggested by Flood and Fiala-
Medioni (1982). This results in a pressure drop estimate of 0.7 mm
(J0rgensen's estimate) versus 3.0-4.5rnm of water for the mesh size
suggested by Flood and Fiala-Medioni (1982). Even with the lower esti-
mate, the pressure drop across the net in the tubicolus worm is several
times greater than in the ascidian.
A field study of the pumping activity of the facultative sieving suspen-
sion feeder Nereis diversicolor was reported by Videl et al. (1994), who
used a phototransducer system to record long-term seasonal patterns of
pumping activity indicating filtration activity. Observations showed that
the worms also go to the sediment surface to deposit-feed on benthic
diatoms just after low tide at one, but not the other, Danish location
studied. During the summer, Nereis was suspension feeding for 50%-
100% of the time, and this behavior seemed to be triggered by a thresh-
old chlorophyll a concentration of l-Sjig-L"1. During the spring and
autumn when chlorophyll a was scarce, the filtration activity occurred
only 5%-20% of the time.
Facultative active suspension feeders

In some benthic suspension feeders, the filtration-feeding behavior
changes with velocity so that at low velocities active filtration occurs,
while above a critical higher velocity passive suspension feeding
occurs. Although LaBarbera (1984) lists some ascidians - e.g. Styela
monteryensis (Fig. 4.9), some brachiopods, and some sponges - in this
category, their filtration/feeding responses to velocity have not been
adequately examined. The status of these forms is considered to be
unclear, although tentatively they are placed in a separate combined
passive-active category (discussed later, in the section, Combined pas-
sive-active suspension feeders). The taxa best known as facultative active
suspension feeders are the cirripede barnacles, but some decapod
Crustacea also have representatives in this category (Trager et al. 1992).
Barnacles
It was determined by Crisp and Southward (1961) that changes in veloc-
ity can alter the type of feeding behavior of a number of adult cirripede
PASSIVE
ACTIVE FEEDING
FEEDING
C J
C>3 cm-s" 1 )
-\ I 1
forward stroke recovery stroke switch
Figure 4.11 Active and passive feeding in Semibalanus balanoides (Trager et al.
1990).
barnacles. They described a fast-beat rhythmic sweeping of the cirri

through slowly moving seawater and a non-rhythmic full extension of the
cirri held rigidly opposed to the direction of flow when the ambient flow
was more energetic. These different means of suspension feeding
correspond to active suspension feeding at low velocity and passive
suspension feeding at higher velocities. Barnacles, such as Semibalanus
balanoides, are found on those rocky intertidal Atlantic* Ocean shores
where tidal currents and wind-wave activity are usually energetic and
are constantly changing. In a flume study of this species, Trager et al.
(1990) investigated the effect of velocity near the barnacle of up to
9 cm s"1 on feeding behavior. The recirculating flume (capacity 15 L) had
a sophisticated electronic controller which could reverse the propeller,
thereby creating oscillating flows with sine wave acceleration or decel-
eration. Velocity was measured from dark field, laser optical pathway
video frames of sestonic particle movement taken at known intervals.
The mean velocity at which adult cirripedes changed from active to
passive feeding was determined (Fig. 4.11) to be 3.10 0.73 cm-s 1

(SD). The actual range of switch velocities from 22 experiments was
1.84-4.83 cm -s"1. From Fig. 4.11, it can be seen that the active cirral
beating involved rhythmic movements in the same direction as the slow
flows (individuals positioned carinorostrally to theflow).In passive feed-
ing, the cirral fan is held rigidly into the flow so that the concave surface
is opposed to it. If the flow direction and position of S. balanoides were
as described previously, passive feeding involves reversal of the cirral fan
through 180. Trager et al. (1990) could make the cirripedes "dance"
(change the orientation of their cirral fans) by rapidly changing the flow
direction in the flume. This suggests that the response is controlled by
the nervous system with accelerating or decelerating flows acting as
the stimulus to cause the reflex responses observed. Further work by
Trager, Hwang, and Strickier (1992) using a recirculating computer-
controlled oscillating flow tank confirmed that in two species of barna-
cles, S. balanoides and Sarignium milleporum (from the Red Sea), flow
oscillation frequencies in the range 0.16-0.65 Hz, with flow velocities
ranging from 6.2 to 12.1 cm -s"1, stimulated reorientation of the cirral
fan. Reorientation occurred during passive feeding so that the concave
surface of the fan always faced upstream, where it is known to be most
efficient at capturing suspended particles. Behavioral anticipation of
flow change appeared to be based on water deceleration that occurred
during the oscillation cycle. Trager et al. (1992) also showed that barna-
cles subjected to increasing oscillation frequencies spent less time
in passively filtering and relatively more in removing particles filtered
and in re-orienting the fan. Thus, the ratio between the time spent
feeding to non-feeding activity depended on the wave oscillation fre-
quency so that at high oscillation frequencies less time was available for
filtration.
Hunt and Alexander (1991) have used an endoscope developed for
medical research to observe movements of the mouth parts and cirral fan
of a tropical cirripede, Tetraclita squamosa, collected from the east coast
of Australia. They were able to show that a wide range of particle sizes,
including zooplankters and microplankton, were filtered and utilized as
food. They include a description of the ultrastructure of the thoracic
appendages and how they are used in capturing the various prey sizes. A
more detailed description is provided of six different cirral activity pat-
terns, including that of the passive and active cirral fan behavior de-
scribed by previous authors.
v.
Figure 4.12 Diagrammatic view of a model sponge in a flow where a benthic
boundary layer is present (Vogel and Bretz 1972).
Combined passive-active suspension feeders

Unlike in facultatively active suspension feeding, in which ambient
velocity conditions provide the environmental switch to signal
change from, or to, passive feeding, in combined passive-active suspen-
sion feeding the passive and active mechanisms appear to continue
simultaneously.
Sponges
Model studies were undertaken by Vogel and Bretz (1972) in sponges,
which have many small incurrent openings through the body wall. These
openings connect to the spongocoel and an excurrent opening or
osculum which opens dorsally in the animal so that it is high in the
benthic boundary layer (Fig. 4.12). Besides an active flagellar pumping,
passive flow in the sponge is induced by the Bernoulli effect resulting
from the velocity and consequent pressure differences at the incurrent
and excurrent openings caused by their different positions within the
benthic boundary layer. Experiments with live erect sponges
Halichondria bowerbanki obtained from the Massachussets coast, were
made at ambient flows from 0 to lOcm-s"1 (Vogel 1974). Flow measure-
ments at 4 mm into the osculum and 3 cm in front of the sponge were
made to record the osculum and ambient velocities, respectively. By

immersing live sponges in freshwater, it was possible to inhibit
choanocyte pumping. At zero velocity, live Halichondria of 3.1cm
height x 1.2 cm diameter produced an internal flow measured at the
osculum of 2.5 cm s"1. As ambientflowswere increased, both normal live
and choanocyte-inhibited live Halichondria increased internal flows at
the same rate proportional to ambient velocity. At ambient flows
>6cm-s^, most of the total flow through the spongocoel was due to
passive flow induction. Vogel (1974) thought that active pumping in
Halichondria was unaffected by ambient flow and merely added on to
the passive flow, which was directly related to the Bernoulli effect and
ambient velocities. Further results using similar methods (Vogel 1977),
but on eight species of live sponges obtained from Bermuda, showed
similar results. In this case, the velocity measurements were made in the
field at depths of 1-2.5 m, without removing the sponges. The osculum
flow rates (when ambient flow = 0) obtained were much higher than
obtained in the lab studies with Halichondria. For example, in Verogia
fistularis and Haliclona viridis, the osculum flow rate was 21cm-s~1,
whereas in Ircinia fasciculata it was llcm-s" 1 . These results may have
differed from the laboratory ones because of difficulties in handling the
sponge and in simulating suitable environmental conditions.
Riisgard et al. (1993) examined the effects of temperature on filtration
rates in the sponges Halichondria panicea and Haliclona urceolus by
using a direct and an indirect method. The sponge pump characteristics
were also estimated by a model and methods previously employed
(Famme, Kristensen, Larsen, M0hlenberg, and Riisgard 1986a). The
maximal operating point of the pump was 0.673 mm H2O. The pump
work was less than 1 % of the respiratory output, suggesting a very low
energy cost for filtration in sponges. Because of this low cost, Riisgard et
al. (1993) considered it unlikely that passive increments toflowas postu-
lated by Vogel and Bretz (1972) would be relatively important.
Vogel (1978) has postulated that a now extinct fossil group present in
the Cambrian period - the Archaeocyanthids - may have been purely
passive feeders relying on strong local currents to supply seawater for
passive suspension feeding in the manner described for modern sponges.
Lampshells
LaBarbera (1977) proposed that active suspension feeders such as
lampshells or brachiopods may use ambient flows to "augment or even
supplant their active pump," as has been suggested for sponges. Model
studies showed that induction of water movement through the dead
valves of Terebratulina unquicula, Terebratalia transversa, and Laqueus
californianus did passively occur at ambient flows of 2-4 cm s"1. It is also
known that brachiopods may actively pump at zero flows. However, as
pointed out by LaBarbera (1977), there is no evidence available to show
that the filtration rate increases as ambient velocities are increased.
The bilateral symmetry of brachiopods (Chap. 6) with two incurrent
openings 180 apart allows them to utilize oscillating flows if the ante-
rior-posterior axis of the valves is aligned parallel to the current direc-
tion. There is also evidence that some lampshells can behaviorally
reorient toflowdirection by means of pedicel musculature (see Chap. 6).
LaBarbera (1977) proposed that passive flow augmentation through the
brachiopod lophophore resulted in increased feeding rates and/or in
increased availability of dissolved oxygen and removal of wastes. Vogel
and Bretz (1972) and LaBarbera (1977) suggested that their ambient
velocity passive enhancement theory could apply to other species, e.g.
scallops. An independently derived theory, the inhalant/exhalant pres-
sure field theory, was proposed by Wildish et al. (1987) to explain
behavioral orientation to flow direction by scallops, which might also
apply to brachiopods. Certainly, in future studies, both theories should
be considered in designing experiments and testing specific hypotheses.
According to the inhalant/exhalant pressure field theory applied to
brachiopods, the lampshells would be considered as active suspension
feeders, and the low exhalant flows measured by LaBarbera (1981) in
still water indicate that the ciliary pump is much less effective than in
bivalves. Exhalant velocities of Terebratulina unquicula, Terebratalia
transversa, and Laqueus californianus ranged from 0.2 to 1.2 cm-s"1 and
flow within the trophic fluid transport system was laminar. Referring to
Fig. 6.5, the following orientation mechanisms are possible:
A. When the exhalant flow directly opposes the ambient flow direction and
when the ambient flow pressure exceeds the ciliary pump pressure either
extra work must be done by the pump to overcome it or the valves partially
close (as in scallops) to reduce the differential.
B. With the exhalant downstream with respect to ambient flow direction and as
the flow builds up so that the external pressure tends to force flow across the
ciliary pump at a rate faster than that maintained by the cilia alone, then
either the valves partially close to reduce it or filtration retention efficiency
declines.
C. In this orientation the valves present the maximum amount of drag opposing
the flow.
D. The valves present minimal drag to theflowand, as in C, the inhalant and

exhalant can be oriented optimally for either uni- or bidirectional flows.
These represent suggestions (see Chap. 6, the section, Sessile epifauna,
for an alternative view) and should be worked up further as null and
alternative hypotheses for experimental testing.
Deposit-suspension feeders
Some species of benthic infauna have the ability to switch between
deposit and suspension feeding. Ambient flow velocity is the environ-
mental cue thought to trigger the change of behavior. Thus at low veloc-
ities, these animals feed on particles present on the sedimentary surface,
while at higher velocities, they switch to suspension feeding.
Spionids
The tube-building spionid polychaetes are characteristically able to
switch from deposit to suspension feeding. The switch seems to encom-
pass a range of velocities from <2 to 5 cm s"1 when a worm population is
sampled (Taghon, Nowell, and Jumars 1980; Taghon and Greene 1992).
For Pseudopolydora kempi japonica and Boccardia pugettensis, increas-
ing velocity causes an asymptotic rise in the number of worms which are
suspension feeding. Thus 50% of worms are suspension feeding at -1.7
for the former species and 4.5 cm-s"1 for the latter species (measured
0.3 cm above the bottom). Partly because a wide enough range of veloc-
ities was not used in the flume experiments of Taghon and Greene
(1992), their results do not provide sufficient detail to determine whether
a unimodal feeding response to velocity exists.
Miller, Bock, and Turner (1992) have extended observations to Spio
setosa in a flume oscillating flow. The sinusoidal flows produced in the
working section were characterized by the maximum free-stream veloc-
ity reached in each excursion. The percentage of worms suspension
feeding (that is palp coiling) in still water is zero, rising to -90% at
6.5 cm-s"1 and thereafter dropping off (maximum velocity tested of
-lOcm-s"1). The feeding response shown by populations of S. setosa
appears to be unimodal with respect to maximum oscillation velocity, but
the absence of higher velocity tests prevents drawing conclusions. In
another experiment, the percentage of Spio showing palp coiling was
observed at five maximum free stream velocities (6.8-30cm -s"1). In this
experiment, theflowswere increased every lOmin and the palp coiling of

19 individual Spio observed. Again, a unimodal response is suggested
with a lower percentage of worms coiling at 6.8 and the two highest
velocities tested: 24 and 30cm-s~1. To establish firmly the oscillating
velocity threshold from deposit to suspension feeding and the worms'
responses to higher flows, a wider range of velocities should be tested.
When one, or both, tentacles of spionids are lost to sublethal predation,
Lindsay and Woodin (1995) have shown experimentally that such worms
can still deposit feed directly from sediment surfaces.
Bivalves
Some species of the bivalve molluscan family Tellinacea, although
mainly sediment interface deposit feeders, may, under some environ-
mental conditions, subsurface deposit feed or suspension feed. Levinton
(1991) suggests the following feeding classification for tellinaceans:
Swash rider suspension feeders, intertidal species such as Donax which mi-
grate on sandy beaches to maintain moist and relatively unexposed sediments
Sandy bottom suspension feeders, intertidal and nearshore subtidal species
with relatively large ctenidia specialized for suspension feeding
Sandy-mud bottom switching feeders, intertidal and nearshore subtidal
sandy-mud species such as Macoma and Scrobicularia, some species of which
can switch between deposit and suspension feeding.
Of the latter group, Hughes (1969) studied the suspension feeding of
Scrobicularia plana by using the direct method of measuring pumping
rate in apparatus designed by Drinnan (1964). Filtration rates were
estimated independently by indirect measurement of seston depletion in
100-ml dishes containing one Scrobicularia. In these studies, no attempt
was made to investigate the effect of flow and the environmental factors
causing deposit to change to suspension feeding. In the study by
Levinton (1991), the feeding behavior of three Pacific species of Macoma
from the San Juan Archipelago - M. nasuta, M. secta, and M. unquinata
- was studied with respect to water flow and sediment transport. These
studies were done in an annular flume and showed that deposit feeding
was inhibited by increasing velocities (only three free-stream velocities,
-10, -25, and 33-36cm-s"1 were tested). This involved a decrease in
deposit feeding radius from the siphon hole because the inhalant was
withdrawn as velocity was increased. Levinton (1991) observed a loss of
control of the inhalant siphon (quivering/flopping over) at a free-stream
velocity between 10.5 and 12.0 cm -s"1, suggesting that drag on the siphon
is sufficient to overcome its normal functioning.
Summary
The following criteria can be used for assessing the completeness, or lack
thereof, of our knowledge of the physiological mechanisms of filtration
in benthic suspension feeders:
That velocity has been accurately measured near the appropriate filter or
inhalant opening of the trophicfluidtransport system over a sufficiently wide
range of ambient velocities and directions
That the volumetricflowwhich enters thefilteror inhalant is known
That the retention efficiency of thefilteris known.
Most suspension feeders have not been studied over a sufficiently wide
range of velocities to determine effects on filtration, feeding, or growth
and, consequently, conclusions about the general applicability of the
unimodal function model may be premature. Some hypotheses concern-
ing filtration of benthic macrofauna which require further experimental
testing are summarized in Table 4.5.
Of the few species that have been studied adequately with respect to
velocity, inclusive of passive (the sea pen Ptilosarcus querneyi,
gorgonians of the genus Pseudopterogorgia, the crinoids Oligometra
serripinnia and Antedon mediterrana) and active suspension feeders
(scallops, Argopecten irradians, Placopecten magellanicus, plus a few
species of Bryozoa), all appear to fit the unimodal model. The cause-
effect mechanisms at each response stage of the model are poorly under-
stood and require further experimental investigations. These studies
need access to controlled flows in flumes (see Chap. 2) with well-defined
boundary layer flows which simulate field flows as closely as possible. In
general, in stage a transport limitation of filtration can be expected, in
stage b this effect is absent and ambient velocity has no influence on
filtration, whereas in stage cfiltrationinhibition related to velocity can be
expected. In active suspension feeders, the filtration responses may also
be interactively linked to seston concentration and quality, making them
complex and often difficult to interpret.
As far as we are aware, there have been no studies which investigate
the effect of a wide range of flume-controlled ambient velocities on
filtration or feeding of sieving suspension feeders (Table 4.5, number 1),
and a unimodal response hypothesis probably applies to most groups
considered. The exact nature of the compound passive-active suspension
feeding group is still in doubt because, in order to establish firmly that
passive supply of seawater to the organism is of physiological conse-
quence, it must be demonstrated in live animals that enhanced filtration
associated with enhanced ambient velocity results in increased growth
Table 4.5. Some hypotheses related to filtration of benthic suspension

feeders which need further experimental testing.
Ho The filtration rate of sieving Vogel (1981)

suspension feeders is unaffected by LaBarbera (1984)
velocity
Hx The filtration rate of sieving
suspension feeders is a unimodal
response of velocity
Ho Combined passive-active pumping This analysis
does not enhance filtration/growth
rates
Hi Combined passive-active pumping
enhances filtration/growth in some
sponges and brachiopods
Ho The energetic cost of ciliary pumping Winter (1978)
in bivalves is directly proportional to J0rgensen (1990)
pump power output
Hx The energetic cost of ciliary pumping
in bivalves is small and does not vary
with pump power output
Ho Filtration or growth rate in Vogel and Bretz (1972),
brachiopods and scallops is not LaBarbera (1977)
affected by their orientation to flow Wildish et al. (1987),
Hi Filtration or growth in brachiopods Wildish and Saulnier
and scallops is controlled by their (1993)
orientation with respect to flow
direction
(Table 4.5, number 2). The critical experiments required to establish this
have not yet been carried out. The question of the energetic costs asso-
ciated with ciliary pumping in bivalves (number 3, Table 4.5) is periph-
eral to the arguments presented in this chapter, although we return to
this subject in relation to the evolution of the ciliary pump in Chapter 9
(the section, Ultimate questions).
Another interesting unanswered question concerns the ambient veloc-
ity passive enhancement and inhalant/exhalant pressure field theories
and their possible applicability to both brachiopods and scallops. As
proposed, the hypothesis (Table 4.5, number 4) does not appear to be
incisive in its ability to provide support for one or the other of the two
competing theories. Published data of seston uptake in scallops (Wildish

et al. 1987) do support Hx in hypothesis number 4 (Table 4.5). Utilizing
the principle of multiple hypotheses - if the unimodal response for
scallops outlined earlier is valid - the most likely applicability of the
passive enhancement theory is in the "a part" of the unimodal response,
i.e. from 0 to -Scm-s"1. Thus the goal of an experimental test would be
to determine whether small increases in flow serve only to remove local-
ized seston depletion due to dilution with pre-filtered seawater around
the animal, or whether the flow provides a direct subsidy to the scallop's
active pumping, as is required by the passive enhancement theory.
References
Bayne, B. L., and R. C. Newell. 1983. Physiological energetics of marine
molluscs, pp. 407-515. In K. M. Wilbur (ed.) The mollusca. Vol. 4.
Academic Press, New York.
Bayne, B. L., R. J. Thompson, and J. Widdows. 1976. Physiology I, pp. 121-
206. /n B. L Bayne (ed.) Marine mussels, their ecology and physiology.
Cambridge University Press, London.
Beninger, P. G. 1991. Structures and mechanisms of feeding in scallops:
paradigms and paradoxes, pp. 331-340. In S. E. Shumway and P. A.
Sandifer (ed.) World Aquaculture Society, Baton Rouge, La.
flow on volume flow rate andfilteringefficiency. Biol. Bull. 175: 332-342.
Best, M. A., and J. P. Thorpe. 1983. Effects of particle concentration on
clearance rate and feeding current velocity in the marine bryozoan,
Flustrellidra hispida. Mar. Biol. 77: 85-92.
1986. Effects of food particle concentration on feeding current velocity in six
species of marine Bryozoa. Mar. Biol. 93: 255-262.
Birbeck, T. H., J. G. McHenery, and A. S. Nottage. 1987. Inhibition of
filtration in bivalves by marine vibrios. Aquaculture 67: 247-248.
Birkbeck, T. H., and J. G. McHenery. 1982. Degradation of bacteria by
Mytilus edulis. Mar. Biol. 72: 7-15.
Bricelj, V. M., A. E. Bass, and G. R. Lopez. 1984. Absorption and gut passage
time of microalgae in a suspension feeder: an evaluation of the 51Cr:14C
twin tracer technique. Mar. Ecol. Prog. Ser. 17: 57-63.
Cahalan, J. A., S. E. Siddall, and M. W. Luckenbach. 1989. Effects of flow
velocity, food concentration and particle flux on growth rates of juvenile
bay scallops Argopecten irradians. J. Exp. Mar. Biol. Ecol. 129: 45-60.
Carlisle, D. B. 1979. Feeding mechanisms in tunicates. Sci. Ser. 103, Canada
Inland Waters Directorate.
Crisp, D. J., and A. J. Southward. 1961. Different types of cirral activity of
barnacles. Phil. Trans. R. Soc. (Ser. B) 243: 271-308.
Dai, C.-F., and M.-C. Lin. 1993. The effects of three gorgonians from southern
Taiwan. J. Exp. Mar. Biol. Ecol. 173: 57-69.
Drinnan, R. E. 1964. An apparatus for recording the water-pumping
behaviour of lamellebranchs. Neth. J. Sea Res. 2: 223-232.
Eckman, J. E., C. H. Peterson, and J. A. Cahalan. 1989. Effects of flow speed,
turbulence and orientation on growth of juvenile bay scallops, Argopecten

irradians concentricus (Say). J. Exp. Mar. Biol. Ecol. 132: 123-140.
Fenchel, T. 1980. Suspension feeding in ciliated Protozoa: feeding rates and
their ecological significance. Microb. Ecol. 6: 13-25.
Fiala-Medioni, A. 1978. Filter feeding ethology of benthic invertebrates
(Ascidians). V. Influence of temperature on pumping, filtration and
digestion rates and rhythms in Phallusia mamillata. Mar. Biol. 48: 251-
259.
Flood, P. R., and A. Fiala-Medioni. 1981. Ultrastructure and histochemistry of
the food trapping mucous film in benthic filter feeders (Ascidians). Acta.
Zool. Stockh. 62: 53-65.
Flood, P. R., and A. Fiala-Medioni. 1982. Structure of the mucous filter of
Chaetopterus variopedatus (Polychaeta). Mar. Biol. 72: 27-33.
Foster-Smith, R. L. 1975. The effect of concentration of suspension on the
filtration rates and pseudofaecal production for Mytilus edulis L.,
Cerastoderma edule (L.) and Venerupis pullastra. J. Exp. Mar. Biol. Ecol.
17: 1-22.
1976. Pressures generated by the pumping mechanism of some ciliary filter-
feeders. J. Exp. Mar. Biol. Ecol. 25: 199-206.
Grizzle, R. E., and P. J. Morin. 1989. Effect of tidal currents, seston, and
bottom sediments on growth of Mercenaria mercenaria: results of a field
experiment. Mar. Biol. 102: 85-94.
Hildreth, D. I. 1976. The influence of water flow rate on pumping rate in
Mytilus edulis using a refined direct measurement apparatus. J. Mar. Biol.
Assoc. U.K. 56: 311-319.
Hughes, R. N. 1969. A study of feeding in Scrobicularia plana. J. Mar. Biol.
Assoc. U.K. 49: 805-823.
1978. The biology of Dendropoma corallinaceum and Serpulorbis natalensis,
two South African vermetid gastropods. Zool. J. Linn. Soc. 64: 111-127.
Hunt, M. J., and C. G. Alexander. 1991. Feeding mechanisms in the barnacle
Tetraclita squamosa (Brugiere). J. Exp. Mar. Biol. Ecol. 154: 1-28.
J0rgensen, C B. 1966. Biology of suspension feeding. Pergamon Press,
Oxford.
J0rgensen, C. B., P. Famme, H. S. Kristensen, P. S. Larsen, F. M0hlenberg,
and H. U. Riisgard. 1986a. The bivalve pump. Mar. Ecol. Prog. Ser. 34:
69-77.
J0rgensen, C. B., T. Ki0rboe, F. M0hlenberg, and H. U. Riisgard. 1984. Ciliary
and mucus-net filter feeding, with special reference to fluid mechanical
characteristics. Mar. Ecol. Prog. Ser. 15: 283-292.
J0rgensen, C. B., P. S. Larsen, and H. U. Riisgard. 1990. Effects of
temperature on the mussel pump. Mar. Ecol. Prog. Ser. 64: 89-97.
J0rgensen, C. B., F. M0hlenberg, and O. Sten-Knudsen. 1986b. Nature of
relation between ventilation and oxygen consumption in filter feeders.
J0rgensen, C. B., and K. Ockelmann. 1991. Beat frequency of lateral cilia in
intact filter feeding bivalves: effect of temperature. Ophelia 33: 67-70.
J0rgensen, C B., and H. U. Riisgard. 1988. Gill pump characteristics of the
soft clam Mya arenaria. Mar. Biol. 99: 107-109.
Kerswill, C. J. 1949. Effects of water circulation on the growth of quahaugs
and oysters. J. Fish. Res. Board Can. 7: 545-551.
Kirby-Smith, W. W. 1972. Growth of the bay scallop: the influence of

experimental water currents. J. Exp. Mar. Biol. Ecol. 8: 7-18.
laboratory behaviour and field orientations. Paleobiology 3: 270-287.
1981. Water flow patterns in and around three species of articulate
brachiopods. J. Exp. Mar. Biol. Ecol. 55: 185-206.
1984. Feeding currents and particle capture mechanisms in suspension
feeding animals. Am. Zool. 24: 71-84.
current velocity on a passive suspension-feeder. J. Exp. Mar. Biol. Ecol.
127: 81-103.
Leonard, A. B., J. R. Strickler, and N. B. Holland. 1988. Effects of current
speed on filtration during suspension feeding in Oligometra serripinna
(Echinodermata: Crinoidea). Mar. Biol. 97: 111-125.
coral, Leptogorgia. Biol. Bull. 151: 344-356.
Levinton, J. S. 1991. Variable feeding behaviour in three species of Macoma
(Bivalvia: Tellinacea) as a response to water flow and sediment transport.
Mar. Biol. 110: 375-383.
Lindsay, S. M., and S. A. Woodin. 1995. Tissue loss induces switching of
feeding mode in spionid polychaetes. Mar. Ecol. Prog. Ser. 125: 159-
169.
Manahan, D. T., S. H. Wright, G. C. Stephens, and M. A. Rice. 1982.
Transport of dissolved amino acids by the mussel Mytilus edulis:
demonstration of net uptake from natural seawater. Science 215: 1253-
1254.
McHenery, J. G., and T. H. Birkbeck. 1986. Inhibition of filtration in Mytilus
edulis L. by marine vibrios. J. Fish. Dis. 9: 256-261.
Mohlenberg, F., and H. U. Riisgard. 1978. Efficiency of particle retention in 13
species of suspension feeding bivalves. Ophelia 17: 239-246.
Okamura, B. 1984. The effects of ambient flow velocity, colony size, and
upstream colonies on the feeding success of Bryozoa. I. Bugula stolonifera
Ryland, an arborescent species. J. Exp. Mar. Biol. Ecol. 83: 179-193.
1985. The effects of ambient flow velocity, colony size, and upstream
colonies on the feeding success of Bryozoa. II. Conopeum reticulum
(Linnaeus), an encrusting species. J. Exp. Mar. Biol. Ecol. 89: 69-80.
1987. Particle size and flow velocity induce an inferred switch in bryozoan
suspension-feeding behaviour. Biol. Bull. 173: 222-229.
1990. Particle size, flow velocity, and suspension-feeding by the erect
bryozoans Bugula neritina and B. stolonifera. Mar. Biol. 105: 33-39.
Patterson, M. R. 1991. The effects of flow on polyp-level prey capture in an
octocoral, Alcyonium siderium. Biol. Bull. 180: 93-102.
Randl0v, A., and H. U. Riisgard. 1979. Efficiency of particle retention and
filtration rate in four species of Ascidians. Mar. Ecol. Prog. Ser. 1: 55-59.
Riisgard, H. U. 1988. The ascidian pump: properties and energy cost. Mar.
Ecol. Prog. Ser. 47: 129-134.
1989. Properties and energy cost of the muscular piston pump in the
suspension feeding polychaete Chaetopterus variopedatus. Mar. Ecol.
Prog. Ser. 56: 157-168.
1991. Suspension feeding in the polychaete Nereis diversicolor. Mar. Ecol.

Prog. Ser. 70: 29-37.
Riisgard, H. U., and A. Randl0v. 1981. Energy budgets, growth and filtration
rates in Mytilus edulis at different algal concentrations. Mar. Biol. 61:
227-234.
Riisgard, H. U., S. Thomassen, H. Jakobsen, J. M. Wicks, and P. S. Larsen.
1993. Suspension feeding in marine sponges Halichondria panicea and
Haliclona urceolus: effects of temperature on filtration rate and energy
cost of pumping. Mar. Ecol. Prog. Ser. 96: 177-188.
Robbins, I. J. 1984. The regulation of ingestion rate, at high suspended
particulate concentrations, by some phleobranchiate ascidians. J. Exp.
Mar. Biol. Ecol. 82: 1-10.
Shumway, S. E., T. L. Cucci, L. Gainey, and C. M. Yentsch. 1985a. A
preliminary study of the behavioural and physiological effects of
Gonyaulax tamarensis on bivalve molluscs, p. 389-394. In D. M.
Anderson, A. W. White, and D. G. Baden (eds.) Toxic dinoflagellates.
Elsevier, Amsterdam.
Shumway, S. E., T. L. Cucci, R. C. Newell, and C. M. Yentsch. 1985b. Particle
selection, ingestion, and absorption in filter-feeding bivalves. J. Exp. Mar.
Biol. Ecol. 91: 77-92.
Shumway, S. E., and L. F. Gainey. 1992. A review of physiological effects of
toxic dinoflagellates on bivalve molluscs. Proc. Ninth Int. Malac.
Congress: 357-362.
Sponaugle, S. 1991. Flow patterns and velocities around a suspension-feeding
gorgonian polyp: evidence from physical models. J. Exp. Mar. Biol. Ecol.
148: 135-145.
Sponaugle, S., and M. LaBarbera. 1991. Drag-induced deformation: a
functional feeding strategy in two species of gorgonians. J. Exp. Mar. Biol.
Ecol. 148: 121-134.
Stenton-Dozey, J. M. E., and A. C. Brown. 1994. Short-term changes in the
energy balances of Venerupis corrugatus (Bivalvia) in relation to tidal
availability of natural suspended particles. Mar. Ecol. Prog. Ser. 103: 57-
64.
Taghon, G. L., and R. R. Greene. 1992. Utilization of deposited and
suspended particulate matter by benthic "interface" feeders. Limnol.
Oceanogr. 37: 1370-1391.
Taghon, G. L., A. R. M. Nowell, and P. A. Jumars. 1980. Induction of
suspension feeding in spionid polychaetes by high particulate fluxes.
Science 210: 562-564.
Timko, P. L. 1976. Sand dollars as suspension feeders: a new description of
feeding in Dendraster excentricus. Biol. Bull. 151: 247-259.
Trager, G. C, D. Coughlin, A. Genin, Y. Achituv, and A. Gangopadyay. 1992.
Foraging to a rhythm of ocean waves: porcelain crabs and barnacles
synchronize feeding motions with flow oscillations. J. Exp. Mar. Biol.
Ecol. 164: 73-86.
Trager, G. C, J.-S. Hwang, and J. R. Strickler. 1990. Barnacle suspension-
Videl, A., B. B. Andersen, and H. U. Riisgard. 1994. Field investigations of
pumping activity of facultatively filter feeding polychaete Nereis
diversicolor using an improved infrared phototransducer system. Mar.
Ecol. Prog. Ser. 103: 91-101.
Vogel, S. 1974. Current-induced flow through the sponge, Halichondria. Biol.
Bull. 147: 443-456.
1977. Current-induced flow through sponges in situ. Proc. Nat. Acad. Sci.,
U.S.A. 74: 2069-2071.
1978. Organisms that capture currents. Sci. Am. 239: 128-139.
1981. Life in moving fluids: the physical biology of flow. Willard Grant
Press, Boston.
Vogel, S., and W. L. Bretz. 1972. Interfacial organisms: passive ventilation in
the velocity gradients near surfaces. Science 175: 210-211.
Walne, P. R. 1972. The influence of current speed, body size and water
temperature on the filtration rate of five species of bivalves. J. Mar. Biol.
Assoc. U.K. 52: 345-374.
Ward, J. E., H. K. Cassel, and B. A. MacDonald. 1992. Chemoreception in the
sea scallop Placopecten magellanicus (Gmelin). I. Stimulatory effects of
phytoplankton metabolites on clearance and ingestion rates. J. Exp. Mar.
Biol. Ecol. 163: 235-250.
Ward, J. E., and N. M. Targett. 1989. Influence of marine microalgal
metabolites on the feeding behaviour of the blue mussel Mytilus edulis.
Mar. Biol. 101: 313-321.
Warner, G. F. 1977. On the shapes of passive suspension feeders, p. 567-576.
In B. F. Keegan, P. 6. Ceidigh, and P. J. S. Boaden (eds.) Biology of
benthic organisms. Pergamon Press, Oxford.
Warner, G. F., and J. D. Woodley. 1975. Suspension-feeding in the brittle star,
Ophiothrixfragilis. J. Mar. Biol. Assoc. U.K. 55: 199-210.
Wells, G. P., and R. P. Dales. 1951. Spontaneous activity patterns in animal
behaviour: the irrigation of the burrow in the polychaetes Chaetopterus
variopedatus Renier and Nereis diversicolor. J. Mar. Biol. Assoc. U.K. 29:
661-680.
Widdows, J., P. Fieth, and C. M. Worrall. 1979. Relationships between seston,
available food and feeding activity in the common mussel Mytilus edulis.
Mar. Biol. 50: 195-207.
Wildish, D. J., and D. D. Kristmanson. 1985. Control of suspension-feeding
bivalve production by current speed. Helgol. Wiss. Meeresunters. 39: 237-
243.
Wildish, D. J., D. D. Kristmanson, R. L. Hoar, A. M. DeCoste, S. D.
McCormick, and A. W. White. 1987. Giant scallop feeding and growth
responses to flow. J. Exp. Mar. Biol. Ecol. 113: 207-220.
Wildish, D. J., D. D. Kristmanson, and A. M. Saulnier. 1992. Interactive effect
of velocity and seston concentration on giant scallop feeding inhibition. J.
Exp. Mar. Biol. Ecol. 155: 161-168.
Ecol. 142: 213-220.
Wildish, D. J., and A. M. Saulnier. 1992. The effect of velocity and flow
direction on the growth rate of juvenile and adult giant scallops. J. Exp.
Mar. Biol. Ecol. 155: 133-143.
1993. Hydrodynamic control of filtration in the giant scallop. J. Exp. Mar.
Biol. Ecol. 174: 65-82.
Winter, J. E. 1978. A review on the knowledge of suspension-feeding in
lamellibranchiate bivalves with special reference to artificial aquaculture
systems. Aquaculture 13: 1-33.
Young, C. M., and L. F. Braithwaite. 1980. Orientation and current-induced
flow in the stalked ascidian, Styela montereyensis. Biol. Bull. 159: 428-440.
5
Mechanisms of seston capture
It is the purpose of this chapter to address the second stage of suspension

feeding, as defined in Chapter 4: specific particle capture mechanisms at
the appropriate suspension-feeding collecting surface. The theory for
our approach to this is borrowed from physics and engineering literature,
inclusive of sieving and the aerosol theory mechanisms. We are espe-
cially indebted to reviews by Rubenstein and Koehl (1977), LaBarbera
(1984), and Shimeta and Jumars (1991), which help interpret this in a
biological context. We also include a novel mechanism involving
hydromechanical shear forces at the bivalve gill proposed by J0rgensen
(1981a, 1983, 1990), which is not part of the classical particle capture
theory. Other topics discussed here include the nature of seston, how
sestonic particles reach the collector, and a brief consideration of those
species where calculations of seston encounter efficiencies according to
aerosol theory have been made.
Nature of seston
The type of particles which may be collected by suspension feeders are
highly variable and dependent on the local conditions. Non-viable parti-
cles of silt, clay, sand, and detritus may be processed for the attached
microbiota. Viable particles include bacteria, phytoplankton, inverte-
brate larvae, and eggs, many of which are weakly motile. Collectively,
this material is referred to as seston (Fig. 5.1); it ranges in size from less
than 1 \xm up to 1000 |iim or more.
In order to understand the mechanics of collection of sestonic particles
properly, their size, shape, density, and settling velocities should be
known. Shapes range from spherical to very complex. Inorganic particles
are of densities often well known from their mineral content, and settling
170
Mechanisms of Seston Capture 111
BIVALVES SESTON
1 m -T- 2 log cm
Inertial
forces
dom nate
ADULTS
1 cm 4 - 0
1 mm - 1 Macroplankton
JUVENILES
LARVAE
.EGGS. Microplankton
Viscous
forces -3 Nanoplankton_.
dominate
, Bacteria
1 pm - 4 Ultrananoplankton
---5
Virus
J6
Figure 5.1 Comparative sizes of suspension-feeding bivalves and their potential

food (= seston) (Wildish and Kristmanson 1993).
velocities can often be estimated from Stokess law. Typically, particles of

biological origin have density close to that of seawater. It is particularly
difficult to generalize about their settling velocities. Even for a particular
species, density will vary between individuals and may be modified by
changes in lipid content or ionic concentrations of the cell.
A compilation of literature on the sinking rates for phytoplankton
(Parsons, Takahashi, and Hargrave 1977) includes living phytoplankton
(25 species), 0-30m-d"1, and phytoplankton, dead, intact (10 species):
<l-510m-d~1. This range gives an indication of the variability of this
parameter.
Swimming speeds of motile species present in the plankton are also
variable and resistant to generalisations. Some small flagellates swim at
a speed of the order of a body length per second, large dinoflagellates at
25-50 times body length per second (Parsons et al. 1977). In most pub-
lished accounts of the mechanics of particle collection discussed here,
it has been assumed that the particles are of low motility and move
with the fluid streamlines without significant settling. This assumption
Table 5.1. Typical seston particles and encounter rate estimates.
Particle g-cnT3 cm-s 1

Medium clay 1.0 1.630 8.9 x 10"5
Medium silt 20 1.630 3.5 x 10-2
Fine sand 150 1.630 2.0
Organic-mineral aggregate 100 0.088 5.0 x 10-2
Bacterium 1.0 0.050 2.7 x 10"46
Small phytoplankton 10 0.069 3.4 x 10-
Large phytoplankton 100 0.009 5.0 x 10-3
Invertebrate larvae 500 0.0073 0.1
Marine snow 5000 6.5 x 10-5 8.8 x 10-2
Source: Shimeta and Jumars (1991).
seems to be generally reasonable but should not be made without due

consideration.
Shimeta and Jumars (1991) have summarized the properties of typical
seston particles used in their analysis of examples of particle capture
which will be discussed later (Table 5.1).
Fluxes, encounter rates, and collection efficiencies

If we consider an object fixed in a steadyflowof water containing seston,
then the undisturbed flow upstream of the object can be characterized by
a mean velocity, U (convenient units of centimetres per second [cm s"1]).
For the purposes of this presentation, the effects of shear and turbulence,
though typical of the benthic boundary layer, will not be considered. Our
purpose is to define the relationships among flux, encounter rates, and
collection efficiency.
The concentration of seston in the approaching stream, c (the number
of particles per unit volume), is assumed to be constant in time and
space. Then, thefluxof seston at a point upstream of the organism can be
defined as the product of velocity and seston concentration, cU. The
frontal area of the object, A, is the projection of its outline on a plane
normal to the flow vector. The engineering definition of the collection
efficiency (e) of a suspension feeder would be the ratio of the number of
particles striking it to the number which would strike it if the stream were
not diverted (Dorman 1966). This definition is based on the assumption
that all particles which strike the collector are retained. In engineering
applications, an adhesion coefficient (zA) is defined to allow for any
resuspension of particles so that the overall collection efficiency is the
product of the adhesion coefficient and the engineering collection
efficiency.
The rate at which particles are collected is, therefore, given by
ECUAEA
Shimeta and Jumars (1991) argue that the engineering efficiency of

biological filters is a less useful term than the encounter rate, defined as
the rate at which particles reach the surface of the collector:
F=zcUA (5.1)
The advantage of using encounter rates is that this quantity is a direct
measure of the potential food available to the organism and takes into
account the seston concentration. The rate at which food is collected is
obtained by multiplying the encounter rate by the retention coefficient
(Shimeta and Jumars 1991), equivalent to the adhesion coefficient used
in engineering.
The clearance rate, R, as used in Chapter 2, is generally defined as the
volume of fluid cleared of particles per unit time. This term is equivalent
to the encounter rate times the retention coefficient divided by the
upstream concentration, or
Following Shimeta (1993) we will also use a clearance rate for a circular
cylinder defined in units of volume cleared per unit time per unit length
of cylinder.
The passive suspension feeder is often configured as a large number of
individual collecting elements in a supporting structure. A typical exam-
ple is the sea pen (see the section, Passive suspension feeders, Chap. 4).
Particle collection theory is mainly based on the analysis of capture by
isolated elements of simple geometry or by filter beds composed of layers
of cylinders or spheres. The preceding definitions were written with a
single solid obstacle such as a cylinder in mind, but with suitable reserva-
tions they can be applied to the whole structure, in which case the
collection efficiency represents the sum of the contributions of each of
the collecting elements in the entire organism. The analysis can also be
applied to a single collecting element. Each collecting element is charac-
terised by its own approach velocity and upstream concentration, and an
estimate of its collection efficiency may be made from theory, as will be

detailed later. It is important to note, however, that the collection effi-
ciency of the organism as defined is in general different from that of the
individual collecting elements and is not necessarily easily estimated
from the former.
Mechanics of capture
Suspension feeders must process large volumes of water to extract suffi-
cient seston as food. The food is in the form of particulate matter, usually
in low concentrations, often less than a few milligrams per liter (mg L"1),
and must be removed from the water being processed. The means by
which this is done by benthic suspension feeders are many and varied.
We have discussed the classification of the major methods in Chapter 4.
Classification may also be based on the mechanisms by which the parti-
cles are removed (LaBarbera 1984):
Biological, involving active response of the animal
Physical
The latter can be further classified according to the dominant mecha-

nisms involved. Another mechanism of particle segregation which does
not depend on a biological response or on physical contact with a biologi-
cal structure has been proposed by J0rgensen (1983). It has been argued
that in the bivalve gill, particles migrate across streamlines in pulsatile
creeping flows close to the frontal surface of the gill. It has been sug-
gested by Nielsen et al. (1993) that this is a third category, fluid/mechani-
cal, which should be added to LaBarbera's scheme.
Two conflicting demands must be satisfied to handle thisfluidprocess-
ing problem: pressure losses must be small and, at the same time, a large
proportion of the particles present in the stream to be processed must be
captured. Particulate matter should be removed from the water as effi-
ciently as possible in order to minimize the amount of water which has to
be handled. The most important component of seston, in terms of its
contribution to the ecology of coastal waters, is the primary producers,
the phytoplankton. Seston particles, inclusive of phytoplankton, are gen-
erally small, and the need for high overall particle removal efficiency
would suggest that the filtration structures would be of small dimensions
and, as a consequence, of high flow resistance. In the case of many
passive filter feeders, for instance, an array of small collecting elements
is presented to the flow. The array must be close enough together to be
an efficient collector without unduly restricting the flow through it. A

balance must be struck between the two competing effects.
The same optimization problem arises in the design of filters by engi-
neers for purposes such as air cleaning. For example, filters used for
cleaning air in domestic heating systems are made of beds of loosely
packed fibres. The designer must decide on the diameter of the fibres
used and the density at which they are packed in order to achieve a
desired level of collection efficiency. The criterion to be satisfied is that
the design gives a minimal total cost for the performance specified. The
costs expressed on an annual basis arise from an amortized charge to pay
for the construction of the filter and the cost of energy required to force
the air across the filter. Broadly speaking the designer looks for a balance
between a large and loosely packed filter and a small, more tightly
packed version. Fibre diameter must also be specified. Small-diameter
fibres will give good collection efficiency but will require more energy for
gas transport, so again a compromise is necessary in the setting of this
parameter of the design.
In the same way, the structures which suspension feeders use to collect
seston can only be effective if a balance is reached between the conflict-
ing demands of high removal efficiency and the costs of building and
maintaining collecting structures. The cost associated with pumping
seawater to the filtration surfaces will play a part in the metabolism of an
active suspension feeder, but its relative importance is disputed (see
Chap. 9, the section, Ultimate questions).
The theoretical analyses which guide the industrial designer of filtering
devices were largely developed to solve problems of gas/solid sepa-
rations including gas cleaning, industrial hygiene, and medicine. In
the biological literature, these developments are generally called the
aerosol theory, we will continue to use this term. Many of these findings
can help us begin to understand the mechanisms which might be impor-
tant in suspension feeding, as was first pointed out by Rubenstein and
Koehl (1977). Although the medium of interest for us is water, not
air, much of the aerosol theory can be applied with little or no
modification.
The student of suspension feeding, however, faces a more difficult
problem than the physical scientist or engineer, who generally deals with
steady flows, rigid structures, well-defined geometry, relatively simple
particulate properties, and no feedback from the filtering system to
changes in external variables such as velocity and concentration. Biologi-
cal suspension feeding systems are more complicated in structure and
often surprising in their subtlety of operation. Filter elements may be

much more complex in their geometry than the simple shapes, e.g. circu-
lar cylinders and spheres, commonly addressed by aerosol theory. Bio-
logical filtering elements may assume different orientations with changes
in velocity (Harvell and LaBarbera 1985) and may systematically sweep
the fluid around them (Trager, Hwang, and Strickler 1990). Secondary
flows may be exploited to divert food particles to collecting surfaces as,
for instance, in the bivalve filtering system (J0rgensen 1981a). In addi-
tion, as pointed out by Leonard (1989), seston capture rates not only are
determined by fluid mechanics, but also may depend on "complex
behavioral modulation of the capture response." Notwithstanding these
difficulties in the extension of aerosol theory to biological problems,
insights into the limitations to feeding rates which may control the
growth of filter feeders may sometimes be obtained by the use of these
theoretical principles.
Sieving
Of all the physical mechanisms, sieving is the most obvious and simple to
understand. Water containing seston is passed through a porous obstacle
and particles larger than the passages forfloware retained. Although this
is a simple concept, the reality is more complicated. Most suspension
feeders utilize seston of a wide range of sizes and while large particles
may be sieved, smaller particles will also be separated simultaneously at
lesser collection efficiencies by other mechanisms. There are many ex-
amples, however, of suspension feeders in which sieving is the dominant
mechanism. The use of mucous nets by ascidians and polychaetes has
been discussed in the previous chapter (see the section, Sieving suspen-
sion feeders). These nets are used within an animal, and the pressure
differences required to pump water through the mesh have been calcu-
lated and in some cases reconciled with the mesh sizes inferred from
particle retention efficiency measurements.
In some freshwater aquatic insect larvae, seston is collected by pre-
senting a filter net to a passing flow. Loudon (1990) and Loudon and
Alstad (1990) have discussed the mechanics of particle collection by
freshwater caddisfly larvae. The filter is external to the animal, and the
seston particles after capture are collected by the forelimbs for passing to
the mouth. These filtering nets are typically arrays of cylindrical obsta-
cles and separate particles both larger and smaller than the openings
between the elements of the filter. Silvester (1983) and Cheer and Koehl
(1987) have analyzed the fluid mechanics of flows through biological

filters and have described some applications in suspension feeding.
A pelagic tunicate, Oikopleura vanhoefferi, uses an unusual form of
sieving to separate particles (Morris and Diebel 1993). Instead of collect-
ing particles, a mesh system is used to remove water from the inlet
stream to concentrate the particles before acceptance by the food tube.
The stream is, so to speak, de-watered. This process is analogous to that
of a man-made tangential, or cross-flow, filter. Concentration factors of
100-1000 were measured in these experiments.
The aerosol theory and benthic suspension feeders

Many suspension feeders do not remove seston from water by a simple
sieving mechanism. When the structures which effect the separation are
examined, it is seen that they are not formed as nets, and the gaps
between collecting elements are often larger than particles which are
observed to be collected efficiently. There are obviously other mecha-
nisms at work.
In gas-cleaning technology, much the same situation is met. When
particles are removed by filters which consist of beds of fibres, the pas-
sages in the filter are generally much larger than the smallest particles
collected. Particles are collected within the bed, rather than preferen-
tially at the surface, as would be the case if the passages were smaller
than the particles. The fibres in the bed must collect particles by mecha-
nisms other than sieving. These involve aerosol mechanisms occurring
within the fibre bed filter, as pointed out by Patterson (1991). The basic
principles of aerosol filtration have been reviewed by many authors,
including Pich (1966), Licht (1980), and Fuchs (1986).
Aerosol mechanisms should be applicable to the similar problem of
suspension feeding by attached benthic animals. The particles of interest
are of similar size, although their densities are much closer to that of the
fluid, and the collecting elements of suspension feeders are often of
cylindrical shape and in the same size range as aerosol filters. The veloci-
ties are generally lower in biological systems, but water has a kinematic
viscosity about V15 that of air so that the Reynolds numbers of both types
of applications are often in the same range. In fact, the creeping flow
regime, which is usually encountered at the scale of the collecting ele-
ments of biological systems, is perhaps the best understood Reynolds
number range of aerosol theory. The application of these ideas to sus-
pension feeding in benthic animals was developed systematically in the
important paper of Rubenstein and Koehl (1977). They showed that by

considering in detail the mode of particle capture by an isolated fibre, it
was possible to determine the dominant mechanisms and to "provide
insights for those investigating the efficiency of various modes of filter
feeding and the mechanisms of size-selective filter feeding."
Consider the problem of a fluid stream containing suspended particles
which flows towards an isolated circular cylinder oriented with its axis
normal to the flow vector. Some of the particles in the approaching
stream will strike the obstacle, and if the adhesion or retention coefficient
is unity, these particles will be collected. The engineering efficiency of
collection must be predicted, and this is done according to equation (5.1).
In order to analyze seston encounter efficiencies, some simplifying
assumptions are necessary. The particles are assumed to be much smaller
than the obstacle and to be evenly distributed in the approaching stream.
The surface of the obstacle is assumed to be smooth and the flow to be
steady, uniform, and turbulence-free. The kinematics of the flow around
the obstacle, i.e. the pattern of streamlines, is assumed to be known. The
adhesion coefficient is assumed to be unity. Emphasis will be placed on
the low Reynolds number case because of its practical importance in
suspension feeding. Additional assumptions may be added in specific
cases as the arguments are developed.
Each of the five mechanisms of capture of particles from aerosols
(direct interception, gravitational deposition, inertial impaction,
Brownian diffusional deposition, electrostatic attraction) will be exam-
ined in turn, and the functional dependencies of the collection mecha-
nisms will be described. Equations for the engineering efficiency and the
encounter rate will be developed. As well, indices will be defined which
will provide a simple means of judging whether or not the conditions
under investigation are favorable to the mechanism. The treatment fol-
lows that of Pich (1966), Rubenstein and Koehl (1977), Spielman (1977),
and Shimeta and Jumars (1991).
Direct interception
Consider a particle of neutral buoyancy in a steady flow around a right
circular cylinder. This particle will move with the fluid if there are no
other forces acting on it. Its path will coincide with the fluid streamline it
was on upstream of the cylinder. It will collide with the cylinder if this
stream line comes within one particle radius of the surface of the cylin-
der. If the retention efficiency is unity, the particle will be captured. This
mode of capture is called direct interception.
Direct interception by a right circular cylinder is shown schematically

in Fig. 5.2A, with the flow assumed to be steady and two-dimensional.
Well upstream of the cylinder, the flow is uniform, meaning that the
velocity does not vary either in the flow direction or in the cross-flow
direction. The motion of the fluid is visualized by the use of "stream-
lines," which represent the trajectories of infinitesimal fluid elements
marked at the upstream edge of the diagram. Molecular diffusion is
ignored. The streamlines must be everywhere tangent to the velocity
vector, so that fluid cannot cross streamlines. As a consequence of the
law of conservation of mass, the product of the average velocity between
adjacent streamlines and their spacing is everywhere constant and
equivalent to the volumetric flow per unit depth into the paper between
streamlines. By convention, the streamlines are evenly spaced in the
uniform flow upstream of the obstacle: i.e. the "stream tubes" defined by
adjacent streamlines all carry equal volumetric flows. Streamlines which
come closer indicate an accelerating flow, whereas any reduction in
velocity will result in the streamlines' diverging.
The capture efficiency and the encounter rate can be deduced if the
streamlines are defined and it is assumed that the particles follow them
without modifying the flow. All particles (diameter dp) carried in the
stream tube defined by the streamlines which pass above and below the
cylinder (diameter dc) at a distance of one particle radius will be removed
(see Fig. 5.2A). This defining streamline (the "limiting streamline") has
a position far upstream with respect to the axis of flow which is defined
by the cross-flow distance (8). Since the width of the stream tube which
would have flowed through the space occupied by the cylinder in its
absence is dc, the cylinder diameter, the engineering collection efficiency
for direct interception alone, ER, must be
The encounter rate for a cylinder of length lc is given by
FR = cUlcd(j-)=cUlc8 (5.3)
and a convenient index to indicate the relative efficiency can be defined

as
NR = ER (5.4)
Figure 5.2 Schematics of capture of sestonic particles of diameter dp by a

cylinder of diameter dcin a flow of Re c <l: A, direct interception; B, inertial
impaction.
The distance, 8, will be dependent on the flow field around the cylinder.
If the spacing of the streamlines is assumed to be unchanged as the flow
passes around the cylinder, 8 can be replaced by the particle radius,
giving
This approach is used by Rubenstein and Koehl (1977) and Shimeta and
Jumars (1991) as a first approximation to the solution of the problem. In
creeping flow, however, the fluid velocity near the cylinder is less than
the upstream velocity. By the principle of conservation of mass, the
volumetric flow between any two streamlines must be constant every-
where in the flow field. It follows that the spacing between the stream
line to the stagnation point and the limiting stream line must increase in
the downstream direction since the average velocity in this stream tube
decreases.
A comment on the mechanics of steady flow around submerged cylin-
ders is needed at this point. The fluid flowing around the cylinder is
subject to forces which arise from the inertia of the fluid elements and
also as a result of the viscosity of the fluid, which resists any deforma-
tions. If the inertial forces are small with respect to the viscous forces, it
may be possible to ignore them and find a solution for the streamlines
dependent on the viscous properties of the fluid. The criterion for this
condition is that the "Reynolds number" of the flow is small. The
Reynolds number is defined as
where D and U are, respectively, a characteristic size parameter, here

conveniently the diameter of the cylinder or sphere and the upstream
velocity, and x> is the kinematic viscosity of the test fluid used. For
Reynolds numbers well below unity for spheres, the flow can be repre-
sented by a "creeping flow" solution. In the case of cylinders, the as-
sumption that inertial forces are everywhere much less than viscous
forces at low Reynolds numbers raises mathematical difficulties. An
approximate solution accurate near the surface of the cylinder has been
provided by Lamb (1962).
A more accurate formulation of the problem can be made if the Lamb
solution for the velocity field around the cylinder at low Reynolds
number is used. For the flow close to the cylinder, the following equa-
tions apply (Spielman 1977):
E iA 'dA2
--- \t
and
FR = 2AFcUlcd2pd;x (5.7a)
where
AF = 0.5(2 -lnfRe,))"1 (5.7b)
The term AF is a dimensionless coefficient and is here evaluated for the

specific case of an isolated right circular cylinder in a low Reynolds
number flow.
Unfortunately, this solution of the problem is not the whole picture.
As the particle approaches the surface of the collector, the implicit
assumption in this treatment that the presence of the particle does not
alter the flow around the cylinder becomes less justifiable. It can be
expected that the particle's approach to the surface will be restrained by
the outflow of fluid displaced by the particle. In addition, as the particle
approaches the collector it is subject to attractive van der Waals forces,
which tend to counteract the effect of fluid drainage around the particle.
Contact with the surface is thus mediated by intersurface forces. These
complications are described by Spielman (1977) and will be discussed in
a later section.
Gravitational deposition
A particle which has a greater density than the surroundingfluidwill sink
at its terminal settling velocity, ws, and its trajectory will cross the fluid
streamlines, resulting in capture if the surface of the collector is reached.
The encounter rate for a horizontal cylinder is given (Shimeta and
Jumars 1991) by
FG = cwslc(dc+dp) (5.8)
Most particles of interest will settle at a Reynolds number considerably
smaller than 0.2, for which the settling velocity can be calculated from
Stokess law by
where
pp = density of the particle
p = density of the fluid
|i = absolute viscosity of the fluid
g = gravitational constant
given that the particle is spherical and of known density. The encounter
rate for particles small with respect to the obstacle is then approximately
proportional to the square of the particle diameter and is independent of
the ambient velocity. Equation (5.9) cannot be used if the density of the
particle is not precisely known and inaccuracies may arise if the particle
deviates from a spherical shape.
The engineering efficiency can be calculated as a fraction of the up-
stream flux of particles:
and the efficiency index can be taken to be
f
which can be considered as being equivalent to the efficiency for particles
small with respect to the obstacle, or as a measure of the drift of particles
across the fluid streamlines towards the upper surface of the obstacle.
Inertial impaction
When fluid containing particles approaches an obstacle, the streamlines
curve and particles of densities greater than the fluid will tend to move in
straight lines because of their inertia. These particles are more likely to
be captured. This mechanism of capture is called inertial impaction and
is shown schematically in Fig. 5.2B.
The importance of inertial impaction is determined by the stopping
distance of the particle in the fluid, which is defined as the distance the
particle, after release in the fluid at a certain velocity, will move before it
stops. The stopping distance of interest here is that for the approach
velocity of the fluid. It is given by
(5,2)
An index of efficiency for inertial impaction may be formed by the ratio

of the stopping distance to the diameter of the collecting cylinder. This
ratio is called Stokes number.
NI=-s- = Stokes number (5.13)

dc
For low values of Rec, it can be argued (Shimeta and Jumars 1991) that
if the stopping distance is less than the radius of the cylinder, the width
of the stream tube cleaned of particles is increased by twice the stopping
distance and so the encounter rate, in excess of that by direct intercep-
tion, is given by
Ft = 2cUlslc (5.14)
If, on the other hand, the stopping distance is much greater than the
radius of the cylinder, the particles will continue in a straight line as the
streamlines curve around the obstacle and all the particles in a stream
tube of the diameter of the cylinder will be collected, giving
F, = cUdJe (5.15)
At other values of the stopping distance, prediction of the encounter rate
is more difficult. It is determined by Stokes number and also by the
Reynolds number of the flow around the cylinder. At Reynolds numbers
less than about 0.2, the efficiency can be estimated by hydrodynamic
arguments. Davies and Peetz (1955) have presented a plot of efficiency
against Stokes number for an Rec of 0.2 with the particle to cylinder
diameter ratio as a parameter. They have also predicted efficiency at
higher Reynolds numbers.
Inertial impaction is of considerable practical importance over a wide
range of applications in aerosol science at higher Reynolds number
flows. Langmuir and Blodgett (1946) analyzed this problem in a study of
aircraft wing icing and provided predictions of collection efficiency as a
function of Stokes number. Brun et al. (1955) refined and verified these
calculations. Experimental measurements of collection efficiencies of
cylinders, ribbons, and discs which were made by May and Clifford
(1967) at Rec of 165-8500 generally verify the theoretical predictions
for cylinders. These results are important in aerosol mechanics but are
of limited use in the understanding of suspension feeding by benthic
animals, which are generally characterized by smaller values of Nj and
Collection by Brownian diffusion

Particles of interest in suspension feeding are often small enough to be
affected by Brownian diffusion. These particles will move off the stream-
lines and some will strike the collecting surface. A diffusion coefficient
for non-motile particles can be estimated from
(5.16)
3n\idp
where K is Boltzmann's constant, Tis temperature (Kelvin), and \i is the
absolute viscosity of the fluid.
One way of evaluating the relative importance of the diffusion mecha-
nism in particle collection is to determine the Peclet number
Pe = ^ (5.17)
where U is the approach velocity and dc is the diameter of the cylinder,

here taken to be the collector of interest. The Peclet number is an
indicator of the ratio of the advective flux of particles from upstream and
the diffusive flux to the cylinder. Pich (1966), Rubenstein and Kohl
(1977), and others have used the reciprocal of the Peclet number as an
index for the intensity of collection by diffusion: ND = 1/Pe.
The analysis of the mechanics of diffusion of particles to a collecting
surface has been done in an approximate way by assuming that the
particles can be represented by points and that they are present in
sufficient numbers that their concentration can be assumed to approxi-
mate a continuum. Given these assumptions the problem can be
analyzed in a similar manner to that for problems in the transfer of
dissolved solutes or heat. The simplest case is of diffusion to a cylinder or
sphere in still water. The Peclet number for this case is zero. The solu-
tions are well known (Crank 1956) and give the concentration distribu-
tion in space and time. The solution is time-dependent since the particles
must be drawn from increasing distances from the collector as the sur-
rounding layers are depleted. It is of limited interest in the understanding
of benthic suspension feeding.
A more relevant problem is that of diffusion with flow past the collec-
tor, and the mechanism can be analyzed as a combination of diffusion
and advection by the passing fluid (Pich 1966). If it is assumed that the
particles are present at sufficient concentration that they can be treated
as a continuum, their mass balance over an infinitesimal fluid element is

given by
^ + MVC DVC (5.18)

ot
where c is the concentration of particles at this point, D is the diffusion
coefficient for the particles, and u is the velocity vector at the position of
the fluid element (Crank 1956). This equation may be understood as a
combination of Fick's second law of diffusion (the first and last terms)
and a term expressing the advection of material by the local velocity.
For this equation to be solved the velocity field around the cylinder
must be known. Solutions have been developed for low Reynolds
number flows and for potential flows (see Pich 1966; Shimeta and Jumars
1991). The separation efficiency for the case of low Reynolds number is
^ , (5-19)
There is some uncertainty about the value of the coefficient (Pich 1966).
Shimeta (1993) gives the following expression for the encounter rate
for a cylinder in low Reynolds number flow:
FD = 1.17 D2/3clc(Urcf3 (2 -lnRe c )~ V 3 (5.20)
This result predicts that the encounter rate due to Brownian diffusion
increases with the diffusion coefficient, as would be expected, but also
with the velocity, which is less obvious. The efficiency of collection is
inversely related to velocity, but since increased velocity delivers more
particles to the region of the cylinder the net effect is to increase the
encounter rate.
Electrostatic attraction
If there is an attractive electrostatic force between a particle and a
collecting surface, the particle will move across streamlines and is more
likely to be collected. This is an important mechanism in gas cleaning,
which is exploited in electrostatic precipitators and in fabric filtration.
Particles in seawater, however, have a "surface charge that is neutralized
locally by counterions that concentrate on the solution side to form an
electrical double layer" (Spielman 1977). The electrical fields extend to
only about 0.02 Jim from the surface, and it is reasonable to expect that in
marine conditions, electrostatic forces will determine the adhesion of the

particle to the collecting surface but that other collecting mechanisms
will be of more importance in getting the particle in close proximity to
the collecting surface.
Shimeta and Jumars (1991) present equations for the encounter rate
and the efficiency index for electrostatic attraction, which are given here
in order to complete the set of relationships for the aerosol mechanisms.
If Q is the charge per unit length of collector and q is the charge on a
particle of diameter dp, the encounter rate for an infinite circular cylinder
is given by
F = 4cQqlc
E
(3^)
and the efficiency index by
4Qq
*B = ,- .^ (5-22)
These equations were derived for particles and collecting surfaces which
are smooth, in the sense of having irregularities much smaller than
0.02 |im, conditions which are not commonly encountered in benthic
suspension feeding. Although the equations may not be directly applica-
ble in many cases, the electrostatic mechanism may still be important in
the final stages of collection as the particle approaches the surface. How
this mechanism interacts with direct interception will be discussed in the
next section.
In freshwater applications, because of the much lower ionic strengths
involved, electrostatic attraction should not be dismissed without careful
consideration (Shimeta and Jumars, 1991). Gerritsen and Porter (1982)
showed that relative ingestion of three sizes of polystyrene spheres by
Daphnia magna could be modified if the surface charges on the particles
were neutralized or if a non-ionic surfactant were added to reduce
wettability.
Combined mechanisms
Each of the mechanisms of capture has been analyzed on the assumption
that the other mechanisms are of negligible importance. Particles which
are massive enough to separate by gravity, for example, will often also
have enough inertia to separate from curvilinear streamlines and so be
removed by inertial impaction simultaneously. In many, if not most,
cases of practical interest, more than one mechanism may contribute.

For some solutions which have been derived for the aerosol case, see
Pich (1966). In the case of liquids, these solutions may not always be
useful because of the differences between liquids and gaseous systems in
the final stages of capture, viz. the complications which arise from the
drainage of fluid from the gap between the particle and the collecting
surface and the van der Waals electrostatic forces which counter this
resistance.
Particles which are not of neutral buoyancy will be influenced by
gravity and inertial effects. Yoshioka and Kansoka (1972) have analyzed
the combined effects of gravitation, interception, and inertial separation
in gas-solid systems. Theoretical equations for collection efficiencies at
Rec < 1 are presented and their experimental measurements are consist-
ent with the theory. If particles are of neutral buoyancy, gravitational
capture and inertial separation are not factors. Direct interception may
be influenced by electrostatic forces and by Brownian diffusion.
The theoretical analyses of interception in liquids have been extended
to cover some of the cases where two or more processes contribute to the
collection of particles. We will briefly look at approaches to the solutions
of this problem for low Reynolds numberflowaround a cylinder. We can
approach the problem of combined Brownian diffusion and direct inter-
ception it in at least two ways. The fundamental analysis of particle
migration can be carried out with more than one mechanism in mind. For
instance, Langmuir (as cited by Dorman 1966) has modified his equation
for diffusion of a particle to allow for direct interception by assuming
that the particles diffuse, not to the surface of the collector, but to an
imaginary shell at a distance of one particle radius from the collector.
Another approach is to assume that each mechanism is operating inde-
pendently and that, in the case of the total collection efficiency, it is the
linear sum of the efficiencies of each mechanism. Fuchs (1964) and
others have pointed out that in fact considerable error can arise from this
view of the problem when each mechanism makes a significant contribu-
tion. It is, however, useful in identifying the ranges of variables in which
one or the other mechanism is dominant.
The second approach was used by Shimeta (1993) in order to quantify
the dependence of encounter rate on the size of non-motile particles. In
the circumstances of interest to us, particle size may range widely. Poten-
tial prey may be as small as colloids and viruses, that is under 0.1 |um, with
successively larger bodies as indicated in Fig. 5.1. If it is assumed that the
encounter rates for direct interception and Brownian diffusion are addi-
0.01 0.1 1 10
Dp, micrometres
Figure 5.3 Combined collection by direct interception and Brownian move-
ment. Clearance rates for a right circular cylinder in milliliters cleared per hour
per centimeter of length versus particle diameter. Velocity of approaching
stream: lower three curves, 0.01 cm-s"1; upper three curves, 1.0 cm-s"1, solid,
dotted, and dashed lines, cylinder diameters of 10, 50, and 100 (im; dp<0.1 dc\
Re c <1.0.
tive, equations (5.6) and (5.20) can be added to give the combined
encounter rate. Following the procedure of Shimeta (1993), this calcula-
tion has been done for cylinders of three diameters at velocities of 0.01
and 1.0cm-s"1 (Fig. 5.3). The Reynolds number of the flow around the
largest cylinder at the highest velocity is unity. The ordinate is the clear-
ance rate per unit of length of the cylinder, equivalent to the encounter
rate divided by concentration of particles and collector length. This
definition of clearance rate, based on a unit length of the collector, is
different from the clearance rate, R, defined in Chapter 2, which is based
on a single animal, or 5, the seston removed by a unit weight of a
suspension feeder. The diffusion coefficient has been calculated from
equation (5.15) at 20C. The curves have been terminated where the
particle diameter exceeds 10% of the collector diameter.
In the Brownian diffusion-dominated regime, the predicted clearance
rate is inversely related to particle size as a consequence of the depend-

ence of the diffusivity on the inverse of the particle diameter. The mini-
mums occur because of the increasing importance of direct interception
at higher velocities and its second-power dependency on particle diam-
eter. The curves show the importance of diffusion to the collection of
seston of the size range of bacteria or smaller in the range of sizes and
velocities used in preparing the plot. For smaller obstacles at lower
velocities it is clear that diffusion will be more significant than direct
interception. Seston larger than about 10|iim will more likely be collected
by direct interception unless the velocity is extremely low. Motile parti-
cles will behave differently. Their apparent diffusivity will likely increase
with particle size so that the plot of encounter rate against particle
diameter will not generate a minimum. The diffusion-dominated size
range will extend to higher particle diameters.
The combination of electrostatic forces and direct interception has
also been considered in some detail. As mentioned earlier, the modern
theory of direct interception in liquids accounts for the hydrodynamics of
the drainage flow from under the particle as well as the influence of
surface forces (Spielman 1977). The interaction of the two mechanisms is
described by the adhesion number 7Vad, which is defined as
N
ad
where
Q = Hamaker constant
rc = radius of the collector
rp = radius of particle
The term AF describes the flow field around the collector, equation
(5.7b). The term Q is a constant of proportionality in Hamaker's equa-
tion for the force between a spherical particle and a much larger collec-
tor. It depends on the properties of the molecules of the two surfaces and
of the liquid medium (Spielman 1977). Monger and Landry (1991) have
used a value of 3.7 x 1CT14 erg for general lipid bilayer systems in
seawater. The adhesion number can be interpreted as the ratio of van der
Waals attraction to hydrodynamic forces.
Spielman and Fitzpatrick (1973) solved the equation for the limiting
particle trajectory around a cylinder in the presence of an attractive
force. By knowing this trajectory, the collection efficiency can be calcu-
O >-
LU o
ccz
Q LU
11
LU Q_
LLJLU
E LU
I I ' ' 'I
1CT5 10"4 10" 3 10"2 10~1 10 101

ADHESION NUMBER
Figure 5.4 Normalized capture efficiency versus adhesion number for capture
of neutrally buoyant particles by a cylinder: direct interception (Eq. 5.7) and van
der Waals attraction. Dashed line (Eq. 5.21), capture by van der Waals attraction
only (Spielman 1977).
lated. Their results were presented in the form of a dimensionless plot of

the normalized capture efficiency vs. the adhesion number (Fig 5.4). The
normalized capture efficiency is the predicted capture efficiency divided
by that of the classical theory equation (5.5). It is interesting to note that
the classical theory does not agree with the more advanced theory except
coincidentally at one value of the adhesion number: i.e. there is no
asymptotic convergence of the two treatments. At low adhesion numbers
the capture efficiency is lower than predicted by the classical theory,
primarily because of hydrodynamic drainage from under the particle. At
high adhesion numbers, the electrostatic effects become dominant and
the capture efficiency is given by
J. (5.23)
These theoretical results for capture by cylinders were compared with

experimental measurements by Schrijver, Vreeken, and Wesselingh
(1981), who used a glass cylinder of 250-Lim diameter and latex particles
ranging in size from 0.8 to 25.7 Lira. The working fluid was an aqueous
electrolyte solution so that the only important surface force was the
London-van der Waals force. These particles were not of neutral buoy-
ancy. The cylinder was mounted in a vertical position to prevent gravita-
tional collection for measurements of collection efficiency. The results

fell within a factor of 2 from the theoretical prediction and confirmed the
general form of the relationship derived by Spielman and Fitzpatrick,
although the interpretation of the results was complicated by a signifi-
cant contribution of Brownian diffusion of the smallest particles.
Spielman and co-workers have also derived similar results for collec-
tion by spheres and a rotating disc (Spielman 1977). The effect of gravity
has also been included (Spielman and Fitzpatrick 1973). Monger and
Landry (1990) have used a similar approach to analyze the capture of
bacteria by marine flagellates. Shimeta and Jumars (1991) have argued
that the assumptions of smooth collecting surfaces and steady velocity
are not satisfied in this case and that this refinement of interception
theory is unnecessary.
Clearance rates were earlier calculated for the combined mechanisms
of direct interception and Brownian diffusion for some typical conditions
(Fig. 5.3). A similar calculation has been made by using Spielman and
Fitzpatrick's treatment for direct interception with an allowance for
surface forces (Fig. 5.5), using Q = 3.7 x 10~14. The plots are for two
velocities, 0.01 and 1.0 cm-s"1, and two cylinder diameters, 10 and
lOOjum. One of the curves from Fig. 5.3 has been superimposed on the
plot to show the relative importance of Brownian diffusion. If the plots
are compared it is seen that particles of the typical size of bacteria (say
diameters about 0.5 jum) or smaller are cleared more rapidly by
Brownian diffusion (curve on the upper left of Fig. 5.5) than by direct
interception and electrostatic attraction, particularly at lower velocities.
The curves on the right hand side of both graphs represent collection
dominated by direct interception, but these have been calculated differ-
ently and the clearance rates are not consistent for Figs. 5.3 and 5.5. The
classical theory, which does not incorporate surface forces, was used for
Fig. 5.3, and the clearance rates of Fig. 5.5 should be considered to be the
more accurate estimates.
There is one important caveat to be raised regarding these results. The
implicit assumption in the treatment of surface forces is that the surfaces
of the cylinder and of the spherical particles are smooth. Any surface
roughness or asperities of the order of a few hundred angstroms or more
will change the rate at which the particle approaches the surface, as will
any departures from sphericity of the particle. Collecting surfaces and
seston particles are more likely to violate these geometric constraints
than not.
C/D
LU
10"
LU
o
0.01 cm S'1
5 1Q- E 5
o
10-
0.01 0.1 1 10
PARTICLE DIAMETER, micrometres
Figure 5.5 Clearance rates for right circular cylinders of diameter lOjLim (solid)
and 100|Lim (dotted) at velocities of 0.01 and 1.0 cm-s"1, attributable to direct
interception and electrostatic attraction. The dashed line is taken from Fig. 5.4
for direct interception, Brownian diffusion for U= lcm-s"1; dc = lOjim.
Assessing the relative importance of capture mechanisms

Aerosol theory suggests a simple method of assessing the relative impor-
tance of direct interception, Brownian diffusion, and electrostatic attrac-
tion in the collection process. The efficiency indices which have been
defined for each mechanism provide a simpler means of estimating which
is likely to be the predominant mechanism in a specific case (Ranz and
Wong 1952; Shimeta and Jumars 1991). These have been collected in
Table 5.2. In many practical situations, substitution of reasonable values
of the key parameters into these ratios can quickly identify the relative
importance of each of the collection mechanisms. As the preceding
discussion has suggested, however, more than one mechanism may be
important. In addition, relative encounter rates are not necessarily re-
flected well by the efficiency indices, as has been discussed by Shimeta
and Jumars (1991), and the ratios should be interpreted with caution.
Table 5.2. The five mechanisms of the aerosol theory: efficiency indices, engineering collection efficiencies, and
encounter rates for infinite right circular cylinder at low Reynolds number.*
Mechanism Efficiency index Engineering efficiency Encounter rate

2 2
Direct interception NR = dpd;1 ER = 2AFd d; FR = 2ApcUlc dp2dc~l
(classical theory)
Inertial impaction E, = 2lsd,r1 if l,d c F^lcUlX
E, = 1.0 if /, dc Fj = cUdclc
1
Gravitational Ea=WsU->(l + dpdc- ) F
G=CWslc{dc+dp)
deposition
Brownian diffusional ND = Dir1d<r1 ED=2.92{2-lnRecyV3{UdjDy2p FD = l.llD2^clc(Urcf(2-\n Rec)~1/3
deposition
Electrostatic .U)-1 E E = NE
attraction
a
Symbols defined in text.
Additional insight into this question has been provided by Shimeta

and Jumars (1991), who have calculated encounter rates and efficiency
indices for a number of particles of importance to suspension feeders,
which range in size from 1 \im (medium clay) to 2500 jim (marine snow).
Collecting surfaces are circular cylinders of sizes of practical interest.
They have presented their results in the form of plots of clearance rates
and of encounter rates and efficiency indices normalized by those for
direct interception alone as a function of ambient velocity. These plots
are restricted to Rec < 1. The rate of direct interception has been calcu-
lated using equation (5.5).
One of their plots has been reproduced here (Fig. 5.6). Clearance rates
per centimetre of cylinder are plotted against ambient velocities from
0.1 to SOcm-s"1. In plot A, the clearance rates attributable to direct
interception for nine particles of relevance to benthic suspension feeders
(Table 5.2) are shown. The clearance rate is directly dependent on the
size of the particles and the velocity but is independent of collector
radius for this mechanism. The plot is for Rec < 1, which implies that at
the highest velocity of SOcm-s"1 the plot should not be used if the diam-
eter of the collector exceeds 2|im. These plots were derived on the
assumption that the particles are much smaller than the diameter of
the collector. In the case of medium silt, for instance, the particle diam-
eter is 20|Lim. At Scm-s"1, the maximum acceptable cylinder diameter,
according to the Reynolds number limit, is also 20\im so that the theory,
devised for particles small with respect to the cylinder, is really only
applicable to velocities well under 5cm-s-1. Plot B shows clearance
rates attributable to inertial impaction on a cylinder of diameter of
50|im. The dashed portions of the lines denote violation of the low
Reynolds number condition. As would be expected the largest clearance
rates are seen for the particles with the largest settling velocities. The
smaller particles are little affected by inertial forces. At a velocity of
1 cm s"1 the clearance rate for direct interception of small phytoplankton
cells is almost two orders of magnitude greater than for inertial intercep-
tion. The combined clearance rates due to inertial impaction and gravi-
tational deposition are shown on plot C (dc = 50\im) and plot D
(dc = 1000 |im) . The horizontal sections of the curves show where gravi-
tational deposition is dominant. Particles with the highest settling rates
have the highest clearance rates, usually comparable with the direct
interception rates, especially at the higher velocities. Clearance rates
due to Brownian diffusional deposition are shown on plot E for the
smallest particles and collectors of various diameters. They are not neg-
10
i A B
1 0n- -^3 O
10 1 -
3
A
10 - ' * ' I1
icr2-
CO
10"5-
104- 10 7 -
i 1 0 -6.
o iff9
I
o I
C/)
CO DC
E
o c 3 L 9...,
LU n
10-
D
LU irT3 -2
10 - *o ^^ 10 - r.-.-.-.-A-------"^7
[
'-''','
DC 10~5- 4 ,/. sA
LU
LUI 10" -
O 10~7- I I I o 10" 0- .1
6
I
1
I
10 50
I
<
cc VELOCITY, cm
LU 10 -i r 1 Clay
1 1 5 ( r c _ 5 ( )|im) 2 Silt
o 10 -
^^-^ L^-1 5(rc = 5 |im) 3 Sand
= 50 |im) 4( DMA
^ * *
10~ -7 (rc = 50[im) 5 Bacterium

6 Small Phytoplankton
irf8 7 Large Phytoplankton
0.1 1 10 50 8
VELOCITY cm s"1 9 Marine Snow
Figure 5.6 Capture of typical seston particles by a circular cylinder (expressed

as clearance rates per unit length against velocity) for the mechanisms of the
aerosol theory: A, direct interception; B, inertial impaction; C and D, inertial
impaction and gravitational deposition with dc = 50 urn (C) and 1000 urn (D); E,
Brownian diffusional deposition (rc = radius of cylinder); OMA, organic matter
aggregate. Lines are either ended or dotted if Rec > 1 (Shimeta and Jumars 1991).
ligible with respect to direct interception, especially at the lowest

velocities.
Although these plots are not applicable to all situations, a methodol-
ogy for assessing the relative importance of the collection mechanisms
for a particular case is clear from Shimata and Jumars's presentation.
This approach is more illuminating than a simple use of indices. Note
that a more accurate method of calculating the direct interception rates
could have been used, as has been shown by Shimeta (1993), who used
equation (5.7) rather than equation (5.5).
Higher Reynolds numbers

The primary collecting elements of suspension feeders are typically small
and of cylindrical shape. It has been argued that in a majority of cases the
velocities are such that they operate in the creeping flow regime
(LaBarbera 1984). This appears to be so for many active benthic suspen-
sion feeders and for planktonic suspension feeders. Many of the passive
benthic suspension feeders, however, can be shown to have primary
collecting elements which commonly experience Reynolds numbers in
excess of unity (Shimeta and Jumars 1991). For example, the pinnules of
the soft coral Alcyonium siderium are about 380 jam long and 60jim in
diameter spaced at gaps of about 230 Jim (Sebens and Koehl 1984). The
velocity past the pinnule need only exceed 1.67 cm s"1 for Rec to be over
unity, which is well within the range of velocities observed in nature.
The relationships developed for encounter rates for the creeping flow
case cannot be extended to Rec greater than about 0.1 without some
error. These relationships are based on assumptions about the form of
the streamlines of the flow around a cylinder which are derived from
Lamb's solution. Experimental drag force measurements, which reflect
the shape of the streamlines, show that at Rec = 1.0 the Lamb solution
overpredicts drag by about 25%. The streamlines, symmetrical about the
cylinder in creeping flow, become compressed on the upstream side and
relaxed downstream. The critical trajectory for direct interception will be
different, consistent with a greater collection efficiency. At Rec of about
10, a pair of attached vortices form on the downstream side of the
cylinder. The form of the streamlines around a long circular cylinder at
Rec > 1 can be estimated (see, for instance, Davies and Peetz (1955) for
a calculation at Rec = 10). At Rec > 40, vortices form alternately on each
side of the downstream face and are shed periodically. At higher veloci-
ties the wake becomes turbulent.
From this description of the hydrodynamics of thisflow,it is clear that
the encounter rate equations lose their predictive value at higher
Reynolds numbers and that the functional dependencies of encounter
rate on collector size and velocity, used to deduce the relative impor-
tance of mechanisms, are not reliable. For a full discussion of the details
of particle collection at higher Reynolds numbers, see Shimeta and
Jumars (1991).
Other geometries
Encounter rate predictions for spherical collectors under low Reynolds
number conditions can be found in Shimeta and Jumars (1991). Cylin-
ders of non-circular cross sections are not amenable to simple analysis,
except in the case of gravitational collection, because of the lack of
hydrodynamic solutions for the flow around them, or because of their
complexity if they exist.
Assemblages of collectors
Many passive suspension feeders do not present single collecting ele-
ments to the flow but instead use arrays in a more or less parallel
configuration. For example, consider the pinnules on the tentacles of the
sea pen (Best 1988), the octocorals (Patterson 1991a), the comatulid
crinoids (Mayer 1979), and the stalkless crinoid Antedon mediterranea
(Leonard 1989). Active suspension feeders may also use similar arrays,
as is the case with bivalve molluscs. The first effect of introducing addi-
tional collecting elements is to modify the velocity field near the surface
of the collectors. The term AF, which was defined in equation (5.7a) for
the case of flow around an isolated circular cylinder, is modified for
arrays of cylinders (Spielman 1977). Streamlines will be compressed and
direct interception efficiency will be improved. The second effect is that
the bulkflowwill be influenced. In the case of passive collectors present-
ing an array of collecting elements to a stream, the flow between the
collecting elements will be reduced and more of the approaching fluid
will be diverted around the array if the elements become more closely
spaced.
Active suspension feeders will be required to supply pumping energy
to move fluid through the filtering apparatus at a rate proportional to the
superficial velocity and the area of collecting surface presented under
low Reynolds number conditions.
Approximate solutions to the problem of the filtering of aerosols by
beds of cylinders have been developed for low Reynolds number flows.
The approach has been to simplify the problem to one in which the
particles are assumed to be small with respect to the diameter of the
collecting cylinders so that the velocity distribution need only be known
close to the surface of the cylinder. The influence of the neighboring
cylinders is reflected in the value of the term AF introduced earlier for
isolated cylinderflowsin equation (5.7b). This term can be related to the
volume fraction of fibres in the filter bed. The equations developed for
capture by an isolated fibre can then be applied with the new value of AF
(Spielman 1977). The functional dependencies of encounter rates by the
aerosol mechanisms on parameters such as velocity, particle diameter,
and collector diameter are unchanged.
Interpreting mechanisms from parameter sensitivity

It is sometimes argued that the dominant collecting mechanism can be
identified if the dependency of collection efficiency or encounter rate on
changes of certain independent variables is observed. For instance, if an
experiment is carried out using particles of different sizes with all other
variables fixed, the efficiency of collection should increase as dp, equa-
tion (5.6), if direct interception is the dominant mechanism. If Brownian
diffusion, on the other hand, is dominant, the dependency is to the -2/3
power, equation (5.19). In principle, this should provide a means of
identifying dominant mechanisms, but inspection of Table 5.2 shows that
relationships are not that straightforward in many instances. The de-
pendence of efficiencies on velocity, in particular, is complicated by the
term AF, which must be known for the flow past the collecting elements
and which may not be easily determined in either laboratory or field
experiments. In practice, the indices should be used with caution and
calculations of encounter rates for model systems similar to the organism
under study are a preferred method of analysis.
Ambient flow affects particle capture

Our treatment has been based on the assumption that collecting ele-
ments are rigid cylinders in a steady flow, conditions which are by no
means universal. Benthic suspension feeders often live in wave-driven
and turbulent environments, and the collectors experience periodic and
randomly varying velocities. They may also be in shear zones, as opposed
to uniform flow zones. The scale of the turbulent motions may be larger
or smaller than the characteristic size of the collector, depending on the
energy viscous dissipation rate of the turbulent flow and the collector
size. Collection in wakes of either whole organism or single collecting
elements may also be important. All of these hydrodynamic factors will
affect the rate of collection of particles. In addition, collecting elements
are often not rigid, but either move with the motion of the water, periodi-
cally sweep, or respond behaviorly to the presence of seston or velocity
(see Chap. 4).
The adhesion or retention coefficient has been assumed to be unity in
this discussion. In reality little is known of its magnitude, which depends
Table 5.3. Passive suspension feeders for which aerosol theory seston
encounter efficiency index calculations are available.
Scientific name Major taxa Reference

Alcyonium siderium Soft coral Sebens and Koehl (1984)
A. siderium Soft coral Patterson (1984, 1991a,b)
Briarium abestimum Soft coral Lasker (1981)
Pseudoplexaura porosa Soft coral Lasker (1981)
Pseudopterogorgia americana Soft coral Lasker (1981)
1Alcyonium sp. Soft coral McFadden (1986)
Meandrina meandrite Scleractinian coral Johnson and Sebens
(1993)
Ophiopholis aculeata Brittle star LaBarbera (1987,1984)
Ptilosarcus gurneyi Sea pen Best (1988)
Oligometra serripinnia Feather star Holland et al. (1986)
Florometra serratissima Crinoid Byrne and Fontaine
(1981)
on the balance of hydrodynamic forces and electrostatic forces. It is then

dependent on the flow, the particle and collector geometry, and the
chemistry of the surfaces. It is not necessarily constant for any given
combination of seston and collecting element and can be expected to be
velocity-sensitive.
Seston capture mechanisms

Five major categories of suspension feeders were recognized in Chapter
4: passive, active (inclusive of sieving and non-sieving forms), facultative-
active, combined passive-active, and deposit-suspension switching spe-
cies. Of these, only the first two groups have been cursorily examined as
to which of the aerosol theory seston capture mechanisms was predomi-
nant. These include passive and a few active suspension feeders.
Passive suspension feeders

A list of the species for which data involving seston encounter efficiency
calculations have been made is shown in Table 5.3. The species involved
are either corals or echinoderms. A review of much of this work is
presented by Shimeta and Jumars (1991). They point out that LaBarbera
(1978) used an inappropriate way of calculating the efficiency of seston
capture by direct interception. Further analysis by LaBarbera (1984),
incorporating the hydrodynamic retardation model of Spielman (1977)
and allowance for surface forces in seston capture by the brittle star, was
also criticized by Shimeta and Jumars (1991) on several grounds. Among
those grounds was that the hydrodynamics of the flow near the collector
was not fully defined, that simple measures of mainstream flow are
insufficient to characterise the appropriate velocity to the collector, and
that streamlines around the collector depend on the Reynolds number,
which, in the case of the brittle star, was >1 and therefore strictly not
covered by aerosol theory.
Most of the seston encounter efficiency results obtained by authors
shown in Table 5.3 suggest that direct interception was the predominant
seston capture mechanism. Patterson (1984,1991a), for example, found
that the details of the capture process in Alcyonium siderium were more
complicated than might be visualised by a simple application of aerosol
mechanisms. Capture sites were not evenly distributed about the colony,
and the frequency and location of capture were influenced by the veloc-
ity and turbulence of the ambientflow.Patterson also observed that cysts
which were caught on the aboral side of the tentacle were usually lost
before transfer to the mouth, leading him to suggest that large particles
are more likely to be caught effectively by the tentacles on the down-
stream side of a polyp. The efficiency indices were calculated by using
the tentacle width as a characteristic dimension, typical velocities near
the tentacle, and a settling velocity of cysts of 0.08 cm s"1. The indices for
a velocity of 30cm-s~1 were as follows:
Gravitational deposition 0.003
Direct interception 0.67
Inertial impaction 4 x 10~5
Diffusive deposition 8 x 10"12
For a velocity of 3 cm- s"1, the direct interception index was unchanged,
the gravitational and diffusive deposition indices were increased by a
factor of 10, and the inertial impaction index was reduced by a factor of
10. The indices suggest that direct interception was the dominant mecha-
nism at velocities commonly encountered in nature. It should be empha-
sised, however, that some of the assumptions used in the development of
the indices are violated. The efficiency of collection of cysts was also
directly measured at flow velocities ranging from 2.7 to 19.8 cm-s"1 and
found to decrease with velocity, qualitatively consistent with a dominant
CAPTURED CAPTURED
L = 331 201 L = 593 284
250 -
N = 452 N = 452
125
DC
LU
GO
AVAILABLE PLANKTON ^> AVAILABLE PLANKTON
400 - /_ = 713620 /_ = 2021 8 1 7 ,
A/ = 1017 N =1017
LU
DC
200 -
1500 3000
SIZE, Mm
Figure 5.7 Size distributions of particles captured by Alcyonium siderium and
of the available plankton: L, length 1 SD; plots on left, based on number of
items; plots on right, based on biomass (Sebens and Koehl 1984).
mechanism of direct interception (equation [5.6], the weak inverse de-

pendence on velocity appearing in the AF term). Sebens and Koehl
(1984) showed that the same species of soft coral was a benthic
zooplanktivore selecting smaller prey, similar in size to the tentacles
(width of 300 |nm), but much larger than the pinnule diameter of 60 jura
(Fig. 5.7). A regression of collection efficiency against the Reynolds
number of polyps (based on the diameter of the oral dise and the velocity
lcm above it) gave a dependence on Reynolds number to the power
-1.55. This result suggests that in this case particle capture is influenced
by factors in addition to those of the aerosol theory. Perhaps most
important is the observation that many particles which struck the surface
of the tentacles were not retained; Patterson (1991a) noted the need for
additional study of this process. Changes in velocity also change the
orientation of the polyps in the stream, resulting in differences in collec-
tion efficiency between upstream and downstream polyps.
Other possible aerosol mechanisms reported by the authors listed in
Table 5.3 include gravitational deposition at low velocities in the coral
Meandrina meandrites (Johnson and Sebens 1993) and strong adhesion
of particles in feather stars (Holland et al. 1986).
Active suspension feeders

The active suspension feeders which have been examined to determine
the mechanism of seston capture include only two categories: sieving
suspension feeders and suspension feeding bivalve molluscs.
Sieving suspension feeders

The polychaete Chaetopterous variopedatus generates aflowthrough its
U-shaped tube by muscular contractions of its fan-shaped notopodia (see
Fig. 4.10). A mucus net used for feeding is generated continuously within
its tube at the twelfth segment and is rolled up and eaten at 15-min
intervals (J0rgensen et al. 1984). The net is made of longitudinal fila-
ments estimated to be about 0.1 |um thick and transverse filaments of
smaller diameter. The mesh size is 0.76 0.96 by 0.46 0.12 um (Flood
and Fiala-Medioni 1982). Velocity through the mesh was estimated to be
3mm-s 4 . Particle retention efficiency was measured by J0rgensen et al.
by two different makes of electronic particle counters, which gave differ-
ent results in the size range of interest. Retention of 90% was seen at a
particle size of 1.2-1.7 Jim, depending on the measurement method used.
Bacteria of about 1 jum were retained at about one-third efficiency, as
measured by fluorescence microscopy. The dominant collection mecha-
nism is clearly sieving, but there is no evidence of how, if at all, smaller
bacteria (say about 0.5 (im) are captured. These particles will be captured
by direct interception and Brownian diffusion. As the seston is 5-10
times the filament diameter size, the aerosol theory is strictly
inapplicable.
Similar measurements were made of particle retention by the gastro-
pod Crepidula fornicata, which generates a current through the mantle
cavity by the lateral cilia on the gillfilaments(J0rgensen et al. 1984). Two
mucus nets, one coarse and at the entrance to the mantle cavity, the
other fine and transported across the gill surface, collect particles. The
morphology of the fine net is not known but it is better than 90% efficient
at removing particles larger than 1.2 (im. The pressure drops across the
nets of both of these mucus net feeders were estimated and seemed to be
consistent with the capacity of the pumping mechanisms.
J0rgensen et al. (1984) also described particle capture by mucus nets of
the solitary ascidians Ascidia obliqua, A. Virginia, and Ciona intestinalis.
Seawater flow in the trophic fluid transport system of ascidians is de-
scribed in Chapter 4, the section, Sieving suspension feeders; also see
Fig. 4.8. A mucus net is continuously formed and drawn across the
pharynx wall. These nets become leaky to particles in the 1- to 2-um size
Suprabranchial Exhalant jet

cavity [2 cm-s" 1 )
Mantle cavity
B
F LF L
Mantle -Filter Pump-Suprabranchial

cavity cavity
Figure 5.8 A, schematic of water flow through Mytilus edulis; B, mean flow
streamlines through three gill filaments: F, frontal cilia (beating to the plane of
the diagram; LF, laterofrontal cirri (effective stroke to the left); L, lateral cilia
(effective stroke to the right) (Silvester and Sleigh 1984).
range, depending on the species, as is consistent with a sieving mecha-

nism from what is known of mesh sizes (Flood and Fiala-Medioni 1981).
Bivalves
Although there is a long history of scientific study of the mechanism of
seston capture in suspension-feeding bivalve molluscs, there is still no
consensus on how it is achieved. The earliest view was that seston was
removed from inhalant water by passing it through sievelike structures
or gills. Mucus was thought to contribute to this mechanism by removing
the already captured seston and/or providing the actual sieving net;
hence, the mucociliary sieving theory. The ciliary pump in active suspen-
sion feeders consists of ciliary bands with seston collected by another set
of cilia beating in the direction of the mouth (J0rgensen et al. 1984). The
seston may be collected upstream of the ciliary pump, as in bivalves, or
downstream, as in sabellid polychaetes (J0rgensen et al. 1984). The
mechanisms by which sestonic particles leave the viscous stream through
pinnules and gill filaments to enter the surface flow to the mouth are still
a matter of contention. The trophic fluid transport system of the blue
mussel Mytilus edulis is shown schematically in Fig. 5.8A.
The sieving hypothesis generally considered the laterofrontal cirri the
location where sieving occurred (Fig. 5.8B). The Reynolds number of the
flow, based on the size of the laterofrontal cirri, is of the order of 10~4
(J0rgensen 1983). Each laterofrontal cirrus is in fact a compound struc-
ture consisting of 22-26 pairs of cilia arranged in two parallel but alter-
nating rows (Moore 1971; Owen 1974). The spacing of the cirri is about
2.0-3.5 |im, and the branched cilia are separated by about 0.6 |um.
Silvester and Sleigh (1984), Silvester (1988), Jones et al. (1992), and
J0rgensen et al. (1988) provide a discussion of the dimensions of the cirri
and the interfilamentary channel. The cirri are arranged in alternate rows
so that the branched cilia overlap slightly to form a barrier to the passage
of particles. The alternate rows have an effective stroke of 90 directed
towards the frontal cilia, one row being in position to screen water
coming through the ostium, while the other, a half cycle out of phase, is
in the recovery stroke (Owen 1974). The gills of Mytilus edulis have been
seen to remove particles larger than about 5|iim with close to 100%
efficiency. Particles of 1-2 jam are removed with about 50% efficiency
and particles smaller than l|um are removed with about 10% efficiency
(J0rgensen 1975).
J0rgensen (1975) was among the first to doubt the sieving hypothesis
in bivalve molluscs. He pointed out that observations on the dependence
of retention efficiency on particle size were not consistent with the mesh
openings provided by the laterofrontal cirri. In addition, the pressure
drop expected across the laterofrontal cirri in a position which blocked
the flow in the interfilamentary passage was too high to be supplied by
the ciliary pump (J0rgensen 1981a). Silvester and Sleigh (1984) chal-
lenged the pressure loss argument on two grounds: that the cilia formed
a more complicated organ than the series of parallel cylinders assumed
by J0rgensen (1975), leaving larger openings for the passage of inhalant
water, and that the velocities calculated by J0rgensen in the
interfilamentary canals were too high. Observed particle retentions were
reconciled by a sieve model having openings of a range of dimensions

and separation by sieving, without contributions from aerosol theory
mechanisms.
Because of his doubts with regard to the sieving theory in mussels,
J0rgensen (I981a,b) proposed an alternative one: the hydromechanical
shear force theory. According to this proposal, the laterofrontal cirri
function as flow modifiers rather than filters. This is consistent with the
theoretical analysis by Cheer and Koehl (1987) that in viscous flow,
bristled appendages act more as paddles than as rakes, with the cirri
moving water rather than sieving it. The seston particles are separated
from the main trophic fluid flow by hydrodynamic forces which move
particles across shear zones (see also J0rgensen 1983, 1990) and which
are well known in fluid mechanics (Leal 1980). The three-dimensional
nature of the flows in front of the gills (J0rgensen 1990) and, in particu-
lar, the interactions between the pulsating flows generated by the
laterofrontal cirri and frontal cilia are not clear. There is very little
guidance available from established fluid mechanics to confirm or refute
the hydromechanical shear force theory. Silvester and Sleigh (1984)
argued against it on the basis that the order of magnitude of the lateral
forces available for diverting particles was too low to explain the particle
separation achieved by blue mussels.
Work on the blue mussel pump by J0rgensen (1982) and J0rgensen et
al. (1986,1988) (and see Chap. 4) has provided further evidence against
the sieving hypothesis in bivalve suspension feeders. The pressure drop
across the laterofrontal cirri cannot be estimated correctly if they are
assumed to be arrays of cylinders blocking the interfilamentary canals, as
proposed by Silvester and Sleigh (1984). The 2-|nm spaces between alter-
nating cirri beating in phase would suggest that a fraction of the water
bypasses the cirri and is not filtered. The mussel pump model suggests
that only the distal half of the lattice is passable and the forward stroke
of the cirri will produce an additional loss component. The pressure drop
across the laterofrontal cirri cannot be measured directly, but the resist-
ance the cirri represent can be removed experimentally by adding
serotonin to the water in low concentrations. Serotonin is a nerve trans-
mitter in the bivalve gill, which, when added to ambient water, causes the
laterofrontal cirri to remain more or less fixed at the end of the stroke,
thus eliminating their flow resistance (J0rgensen 1990). Comparing the
pump characteristic for the system with and without serotonin, it was
concluded that the frictional resistance was not measurably affected by
the removal of the laterofrontal cirri. This result would support the
contention that most of the flow is not sieved by the cirri. When the
laterofrontal cirri are immobilized by serotonin, the retention efficiency
of algal particles smaller than 6|im is drastically decreased but that of 14-
lam particles is little changed. J0rgensen et al. (1988) suggest that the
laterofrontal cirri do not act as sieves of the through current, but rather
as modulators in the process of particle retention, improving its effi-
ciency for the smaller particles. Recent work by Nielsen et al. (1993) has
lent support to this view. Videographs of particle tracks in the vicinity of
an isolated gill filament showed a large fraction of sestonic particles
passing from the through to the frontal currents without any contact with
cilia. This diversion of particles was largely absent when the laterofrontal
cirri were immobilized. In bivalve taxa other than mussels, the
laterofrontal cilia may be absent or less well developed, and this is
considered by Riisgard (1988) and Nielson et al. (1993) to be an adapta-
tion concerned with improving the capture of smaller particles.
Summary
It is clear that it is usually possible to identify the dominant aerosol
mechanism(s) operating for a specified passive benthic suspension
feeder after a relatively cursory examination of the parameters of the
problem. The more ambitious aim of calculating the efficiency of collec-
tion or the encounter rate for a whole animal from observations of
animal morphology, fluid flow conditions, and type and quality of seston
collected has rarely been successfully completed in the same animal.
Although of practical interest, further work of the latter kind may not be
very helpful in advancing our understanding of the mechanisms of sus-
pension feeding, which are the focus of this chapter.
Some of the details of the mechanisms of suspension feeding have
been clarified by the use of the aerosol theory, since their introduction to
marine biology in 1977, but there have not been as many advances as
might have been reasonably expected. Some of the reasons for limited
success of the aerosol theory are described later. One limitation is that it
is only well developed for the case of capture by long circular cylinders
of smooth surfaces at low Rec. Although the low Rec condition is often
satisfied, the collecting elements are often not circular in cross section
and the surfaces may have roughnesses and protuberances. Nor are they
always oriented normal to theflowvector and they may beflexiblerather
than rigid. In addition, the theoretical efficiency predictions are devel-
oped on the assumption that the seston diameter is much smaller than
the collector diameter. In most of the field and laboratory cases studied,
seston sizes were widely distributed and frequently overlapped the diam-
eter of the collector. This may not be a difficulty in identifying dominant
mechanisms, however, since it is inevitable that seston particles close in
size to that of the collector will be captured mainly by interception.
Real flow situations are seldom as simple as the idealized flows treated
by the theory. Nearby structures may alter the streamlines about the
collector, although this effect may be compensated for by the analyses
developed for fibre filter beds. Oscillating flows typical of wave environ-
ments or the stochastic flow caused by turbulence may alter the encoun-
ter rates and have not been fully studied. In passive suspension feeders,
the collecting elements may see only a portion of the flow which ap-
proaches the whole animal. Much of theflowmay be diverted around the
animal. Some will flow through but not encounter the filtering elements,
passing between the tentacles, for instance. This means that the flux of
seston based on the approaching flow overestimates what the encounter
rate, summed over all the collecting elements, actually is.
Passive suspension feeders have proved generally easier to study than
active suspension feeders. The following conclusions can be drawn:
Much of the work that has been reported on benthic suspension feeders up to
1995 had dealt with animals with primary collecting elements close to cylin-
drical in shape (though there are many exceptions) and with a flow around
them of Rec order of 1 or less. On the other hand, the preferred seston usually
is not significantly smaller (say by an order of magnitude) than the diameter
of the collectors so that this constraint on the theory is not satisfied.
Direct interception is an important mechanism for most non-motile organic
particles captured by passive suspension feeders, but if the modernized
theory of interception is applied, surface properties must be known. This
difficulty may be more real in theory than in practice in view of fact that
surface irregularities of typical collection elements are often larger than the
characteristic distance over which the surface forces are important (order of
0.02 urn). It is unfortunate that the theory for direct interception does not
reach an asymptotic state at low values of the adhesion number so there is no
limiting expression, as there is for the other extreme of very large intersurface
forces. The "classical" expression for collection efficiency, equation (5.7),
should give a reasonable estimate for many situations.
Gravitational separations may be important for certain particles which have
a significant sinking rate, but few reports are found in the literature.
Brownian diffusion is an important mechanism for collection of particles of
the size of the order of 1 jam, e.g. bacteria.
At the typical scales and Reynolds numbers encountered, inertial impaction
is probably not a significant mechanism for living organic seston. Inorganic
particles are often much more dense and, although not common in the litera-
ture reviewed, some inertial effects might be expected under favorable cir-
cumstances of particles with large stopping distances in rapid flows.
The ciliary, flagellate, and muscular pumps in suspension feeders char-

acteristically operate in trophic fluid transport systems where viscous
flow is present. Clearance measurements show that many benthic active
suspension feeders efficiently remove particles smaller than the gaps
between the filter elements so that other mechanisms than sieving are
probably involved. Detailed study of the blue mussel has shown that the
bivalve gill is not a static filter, but is characterized by pulsatingflowsand
moving cilia and cirri at very low Reynolds numbers. A mechanism other
than that included in the aerosol theory may be involved, as has been
described in our brief account of the controversy between mucociliary
and particle migration by shear force theories. Few other active benthic
suspension feeders have been studied in enough detail to determine their
mechanisms of seston capture.
Finally, the theory associated with sieving, or aerosol mechanisms, has
not yielded experimentally testable hypotheses by benthic biologists.
This applies also to the complex theory of hydromechanical shear force
proposed by J0rgensen (1981a) in mussels. Consequently, we have not
found in the work concerned with seston capture in benthic suspension
feeders any key hypotheses around which this chapter could be con-
structed. Thus, this chapter is different from Chapters 3 through 8. Per-
haps the way forward depends on further development of a technology
concerned with observation and manipulation of living microscale col-
lection surfaces, including local viscous flows throughout the trophic
fluid transport system. Then the framing of key hypotheses, particularly
for active suspension feeders, could lead to the determination of the
mechanisms of seston capture.
References
flow on volume flow rate and filtering efficiency. Biol. Bull. 175: 332-342.
Brun, R. J., W. Lewis, P. Perkins, and J. Serafini. 1955. Impingement of cloud
droplets on a cylinder and procedure for measuring liquid-water content
and droplet sizes in supercooled clouds by rotating multicylinder method.
Nat. Adv. Comm. Aero. Report 1215. NACA, Washington, D.C.
Byrne, M., and A. R. Fontaine. 1981. The feeding behaviour of Florometra
serratissima (Echinodermata: Crinoida). Can. J. Zool. 59: 11-18.
Cheer, A. Y. L., and M. A. R. Koehl. 1987. Paddles and rakes:fluidflow
through bristled appendages of small organisms. J. Theor. Biol. 129: 17-
39.
Crank, J. 1956. The mathematics of diffusion. Clarendon Press, Oxford.
Davies, C. N., and C. V. Peetz. 1955. Impingement of particles on a transverse
cylinder. Proc. R. Soc. A. 234: 269-295.
Dorman, R. G. 1966. Filtration, p. 195-222. In C. N. Davies (ed.) Aerosol

science. Academic Press, London.
Flood, P. R., and A. Fiala-Medoni. 1981. Ultrastructure and histochemistry of
the food trapping mucous film in benthic filter-feeders (Ascidians). Acta.
Zool. (Stockholm) 62: 53-5.
1982. Structure of the mucous feeding filter of Chaetopterous variopedatus
(Polychaeta). Mar. Biol. 72: 27-33.
Fuchs, N. A. 1964. The mechanics of aerosols. Pergamon Press, New York.
1986. Methods for determining aerosol concentration. Aerosol Sci. Tech.
5(2): 123-143.
Gerritsen, J., and K. G. Porter. 1982. The role of surface chemistry in filter
feeding by zooplankton. Science 216: 1225-1227.
Harvell, C. D., and M. LaBarbera. 1985. Flexibility: a mechanism for control
of local velocities in hydroid colonies. Biol. Bull. 168: 312-320.
Holland, N. D., J. R. Strickler, and A. B. Leonard. 1986. Particle interception,
transport and rejection by the feather star Oligometra serripinna
(Echinodermata: Crinoidia), studied by frame analysis of videotapes. Mar.
Biol. 93: 111-126.
Johnson, A. S., and K. P. Sebens. 1993. Consequences of a flattened
morphology: effects of flow on feeding rates of the scleractinian coral
Meandrina meandrites. Mar. Ecol. Prog. Ser. 99: 99-114.
Jones, H. D., O. G. Richards, and T. A. Southern. 1992. Gill dimensions,
water pumping rate and body size in the mussel Mytilus edulis L. J. Exp.
Mar. Biol. Ecol. 155: 213-237.
J0rgensen, C. B. 1975. On gill function in the mussel Mytilus edulis L. Ophelia
13: 187-232.
1981a. A hydromechanical principle for particle retention in Mytilus edulis
and other ciliary suspension feeders. Mar. Biol. 61: 277-282.
1981b. Feeding and cleaning mechanisms in the suspension feeding bivalve
Mytilus edulis. Mar. Biol. 65: 159-163.
1982. Fluid mechanics of the mussel gill: the lateral cilia. Mar. Biol. 70: 275-
281.
1983. Fluid mechanical aspects of suspension feeding. Mar. Ecol. Prog. Ser.
11: 89-103.
1990. Bivalve filter feeding: hydrodynamics, bioenergetics, physiology, and
ecology. Olsen and Olsen Press. Fredensborg, Denmark.
J0rgensen, C. B., P. Famme, H. S. Kristensen, P. S. Larsen, F. Mohlenberg,
and H. U. Riisgard. 1986. The bivalve pump. Mar. Ecol. Prog. Ser. 34: 69-
77.
J0rgensen, C. B., T. Kiorboe, F. Mohlenberg, and H. U. Riisgard. 1984. Ciliary
and mucus-net filter feeding, with special reference to fluid mechanical
characteristics. Mar. Ecol. Prog. Ser. 15: 283-292.
J0rgensen, C. B., P. S. Larsen, F. Mohlenberg, and H. U. Riisgard. 1988. The
bivalve pump: properties and modelling. Mar. Ecol. Prog. Ser. 45: 205-
216.
LaBarbera, M. 1978. Particle capture by a Pacific brittle star: experimental test
of the aerosol suspension feeding model. Science 201: 1147-1149.
1984. Feeding currents and particle capture mechanisms in suspension
feeding animals. Am. Zool. 24: 71-84.
Lamb, H. 1962. Hydrodynamics. 6th ed. Cambridge University Press,
Cambridge.
Langmuir, I., and K. B. Blodgett. 1946. U.S. Army Air Forces Tech. Rept.
5418. U.S. Office of Technical Services. PB 27565.
Lasker, H. L. 1981. A comparison of the particulate feeding abilities of three

species of gorgonian soft coral. Mar. Ecol. Prog. Ser. 5: 61-67.
Leal, L. G. 1980. Particle motion in a viscous fluid. Annu. Rev. Fluid Mech.
12: 435-476.
current velocity on a passive suspension-feeder. J. Exp. Mar. Ecol. 127:
81-103.
Leonard, A. B., J. R. Strickler, and N. D. Holland. 1988. Effects of current
speed on filtration during suspension feeding in Oligometra serripinna
(Echinodermata: Crinoidea). Mar. Biol. 97: 111-112.
Licht, W. 1980. Air pollution control engineering: basic calculations for
particulate collection. Marcel Dekker, New York.
Loudon, C. 1990. Empirical test of filtering theory: particle capture by
rectangular-mesh nets. Limnol. Oceanogr. 35: 143-148.
Loudon, C, and D. N. Alstad. 1990. Theoretical mechanics of particle capture:
predictions for hydropsychid caddisfty distributional ecology. Am. Nat.
135: 360-381.
May, K. R., and R. Clifford. 1967. The impaction of aerosol particles on
cylinders, spheres, ribbons and discs. Ann. Occup. Hyg. 10: 83-95.
Mayer, D. L. 1979. Length and spacing of the tube feet in crinoids
(Echinodermata) and their role in suspension feeding. Mar. Biol. 51: 361-
369.
McFadden, C. S. 1986. Colony fission increases particle capture rates of a soft
coral: advantages of being a small colony. J. Exp. Mar. Biol. Ecol. 103: 1-
20.
Monger, B. C, and M. R. Landry. 1990. Direct interception feeding by marine
zooflagellates: the importance of surface and hydrodynamic forces. Mar.
Ecol. Prog. Ser. 65: 123-140.
Moore, H. J. 1971. The structure of the latero-frontal cirri on the gills of
certain lamellibranch molluscs and their role in suspension feeding. Mar.
Biol. 11: 23-27.
Morris, C. C, and D. Diebel. 1993. Flow rate and particle concentration within
the house of the pelagic tunicate Oikopleura vanhoefferi. Mar. Biol. 115:
445-452.
Nielsen, N. F., P. S. Larsen, H. U. Riisgard, and C. B. J0rgensen. 1993. Fluid
motion and particle retention in the gill of Mytilus edulis: video recordings
and numerical modelling. Mar. Biol. 116: 61-71.
Owen, G. 1974. Studies on the gill of Mytilus edulis: the eu-latero-frontal cirri.
Proc. R. Soc. London (B) 187: 83-91.
Parsons, T. R., M. Takahashi, and B. Hargrave. 1977. Biological
oceanographic processes. 2nd ed. Pergamon Press, Oxford.
Patterson, M. R. 1984. Patterns of whole coral prey capture in the octocoral
Alcyonium siderium. Biol. Bull. 167: 613-629.
1991a. The effects of flow on polyp-level prey capture in an octocoral,
Alcyonium siderium. Biol. Bull. 180: 93-102.
1991b. Passive suspension feeding by an octocoral in plankton patches:
empirical test of a mathematical model. Biol. Bull. 180: 81-92.
Pich, J. 1966. Theory of aerosol filtration by fibrous and membrane filters, p.
223-285. In C. N. Davies (ed.) Aerosol science. Academic Press, London.
Ranz, W. E., and J. B. Wong. 1952. Impaction of dust and smoke particles on
surface and body collectors. Ind. Eng. Chem. 44: 1371-1381.
Riisgard, H. U. 1988. Efficiency of particle retention and filtration rate in 6
species of northeast American bivalves. Mar. Ecol. Prog. Ser. 45: 217-
223.
Rubenstein, D. I., and M. A. R. Koehl. 1977. The mechanisms of filter feeding:
some theoretical considerations. Am. Nat. I l l : 981-994.
Schrijver, J. H. M., C. Vreeken, and J. A. Wesselingh. 1981. Deposition of
particles on a cylindrical collector. J. Coll. Interf. Sci. 81: 249-256.
Sebens, K. P., and M. A. R. Koehl. 1984. Predation on zooplankton by the
benthic anthozoans Alcyonium siderium (Alcyonacea) and Metridium
senile (Actiniaria) in the New England subtidal. Mar. Biol. 81: 255-271.
Shimeta, J. 1993. Diffusional encounter of submicrometer particles and small
cells by suspension feeders. Limnol. Oceanogr. 38: 456-465.
Shimeta, J., and P. A. Jumars. 1991. Physical mechanisms and rates of particle
capture by suspension feeders. Oceanogr. Mar. Biol. Ann. Rev. 29: 191
257.
Silvester, N. R. 1983. Some hydrodynamic aspects of filter feeding with
rectangular-mesh nets. J. Theor. Biol. 103: 265-286.
1988. Hydrodynamics of flow in Mytilus gills. J. Exp. Mar. Biol. Ecol. 120:
171-182.
Silvester, N. R., and M. A. Sleigh. 1984. Hydrodynamic aspects of particle
capture by Mytilus. J. Mar. Biol. Assoc. U.K. 65: 859-879.
Spielman, L. 1977. Particle capture from low-speed laminar flows. Annu. Rev.
Fluid Mech. 9: 297-319.
Spielman, L. A., and J. A. Fitzpatrick. 1973. Theory for particle collection
under London and gravity forces. J. Coll. Interf. Sci. 42: 607-623.
Trager, G. C, J.-S. Hwang, and J. R. Strickler. 1990. Barnacle suspension
Wildish, D. J., and D. D. Kristmanson. 1993. Hydrodynamic control of bivalve
filter feeders: a conceptual view, p. 299-324. In R. F. Dame (ed.) Bivalve
filter feeders in estuarine and coastal ecosystem processes. Springer-
Verlag, Berlin.
Yoshioka, N., and C. Kansoka. 1972. Collection efficiency of an aerosol by an
isolated cylinder: gravity and inertia predominant region. Chem. Eng.
(Japan) 36: 313-319.
6
Behavioral responses to flow
The life form of benthic suspension feeders includes both infauna and
epifauna, and as Table 6.1 suggests, the majority are epifauna. The
central problem for both sessile and tube-living epifauna is that they
must grow upwards into the benthic boundary layer to reach good qual-
ity, or higher quantities of, seston, both of which may be height or
velocity dependent (Muschenheim 1987a,b). Yet, these suspension feed-
ing epifauna may be restricted by the increasing velocity encountered
higher in the benthic boundary layer, because it increases drag and the
likelihood of dislodgement for them. By contrast, infauna are less sus-
ceptible to dislodgement, which does not occur unless a substantial part
of the sediment is swept away.
Behavioral responses to flow include rheotaxis, which is a directed
response to flow direction involving locomotion or muscular turning of
body parts, and rheokinesis, which is a non-directed response causing
random movement proportional to flow velocity. Rheotropisms are
debatably behavioral responses but fit the definition of behavior in
Carthy (1958); that is they involve flow-induced differential growth per-
ceived as twisting stresses by the axial support system, e.g. of gorgonian
corals which have semi-conductor properties that provide electrical
stimuli for skeletal secreting cells (Wainwright and Dillon 1969). The
result is greater skeletal growth on the stimulated side.
This chapter is organized around two general responses of suspension
feeders to hydrodynamics. The first involves only non-radially symmetri-
cal suspension feeders which respond behaviorally so as to position
themselves toflowin order to optimize filtration and, hence, growth. The
second is suspension feeding epifauna, which respond behaviorally to
flow in order to minimize lift and drag on their body parts. In addition to
this, we present a preliminary discussion of the phenomenon of aggrega-
213
Table 6.1. Life form categories of benthic suspension feeders.
Life form Examples

Infauna Some Polychaeta, e.g. Chaetopterus variopedatus
Some Bivalvia, e.g. soft shell clam Mya arenaria,
cockle Cardium edule
Phoronidea - Phoronopsis virida
Sessile epifauna Cnidaria: Hydrozoa - hydroids
Cnidaria: Anthozoa - gorgonian corals
Cnidaria: Antipatharia - octocorals
Bryozoa - bryozoans
Brachiopoda - lampshells
Echinodermata: Crinoidea - stalkless crinoids
Porifera - sponges
Ascidiacea - sea squirts
Cirripedia - barnacles
Tube-living epifauna Some Polychaeta and Crustacea (see Table 6.3)
Free-living epifauna Some Echinodermata, e.g. feather stars (Crinoidea),
sand dollars (Echinoidea)
Some decapod Crustacea, e.g. sand crabs (Hippidae),
porcelain crabs (Porcellanidae)
Some Bivalvia, e.g. scallops (Pectinidae), mussels
(Mytilidae)
tion among benthic suspension feeders, inclusive of a consideration of

the proximate causes of the behavior. Although only a few free-living
epifaunal suspension feeders have developed any swimming ability, we
have also considered this behavior in relation to flow.
Optimizing growth by orientation to flow

Infauna
Suspension feeding burrowing polychaetes and bivalves are intimately
connected with water movement because they depend on moving
seawater for distributing sexual products and for obtaining food, by
drawing it into either their burrows, e.g. Chaetopterus variopedatus, or
their siphon tubes, e.g. the soft-shell clam. They also depend on water
movement to replenish the otherwise seston-depleted seawater which
would build up around them in static conditions (Creutzberg 1975).
Many suspension-feeding infaunal polychaetes have radially symmetri-
Behavioral Responses to Flow 215
Flood
Figure 6.1 The clam My a arenaria: in life the valves are buried with only the tips
of the siphon reaching the sediment surface.
cal feeding structures, e.g. many sabellids, and so orientation with re-
spect to a particular current direction is not usually of concern. This is
not the case in bilaterally symmetrical infauna, e.g. bivalves, where indi-
vidual orientation to current direction may be of more importance.
It was noticed by Vincent, Desrosiers, and Gratton (1988) that the
deep-burrowing soft-shell clam Mya arenaria L. was oriented on tidal
mudflats in the Gulf of St. Lawrence such that the clam's sagittal plane
was perpendicular to the current direction (Fig. 6.1). Tidal currents over
the clam flats here were predominantly bidirectional, flood and ebb, and
it was hypothesized that this orientation was a means of optimizing
filtration. In any other orientation with respect to bidirectional currents,
some degree of exhalant mixing with the inhalant, and therefore seston
dilution, would occur, resulting in availability of a lower seston load for
filtration. No field evidence to support this view could be found by
Vincent et al. (1988) when individual orientation was related to growth
and reproductive measures in older clams, and it was suggested that this
was because natural selection was focused on younger individuals.

Whether the observed distribution of clams on the mudflats was due to
differential mortality rates after setting of inappropriately oriented indi-
viduals or to behavioral responses, at or after settlement, was not deter-
mined by Vincent et al. (1988).
Flume experiments undertaken by Monismith et al. (1990) in which a
model siphon pair of a clam connected to a pump and shear velocity and
bottom roughness closely simulated field conditions, provide results sup-
portive of the hypothesis of Vincent et al. (1988) that clam orientation
serves to optimize filtration/feeding. Model experiments show that
boundary layer velocity profiles, siphon height, siphon pair orientation,
and size and exit velocity of the siphons all affect the vertical distribution
of exhalant flow and the fraction of thisflowwhich is refiltered when the
inhalant is downstream. In an inhalant downstream model (dorsal-
ventral axis of clam) in which the inhalant diameter = 3.2mm and exha-
lant diameter = 6.4 mm, the amount of refiltration depends on the exha-
lant jet flow and height of the siphon above the sediment. When the jet
is strong and the siphon is flush with the sediment, little seawater is
refiltered because the jet is not deflected close to the siphon openings.
Conversely, when the jet is weak, such a deflection does occur, caused by
the cross-flow, and results in up to 35% refiltration. In similar conditions,
except that the siphon height is adjusted to be 10 mm above the sedi-
ment, much less of the exhalant is refiltered. Monismith et al. (1990)
concluded that the optimum conditions for clams to take advantage of bi-
directional tidalflowswas with the siphons oriented with their centres on
a line normal to the flow.
In further flume studies in which siphons of a bivalve reef were mod-
elled, O'Riordan, Monismith, and Koseff (1993, 1995) presented flume
data simulating either Tapes japonica or Potamocorbula amurensis from
San Francisco Bay. The experiment involved simulated siphons of each
bivalve jetting or withdrawing realistic amounts of seawater in a flume
boundary layer. At realistic siphon diameters for each species, simulated
by appropriately sized plastic tubing and at field observed densities of
bivalves, the amount of exhaled seawater which was refiltered could be
determined. For T. japonica at a density of 625 -rrf2 the maximum
refiltration ratio was 18%, and for P. amurensis at 11,025 -nT2, this value
reached 48%. The major factors influencing refiltration were bivalve
density and increases in relative shear velocity. Two lesser factors which
influenced refiltration were siphon height and cross-flow Reynolds
number (O'Riordan et al. 1995).
In growth experiments with the hard clam Mercenaria mercenaria in a

laboratory-induced oscillating flow, Turner and Miller (1991a) found
that the clam siphons were randomly oriented in the flume sediment,
thus not supporting the optimum orientation theory of Vincent et al.
(1988) and Monismith (1990). However, in Turner and Miller's (1991a)
work, the maximum growth period was 12 d and the hard clams bur-
rowed into the sediment while the flume was being filled, i.e. when no
directional cues were available to guide behavior. Field observations and
a more intensive experiment directed to individual hard clam orientation
in relationship to flow direction are required. Dolmer, Karlsson, and
Svane (1994) also found that the blue mussel Mytilus edulis L. did not
undertake rheotactic behavior. These experiments involved underwater
stereophotography of undisturbed mussels, as well as laboratory experi-
ments in a recirculating flume, and showed that foot and byssus thread
movements were unrelated to the current direction at velocities of 0-
20cm-s"1.
The phoronidean Phoronopsis psammophila from sublittoral sandy
sediments in the Gulf of Marseilles was studied by Emig and Becherini
(1970) in a recirculating raceway. When subjected to a unidirectional
flow, the lophophore, mouth, and surrounding cone of tentacles of this
phoronidean are turned to oppose the flow, and if the flow is reversed,
the lophophore orientation is also reversed. In this case, the response is
a rheotactic one, probably associated with the optimum orientation for
filter feeding, although feeding success was not determined. Johnson
(1988) studied another phoronidean, P. viridis, present on intertidal
sands of Bodega Bay, California. Phoronideans live in an infaunal tube
and, during feeding, extend the anterior part of their body, which carries
the terminal lophophore. The latter is a ciliated crown of tentacles which
creates an active feeding current when P. viridis is feeding in still
seawater. Both live animals and models were studied in a recirculating
flume at velocities measured 4 cm above bottom at up to lOcm-s"1. Since
the flume boundary layer was not characterized in this study, it was not
clear whether or not the measured velocities were in the free-stream
flow or within the flume benthic boundary layer. Flow around the
phoronideans was visualized with fluorescein dye and photography. It
was shown by Johnson (1988) that P. viridis models involved sediment
scouring around their anterior body, where it emerged from the sedi-
ment, causing entrainment upwards towards the lophophore because the
body was increasingly deformed in the downstream direction asflowwas
increased. As was pointed out, this is similar to the results for tube-living
FLARE
HOOD
Figure 6.2 Structure of aggregations of the upper part of Phragmatopoma

californica tubes (Thomas 1994); note thatflareand hood are ~7mm above the
sediment-water interface. End-on and sectional views of an individual tube on
right.
polychaetes found by Eckman, Nowell, and Jumars (1981) and Carey

(1983) discussed later in the section, Tube normal to flow. However,
Johnson (1988) did not demonstrate enhanced feeding in live P. viridis or
adjustment of its body to optimize feeding, as demonstrated by P.
psammophila (Emig and Becherini 1970).
The sabellarid polychaete Phragmatopoma calif ornica is an infaunal
species of Californian shores which extends its tentacles into the water
column for passive filter feeding. These worms occur in dense
aggregations at suitable locations, and Thomas (1994) noticed that P.
californica built two types of tube extensions, flares and hoods, which
protruded ~7mm above the sediment-water interface (Fig. 6.2). Field
observations suggested that tubes with flare extensions were associated
with wave-exposed, whereas hood tubes were associated with wave-
protected localities. The open face of the hood (Fig. 6.2) was shown by
field observations to face the dominant field current direction. Tube
extensions were also rapidly replaced after damage, suggesting their
importance to the worm, although repair was slower in the case of flare
extensions (Thomas 1994). Flume studies showed that hoods and flares
resulted in a decrease of fluid exchange with seawater above the colony;
thus the extensions were not linked to possible food depletion above the
worms. In flume experiments hood orientation with respect to unidirec-
tionalflowsor complete removal of the hood showed that the hood acted
to reduce tentacle deflection at higher flows when present at its normal
field orientation. Since deflection of filtering elements by flow in other
suspension feeders (Chap. 4) results in a reduction in filtering efficiency,
it is reasonable to suppose that the extensions assist the worms to filter
feed at higher limiting velocities. Further experiments involving filtration
or growth are necessary to test this hypothesis experimentally.
Sessile epifauna
The coral sea fans (Gorgonacea) are frequently variable in growth form
with a particular morphology dependent on water movement. Theodor
(1963) reported water movement-related morphology in the Mediterra-
nean gorgonian Eunicella stricta. This coral has a concave form in strong
wave surges, a fan shape in moderate flows, and a whip form in lesser
flows. Leversee (1976) reported two growth forms of the sea whip
Leptogorgia virgulata in North Carolina waters: a bushy form where
directionally unpredictable, turbulent currents were present, and fan-
shaped colonies where predictable, bidirectional currents were found.
The Caribbean reef coral Agaricia agaricites, studied by Helmuth and
Sebens (1993), grew in three forms: bifacial, horizontal unifacial, and
vertical unifacial (Fig. 6.3). These growth forms were related to water
movement, which was inversely related to depth. Thus, the bifacial form
was found at the greatest depths, where the water movement was least,
and appears to be locally adapted by opposing rather than being parallel
to the flow, as in the unifacial forms. Other authors who have studied
corals report that the fan is generally opposed, or perpendicular to, the
major bidirectional flow occurring at a particular field location.
Leversee (1976) was the first to demonstrate that the orientation of
fan-shaped gorgonian corals with respect toflowdirection was important
in determining the rates of prey capture in this passive suspension feeder.
Experimental tests were made in a 60-L recirculating flume in which the
fan-shaped form of Leptogorgia virgulata was presented to the unidirec-
tional flow (at a velocity of 4.0 0.05 cm-s"1), either opposed or parallel
to it. Artemia nauplii at a concentration of 20 3 L"1 were used as food,
and the rate of depletion of the nauplii monitored over a 3-h period. Of
FLOW
Horizontal unifacial Bifacial upright Vertical unifacial
Figure 6.3 Colony form of Agaricia agaricities in the Caribbean (Helmuth and
Sebens 1993). Amorphous encrusting forms may also be present, but are not
shown in the diagram.
six colonies tested, four showed significantly greater feeding when op-
posed versus when parallel to the flow. Similar results were obtained by
Helmuth and Sebens (1993) with the scleractinian coral Agaricia
agaricites in unidirectional flume flows of 3 and 6cm-s"1 and food sup-
plied as Artemia cysts at 1000 cysts-L"1. Results suggest that at 3cm-s~1,
significantly more cysts were captured by the opposed coral versus that
with the fan parallel to the flow. Although more than twice the number
of Artemia cysts were captured by opposed compared to parallel corals at
a speed of 6cm-s~1, this difference was not statistically significant. A
review of the shapes produced by growth of some passive suspension
feeders in different locations is presented by Warner (1977).
Helmuth and Sebens (1993) compared the feeding rate of three
growth forms of Agaricia agaricites found in Caribbean waters (Fig. 6.3).
Their results reveal that for a given unit area of feeding surface of this
colonial coral, more polyps were present in unifacial (horizontal
12.0 3.3, vertical 13.7 3.2 calices cm"2) than in bifacial (8.1 2.0
calices cm"2) forms, and that this led to higher feeding rates when ex-
pressed either on a per polyp or unit area basis. We have reanalyzed their
results in terms of additional data (total number of cysts captured for the
whole colony) supplied by Helmuth (pers. commun. 1995). The addi-
tional results (Table 6.2) suggest that the bifacial upright colony feeds
best at lower ambient velocities with a nominal feeding peak -IScm-s"1,
compared to the vertical unifacial colonies, which feed best at higher
velocities and with a peak at >30cm*s~1. The latter results are consistent
with the known field distribution of the water movement-related
Table 6.2. Food capture rates of Agaricia agaricites.
Bifacial Horizontal unifacial Vertical unifacial
Whole Whole Whole

colony colony colony
Velocity no. of no. of no. of
cm s"1 Cysts/cm2 cysts Cysts/cm2 cysts Cysts/cm2 cysts
3 2.5 171 6.5 284 3.0 144

6 3.0 242 11.0 371 5.0 224
18 8.0 508 15.0 514 15.0 615
30 3.0 293 15.0 623 38.0 1320
50 1.0 87 16.0 739 6.0 297
Results are adjusted to the same ambient cyst concentration, 1 cyst-ml \

Source: Estimates from (1) mean, unit surface area values as given in Fig. 4 of
Helmuth and Sebens (1993), and (2) total number of cysts captured per
colony, based on unpublished results (Helmuth, pers. commun.).
morphotypes: bifacial, deep, low velocity or unifacial, shallow, higher

velocity in Caribbean waters down to 20 m (Helmuth and Sebens 1993).
Similar to the conclusions reached by Helmuth and Sebens (1993) based
on unit area, analysis based on whole colony feeding suggests that the
bifacial colony performs better at intermediate velocities. At IScm-s"1,
the bifacial colony can feed with efficiency equal to that of the horizontal
unifacial colony and nearly as well as the vertical unifacial form.
The hypothesis tested by Abelson, Miloh, and Loya (1993) was that
the shape of passive suspension feeders was critical in determining the
quality and quantity of sestonic particles available as food at different
levels within the benthic boundary layer. Two general categories, based
on the height to width ratios (= slenderness ratio [SR]), were recognized
by Abelson et al. (1993): high SR, high branching corals, and low SR,
with flattened or massive corals or sponges. High SR corals protrude to
a greater height within the benthic boundary layer than low SR corals.
The latter encounter increased seston concentration due to wake
resuspension from the sediment. Because low SR corals encounter high
fluxes of bedload material, they may be specialized in this type of feed-
ing. On the basis of a hydrodynamic model constructed during this study,
sestonic particle concentration and particle settling velocity were deter-
mined around model passive suspension feeders placed in the field and
these measurements supported the model predictions.
A B C D
Figure 6.4 Orientation of a sea fan to ambient flow from left to right: A, fan
blade parallel to theflow;B, blade at a low angle; C, blade at a larger angle to the
flow; D, blade perpendicular to the flow; B and C, unequal twisting stresses,
indicated by filled arrows, are present (Wainwright and Dillon 1969).
How then can these growth patterns be stimulated by flow? Wain-

wright and Dillon (1969) studied two sea fans, Gorgonia ventalina and G.
flabellum, which lived in patches at depths of 2-10 m on the seaward reefs
of the Florida Keys, United States. Like the authors mentioned earlier,
they found that the majority of larger fans were oriented in the same
plane (although how this was related to water movements was not deter-
mined). They observed a large number of fans of different size - from 10
to 90 cm in height - and showed that, whereas the smallest colonies were
randomly oriented, the largest ones were aligned in the same plane.
Wainwright and Dillon (1969) suggested that as they grow in size and
become more influenced by the greater velocities that they reach higher
in the benthic boundary layer, twisting (= torque) stresses (Fig. 6.4) due
to bidirectional flows are used by the colony to bring about colony
orientation opposed to this flow. Twisting stresses occur when an un-
equal area on one side of the colony is presented to the flow, causing
greater force to be applied there. If the colony was seeking to minimize
drag, then Fig. 6.4A would be the optimum orientation strategy, but,
since D in Fig. 6.4 is the morphology realized in larger sea fans, then drag
must normally be of little concern. Both A and D in Fig. 6.4 are stable
orientations for which unidirectional flows can cause no colony twisting.
The flexible sea fan blade of Gorgonia consists of a vertical, thickened
axial support that extends from a holdfast to the blade. The stiffened
axial support is thicker at its base and tapers to the top of the blade. As
a result of twisting forces in B and C (Fig. 6.4), the axial support displays
creep. As a result of the secretion of new skeletal material around that

already present, the sea fan is held in its new position. Evidence of
changes in fan orientation which occurred during growth were seen by
Wainwright and Dillon (1969) on sectioning the blade. The axial fan of
gorgonian sea fans consists of calcite and fibrous gorgonin containing a
collagenlike material, probably supplemented by soluble organic matter.
These materials have piezoelectric and semi-conductor properties that
can provide electrical stimuli to cells to cause them to secrete the skeletal
supporting material.
Brachiopods from Washington on the U.S. Pacific coast were used in
laboratory experiments in 15-L recirculating water tunnels by LaBarbera
(1977). Experiments with orientation of dead shells in flow and tubing
carrying a dye which was released from a model lophophore within the
shell confirmed theflowpathways of ambient seawater shown in Fig. 6.5.
In life, these flows are generated by ciliary pumping of the lophophore
with seawater being drawn in via the right and left lateral margins and
exiting at the anterior shell gape. Of the four orientations shown in Fig.
6.5, D is the hydrodynamically optimum position - that is, when the
hinge axis is opposed to the flow direction. With this orientation bidirec-
tional flows, which are a common occurrence in the field, can be accom-
modated because the ambient flow simply changes from the left to the
right lateral margin. Some brachiopods are able to adjust their position
with respect to flow direction, and the rheotactic behavior observed may
be linked to ambient flow enhancement of filtration, as proposed by
LaBarbera (1977). Alternatively, we suggest that brachiopods do not
experience ambient flow enhancement and may respond to dynamic
pressure differences between inhalant and exhalant, as do scallops (see
the sections, Bivalves, and Lamphells, Chap. 4). According to this view,
dynamic pressure differences between inhalant and exhalant cause them
either to reduce pumping, to work harder to effect the same rate, or to
change their position behaviorally with respect to flow - all in order to
optimize feeding rates.
LaBarbera (1977) found that of two species of articulate brachiopods
he could observe in the field off San Juan Island, Washington, both
preferred the D orientation with approximately half facing one way and
the other half facing the opposite way, consistent with the flows at the
site being bidirectional. Experiments with live brachiopods placed in the
water tunnel so that the initial orientations were either as A or as B in
Fig. 6.5, demonstrate that two species, Laqueus californianus and
Terebratulina unquicula, actively reoriented to position D. In the proc-
Side Top
Figure 6.5 Possible orientations indicated by A, B, C, or D of a brachiopod

relative to ambient flow as seen from the side and top of the animal: small arrows,
activity by the ciliary pump; heavy arrows, ambient flow with the length suggest-
ing the velocity (redrawn from LaBarbera 1977).
ess, both could traverse an arc as great as 120 to achieve their preferred
orientation. Although the mechanism was not studied by LaBarbera, he
suggested that the pedicle adjustor muscles were involved although the
sensory organs and nervous pathways were not identified. In L.
califbrnianus, movement of the pedicle was preceded by three to five
coughs (a rapid sharp closure followed by a full valve opening), whereas
the rheotactic behavior of T. unquicula did not include coughing but
rapid rotation through 60 to 90 with the valves closed or slightly
opened or with the valves open and a slower rotation, about 1 s per 30
through a final arc of 20 to 40. One other species, Hemithryris psittacea,
showed similar rheotactic behavior but could only adjust through 45,
whereas a fourth species, Terebratulina transversa, never showed read-

justments of body positions during the flume experiments. So possibly a
non-behavioral mechanism is involved in achieving the preferred D po-
sition with respect to field flow in T. transversa.
Threshold current speeds for rheotaxis in L. californianus (-1.5 cm s"1)
and T. unquicula (-0.75 cm -s"1) were similar to the excurrent velocity
measured with a thermistorflowmeter, suggesting that dynamic pressure
is the factor responded to in this process. Eshelman and Wilkens (1979)
observed that T. transversa in Pacific waters of the British Columbia,
Canada, coast were -70% in position D and -30% in position C, in line
with LaBarbera's (1977) preferred orientation hypothesis. T. transversa
was absent from sites without a current flow or with currents that were
too extreme ^lSScm-s" 1 ).
In a study of the mechanism of brachiopod orientation to ambient
flows (LaBarbera 1978), active rheotactic behavior was considered to be
achieved by muscular activity. A structural study of the pedicle muscula-
ture suggested that the dorsal adjustors were primarily involved in gen-
erating torque around the long axis of the pedicle to resist the torque
applied by the ambient flows. The latter was measured in model
brachiopods at flows up to 14cm-s"1. LaBarbera (1978) found that the
torque around the pedicle was low and that the dorsal pedicle adjustor
muscles were easily capable of resisting this force, although sufficient
data were not obtained to estimate the energetic cost of this behavior.
A common North Atlantic barnacle, Semibalanus balanoides, prefer-
entially settles infinegrooves on hard substrates and orients positively to
light. Immediately after settlement and metamorphosis, orientation may
be changed by rheotropic response (Crisp and Stubbings 1957) where
flows are predominantly unidirectional. This brings the rostrum to op-
pose the major flow direction and involves torsional growth resulting
from mechanical strain unequally distributed on the barnacles' body.
Crisp (1960a) showed that S. balanoides oriented with the rostrum to the
flow grew better in wet weight than those in which the orientation was
experimentally varied by 90 or 180. A whale barnacle, Cornula
diadema, is present as an ectocommensal with the rostrum opposed to
the flow as the whale swims forward (Crisp and Stubbings 1957) similar
to the orientation of S. balanoides. According to Crisp (1960b), S.
balanoides can move up to 2 cm away from the original settlement site
under direct competitive pressure from faster growing, larger barnacles
of the same or different species. Trager, Hwang, and Strickler (1990) and
Trager et al. (1992) have shown that flow provides important cues in the
Semibalanus
balanoides =>
Elminius
modestus =>
Light
t
Balanus
improvisus =>
Balanus
amphitrite
Balanus
trigonus
=>
orientation of Direction Direction of

cyprid of rotation cirral beat
Figure 6.6 Diagrammatic representation of cyprid orientation, direction of ro-

tation, and direction of adult cirral beat forfivespecies of barnacles (based on
Crisp 1953; Ayling 1976).
feeding behavior of barnacles, such as S. balanoides (see the section,

Barnacles, Chap. 4). This species switches from active to passive feeding
at -3.1 cm-s"1, and in oscillating flows, seawater deceleration is the cue
for switching the cirral fan orientation through 180 so that the barnacle
can anticipate the maximum wave surge energy on the concave surface of
the fan during passive feeding.
The common barnacle of New Zealand, Balanus trigonus, shows some
important differences from S. balanoides. Whereas the latter is a
facultatively passive/active suspension feeder, B. trigonus is an obligate
active suspension feeder with cirral beats of 2-3 beats -s"1, whatever the
ambient flow velocity (Ayling 1976). At low flows, B. trigonus cirral
beating is opposed to the flow, but at higher flows (~30cm-s~1), the cirri
are reversed and beat in the same direction as the flow. A further
difference is that the orientation of the body, defined by the rostrocarinal
axis, is initially at right angles to the flow, whereas in S. balanoides, the
axis is aligned in the same direction as the flow. Thus, in B. trigonus the
cirral fan must swivel through only 90 (Fig. 6.6) to oppose flows from
either direction. For S. balanoides in oscillatingflows,the cirral fan must
be swivelled through 180 to oppose the flow when it approaches the
barnacle from the carina end. There appear to be no comparative growth
studies for B. trigonus at different orientations to prove the contention of
Ay ling (1976) that the preferred orientation toflowallows this barnacle
to minimize energetic costs involved in suspension feeding.
Otway and Underwood (1987), working with an Australian intertidal
barnacle, Tesseropora rosea, showed that its preferred orientation was
with the rostrum and cirral fan opposed to the main flow direction, as in
S. balanoides. Major flow in this habitat was the wave backwash. In
growth experiments on plates cut out of the solid rock, the plates were
rotated through 90 or 180 and compared to controls which were undis-
turbed or prepared exactly as the treatments. Results showed that signifi-
cant depressions of wet weight occurred in rotated plates in 15- to 19-mo-
long experiments, but there was no effect on the rate of shell growth,
mortality, or egg mass production.
Preliminary descriptions of the behavior of stalked barnacles Trilasmis
fissum hawaiense (family Poecilasmatidae), which are epizoic ecto-
commensals on the mouth parts of spiny lobsters from Hawaii, have been
given by Bowers (1968). The behavior includes movement of the
cirral net and other movements of the peduncle in response to water
movements.
Tube-living epifauna
Macrofauna which build tubes that project into the benthic boundary
layer are usually, although not exclusively, suspension feeders. On the
basis of their adult tube dimensions, H, height above the sediment-water
interface, and tube diameter, Z), opposing the majorflowvector, the ratio
H/D (using consistent units for H and D) can be used to characterize the
main types (Table 6.3) as follows:
Tube normal to flow H/D = 2-10

Truncated cone tube H/D < 2
Spar buoy tube H/D > 20
In addition, some tubes have one, or both, open ends opposed to bidirec-
tional flow; the ratio H/D is not applicable to them. The spar buoy tube
is so named because it is articulated near its base and can passively move
Table 6.3. Tube height, H, cm, above the sediment-water interface and maximum diameter, Y>, cm.a
Tube dimensions
Species Higher taxa H D H/D Reference
Tube normal to flow
Lanice conchilega Polychaeta, Terebellidae 2.0 0.5 4 Carey (1983)
Pseudopolydora paucibranchiata Polychaeta, Spionidae 3.5 0.5 7 Levin (1982)
Streblospio benedicti Polychaeta, Spionidae 3.5 0.5 7 Levin (1982)
Eudistylia vancouveri Polychaeta, Sabellidae 12.5 1.7 7 Merz (1984)
Ampelisca abdita Amphipoda, Ampeliscidae 0.5-1.0 0.2-0.3 3-5 Mills (1967)
Ampelisca vadorum Amphipoda, Ampeliscidae 0.4-1.0 0.2-0.5 2-5 Mills (1967)
Haploops fundiensis Amphipoda, Ampeliscidae 1.4 0.5 3 Wildish and Lobsiger (1987)
Tube opposed to flow
Net-spinning caddis larvae Trichoptera, Hydropsychidae Edington (1968)
Ampithoe valida Amphipoda, Ampithoidea Borowosky (1983)
Tanais cavolinni Tanaidacea Borowosky (1983)
Truncated cone tube
Spio setosa Polychaeta, Spionidae 5.0 5.0 ~1 Muschenheim (1987a)
Fabricia limnicola Polychaeta, Sabellidae 0.6 0.4 ~1 Levin (1982)
Mesochaetopterus Sagittarius Polychaeta, Chaetopteridae 0.9 0.6 -1 Bailey-Brock (1979)
Phyllochaetopterus verrilli Polychaeta, Chaetopteridae 0.9 0.6 ~1 Bailey-Brock (1970)
Spar buoy tube
Potamilla neglecta Polychaeta, Sabellidae 1.5-2.5 0.08 19-31 Wildish and Lobsiger (1987)
Complex tube shape
Diopatra cuprea Polychaeta, Onuphidae 2-5 0.5-0.8 2-10 Myers (1972)
a
In some cases, tube dimensions were estimated from the original figures.
in whatever direction the flow dictates. In this respect, it is like the

navigational buoy of this name, which is common where a considerable
tidal prism is available.
Anyone who has watched snow being scoured around the base of a
tree, or seawater flow scouring sediments at the base of a pier piling, is
familiar with the general effects to be expected from flow around animal
tubes. Isolated tubes which project into the logarithmic part of the
benthic boundary layer cause momentum transfer, sediment deposition,
and sediment entrainment. A theoretical and experimental flume study
using a rigid circular cylinder as a tube mimic (Eckman and Nowell 1984)
provides a good basis for those considering flume experiments with live
tube-living epifauna. The range of tube dimensions tested by Eckman
and Nowell (1984) was H/D^ 1.0-6.7, so their results are applicable to
most species listed in Table 6.3, although not to the thin tube of
Potamilla neglecta. Boundary shear stresses at tube height, which vary
with ambient velocities, and to a lesser extent tube diameter are the
critical factors involved in the small-scale geophysical effects seen. An
interesting prediction from the Eckman and Nowell model is that if the
tube is squat and H < 100 times the viscous sublayer thickness, a net
deposition of sediment may occur if the boundary shear stresses are low
enough to prevent scour in the downstream vortex.
Tube normal to flow

Laboratory studies in a recirculating flume by Eckman et al. (1981) with
the polychaete Owenia fusiformis collected from False Bay, Washington,
showed that isolated tubes caused sediment scouring around their base,
which was assisted by tube bending. The scoured pits became traps for
bedload transport on which Owenia could feed. In a flume study of
Lanice conchilega collected from the Tay estuary, Scotland, Carey (1983)
showed that the tube wakes (Fig. 6.7) resuspended material in such a way
that the low density particles were available for downstream tentacle
capture. The tube top was extended into two semi-circular plates which
were arranged normal to the major flow direction and supported a
branching fringe constructed of sand particles which was presumed to
support the tentacular crown of Lanice during feeding. It is likely that
these structures have importance in protecting and regulating the local
flow environment near the tentacles.
Tube reconstruction of the sabellid polychaete Sabellastarte magnifica
has been observed in the laboratory after removal of the worm from its
old tube (Fitzsimons 1965). Beginning with the worm horizontal on the
Figure 6.7 Lanice conchilega tubes; flow patterns and sediment scouring near
the tube are based on time-lapse photography of egg white particles (Carey
1983); A, side elevation; B, from downstream; large arrow, flow direction.
substrate, the tube is eventually bent up at 90to the substrate so that the
finished tube is normal to ambientflows.Tube building occurs at the rate
of 3-4 mm-d"1 in laboratory conditions.
No laboratory flume studies have yet been made of ampeliscid
amphipods (Fig. 6.8), which have a tube with an elliptical cross section.
The study by Eckman and Nowell (1984) considered only rigid right
circular cylinders. Enequist (1949) and Mills (1967) have described
ampeliscid tube building and feeding, aided by observations in aquaria
Antenna current
Opening Food
accumulates
Approximate_
Pleopod current
sediment level
Membranous
area
Figure 6.8 Ampelisca abdita amphipod: position adopted for feeding and com-
plete tube (Mills 1967).
but without simulation of water movement. Tube building began after a

freshly disturbed Ampelisca abdita had begun to burrow into the
substrate, when it turned ventral side up and began secreting threads
from glandular tissue on peraeopods 1 and 2 through ducts on the dactyls
(Mills 1967). The threads were stretched around the body by the
urosome and uropods and covered by sand particles placed there by the
gnathopods. The inner lining of the finished tube consisted of smooth,
whitish glandular secretion to which fine particles were stuck on the
outside of the tube. It was noticed by Mills (1967) that both A. abdita and
A. vadorum tubes were oriented so that the widest part of the tube
opposed the major flow direction on intertidal sands. Tube orientation of
Haploops fundiensis found at 80-m depth in the Bay of Fundy brought up
in undisturbed grabs or indicated in underwater photographs, showed
that the tubes were all oriented in the same plane and that the flat surface
opposed the flow (Wildish and Lobsiger 1987). The elliptical tube section
of H. fundiensis, 0.5 cm wide x 0.2 cm thick, has openings at the top and
bottom and is quite flexible. No flows are induced in the tube, and the
end inserted in the substrate is plugged with sediment.
During feeding, ampeliscid amphipods sit ventral side up at the top of
their tube (Fig. 6.8), with the body along the long axis of it, braced by the
posterior peraeopods and with the first two pairs of peraeopod dactyls
hooked over the tube lip. If the animal moves lower into the tube, the lips
close after it, thus preventing outside access. Three feeding methods are
possible (Enequist 1949; Mills 1967): passive filtration, holding long
setose first and second antennae into the flow; whirling, movements of
both second antennae beating in synchrony, which creates a flow across
the gnathopods where particles are filtered, with the feeding flow aug-
mented by pleopod beating, which also drives the flow from anterior to
posterior; and deposit feeding, which is possible when the long second
antennae are scraped across the sediment surface. The specificity of
these feeding methods is not clear, although Mills (1967) considers A
vadorum and A. abdita as mainly "whirling" feeders, whereas Haploops
fundiensis must be a passive filterer or whirler since the second antennae
are not long enough to reach the sediment for deposit feeding from its
position in its tube.
Tubes opposed to the flow

Although some freshwater organisms may orient their feeding structures
to take advantage of ambient flow in passive filtration, e.g. the net
spinning caddis larvae (Table 6.3), whose pocket-shaped nets on stones
are opposed to freshwater stream flow, we know of no marine tube
builders which oppose the flow to filter passively. Perhaps the amphipod
Ampithoe valida and tanaid Tanais cavolinii, which both have open-
ended tubes, oriented horizontally on the substrate surface (Borowsky
1983), are examples of this type.
Truncated cone tubes

Short, squat tubes are built by cementing particles to the rim of the tubes,
which grow upwards as sedimentation proceeds. Polychaetes with trun-
cated cone tubes (Table 6.3) include Spio setosa, which have been stud-
ied by Muschenheim (1987a, b) in unidirectional laboratoryflumeflows.
In nature, tubes are often broken off to the sediment surface level by
wind-wave effects. By comparing the suspension feeding of broken tube
worms, feeding at 0-2 cm above the bed, with that of intact tube worms
feeding at 4-6 cm above the bed, Muschenheim (1987a) showed that the
latter were feeding on better quality food, as evidenced by a greater
Table 6.4. Effect of ambient velocity on helical coiling in

Pseudopolydora kempi japonica palps.a
Velocity (cm s"1) Number of observations Median coils per palp

5 10 1.00
12 8 1.25
19 6 2.25
26 9 2.00
a
Velocities are measured at palp height.
Source: Data from Taghon et al. (1980).
proportion of enriched organic mineral aggregates in gut contents.

Spionid polychaetes are considered to be deposit-suspension feeders,
using a pair of long palps for feeding. Taghon et al. (1980) conducted
flume studies with a spionid, Pseudopolydora kempi japonica, over a
unidirectional velocity range of 0 to ~16cm-s"1, measured at 0.4 cm
above the bed. Behavior of the worms was observed under low power
magnification. As ambient velocity increased, the proportion of worms
deposit feeding steadily decreased. Worms suspension feeding held
their palps in a downstream helical coil with the number of turns increas-
ing with ambient velocity (Table 6.4). The rheotactic behavior of palp
coiling is directly proportional to velocity - at least in the range up to
20cm-s~1 - and can be shown to increase the surface area of the tentacle
facing theflow.Further experiments with narcotized worms showed that
the behavior was absent when the muscular and nervous systems were
inactivated, suggesting that a coordinated behavioral response was
involved.
Turner and Miller (1991b) have provided a laboratory flume study of
the chaetopterid polychaete Spiochaetopterus oculatus subjected to oscil-
latory flow. The water tunnel used permitted maximum oscillatory flows
of 5-40 cm-s"1 with amplitudes of 8-32 cm and periods of 4-25 s. This
chaetopterid is a facultative deposit-suspension feeder like the spionids
mentioned. It has a pair of long palps which touch the sediment during
deposit feeding and transfer the particles collected along the ciliated
groove to the mouth. During suspension feeding, the palps become
helically coiled, the degree of coiling increasing with higher ambient
velocities. Exposure to oscillatingflowsshowed that the worm rotated the
palps at flow reversal so as to keep the ciliated food groove on the palps
opposed to the flow. The time spent suspension feeding was related to
ambient velocity so that it predominated at maximum flows >15cm-s"1.
Deposit feeding predominated at maximum flows <10cm*s~1, and a
change to suspension feeding appeared to be related to the inability to
keep the palp tip on the sediment surface at velocities >10cm-s~1. The
relevant hydrodynamic variable to characterize the transition could be
the [bed] sediment shear stress. In Barnes's (1964) study of the same
species, the worms also spent time building their tubes, which involved
withdrawing their palps and exposing the fourth parapodia for this pur-
pose or tube cutting as velocities increased by means of the fourth setae.
Tube building in Spiochaetopterus was maximal in motionless seawater,
and within 15 min of initiating oscillating flow, tube cutting also started
(Barnes 1964; Turner and Miller 1991b). Reducing the height of the tube
is presumably related to adjusting the worm's feeding ambit to one where
good quality seston is available.
Spar buoy tubes

A sabellid polychaete, probably Potamilla neglecta, was the first benthic
suspension feeder recognized as belonging to the group of spar buoy
tubes. It was studied by underwater time-lapse photography at an 80-m-
deep station in the Bay of Fundy (Wildish and Lobsiger 1987). Because
these sabellids have long, very thin tubes, conventional grabs are inad-
equate for sampling and underestimates their density. Time-lapse pho-
tography (10- to 12-min intervals) over a 3-d period showed that the tube
positions passively followed the current direction indicated by a current
meter attached to the camera frame. This suggests that the tubes are
articulated near their base and that the conical, branchial crowns are
always able to feed in the same downstream position, whatever the flow
direction.
The active suspension feeding sabellid polychaete Sabella penicillus
studied experimentally by Riisgard and Ivarsson (1990) may also have a
spar buoy tube, although it has a much greater tube height (^7.5 cm)
than Potamilla spp. Sabella has a crown-filament pump in which the
laterofrontal cilia on pinnules borne from the filaments represent the
active pump.
Complex tube shapes

The polychaete Diopatra cuprea, which is found intertidally and
subtidally from Cape Cod to Florida, has an inverted J-shaped tube
(Myers 1972). Its tube extends 2 (1-5 range) cm above the sediment
surface, depending on local sediment erosion/deposition near the tube.
Diopatra can reduce the tube height by cutting if erosion is marked and
build the tube upwards if considerable deposition of sediment occurs.
Where currents are high, greater worm densities are present and the
worms can orient the tubes behaviorally to within a 90 arc, so that
the long axis of the J, seen from above, is at right angles to theflow.The
worms may also rebuild the tubes so that a better orientation to flow is
obtained (Myers 1972). Although it was suspected that the tube orienta-
tion was connected with feeding, no experimental results confirming this
were presented.
A flume study of Diopatra cuprea was undertaken by Luckenbach
(1986) to determine the erodibility of sediments affected by tubes in flow
or by associated deposit-feeding macrofauna. During these experiments,
it was not stated how the tubes were oriented with respect to the unidi-
rectional flume flow. Those sediments with high macrofaunal densities,
with or without Diopatra tubes, were more erodible than consolidated
sediments with fewer macrofauna, suggesting that the tubes were unim-
portant in sediment stabilization, at least at friction velocity values em-
ployed during these experiments (Luckenbach 1986).
Free-living epifauna
The scallops (Pectinidae) are mobile epifauna with some swimming ca-
pabilities, which are more fully discussed later in the section, Swimming.
Because their body plan is bilaterally symmetrical with inhalant and
exhalant openings at specific positions on the body (Fig. 6.9) - as in the
brachiopods observed earlier - it is plausible to propose that rheotaxis
can optimize their orientation with respect to ambient flows.
Observations by SCUBA of natural populations of Pecten maximus in
Strongford Loch, Northern Ireland (Hartnoll 1967), showed that the
majority of scallops opposed the predominantly unidirectional flow
(59% of individuals at 45 to 135; see Fig. 6.9C). Mathers (1976) also
observed P. maximus on the west coast of Ireland, at two locations where
the tidal flows reversed on flood and ebb tides. Here it was found that
approximately half of the scallops opposed the ebb and half opposed the
flood currents with the ventral margin. The digestive histology of flood
and ebb subpopulations showed that digestion was offset by 6h, suggest-
ing that each group fed on either the flood or ebb tide only, but not on
resillum
ANTERIOR
270 c
Figure 6.9 Scallop valve orientation and morphology: A, left upper valve from
above; B, right, lower valve, inside view; small arrows, inhalant; large arrow,
exhalant; C, valve orientation with respect toflowdirection (large arrow).
both (Mathers 1976). Giant scallop (Placopecten magellanicus) beds in

the Gulf of St. Lawrence studied by Caddy (1968) had most individual
scallops opposing the largely unidirectional flow.
Evidence that scallop orientation toflowdoes affect filtration rate, and
hence feeding and growth, is provided by Wildish et al. (1987) and
Wildish and Saulnier (1992). Experiments with scallops glued in place in
a Blazka respirometer at three flow velocities (Table 6.5) showed that
giant scallops with their exhalant opening directly facing the ambient
flow at 225 had significantly lower seston uptake rates at both 10 and
18 cm s"1 than in the other treatments. Filtration rates for all orientations
at Slcm-s"1 were significantly less than at lOcm-s"1. Similar results are
obtained if giant scallop growth is measured. Multiple-channel flume
growth experiments (Table 6.6), in which four concurrent treatments are
possible with natural seston concentrations which are the same in each
channel, show that at high velocities (22.5 cm s"1) growth is significantly
Table 6.5. Giant scallop seston uptake rates, jigChla -gwet^h1, mean
of two replicate experiments at each treatment?
Velocity, cm-s 1
Scallop 10 18 31
orientation
(see Fig. 6.9C) ng-g'-h-1 % ng-g'-h-1 % Hg-g'-h-
1
%
45 0.77 100 0.78 101 0.24 31

135 0.96 100 0.76 80 0.49 51
225 0.54 100 0.22 51 0.07 13
315 0.77 100 1.44 188 0.53 70
Conditions: temperature = 9.5 0.2C (SE); initial food concentration = 104

cells-ml"1.
Source. Wildish et al. (1987).
Table 6.6. Multiple-channel flume growth experiments in which

individual giant scallops were glued to the flume floor with the
orientation shown?
Treatment Percentage daily growthb SE
Speed (cm s'1) Scallop orientation () Valve height (h) Wet weight (W)
22.5 225 0.103 0.007 0.263 0.020

22.5 45 0.125 0.004 0.344 0.015
3.6 225 0.149 0.005 0.321 + 0.011
3.6 45 0.135 0.006 0.336 0.014
a
Conditions: 28-d duration starting on November 5, 1986, mean
temperature = 7.6C with range of 6.1-9.5; mean Chla = 0.65 (range 0.48-
0.99) ug/L, 30 juvenile giant scallops per treatment.
b
100 (hx - hoyhoN, 100 (Wt - W0)/WyV, where N = time in days.
Source: Wildish and Saulnier (1992).
limited when the exhalant faces the flow (225), but this effect is absent
at low ambient velocities (3.6 cm-s"1) for the same orientation.
Behavioral cues which must be involved in scallop orientation
behavior are not well known, although rheotaxis may be involved. The
ability of scallops to spin around by directing a jet of water from only one
side of the animal (Caddy 1968) may help to explain how it is done. An
alternative hypothesis that scallop orientation toflowis a passive process
brought about by ambient water movement has been proposed by Grant

et al. (1993).
Natural populations of feather stars (Echinodermata, Crinoidea)
found on the Great Barrier Reef were observed by time-lapse cinema-
tography (Meyer et al. 1984). It was observed that the current regime
caused changes in body posture (rheotaxis). Feather stars such as
Comanthus bennetti and Himerometra robustipinna were able to crawl
for periods of ~10min at a very slow speed (1 arm length min"1) to
achieve a favoured position for passive filter feeding. The body position
adopted at different flows was predominantly fan-shaped when flows
were maximal, but transitional, meridional, and hidden positions were
also seen. No experimental studies related the body position to feeding
success, however, and the exact purpose of the rheotaxic behavior thus
remains conjectural.
The feeding behavior of porcelain crabs - Petrolisthes leptochelys from
the Red Sea and P. oschimai from the Indian Ocean - was shown by
Trager et al. (1992) to be controlled by oscillating flows. Setae on the
third maxillipeds are used by porcelain crabs as afilter.In very low flows,
the paired third maxillipeds are rapidly swept through seawater, but, at
higherflows,they are used passively much as they are in other facultative
active suspension feeders such as barnacles. Also as in barnacles, it is the
concave face of the filter presented to the flow. In low oscillating flows
(0.16 Hz), both third maxillipeds acted in concert during passive feeding.
At higher oscillating flows (>0.58Hz), the third maxillipeds were ex-
tended alternately so the left and right sampled the flow on alternate
pulses (Trager et al. 1992).
Avoiding flow dislodgement

Macrobenthic suspension feeders can be dislodged by wind-wave action
and tidal flows. Experimental or observational studies which seek to
understand the mechanisms and behavioral responses which are con-
cerned with minimizing dislodgement are considered here.
Sessile epifauna
The greatest hydrodynamic forces exerted on macrobenthos occur on
intertidal rocky shores. The reef coral Acropora reticulata was shown by
Vosburgh (1977) to be limited to depths >7.7m in Eniwetak Atoll, where
it escaped the most violent breaking waves. Drag and mechanical tests
Table 6.7. Force coefficients for intertidal rocky shore suspension-

feeding organisms.^
Species Re cD cL CM
5
Semibalanus cariosus 10 0.52 N/A 1.31
Mytilus californianus - end-on 105 0.20 1.20
- broadside 105 0.80 2.00
Models
Sphere 105 0.47 0.4 1.67
Cylinder 105 0.73 N/A 2.00
a
CD = drag coefficient, CL = lift coefficient, CM = inertia coefficient. (See the
section, Free-living epifauna).
Source: Denny (1985).
were performed in a tow tank on air-dried specimens to determine the

forces required for coral breakage. The extreme hydrodynamic forces
predicted at these depths on the windward side of Eniwetak Atoll were
less than the critical values for coral breakage.
Denny (1982, 1985) constructed a wave force telemetry system suit-
able for field operations and made observations on individual wave
forces through many tidal cycles at one exposed location on the coast of
Washington. Denny (1985) found that 2- to 4-m-breaking-height waves
at this location produced maximum velocities of at least 8m-s 4 and
accelerations of at least 400 m s"2 near the substratum. Denny (1985) has
interpreted the field data obtained in terms of the force which will be
experienced by a few representative intertidal organisms for which data
are available (Table 6.7) where the Re values are in the range 104-106.
The forces are of three types: drag, wave acceleration reaction, and lift
(see Vogel 1981a, b). For example, exposure to shear forces in breaking
waves has been estimated by Denny (1985) by calculating the force on
the basis of drag alone and ignoring the acceleration reaction, which
contributes <15% to the total force. For waves up to a breaking height of
10 m, Semibalanus experiences lower shear forces than do mussels. It is
of interest that the bilaterally symmetrical mussel has a much greater
exposure to shear force if it is oriented broadside rather than end on to
theflow.From Table 6.7, it can be seen that drag coefficients are high for
the intertidal organisms measured in comparison to those of streamlined
shapes, e.g. of scallops, for which precise data are unavailable. Streamlin-
ing is only effective in reducing drag of an animal when it is optimally
oriented with respect to the flow. On the wave-swept rocky shore, flow
directions are unpredictable and therefore streamlining is probably not a
realistic option. As Denny (1985) points out, mussels may be present in
densely packed reefs where the forces are shared by all rather than by an
individual mussel. Some intertidal organisms, e.g. Semibalanus cariosus,
arefirmlycemented to the substrate by a special adhesive to resist the lift
and drag forces involved.
Adhesive forces involved in Semibalanus balanoides attachment have
been studied by Yule and Crisp (1983) and Yule and Walker (1987) by
using strain gauges. Eckman, Savidge, and Gross (1990) used an indirect
method of determining drag and detachment forces for Balanus
amphitrite from shear forces in a small flume. The latter study suggests
that the drag and detachment forces associated with Balanus are lower
than those determined directly with strain gauges in Semibalanus.
Whether this is a real difference consistent with the more exposed life-
style of Semibalanus or is due to differences in measurement technique
awaits further standardization of both methods.
In terms of the preceding introduction to forces commonly present on
wave-swept rocky shores, we must ask whether behavioral responses to
flow are involved in helping the organism to avoid flow dislodgement.
Adhesive attachment of benthic organisms on a wave-swept rocky shore
is a common solution to dislodgement, e.g. in barnacles (see earlier
discussion), or mussel byssus threads which are glued to the substrate
(Waite 1983). Skeletal stiffening by spicule formation, e.g. in sponges
and cnidarians (Koehl 1982), and strengthening by coral formation or
development of various forms of body elasticity which allow form
changes proportional to velocity, resulted in reduced drag (Koehl
1911 di). Most of these must be considered as structural adaptations to
resist the forces generated by flow.
Size may also represent a refuge from the worst effects of orbital
wind-wave action, e.g. for the sea anemone Metridium senile (Anthony
and Svane 1994). Large numbers of small sea anemones of pedal disc
diameter of 0.3 cm are found where wave exposure is strong versus
fewer, much larger, 2.5- to 3.5-cm-pedal-disc-diameter individuals in
subtidal lower wave-exposed habitats.
Behavioral responses may be involved (Koehl 1984) in withstanding
moving water; they include adjusting the orientation of the individual to
minimize lift and drag. One of the characteristics shown by the sea
anemone Anthopleura xanthogrammica which live in exposed intertidal
surge channels of rocky shores on the U.S. West Coast was found by
Koehl (1977b) to involve a behavioral response to flow. In the field,

Koehl showed that there was a significant height difference between
individuals from more exposed (2.1-cm) and those from protected (5.6-
cm) sites. Subsamples of exposed and protected anemones were placed
in a laboratory sea table where the flows were maintained at a constant
low level of <5 cm s"1. After 8 d, the exposed individuals had significantly
increased in height to 4.3 cm, whereas those from the protected site
showed no significant change. This suggests that the response was
rheotropic and that excessive drag is prevented by maintaining a low
profile in the benthic boundary layer. This was possible because this
anemone is not a suspension feeder but feeds on dislodged mussels
which fall onto the oral disc. Drag forces measured on live and model
anemones showed that drag was greatest on the tallest ones. A retracted
sea anemone also had a streamlined shape consistent with waves coming
from any direction.
Changes of body shape or orientation of sessile epifauna with respect
to ambient flow which involve muscular movement or locomotion, e.g.
withdrawal of the tentacles of scleractinian coral polyps (Hubbard 1974)
or locomotion by normally sessile bivalves such as mussels, may be
related to the flow forces they experience. We could find no other de-
tailed studies of rheotactic behavior by sessile epifauna associated with
avoiding flow dislodgement.
Free-living epifauna
Arnold and Weihs (1978) developed a hydrodynamic model for plaice
rheotaxis which is applicable to most benthic macrofauna where
Re = 10,000 - 1,000,000 (where Re is defined by multiplying body length
times free-stream velocity divided by the kinematic viscosity of sea-
water) as a means of analyzing the effect of flow forces in dislodging
free-living epifauna. In the model, the fish is assumed to be dislodged by
the flow when the drag force on it is just balanced by the static friction
force between the fish and the bottom. The static friction force is propor-
tional to the difference between the fish's buoyant weight and the lift
force generated by the flow over the fish.
An experimental study of mussel dislodgement of both Mytilus
californianus and M. edulis was made by Witman and Suchanek (1984),
using a direct drag force measurement system. The force that was re-
quired to dislodge the mussel by breaking the bysuss threads was re-
corded. It was found that attachment strengths increased with shell area
and varied with position in the mussel reef and with the degree of wave
exposure of the habitat. The measured attachment strength of the byssus
threads was found to be proportional to the degree of wave exposure of
the mussel's habitat. The Re under flume flow conditions and based on
the valve length of the mussels used by Witman and Suchanek (1984)
ranged from 9000 to 21,000, but with simulated epifloral attachments of
considerable length, this increased to 53,000-27,000,000. Thus, mussels
that had been overgrown by kelp experienced flow-induced drag forces
that were two to six times greater than on the mussels alone, and this
explains why, during storms, mussels could be dislodged. Laboratory
experiments by Young (1985) were designed to determine the effect of
various environmental variables on the rate of byssus formation by M.
edulis. The most important factor was found to be periodic agitation,
simulating wave surges, which caused 15.8 threads d"1 for each mussel to
be produced. This rate was twice that of any other factor tested such as
temperature, salinity, tidal submergence, or season, although all of those
affected byssus thread formation. In agreement with results of Witman
and Suchanek (1984), Young (1985) found that wave-exposed mussels
had more byssus threads (87 per mussel) than those from sheltered sites
(48 per mussel).
Preliminary studies using the Arnold and Weihs (1978) model with
giant scallops Placopecten magellanicus have been presented by Wildish
and Saulnier (1993). In these experiments, the scallops were placed on
the smooth bottom of a flume and the velocities at which they were first
dislodged observed.
In order to study the hydrodynamics of the scallop, it is necessary to
study their body structures. The geometry of giant scallop valves (Fig.
6.10) shows that the valve length is less than the valve height at <10cm,
but is greater at >10-cm valve height. In natural conditions, the giant
scallop opposes the flow with its ventral edge. The frontal area Af pre-
sented to this flow (Fig. 6.10) is irregularly ellipsoidal in cross section,
and the lower, right valve is flatter than the upper, left valve. Because of
this irregularity, Af was determined empirically by Wildish and Saulnier
(1993) from photographic cutouts calibrated to area (rather than by
calculation of n\w). Seen from the posterior edge (Fig. 6.10), the maxi-
mum valve width does not occur centrally but more dorsally as shown.
Because of the choice by the scallop to oppose the flow with the ventral
edge first, i.e. with the "trailing edge" (Fig. 6.10) rather than the "blunt"
end, it is not optimally streamlined for lift (like the aeroplane wing). The
potential hydrodynamic instability of this orientation has been discussed
Left upper valve

dorsal
anterior
ventral view posterior view

h
Figure 6.10 Geometry of the adult giant scallop valves showing the measuring
origins adopted: h, valve height; /, valve length; w, maximum valve width or
thickness; g, maximum valve gape.
by Grant, Emerson, and Shumway (1993). Its common occurrence in the

field seems unrelated to optimum feeding orientation because Wildish et
al. (1987) showed that this was at 315 (exhalant 90 toflow)and in a field
test at high flows (>10cm-s~1) scallops grew better at 270 (exhalant 45
to flow) than opposing the flow at 90.
During flow over scallop valves, according to Bernoulli's principle, a
decrease in pressure occurs above the left valve as a result of the faster
flows there. The pressure on the right valve, which may be partly re-
cessed in the sediment, is higher, resulting in lift. Lift forces act in the
geometric centre of the scallop valve and tend to push the ventral edge
of the valve upwards into the flow. When the negative buoyancy is
balanced by the lift force (FL), the scallop will be lifted off the sediment.
The velocity at which this occurs (VL) was called the "lift-off speed" by
Arnold and Weihs.
It is possible that at velocities less than required to lift the scallop,
viscous and form drag on the valves will exceed the restraining force
exerted on the left valve by the sediment, and the scallop will slip down-
stream. At this slip speed, Arnold and Weihs (1978) show
FD=K(W-FL) (6.1)
where FD is the drag force, K is the static friction coefficient, and W is the
submerged weight of the scallop.
The lift is defined as
FL = 0.5psAfCLU2 (6.2)
where ps is the density of seawater, Af is the frontal area of the scallop

seen from the ventral edge, CL is a non-dimensional lift coefficient, and
U is the ambient free-stream velocity. The drag is determined from
D = 0.5psAfCDU2 (6.3)
where CD is the frontal drag coefficient. Substituting and rearranging,
- ^ - T = CL+-^- (6.4)
?AfU] %
The term on the left of the equation where Us is the slip speed, cm-s"1,
determined experimentally, should be nearly constant for individuals
with valves of geometrically similar shapes and sizes yielding Re > 104.
Re is defined here by valve height and local free-stream velocity. The
ratio on the left side of (6.4) can be referred to as Arnold and Weihs's
ratio and should be useful in correlating slip velocities with size of
benthic animals in which Re > 104.
For live, unrestrained giant scallops (n = 17), &US = 11-37 cm s"1 of 2.7
to 12.0 cm valve height of animals was determined by Grant et al. (1993)
on a smooth acrylic floor. Independent data for the same species of live
scallop, but with the valves restrained with a rubber band at the anterior
end so that the band did not touch the smooth acrylic floor of the
Saunders and Hubbard recirculating flume were obtained by Wildish and
Saulnier (1993). The scallops (n = 26) were of 2.6- to 10.4-cm valve height
and gave a Us = 9-32 cm-s"1 (Wildish and Saulnier 1993), similar to the
results obtained by Grant et al. (1993). In addition to live giant scallops,
Wildish and Saulnier (1993) undertook experiments on model scallops
consisting of the valves of live specimens used in the previous experi-
ments, but with all wet tissue removed and the valves dried. Each set of
o
30 -
Q?
CO
20-
Q
LU
LU -
K
Q_
CO
10-
CO oo/
0-
0 2 4 6 8 10 12
VALVE HEIGHT, cm
Figure 6.11 Valve height of the giant scallop as a function of slip speed: open
circles, live, shut valves, filled circles, dead, open valves (Wildish and Saulnier
1993).
hinged valves was then adjusted to a buoyant weight equal to that in life
by the addition of plasticine and the valves propped open to simulate the
observed maximal gaping in life - g in Fig. 6.10.
Results suggest that the slopes of Us as a function of valve height (Fig.
6.11) for live and model scallops are significantly different. Larger indi-
viduals (>8cm valve height) in which the valve is maximally gaped have
anonymously low slip speeds. This is linked to the increased frontal area,
which increases drag. In scallops of sizes >8 cm valve height, frictional
resistance with the bottom and buoyant weight is less efficient in over-
coming drag and lift. Possibly the behavioral phenomenon of recessing in
older scallops may be linked to this physical effect resulting from
allometric changes in the animal (Wildish and Saulnier 1993). Recessed
scallops are subject to lower drag forces because much of the body may
be beneath the benthic boundary layer. Additionally, the sides of the

recessed pit will prevent slipping although not affect the lift forces suffi-
cient to cause the ventral valve edge to flip upwards.
Recessing by Pecten maximus when placed in a large seawater tank
(flow characteristics not specified) on an 8-cm-deep bed of sand has been
described by Baird (1958). The scallop jetted water from the right side
near the auricle, causing the animal to swing around and the valves to
become inclined at 60 with the bottom. The jet of water is repeated by
opening and closing of the valves, which serves to blow away the sand.
By an extra large jet, the scallop is lifted and settles into the pit it has
excavated. From within the pit, it is enlarged by expelling water forcibly
around the mantle edge, as described by Caddy (1968), and this causes,
typically, a coating of sediment to appear on the upper, left valve. Older,
larger individuals of Placopecten magellanicus were more likely to be
recessed than smaller individuals. Recessing is also a common phenom-
enon in Amusium pleuronectes (Morton 1980) and Pecten ziczac (Waller
1975).
Some species of Australian terebratellid brachiopods are specially
adapted for sand bottoms where very high wave surge velocities may be
present (Richardson and Watson 1975). Magadina cumingi is such a
species, which is free living and not attached, as in most brachiopods.
The dorsal and ventral pedicle adjustor muscles act antagonistically to
extend and retract the pedicle, which is used as a pogo stick to raise the
animal off the bottom. Few details of the behavior of this interesting
brachiopod were given.
Although most sand dollars (Echinodermata, Echinoidea) are deposit
feeders (Telford and Mooi 1987) and thus do not warrant inclusion here,
they have been studied by hydrodynamic methods and concepts similar
to those presented here. Telford (1983) believed that in the sand dollar
Mellita quinquiesperforata, the lunules present act as lift spoilers, that is
reduce lift by inducing flow through the lunules, which tends to reduce
the pressure differential between the top and underside of the animal.
Lift spoilers increase the ambient velocity, which induces slip and critical
lift-off in this sand dollar. Dendraster excentricus (deposit-suspension
feeder), which lives in wave surge channels of the western U.S. coast,
avoids lift behaviorally by partially burrowing into the sand with the
result that its thin, flat body is upright in the sediment with the result that
all individuals are oriented parallel to the flow direction (Merrill and
Hobson 1970). They recognized bay, tidal channel, and outer coast
populations and the primary importance of water movement in their
distribution. At moderate levels of tidalflowand wave surge, the highest

densities were found in the "inclined" position - that is with the anterior
end of the test inserted in the sand and the thin test profile parallel to
ambient flows. In lesser flows, they opposed the flow or assumed a
horizontal position on the sand. During storm wave conditions,
Dendraster burrowed into the sediment. The "inclined" behavior of this
species would serve both to optimize feeding and to prevent washout by
waves. O'Neill (1978) showed that dense aggregations of the sand dollar
also benefitted feeding efficiency when they were passively suspension
feeding. At flows >10cm-s~1, the inclined position improves the feeding
success of the oral surface tube feet compared to that of the normal
horizontal position. This is because the cambered shape of the test causes
the streamlines between adjacent individuals to compress. Individuals
which are packed closely take advantage of local streamline curvature to
enhance particle capture for each other (O'Neill 1978).
Epifauna and aggregation

Evidence is presented here that aggregation among benthic suspension
feeders does commonly occur. In investigating the proximate causes of
aggregation, a number of general hypotheses should be considered (Ta-
ble 6.8). For those suspension feeders which are sessile after larval at-
tachment, the factors influencing larval recruitment are critically impor-
tant (see Chap. 3) and presumably linked to the flow-related advantages
in numbers 1 and 2 of Table 6.8. Other species can move away and
reattach at a new and perhaps better location, e.g. juvenile mussels. Yet
others are colonial: the original larval settler undergoes a sexual repro-
duction to create a colony at the original settlement site, e.g. some
sponges, corals, and ascidians. Because larval recruitment and its possi-
ble link to improved reproductive success (number 3) and larval recruit-
ment linked to predator protection (number 4, Table 6.8) are not directly
related to flow, they are not considered in detail. Nevertheless, the two
latter general hypotheses must be part of the synthetic phase of thinking
when the topic of aggregation is considered.
According to the first hypothesis of Table 6.8, aggregated groups of
individuals occupy areas where a specialized feature of the hydrody-
namic regime provides a locally higher flux of seston past the group of
suspension feeders. The spatial extent of the area occupied by the group,
or patch, is determined by the spatial limits dictated by the hydrodynam-
ics and the absolute seasonal seston flux values, which also dictate the
Table 6.8. Hypotheses concerned with aggregation in suspension-

feeding macrofauna.
No. Hypothesis Reference

1 Ho Larval recruitment to a group is not Fager (1964)
related to the location's being optimal Okamura (1986)
for suspension feeding
Larval recruitment to a group is
related to the location's being optimal
for suspension feeding
2 Ho Larval recruitment to a group is not Barkai (1991)
related to providing protection from
energetic water movement
Larval recruitment to a group allows
protection for some individuals from
energetic water movement
3 Hn Larval recruitment to a group is Crisp (1979)
unconnected with reproductive Okamura (1986)
success
Larval recruitment to a group results
from the group's being optimal for
reproductive success
4 Ho Larval recruitment to a group is not Okamura (1986)
related to protection from predators
Larval recruitment to a group results
in better protection from predators
carrying capacity (by addition or death of individuals) for the patch. At

high carrying capacities where densities may reach very high levels,
behavioral mechanisms may serve to limit densities, as shown by
Johnson (1959) in Phoronopsis viridis. In densely packed bivalve mollusc
reefs the nearest neighbor distance may be so small that the valves are
not able to open completely, thereby imposing spatial limits to growth
(Bertness and Grosholz 1985).
Scallops Placopecten magellanicus in intensive culture (Wildish et al.
1988) interact behaviorally ("biting syndrome") at high densities by
locking of valves, which can prevent feeding or involve adductor muscle
damage. Cote, Himmelman, and Clarereboudt (1994) grew juvenile scal-
lops in pearl nets with and without dummy scallops made by gluing
valves of dead scallops together. The pearl nets were suspended in the
southwestern Gulf of St. Lawrence and left for a summer's growth. The
results obtained by Cote et al. (1994) clearly showed that at up to a

maximum density of 250 scallops per net (=1563 scallops m"2) tested, the
dummy scallops did not decrease the growth of 25 scallops per net as did
living scallops up to this density. This suggests that the major effect was
caused by either food depletion or behavioral interaction, e.g. by inter-
ference competition in cages, but not directly by space limitation.
There are some suggestions in the literature that special and localized
hydrodynamic conditions controlling seston flux are linked to aggre-
gations of sessile suspension feeders. Fager (1964) described a reef of the
polychaete tube-dwelling Owenia fusiformis, which was 60 x 150 m long
with its long axis to the beach and at a depth of 6 m. The reef was
associated with a rip current which could serve to concentrate both the
seston supply and Owenia larvae at this site. In studies undertaken by
Bosence (1979) in an Irish loch, the serpulid polychaete Serpula
vermicularis was substrate limited. The larvae initially settled on rare
rock outcrops, but after this substrate was covered, social settlement
occurred, resulting in the building of extensive reefs. The reefs were
formed by subsequent waves of larvae which settled on the adult tubes of
serpulids, forming bushlike aggregations of tubes. In Narragansett Bay
blue mussel reefs were described by Nixon et al. (1971) as developing
across narrow inlets on bars near the mouths of tidal channels and on
muddy sandflats where energetic tidal currents rather than wind-wave
energy were the major source of water movement.
From the preceding and similar further examples available in the
literature, we conclude that the evidence that location is important for
the development of patches of suspension feeders is circumstantial and
weakly supportive of hypothesis 1 in Table 6.8. However, very little
information on the behavioral mechanisms which must be involved is
available.
One form of aggregation among sessile suspension feeders involves
coloniality, e.g. in some hydroids, sponges, bryozoans, and ascidians. In
these cases, it is probably difficult to determine a proximate cause, or
causes, of coloniality. Two examples will be given to illustrate this diffi-
culty. The larvae of the colonial ascidian Botryllus schlosseri can recog-
nize genetically similar forms and settle gregariously so as to increase the
chances of fusion as adults (Grosberg and Quinn 1986). This forms a
larger, more competitive type of colony. Genetically dissimilar forms
which meet in this way are unable to fuse and instead act as spatial
competitors. The second example is an interesting type of coloniality
discussed by Hughes (1979) among vermetid prosobranch gastropods.
Vermetids are sedentary forms with a loosely coiled (= vermiform) shell
which is cemented to a hard substratum. Many vermetids form loose

aggregations which may not qualify as true colonies, although the South
African Dendropoma corallinaceum seems to be a true colonial form,
producing dense sheetlike aggregations on rock surfaces which are po-
tent competitors for space, although it appears to tolerate mutualistically
the presence of other encrusting forms such as the calcareous alga
Lithothamnion sp. (Hughes 1979). D. corallinaceum is found in a narrow
zone near mean low water springs and suspension feeds by means of a
mucous net. In this species, the nets overlap and coalesce to form a
communal net, so that individuals, when they haul in the net, may be
feeding on part of a net secreted by their nearest neighbor. Hughes
(1979) considered that vermetid aggregation was linked to successful
fertilization by waterborne sperm and/or the use by juveniles of the
presence of adults to indicate a suitable niche for settlement.
Evidence that many epifaunal suspension feeders do live together in
aggregated groups or reefs comes from standard grab or core sampling
(Table 6.9), where the area sampled is usually 0.1m2 and maximally
1.0 m2. The examples included in Table 6.9 were selected because they
had high densities or biomasses and it can be assumed that the grab
sampled within an aggregated group. In no individual reports in Table
6.9 was the sample plan area varied, and, therefore, the type of spatial
distribution (random, regular, or contagious [Elliot 1977; Thrush 1991])
at all scales remains unknown. Another difficulty with the data shown in
Table 6.9 is that the densities given may include a preponderance of very
young, small individuals. This becomes important if there are no accom-
panying biomass data, as proved to be the case in many of the studies
cited.
Despite the obvious sampling problems, some impressively high num-
bers and biomasses are shown in Table 6.9. Thus, suspension-feeding
bivalves may reach a dry weight standing crop of 21 kg m~2. We conclude
that for a range of species, available evidence is consistent with their
living closely together as groups. Tube-living amphipods have high
densities, but because each individual is small, the population biomass
is modest. All the amphipods of Table 6.9 are probably deposit-
suspension feeders (Enequist 1949), suspension feeding when water
movement is sufficiently energetic by means of rigidly held, feathery
antennae for passive suspension feeding. The antennae can, in some
cases, also be used to surface deposit feed if flows slacken. Only the
suspension-feeding crinoids and ophuroids (Warner 1979) appear to be
present at significant biomasses (unfortunately no data are available in
this reference) because the adults link arms and passively filter feed by
extending their arms into the boundary layer. This has the effect of
anchoring them to resist energetic water movements and to create turbu-
lence in the benthic boundary layer and hence increase the feeding
potential.
An example of clumping behavior, which provides protection from
wave effects able to dislodge sessile epifauna, comes from a study of
sublittoral holothurians on the west coast of South Africa (Barkai 1991).
The holothurians involved were all passive suspension feeders. In loca-
tions where wave exposure was high, Thy one aurea (maximum density
377 nT2) and Pentacta doliolum (maximum density 464 m~2) occur to-
gether in dense aggregations. Thyone, which has degenerate tube feet,
insinuates its body beneath Pentacta, utilizing it as a means of attach-
ment. In more protected sites, Thyone creates separate clumps but
Pentacta is absent there. Flume studies showed that the strength of
attachment and resistance to increasing velocity as lift-off speed, Us - the
velocity at which a holothurian was swept away - were much higher in
P. doliolum than T aurea. Barkai (1991) considered that Thyone
commensally used the Pentacta colony to extend its range from protected
sites to more exposed ones. Model sea cucumber experiments in flumes
in which grouped and singular models were compared showed that Us of
grouped models was often two times higher than that of singular models.
The precise behavioral mechanisms involved in aggregation of these two
holothurians are unknown.
Usually, clumps are of individuals of a single species, e.g. polychaete or
amphipod tubes. Nowell and Church (1979) provide tube model experi-
ments in a flume that predicts tube density conditions which permit
reduction of drag within the clump. When tube density is sufficient to
give approximately one-twelfth the plan area of the bottom, the flow
changes to a "skimming flow" regime. In skimming flow over tube
clumps, much less kinetic energy reaches the bottom, and, under these
conditions, the flow-sediment interaction is stabilized. As pointed out by
Eckman et al. (1981), the Nowell and Church model does not accurately
predict stabilized versus de-stabilizing tube-sediment relations in field
conditions. They suggested that the discrepancy between field and labo-
ratory results was due to the common occurrence in the field of sediment
binding by mucus from bacteria or benthic diatoms. Eckman et al. (1981)
point out that a purely hydrodynamic cause of reduced kinetic energy
reaching the sediment as a result of tube clumping in field conditions has
not been demonstrated.
Table 6.9. Density and biomass among aggreging epifaunal suspension feeders.*
Density: no*m 2 Biomass, g d r y m 2
Taxa Location Mean Max Mean Max Reference
Bivalve molluscs
Modiolus modiolus Glacial till, Bay of Fundy, - 510b - 3,038b Peer et al. (1980)
Canada 281 1,023 Wildish and Peer (1983)
Abra alba Muddy sand, Consolidated 1,235 Warwick and Davies (1976)
substrate, Mud, Bristol 121
Channel, UK 70
Mesodesma deauratum Fine-medium sand, Grand 488 l,550 b 6,485 21,030b Hutcheson and Stewart
Banks, Nfld., Canada 3,010 12,000b 939 4,697b (1994)
Crassostrea virginica Estuaries of southeastern 4,077 214 Dame et al. (1984)
USA
Mytilus edulis Narragansett Bay, Rhode 2,139 10,962 Nixon et al. (1971)
Island, USA
Baltic Sea 36,000 158,000 101 Kautsky (1982)
Newfoundland, Canada 596 1,101 Thompson (1984)
Morecambe Bay, UK 1,234 Thompson (1984)
Perna perna South Africa 826 1,284 Thompson (1984)
Echinoderms
Amphiura filiformis Silt-sand burrower of 2,200 Warner (1979)
eastern Atlantic
Antedon bifida Outcropping rock, 1,200 Warner (1979)
Torbay, UK
Ophiothrix fragilis High tidal current flows 2,196 Warner (1979)
on various substrates
Polychaetes Fine sand-silt, La Jolla 500 15,000 Fager (1964)
Owenia fusiformis Bay, USA
Lanice conchilega Fine-medium sand, 20,200 1,094 Buhr and Winter (1977)
Weser estuary, FRG
Spio setosa Sand beach, Halifax 408 2,002 61 109 Muschenheim (1987a)
Harbour, Canada
Eudistylia vancouveri Rock, San Juan Is., 2,576 Merz (1984)
Pacific coast, USA
Phyllochaetopterus Fringing reef, Oahu, 62,400 Bailey-Brock (1979)
verrilli and Hawaii, USA
Mesochaetoptems
Sagittarius
Amphipods
Haploops fundiensis Silt-clay soft sediment, 376 923 0.35-0.53 0.96 Wildish (1984)
80-m depth, Bay of
Fundy, Canada
Ampelisca abdita Fine sand, Atlantic coast 1,360 73,000 Mills (1967)
Ampelisca vadorum Very coarse sand, 1,307 1,885 Mills (1967)
Atlantic coast,
N. Amer.
a
Wet biomass converted to dry by dividing by 3.33.
b
Maximum from single grab sample.
Distance,cm 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Water level
0.5 second
Figure 6.12 Swimming sequence of Amusium pleuronectes as recorded by a

cine camera (Morton 1980). The arrows indicate the timing of muscle adductions.
Swimming
The pectinids are the bivalves with the best developed ability to swim,
although Stanley (1970) mentions solenids, solemyids, and cardids as
other bivalves with some locomotory abilities. Among scallops, two
types of locomotion can be recognized (Moore and Trueman 1971): a
predator escape reaction, in which a limited number of adductions of the
valves moves the animal as a result of the rapid expulsion of pallial water
which pushes the animal in the opposite direction from the point of
predator attack, and a complex swimming behavior.
Field studies have described complex swimming sequences in
Placopecten magellanicus (Caddy 1968) and Amusium pleuronectes
(Morton 1980) as (1) a rise from the bottom, (2) level flight, and (3)
passive sinking (Fig. 6.12). During the rise and level flight stages of
swimming, periodic contractions of the adductor muscle occur, depend-
ing on the length of the flight, the species, and possibly the size of the
individual. The timing of valve adduction is shown in the swimming
sequence of A. pleuronectes shown in Fig. 6.12. The force powering
swimming in scallops is provided by periodic jet propulsion. This in-
volves valve gaping to take in pallial water, closing the mantle edge
around the water, followed by rapid contraction of the adductor muscle,
which forces it out in two vectored jets dorsally on either side of the
auricles. Valve opening is usually described as being due to elasticity of
the ligament or resilium (Demont 1990), although Vogel (1985) makes a
good case for flow-induced forces augmenting the ligament in reopening
the valves. According to this argument, dynamic pressure on the right
valve ventral edge will be high and on the left valve low, thereby creating
pressures which tend to open the valves. Calculations made by Vogel

(1985) based on a gape angle of -13 and at a swimming speed of
SOcm-s"1 suggest that the opening moment is -22% of that due to liga-
ment elasticity. Following this is a glide phase in which the valve margins
are closed before initiation of another adduction jet propulsion se-
quence. It is local muscular contractions of the mantle which helps to
produce the vectored jet propulsive forces, which are principally pow-
ered by adductor muscle contractions. Stephens (1978) has studied the
nervous control mechanisms involved in swimming for Argopecten
irradians. The jet propulsion rhythm is maintained during rise and level
flight phases by neuronal feedback from stretch receptors in the adduc-
tor muscle which are activated during valve opening.
The pressure pulses involved in swimming of Chlamys opercularis
have been recorded with a pressure transducer (Moore and Trueman
1971). The transducer consisted of plastic tubing and a hypodermic
needle sealed into the valve so that it opened into the pallial cavity.
Maximum pressures measured for normal swimming were 5-6 cm of
water at the relatively low swimming speed (-SOcm-s"1) of this species.
Escape reaction movements involved periodic pulses of about half the
pressure found in complex swimming by Chlamys (Moore and Trueman
1971).
Lift was determined experimentally in four scallops, Chlamys
islandica, C. opercularis, Placopecten magellanicus, and Pecten maximus,
by Gruffyd (1976). Measurements were made on model scallops of each
species at zero flow, equal to the buoyant weight of the live animal, and
at increasing velocities until the model became unstable as the lift gener-
ated by the flume flow approached the model weight. The results were
dependent on scallop size and the angle of incidence, that is the angle of
inclination of the ventral valve tips with the horizontal. Unfortunately,
the frontal area, Af, at different angles of incidence of the valves was not
determined so that coefficients of drag could not be determined with
these data. Re-examination of Gruffyd's results by Thorburn and
Gruffyd (1979) when the angle of incidence was 0 showed that small
amounts of lift were always generated, rather than the zero lift previ-
ously reported. The species examined by Gruffyd (1976) showed in-
creased lift as this angle was increased. Maximum lift was produced at
angles of incidence of -20 in all species, except P. magellanicus, in which
lift increased at angles at least up to 60. During initiation of swimming
behavior, the angle of attack of the body is important in achieving
sufficient lift-off from the substrate. In a similar model study of Chlamys
opercularis, Thorburn and Gruffyd (1979) showed that allometric

changes in valve shape were insufficient to provide lift to counteract
increasing weight with growth. This is because of the mass/surface area
law, which states that mass increases according to h3 while surface area
increases according to h2.
The lift and drag forces of plaster casts of six species of scallop have
been determined in flume studies by Millward and Whyte (1992). New
lift and drag coefficients using the plane form area of the valves, rather
than frontal area, were determined. Typical CL ranges from <0.1 to 0.8
for all species with stall angles of incidence of -15 for planoconvex
shells, e.g. Pecten alba, and -25 for others (biconvex valves). Millward
and Whyte (1992) point out that because of the use of an inconsistent set
of units, the calculated values of Gruffyd (1976) and Thorburn and
Gruffyd (1979) are two orders of magnitude greater than they should be.
These authors also found a correlation between CD and the thickness
ratio of the valve - equivalent to Ap as found also by Wildish and
Saulnier (1993) in Placopecten magellanicus.
Flow patterns over the valves were studied (Thorburn and Gruffyd
1979) by visualizing natural particles present in seawater by dark field
photography. The results suggest that a smooth attached flow over the
left valve becomes a turbulent eddying wake at angles of attack >20 and
at ambient velocities >25cm-s"1. The stability of C. opercularis during
swimming was enhanced by the presence of auricles, which, when re-
moved, caused instability. Auricles were also thought to increase lift at
low swimming speeds.
Hydrodynamic characteristics of scallops vary as a size-related phe-
nomenon because of body allometric changes which occur during
ontogeny. Although this was recognized by Gould (1971), the first stud-
ies of swimming behavior which incorporated size as a variable were
those of Morton (1980), Joll (1989), and Dadswell and Weihs (1990).
Measures required to determine the hydrodynamic characteristics of a
scallop valve (see Fig. 6.10) include valve height (/z), maximum thickness
or valve width (w), valve length (/), and valve area (a). From Dadswell
and Weihs (1990), the Re of the scallop body can be expressed as
Re = - ^ (6.5)
where U = ambient velocity and u = kinematic viscosity of seawater at a

given temperature. The fineness (F) of the scallop valves, which is re-
lated to drag encountered, is given by
Table 6.10. Pectinid swimming speeds estimated from original scatter

plots scaled against valve height*
Swimming
speed, cm s"1
h
Species cm xlO 4 F Mean Max Reference
Chlamys islandica 4.0 25 Gruffyd (1976)

7.0 35
C. opercularis 6.7 21 34 Thorburn and
Gruffyd (1979)
Amusium pleuronectes 6.5 45 70 Morton (1980)
10.0 37 40
A. balloti 5.5 68 80 Joll (1989)
10.0 90 100
Placopecten 5.5 2.7 4.4 1.19 49 68 Dadswell and
magellanicus 10.0 4.0 3.3 1.32 40 53 Weihs (1990)
*h = valve height, F-fineness,Ra = aspect ratio.
(6.6)
w
The aspect ratio of the scallop valves is given by
(6.7)
and the available lift, L, can be calculated from
(6.8)
where CL is the coefficient of lift for low aspect ratio wings obtained from
Hoerner (1975), 0.4 <CL< 0.8.
Complete data are available only for Placopecten magellanicus (Table
6.10). The data for this species show that the fineness ratio fits a near-
optimum shape, as does the relationship x/h = 0.6 in giant scallop adults
(where x = distance from ventral edge to the widest point in cross sec-
tion) for streamlining and fast swimming powers (Arnold and Weihs
1978). As pointed out in the section, Free-living epifauna, the blunt
80
70
EED, cm-s" 1
60
50
40-
CO
O
2 30
^ 20
CO
10
20 40 60 80 100 120
L, mm
Figure 6.13 Swimming speed of Placopecten magellanicus as a function of valve
height: vertical lines, sample range; points, means, horizontal lines, one standard
deviation; curved line,fittedby eye (Dadswell and Weihs 1990).
trailing edge (dorsal side) opposed to the flow would be the optimum
hydrodynamic shape. During the ontogeny of the giant scallop, the ad-
ductor grows toward the mid-centre point of the valve, a process which
is complete by h = 5.5 cm (Dadswell and Weihs 1990). A similar ventral
displacement of the adductor muscle in relation to the valves is not
complete until h = 8.0 cm.
Swimming speeds shown in the table are approximate estimates made
from scatter plots of the original data made under non-standard environ-
mental conditions (e.g. temperature). Another difficulty in comparing
the swimming speeds shown in Table 6.10 is that they were determined
in different ways. Thus, Gruffyd (1976) and Morton (1980) measure the
time taken over a set distance during level flight swimming, while Joll
(1989) and Dadswell and Weihs (1990) use the total distance travelled
divided by the time taken from the initiation of stage (1) to the termina-
tion of stage (2) - that is the rise and level flight stages - but exclude the
passive sinking stage (3). Nevertheless, it is clear that of the swimming

speeds of the species so far examined (Table 6.10), the fastest is the
Australian saucer scallop A. balloti. The effect of scallop size on swim-
ming speed is reported by Joll (1989) for A balloti and by Dadswell and
Weihs (1990) for P. magellanicus (Fig. 6.13) with a drop in performance
of all larger individuals. In fact, Dadswell and Weihs (1990) calculated
that the largest giant scallops, h > 11.0 cm, may be unable to generate
enough lift to overcome their buoyant weight. Presumably, it is only
higher ambient flows and increase in the angle of attack, both of which
increase lift, which permit the larger scallops to initiate swimming
behavior.
Other aspects of swimming performance, such as the distance trav-
elled and the number and rapidity of adductions per swimmingflight,are
obviously important in determining the value of swimming to individu-
als. Although data for distance travelled per swimming flight and adduc-
tion rate are available for Amusium (Morton 1980; Joll 1989), there are
insufficient data for other species to allow worthwhile comparison. De-
spite assertions by Morton (1980) that larger Amusium pleuronectes
swim for a longer time (and hence distance), his own data of swimming
time as a function of valve height have a non-significant r value. Simi-
larly, for the relationship between the number of adductions per swim-
ming flight, the relationship with valve height is non-significant. Data
presented by Joll (1989) for Amusium balloti of swimming distance
versus valve height were significantly correlated on log-transformed
data. One group of scallops tested when ambient flow was zero showed
significantly lower swim distances than when ambient flow was present.
The adduction rate of juvenile Placopecten magellanicus (h = 0.04-
0.35 cm) during swimming is positively related to temperature with a
Q10 3 adductions s"1 measured over the temperature range 9.0 to
14.5C (Manuel and Dadswell 1991). These data suggest that scallop
swimming performance comparisons must consider temperature as well
as individual size and ambient flow conditions. Manuel and Dadswell
(1993) studied swimming in relation to size of juvenile P. magellanicus.
At a Reynolds number <3 x 103, equivalent to a valve height <12-16mm,
the amount of swimming decreased with size. This critical size below
which swimming was also inefficient hydrodynamically may explain the
use of byssus attachment by the smallest juveniles of the giant scallop.
The type of sedimentary substrate on which adult bay scallops
Argopecten irradians are placed determines the probability that they will
swim (Winter and Hamilton 1985). Bay scallops were significantly more
likely to swim when placed on clean sand than on seagrass beds, the
preferred habitat. Distances swum by bay scallops depended inversely
on the weight of fouling organisms, as simulated by adding 2-8 g of epoxy
putty to the left valve. The predacieous gastropods Fasciolaria tulipa and
Murex pomium always caused the bay scallop to swim when the tentacles
came into contact with the snail's foot. The direction of swimming by the
scallop was controlled by the location on the mantle edge contacted by
the predator - the escape reaction was directly away from the source of
stimulation (Winter and Hamilton 1980).
Summary
The foregoing shows that orientations to flow by a range of suspension
feeders, e.g. infaunal clams, phoronidians, brachiopods, barnacles, and
epifaunal bivalves, do take place. Research required to establish that
suspension feeders orient rheokinetically or rheotactically to enhance
their feeding includes observations of the behavior in field and lab and
experimental confirmation that the response is to flow and that the
individual derives trophic advantage from the preferred orientation.
Other species grow with respect to flow direction - rheotropisms -
thereby optimizing their feeding and growth. Research required to es-
tablish that rheotropic growth is occurring includes field observations
which establish a preferred orientation with respect to flow and flume
experiments which establish the effect of flow on growth. Although
relatively few suspension feeders have locomotory capabilities, those
that do may use them to increase their feeding opportunities. The re-
search required to establish this mechanism is often difficult to complete
satisfactorily. Steps involved in the process include field observations of
suspension feeder movement; experimental confirmation of sensory
cues, which frequently will not be flow-related ones; and some experi-
mental verification that restrained animals grow more slowly than unre-
strained ones.
Tube-living epifauna behavior has been little studied and so of the five
types of tube recognized here (there may well be others), studies have
only been started in one of the groups. The research required includes a
careful field observation of normal behavior in the tube, hydrodynamic
experiments to make clear the flow-tube interaction and the way ethol-
ogy is linked to it and that improved feeding or growth results from the
observed behavior. Abelson et al. (1993) used a physical model for
passive suspension feeders such as corals and sponges to predict the
quality and quantity of seston encountered. The "slenderness ratio"

employed to characterise coral and sponge bodies uses the same height/
diameter ratio used in this chapter for tube-living epifauna. The model
predicts that high slenderness-ratio bodies ( high H/D) encounter
higher fluxes of finer, good quality particles, whereas low slenderness
ratio bodies (^ low H/D) encounter coarser, poorer quality bedload
particles. It is interesting to speculate why the ampeliscid amphipod tube
is elliptical and oriented with the long axis opposed to the major flow
vector. One possibility is that the amphipod feeds in the wake down-
stream of the tube to take advantage of resuspended sedimentary
particles.
From the results discussed in this chapter, it will be clear to the reader
that insufficient evidence is available for much of the behavior discussed
to link it to one or another of the responses presented at the beginning of
this chapter. In formulating future hypotheses about the behavior of
suspension feeders and flow, optimizing feeding and avoiding flow dis-
lodgement could be alternative hypotheses (Ho and H^ or competing
multiple hypotheses. Okamura (1990) pointed out that many suspension
feeders have alternative or parallel forms of nutrition. The importance of
these, e.g. of uptake of dissolved organic matter directly from seawater,
is still in doubt, although for some cases, e.g. photosynthetic microalgae
in coral reefs and chemoautotrophic bacteria in deepsea pogonophorans,
it is well established (Okamura 1990). All potential forms of nutrition
need to be considered when constructing hypotheses regarding suspen-
sion feeding and flow.
Many ways of overcoming theflowdislodgement problem for epifauna
do not involve behavior at all, but rather some form of morphological-
biochemical solution, e.g. adhesives from antennular secretions of barna-
cle cyprids and byssus threads in bivalves. Some solutions do, however,
involve behavioral adaptations, and the following were recognized: ori-
entation of the body to minimize drag, clumping of individuals to reduce
individual effects of drag, and recessing of scallops within the sediment
to minimize drag. Research required for these studies includes a field or
lab description of the flow avoidance behavior as well as a complete
hydrodynamic model study to show which orientation is "best" in regard
to feeding, slip or lift-off speed, individual drag within a clump versus an
isolated individual, and slip or lift-off speeds of non-recessed versus
recessed scallops. An example of the difficulty in deciding whether orien-
tation of free-living epifauna is passive or active is provided by the
scallop (Grant et al. 1993). If the response is active, it is not clear to what
the scallop is responding - is it feeding optimization or flow avoidance?

If the response is not concerned with feeding, as seems likely from the
available data, an alternative possibility is that in bidirectional tidal flows
the 90 to 270 axis is the best orientation to prevent being passively
swept away. Another possibility is that the ventral edge of the valves
opposing the flow may ensure the maximum lift available to initiate
swimming, although this would only work for one tidal direction. These
hypotheses deserve further experimental testing. That little research has
been undertaken in the laboratory is surprising because much of it can be
completed with the aid of model suspension feeders in flume studies,
thereby simplifying it.
For many benthic suspension feeders, larvae represent the dispersal
agent for the species. Yet, as we have shown (Chap. 3) for some suspen-
sion feeders, they may settle gregariously. Of four general hypotheses
shown in Table 6.8 and proposed to determine the mechanisms involved
in aggregation, none has received enough attention to determine their
relative importance. Clearly, more observational and experimental work
is required before general conclusions can be reached.
Swimming locomotion may be part of a predator avoidance reaction in
some epifaunal bivalves, although many authors consider scallop swim-
ming to have been refined for additional functions, e.g. in effecting
reproductive dispersal, finding mates, choosing better feeding condi-
tions. Predator avoidance behavior can readily be confused with flow-
related behavior. Some idea of the sensory cues used to initiate the reflex
responses involved in escaping predators is required, as well as the
specificity of the stimulus and its threshold level. When behavioral stud-
ies involving swimming are contemplated, the alternate hypothesis of
predator avoidance should be considered along with the flow-related
one.
The complex behavior of swimming should ideally be studied in con-
trolled environmental conditions (e.g. flow, temperature, salinity) as a
function of body size, with swimming performance expressed as speed,
distance travelled, and number of adductions per swimming flight. Rela-
tively few studies have met these rigorous criteria, although it is clear
that there is an optimum swimming performance by mid-sized individu-
als. The older and larger scallops swim at lower speeds than their smaller
confreres. Morton (1980) thought that this was because of senescence in
the adductor muscle with age and a consequent change in behavior. In
contrast, Gould (1971) and later writers, particularly Dadswell and
Weihs (1990), who linked ontogenic changes to hydrodynamic character-
istics, consider that ontogenic changes are responsible for the size-
related swimming speed differences. In particular, the weight, which
scales according to h3 during ontogeny, outstrips the lift available, which
only increases as h2.
References
Abelson, A., T. Miloh, and Y. Loya. 1993. Flow patterns induced by substrata
and body morphologies of benthic organisms and their roles in
determining availability of food particles. Limnol. Oceanogr. 38: 1116
1124.
Anthony, K. R. N., and I. Svane. 1994. Effect of flow-habitat on body size and
reproductive patterns in the sea anemone, Metridium senile, in the
Gullmarsfjord, Sweden. Mar. Ecol. Prog. Ser. 113: 257-269.
Arnold, G. P., and D. Weihs. 1978. The hydrodynamics of rheotaxis in the
plaice (Pleuronectes platessa L.). J. Exp. Biol. 75: 147-169.
Ayling, A. M. 1976. The strategy of orientation in the barnacle Balanus
trigonus. Mar. Biol. 36: 335-342.
Bailey-Brock, J. H. 1979. Sediment trapping by chaetopterid polychaetes on a
Hawaiian fringing reef. J. Mar. Res. 37: 643-656.
Baird, R. H. 1958. On the swimming behaviour of scallops (Pecten maximus
L.). Proc. Malacological Soc. 33: 67-71.
Barkai, A. 1991. The effect of water movement on the distribution and
interaction of three holothurian species on the South African west coast.
Barnes, R. D. 1964. Tube building and feeding in the chaetopterid polychaete,
Spiochaetopterus oculatus. Biol. Bull. 127: 397-412.
Bertness, M. D., and E. Grosholz. 1985. Population dynamics of the ribbed
mussel, Geukensia demissa: the costs and benefits of an aggregated
distribution. Oecologia (Berlin) 67: 192-204.
Borowsky, B. 1983. Reproductive behaviour of three tube-building peracarid
crustaceans: the amphipods Jassa falcata and Amphithoe valida and the
tanaid Tanais cavolinii. Mar. Biol. 77: 257-263.
Bosence, D. W. J. 1979. The factors leading to aggregation and reef formation
in Serpula vermicularis L., p. 299-318. In G. Larwood and B. R. Rosen
(eds.) Biology and systematics of colonial organisms. Academic Press,
London.
Bowers, R. L. 1968. Observations on the orientation and feeding behaviour of
barnacles associated with lobsters. J. Exp. Mar. Biol. Ecol. 2: 105-112.
Buhr, K.-J. and J. E. Winter. 1977. Distribution and maintenance of a Lanice
conchilega association in the Weser estuary (FRG) with special inference
to the suspension-feeding behaviour of Lanice conchilega, p. 101-113. In
B. F. Keegan, P. 6'Ceidigh, and P. J. S. Boaden (eds.) Biology of benthic
organisms. Pergamon Press, Oxford.
Caddy, J. F. 1968. Underwater observations on scallop (Placopecten
magellanicus) behaviour and drag efficiency. J. Fish. Res. Board Can. 25:
2123-2141.
Carey, D. A. 1983. Particle resuspension in the benthic boundary layer by a
tube-building polychaete. Can. J. Fish. Aquat. Sci. 40 (Suppl.): 301-308.
Carthy, J. D. 1958. An introduction to the behaviour of invertebrates. George

Allen & Unwin, London.
Cote, J., J. H. Himmelman, and M. R. Clarereboudt. 1994. Separating effects
of limited food and space on growth of the giant scallop, Placopecten
magellanicus in suspended culture. Mar. Ecol. Prog. Ser. 106: 85-91.
Creutzberg, F. 1975. Orientation in space: animals. Invertebrates, p. 555-653.
In O. Kinne (ed.) Marine ecology. Vol. II. Part 2. Wiley, London.
Crisp, D. J. 1953. Changes in the orientation of barnacles of certain species in
relation to water currents. J. Anim. Ecol. 22: 331-343.
1960a. Factors influencing growth rate in Balanus balanoides. J. Anim. Ecol.
29: 95-116.
1960b. Mobility of barnacles. Nature 188: 1208-1209.
1979. Dispersal and re-aggregation in sessile marine invertebrates,
particularly barnacles, p. 319-327. In G. Larwood and B. R. Rosen (eds.)
Biology and systematics of colonial organisms. Academic Press, London.
Crisp, D. J., and H. G. Stubbings. 1957. The orientation of barnacles to water
currents. J. Anim. Ecol. 26: 179-196.
Dadswell, M. J., and D. Weihs. 1990. Size-related hydrodynamic characteristics
of the giant scallop, Placopecten magellanicus (Bivalvia: Pectinidae). Can.
J. Zool. 68: 778-785.
Demont, M. E. 1990. Tuned oscillations in the swimming scallop, Pecten
maximus. Can. J. Zool. 68: 786-791.
Denny, M. W. 1982. Forces on intertidal organisms due to breaking ocean
waves: design and application of a telemetry system. Limnol. Oceanogr.
27: 178-183.
1985. Wave forces on intertidal organisms: a case study. Limnol. Oceanogr.
30: 1171-1187.
Dolmer, P., M. Karlsson, and I. Svane. 1994. A test of rheotactic behavior of
the blue mussel Mytilus edulis L. Phuket. Mar. Biol. Cent. Spec. Publ. 13:
177-184.
Eckman, J. E., and A. R. M. No well. 1984. Boundary skin friction and
862.
Eckman, J. E., A. R. M. No well, and P. A. Jumars. 1981. Sediment
destabilization by animal tubes. J. Mar. Res. 39: 361-31 A.
Eckman, J. E., W. B. Savidge, and T. F. Gross. 1990. Relationship between
duration of cyprid attachment and drag forces associated with detachment
of Balanus amphitrite cyprids. Mar. Biol. 107: 111-118.
Edington, J. M. 1968. Habitat preferences in net-spinning caddis larvae with
special reference to the influence of water velocity. J. Anim. Ecol. 37:
675-692.
Elliot, J. M. 1977. Some methods for the statistical analysis of samples of
benthic invertebrates. Freshw. Biol. Assoc. Sci. Publ. 25.
Emig, C. C, and F. Becherini. 1970. Influence des courants sur l'ethologie
alimentaire des phoronidians. Etude par series de photographies
cycliques. Mar. Biol. 5: 239-244.
Enequist, P. 1949. Studies on the soft-bottom amphipods of the Skagerrak.
Zool. Bidr. Upps. 28: 297-492.
Eshleman, W. P., and J. L. Wilkins. 1979. Brachiopod orientation to current
direction and substrate position (Terebratalia transversa). Can. J. Zool. 57:
2079-2082.
Fager, E. W. 1964. Marine sediments: effects of a tube-building polychaete.

Science 143: 356-359.
Fitzsimons, G. 1965. Feeding and tube-building in Sabellastarte magnified
(Shaw) (Sabellidae: Polychaeta). Bull. Mar. Sci. 15: 642-671.
Gould, S. J. 1971. Muscular mechanics and the ontogeny of swimming in
scallops. Palaeontology 14: 61-94.
Grant, J., C. W. Emerson, and S. E. Shumway. 1993. Orientation, passive
transport and sediment erosion features of the sea scallop Placopecten
magellanicus in the benthic boundary layer. Can. J. Zool. 71: 953-959.
Grosberg, R, K., and J. F. Quinn. 1986. The genetic control and consequences
of kin recognition by the larvae of a colonial marine invertebrate. Nature
322: 456^59.
Gruffyd, L. D. 1976. Swimming in Chlamys islandica in relation to current
speed and an investigation of hydrodynamic lift in this and other scallops.
Norw. J. Zool. 24: 365-378.
Hartnoll, R. G. 1967. An investigation of the movement of the scallop. Helg.
Wiss. Meeresunters. 15: 523-533.
Helmuth, B., and K. Sebens. 1993. The influence of colony morphology and
orientation to flow on particle capture by the scleractinian coral Agaricia
agaricities (Linnaeus). J. Exp. Mar. Biol. Ecol. 165: 251-278.
Hoerner, S. F. 1965. Fluid-dynamic drag. Midland Park, New Jersey.
Hubbard, J. A. E. B. 1974. Scleractinian coral behaviour in calibrated current
environment: an index to their distributional patterns. Proc. Second Int.
Coral Reef Symp. Great Barrier Reef Comm. 2: 107-126.
Hughes, R. N. 1979. Coloniality in Vermetidae (Gastropoda), p. 243-253. In
G. Larwood and B. R. Rosen (eds.) Biology and systematics of colonial
organisms. Academic Press, London.
Hutcheson, M. S., and P. L. Stewart. 1994. A possible relict population of
Mesodesma deauratum (Turton): Bivalvia (Mesodesmatidae) from the
southeast shoal, Grand Banks of Newfoundland. Can. J. Fish. Aquat. Sci.
51: 1162-1168.
Johnson, A. S. 1988. Hydrodynamic study of the functional morphology of the
benthic suspension feeder Phoronopsis viridis (Phoronida). Mar. Biol.
100: 117-126.
Johnson, R. G. 1959. Spatial distribution of Phoronopsis viridis Hilton.
Science, N.Y. 129: 1221.
Joll, L. M. 1989. Swimming behaviour of the saucer scallop Amusium balloti
(Mollusca: Pectinidae). Mar. Biol. (Berl.) 102: 299-305.
Kautsky, N. 1982. Growth and size structure in a Baltic Mytilus edulis
population. Mar. Biol. 68: 117-133.
Koehl, M. A. R. 1977a. Mechanical design of cantilever-like sessile organisms:
sea anemones. J. Exp. Biol. 69: 127-142.
1977b. Effects of sea anemones on the flow forces they encounter. J. Exp.
Biol. 69: 37-105.
1982. Mechanical design of spicule-reinforced connective tissues: stiffness.
J. Exp. Biol. 98: 239-267.
1984. How do benthic organisms withstand moving water? Am. Zool. 24:
57-70.
laboratory behaviour and field observations. Paleobiology 3: 270-
287.
1978. Brachiopod orientation to water movement: functional morphology.
Lethia 11: 67-79.

coral Leptogorgia. Biol. Bull. 151: 344-356.
Levin, L. A. 1982. Interference interactions among tube dwelling polychaetes
in a dense infaunal assemblage. J. Exp. Mar. Biol. Ecol. 65: 107-119.
Luckenbach, M. W. 1986. Sediment stability around animal tubes: the roles of
hydrodynamic processes and biotic activity. Limnol. Oceanogr. 31: 779-
787.
Manuel, J. L., and M. J. Dadswell. 1991. Swimming behaviour of juvenile giant
scallop, Placopecten magellanicus, in relation to size and temperature.
Can. J. Zool. 69: 2250-2254.
1993. Swimming of juvenile sea scallops, Placopecten magellanicus (Gmelin):
a minimum size for effective swimming. J. Exp. Mar. Biol. Ecol. 174: 137-
175.
Mathers, N. F. 1976. The effects of tidal currents on the rhythms of feeding
and digestion in Pecten maximus L. J. Exp. Mar. Biol. Ecol. 24: 271-
284.
Merrill, R. J., and E. S. Hobson. 1970. Field observations of Dendraster
excentricus, a sand dollar of western North America. Am. Midi. Nat. 83:
594-624.
Merz, R. A. 1984. Self-generated versus environmentally produced feeding
currents: a comparison for the sabellid polychaete Eudistyla vancouveri.
Biol. Bull. 167: 200-209.
Meyer, D. L., C. A. LaHaye, N. D. Holland, A. C. Arneson, and J. R.
Strickland. 1984. Time-lapse cinematography of feather stars
(Echinodermata: Crinoidea) on the Great Barrier Reef, Australia:
demonstrations of posture changes, locomotion, spawning and possible
predation by fish. Mar. Biol. 78: 179-184.
Mills, E. L. 1967. The biology of an ampeliscid amphipod crustacean sibling
species pair. J. Fish. Res. Board Can. 24: 305-355.
Mill ward, A., and M. A. Whyte. 1992. The hydrodynamic characteristics of six
scallops of the superfamily Pectinacea, class Bivalvia. J. Zool. Lond. 227:
547-566.
Oceanogr. 35: 680-696.
Moore, J. D., and E. R. Trueman. 1971. Swimming of the scallop Chlamys
opercularis (L.). J. Exp. Mar. Biol. Ecol. 6: 179-185.
Morton, B. 1980. Swimming in Amusium pleuronectes (Bivalvia: Pectinidae). J.
Zool. London 190: 375-404.
Muschenheim, D. K. 1987a. The role of hydrodynamic sorting of seston in the
nutrition of a benthic suspension feeder, Spio setosa (Polychaeta:
Spionidae). Biol. Oceanogr. 4: 265-288.
1987b. The dynamics of near-bed seston flux and suspension-feeding
benthos. J. Mar. Res. 45: 473-496.
Myers, A. C. 1972. Tube-worm-sediment relationships of Diopatra cuprea
(Polychaeta: Onnpulidae). Mar. Biol. 17: 350-356.
Nowell, A. R. M., and M. Church. 1979. Turbulent flow in a depth limited
boundary layer. J. Geophys. Res. 84: 4816-4824.
Okamura, B. 1986. Group living and the effects of spatial position in
aggregations of Mytilus edulis. Oecologia (Berlin) 69: 341-347.
1990. Behavioural plasticity in the suspension feeding of benthic animals, p.

637-660. In R. N. Hughes (ed.) Behavioural mechanisms of food
selection. NATO ASI Ser. Vol. G.20. Springer-Verlag, Berlin.
O'Neill, P. L. 1978. Hydrodynamic analysis of feeding in sand dollars.
Oecologia 34: 157-174.
O'Riordan, C. A., S. G. Monismith, and J. R. Koseff. 1993. An experimental
study of concentration boundary layer formation over a bed of model
bivalves. Limnol. Oceanogr. 38: 1712-1729.
O'Riordan, C. A., S. G. Monismith, and J. R. Koseff. 1995. The effect of
bivalve excurrent jet dynamics on mass transfer in a benthic boundary
layer. Limnol. Oceanogr. 40: 330-344.
Otway, N. M., and A. J. Underwood. 1987. Experiments on orientation of the
intertidal barnacle, Tesseropora rosea (Krauss). J. Exp. Mar. Biol. Ecol.
105: 85-106.
Peer, D., D. J. Wildish, A. J. Wilson, J. Hines, and M. Dadswell. 1980.
Sublittoral macro-infauna of the lower Bay of Fundy. Can. Tech. Rep.
Fish. Aquat. Sci. 981: 74.
Richardson, J. R., and J. E. Watson. 1975. Locomotory adaptations in a free-
lying brachiopod. Science 189: 381-382.
Riisgard, H. U., and N. M. Ivarsson. 1990. The crown filament pump of the
suspension feeding polychaete Sabella penicillus: filtration effects of
temperature and energy cost. Mar. Ecol. Prog. Ser. 62: 249-257.
Stanley, S. M. 1970. Relation of shell form to life habits of the Bivalvia
(Mollusca). Geol. Soc. Am. Inc., Memoire 125. Geological Society of
America, Boulder, Colo.
Stephens, P. J. 1978. The sensitivity and control of the scallop mantle edge. J.
Exp. Biol. 75: 203-221.
Taghon, G. L., A. R. M. Nowell, and P. A. Jumars. 1980. Induction of
suspension feeding spionid polychaetes by high particulate fluxes. Science
210: 562-564.
Telford, M. 1983. An experimental analysis of lunule function in the sand
dollar Mellita quinquiesperforata. Mar. Biol. 76: 125-134.
Telford, M., and R. Mooi. 1987. The art of standing still. New Scientist 114:
36-39.
Theodor, J. 1963. Contribution a l'etude des Gorgones. III. Trois formes
adaptives d'Eunicella stricta en fonction de la turbulence et du courant.
Vie Milieu 14: 815-818.
Thomas, F. I. M. 1994. Morphology and orientation of tube extensions on
aggregations of the polychaete annelid Phragmatopoma californica. Mar.
Biol. 119: 525-534.
Thompson, R. J. 1984. Production, reproductive effort, reproductive value and
reproductive cost in a population of the blue mussel Mytilus edulis from a
subarctic environment. Mar. Ecol. Prog. Ser. 16: 249-257.
Thorburn, I. W., and L. D. Gruffyd. 1979. Studies of the behaviour of the
scallop Chlamys opercuralis (L.) and its shell in flowing seawater. J. Mar.
Biol. Assoc. U.K. 59: 1003-1023.
Thrush, S. F. 1991. Spatial patterns in soft bottom communities. Trends Ecol.
Evol. 6: 75-78.
Trager, G. C, D. Coughlin, A. Genin, Y. Achituv, and A. Gangopodyay. 1992.
Foraging to a rhythm of ocean waves: porcelain crabs and barnacles
synchronize feeding motions with flow oscillations. J. Exp. Mar. Biol.
Ecol. 164: 73-86.
Trager, G. C, J. -S. Hwang, and J. R. Strickler. 1990. Barnacle suspension

Turner, E. J., and D. C. Miller. 1991a. Behaviour and growth of Mercenaria
mercenaria during simulated storm events. Mar. Biol. I l l : 55-66.
1991b. Behaviour of a passive suspension feeder (Spiochaetopterus oculatus)
(Webster) under oscillatory flow. J. Exp. Mar. Biol. Ecol. 149: 123-137.
Vincent, R., G. Desrosiers, and Y. Gratton. 1988. Orientation of the infaunal
bivalve My a arenaria L. in relation to local current direction on a tidal
flat. J. Exp. Mar. Biol. Ecol. 124: 205-214.
Vogel, S. 1981a. Life in moving fluids: the physical biology of flow. Willard
1981b. Behaviour and the physical world of an animal, p. 179-198. In P. P.
G. Bateson and P. H. Klopfer (eds.) Perspectives in ethology, Vol. 4.
Plenum Press, New York.
1985. Flow-assisted shell reopening in swimming scallops. Biol. Bull. 169:
624-630.
Vosburgh, F. 1977. The response to drag of the reef coral, Acropora reticulata,
p. 477-482. In D. L. Taylor (ed.) Proc. 3rd Intl. Coral Reef Symp. Univ.
of Miami, Miami 1.
Wainwright, S. A., and J. R. Dillon. 1969. On the orientation of sea fans
(genus Gorgonia). Biol. Bull. 136: 130-139.
Waite, J. H. 1983. Adhesion in bysally attached bivalves. Biol. Rev. 58: 209-
231.
Waller, T. R. 1975. The behaviour and tentacle morphology of pteriomorphan
bivalves: a motion picture study. Bull. Am. Malacol. Union, p. 7-13.
Warner, G. F. 1977. On the shapes of passive suspension feeders, p. 567-576.
In B. F. Keegan, P. O. Ceidigh, and P. J. S. Boaden (eds.) Biology of
benthic organisms. Pergamon Press, Oxford.
1979. Aggregation in echinoderms, p. 375-396. In G. Larwood and B. R.
Rosen (eds.) Biology and systematics of colonial organisms. Academic
Press, London.
Warwick, R. M., and J. R. Davies. 1976. The distribution of sublittoral
macrofauna communities in the Bristol Channel in relation to substrate.
Est. Coast. Mar. Sci. 5: 111-222.
Wildish, D. J. 1984. Secondary production of four, sublittoral, soft sediment
amphipod populations in the Bay of Fundy. Can. J. Zool. 62: 1027-
1033.
Wildish, D. J., and U. Lobsiger. 1987. Three-dimensional photography of soft
sediment benthos, S.W. Bay of Fundy. Biol. Oceanogr. 4: 227-241.
Wildish, D. J., and D. Peer. 1983. Tidal current speed and production of
benthic macrofauna in the lower Bay of Fundy. Can. J. Fish. Aquat. Sci.
40: 309-321.
Wildish, D. J., and A. M. Saulnier. 1992. The effect of velocity and flow
direction on the growth of juvenile and adult giant scallops. J. Exp. Mar.
Biol. Ecol. 155: 133-143.
Wildish, D. J., and A. M. Saulnier. 1993. Hydrodynamic control of filtration in
Wildish, D. J., A. J. Wilson, W. Young-Lai, A. M. DeCoste, D. E. Aiken, and
J. D. Martin. 1988. Biological and economic feasibility of four grow-out
methods for the culture of giant scallops in the Bay of Fundy. Can. Tech.
Rep. Fish. Aquat. Sci. 1658.
Winter, M. A., and P. V. Hamilton. 1985. Factors influencing swimming in bay
scallops, Argopecten irradians (Lamarck, 1819). J. Exp. Mar. Biol. Ecol.
88: 227-242.
Witman, J. D., and T. H. Suchanek. 1984. Mussels in flow: drag and
dislodgement by epizoans. Mar. Ecol. Prog. Ser. 16: 259-268.
Young, G. A. 1985. Byssus-thread formation by the mussel Mytilus edulis:
effects of environmental factors. Mar. Ecol. Prog. Ser. 24: 261-271.
Yule, A. B., and D. J. Crisp. 1983. Adhesion of cypris larvae of the barnacle,
Balanus balanoides, to clean and athropodin treated surfaces. J. Mar.
Biol. Assoc. U.K. 63: 261-271.
Yule, A. B., and G. Walker. 1987. Adhesion in barnacles, p. 389-402. In A. J.
Southward (ed.) Barnacle biology. A. A. Balkema, Rotterdam.
7
Benthic populations and flow
The concept of hydrodynamic limitation of macrobenthic animal

populations arose from an attempt to explain the importance of flow
to benthic populations consisting of two key trophic groups: the deposit
and suspension feeders (Wildish 1977). Consistent with the aims
outlined in Chapter 1, we are concerned primarily with the latter
trophic group, and deposit feeding macrofauna are not dealt with in
detail here.
We will first consider the historical background of the benthic limita-
tion byflowconcept, as well as a special theory associated with it: trophic
group mutual exclusion, and hypotheses derived from the latter. We
present turbulent mass transfer models relating seston supply and uptake
by populations of suspension feeders with both field and laboratory
experiments to test the theory. We also present field evidence that sus-
pension feeding populations can be impoverished by high energy water
movement events. For completeness, we briefly consider two possible
competing theories of trophic group mutual exclusion which propose
that biotic interactions between deposit and suspension feeders explain
the exclusion of one of the groups.
Benthic limitation by flow

The concept to which the general theory of benthic limitation by flow
belongs is that of limiting factors (Wildish 1977). Limiting factor theory
has a long history, beginning with Liebeg's (1840) law of the minimum
(see in Odum 1971) and Blackman's (1905) ideas on optima and limiting
factors. Both of these authors were dealing with terrestrial plants, and
the application of limiting factors in benthic ecology occurred later (e.g.
Levinton and Lopez 1977; Wildish 1977). According to the limiting fac-
270
Benthic Populations and Flow 271
Table 7.1. Trophic group classification of benthic

macrofauna.
Letter code Trophic group

S Suspension feeding
D Deposit feeding
C Carnivores
O Omnivorous scavengers
A Algal scrapers
Source: Wildish (1985).
tor concept, multiple interactive physical and biological factors regulate

macrofaunal community composition, biomass, and productivity.
Trophic group mutual exclusion (Wildish 1977), a special theory of
benthic limitation, depends on key types, or trophic groups, of
macrofauna: the deposit and suspension feeders. Mutual exclusion arises
because each group responds differently to flow, as is described in detail
later in this section. The trophic group concept originated with Hunt
(1925), who described the feeding methods of the benthic animals taken
in grab samples off the fishing grounds near Plymouth, United Kingdom.
Trophic group designations such as those shown in Table 7.1 have subse-
quently been used to generalize the functional nature of the benthic
communities with which ecologists are working (e.g. Sanders 1956,1960;
Rhoads and Young 1970; Pearson 1971; Wildish 1977; Neyman 1979;
Pearson and Rosenberg 1987).
The trophic group concept has not been without modern critics; cogent
criticisms raised by Dauer, Maybury, and Ewing (1981) are based on the
following points:
Designations of trophic groups are not available for some species.
Some benthic animals are capable of facultative switching from/to suspension
and deposit feeding (Chap. 4).
The limiting physical factors in suspension feeding production remain un-
known in many cases.
Environmental factors not related to feeding, and hence trophic type, may be
critical in controlling local density and productivity of suspension feeders.
Although these criticisms are warranted in some cases, we believe that
they can be resolved by further research. Thus, experimental research in
laboratory flumes is required to answer the first three points. Facultative
suspension-deposit feeders may actually represent only a small fraction
of total benthic animals. Such species usually occur in hydrodynamic

conditions where the erosional/depositional nature of the sediments and
associated water flow patterns is transitional. They can be counted as the
feeding method commonly used in the local conditions. That non-trophic
factors - the last item in the list - may be critical in controlling produc-
tion has been incorporated in the model of unimodal response presented
herein and can be examined as benthic community impoverishment
hypotheses (see Table 7.2 and the section, Water movement impoverish-
ment). As benthic biology matures as a discipline, we would expect the
focus to switch from the arbitrary generalization of the trophic group to
the individual species.
Seed and Suchanek (1992) considered a comprehensive range of
physical and biological environmental factors of importance in control-
ling population ecology of blue mussels. This methodology is probably
applicable to other suspension feeders. The major concern of this chap-
ter is to consider three hydrodynamic mechanisms recognized as signifi-
cant limits and, hence, controls on species composition, biomass, and
productivity. They include the hydrodynamics associated with larval dis-
persal and recruitment (see also Chap. 3), turbulent seston supply to
suspension feeders as food (Wildish 1977), and seawater velocities suffi-
cient to cause macrofaunal impoverishment (Wildish and Kristmanson
1979). Such controls depend directly on water movement and result in
the types of benthic communities diagrammed in Fig. 7.1. In Fig. 7.1,
water movement is the independent variable which, increasing from left
to right, causes flow-dependent effects on benthic animals as well as
sedimentary deposition and natural selection of the proportions of de-
posit-suspension feeders and, hence, type of benthic community (de-
posit or mixed). The highest levels of suspension feeding benthic produc-
tion only occur where a suitable stable substrate is present.
Mutual exclusivity of the two key trophic groups arises because they
respond differently to water movement forces, as depicted in Fig. 7.1.
Deposit feeders are directly and indirectly affected by flow in their
physiology, behavior, and ecology. Thus, infaunal Macoma spp.
(Levinton 1991) and surface deposit feeders (Miller, Bock, and Turner
1992) may be inhibited by flow during feeding; locomotion of surface
deposit feeders may be inhibited (Miller et al. 1992) and burrowing
enhanced (Levinton et al. 1995) by increasing flows. By their burrowing
and feeding activities, deposit feeders influence sediment-water rela-
tions (Davis 1993). Such activities cause sediment to be ejected into the
overlying seawater, decrease silt-clay cohesiveness, increase sediment
Table 7.2. Hypotheses concerning deposit or mixed benthic

macrofaunal communities which are referable to the special theory of
trophic group mutual exclusivity.
No. Hypothesis Reference

1 Ho Sestonic food does not become Wildish and
limiting above a suspension feeding Kristmanson (1979)
bed
H, Downstream seston depletion may
occur above a suspension feeding
bed
2 Ho Suspension feeder production is Wildish and Peer
unrelated to U or potential for (1983)
turbulent and advective transfer Wildish and
H, Suspension feeder production is Kristmanson (1985)
positively related to U or potential
for turbulent and advective transfer
3 Ho Suspension feeder production is Wildish and
unaffected by energetic water Kristmanson (1979)
movement Warwick and
H, Suspension feeder production in Uncles (1980)
soft sediments can be impoverished
by high energetic movement
4 Ho Deposit feeder production is Rhoads (1973)
unrelated to flow velocity Rhoads et al. (1978)
H, Deposit feeder production is
inversely related to the erosional
power of flow
5 Ho Population density does not affect Wildish and
seston uptake or population Kristmanson (1979),
growth of suspension feeders Broom (1982)
H, Population density determines the Frechette and
seston uptake or population growth Bourget (1985)
rate of suspension feeders
6 Ho Suspension feeder reef path length Wildish and
is unrelated to potential for Kristmanson (1979)
turbulent and advective transfer
or tidal excursion length
Suspension feeder reef path length
is determined by potential for
turbulent and advective transfer
and tidal excursion length
WATER
MOVEMENT
SEDIMENT Silt/Clay Net Net Erosional Highly Erosional Rock

Depositional
SEDIMENT
PROCESSES
Bioturbation Burrown
Deposition
^ I. > / , "^Solutes
Physicar _ A I . v. A\
resuspensiqn ôtfbation
.;.;. .1 E r o s i o n - d e p o s i t i o n
cteria cycle
Fa
BENTHIC
W/////////////A Suspension
COMMUNITY Deposit Mixed Mixed Mixed Feeding
BENTHIC
PRODUCTION
Deposit --^"^ Suspension
Figure 7.1 Diagrammatic representation of benthic communities of the continental shelf along a gradient of
increasing water movement energy.
water content, and thereby increase the possibility of physical

resuspension. The degree of bioturbation activity, coupled with benthic
shear stress, were considered by Davis (1993) to be most important in
determining the type of deposit feeding animals present. Deposit feeders
are found on all types of soft sediment, although they reach their highest
production levels where water movements are minimal (Fig. 7.1) and
where gravity settling mechanisms dominate advective flux. In such con-
ditions, a silt-clay sediment is present and active bioturbation by burrow-
ing deposit feeders is high, resulting in resuspension of the deposits, with
subsequent redeposition close to the original bioturbaters. Those deposit
feeders associated with net erosional sediments cannot exploit
bioturbation because most of the resuspended material would be trans-
ported away. Deposit feeders depend trophically on the microbial flora
attached to sedimentary particles (Newell 1965; Levinton and Lopez
1977; Jumars and Nowell 1984a). The microbial bacteria involved are
aerobic forms limited by substrate and oxygen availability within sedi-
mentary deposits. But on resuspension by biological or physical means
the bacteria can grow and multiply (Nickels, King, and White 1979;
White et al. 1979; Anderson and Meyer 1986; Muschenheim, Kepkay,
and Kranck 1989; Ritzrau and Graf 1992) to increase their contribution
to an enhanced food supply for deposit feeders. Because of the presumed
energetic cost of bioturbation, it would be wasteful to resuspend sedi-
ment in this way, only to have it carried away by flow. On highly ero-
sional sediments (Fig. 7.1), only a few deposit feeders are found; they are
specially adapted for deep burrowing to escape the periodic sediment
erosion which is characteristic of these sediments.
By contrast, suspension feeders require a higher rate of water move-
ment (Fig. 7.1), and their biomass and productivity depend positively on
tidal or wind-wave activity. In soft sediments, suspension feeders tend to
bind the sediment to prevent washout and produce at rates related to the
turbulent seston supply. The ability of benthic organisms to de-stabilize
or stabilize sediments was considered by Jumars and Nowell (1984b),
who concluded that no classification was yet possible because of lack of
data. Flume experiments with a wide range of benthic organisms in
realistic simulated sediment-water interface conditions are required, but
not yet available, to achieve a satisfactory empirical classification.
For soft sediments there is an upper limit beyond which seston flux
cannot further increase productivity because energetic water movement
may result in feeding inhibition and eventual washout and impoverish-
ment. Some suspension feeders, e.g. barnacles with specialized attach-
0.001 0.005 0.01 0.050.1 0.5 1 5 10 50 100

PARTICLE SIZE, mm
Figure 7.2 Relationship between current velocity and its ability to erode, trans-
port, and deposit sedimentary particles of various sizes (after Perkins 1974).
ment mechanisms, may be able to survive on stable hard substrates (Fig.

7.1) at extremely high flows.
Specific hypotheses referable to the trophic group mutual exclusion
theory which concerns suspension feeders are dealt with in this chapter,
as shown in Table 7.2, and are considered in the order shown. It is
important in developing the mutual exclusion concept to determine what
constitutes food and to understand the mechanism of capture. However,
it is not necessary to use subcategories in trophic group typing only to
determine whether or not the benthic organism is suspension or deposit
feeding. Sanders (1960) and Rhoads and Young (1970) were among the
first to stress the fundamental importance of suspension and deposit
feeding in food web process studies. Inputs of food reach the benthic
boundary layer and are transferred by advection and diffusion to suspen-
sion feeders or transferred as sedimentary particles after settlement,
which are ingested by deposit feeders.
The relationship between benthic communities and substrates in Fig.
7.1 is really more complex than shown because it is medium sand-sized
particles (250- to 500-|um median particle diameter) which are suscepti-
ble to erosion at the lowest velocities (Fig. 7.2). By contrast, clay-sized
particles (4- to 62-jim median particle diameter) do not erode unless the
velocities reach nearly an order of magnitude greater than for medium
sand-sized particles. Depositional rates, on the other hand, are faster for
sand- than for clay-sized particles because the larger, more dense sand
particles settle at a greater rate than smaller, lighter clay particles. Be-
cause of biological activity involving binding of surficial sediments
(Madsen, Nilsson, and Sundback 1993), the figured critical sediment
erosion velocities in Fig. 7.2 are approximate.
Hydrodynamic modelling of suspension feeding

populations
The earliest attempt to apply a turbulent mass transfer model to the
problem of seston supply and uptake by a bivalve reef was made by
Wolff et al. (1976). Wildish and Kristmanson (1979) independently de-
veloped a turbulent mass transfer model very similar to the earlier one.
In this formulation, the rate at which seston is supplied to the reef is
assumed to be limited by turbulent convection near the top of the reef.
This rate is defined as NA (g seston/h^m2 of reef). The consumption of
seston per unit area of reef is given by
NA = PRaC (7.1)
where
C is the seston concentration near the inhalant opening, or filtration
surface, g m3
P is the density of suspension feeders, number m"2
R is the average filtration rate of each individual in the population,
m^h"1
a is the average filtration efficiency (0% to 100%) of each individual
in the population
The term PRa is equivalent to the more familiar term clearance rate,
defined as cubic metres cleared per hour per square metre of reef. It has
the units of velocity, m h"1. The model shown in (7.1) does not deal with
seston deposition or resuspension. The flux is driven only by the suspen-
sion feeding of the benthos, which creates a concentration deficit near
the reef, and turbulent transport. Assumptions necessary are that the
suspension feeder reef is of semi-infinite extent; that the benthic bound-
ary layer is well developed, that is, that the velocity and concentration
profiles do not change in the downstream direction; and that there is no
mass transfer resistance attributable to a laminar sublayer.
The limiting turbulent transport rate can be estimated by assuming

that seston is transported by the same mechanisms as is momentum.
Bottom shear stress data as in Sternberg (1968) can be used to develop
a flux equation:
NA=yU(C0-C) (7.2)
where
Co is the upstream seston concentration, g m3

y is a dimensionless hydrodynamic term approximately equal to 0.003
and dependent on roughness
U is the velocity at the top of the benthic boundary layer, m hf1
Combining (7.1) and (7.2) we obtain
V ;
C yU
where the left-hand term is a measure of seston depletion. The right-

hand term in (7.3), which we will call the seston depletion index (SDI), is
a useful indicator of whether depletion is an important factor in a par-
ticular locality. The SDI is dimensionless and represents the ratio of the
total clearance rate of all suspension feeders occupying a unit area of
bottom to the rate at which turbulence can deliver seston to them.
Equation (7.3) can also be written as
^ = 1 + SDI (7.4)
SDI can be calculated for a particular site if the total clearance rate can
be estimated and typical velocities of the bulk flow are known. If "signifi-
cant" seston depletion is defined as C < 0.9C0, then it can be expected to
be observed if SDI > 0.11.
Data for sublittoral mussel reefs suggest that depletion effects can be
expected in the field. J0rgensen (1990) provides clearance rate data for
natural blue mussel reefs of the highest density found in the North Sea
and Canada as from 6-12m-h"1. Dare (1976) reported that a blue mussel
reef in Morecambe Bay, United Kingdom, of 1.4kg of soft tissue per m2
of reef had a total clearance rate of 12 m-h"1. Using this value for PRa
and y= 0.003, SDI can be calculated for a range of velocities (Table 7.3).
These results show that the highest density mussel reefs will deplete the
benthic boundary layer of seston at typical tidal velocities. Other
Table 7.3. Dependence of SDI on velocity for mussel reefs at high

density (1.4kg dry body weight/m2).a
U, cm-s"1 5 10 20 50 100 500

SDI 22 11 5.5 2.2 1.1 0.2
a
For the sublittoral of Morecambe Bay, UK, at maximum clearance rate of
12 m-IT1.
Source: Data reported by Dare (1976).
sublittoral mussel reefs show similar total clearance rates (J0rgensen

1990). Values found by Smaal et al. (1986) in the Oosterschelde, Nether-
lands, were lower: a maximum total clearance rate per square meter for
cultured reefs of blue mussels of 0.58 m - h 1 and for cockles Cardium
edule of 0.2 m-tf 1 . Under the Oosterschelde conditions, water samples
showed depletion of seston over the mussel reef, but not the cockle reef.
The preceding analysis is based on the idea that a balance has been
achieved between the two processes of turbulent mass transfer of seston
and its consumption at the boundary. This condition is presumably satis-
fied if the population of filter feeders is evenly distributed over a large
area and the analysis is applied at a point well downstream of the leading
edge of the reef. Upstream of this hypothetical point, the two processes
may not be in balance and the concentrations of seston near the bottom
will vary with depth and distance from the leading edge of the reef. The
distribution of seston (the "concentration boundary layer") will be as
shown in Fig. 7.3. According to Fig. 7.3, upstream mussels will receive
more seston, and, therefore, growth rates will be higher than for down-
stream ones where seston depletion is well developed. At locations
where tidal flows regularly reverse, e.g. in estuaries, growth of individu-
als will be highest at the edges rather than the middle of the reef as a
result of the seston depletion effect. O'Riordan, Monismith, and Koseff
(1993) have presented a concentration boundary layer study based on a
reef of model bivalves (625 animals m~2 simulating Tapes japonica) in a
recirculating flume at realistic pumping rates to characterize the seston
depletion effect at different points in the reef (see also the section,
Infauna, Chap. 6).
To predict under what circumstances seston depletion will become
important in controlling population filtration, and hence growth, a solu-
tion of the mass conservation equation is required:
C(x,o)
WMMmMMMM/W/MM/Mm
x=o
77/7T/
Figure 7.3 The concentration boundary layer which develops over a blue mussel reef in a unidirectional flow:
Z, water depth; X, mussel bed length; Co, initial seston concentration.
dc TT dc d ( dc | . , .
+ (7 - sinks + sources = 0 (7.5)
dt dx dzy dz)
where
c is the concentration of seston
U is the velocity at (x, z) and
Em is the vertical mixing coefficient
Equation (7.5), called the convective-diffusion equation, is of time aver-
aged form and does not consider changes in the cross-stream direction.
The first term represents temporal change in concentration of seston at
the point (x, z). The second term represents the change of advection of
seston by the average flow in the downstream direction. The third term
describes the vertical diffusion of seston due to turbulence. Several au-
thors, including Wildish and Kristmanson (1984), Smaal et al. (1986), and
Frechette, Butman, and Rockwell Geyer (1989), have proposed solu-
tions for the convective-diffusion equation.
Wildish and Kristmanson (1984) suggested an analysis of a developing
flume boundary layer in which the velocity boundary layer thickened in
the downstream direction so that velocity profiles could not be assumed
to be fully developed. The concentration boundary layer was assumed to
be well mixed in the vertical direction and to have the same thickness as
the velocity defect layer. The mean concentration in the boundary layer
was calculated by integrating the differential mass balance. This analysis
was extended by Verhagen (1986) and Smaal et al. (1986), who provided
an analytical solution and introduced a more realistic treatment of the
seston within the boundary layer. For the case of seawater >3 m deep
where steady unidirectional flows prevailed, using realistic profiles of
mean velocity and vertical eddy diffusivity, Verhagen (1986) also pre-
dicted the seston depletion to be expected at various downstream dis-
tances along the reef (Fig. 7.4). The abscissa can be considered as the
ratio of the velocity generated by the mussels (the number of mussels per
unit reef area times the volumetric flow per mussel) to the mean stream
velocity. It is therefore similar to the ratio SDI earlier proposed as a
measure of the tendency to depletion. An analytical solution is also
provided:
Co U (iDj
0.6- x = 1000m
0.4-
O x = 500
\-
LLJ
a.
LJJ
Q 0.2-
= 333
1.0x10"3
Figure 7.4 Seston depletion as a function of PRalV (Verhagen 1986), where X

is the mussel reef path length in meters.
where
x is the distance from the upstream boundary of the reef
C(x, 0) is the seston concentration at the inhalants
Dx is the vertical eddy diffusivity near the bottom
The units of the turbulent flux term are as in (7.1) and (7.2).
This solution is limited to small values of NRa5/D1 where 8 = seston
boundary layer thickness. Note that the depletion is directly propor-
tional to the product of two dimensionless groups, the abscissa of Fig. 7.4,
which depends on the population density and the current speed, and the
square root term, which is of the same form as the Peclet number defined
in Chapter 5 to express the ratio of advective transport to diffusional
transport. Depletion occurs when seston is moved rapidly from the
boundary by pumping, and smearing by turbulent diffusion cannot
smooth out the gradients thus created. The Peclet number term is deter-
mined by the mechanics of the flow and will become smaller if the
DISTANCE FROM MUSSEL BED LEADING EDGE,m

0 20 40 60 80
Figure 7.5 Model generated contours of seston concentration in a unidirec-

tional flume flow in arbitrary fluorescence units: conditions: U=15cTn-s~1 at
Z = 2.25 m; Z0 = 0.1cm, t/, = 0.78 cm-s"1; mussel density = 430 m"2, dry weight
biomass = 497g-nT2, bed length - 100 m; mussel inhalant height above the bot-
tom is 3 cm (from Frechette et al., 1989).
boundary is made rougher. Verhagen (1986) also provided a solution of

the shallow water (<3 m deep) case in which it can be assumed that the
seston is well mixed in the vertical direction and the z-dimension can be
neglected.
Frechette et al. (1989) solved the convective-diffusion equation nu-
merically to analyze the vertical and horizontal distribution of natural
seston over a blue mussel reef. The mean velocity profile was developed
by using field measurements, and the vertical eddy diffusivity was as-
sumed equal to the eddy viscosity. Their treatment thus differs from that
of Verhagen, who assumed that depletion near the bottom was domi-
nated by the mixing produced by the exhalant streams from the mussels.
The sinking rate of phytoplankton was assumed to be small enough to be
ignored. The rate of consumption of phytoplankton by the mussels was
estimated from physiological chamber tests and from measurements of
the size-frequency distribution and total biomass of mussels in the reef.
The unit area clearance rate found in this way was 6 m -h"1. Their results
for a model reef similar to their field site are shown in Fig. 7.5. In this
simulation, the phytoplankton is represented by equivalent fluorescence
units, with the upstream value 300. The velocity at 2.25 m depth is
0.15 m-s"1. The effect of depletion is clearly shown. Time series mea-
surements at a sampling station near the centre of a mussel reef about
100 m long in the axis of the flow showed seston depletion at a depth of
0.05 m of approximately 50% at similar current speeds. Although the
model represents a simplification of the field conditions, the extent of the
agreement between the model and the fluorometric measurements is
impressive.
Seston depletion effect

An early demonstration of seston depletion was provided by Glynn
(1973a, b) over a Caribbean coral reef made up of patches 3-50 m wide
and where the depths ranged from 1 to 1.5 m. The purpose of this study
was to test quantitatively whether the reef removed substantial amounts
of seston to determine whether the reef relied mainly on symbiotic
photosynthesis or suspension feeding of plankton as the major contribu-
tor to its energy needs. Wind-wave energy moved seawater across the
Porites reef flat and resulted in effective vertical mixing. Surface samples
were taken at the upstream edge, on the reef, and at a downstream
position landward. Glynn (1973b) was able to show, by weekly sampling
at the same three locations across the reef for a 2-year period, that
numerically diatoms could be reduced by 91% and zooplankton by 60%
in passing across it.
Buss and Jackson (1981) designed a field experiment to measure
seston depletion of seawater passed through a 2.4-m-long, flowthrough
box by aligning the open ends in the mean flow direction in Discovery
Bay, Jamaica, at 40 m depth. Each box contained panels on which various
densities of bryozoans and sponges were allowed to grow during the 26-
mo-long experiment. Upstream and downstream seawater samples were
taken simultaneously through corked holes in the upper wall of the box
by SCUBA divers. Samples taken with a 25-ml syringe were analyzed for
densities of bacteria and phytoplankton (mostly flagellates). Measure-
ments of seston depletion were made at different times during the ex-
periment. Because different percentages of cover had developed on the
test panels at these times, suspension feeder density was varied. Results
suggested that percentage cover by sponges and Bryozoa was the pri-
mary factor determining seston depletion, and if these two variables
were plotted on logarithmic coordinates, they were linearly related.
During this experiment, equal numbers offlowboxes were set up so that
one end was blocked off in one series of boxes as controls, and in the
treatment boxes both ends were left open to test the effect of flow.
In the first study to use adequate hydrodynamic field measurements,

Frechette and Bourget (1985) showed that an intertidal blue mussel reef
in the St. Lawrence estuary could produce seston depletion effects within
the benthic boundary layer. By direct sampling at 3 to 4 cm height above
the mussel reef, fluorometric measurements showed that seston deple-
tion during the first week of observation was negatively correlated with
current velocity. During the next week, seston depletion was only ob-
served for half of the sampling time and was negatively correlated with
wave energy. Frechette and Bourget (1985) thus observed a temporal
change in the presence/absence of a seston depletion effect, depending
on the seston concentration as well as the type of water movement
energy available. During the first week, seston depletion was associated
with tidal flows; during the second week, this association was partially
destroyed by wind-wave-caused oscillating flows which resulted in
resuspension and better mixing.
In a field study on an intertidal sand flat in Shark Bay, Western
Australia, the ability of infaunal suspension feeding bivalves to deplete
the flooding tidal seawater was tested by Peterson and Black (1991).
Seawater samples were pumped, approximately isokinetically, from 5 to
10 cm above the bottom over a 205-min period in the same water mass
(indicated by rhodamine B dye marker). Chlorophyll a had declined by
25% during this period.
A few laboratory experiments have also been designed to demonstrate
that populations of suspension feeding animals can deplete the seston as
it passes over the reef, as is required by all hydrodynamic models dis-
cussed in the section, Hydrodynamic modelling of suspension-feeding
populations. Wildish and Kristmanson (1984) used an outdoor flow-
through flume of 5 m length which was supplied with unfiltered seawater
from a submersible pump. Seasonal seston concentrations and tempera-
tures in the Bay of Fundy were not modified in the flume during the
experiment except by the feeding activities of mussels. Sampling was
from 1 cm above the flume bottom at the inlet end and 3.6m away at the
downstream end of the mussel reef. Sampling was approximately
isokinetic since the pumping rate used was adjusted to be isokinetic at
4cm-s"1. Results suggest that significant ATP-seston depletion occurred
in six of eight experiments (Table 7.4), whereas for bacterial numbers
only three of eight experiments showed significant depletion.
Experiments undertaken by Butman et al. (1994) in a large
recirculating flume containing a 5-m-long reef of blue mussels showed
that seston depletion could be detected in one experimental run al-
though not in two others. For the experimental run showing seston
Table 7.4. Depletion of seston from natural seawater by adult horse

mussels Modiolus modiolus (Expts. 1-3) and blue mussels Mytilus
edulis (Expts. 4-8) over reef of 3.6m path length.*
Bacterial
Bulk flow Temperature Mussel density ATP-seston nos. %
Expt. cm s"1 C number m2 % depletion15 depletion
1 3.92 7.5 164 -36 C -39C

2 3.78 7.7 164 -80 c -39 C
3 3.35 7.8 164 +3 -53 C
4 4.10 13.6 2123 -58 C +7
5 4.47 12.8 2123 -46 C -36
6 3.20 12.7 2123 -81 C -23
7 4.18 12.2 1062 -16 +11
8 3.84 10.9 1062 -84C
a
Wet biomass 3.672kgm" 2 (Expts. 1-3), 4.848kg-m"2 (Expts. 4-6), 2.424kgm" 2 (Expts.
7-8).
b
ATP, adenosine triphosphate.
c
P<0.5.
Source: Wildish and Kristmanson (1984).
depletion, the vertical distribution of seston, as chlorophyll a, was as

predicted by the model of Frechette et al. (1989). Unfiltered seawater
and natural seston were used in these experiments with high mussel
densities and biomasses. Low concentrations of natural seston may be
the explanation for the low filtration activity of the mussel reefs in the
experiments in which seston depletion was absent.
Even though the experimental flume results discussed are for a devel-
oping boundary layer, the SDI measure appears to be robust enough to
give reasonable values (Table 7.5), which are all >1, indicating that
seston depletion is feasible. This is in agreement with independent con-
centration boundary layer measurements, which were made in both stud-
ies shown in Table 7.5.
Flow and suspension feeder growth and production

The study of energy exchanges which occur between communities or
populations of suspension feeders at the sediment-water interface and
the water column can involve a detailed understanding of consumption,
absorption, excretion, assimilation, production, respiration, and gonadal
output (Crisp 1984). In this account, we have avoided the detail and used
Table 7.5. Free-stream velocities and calculated SDI values for some
flume experiments with mussels.^
Biomass, u,
Species kg A F D W m " 2 cm s"1 SDI Reference
Mytilus edulis 0.245 8 3.3 Butman et al. (1994)

Mytilus edulis 0.245 19 1.3 Butman et al. (1994)
Mytilus edulis 0.243-0.485 3.2-4.5 3.3-8.7 Wildish and
Kristmanson (1984)
Modiolus modiolus 0.367 3.4-3.9 2.1-2.5 Wildish and
Kristmanson (1984)
a
The following were used in the calculations: 7?=2.83L-h x for M. edulis and
R = 10.98 L-h" 1 for M. modiolus (Wildish and Kristmanson 1984, based on
J0rgensen 1990); y=0.006 (Butman et al. 1994); wet weight converted to
ash-free dry weight (AFDW) by multiplying by 0.1.
simple measures in describing the energetics of suspension feeders as

follows:
Biomass, or amount of living substance measured as wet or dry weight per
unit area (g-nT2). The wet measure uses crude units inclusive of non-living
tissues such as valves. Biomass is here used synonymously with "standing
crop" or "standing stock."
Production, the rate of change of biomass from that part of the assimi-
lated food that is retained and incorporated in the biomass of the population
(g-m^-y 1 ). Our measure does not separate gonad from somatic growth. We
use short-term direct individual growth rate estimates (g-d"1) rather than
actual observations of population growth. The latter involves integration of
survivorship-biomass curves for each year-class (Crisp 1984).
Model predictions from the previous section suggest that seston deple-
tion above a suspension feeder bed will be affected by turbulent transfer
processes; the most important variables are velocity, seston concentra-
tion, and bed roughness. Other variables which determine whether
seston depletion occurs are reef path length, bed density, and average
body size of suspension feeders. Diagrammatic representation of the
effect of different velocities over a mussel reef shows how population
growth may be increased if the ambient flow rate is increased (Fig. 7.6),
assuming that depth is shallow and that the benthic boundary layer
intersects the seawater surface.
In Chapter 4, we saw how some active suspension feeders filter
optimally in particular environmental conditions, e.g. velocity and seston
concentration. If these suspension feeders arefilteringat reduced rates in
Number Mean Total

mussels growth per growth
mussel
A
0
' ' *w
10 0.08 0.8
^ mmmi
10 0.05 0.5
^ T . 10 Qm o.3
Figure 7.6 Effect of unidirectional flow at different velocities on mussel growth

(arbitrary units): A, high; B, medium; C, low; mussel density and seston concen-
tration at the upstream end are constant. The arrow indicates the point at which
NA becomes limiting.
environmental conditions below their optimum, e.g. in seston concentra-

tion, then further increases in seston concentration have the potential to
increase downstream growth rate, beyond the point indicated by the
arrow in Fig. 7.6, where NA is limiting. If there is a permanent increase in
seston flux, then the point where NA becomes limiting also moves down-
stream in the reef. Considering a longer temporal scale, it may also allow
an increased reef length after immigration or settlement of new recruits.
Wildish and Kristmanson (1985) investigated the effect of current
velocity on growth rates of blue mussels in a Kirby-Smith growth tube
apparatus. Although this apparatus does not simulate the benthic bound-
ary layer as it occurs in the sea, mussels are common fouling organisms
in pipes carrying seawater. Natural seston was supplied concurrently to
eight tubes, all with seawater flowing at different velocities, with each
tube containing 10 mussels. The results (Fig. 7.7) show that growth was
asymptotic with respect to velocity and that up to ~2cm-s~1 growth
increases with velocity. Examination of individual growth showed that
CD 0 . 3 -
en
CD
Z3
E (
/
(/
c
oS /
CD
CE-0.1-
0 1 2 3 4
(7, cm s'1
Figure 7.7 Blue mussel growth rates (mean and range) as a function of velocity
(/, cm s"1) in 31-d-long concurrent experiments in Kirby-Smith growth tubes; 10
mussels per tube 364 mussels-m~2 with each mussel approximately 2.54 g wet
weight (Wildish and Kristmanson 1985); natural seawater from the Bay of
Fundy, Canada.
upstream (the first five mussels in the tube), mussels grew better than
downstream ones (Table 7.6) at flows <2cm-s~1, but at >2cm-s"\ there
was no significant difference between upstream and downstream indi-
vidual mussel growth. The theory predicts that not only velocity but also
seston concentration in relation to the composition of the suspension
feeder bed (mean size, density, and reef path length) affected average
growth, so this result is not universally valid. Thus, if the seston concen-
tration were increased or mussel density reduced, quite different growth
results could be expected.
To understand fully the ecological effects of flow on suspension
feeding populations, a wide flow range must be tested. Flow velocities
from -1-24 cm-s"1 were provided in a multiple-channel flume (Chap. 2)
with four separate channels in giant scallop growth experiments (Wildish
et al. 1987). Each channel had the same seston concentration but a
Table 7.6. Growth rates of blue mussels as a function of velocity in a

concurrent 31-d-long experiment*
Mean growth, g wet d l mussel l

Mean velocity
cm s"1 Overall, 1-10 Upstream, 1-5 Downstream, 6-10
0.10 0 0.05 -0.05

0.13 -0.06 0.02 -0.15
0.40 0.22 0.33 0.15
0.45 0.13 0.44 -0.13
0.75 0.12 0.26 -0.01
1.95 0.27 0.30 0.31
3.50 0.28 0.30 0.30
3.89 0.29 0.34 0.27
a
Natural seston at a constant density of 10 mussels per tube = 364 mussels-m 2.
different velocity during each experiment. The results of six experiments

at different seasonal times (Fig. 7.8) confirmed the unimodal response to
velocity expected from physiological experiments on filtration (Chap. 4).
At the scallop density tested in each channel of -27 adult scallops m~2,
there were no differences in growth between the inlet and outlet ends of
the flume, so a seston depletion effect was absent. This is because low
values of PRa were associated with this density of scallops.
Earlier growth results with bay scallops obtained by Kirby-Smith
(1972) can also be explained as a unimodal response to velocity, although
because of the complexity of theflowin the tube and because volumetric
flow estimates were used to calculate velocities, they are not directly
comparable. One way in which growth experiments under field condi-
tions can be misleading is if for a substantial part of the time, strong tidal
flows are directly growth inhibiting. This possibility was examined with
giant scallops in a multiple-channel flume, but with a variableflowso that
various periods of inhibitory flows (>12cm-s~1) and optimum flows
(~4cm-s"1) were present (Wildish and Kristmanson 1988). Results (Ta-
ble 7.7) suggest that if between 33% and 66% of the time the flows are
optimal, then the scallops can adjust and produce growth rates equiva-
lent to those of controls which 100% of the time are at optimum flows.
Grizzle and Morin (1989) describe a field growth experiment of the
hard clam Mercenaria mercenaria in a tidal lagoon in New Jersey, United
States. Their 3 x 3 factorial experiment was designed to test, over a
0.4
o
o
0.3 -
0.2 -
0.1 -
0-
-0.10 i i i i i ih
0.24 - B
o 0.20 -
\
o
0.16 -
-fW 0.12 - o o*-^.
0.08 -
L
0.04- \
\
0- o
10 20
U, cm s'
Figure 7.8 Giant scallop Placopecten magellanicus specific daily growth rates of
adults (60 to 75 g wet weight) determined in 25- to 33-d-long experiments at
different times of the year, as a function of velocity (U, cm-s"1). Four channels
each with 16-24 scallops^ 18-27-m~2 were run concurrently in a multiple-
channel flume using natural seawater from the Bay of Fundy, Canada: A,
W = total wet weight, g; B, L = valve height, cm; N = number of days in experi-
ment (Wildish et al. 1987).
15-wk summer growth period, the effect of sediments (mud, sand, and
undisturbed control) at three different sites in the local area. Current
speed was measured at each site for such a small portion of the experi-
mental period that it is possible that the data are not representative of
conditions during the experiment. Grizzle and Morin (1989) concluded
that, whereas sediments had no effect on clam growth, site differences
did. Because environmental variables such as tidal velocity and seston
concentration were not controlled during the experiment, we consider
Table 7.7. Effect of inhibitory flows as percentage of total

experimental time available on mean growth rates of giant scallops
Placopecten magellanicus.
Percentage daily valve

Treatment: percentage growth: valve height Percentage daily biomass
of time flow is inhibitory (cm) scallop"1 d"1 growth g wet scallop"1 d"1
100 0.053 0.01 0.37 0.02
60 0.056 + 0.01 0.36 + 0.03
33 0.104 0.01 0.41 0.02
0 0.095 0.01 0.41 0.02
that the positive relationship established between seston flux rate and
hard clam growth, based on three data points, may be fortuitous. Other
environmental variables such as seston quality, dissolved oxygen levels,
or inhibitory wind-wave effects could explain the higher growth rates at
one particular site. We concur with Grizzle and Lutz (1989) that there is
a need to determine the physiological responses of hard clams to velocity
and seston density in more controlled experiments.
Other authors (e.g. Cahalan, Sidall, and Luckenbach 1989; Wildish
and Saulnier 1993) have shown that feeding is not simply related to
sestonflux,but rather is interactively related to seston concentration and
velocity. In passive suspension feeders, too, Leonard (1989, and see
Chap. 4) found that in the crinoid he studied, the feeding responses were
to prey concentration and velocity separately.
In a 3-mo-long field experiment in the northern Gulf of Mexico, Judge,
Coen, and Heck (1992) tested the effect of seawater flow on hard clam
growth rates. The water depth was 0.6m and tidal range small (<15cm).
Wind-wave effects were an important driving force for flows at their
study locality. An open-ended four-channel flume, without a bottom,
was built in the field. The flume was 7 m long with 1.2-m-high walls, and
the flows within each channel were controlled by the width of the en-
trance. Comparative flows in each channel were assessed on one day so
that three treatments of 2% and 40% reduction, or 65% increase in
flows, were provided. Velocity was not measured continuously through-
out this experiment, although spot-check determinations suggested a
range of 0-27 cm-s"1. Six Mercenaria mercenaria were placed in the
middle of each channel in a plot of 0.25 m"2 (=24 clams m~2), but whether
other suspension feeders were present in the natural sand of the rest of
the flume was not clear. Since growth rates between all treatments and
the control were not significantly different, Judge et al. (1992) concluded
that velocity, independent of seston depletion, does not influence hard
clam growth. We believe that this conclusion is not justified by the
methods used, particularly because we know nothing of the temporal
distribution of flow throughout the experiment. This is important be-
cause of the stochastic nature of wind-wave effects and the demon-
strated ability of bivalves to adjust and compensate for the varying flow
periods (Wildish and Kristmanson 1988).
Comparative growth studies between siphonate (M. mercenaria) and
non-siphonate {Crassostrea virginica) bivalves were conducted by Griz-
zle, Langan, and Howell (1992) in a multiple-channel flume. Unfortu-
nately, only a limited range of free-stream velocities (<8cm-s~1) were
tested so the results must be regarded as preliminary. For 1991 results,
oyster growth (valve increments) was unimodally related to velocity with
a peak -lcm-s" 1 , while hard clam growth over the same flow range
showed only a and b of the unimodal response, with a broad b peak
between 2 and 4.5 cm-s"1.
Laboratory and field studies in Narragansett Bay, on the eastern coast
of the United States, were designed to determine the effect of flow,
seston concentration, and temperature on the feeding and population
biology of the acorn barnacle, Semibalanus balanoides (Sandford et al.
1994). In simple flow tank studies, the acorn barnacles were exposed to
unidirectionalflowsin the range 0-21 cm s"1 and at temperatures of 10-
25C, which were typical of local conditions. Barnacle feeding was in all
cases passive filtering, except at zeroflow,where slow cirral beating was
observed. The percentage of barnacles passively filtering was shown to
increase up to aflowof 21 cm s"1 and to peak at a temperature ~15C for
both adults and juveniles. Locally available seawater with its naturally
occurring seston resulted in variations in the percentage of barnacles
feeding so that the locality with the highest chlorophyll a values had most
barnacles feeding. These results were generally confirmed by growth
experiments in various local conditions. The study by Sandford et al.
(1994) might have benefitted from the use of an integrated environmen-
tal/physiological model for feeding and growth as suggested for the giant
scallop (Chap. 4, the section, Bivalves). The limited range of unidirec-
tional flows tested with respect to barnacle feeding and growth and
the absence of simulated wind-wave action, render laboratory results
obtained by Sandford et al. (1994) in respect to water movement

preliminary.
So far, the measures which have been used to discuss the effect of flow
are short-term growth rates. It should be realized that we cannot simply
multiply mean growth rates by population density to arrive at a satisfac-
tory secondary benthic production estimate. Not only are these changes
in growth rate seasonal in boreal climates, but positive and negative
changes to population numbers must be accounted for if an accurate idea
of annual production is to be obtained. This involves increases in density
due to recruitment (larval settlement, migration) or decreases caused by
mortality or emigration. It is these population changes which allow
benthic suspension feeder production to adjust to the several orders of
magnitude difference in carrying capacity between local habitats. There
is some evidence that larval settlement occurs preferentially where reefs
of suspension feeders are already established and that migration by
juveniles to adult reefs may occur (Chap. 3).
Emigration of adult Sanquinolaria nuttalli, a deep-burrowing
suspension feeding bivalve of the California coast, was noted after trans-
plant experiments involving high densities of two other deep-burrowing
species (Peterson and Andre 1980). Mortality rates of the blue mussel
Mytilus edulis were reported by Incze et al. (1980) from a Maine, United
States, estuary during a period of high ambient temperatures. The deaths
were thought to result from a sudden crash in the phytoplankton popu-
lation and hence reduction in ration coincident with metabolic stress due
to high temperature.
We have another way of demonstrating that velocity is a key variable
in controlling suspension feeder production: the use of field-observed
standing crops or crude secondary production data calculated from
them. The results of benthic sampling by grab or corer can be expressed
as the species number, as biomass (g wet-mf2), or, if an estimate of the
P:B ratio for each species is available, as crude secondary production
(g wet-m^-y"1) by methods described in Wildish and Peer (1983). For
five estuaries and at seven stations within the Bay of Fundy, Wildish and
Kristmanson (1979) partitioned benthic biomass into deposit and sus-
pension feeding (Fig. 7.9). Deposit feeding biomass was inversely related
to mean tidalflows,whereas suspension feeding biomass was a unimodal
function of velocity. This field result is in general agreement with the
model proposed in Fig. 7.1.
Purdy (1964) reanalyzed H. L. Sanders's benthic data from Buzzards
Bay and Long Island Sound showing that the proportion of deposit
100-
50-
CO
Q 10-
DQ
5^
10 20 30 40 50
U av.
Figure 7.9 Deposit feeder () and suspension feeder (o) biomass (gwet
weight m"2) at seven estuarine or nearshore locations within the southwestern
Bay of Fundy, Canada (Wildish and Kristmanson 1979): x-axis, mean current
velocity as cm-s"1 over one tidal period at 1-2 m above the sediment-water
interface.
feeder biomass was positively related to percentage of clay in the sedi-

ment samples, while the proportion of suspension feeders was inversely
related to the percentage of silt and clay in the sediment. These field
results are also generally supportive of the concept shown in Fig. 7.1.
The suspension feeder production in the lower Bay of Fundy - mostly
horse mussel, Modiolus modiolus - was positively correlated with veloc-
ity (Wildish and Peer 1983). The mean tidal velocity at each station was
determined from a numerical model of circulation in the Bay (see
Wildish and Peer 1983). From a subset of these data, reanalysis by
Wildish and Kristmanson (1993) showed that the important variable
controlling the greater than three orders of magnitude in production was
velocity. Velocity, U in m-s"1, predicted production in the lower Bay of
Fundy as
Log10 benthic production, g m 2 y x = -1.155 + 5.862 [/,
n = 29, r = 0.85, p>99% (7.7)
Table 7.8. Bay of Fundy suspension feeder

production for the equation log10 x=a+by.a
Location a b r N
Lower Bay -1.38 0.06 0.84 33

Minas Basin 1.42 0.01 0.36 34
ChignectoBay -0.71 0.02 0.25 35
a
x = g wet wt m"2 yr"1; y = mean velocity, cm s"1.
Source: Wildish et al. (1986).
by omitting 6 stations with reduced, or no, suspension feeder production

from the total of 35 available. In a continuation of the field work in the
upper Bay of Fundy - primarily Chignecto Bay and Minas Basin -
Wildish, Peer, and Greenberg (1986) found that the relationship for
these upper Bay stations became non-significant (Table 7.8). Despite this
clear indication that the current speed alone was not a sufficiently com-
plete indicator of production in the Bay of Fundy for suspension feeders,
this finding does circumstantially support the views of many earlier bi-
ologists (e.g. Kerswill 1949) that production was influenced by water
movements.
Wind-wave energy sufficient to cause sediment resuspension could be
one way in which particles of nutritive value to suspension feeders could
replenish depleted layers within the benthic boundary layer. Frechette
and Grant (1991) conducted a field experiment with blue mussels
Mytilus edulis to test the relative importance of seston depletion and
resuspension events on growth rates at two intertidal locations, 0 and 1 m
off bottom, in the St. Lawrence estuary. Growth rates were estimated in
two ways: by tissue weight changes over time and by scope for growth
predictions based on filtration rates and calculations performed on calm,
windless days. The former measure should integrate growth due to
seston depletion and resuspension, while the latter should yield pre-
dicted growth rates free of resuspension events. Growth responses at
both sites showed that summer growth rates were not augmented by
resuspension phenomena.
Bock and Miller (1994) showed in a field study on an intertidal sand
flat in Delaware, United States, where water movement was dominated
by wind-wave effects, that growth of Mercenaria mercenaria was corre-
lated with the percentage of organic matter in wave-resuspended seston
Mean
Number growth per Total
mussels mussel growth
20 0.02 0.4
wwm/f,
10 0 0 5
/// //
Figure 7.10 Effect of mussel density on growth (arbitrary units):flowvelocity

and seston concentration are equal at the upstream ends of A, B, and C; arrow,
point where NA becomes limiting.
rather than with total seston concentration. The disparity between this
and the previous results suggests a species-specific difference due to
resuspended sediment and its effect on bivalve growth.
Effects of population density

The effect of density on suspension feeder growth predicted by hydrody-
namic models of suspension feeding is shown diagrammatically in Fig.
7.10. The model applies where the benthic boundary layer intersects the
surface and turbulent transfer rates near the suspension feeder become
limiting to growth along a mussel reef, as occurs in pipe orflumeflows.
Mean growth rate per animal is inversely related to density.
Experimental results in a 7-cm-diameter Kirby-Smith growth tube
(Wildish and Kristmanson 1985) at various densities of blue mussels
Mytilus edulis confirm this prediction at least at one low flow speed,
0.09cm-s~1 (Fig. 7.11B). At a higher velocity - l^cm-s" 1 - at all densi-
ties tested and at fixed path length of 1.5 m for this experiment, all
0.7 A
el" 0.6-
CO
LO
Z3
E 0.5-
day1
0.4-
"S
0.3-
r
0.2-
oDC
CD
200 400 600 800 1000

2
mussels m'
Figure 7.11 Blue mussel growth rates, mean and range, as a function of density
in 34-d-long concurrent experiments in Kirby-Smith growth tubes. Natural
seawater from the Bay of Fundy, Canada: A, 1.44CH1-S"1; B, 0.09cm-s"1 (Wildish
and Kristmanson 1985).
mussels grew at the maximum possible rate, suggesting that no seston

depletion occurred, given the seston concentration available (Fig.
7.11A). The laminar flow speeds chosen for this experiment, however,
render the results strictly outside the theory, which refers to turbulent
flows.
Broom (1982), in field studies of a locally exploited cockle, Anadara
granosa from Malaysia, showed that growth rates, as mean wet weight
per square metre, are inversely related to the log of density. Field experi-
ments in which Anadara of different year-classes were stocked at differ-
ent densities confirmed that density did influence growth rates. A
number of bivalve density manipulation experiments all showed that the
percentage of growth relative to a low density control decreased in
inverse proportion to the stocking density (Table 7.9). Study localities
include the U.S. West Coast (Peterson 1982) and East Coast (Eldridge,
Table 7.9. Field experimental studies of bivalves in which densities were

manipulated, showing individual valve growth rates inversely
proportional to density.
Maximum
Density range, Bed path growth
Species number m1 length, m decrease, % Reference
Andara granosa 125-2500 10.0 ?-70 Broom (1982)
Protothaca staminea 5-96 1.0 38-49 Peterson (1982)
Chione undatella 0.5-48 ]L.O 38-49
Circe lenticularis 5-30 ]L.O 40-90
Anomalocardia 5-30 1L.O -70 Peterson and Black
squamosa (1987)
Callista impar 10-60 ]L.O 10-30
Katelysia scalarina 20-160 ]L.O 35 Peterson and Black
(1988)
K. rhytiphora 20-160 1.0 40
Mercenaria 290-1159 0.6-1.2 50 Eldridge et al. (1979)
mercenaria
M. mercenaria 10-80 1.0 18 Peterson and Beal
(1989)
Eversole, and Whetstone 1979; Peterson and Beal 1989) and Australia
(Peterson and Black 1987, 1988).
Field experiments involving intertidal blue mussels Mytilus edulis have
shown that for patches of mussels, the individual "edge" mussels grew
faster than the control ones (Okamura 1986; Newell 1990). Here, the
oscillatingflowsof the shore enable "edge" mussels to be upstream for at
least some of the time, while the central mussels experience the seston
depletion effect and thus grow at slower rates.
Water movement impoverishment

In a conventional benthic grab sampling study of Bay of Fundy estuaries,
limited to soft sediment and depths of 8-35 m, referred to earlier, Wildish
and Kristmanson (1979) described three distinct communities: mixed,
deposit, and impoverished. The basic distinction is between mixed and
deposit feeding; thus the mixed community consists of both suspension
and deposit feeders, whereas both other types contain no suspension
feeders at all. The characteristic feature of an impoverished community is
that it has high water velocities and high turbulentfluxesof seston, yet no
suspension feeders (see Fig. 7.9). Subsequently, Wildish and Peer (1983)
described specialized species of deposit-feeding benthic macrofauna -

usually deep burrowers - capable of resisting sediment erosion by tidal
or wind-wave action. In the Bay of Fundy, deep-burrowing polyc-
haetes such as Spiophanes bombyx, Scoloplos armiger, and Scolelepis
squamatus; a rapid-burrowing amphipod, Psammonyx nobilis; and a mas-
sively built sand dollar able to burrow, Echinarachnius parma, are exam-
ples of species associated with water movement conditions which cause
impoverishment of suspension feeders. Specialized water movement-
impoverished macrofaunal communities of soft sediments are generally
of low biomass and production, e.g. in the Bay of Fundy (Wildish and
Peer 1983). The few species involved are deposit feeders specialized for
high energy niches and also do not conform to the exclusion principle
outlined in Fig. 7.1. Some field sampling evidence supporting this charac-
terization is shown in Fig. 7.9.
In their field study of benthic macrofauna in the Bristol Channel,
Warwick and Uncles (1980) related classically defined communities to
sediment bed stress. Bed shear stress, x0, for each location was deter-
mined from
XO=|YP^L (v.8)
where y is a dimensionless sediment friction coefficient and p is the

density of seawater, usually expressed in units of dynes per square cen-
timetre. The maximum tidal current velocities of stations within the
Bristol Channel were obtained from a numerical model. Results sug-
gested that each community was associated with a particular set of bed
stress values so that a map of bed stress values closely followed that of
the independently defined communities. Associated with the highest
shear stresses of 22 dynes-cm"2, Warwick and Uncles (1980) describe a
reduced soft bottom community similar to impoverished soft bottom
communities of the Bay of Fundy. Use of x0 as the independent variable
in this study is similar to the use of mean velocity by Wildish et al. (1986),
as described earlier, with the exception that this study made no allow-
ance for bottom roughness implicit in the term y.
One type of water movement which has not yet been discussed is direct
wind-wave activity on benthic communities. Wind effects are generally
discrete, stochastic events which may have direct influence on the bottom
at depths down to ~100m (Emerson 1989). As for tidal currents, we can
expect both positive effects of wind-wave activity on benthic production
as well as negative ones, e.g. storm damage to benthic communities (Yeo
and Risk 1979; Wildish 1980; Dobbs and Vozarik 1983; Shanks and
Wright 1986; Peterson et al. 1989; Nehls and Thiel 1993). Emerson (1989)
presented a statistical study of published secondary benthic production
data which included biotic (phytoplankton primary production, benthic
primary production, meiofaunal, macrofaunal, and total benthic second-
ary production) as well as physical variables (temperature, salinity,
depth, tidal height, sediment type, wind stress, and two wind shelter
indices). From 201 published data sets, a Pearson correlation matrix
determined that wind stress was weakly inversely correlated with total
(macrofauna, meiofauna, microflora) and macrofaunal secondary pro-
duction (r2 = 0.32 and 0.12, respectively). Benthic production was nega-
tively correlated with wind stress, suggesting that the overriding effect of
this variable in the benthic environment is negative, i.e. that increasing
wind stress at a locality results in decreased production. Further multiple
regression analysis on the log-transformed benthic secondary production
data (Emerson 1989) showed that wind stress, tidal height, wind shelter
indices, and seawater temperature explained -90% of the variance in
total secondary benthic production. This a posteriori analysis suggests
that water movement variables are major controls on benthic energy
flow, as is postulated in the trophic group mutual exclusion theory.
A spatial survey of the benthic macrofauna (biomass, numbers of
species) of high energy sediments on the coast of South Africa, inclusive
of intertidal and nearshore sublittoral regions, could be interpreted as
benthic limitation by water movement (McLachlan, Cockcroft, and
Malan 1984). In the wave breaking region of the high energy surf zone
transect, the extreme turbulence prevents any macrofauna from becom-
ing established, and it is only when more stable sediment-water inter-
faces occur that macrofaunal biomass and species numbers increase.
Emerson and Grant (1991) have observed one way in which wind-
wave activity could affect production on intertidal soft shell clam {Mya
arenaria) flats. Juvenile clams were eroded and transported across the
flats coincident with peaks of bedload transport associated with wind-
wave activity.
Biological interactions in mixed benthic communities

The trophic group amensalism hypothesis of Rhoads and Young (1970)
suggested that deposit feeding infauna exclude other species, particularly
suspension feeders, by their feeding and burrowing activities, which also
result in resuspension of sedimentary particles. High sedimentary
particulate loads cause clogging of the filtration mechanism of suspen-

sion feeders and hence results in reduced growth and/or increased mor-
tality rate. Posey (1990) has reviewed the scant evidence pertaining to
trophic group amensalism. The deposit feeding ghost shrimp Callianassa
sp. in tropical lagoons caused reduced growth rates of suspension
feeders, such as corals or polychaetes (Aller and Dodge 1974), by its
bioturbating activities. Another species of Callianassa present in a Cali-
fornia lagoon apparently restricted the introduced suspension-feeding
hard clam Mercenaria mercenaria by its bioturbating activity (Murphy
1985). Field and laboratory experiments undertaken by Murphy (1985)
confirmed that hard clam growth was reduced by sediment resuspension
produced by the ghost shrimp.
Rhoads and Young (1970) realized the limited applicability of their
theory because it did not adequately describe the factors that actually
separate deposit feeding and suspension feeding communities. Examples
of field studies of macrobenthic distribution which are at odds with
trophic group amensalism include those of Posey (1986), who found that
dense populations of the ghost shrimp C. californiensis were positively
correlated with a suspension feeder on the Oregon, United States, coast
and Dittmann (1990), who studied a blue mussel (Mytilus edulis) reef on
the island of Sylt in the North Sea and showed that this species facilitates
the establishment of a rich deposit feeding infauna associated with
hydrodynamic protection within the reef and which fed on the faecal
wastes produced by the mussels. A similar rejection of trophic group
amensalism is presented in the review of Snelgrove and Butman (1994).
The adult-larval interaction theory of Woodin (1976) was designed to
explain the occurrence of deposit feeders, suspension feeding clams, and
tube builders on the U.S. West Coast. The mechanism proposed was that
burrowing deposit feeders disturbed the sediment to the extent that
epifaunal tube bases and siphon holes of other species were disrupted,
thus inhibiting their growth and development. In addition, recently set-
tled larvae were consumed by the deposit feeders, thus helping to ensure
that clams and tube builders could be excluded from these highly re-
worked sediments. The suspension feeding of clams served to remove
settling larvae, as did the predatory and defaecatory activities of tube
builders, thus tending to exclude deposit feeders.
Posey (1990) reviewed the evidence for the adult-larval interaction
hypothesis, listing 12 studies which provide some evidence in favor of the
alternate hypothesis that larval settlement is inhibited by feeding and
locomotion of adult infauna. The work of Elmgren et al. (1986) in the
soft sediments of the Baltic Sea with two deposit feeding amphipods
{Pontoporeia affinis and P. femorata) and the facultative suspension
feeding bivalve Macoma balthica may be cited as an example. In the
Baltic, Macoma is absent or rare in those areas where dense populations
of Pontoporeia sp. occurred. Experimental tests in flowthrough aquaria
containing sediments (2-L glass jars) in which adult P. affinis and newly
settled spat of M. balthica were placed, showed that the latter decreased
with increasing densities of Pontoporeia. Apparently, P. affinis caused
direct physical injury to the spat, presumably ingesting the soft tissue,
although in nature the spat are not an important food source for
Pontoporeia.
Olafsson (1989) studied two contrasting localities in the Baltic contain-
ing Macoma balthica on the coast of southern Sweden. The first was in a
sheltered bay with limited water movement in a Zostera meadow where
the Macoma were deposit feeders. The second locality was exposed and
had a sandy sediment and more energetic water movement - mainly by
wind-wave action - where adult Macoma were primarily suspension
feeding. Adult density manipulation experiments in thefieldshowed that
juvenile Macoma growth and density levels were negatively affected at
the first, but not the second, location. Juvenile Macoma in both habitats
are deposit feeders, and thefieldresult was explained by dietary resource
overlap, which is absent where the water movements are more energetic
and where adult Macoma are suspension feeders. Experimental tests
suggested that Hydrobia sp., another deposit feeder, was negatively af-
fected by high densities of deposit feeding Macoma, whereas high densi-
ties of suspension feeding Macoma had no effect on deposit-feeding
oligochaetes of the second, water movement-energetic, sand habitat.
Andre and Rosenberg (1991) in manipulative field experiments in-
volving adjusting the densities of Cerastoderma edule and Mya arenaria
in the Baltic showed that adults reduced the settlement of bivalve larvae
and hence reduced the recruitment of new individuals proportional to
adult density. In a further laboratory flume study (Andre et al. 1993),
these findings were confirmed. Survival of settling larvae was propor-
tional to adult density of C. edule and increases in flume velocity caused
only a slight rise in predation risk.
There is as much evidence against the adult-larval interaction hypoth-
esis as there is for it (Posey 1990). Thus, deposit feeders, tube builders,
and suspension feeding clams do not always occur in discrete communi-
ties and may occur together in many natural situations (Posey 1990).
A direct field test involving manipulated densities of deposit- and
suspension-feeding clams and their effect on co-existing tube dwelling

and infaunal deposit-feeding polychaetes suggested no exclusion effects
(Hines, Posey, and Haddon 1989). Posey (1990) concluded that for a
number of communities examined, the predictions implied in the adult-
larval interaction hypothesis were not met and other factors must be
involved.
Summary
The importance of benthic production limitation by flow, inclusive of
trophic group mutual exclusion, has been emphasized in recent reviews
(e.g. Olafsson et al. 1994; Hall 1994). However, flow influences food
supply to suspension feeders not only by controlling the turbulent supply
of seston, but by direct water movement impoverishment of both suspen-
sion and deposit feeders. The mechanisms involved have, by comparison,
received relatively little study and could include larval incompetency,
and adult washout during storm events.
Using the well-developed theory of hydrodynamics to help formulate
specific hypotheses about suspension feeders at all hierarchial levels has
proved to be, and we believe will continue to be, most rewarding. At the
level of the population the seston depletion effect (number 1, Table 7.2)
has been validated by both downstream reductions in seston concentra-
tion and downstream reductions in growth. Howflowand seston concen-
tration interact to produce increased production has begun to be docu-
mented for a few species (number 2, Table 7.2). The references cited for
number 2, Table 7.2, are observational confirmation of Hl5 although
experimental results (e.g. Wildish and Kristmanson 1985) which also
support it are available. More experimental results to cover a wider
range of variables (velocity, seston concentration) are needed. Another
ecological mechanism which may be important in the population re-
sponse to increases in flow involves the addition of suspension feeders,
by immigration and larval recruitment, and their subtraction, by emigra-
tion and death. All of these mechanisms allow three to four orders
of magnitude variation in suspension feeder production in the natural
marine environment, dependent largely on velocity and seston
concentration.
Of the four other hypotheses outlined in Table 7.2, there is field
evidence that the growth and production of suspension feeders can be
limited by energetic water movement (number 3), although the mecha-
nisms remain to be investigated in detail. The effect of suspension feeder
density on seston uptake or growth rates of populations (number 5) has

been demonstrated, although this is not the case with the effect of vari-
able reef path lengths on these rates (number 6). Because it is outside the
scope of this book, we have not considered the relationship between
deposit feeder production and the erosional power of flow (number 4),
although pertinent studies have been undertaken (e.g. by Rhoads 1973;
Rhoads, Yingst, and Ullman 1978; Wainwright 1987, 1990; Davis 1993).
The trophic group amensalism hypothesis of Rhoads and Young
(1970) has received little support despite some efforts to test it and
should be discarded as a general explanation (see also Snelgrove and
Butman 1994). Similarly, the larval interaction hypothesis of Woodin
(1976) has received some support as well as contradictory evidence from
field studies. By comparison with the hydrodynamic hypotheses pre-
sented in Table 7.2, the two biotic interaction hypotheses are limited in
application and are not central to the exclusion principle operative
among suspension and deposit feeders in mixed benthic communities.
Wildish and Kristmanson (1993) recognized an apparent paradox
when comparing physiological responses of filtration at variable flows
(Chap. 4) to the ecological or population responses to flow presented in
this chapter. The blue mussel data of Fig. 7.12 are for the filtration rate
of an individual, while the horse mussel data are for the annual produc-
tion of mussels from many different localities in the Bay of Fundy. Figure
7.12 is a depiction of this paradox where, although the two y-axis values
are incommensurate for the reasons just mentioned, the x-axis is the
same for both y variables. The latter was achieved by measuring flows
near blue mussel inhalants and by recalculating the horse mussel ambi-
ent velocities at the inhalant level from free-stream averages (Wildish
and Kristmanson 1993). If the filtration rate of an individual suspension
feeder is considered to indicate feeding and thus growth rate, one might
expect more concordance between the physiological and ecological data
than is shown in Fig. 7.12. We believe that the paradox can be described
better as an emergent property of the population (reef or aggregation) of
suspension feeders. Salt (1979) provided the following operational
definitions:
"An emergent property of an ecological unit is one which is wholly unpredict-
able from observation of the components of that unit."
"An emergent property of an ecological unit is only discernible by observa-
tion of that unit itself."
By these considerations the average filtration and feeding rate of a
population as a function of velocity cannot be determined by studying
W 50 n
DC -3.0 Q
40-
O T
"D 8i
"2.0 Q w
ILTR,
CD 3 0 - DC =
|
u_ CO
-1.0
1
1 c:
o CO
20-
LU o
CO
CO I F
i O 10- LLJ
CO
^
LU
Z)
DQ
10 20 30 40
înhalant
Figure 7.12 Blue mussel Mytilus edulis filtration rates, as percentage chloro-
phyll a consumed per hour. Dashed line (based on Wildish and Miyares 1990);
solid line, horse mussel Modiolus modiolus production (Wildish and Peer 1983).
The x axis shows ambient velocities at inhalant level for both species. Chla,
chlorophyll a.
the physiological feeding rate of one or a few individuals even though the
flow conditions used are the same. The change in hierarchical level from
the individual to the population is, of course, crucial in appreciating the
nature of this phenomenon. If feeding rates are an emergent property of
a population of suspension feeders, then methods presently used to
determine population feeding rates and hence growth, e.g. in bivalves
(Frechette et al. 1989; Frechette and Grant 1991), must be misleading.
This is because population level phenomena, such as growth, determined
in this way exclude recruitment mortality and the seston depletion effect,
and so cannot reliably predict population growth.
References
Aller, R. C, and R. E. Dodge. 1974. Animal-sediment relations in a tropical
lagoon, Discovery Bay, Jamaica. J. Mar. Res. 32: 209-232.
Anderson, F. E., and L. M. Meyer. 1986. The interaction of tidal currents on a
disturbed intertidal bottom with a resulting change in particulate matter

quantity, texture and food quality. Estuar. Coastal Shelf Sci. 22: 19-30.
Andre, C, P. R. Jonsson, and M. Lindegarth. 1993. Predation on settling
bivalve larvae by benthic suspension feeders: the role of hydrodynamics
and behaviour. Mar. Ecol. Prog. Ser. 97: 183-192.
Andre, C, and R. Rosenberg. 1991. Adult-larval interactions in the
suspension-feeding bivalves Cerastoderma edule and Mya arenaria. Mar.
Ecol. Prog. Ser. 71: 227-234.
Blackman, F. F. 1905. Optima and limiting factors. Ann. Bot. 19: 281-295.
Bock, M. J., and D. C. Miller. 1994. Seston variability and daily growth in
Mercenaria mercenaria on an intertidal sandflat. Mar. Ecol. Prog. Ser. 114:
117-127.
Broom, M. J. 1982. Analysis of the growth of Anadara granosa (Bivalvia:
Arcidae) in natural, artificially seeded and experimental populations. Mar.
Ecol. Prog. Ser. 9: 69-79.
Buss, L. W., and J. B. C. Jackson. 1981. Planktonic food availability and
suspension-feeder abundance: evidence of in situ depletion. J. Exp. Mar.
Biol. Ecol. 49: 151-161.
Butman, C. A., M. Frechette, W. Rockwell Geyer, and V. R. Starczak. 1994.
Flume experiments on food supply to the blue mussel Mytilus edulis L. as
a function of boundary layer flow. Limnol. Oceanogr. 39: 1755-1768.
Cahalan, J. A., S. E. Sidall, and M. W. Luckenbach. 1989. Effects of flow
velocity, food concentration and particle flux on growth rates of juvenile
bay scallops Argopecten irradians. J. Exp. Mar. Biol. Ecol. 129: 45-60.
Crisp, D. J. 1984. Energy flow measurements, p. 284-372. In N. A. Holme and
A. D. Mclntyre (eds.) Methods for the study of marine benthos. 2nd ed.
Blackwell, Oxford.
Dare, P. J. 1976. Settlement, growth, and production of the mussel, Mytilus
edulis 1., in Morecambe Bay, England. Fish. Invest. Ser. II, 28: 1-25.
Dauer, D. M., C. A. Maybury, and R. M. Ewing. 1981. Feeding behaviour and
general ecology of several spionid polychaetes from the Chesapeake Bay.
Davis, W. R. 1993. The role of bioturbation in sediment resuspension and its
interaction with physical shearing. J. Exp. Mar. Biol. Ecol. 17: 187-200.
Dittmann, S. 1990. Mussel beds: amensalism or amelioration for intertidal
fauna? Helgolander. Meeresunters. 44: 335-352.
Dobbs, F. C, and J. M. Vozarik. 1983. Immediate effects of a storm on coastal
infauna. Mar. Ecol. Prog. Ser. 11: 273-279.
Eldridge, P. J., A. G. Eversole, and J. M. Whetstone. 1979. Comparative
survival and growth rates of hard clams, Mercenaria mercenaria, planted
in trays subtidally and intertidally at varying densities in a South Carolina
estuary. Proc. Nat. Shellfish. Assoc. 69: 30-39.
Elmgren, R., S. Ankar, B. Marteler, and G. Edjung. 1986. Adult interference
with postlarvae in soft sediment - the Pontoporeia-Macoma example.
Ecology 67: 827-836.
Emerson, C. W. 1989. Wind stress limitation of benthic secondary production
in shallow, soft-sediment communities. Mar. Ecol. Prog. Ser. 53: 65-77.
Emerson, C. W., and J. Grant. 1991. The control of soft-shell clam {Mya
arenaria) recruitment on intertidal sand flats by beadload sediment
Frechette, M., and E. Bourget. 1985. Energy flow between the pelagic and
benthic zones: factors controlling particulate organic matter available to

an intertidal mussel bed. Can. J. Fish. Aquat. Sci. 42: 1166-1170.
Frechette, M., C. A. Butman, and W. Rockwell Geyer. 1989. The importance
of boundary layer flows in supplying phytoplankton to the benthic
suspension feeder, Mytilus edulis L. Limnol. Oceanogr. 34: 19-36.
Frechette, M., and J. Grant. 1991. An in situ estimation of the effect of wind
driven resuspension on the growth of the mussel Mytilus edulis L. J. Exp.
Mar. Biol. Ecol. 148: 201-213.
Glynn, P. W. 1973a. Ecology of a Caribbean coral reef, the Porites reef-flat
biotope. Part I. Methodology and hydrography. Mar. Biol. 20: 297-318.
1973b. Ecology of a Caribbean coral reef, the Porites reef-flat biotope.
Part II. Plankton community with evidence for depletion. Mar. Biol. 22:
1-21.
Grizzle, R. E., R. Langan, and W. H. Howell. 1992. Growth responses of
suspension-feeding bivalve molluscs to changes in water flow: differences
between siphonate and non-siphonate taxa. J. Exp. Mar. Biol. Ecol. 162:
213-228.
Grizzle, R. E., and R. A. Lutz. 1989. A statistical model relating horizontal
seston fluxes and bottom sediment characteristics to growth of Mercenaria
mercenaria. Mar. Biol. 102: 95-106.
Grizzle, R. E., and P. J. Morin. 1989. Effect of tidal currents, seston, and
bottom sediments on growth of Mercenaria mercenaria: results of a field
experiment. Mar. Biol. 102: 85-94.
Hall, S. J. 1994. Physical disturbance and marine benthic communities: life in
unconsolidated sediments. Ocean. Mar. Biol. Annu Rev. 32: 179-239.
Hines, A. H., M. H. Posey, and P. J. Haddon. 1989. Effects of adult
suspension- and deposit-feeding bivalves on recruitment of estuarine
infauna. Veliger 32: 109-119.
Hunt, O. D. 1925. The food of the bottom fauna of the Plymouth fishing
grounds. J. Mar. Biol. Assoc. U.K. 13: 560-599.
Incze, L. S., R. A. Lutz, and L. Watling. 1980. Relationship between effects of
environmental temperature and seston on growth and mortality of Mytilus
edulis in a temperate northern estuary. Mar. Biol. 57: 147-156.
J0rgensen, C B. 1990. Bivalve filter feeding: hydrodynamics, bioenergetics and
Judge, M. L., L. D. Coen, and K. L. Heck. 1992. The effect of long-term
alteration of in situ water currents on the growth of Mercenaria
mercenaria in the northern Gulf of Mexico. Limnol. Oceanogr. 37: 1550-
1559.
Jumars, P. A., and A. R. M. No well. 1984a. Fluid and sediment dynamic
effects on marine benthic community structure. Am. Zool. 24: 45-55.
1984b. Effects of benthos on sediment transport: difficulties with functional
grouping. Cont. Shelf Res. 3: 115-130.
Kirby-Smith, W. W. 1972. Growth of the bay scallop: the influence of
experimental currents. J. Exp. Mar. Biol. Ecol. 8: 7-19.
Leonard, A. B. 1989. Functional responses in Antedon mediterranea
(Lamarck) (Echinodermata: Crinoidea): the interaction of prey
concentration and current velocity on a passive suspension feeder. J. Exp.
Mar. Biol. Ecol. 127: 81-103.
Levinton, J. S. 1991. Variable feeding behaviour in three species of Macoma
(Bivalvia: Tellinacea) as a response to water flow and sediment transport.

Mar. Biol. 110: 375-383.
Levinton, J., and G. R. Lopez. 1977. A model of renewable resources and
limitation of deposit feeding benthic populations. Oecologia 31: 177-
190.
Levinton, J. S., D. E. Martinez, M. M. McCartney, and M. L. Judge. 1995. The
effect of water flow on movement, burrowing, and distributions of the
gastropod Ilyanassa obsoleta in tidal creeks. Mar. Biol. 122: 417-424.
Madsen, K. N., P. Nilsson, and K. Sundback. 1993. The influence of benthic
microalgae on the stability of a subtidal sediment. J. Exp. Mar. Biol. Ecol.
170: 159-177.
McLachlan, A., A. C. Cockcroft, and D. E. Malan. 1984. Benthic faunal
response to a high energy gradient. Mar. Ecol. Prog. Ser. 16: 51-63.
Murphy, R. C. 1985. Factors affecting the distribution of the introduced
bivalve, Mercenaria mercenaria, in a California lagoon: the importance of
bioturbation. J. Mar. Res. 43: 673-692.
Muschenheim, D. K., P. E. Kepkay, and K. Kranck. 1989. Microbial growth in
turbulent suspension and its relation to marine aggregate formation. Neth.
J. Sea Res. 23: 283-292.
Nehls, G., and M. Thiel. 1993. Large-scale distribution patterns of the mussel
Mytilus edulis in the Wadden Sea of Schleswig-Holstein: do storms
structure the ecosystem? Neth. J. Sea Res. 31: 181-187.
Newell, R. 1965. The role of detritus in the nutrition of two marine deposit-
feeders, the prosobranch Hydrobia ulvae and the bivalve Macoma
balthica. Proc. Zool. Soc. London 144: 25-45.
Newell, C. R. 1990. The effects of mussel {Mytilus edulis, Linnaeus, 1758)
position in seeded bottom patches on growth at subtidal lease sites in
Maine. J. Shellfish Res. 9: 113-118.
Neyman, A. A. 1979. Soviet investigations of the benthos of the shelves of the
marginal seas, p. 269-284. In M. V. Dunbar (ed.) Marine production
mechanisms. Cambridge University Press, Cambridge.
Nickels, J. S., J. D. King, and D. C. White. 1979. Poly-(3-hydroxybutyrate
accumulation as a measure of unbalanced growth of the estuarine detrital
microbiota. Appl. Environ. Microbiol. 37: 459-465.
Odum, E. P. 1971. Fundamentals of ecology. W. B. Saunders, Philadelphia.
Olafsson, E. B. 1989. Contrasting influences of suspension-feeding and
deposit-feeding populations. Mar. Ecol. Prog. Ser. 55: 171-179.
Olafsson, E. B., O. H. Peterson, and W. G. Ambrose. 1994. Does recruitment
limitation structure populations and communities of macroinvertebrates in
marine soft sediments: The relative significance of pre- and post-
settlement processes. Ocean. Mar. Biol. Annu. Rev. 32: 65-109.
O'Riordan, G. A., S. G. Monismith, and J. R. Koseff. 1993. A study of
concentration boundary layer formation over a bed of model bivalves.
Pearson, T. H. 1971. Studies on the ecology of the macrobenthic fauna by
comparison of feeding groups. Vie Milieu 22 (Suppl.): 53-91.
Pearson, T. H., and R. Rosenberg. 1987. Feast and famine: structuring factors
in marine benthic communities, p. 373-395. In J. H. R. Gee and P. S.
Giller (eds.) Organization of communities past and present. Blackwell,
Oxford.
Perkins, E. J. 1974. The biology of estuaries and coastal waters. Academic

Press, London.
Peterson, C. H. 1982. The importance of predation and intra- and interspecific
competition in the population biology of two infaunal suspension-feeding
bivalves, Protothaca staminea and Chione undatella. Ecol. Monogr. 52:
437-475.
Peterson, C. H., and S. V. Andre. 1980. An experimental analysis of
interspecific competition among marine filter feeders in a soft-sediment
environment. Ecology 61: 129-139.
Peterson, C. H., and B. F. Beal. 1989. Bivalve growth and higher order
interactions: importance of density, site and time. Ecology 70: 1390-1404.
Peterson, C. H., and R. Black. 1987. Resource depletion by active suspension
feeders on tidal flats: influence of local density and tidal elevation.
Peterson, C. H., and R. Black. 1988. Density-dependent mortality caused by
physical stress interacting with biotic history. Am. Nat. 131: 257-270.
1991. Preliminary evidence for progressive sestonic food depletion in
incoming tide over a broad tidal sand flat. Estuar. Coast. Shelf Sci. 32:
405-414.
Peterson, C. H., H. C. Summerson, S. R. Fegley, and R. C. Prescott. 1989.
Timing, intensity and sources of autumn mortality of adult bay scallops
Argopecten irradians concentricus Say. J. Exp. Mar. Biol. Ecol. 127: 121-
140.
Posey, M. H. 1986. Changes in the benthic community associated with dense
beds of a burrowing deposit feeder, Callianessa californiensis. Mar. Ecol.
Prog. Ser. 31: 15-22.
1990. Functional approaches to soft-substrate communities: how useful are
they? Rev. Aquat. Sci. 2: 343-356.
Purdy, E. G. 1964. Sediments as substrates, p. 238-271. In J. Imbre and N.
Newell (ed.) Approaches to paleoecology. Wiley, New York.
Rhoads, D. C. 1973. The influence of deposit-feeding benthos on water
turbidity and nutrient recycling. Am. J. Sci. 273: 1-22.
Rhoads, D. C, J. Y. Yingst, and W. J. Ullman. 1978. Seafloor stability in
Central Long Island Sound. Part 1. Temporal changes in erodibility of
fine grained sediments, p. 221-244. In M. L. Wiley (ed.) Estuarine
interactions. Academic Press, New York.
Rhoads, D. C, and D. K. Young. 1970. The influence of deposit feeding
organisms on sediment stability and community trophic structure. J. Mar.
Res. 28: 150-178.
Ritzrau, W., and G. Graf. 1992. Increase of microbial biomass in the benthic
turbidity zone of Kiel Bight after resuspension by a storm event. Limnol.
Oceanogr. 37: 1081-1086.
Salt, G. W. 1979. A comment on the use of the term emergent properties. Am.
Nat. 113: 145-148.
Sanders, H. L. 1956. Oceanography of Long Island Sound 1952-1954. X. The
biology of marine bottom communities. Bull. Bingham Oceanogr.
Collection, Yale Univ. 15: 345-414.
1960. Benthic studies in Buzzards Bay. III. The structure of the soft-bottom
community. Limnol. Oceanogr. 5: 138-153.
Sandford, E., D. Bermudez, M. D. Bertness, and S. D. Gaines. 1994. Flow,
food supply and acorn barnacle population dynamics. Mar. Ecol. Prog.
Ser. 104: 49-62.
Seed, R., and T. H. Suchanek. 1992. Population and community ecology of

Mytilus, p. 87-169. In E. Gosling (ed.) The mussel Mytilus: ecology,
physiology, genetics and culture. Elsevier, Amsterdam.
Shanks, A. L., and W. G. Wright. 1986. Adding teeth to wave action: the
destructive effects of wave borne rocks on intertidal organisms. Oecologia
69: 420-428.
Smaal, A. C, J. H. G. Verhagen, J. Coosen, and H. A. Haas. 1986. Interaction
between seston quantity and quality and benthic suspension feeders in the
Oosterschelde, The Netherlands. Ophelia 26: 385-400.
Snelgrove, P. V. R., and C. A. Butman. 1994. Animal-sediment relationships
revisited: cause versus effect. Ocean. Mar. Biol. Annu. Rev. 32: 111-177.
Sternberg, R. W. 1968. Friction factors in tidal channels with differing bed
roughness. Mar. Geol. 6: 246-260.
Verhagen, J. H. G. 1986. Tidal motion, and the seston supply to the benthic
macrofauna in the Oosterschelde. DHL Report R1310-14.
Wainright, S. C. 1987. Stimulation of heterotrophic microplankton production
by resuspended marine sediments. Science 238: 1710-1712.
1990. Sediment-to-water fluxes of particulate material and microbes by
resuspension and their contribution to the planktonic food web. Mar.
Ecol. Prog. Ser. 62: 271-281.
Warwick, R. M., and R. J. Uncles. 1980. Distribution of benthic macrofauna
associations in the Bristol Channel in relation to tidal stress. Mar. Ecol.
Prog. Ser. 3: 97-103.
White, D. C, W. M. Davis, J. S. Nickels, J. D. King, and R. J. Bobbie. 1979.
Determination of the sedimentary microbial biomass by extractible lipid
phosphate. Oecologia 40: 51-62.
1980. Reproductive bionomics of two sublittoral amphipods in a Bay of
Fundy estuary. Int. J. Invertebr. Rep. 2: 311-320.
1985. Geographical distribution of macrofauna on sublittoral sediments of
continental shelves: a modified trophic ratio concept. J. Mar. Biol. Assoc.
U. K. 65: 335-345.
Wildish, D. J., and D. D. Kristmanson. 1979. Tidal energy and sublittoral
macrobenthic animals in estuaries. J. Fish. Res. Board Can. 36: 1197-1206.
1984. Importance to mussels of the benthic boundary layer. Can. J. Fish.
Aquat. Sci. 41: 1618-1625.
1985. Control of suspension-feeding bivalve production by current speed.
Helgolander. Wiss. Meeresunters. 39: 237-243.
1988. Growth response of giant scallops to periodicity of flow. Mar. Ecol.
Prog. Ser. 42: 163-169.
1993. Hydrodynamic control of bivalve filter feeders: a conceptual view, p.
299-324. In R. F. Dame (ed.) Bivalve filter feeders in estuarine and
coastal ecosystem processes. NATO ASI Ser. Vol. G33. Springer-Verlag,
Berlin.
Ecol. 142: 213-219.
40(Suppl. 1): 309-321.
production in the Bay of Fundy and possible effects of a tidal power

2410-2417.
Wolff, W. J., F. Vegter, H. G. Mulder, and T. Meijs. 1976. The production of
benthic animals in relation to the phytoplankton production: observations
in the saline Lake Grevelingen, The Netherlands. Proc. 10th Eur. Symp.
Mar. Biol. 2: 653-672.
Woodin, S. A. 1976. Adult-larval interactions in dense infaunal assemblages:
patterns of abundance. J. Mar. Res. 34: 25-41.
Yeo, R., and M. J. Risk. 1979. Intertidal catastrophes: effects of storms and
hurricanes on intertidal benthos of the Minas Basin, Bay of Fundy. Can.
J. Fish. Aquat. Sci. 36: 667-669.
8
Ecosystems and flow
Our aim in this chapter is to examine whether water movement is funda-

mentally important in aquatic ecosystems to productivity and materials
cycling in the ways briefly outlined in the next section. The testing of a
suitable ecosystem level null hypothesis in which water movement is
excluded is probably an unrealistic task. Yet, the idea of such a test
should be kept in mind as we examine the real-world ecosystems which
follow.
We note that the thrust of this chapter, ecosystems interacting with
flow, is contradictory to the book's theme. This is because, in most of the
book, we concentrate on suspension feeders and largely ignore the rest
of the ecosystem of which they are a part. Yet, for ecosystem analysis we
must be concerned with all of the biological and physical variables that
are important in energyflow.Thus, for this chapter only, we have relaxed
our definition to mean complete ecosystems including hydrodynamic
effects on macroflora and macrofauna, inclusive of plant canopies, which
are clearly not suspension feeders, but important primary producers.
Our review of the contemporary benthic literature suggests that plant
canopies and bivalve and coral reef ecosystems are best studied from the
hydrodynamic perspective.
The examples we have chosen to include in this chapter are plant
canopies, e.g. seaweed and seagrass ecosystems, and marine/estuarine
environments with suspension feeding bivalve reefs. We have excluded
coral reef ecosystems because at least two recent reviews (Sorokin 1993;
Bakus 1994) have been published, the literature on this subject is vast,
and the authors have no firsthand experience with corals.
Background
The aquatic ecosystem concept of materials cycling, measured as energy,
carbon, or nutrient fluxes, was initiated by freshwater ecologists. Juday
313
(1940) described the annual energy budget of an inland lake in which the
phytoplankton and attached aquatic plants utilized only about 1% of the
available subsurface solar energy during photosynthesis. Such energy
balances depend directly on the first law of thermodynamics - that
energy cannot be created or destroyed. Lindeman (1942), also working
in a lacustrine environment, showed how the food web complexity of an
ecosystem could usefully be simplified by categorizing the constituent
organisms into trophic-dynamic groups, hence, primary producers, her-
bivorous consumers, predators, and decomposers. Lindeman (1942) pro-
duced a diagram of a food cycle (web) for a lake in which the loss in
respiratory energy at each trophic level increased with distance from the
original solar energy supply. This conforms with the second law of ther-
modynamics because the transformation of energy at each trophic level,
e.g. light to chemical energy in plant tissue, involves an energy loss.
Both Juday and Lindeman tacitly assumed that solar radiation was of
most importance in freshwater ecosystems. Munk and Riley (1952) pre-
sented a quantitative theory that nutrient uptake by marine pelagic and
attached plants, e.g. seaweeds and seagrasses, was influenced by flow in
removing mass transfer limitation. Absorption of nutrients was consid-
ered to be dependent on the ambient concentration gradients and veloci-
ties. Thus, Munk and Riley (1952) showed how water movement, up to
a threshold level, was an integral part of materials cycling. The physi-
ological cast of their quantitative model, applicable to plant material
exchanges and water movement, precludes its central use in this chapter,
which is concerned with ecosystem level phenomena.
Odum (1971) and Odum, Finn, and Franz (1979) proposed that water
movement was unimodally related to attached plant biomass or produc-
tivity (Fig. 8.1) in an ecosystem level extension of part of Munk and
Riley's mass transfer limitation theory. Odum (1980) considered that
increases in attached marine plant production were caused by increases
in water movement which allowed the following:
Increased removal of waste gases and greater supply of bicarbonate for
photosynthesis
Increased nutrient fluxavailability to plant absorptive surfaces
Increased water column recycling of plant nutrients
Increased sediment-water interface fluxesleading to greater plant nutrient
cycling
The reference to this as a "tidal energy subsidy" by Odum (1980) is
clearly in error since no energy which subsidizes or enhances the produc-
tion process is directly supplied. Rather, the tidal energy supplied is
Ecosystems and Flow 315
Subsidy
LU
C/D
o
CL
CO
LU
DC Normal Range
LJLJ
Stress - \
CO
O Replacement
O \
LU
WATER MOVEMENT
Figure 8.1 Theoretical ecosystem response to water movement perturbation
(adapted from Odum et al. 1979).
involved in removal of mass transfer limitation, as proposed by Munk

and Riley (1952), and it is this which allows production to proceed at a
higher, up to an optimum, rate.
Odum et al. (1979) recognized in the perturbation theory that ecosys-
tem level effects were a unimodal response of important environmental
variables such as water motion. Ecosystem level stresses resulted if water
movement reached certain levels and could result in physiological stress
of the biota present, replacement of sensitive by hardier species, or death
of the resident biota if the stresses were sufficiently large or persistent.
Marine plant canopies: kelp forests

Kelp forests form the characteristic nearshore vegetation of the hard
substrates of most temperate and boreal latitudes. The dominant kelps of
the world (Mann 1973) include Laminaria of the North Atlantic and
eastern North Pacific and Okhotsk Sea; Macrocystis of the Pacific Ocean
along South and North American shores, which may reach a lamina
length of 40m (Jackson and Winant 1983); and Ecklonia of the southern
tip of Africa and southeast Australia. Species of Macrocystis are also
present at the two latter localities, as well as on the rocky coasts of New
Zealand.
Understorey kelps, where the lamina is usually only a few metres in

length, may also be present and add to the kelp forest productivity. As
pointed out by Eckman, Duggins, and Sewell (1989), understorey kelps
may affect horizontal transport of mass and momentum somewhat dif-
ferently than the larger overstorey kelps.
Hydrodynamics
The theoretical account of Raupach and Thom (1981) concerning turbu-
lence in and above plant canopies provides a good starting point, al-
though these authors were dealing with terrestrial plants in air. Raupach
and Thom (1981) conclude that the subject of canopy turbulence is still
largely empirical. The plant structures which protrude into the boundary
layer, as well as the elasticity of the plants within the canopy, are of
importance in determining aerodynamic as well as hydrodynamic re-
sponses associated with plant canopies.
The fluid dynamics of seawater flow through a bushy red alga,
Gelidium nudifrons from Southern California, was studied by Anderson
and Charters (1982). Gelidium is found on semi-exposed shores, and
field measurements of orbital wave action of 2-10 cm-s 1 which passed
through the branches of the plant emerged as turbulent flow. Laboratory
studies in a low velocity water tunnel specially designed for this purpose
used dyes and a hot film anemometer to characterise flow. Gelidium
plants of 35 cm height and with branches of 0.05 cm thickness, when
tested with unidirectional flows in the water tunnel, damped turbulent
flows <6-12cm-s~1, producing a laminar exitflow.Atflows>6-12cm-s~1,
the flow leaving the plant was always turbulent.
A large kelp forest of 7- x 1-km area near San Diego, California, and
a nearby reference site which was kelp-free were compared by current
meter observations by Jackson and Winant (1983). The dominant kelp
was Macrocystis pyrifera, with some specimens reaching 40 m in stipe
length. These authors tested the null hypothesis that kelp plants had no
effect on flow patterns within the kelp forest (number 1, Table 8.1).
Vector measuring current meters were deployed at three depths in a
central location of the kelp bed and similarly at the reference site.
Records were obtained at both sites for a 7-d period and showed that the
average longshore currents within the kelp forest were -lcm-s" 1 . This
velocity was 43%-54% less than the root mean square velocity at the
reference site, thus supporting Hx in number 1, Table 8.1. Calculations
from these data suggest that seawater residence times within the kelp
Table 8.1. Kelp forest ecosystem hypotheses.
1 Ho Current velocity is independent of kelp Jackson and Winant

presence/absence (1983)
H^ Current velocity is inhibited by the
presence of kelp
2 Ho Sedimentation rates within are similar Eckman et al. (1989)
to those outside the kelp ecosystem
Hj Sedimentation rates within are greater
than those outside the kelp ecosystem
3 Ho Key predators have no effect in Mann (1973)
structuring kelp forest ecosystems Estes and Palmisano
Hx Key predators structure kelp forest (1974)
ecosystems by consuming key
herbivores such as algal grazing sea
urchins
4 Ho Reduced velocities and increased Duggins et al. (1989)
sedimentation rates in kelp forests Eckman and Duggins
have no effect on suspension feeder (1991)
production
Hx Reduced velocities and increased
sedimentation in kelp forests enhance
suspension feeder production or
growth
5 Ho Most POM and DOM derived from Duggins and Eckman
kelp forests are exported (1994)
Hx Most POM and DOM from kelp
forests are utilized within the kelp
forest by pelagic and benthic
suspension feeders
forest were >7 d. Measured velocities were always strongest at the kelp
forest edge, but slowed to average values within 100 m into the kelp
forest. In a further study of the same area, Jackson (1984) placed current
meters with temperature recorders along a transect across the kelp for-
est. As a result of the high drag of the kelp plants, present at densities of
0.02-0.14 plants-m 2 , internal waves which passed through the forest
were considerably dampened (Fig. 8.2), as evidenced by a smoothing of
high frequency fluctuations.
Understorey kelps dominated by Agarum fimbriatum were studied on
'irfVÛir-^^
+530 m
10 11 12 13 14 15 16 17
OCTOBER
Figure 8.2 Continuous temperature records over a four site deployment at
Point Loma kelp forest, near San Diego, California. The first site (-70 m) is 70 m
outside, the next (+120m) is 120m into the kelp, and so on (Jackson 1984).
the San Juan Archipelago, at 48N, 123E, by Eckman et al. (1989). The
relative degree of water movement activity, assessed by the gypsum flux
method, showed that the kelp plants impeded flow and dampened vari-
ability in currents relative to that of a reference location nearby. As with
the California data outlined previously, the effect was accompanied by a
reduction in mass transport and shear within the kelp bed. Because of
the reduced flow within the kelp, Eckman et al. (1989) reasoned that
increased sedimentation should occur (Table 8.1). Sedimentation rates
were measured by collection of the particles on acrylic plastic (Plexiglas)
plates, followed by pumping into deflated plastic bags,filtering,and then
weighing. Because this method did not properly distinguish sedimenta-
tion and resuspension events, a particle impaction experiment was also
employed. It involved glass beads of 22 (im diameter and with a settling
velocity of 0.0276 cm s"1, which were slowly released upstream of micro-
scope slides coated with vacuum grease and on which the beads became
trapped. Results suggested to Eckman et al. (1989) that particle trapping
was reduced in the presence of kelp plants, consistent with a reduction in
the intensity of turbulent mixing within the canopy. The increased sedi-
mentation within kelp canopies is explained by the greater retention
times that particles must spend within the canopy and hence a greater
proportion of particles which must settle. During flood-ebb tidal
changes, settled particles may be resuspended and the resuspension-

sedimentation cycle repeated many times.
These results suggest that flow may pass through and around indi-
vidual or groups of macroalgal plants, depending on the ambient veloc-
ity. In passing through the bed, flow is impeded by the elastic lamina and
stipe. A consequent reduction in mass transport and shear stress results
in increased sedimentation within the macroalgal bed.
Materials cycling
A functional account of the Laminaria-dominated kelp forests of Atlan-
tic Canada is given by Mann (1973). From this and related studies, a
partial carbon budget can be constructed (Fig. 8.3). Such a budget repre-
sents the following:
An estimate of the amount of energy passing through the ecosystem
A simplification of a more complex cycle than is shown in Fig. 8.3
Hypotheses about the pathways of energy, carbon, or minerals in the kelp
ecosystem
Salient features of this budget are the very high primary production
levels achieved by Laminaria longicornis, L. digitata, and Agarum
cribosum, which are based on energetic water movement, principally
wind-wave action, which enhances photosynthesis by supplying and re-
moving gases, bicarbonate, and nutrients. The proportion of dissolved
organic matter (DOM) utilized in the pelagic microbial loop was not
estimated, although it was probably small. Both DOM and particulate
organic matter (POM) are exported from the kelp forest. Estimates of
net export of POM suggest that it is a large proportion of the total
because the principal herbivores, the sea urchin Strongylocentrotus
droebachiensis and periwinkle Littorina littorea, utilized <10% of the
available kelp carbon. High density aggregations of urchins sometimes
occurred and could be very destructive of the kelp plants by biting
through the stipe bases, thus forming cleared patches referred to as "sea
urchin barrens" (Mann 1973). Whether the feeding aggregations of S.
droebachiensis are controlled by decapod key predators, as suggested by
Mann (1973), or by some other mechanism, inclusive of innate urchin
behavior (Hagen and Mann 1994), is still a matter of debate.
In field studies conducted on the Aleutian Archipelago, Alaska, Estes
and Palmisano (1974) chose two study sites: Rat Island at 52N, 178E,
and Near Island at 52N, 174E. The two sites were 400km apart but
[930-1150] TURBULENTTRANSPORT
Figure 8.3 Partial carbon budget, g C m 2 yr \ for a Laminaria-dominated kelp

forest in St. Margaret's Bay, Nova Scotia (based on Mann 1972,1973; Miller and
Mann 1973).
otherwise were similar in physical characteristics, differing only in that

Rat Island had a rich and diverse littoral and sublittoral kelp forest
dominated by Laminaria sp., whereas at Near Island kelps were virtually
absent sublittorally and replaced by a sea urchin barrens dominated by
Strongylocentrotus sp. Estes and Palmisano (1974) suggested that the
marked difference between the kelp-free site at Near and kelp-rich site
at Rat Island was due to the presence of a key predator, the sea otter
Enhydra lutris, at the latter (Table 8.1, number 3). High densities - 20-
30 sea otters km"2 - removed the larger sea urchins, thereby preventing
destructive damage by them to the kelp. Circumstantial field evidence
gathered in support of this hypothesis was limited; it included the follow-
ing: sea urchin densities were much higher at the kelp-free sites; at those
sites where sea otters were present at high density, only small urchins,
less than 35-mm test diameter, were found; and some differences in
higher trophic level species between sites, e.g. the presence of harbour
seals and bald eagles, were related to the higher primary production
available at kelp-rich Rat Island.
Further study at these sites was undertaken by Duggins, Simenstad,
and Estes (1989) to determine the importance of detrital pathways
within coastal ecosystems (Table 8.1, number 4). Comparative field
measurements of growth for two suspension feeders, Balanus glandula
attached to test plates and Mytilus edulis, were made by locating marked
individuals sublittorally within kelp beds at Rat Island and at the same
depth in the kelp-free environment at Near Island. After 1 yr of growth,
the rates were compared (Fig. 8.4); barnacle growth was five times and
mussel growth two to four times greater at the kelp-rich site. Because the
carbon ingested during feeding by these suspension feeders could have
originated from sources other than kelp, e.g. from phytoplankton, it was
necessary to use a method capable of tracing the feeding habits of sus-
pension feeders. Because the 8 12C: 13C signatures of phytoplankton and
kelp were so different, it was possible to indicate the source of the carbon
on which they had been feeding. Of 11 species examined by this tech-
nique from the Alaskan sites, Duggins et al. (1989) found that 10 species
had enriched levels of 8 13C from Rat Island, suggesting that they had
been feeding indirectly on kelp (on average 58% C from kelp). Only
Mytilus edulis did not show enrichment in this way, suggesting that it may
prefer a phytoplankton diet. By contrast, 8 13C enrichment at the Near
Island site was much lower (on average 32% C from kelp).
Both field growth experiments and measurements of carbon isotope
ratios in naturally co-occurring benthic macrofauna of kelp beds led
Duggins et al. (1989) to conclude that the kelp plants themselves are a
significant carbon source for them. Because of the nature of the measure-
ments, the results give no indication about how important this pathway is
to suspension feeders. Nor do they allow an estimation of the relative
importance of kelp carbon export from the kelp forest.
43
KELP-RICH KELP-FREE J
Figure 8.4 Growth experiments on the Aleutian archipelago, Alaska: A,
Mytilus edulis; B, Balanus glandula (from Duggins et al. 1989). Two kelp-rich
sites from Rat Island, one kelp-free site from Near Island. Number above each
bar, sample size.
For the understorey kelp ecosystem on the Washington coast de-

scribed from the physical point of view earlier Eckman and Duggins
(1991) conducted a field manipulative experiment to test hypothesis
number 4 (Table 8.1) further. Four sites, which differed in mean free-
stream velocity from 16 to 24cm-s"1 and, therefore, in sedimentation
rate, were chosen. Kelp and kelp-free treatments were created and main-
tained by translocating or removing kelp plants in plots of 10-20 m2. Field
growth experiments with a range of typical kelp suspension feeders,
attached to either upward- or downward-oriented test plates bolted to
the substrate, established that sedimentation within the kelp bed was
inhibiting for growth. Despite this, upwardly oriented plates with suspen-
sion feeders such as Mytilus sp., Balanus glandula, and a serpulid
polychaete, Pseudochitmopoma occidentalis within the kelp plots all
grew significantly faster than those in adjacent kelp-free plots. For two
additional species, a bryozoan and a sponge, no differences in growth in

kelp and kelp-free plots were evident because other limiting factors were
involved. The results are at least partly consistent with Hx in hypothesis
4, Table 8.1.
In a further field experiment on the Washington coast, Duggins and
Eckman (1994) attempted to de-couple velocity and sedimentation ef-
fects (hypothesis 4, Table 8.1) from those due to in situ production of
DOM and POM from kelps. These experiments involved creating real
and plastic model kelp understorey in 10-m2 plots and testing whether
suspension feeders benefitted trophically from the autochthonously pro-
duced POM and DOM (hypothesis 5, Table 8.1). Of the six species of
suspension feeders tested - Mytilus trossulus, Chlamys hastata, Balanus
glandula, Pseudochitmopoma occidentalism Chelyosoma productum, and
spirorbid polychaetes - only one, B. glandula, grew significantly faster in
the artificial kelp plot. These results support the null hypothesis (number
5, Table 8.1) and are clearly different from the results obtained by
Duggins et al. (1989) in the Aleutian Islands, suggesting that responses
are governed by multiple limiting factors and rendering field experimen-
tation difficult to interpret.
By comparison with the closed steady-state model of a Laminaria-
dominated kelp forest in a small bay on the northwest Atlantic coast
(Fig. 8.3), an ecosystem simulation model of a kelp forest dominated by
Ecklonia maxima and Laminaria pallida on the west coast of South
Africa has been described (Wulff and Field 1983; Wickens and Field
1986). The model simulations suggested that local water movements of
the Benguela current were integrally involved in materials cycling of the
kelp ecosystem. Thus, down-welling was brought about by relaxation of
offshore prevailing southeast winds or by strong onshore winds, and up-
welling caused by strong prevailing southeast winds. During the southern
summer (September-April), the wind patterns favor up-welling; during
winter (May-August), down-welling (Wulff and Field 1983). Under up-
welling conditions, the major consumers' trophic requirements, includ-
ing the mussel Aulacomya ater, are not met, as kelp POM and faeces are
also being exported from the kelp forest. During down-welling, offshore
phytoplankton is continuously being advected through the kelp and
provides a rich source of food for suspension feeders which are part of
the kelp ecosystem. If a closed system model is used for the South
African kelp ecosystem, then a large portion of the nitrogen flow is
depicted to be present in bacteria associated with kelp POM and faeces
(Wickens and Field 1986). Modelling this system as an open one in which
local water movements are included shows that bacteria have insufficient
time to accumulate before being advected from the kelp forest.
Kelp productivity
Overstorey kelps are present on rocky substrates of the nearshore
coastal environment to a maximum depth of -40 m. Hence, they form a
narrow zone close to the shore and are light-limited at greater depths.
Primary production estimates made here suggest very high levels, e.g.
1750 g C-m^-yr"1, for a Laminaria-dominated forest on the Atlantic
coast of Canada (Mann 1973). A West Coast kelp forest dominated by
the sea palm Postelsia palmaeformis is reported (Leigh et al. 1987) to
have an even higher productivity, 3000gC-m"2-yr~1, on the most wave-
exposed parts of the coast of Washington.
A cursory review of the data concerning kelp productivity in relation
to wind-wave exposure suggests apparently conflicting results. Thus,
Laminaria kelp forests were less productive at the most exposed sites on
the Atlantic coast of Canada studied by Gerard and Mann (1979), and
Macrocystis sp. beds grew best at intermediate levels of wave exposure
on the southern coast of Chile (Dayton 1985). Yet, on the coast of
Washington, Postelsia sp., referred to earlier, as well as the shrubby
kelps, Lessoniopsis littoralis, are found only on the most wave-exposed
locations (Leigh et al. 1987). Unfortunately, although Leigh et al. (1987)
determined kelp productivity at a number of different sites on the Wash-
ington coast, the ranges of productivity were lumped and not related
individually to an estimate of wave power, thus rendering the results
unconvincing.
An obvious difficulty in comparing the work of different authors on
kelp productivity is the lack of a commonly used absolute measure of
wave exposure. This problem was considered in Chapter 2 and the
gypsum dissolution method proposed is an inexpensive way of compara-
tively measuring wind-wave activity.
Despite the lack of a wave exposure measure for the data cited, we
believe that the gross differences in exposure suggested are real and
result in the responses by kelp forests suggested. If this is so, the key to
understanding the conflicting results lies in the probability that kelps are
adapted to different amounts of wave exposure, both inter- and
intraspecifically. Indeed, the form of a sheltered versus an exposed frond
of Laminaria longicornis (Fig. 8.5) bears this out. Further examples of
intraspecific lamina form changes, e.g. within Macrocystis sp., which are
Figure 8.5 Growth forms of Laminaria longicornis from the Canadian Atlantic
coast: A, sheltered site; B, exposed site (Gerard and Mann 1979).
dictated by the hydrodynamic conditions experienced, are provided by

Wheeler (1980), and within Nerocystis by Koehl and Alberte (1988). It is
reasonable also to expect that different kelp species will have different
wave exposure optima.
Finally, we consider possible ways that wave energy may influence
kelp productivity. Two major kinds of mechanisms can be identified:
physical effects and interacting physical-biological effects. They are for-
mulated as hypotheses in Table 8.2 with numbers 1-3 representing physi-
cal and numbers 4-5 representing physical-biological effects.
As has already been pointed out, increased diffusion of materials
across plant surfaces results when water movement is increased up to an
asymptotic point. This idea was introduced in marine ecology by Munk
and Riley (1952). Although there is available evidence in both the field
and the laboratory that undirectional free-stream velocities increase this
flux and enhance photosynthesis, we know of no studies where simulated
wave action has been tested for the same effect. An example of a good
laboratory study with unidirectional flows and the kelp Macrocystis
pyrifera is provided by Wheeler (1980), who showed that velocities up to
4 cm s"1 increase the photosynthetic rate of the lamina.
Table 8.2. Hypotheses regarding wind-wave effects on

kelp productivity.

1 Ho Diffusion of gases or nutrients across Leigh et al. (1987)
lamina surfaces is unrelated to wave
activity
Hi Enhanced diffusion of gases or
nutrients across lamina surfaces
results from increased wave activity
2 Ho Wave stirring of kelp fronds has no Leigh et al. (1987)
effect on net photosynthetic rate of Koehl and Alberte
the plant (1988)
Hx Wave stirring of kelp fronds causes
more fronds to be light exposed, thus
increasing net photosynthetic rate of
the plant
3 Ho Wave exposure does not produce Koehl (1986)
shearing stresses to remove all, or
part, of the kelp plant
Hj Wave exposure produces shearing
stresses which can remove all, or
part, of the kelp plant, resulting in its
death
4 Ho Strong wave action has no effect on Connell (1972)
the kelp-herbivore interaction Menge and Sutherland
Hx Strong wave action protects kelps (1976)
from herbivory
5 Ho Strong wave action has no effect on Dayton (1971)
kelps and their competitors Dayton (1975)
Hx Strong wave action allows kelps to Paine (1979)
outcompete other species present
For the other hypotheses shown in Table 8.2 (numbers 2-5), there is
some evidence for the alternative hypotheses, as described in the listed
references.
Marine plant canopies: seagrass meadows

Seagrass meadows represent the common attached macrophytic ecosys-
tem of nearshore soft sediments at temperate and boreal latitudes (Den
Hartog 1970). Over 50 species of seagrass are known; the following

genera are common: Zostera, Spartina, Halodule, Thalassia, Halophilia,
Posidonia, and Enhalus. Seagrasses may be present intertidally and/or
subtidally, and each plant consists of a shoot and leaves with an under-
ground root and rhizomes which serve to stabilize both plant and sur-
rounding soft sediment.
Primary production by seagrass ecosystems is generally high, although
somewhat less than that of overstorey kelps. Thus, for Zostera marina
ecosystems near the low tide mark on Atlantic shores of North America,
the maximum production was -lSOOgC-m^-yr"1, and for a Spartina salt
marsh in the same area near high water mark, the maximum production
found was -897 g C m"2 yr"1 (Mann 1973).
Hydrodynamics
Laboratory studies in a recirculating flume 6 m long by 0.15 m wide by
Fonesca et al. (1982) were designed to determine the effect of Zostera
marina shoot density on flows within the meadow. This work was criti-
cized by Gambi, No well, and Jumars (1990) because the flume did not
simulate dynamically similar conditions to those present in the field, in
particular, the flume was too narrow, resulting in flow blockages when
seagrass shoots were present.
A larger annular flume with an 8-m-long working section, 2 m wide,
was used by Gambi et al. (1990). This flume had a water depth of 25 cm
and a 4-cm-deep sand layer on the flume floor into which young Z.
marina plants were inserted. The eelgrass plants used were -16 cm in
height so that flow could form above the bed. The experimental eelgrass
patch (15 by 100 cm long) was placed where the flume boundary layer
was fully developed and so that the plants were 30 cm from the walls.
Flow was deflected around and over the patch so that there was a sharp
interface near the shoot tips where high shear and turbulence intensity
occurred. Within the eelgrass shoots, velocity was reduced as a function
of shoot density and distance into the meadow (Table 8.3). In these
experiments, the penetration offloweffects into the eelgrass bed is more
marked than the density effect.
Field studies by Ackerman (1983), Fonesca et al. (1983), and Eckman
(1987) also suggest that reduced velocities occur within natural Zostera
meadows. Calculations by Eckman (1987) to determine the flux rates
through a high density (604-1134 plants m~2) and a low density (183-271
plants m"2) eelgrass meadow showed that at slower free-stream veloci-
Table 8.3. Percentages of water flux reduction at

different freestream velocities, shoot densities, and
sampling positions in flume experiments with
Zostera marina.
Downstream sampling positions

Density
(shoots m~2) 25 cm 50 cm 75 cm 100 cm
Freestream velocity 5 cm s
1200 20.7 38.3 43.1 35.2
1000 11.0 31.4 26.7 40.6
800 7.55 19.9 21.4 28.9
600 3.58 12.3 13.7 20.3
400 9.6 18.7 29.7 21.8
Freestream velocity 10 cm-

1200 11.5 30.0 30.7 31.5
1000 5.6 20.6 30.9 31.5
800 12.3 23.0 31.4 32.6
600 0.9 9.8 19.5 20.5
400 8.0 19.4 27.6 22.1
Freestream velocity 20 cm
1200 8.9 16.6 19.9 34.2
1000 10.4 19.6 23.16 21.0
800 2.2 13.4 21.0 18.3
600 5.2 13.9 21.7 19.8
400 2.0 8.3 15.2 14.7
Source: Gambi et al. (1990).
ties, there was a 16% reduction in flux due to the higher density. At a
higher velocity, difference was increased to 52%. The field results do not
agree very well with the flume results shown in Table 8.3, perhaps be-
cause the latter are artifacts of the small bed size tested.
Field observations with a dual axis electromagnetic current meter
suggested to Ackerman and Okubo (1993) that the plants themselves
were generating mechanical turbulence. The frequency generated by the
eelgrass plants of 6.4-8.0 s"1 results because they are hydroelastically
translating fluid energy.
The ecosystem implications of seawaterflowsto and in seagrass mead-
ows include larval supply (Chap. 3), effects on seagrass production,
nutrient exchange between sediment and seawater inclusive of re-
suspension and deposition, direct shear stress on seagrass plant stability,
and flux reduction within seagrass which results in increased sedimenta-

tion and limits suspension feeder production. Some of these are con-
sidered in the following sections, although additional flow-related
phenomena - effects on diffusion of gases across leaf surfaces, seagrass
pollination, and increased seawater turbidity, which results in reduced
light and therefore less photosynthesis - are not because they are consid-
ered more appropriate to discussion of lower hierarchical levels of biol-
ogy than ecosystem level.
Seagrass productivity
The effect of tidal flows on seagrass productivity (Table 8.4, number 2)
was investigated, apparently independently, by Conover (1968) on
eelgrass meadows of the U.S. East Coast and by Odum (1961,1971) and
Steever, Warren, and Niering (1976) on cordgrass meadows on the same
coast.
In Conover's study, eelgrass standing crops of various locations are
presented as a function of the mean hourly tidal velocity averaged over
a complete springs to neaps period (Conover 1968). Where the current
meter was deployed in this study was not stated; presumably free-stream
velocities seaward of the eelgrass meadow are meant. The data for
Charlestown Pond, Rhode Island, are shown in Fig. 8.6. Standing crop
data in Fig. 8.6 are probably a good indication of eelgrass primary pro-
ductivity because herbivores are unimportant in this ecosystem. A char-
acteristic unimodal response to velocity is seen. Conover (1968) sug-
gested that the positive relationship could be explained by an increased
photosynthetic activity as flows increased, and that the inverse relation-
ship was due to an inhibition of metabolism at high flows and/or the
shearing effect of high flow, which prevented the plants from becoming
established. Further results from Texas and from Massachusetts and
Rhode Island along the U.S. East Coast for both seagrasses and
seaweeds (Fig. 8.7) show a similar trend. The data confirm that individual
species of canopy-forming macroalgae have distinctively different re-
sponses to tidal flows, confirming the suggestion made previously for
kelp.
Steever et al. (1976) compared the standing crops of Spartina
alterniflora at various marshes along the Connecticut, United States,
coast. For 10 such marshes in 1971, the standing crop dry weight of
cordgrass was linearly related to mean tidal height (how this was esti-
mated was not stated) with a calculated slope equal to 544 g dry
Table 8.4. Ecosystem hypotheses referable to seagrass meadows.

1 Ho Velocity in seagrass meadows is Fonesca et al. (1982)
unaffected by the presence of seagrass Eckman (1987)
shoots Gambi et al. (1990)
Hx Velocity within a seagrass meadow is
dependent on the seagrass shoot
density
2 Ho Primary production within seagrass Odum (1961, 1971)
meadows is unaffected by tidal factors Conover (1968)
Hx Primary production of seagrass
meadows is a unimodal function of
tidal factors
3 Ho Intermediate pools of sedimentary Kenworthy (1982)
nitrogen are unaffected by seagrass
cover
Hj Intermediate pools of sedimentary
nitrogen are enhanced in seagrass by
sediment trapping
4 Ho Productivity of suspension feeders Kerswill (1949)
within seagrass meadows is Peterson et al. (1984)
independent of shoot density related Eckman (1987)
fluxes
Hx Productivity of suspension feeders
within seagrass meadows is controlled
by shoot density related fluxes
5 Ho POM and DOM arising from the Lin (1989)
seagrass meadow have no effect on
resident suspension feeders
Hx POM and DOM arising from the
seagrass meadow provide additional
food, which enhances suspension
feeder productivity
6 Ho Energetic cost of maintaining Myers (1977)
suspension feeder shore position is Peterson and Beal
unrelated to local velocities (1989)
Hx Energetic cost of maintaining Irlandi and Peterson
suspension feeder shore position is (1991)
directly related to local velocities
weight m 2 for every increase of 1 m in mean tidal range and r = 0.96.

Similar data for other years and sites suggest a similar trend but with
increased scatter, consistent with involvement of other factors besides
tidal height. Although not observed in the data reported by Steever et al.
CL 200 -
O
D C CM
U'E
CD
\
W
,*
D)100 - \
CO
25 50 75
U, cm s"
Figure 8.6 Standing crop of Zostera marina in Charlestown Pond, Rhode Is-
land, 1962-63, as a function of velocity (average lunar period velocity expressed,
cm-s" 1 ) at different sites (Conover 1968).
200
Figure 8.7 Standing crop data for seaweeds and seagrasses for various locations
on the U.S. East Coast where summer temperatures are ~23-28C and salinities
are -29-31 o/oo; dashed line divides lagoon (to left) from channel and coastal
ecosystems: A, Ruppia maritima; B, Thalassia testudinum; C, Stylophora
rhizoides; D, Zostera marina; E, Chondrus crispus; F, Laminaria agardhii; G,
Bryopsis plumosa (Conover 1968).
(1976), the response is considered by these authors to be unimodal; they

cite as an example published data from the Bay of Fundy where the S.
alterniflora plants are dwarfed and mean tidal ranges exceed 8 m. As for
the eelgrass data, the mechanisms involved were not investigated, al-
though Steever et al. (1976) follow Odum (1971) in suggesting that tidal
flows increase the availability of nutrients, such as nitrogen, for cordgrass
growth.
Field studies in a freshwater tidal marsh in Augustine Creek, Georgia,
were undertaken by Odum, Birch, and Cooley (1983) to compare the
productivity of an impounded marsh and one subject to mean natural
tidal amplitudes of 0.5 m. The major plant producer in the marsh was the
giant cutgrass Zizaniopsis miliacea, which produced significantly more
above-ground production, -31%, in tidal than non-tidal conditions in
accordance with Ylx of the unimodal response hypothesis (number 2,
Table 8.4).
Study sites in North Carolina and Rhode Island were used by Fonesca
et al. (1982) to relate field-measured current velocities to functional
aspects of eelgrass ecosystems. One of the study locations was at
Charlestown Pond, worked earlier by Conover (1968), although this
reference is absent from the reference list of Fonesca et al. Current
profiles were made with a bidirectional electromagnetic current meter
for one complete ebb or flood tide (whichever was greater) and the
results corrected to give maximum velocities for surface flows above the
meadow. Surface velocities correlated with meadow mounding, defined
by the height/length ratio of the meadows, h T1, such that at h-l~l> 0.1,
f/max>94crn-s"1, at /i-/"1 = 0.1-0.09, Umax -53-94 cm -s"1, and at
h T1 < 0.09, Umax < 53 cm s"1. The regulation of mounding by strong tidal
currents was earlier described by Scoffin (1970), and Fonesca et al.
(1982) also found that the continuity of seagrass cover was inversely
related to velocity. Unfortunately, this latter work was not compared to
the clear demonstration by Conover (1968) of the effect of tidalflowson
eelgrass productivity.
Sedimentation and nutrient cycling

A partial energy budget for a Zostera marina meadow located in the
Newport River estuary, North Carolina, was presented by Thayer,
Adams, and LaCroix (1975). We have adapted this, using carbon equiva-
lents, in constructing Fig. 8.8. Three sources of primary production are
available in this particular eelgrass meadow - from eelgrass itself 62%
Figure 8.8 Partial c a r b o n budget, g C - m 2 y r \ for a N o r t h Carolina estuary

Zoster a marina seagrass m e a d o w (Thayer et al. 1975).
of total primary production, equivalent to 1545kCal-m 2-yr *; from

phytoplankton -30%; and from benthic microalgae -8%. The two latter
minor primary producers are omitted from Fig. 8.8 in order to simplify
the diagram.
In contrast to the kelp forest budget of Fig. 8.3, the carbon of the
eelgrass meadow passes through a primarily sedimentary detritus cycle
and there are no important herbivores of the seagrass plants. Microbial
degradation of dead seagrass releases nutrients, e.g. nitrogen and phos-

phorous containing compounds, which are available for recycling by
growing seagrass plants. The main export from the meadow may be
partially degraded plant material in POM. In Fig. 8.8, the benthic
epifauna and infauna have been shown as their functional groups, hence
deposit and suspension feeders. The question as to whether suspension
feeders consume POM and its attached microflora, and/or phyto-
plankton largely advected into the meadow, was not determined by
Thayer et al. (1975). The latter authors estimate that the benthic
macrofauna - suspension and deposit feeders - consume an amount of
energy equivalent to 55% of the total available primary production of
the eelgrass meadow.
The result of reduced velocities and turbulence intensity within
seagrass meadows is that sedimentation rates increase. Thus, in the
North Carolina Zostera marina meadow studied by Thayer et al. (1975),
sedimentation rates ranging from 1 to 57 mm yr"1 occurred with a carbon
content on a dry weight basis ranging from 0.7% to 5.6%. The ability
of seagrasses such as Thalassia testudinum in Bimini Lagoon, Bahamas,
to trap and bind sediments was qualitatively studied by Scoffin (1970).
As well as creating increased sedimentation within the meadow, the
seagrass roots and rhizome serve to stabilize the sediments by reducing
erosion events due particularly to wave action.
The silt-clay proportion within the sediments of a seagrass meadow is
a rough indicator of its depositional environment. Where a wide range of
tidal currents is available, as on the U.S. East Coast in studies by Fonesca
et al. (1982), there is a linear relationship between shear stress and
decreasing proportion of silt-clay size particles within the seagrass
meadow sediment. For these reasons, Fonesca et al. (1982) pointed out
that seagrass beds of different h l~l ratios would also have characteristi-
cally different inputs of POM derived from seagrass plants. Thus, where
h r 1 > 0.1 and significant mounding was present, the discrete patches of
seagrass would be sources of POM, whereas where h I'1 < 0.09 and a
greater cover of seagrass was present, such areas would be greater sinks
of seagrass-derived POM.
Nitrogen cycling at Back Sound, North Carolina, was studied by
Kenworthy, Zeiman, and Thayer (1982), who conclusively showed that
the seagrass meadow was significantly enriched by comparison to a ref-
erence location where seagrasses were absent (Table 8.4, number 3).
Both Zostera marina and Halodule wrightii were present and resulted in
elevated levels of silt-clay, organic matter, exchangeable ammonium,

ammonium dissolved in pore water, and total nitrogen within the
meadow. The enrichment effect was strongest at mid-meadow, interme-
diate at the edges or in isolated, mounded patches, and small or absent in
unvegetated reference locations. Although the major mechanism in-
volved in this enrichment of the seagrass meadow involves the hydro-
dynamic-seagrass meadow interaction mentioned earlier, another
mechanism was recognized by Jordan and Valiela (1982) and Bertness
(1984). Ribbed mussels Geukensia demissus present within a Spartina
alterniflora meadow in New England caused a similar enrichment effect.
During feeding, the mussels capture seston advected into the bed, as well
as seagrass-derived POM, and produce faeces and pseudofaeces which
are deposited within the meadow. After aerobic microbial breakdown,
such materials release plant nutrients, e.g. nitrogen and phosphorus
compounds, which can be used by the seagrass plants during photosyn-
thesis. It has been confirmed experimentally that blue mussels Mytilus
edulis also increase Zostera marina growth in this way in the western
Baltic (Reusch, Chapman, and Groger 1994).
Nitrogen cycling within a tropical seagrass meadow in Jamaica domi-
nated by Halodule beaudetti was studied by Blackburn, Nedwell, and
Wiebe (1994), who found active nitrogen and carbon cycling within the
root-sediment environment. Nitrogen-fixing bacteria associated with
the roots, as confirmed by Shieh, Simidu, and Maruyama (1989) for Z.
marina roots, may be a major pathway, although direct uptake from
seawater and after POM decay within sediments may be an alternate
pathway. Blackburn et al. (1994) found an ammonia accumulation within
the sediment which occurred overnight and which disappeared during
the day as a result of active plant uptake. Measured rates of nitrogen
fixation in the tropical sediment were lower than denitrification rates,
rendering a conclusion about the source of the ammonia pool as
problematic.
Nutrient cycling studies clearly require a detailed knowledge of the
mechanisms of transport of nutrients as both imports and exports of the
seagrass ecosystem. Such studies include that by Valiela et al. (1978) in
Great Sippewissett marsh, Massachusetts. This marsh was dominated by
Spartina alterniflora and S. patens. Dissolved plant nutrients could enter
the seagrass ecosystem via ground water or rain or during tidal flooding.
During tidal ebbing, the seawater may transport nutrients away from the
meadow. Seasonal observations and calculations suggest that overall net
SCARP BLOWOUT FLOOR LEEWARD SLOPE
Figure 8.9 Diagrammatic cross section of a crescent-shaped "blowout" in a

Thalassia testudinum meadow at Bath, Barbados (from Patriquin 1975).
flux into or out of the seagrass meadow is complex, but generally is a net
importer during the summer when the Spartina plants are growing and a
net exporter during winter. Wolaver et al. (1983), working on another S.
alterniflora meadow in Virginia, United States, also found pronounced
seasonal differences in fluxes to and from this ecosystem. These studies
were aided by deploying a field flume (see Chap. 2) to aid in sampling
tidal fluxes to and from the seagrass meadow.
The extensive body of work on nutrient fluxes to and from seagrass
meadows has been reviewed by Hemminga, Harrison, and van Leut
(1991). The inputs include nitrogen fixation, sedimentation, and direct
uptake from seawater advected to the marsh. The losses include diffu-
sion and tidal transport away from the meadow, denitrification, export of
POM, loss due to herbivory, and possible transport away from the
meadow, as well as direct exudation of nutrients from seagrass roots and
leaves.
Seagrass stability
Qualitative observations regarding the stability of seagrass meadows
have been made by Scoffin (1970), Patriquin (1975), Orth (1977),
Fonesca and Kenworthy (1987), and Fonesca et al. (1983). In general, the
"blowouts" - crescent-shaped areas of erosion (Fig. 8.9) where water
movement has torn away the seagrass plants - are associated with iso-
lated patches or mounds of seagrass on the seaward face of the meadow.
As classified by Fonesca et al. (1983), these areas are also associated with
the highest water movement energies.
The blowouts are usually associated with extreme wave energy, as may
occur during storms and may result in uprooting and formation of large
drifting mats of seagrasses in the Gulf of Mexico and the Caribbean
(Patriquin 1975). Scoffin (1970) described blowouts in Bimini Lagoon
which were caused by locally high tidal velocities. Preliminary observa-
tions made by Scoffin (1970) in an underwater flume (Chap. 2) suggest
that the shear forces associated with unidirectionalflowsresult in erosion
of sediment around the roots of Thalassia testudinum in the free-stream
velocity range of 50-150 cm-s"1.
Suspension feeder production

Hydrodynamic baffling by the shoots within a seagrass meadow could
result in limits to suspension feeding and hence growth and survival of
suspension feeders present there. This results because the flux of seston
passing through the seagrass meadow is reduced by comparison with
unvegetated adjacent areas and hence reduces the availability of seston
as food (Table 8.4, number 4). Some field tests of hypothesis 4 are
presented in Table 8.5. Of three independent tests involving the hard
clam, one (Kerswill 1949) supports Yix and two (Peterson, Summerson,
and Duncan 1984; Irlandi and Peterson 1991) support Ho. In Argopecten
irradians, and for 1 year of the two studied, Hx is supported, while the
null hypothesis is supported in Anomia simplex. Thus, of five independ-
ent field experiments, only two support the alternative hypothesis that a
reduced seston flux in the seagrass meadow results in reduced feeding
and production.
Clearly the results of Table 8.5 indicate a more complex situation than
suggested by the simple flux hypothesis tested. The low bivalve densities
and short path lengths used (Table 8.5) preclude a population level
seston depletion effect (Chap. 7) causing a downstream reduction in
feeding/growth except, perhaps, in the experiment of Kerswill (1949).
Thus, if feeding effects are involved in the results on bivalve growth rate,
it is individual physiological effects which can be expected, and thus
velocity, seston concentration, or seston quality (Chap. 4) should be the
controlling variable. Unfortunately, little is known about the feeding and
growth response of M. mercenaria to these variables.
Other possibilities which might explain field experimental results to
test the intensity of turbulent mixing on bivalve growth, both within and
Table 8.5. Field experiments to compare the effect of lower turbulent mixing intensities on suspension-feeding bivalve
growth within seagrass meadows compared to an unvegetated nearby reference location.
Tidal flow
(cm- Bivalve Bivalve Bivalve growth rate
density path length
Seagrass meadow Bivalve ^max (no. m"2) (m) Seagrass Reference References
"Eelgrass" Mercenaria mercenaria >20 8 69-101 0.6-1.2 Low High Kerswill (1949)
Zostera marina and M. mercenaria 7 0.7-1.3 9 High Low Peterson et al.
Halodule wrightii (1984)
Z. marina and M. mercenaria 27 0.3 49 1.0 High Zero Irlandi and
H. wrightii Peterson (1991)
Z. marina Argopecten irradians 13-98 ? 22 Low High Eckman (1987)
Z. marina Anomia simplex 13-98 ? 35 Same Same Eckman (1987)
outside a seagrass meadow (Table 8.5), include the food subsidy hypoth-
esis of Table 8.4 (number 5) within seagrass meadows and the physiologi-
cal cost for bivalve feeders to maintain their position on the shore, which
is proportional to local velocities (number 6). Field experiments under-
taken by Lin (1989) in a North Carolina Spartina alterniflora meadow
involved the ribbed mussel Geukensia demissa, which occurred in a 1- to
2-m wide strip on the seaward side. Manipulative experiments involving
singleton or small groups of 10 or 30 mussels were undertaken at differ-
ent tidal levels and positions within the seagrass meadow. Seston deple-
tion effects could be excluded as an explanation for differences in growth
because treatments at the same tidal level gave identical results. At
one of the two sites studied, ribbed mussels grew much faster in the
middle of the seagrass meadow than at the edge, consistent with pres-
ence of a food subsidy (number 5, Table 8.4). Judge, Coen, and Heck
(1993) found that seston concentrations were enriched at <5cm above
the sediment-water interface. The chlorophyll-containing seston in these
samples was made up of 90% pennate diatoms, although an estimate of
non-living POM was not made. For both of these studies the link show-
ing that the living or non-living POM enhanced bivalve growth was not
made.
That the energetic cost of maintaining shore position is higher where
water movement energies are higher and sediments less stable (Table
8.4, number 6) has been suggested repeatedly. Irlandi and Peterson
(1991), for example, were able to show by seagrass-canopy trimming
experiments that this reduced hard clam growth, consistent with in-
creased velocities in these plots and circumstantial evidence to support
Hi in number 6. They also observed that hard clams grew significantly
more slowly at the seagrass meadow edge where higher velocities were
also present.
Because multiple limiting factors are probably involved in interpreting
the Table 8.5 results, one should also consider biotic factors. The effect
of seagrass beds on predator-prey relations was reviewed by Orth, Heck,
and van MontFrans (1984). Effects could be separated into those involv-
ing the seagrass root rhizome and others involving the shoots. Peterson
(1982) showed that seagrass meadows provide sufficient below-ground
structural complexity to hinder the efficiency of whelk predators,
Busy con sp., capturing Mercenaria mercenaria. Siphon nipping of the
suspension-feeding bivalve Protothaca staminea by predaceous fish was
found to be 9-11 times greater outside the seagrass meadow than within
it (Peterson and Quammen 1982). Simulated siphon nipping of M.
Table 8.6. Dry weight o/Zostera marina and density 0/Mercenaria

mercenaria in six replicate samples of 1225m2 each, within the seagrass
and at nearby sandflat reference locations.
Average g dry weight of Average Mercenaria

seagrass m"2 no .m~2
Sample
replicate no. Seagrass Sandflat Seagrass Sandflat
1 41.6 0.0 9.68 1.44

2 113.6 0.0 10.24 2.00
3 164.8 0.0 7.32 1.88
4 78.4 0.0 9.12 1.00
5 57.6 0.0 9.12 1.32
6 64.0 0.0 8.76 1.88
Average 86.8 0.0 9.04 1.60
Source: Peterson et al. (1984).
mercenaria was shown to have different impacts on growth, dependent

on whether the clams were present within a seagrass meadow (Coen and
Heck 1991). Nipped clams within the meadow grew more slowly than
nipped clams on sand. The siphon nipping effects on growth included
reduced feeding opportunities and costs of regenerating the siphon in
addition to the actual tissue removal by the fish.
Although it is not possible to conclude exactly what mechanisms are at
work in the experiments outlined in Table 8.5, the following are some
possibilities. Seston quality and quantity are greater within the meadow
and result in higher hard clam growth there (Table 8.6). Alternatively,
higher flows at the reference location may inhibit growth directly or
involve higher energetic costs in maintaining position in the less stable
sediments. This is consistent with the canopy trimming experiment of
Irlandi and Peterson (1991), which caused a reduction in growth of hard
clams commensurate with increased velocities. Predator responses
within a vegetated area resulting in decreased predation cannot be ruled
out. In Eckman's (1987) 2-yr Long Island study, in one of the years
worked, the alternate hypothesis (number 4, Table 8.4) was supported
for a location characterized by slow velocities. At a location where
velocities were faster, clam growth for both low and high eelgrass
shoot density treatments was low. Another possibility was that the
seagrass stems provide a refuge from predators which move on the
sediment-water surface, e.g. crabs (Pohle et al. 1991; Ambrose and

Irlandi 1992).
Bivalve reefs
The potential for suspension feeding bivalves to form dense
aggregations, or reefs, occurs in suitable locations throughout the world.
The reefs are most common at intertidal or nearshore subtidal locations,
particularly in estuaries or other areas of high productivity or tidal en-
ergy. Common bivalves which form reefs include the mytilid mussels and
oysters (Ostreidae).
Hydrodynamics
We know of no studies that have set out to determine specifically the
hydrodynamic variables within a natural bivalve reef, although some
have done this incidentally in determining filtration rates of the bivalves.
Such studies include those of Frechette and Bourget (1985) and
Frechette, Butman, and Rockwell Geyer (1989) in the field, Butman et
al. (1994) in a laboratory flume, and Monosmith et al. (1990) for model
bivalve populations in a flume.
The study by Frechette et al. (1989) concerned an intertidal mussel reef
which was 30 m wide by 100 m long in the St. Lawrence estuary, where the
tidal range was -4.8 m. The average biomass for the Mytilus edulis reef
was 500 g dry-m"2. Seawater sampling to determine phytoplankton con-
centration and measure filtration rates by standard physiological meth-
ods was at three heights above the reef: 1.0,0.5, and 0.02-0.05 m. Samples
were drawn continuously by a centrifugal pump at 3.5L-min~1 in a 1.3-
cm-inner-diameter (ID) plastic hose. The current velocities were deter-
mined by continuously recording Aanderaa current meters fitted with
salinity and temperature probes. Current meter rotors measured veloci-
ties at 1.0 and 0.01-0.1 m above the reef. By using benthic boundary layer
theory (Schlichting 1968) and velocity data, the logarithmic velocity
profile was constructed; from it the friction velocity, /*, and bottom
roughness parameter, Zo, were determined. From their detailed hydrody-
namic model analysis, Frechette et al. (1989) were able to show that the
commonly observed tendency for the mussels to form hummocks by
overgrowing other mussels may result in bottom roughness parameter
increases sufficient to increase turbulent transfer of seston to the reef
significantly.
Reef metabolism
Studies which have measured bivalve reef metabolism (Table 8.7) are
considered here in detail. The aim is to determine whether bivalve reef
oxygen uptake data support the conclusions of Chapters 4 and 7 that
increased seawater velocity can enhance bivalve growth by increasing
the turbulent supply of sestonic food. The results summarized in Table
8.7 are of total metabolism of the reef, inclusive of photosynthetic oxy-
gen production, e.g. from phytoplankton in seawater or diatoms on the
reef surface. The only seasonal study of metabolism of which we are
aware is presented by Dame et al. (1992) for a South Carolina oyster
reef. This study found net oxygen production in winter when macroalgae
were present and denoted by a minus sign in Table 8.7; during the
summer, the oyster reef was strongly heterotrophic with a net oxygen
demand (BOD). Physiological data for oyster oxygen uptake rates dis-
cussed in Dame, Spurrier, and Zingmark (1992) suggest that the oysters
are only responsible for 10% of the annual reef oxygen uptake. Other
uptake sources are aerobic bacteria, meiofauna, other macrofauna, and
the chemical oxygen demand (COD) of the reef sediments.
During summer observations of a Narragansett Bay mussel reef,
Nixon et al. (1971) attempted to determine the effect of tidal velocity on
reef metabolism. Dissolved oxygen determinations in triplicate were
measured by Winckler titration. Seawater samples were obtained up-
stream of the seaward edge of the reef and downstream in the dominant
tidal direction, some 20-25 m away. Depth at which the samples were
taken was not stated, and a crude measure of velocity was made by
observing dye or drift bottle movements, so that velocities may be inac-
curate. The data in Fig. 8.10 were obtained by subtracting the down-
stream from the upstream dissolved oxygen onfloodingtides. The differ-
ence - assuming a well mixed tidal prism above the mussel reef -
represents the oxygen uptake calculated for the area over which the
water mass has moved in its passage between the two sampling points.
The instantaneous rate of BOD, /, must be corrected for the volumetric
flow rate which passes over the reef (Chap. 2, the section, Field flows).
Do the results of Fig. 8.10 support the conclusions of Chapters 4 and 7
and the thesis that water movement energy is a community metabolism
multiplier (Table 8.8, number 1)? Assuming that the reef filtration rates
are directly proportional to metabolic uptake of oxygen, the results do
support the theoretical curve of Fig. 4.1 for a bivalve with inhalant
normal to theflowand also the empirical fits of Figs. 7.7 and 7.8 based on
Table 8.7. Comparison of independent studies to measure bivalve reef metabolism.
Variable Nixon et al. (1971) Dankers et al. (1989) Dame et al. (1992)
Location Narragansett Bay, Rhode Island of Texel, North Inlet, South Carolina,
Island, U.S.A. Netherlands U.S.A.
Dominant species Mytilus edulis M. edulis Crassostrea virginica
Temperature (C) 20.1-23.5 14-16 8-30
Salinity (o/oo) 27-30 ? 15-35
Reef area (m2) 100 ? 600
Path length of reef measured (m) 20-25 10 10
Bivalve density (m~2) 6261 7 7
Bivalve dry meat biomass (g m~2) 1523 818 196
Reef respiration rate (g O2 m"2 h"1) 0.2-2.8 1.2-3.9 -1.5 to +4.0
8-i

v 6-
CM

' 4-
C\J
o
oO)
^r 2 -
o
>
0-
10 20 30 40 50 60 70
1
I/, cm s"
Figure 8.10 Instantaneous rate of oxygen uptake (/) of a Narragansett Bay blue
mussel reef as a function of free-stream velocity, U, cm-s"1: solid dots, field
derived data; open dots, recalculated laboratory data. The fitted hyperbola has
r = 0.83, with single arrowed point excluded (from Nixon et al. 1971).
population growth of mussels and scallops. There is likely to be an upper

limit to the hyperbola of Fig. 8.10 beyond which oxygen uptake of the
reef declines. One point - shown in Fig. 8.10 with an arrow and which
was omitted from the regression fitting of Nixon et al. (1971) - could
indicate a limiting flow consistent with the oxygen uptake's being
unimodal. Without knowing the blue mussels' physiological responses
(filtration and metabolic rates) to flow, it is impossible to interpret fully
the results of Nixon et al. (1971). A further difficulty in interpretation is
that the proportion of reef oxygen demand due to mussels was not
determined, although because the mussel biomass involved (Table 8.7) is
approximately eight times higher than for the oyster reef, it is likely to be
higher than the 10% found for the latter.
In the study by Dame et al. (1992), a benthic ecosystem tunnel (Chap.
2, the section, Field flows) was used to obtain metabolic and nutrient flux
rates. Typical results of a flood-ebb cycle are shown in Fig. 8.11. During
the flood tide (I/~7-13cm-s" 1 ), the net uptake of oxygen was 1.1-
3.2gO2-m"2-rT1, and during the ebb tide (t/~8-25cm-s" 1 ), a net release
of up to 3.5 g O2 m"2 h"1 occurred. A net release of oxygen did not begin
until the ebb velocity exceeded 15cm-s"1. During these observations,
c/)
1525 1730 2126

TIME, h
Figure 8.11 Community respiration rate of a South Carolina oyster reef during
a flood-ebb cycle in summer 1983: A, tidal velocity; B, measured dissolved
oxygen concentration of experimental tunnel, C, / = instantaneous rate of oxy-
gen uptake (Dame et al. 1992): , inlet; o, outlet.
oysters fed on the flood tide but not on the ebb, when much of the seston
had already been removed by the oysters'filtering.This interpretation by
Dame et al. (1992) is consistent with other field studies which linked tidal
cycles with feeding as indicated by digestive cycles. Thus, mussels M.
edulis (Langton 1977) and sublittoral Pecten maximus (Mathers 1976)
fed on the flood tide. In the latter case, two distinct subpopulations of
scallops were present - one facing into and feeding on the flood, the
other away from the flooding current and feeding on the ebb current.
Histological studies by Mathers (1976) showed that the digestive physiol-
ogy of each subpopulation was ~6h out of phase, supporting a tidal
periodicity of digestion. Langton and Gabbott (1974) also described a
tidal periodicity of feeding in oysters Ostrea edulis grown near low tide
on the shore at Menai Bridge, North Wales, by monitoring the crystalline
style physiology.
In the study described by Dankers, Dame, and Kersting (1989), two
methods of measurement were compared: a 200-m2 concrete tank sup-
plied with mussels and natural seawater and part of a natural blue mussel
reef enclosed by the tunnel method of Dame, Zingmark, and Haskin
(1984). Results of the former gave average net uptakes of 0.5-1.0g
O2 m~2 h"1, which were somewhat lower than similar values for a natural
mussel reef (Table 8.7). The discrepancy is explained by the sensitivity of
the results to the timing of measurements. Hence, net oxygenfluxeswere
low during the day when photosynthetic production of oxygen was oc-
curring. Nighttime observations were correspondingly high because this
production was absent.
An independent estimation of mussel oxygen uptake was made by
Dankers et al. (1989) from standard physiological data linked to biomass.
This yields a value of 0.37mgO2-g~1*h~1 equivalent to an uptake of
0.3 g O2 m~2 h"1 by the mussel reef. This represents only 8% to 25% of
the day-night dissolved oxygen flux, as determined by the tunnel
method.
Pelagic-benthic coupling
A direct link between the surface primary productivity of the oceans and
benthic productivity, or biomass, of the same area was proposed by
Rowe (1971). From published data he showed that surface primary pro-
ductivity ranked second only to depth in controlling benthic biomass.
Thus, carbon fixed photosynthetically in surface waters, and after pas-
sage through a pelagic food web, reached benthos in the form of faecal
pellets (Hargrave 1973). Rowe et al. (1975) also showed that the ex-
change was a two-way process with benthically regenerated plant nutri-
ents, e.g. various forms of nitrogen, recycled to primary producers in
shallow surface coastal waters. It was thus possible to speak of pelagic-
benthic coupling involving positive-negative feedback loops between
primary and secondary benthic producers.
Typically, energy or carbon budgets are calculated on the basis of unit
area - usually 1 m2 - close to the area actually sampled by a randomly
placed grab or corer used in estimating soft sediment benthic biomass
and hence secondary benthic production. Primary production, too, is
estimated on the basis of unit area or volume. Thus, the first efforts to
link pelagic and benthicfluxeswere made on the basis of small vertically
linked areas. A simple regression model proposed by Hargrave (1973)
was successful in predicting the magnitude of the annual sediment oxy-
gen demand on the basis of the amount of pelagic primary production
and the mixed layer depth of the water column at the same site.
Yet, some authors found discrepancies in balancing energy budgets if
a tight vertical link between pelagos and benthos was assumed. Thus,
Bernard (1974), who described a partial energy budget for an oyster
Crassostrea gigas reef in British Columbia, showed that the energy avail-
able from the local primary production was much less than the oysters'
allocation of energy to the annual production of gametes. Emerson,
Roff, and Wildish (1986) provided a partial carbon budget based on one
site in the Bay of Fundy (Fig. 8.12) - a macrotidal estuarine ecosystem in
eastern Canada. Pelagic production in the Bay over depositional LaHave
clay sediments provided an input of 60gC-m~2-yr~1 from a total
phytoplankton primary production of 133 g C m~2 yr"1 to the benthos.
All of this input could not be utilized by macrofauna or accounted for by
the measured sediment burial rates. It was suggested (Emerson et al.
1986) that the pelagic input, consisting of equal amounts of detritus,
copepod faeces, and microalgae, was tidally transported by tidal excur-
sions of -20 km to a harder sediment on which a productive horse mussel
Modiolus modiolus reef was situated (Table 8.8, number 2). The maxi-
mum production of the macrofauna here was 148gC-m~2-yr"1 at depths
of 60-100 m. Assuming a euphotic depth of 20-40 m, this exceeds the
mean primary production of phytoplankton in the column of seawater
above this part of the lower Bay. It also provides circumstantial evidence
for an alternative source of food by turbulent tidal transport, other than
the direct vertical link between pelagos and benthos proposed by Rowe
(1971). A further example of a partial carbon budget estimate on the
tidal flats of Konigshafen in the German Wadden Sea was provided by
Asmus and Asmus (1990). It also shows the inadequacy of local
autochthonous primary production in the seawater above a blue mussel
reef to provide sufficient carbon sources for the mussel production real-
ized. Estimates of primary production by local phytoplankton amount to
73 g C m~2 yr"1 compared to secondary production of the mussel reef of
264gC-m~2-yr~1. This led Asmus and Asmus (1990) to conclude that the
high suspension feeder production is maintained by tidal transport of
seston from the North Sea or other parts of the Wadden Sea.
HCO3 "MICROBIAL LOOP"

IMPORT
NUTRIENTS- ONI-
-HPHYTOPLANKTON
DOM BACTERIA
i33r>. [12-66]
POM POM
COPEPODS CILIATES FLAGELLATES
16
HERRING
Figure 8.12 Partial carbon budget for the Bay of Fundy ecosystem on LaHave
clay sediments at 50-80m (Emerson et al. 1986): annual production, gC-nT2-
yr"1; bracketed figure estimated by the method of Azam et al. (1983).
The idea that bivalve reefs were an important part of pelagic-benthic

coupling (Table 8.8, number 3) was proposed by Dame et al. (1980) and
supported by an analysis of input-output energy flow analysis of a South
Carolina oyster reef (Dame and Patten 1981). The concept was further
developed on a South Carolina oyster reef dominated by Crassostrea
Table 8.8. Ecosystem level hypotheses referable to bivalve reefs

and flow.

1 Ho Water movement energy has no Odum (1971)
effect as a community Nixon et al. (1971)
metabolism multiplier
Hx Increasing water movement
energy acts as a community
metabolism multiplier
2 Ho Tidal or wind-driven transport of Odum (1971)
seston has no effect on bivalve Emerson et al. (1986)
reef production
Hx Tidal or wind-driven transport of
seston may support bivalve reef
production in excess of locally
available resources
3 Ho Bivalve reefs are unimportant in Dame et al. (1980)
materials cycling between pelagic
and benthic subsystems
Hx Bivalve reefs represent a major
coupling between pelagos and
benthos and hence in materials
cycling between these subsystems
4 Ho Bivalve reefs have no effect on Dame and Dankers (1988)
retaining reactive nutrients in Dame and Libes (1993)
estuaries
Hx The net effect of bivalve reefs is
to retain reactive nutrients, such
as nitrogen, within estuaries
5 Ho Bivalve reefs do not influence the Cloern (1982)
development of microalgal Officer et al. (1982)
blooms in eutrophic estuaries Nichols (1985)
Hx Bivalve reefs influence the
development of microalgal
blooms in eutrophic estuaries
virginica as a subsystem of an estuarine ecosystem (Dame et al. 1984). A

primary concern in making their measurements was to design samplers
which could estimate the flux of materials, particularly the importance of
volumetric flow to this calculation. Various benthic tunnels or open
flumes for field use have been designed (Chap. 2, the section, Field flows)
for this purpose. From tidal velocity measurements made within the
tunnel, the volumetric flow could be calculated and hence the flux of
materials determined. Dame et al. (1984) showed that the oyster reef
reduced particulate organic carbon and chlorophyll a in thefloodingtidal
prism and added significant amounts of excreted ammonia as it ebbed.
Using the benthic ecosystem tunnel, Dame et al. (1992) have determined
the annual flux of ammonia to and from the oyster reef with a net annual
release of 124.8 g-m^-yr"1. In another study, Dame and Libes (1993)
investigated plant nutrient fluxes after experimental manipulations in-
volving removing oysters in some plots. Because the recycling rates of
the oysters were faster than the tidal flushing rates, nutrients were recy-
cled in situ. The controlling feedback loop for nitrogen was suggested
to be the exchange between phytoplankton and oysters, the latter retain-
ing nitrogen within the body.
Field studies on Dutch and German Wadden Sea blue mussel reefs
(Dame and Dankers 1988; Prins and Smaal 1990; Asmus, Asmus, and
Reese 1990; Smaal and Prins 1993) have confirmed hypothesis 4 (Table
8.8): that large quantities of seston are filtered from seawater; faecal and
pseudofaecal biodeposits are produced in large quantities, the latter are
rapidly remineralized, resulting in release of dissolved and inorganic
nutrients in seawater, and this results in stimulation of further
phytoplankton primary production as well as export of nutrients at cer-
tain times of the year. The results are also consistent with H2 of number
3 (Table 8.8), that is that the reef represents a major coupling between
pelagos and benthos.
Nitrogen cycling was investigated by Kaspar et al. (1985) at a green
mussel Perna canaliculus aquaculture site in the Marlborough Sounds,
New Zealand. The main pathways of the nitrogen cycle considered are
shown in Fig. 8.13. Losses of nitrogen occurred when mussels were
removed at harvest and by denitrification processes in sediments and
seawater. Additions occurred from mussel excretion, mainly of ammo-
nia, and the microbial transfer of various organic nitrogen sources which
were present within the sediments. Nitrogen fixation may also have been
important but was not considered in this study. Although it was clear that
there was considerable tidal energy available at the farm, e.g.
Umax = 111 cm-s"1 and a tidal prism of 3-4m, no estimate of tidal trans-
port of nitrogen into or out of the farm area was included in this study.
By comparison with surface sediments at a nearby reference site, signifi-
cantly higher denitrification and nitrogen mineralization rates were
present at the farm.
Figure 8.13 Nitrogen cycle of a green mussel Perna canaliculus farm, Marl-
borough Sounds, New Zealand (Kaspar et al. 1985).
Eutrophication and bivalve reefs

In temperate eutrophic estuaries such as San Francisco Bay on the U.S.
West Coast, the absence of phytoplankton blooms is contrary to the
frequent observation that hypernutrification results in blooms and in-
creases primary production (Neilson and Cronin 1981). Nevertheless,
unexpectedly low densities of phytoplankton do occur in some eutrophic
estuaries, and it has been proposed (Table 8.8, number 5) that the
phytoplankton are limited by the grazing power of extensive reefs of
suspension feeding bivalves as in San Francisco Bay. The presence of the
introduced soft-shell clam Mya arenaria there at densities up to 20,000
clams m~2 was sufficient to filter the whole water column once per day on
the basis of laboratory filtering rate estimates (Nichols 1985). It is clear
that San Francisco Bay and the Sacramento-San Joaquin river system
which drains into it have been highly modified by a range of human
activities (Nichols et al. 1986). In 1986, a new invasion of an exotic
suspension feeding bivalve, the Asian clam Potamocorbula amuriensis,
began and rapidly spread throughout the estuary, reaching densities of

>10,000 clams-m"2 (Carlton et al. 1990). Since the San Francisco Bay
ecosystem is clearly not in a steady state, decisions on the original hy-
pothesis number 5 (Table 8.8) are problematic. Multiple causes of chlo-
rophyll a variability in the landward end of San Francisco Bay, and
including hydrodynamic causes as well as losses to benthic grazers, have
been considered by Jassby and Powell (1994).
Another eutrophic estuary, the 180-km-long Potomac on the U.S. East
Coast, opens into Chesapeake Bay and has a freshwater tidal zone, or
"stuw zone," of 60km (Cohen et al. 1984). After the introduction of
another exotic, the clam Corbicula fluminea, into the freshwater stuw
zone, densities reached -1500 clams-m"2 with a maximum biomass of
3 kg nT2. In 1980 and 1981, a phytoplankton sag curve of 40% to 60% less
density than that of adjacent areas spread for 6-8km within the stuw
zone. The sag coincided with the maximum density of the clam distribu-
tion, thus providing circumstantial evidence to support hypothesis
number 5 in Table 8.8. Other possible explanations for the phytoplankton
sag curve, such as decreases in freshwater discharge, which increase the
retention time of the stuw zone; the filtering power of zooplankton; the
presence of toxic substances, which inhibit phytoplankton growth; or the
lack of plant nutrients, which limits phytoplankton growth, were not
supported by direct observation. Because all of the water in the affected
zone could have been filtered by the clams in 3-4 d, similar to the resi-
dence time there, this is circumstantial evidence supporting the hypoth-
esis that the clams' grazing was responsible for the phytoplankton sag
curve.
Hypernutrification may result in stimulation of species of phyto-
plankton that are trophically unacceptable to benthic grazers and hence
unable to increase productivity through the pelagic-benthic link. As an
example, the "brown tide" blooms along the U.S. East Coast caused
widespread mortality rates of blue mussels (Cosper et al. 1987). The
brown tide bloom consisted of up to 106 cells-ml 1 of a 2-|Lim-long
chrysophycean, Aureoccus anophagefferens, which was shown by Tracey
(1988) to inhibit feeding by M. edulis in laboratory tests. Because of the
persistence of the brown tide, it was suggested that the mussels died of
starvation. Prins, Dankers, and Smaal (1994) have reported reduction in
feeding caused by presence of the haptophycean Phaeocystis sp., which
dominated the phytoplankton for 2 mo in the spring in the Netherlands.
Blue mussels placed in a concrete flume 40 x 5 m wide were supplied with
natural seawater pumped from the North Sea and showed reduced
growth during the Phaeocystis bloom.
We should expect more instances where benthic grazers are unable to

utilize the enhanced production of eutrophication. Possibly the study
reported by Loo and Rosenberg (1989) in the eutrophic Lalholm Bay,
Kattegat, is a case in point. The nearshore benthos here was dominated
by grazers - Cardium edule and My a arenaria - and population filtration
experiments with local seawater showed that they filtered at suboptimum
rates. Loo and Rosenberg (1989) were unable to pinpoint the cause,
although low rates of turbulent transfer from the surface to the bottom
were suggested. The study reported by Peterson, Irlandi, and Black
(1994) in a coastal lagoon on the south coast of Australia could also be
an example. Here, populations of two suspension feeding bivalves,
Katelysia scalarina and K. rhytiphora, suddenly crashed, coincident with
the onset of eutrophication. Unfortunately, examination of phyto-
plankton was not carried out during the study.
Summary
Two common features can be discerned in our review of the three
distinct ecosystems we have chosen for inclusion here: kelp forests
seagrass meadows, and bivalve reefs. The first is that there is evidence in
the perturbation theory for the suggestion by Odum et al. (1979) that
ecosystem productivity is a unimodal function of water motion. The
second common feature is that each species of kelp, seagrass, and bivalve
appears to have a characteristic unimodal response to water movement
with distinct tolerance ranges of water movement, as found by Conover
(1968) for some species of kelps and seagrasses. In future work, it should
be possible to utilize this relationship to predict water movement ener-
gies and perhaps also secondary heterotrophic productivity of suspen-
sion feeders.
Ecosystem analysis requires consideration and resolution of the tem-
poral and spatial scales to be used in the study. It is simply not sufficient
for a biologist to set seasonal and geographic limits arbitrarily. This
should be done after a scoping study involving a biologist and a physical
oceanographer. This point of view is forced on us by the examples
reviewed here, which include the South African kelp ecosystem studies
reported by Wulff and Field (1983), the Bay of Fundy horse mussel reef
studies reported in Emerson et al. (1986), and the Wadden Sea blue
mussel reef studies of Asmus and Asmus (1990). These reefs were shown
by Asmus (1994) to depend trophically on turbulent advective transfer of
seston from sources external to the reef subsystem. Recent studies, e.g.
by Grant et al. (1993), use an interdisciplinary approach and ecosystem
simulation modelling, in which the adjacent hydrodynamics is included,

to represent fluxes and flows within a blue mussel farm in Nova Scotia,
Canada.
Despite the interest in suspension feeding bivalves which occur within
seagrass meadows, it has proven difficult to conclude why bivalve growth
or productivity is so different within and outside the meadow. One
obvious reason for this difficulty is that the basic physiological response
of the bivalves present in seagrass meadows toflowis unknown, and this
should be rectified. Transplant field experiments of bivalves to seagrass
meadows and adjacent reference areas without seagrass may involve
complexities which render them difficult to interpret. This is because
multiple factors may be involved in each seagrass meadow, so it is likely
that determining the cause may prove to be elusive, but possible, if a
combination of different hierarchical levels of the scientific method is
employed.
We believe that measurement of the instantaneous rate of oxygen
uptake of bivalve reefs is a valuable independent metabolic determina-
tion for ecosystem analysis. Further development of the methods to be
used in its determination is required (see Chap. 2), and its results should
be compared with those of the more conventional sediment-oxygen
demand enclosed chamber technique. Methodology development is also
required in measuring enzyme activity and growth as a means of estimat-
ing instantaneous production relatable to instantaneous oxygen uptake
rate.
The modelling study of Koseff et al. (1993) shows that a
phytoplankton bloom in estuaries like San Francisco Bay depends on a
key physical factor - stratification. The latter phenomenon confines
phytoplankton cells in the photic zone, where photosynthesis proceeds
rapidly, and also limits the vertical transfer of pelagos to the benthos.
Some limited wind-wave mixing of stratified surface water is required
for optimal bloom formation. During stratification events, the
phytoplankton cells are effectively de-coupled from the benthic grazers.
With regard to the hypotheses presented in this chapter (Tables 8.1,
8.2, 8.4), we consider that further experimental work is required to
reject Ho in number 1, Table 8.1, and numbers 5 and 6 in Table 8.4. Some
doubt, due to either conflicting results or inconclusive rejection of Ho,
means that further experimental work is required with numbers 2,3, and
4 in Table 8.1; numbers 4 and 5 in Table 8.2; and numbers 2 and 4 in
Table 8.4. For the remaining six hypotheses, we conclude that Hx
has been sufficiently supported in experimental tests. Although some
evidence in support of Hx for all hypotheses in Table 8.8 is given, we

believe that further experimental conformation is necessary in geo-
graphically separate locations before these propositions can be robustly
accepted.
References
Ackerman, J. D. 1983. Current flow around Zostera marina plants and flowers:
implications for submarine pollution. Biol. Bull. 165: 504.
Ackerman, J. D., and A. Okubo. 1993. Reduced mixing in a marine
macrophyte canopy. Funct. Ecol. 7: 305-309.
Ambrose, W. G., Jr., and E. A. Irlandi. 1992. Height of attachment on
seagrass leads to trade-off between growth and survival in the bay scallop
Argopecten irradians. Mar. Ecol. Prog. Ser. 90: 45-51.
Anderson, S. M., and A. C. Charters. 1982. A fluid dynamic study of seawater
flow through Gelidium nudifrons. Limnol. Oceanogr. 27: 399-412.
Asmus, H. 1994. Benthic grazers and suspension feeders: which one assumes
the energetic dominance in Kongigshafen? Helg. Meeresunters. 48: 217-
231.
Asmus, H., and R. M. Asmus. 1990. Trophic relationships in tidal flat areas: to
what extent are tidal flat areas dependent on imported food? Neth. J. Sea
Res. 27: 93-99.
Asmus, H., R. Asmus, and K. Reese. 1990. Exchange processes in an intertidal
mussel bed: a Sylt-flume study in the Wadden Sea. Ber. Biol. Anst.
Helgoland. 6: 1-79.
Azam, F., T. Fenchel, J. G. Field, J. S. Gray, L. A. Meyer-Reil, and F.
Thingstad. 1983. The ecological role of water-column microbes in the sea.
Backus, G. J. 1994. Coral reef ecosystems. A.A. Balkema, Rotterdam.
Bernard, F. R. 1974. Annual biodeposition and gross energy budget of mature
Pacific oysters, Crassostrea gigas. J. Fish. Res. Board Can. 31: 185-190.
Bertness, M. D. 1984. Ribbed mussels and Spartina alterniflora production in a
New England salt marsh. Ecology 65: 1794-1807.
Blackburn, T. H., D. B. Nedwell, and W. J. Wiebe. 1994. Active mineral
cycling in a Jamaican seagrass sediment. Mar. Ecol. Prog. Ser. 110: 233-
239.
Butman, C. A., M. Frechette, W. Rockwell Geyer, and V. R. Starczak. 1994.
Flume experiments on food supply to the blue mussel, Mytilus edulis L.,
as a function of boundary layer flow. Limnol. Oceanogr. 39: 1755-1768.
Carlton, J. T., J. K. Thompson, L. E. Schemel, and F. H. Nichols. 1990.
Remarkable invasion of San Francisco Bay (California, USA) by the
Asian clam Potamocorbula amurensis. I. Introduction and dispersal.
Cloern, J. E. 1982. Does the benthos control phytoplankton biomass in South
San Francisco Bay? Mar. Ecol. Prog. Ser. 9: 191-202.
Coen, L. D., and K. L. Heck. 1991. The interacting effects of siphon nipping
and habitat on the bivalve (Mercenaria mercenaria (L.)) growth in a
subtropical seagrass {Halodule wrightii Aschers) meadow. J. Exp. Mar.
Biol. Ecol. 145: 1-13.
Cohen, R. R. H., P. V. Dresler, E. J. P. Phillips, and R. L. Cory. 1984. The

effect of the Asiatic clam, Corbiculafluminea,on phytoplankton of the
Potomac River, Maryland. Limnol. Oceanogr. 29: 170-180.
Connell, J. T. 1972. Community interactions on marine rocky intertidal shores.
Ann. Rev. Ecol. Syst. 3: 169-192.
Conover, J. T. 1968. The importance of natural diffusion gradients and
transport of substances related to benthic marine plant metabolism.
Bot. Mar. 11: 1-9.
Cosper, E. M., W. C. Dennison, E. J. Carpenter, V. M. Bricelj, J. G. Mitchell,
S. H. Kennster, D. Colflesh, and M. Dewey. 1987. Recurrent and
persistent "brown tide" blooms perturb coastal marine ecosystem.
Estuaries 10: 284-290.
Dame, R. F., and N. Dankers. 1988. Uptake and release of materials by a
Wadden Sea mussel bed. J. Exp. Mar. Biol. Ecol. 118: 207-216.
Dame, R., and S. Libes. 1993. Oyster reefs and nutrient retention in tidal
creeks. J. Exp. Mar. Biol. Ecol. 171: 251-258.
Dame, R., F., and B. C. Patten. 1981. Analysis of energy flows in an intertidal
oyster reef. Mar. Ecol. Prog. Ser. 5: 115-124.
Dame, R. F., J. D. Spurrier, and R. G. Zingmark. 1992. In situ metabolism of
an oyster reef. J. Exp. Mar. Biol. Ecol. 164: 147-160.
Dame, R., R. Zingmark, H. Stevenson, and D. Nelson. 1980. Filter feeder
coupling between the water column and benthic subsystem, p. 520-526. In
V. S. Kennedy (ed.) Estuarine perspectives. Academic Press, New York.
Dankers, N., R. Dame, and K. Kersting. 1989. The oxygen consumption of
mussel beds in the Dutch Wadden Sea. Sci. Mar. 53: 473-476.
Dayton, P. K. 1971. Competition, disturbance and community organization:
the provision and subsequent utilization of space in a rocky intertidal
community. Ecol. Monogr. 41: 351-389.
1975. Experimental evaluation of ecological dominance in a rocky intertidal
community. Ecol. Monogr. 45: 137-159.
1985. The structure and regulation of some South American kelp
communities. Ecol. Monogr. 55: 447-468.
Den Hartog, C. 1970. The seagrasses of the world. North Holland,
Amsterdam.
Doty, M. S. 1971. Measurements of water movement in reference to benthic
algal growth. Bot. Mar. 14: 32-35.
Duggins, D. O., and J. E. Eckman. 1994. The role of kelp detritus in the
growth of benthic suspension feeders in an understorey kelp forest.
Duggins, D. O., C. A. Simenstad, and J. A. Estes. 1989. Magnification of
secondary production by kelp detritus in coastal ecosystems. Science 245:
170-173.
Eckman, J. E. 1987. The role of hydrodynamics in recruitment, growth and
survival of Argopecten irradians (L.) and Anomia simplex (D'Orbigny)
within eelgrass meadows. J. Exp. Mar. Biol. Ecol. 106: 165-191.
Eckman, J. E., and D. O. Duggins. 1991. Life and death beneath macrophyte
canopies: effects of understorey kelps on growth rates and survival of
marine benthic suspension feeders. Oecologia 87: 473-487.
Eckman, J. E., D. O. Duggins, and A. T. Sewell. 1989. Ecology of understorey
kelp environments. I. Effects of kelps on flow and particle transport near

the bottom. J. Exp. Mar. Biol. Ecol. 129: 173-188.
Emerson, C. W., J. C. Roff, and D. J. Wildish. 1986. Pelagic-benthic energy
coupling at the mouth of the Bay of Fundy. Ophelia 26: 165-180.
Estes, J. A., and J. F. Palmisano. 1974. Sea otters: their role in structuring
nearshore communities. Science 185: 1058-1060.
Fonesca, M. S., J. S. Fisher, J. C. Zieman, and G. W. Thayer. 1982. Influence
of the seagrass, Zostera marina L., on current flow. Est. Coast. Shelf Sci.
15: 351-362.
Fonesca, M. S., and W. J. Kenworthy. 1987. Effects of current on
photosynthesis and distribution of seagrasses. Aquat. Bot. 27: 59-78.
Fonesca, M. S., J. Zieman, G. W. Thayer, and J. S. Fisher. 1983. The role of
current velocity in structuring eelgrass {Zostera marina L.) meadows.
Est. Coast. Shelf Sci. 17: 367-380.
Frechette, M., and E. Bourget. 1985. Energy flow between pelagic and benthic
zones: factors controlling particulate organic matter available to an
intertidal mussel bed. Can. J. Fish. Aquat. Sci. 42: 1166-1170.
Frechette, M, C. A. Butman, and W. Rockwell Geyer. 1989. The importance
of boundary-layer flows in supplying phytoplankton to the suspension
feeder, Mytilus edulis L. Limnol. Oceanogr. 34: 19-36.
Gambi, M. C, A. R. M. No well, and P. A. Jumars. 1990. Flume observations
on flow dynamics in Zostera marina (eelgrass) beds. Mar. Ecol. Prog. Ser.
61: 159-169.
Gerard, V. A., and K. H. Mann. 1979. Growth and production of Laminaria
longicornis (Phaeophyta) populations exposed to different intensities of
water movement. J. Phycol. 15: 33-41.
Grant, J., M. Dowd, K. Thompson, C. Emerson, and A. Hatcher. 1993.
Perspectives on field studies and related biological models of bivalve
growth and carrying capacity. NATO ASI Ser. G33: 371^20.
Hagen, N. T., and K. H. Mann. 1994. Experimental analysis of factors
influencing the aggregating behaviour of the green sea urchin
Strongylocentrotus droebachiensis (Miiller). J. Exp. Mar. Biol. Ecol. 176:
107-126.
Hargrave, B. T. 1973. Coupling carbon flow through some pelagic and benthic
communities. J. Fish. Res. Board Can. 30: 1317-1326.
Hemminga, M. A., P. G. Harrison, and F. vanLeut. 1991. The balance of
nutrient losses and gains in seagrass meadows. Mar. Ecol. Prog. Ser. 71:
85-96.
Irlandi, E. A., and C. H. Peterson. 1991. Modification of animal habitat by
large plants: mechanisms by which seagrasses influence clam growth.
Oecologia 87: 307-318.
Jackson, G. A. 1984. Internal wave attenuation by coastal kelp strands.
J. Phys. Oceanogr. 14: 1300-1306.
Jackson, G. A., and C. D. Winant. 1983. Effect of a kelp forest on coastal
currents. Cont. Shelf. Res. 2: 75-80.
Jassby, A. D., and T. M. Powell. 1994. Hydrodynamic influences on
interannual chlorophyll variability in an estuary: upper San Francisco
Bay-Delta (California, U.S.A.). Est. Coast. Shelf Sci. 39: 595-618.
Jordan, T. E., and I. Valiela. 1982. A nitrogen budget of the ribbed mussel
Geukensia demissa and its significance in nitrogen flow in a New England
salt marsh. Limnol. Oceanogr. 27: 75-90.
Juday, C. 1940. The annual energy budget of an inland lake. Ecology 21: 438-
450.
Judge, M. L., L. D. Coen, and K. L. Heck, Jr. 1993. Does Mercenaria
mercenaria encounter elevated food levels in seagrass beds? Results from
a novel technique to collect suspended food resources. Mar. Ecol. Prog.
Ser. 92: 141-150.
Kaspar, H. F., P. A. Gillespie, I. C. Boyer, and A. L. MacKenzie. 1985. Effects
of mussel aquaculture on the nitrogen cycle and benthic communities in
Kenepuru Sound, Marlborough Sounds, New Zealand. Mar. Biol. 85: 127-
136.
Kenworthy, W. J., J. C. Zeiman, and G. W. Thayer. 1982. Evidence for the
influence of seagrasses on the benthic nitrogen cycle in a coastal plain
estuary near Beaufort, North Carolina (USA). Oecologia 54: 152-158.
Koehl, M. A. R. 1986. Seaweeds in moving water: form and mechanical
function, p. 603-634. In T. J. Givnish (ed.) On the economy of plant form
and function. Cambridge University Press, Cambridge.
Koehl, M. A. R., and R. S. Alberte. 1988. Flow, flapping and photosynthesis of
Nereocystis lewkeona: a functional comparison of undulate and flat blade
morphologies. Mar. Biol. 99: 435^44.
Koseff, J. R., J. K. Holen, S. G. Monismith, and J. E. Cloern. 1993. Coupled
effects of vertical mixing and benthic grazing on phytoplankton
populations in shallow turbid estuaries. J. Mar. Res. 51: 843-868.
Langton, R. W. 1977. Digestive rhythms in the mussel, Mytilus edulis.
Mar. Biol. 41: 53-58.
Langton, R. W., and P. A. Gabbott. 1974. The tidal rhythm of extracellular
digestion and the response to feeding in Ostrea edulis L. Mar. Biol. 24:
181-187.
Leigh, E. G., R. T. Paine, J. F. Quinn, and T. H. Suchanek. 1987. Wave energy
and intertidal productivity. Proc. Natl. Acad. Aci. U. S. A. 84: 1314-1318.
Lin, J. 1989. Importance in location in the salt marsh and clump size on
growth of ribbed mussels. J. Exp. Mar. Biol. Ecol. 128: 75-86.
Lindeman, R. L. 1942. The trophic-dynamic aspect of ecology. Ecology 23:
399^18.
Loo, L., and R. Rosenberg. 1989. Bivalve suspension-feeding dynamics and
benthic-pelagic coupling in an eutrophicated marine bay. J. Exp. Mar.
Biol. Ecol. 130: 253-276.
Mann, K. H. 1972. Ecological energetics of the seaweed zone in a marine
bay on the Atlantic coast of Canada. II. Productivity of the seaweed.
Mar. Biol. 14: 199-209.
1973. Seaweeds: their productivity and strategy for growth. Science 182:
975.
Mathers, N. F. 1976. The effects of tidal currents on the rhythms of feeding
and digestion in Pecten maximus L. J. Exp. Mar. Biol. Ecol. 24: 271-284.
Miller, R. J., and K. H. Mann. 1973. Ecological energetics of the seaweed zone
in a marine bay on the Atlantic coast of Canada. III. Energy
transformations by sea urchins. Mar. Biol. 18: 99-114.
Menge, B. A., and J. P. Sutherland. 1976. Species diversity gradients: synthesis
of the roles of predation, competition and temporal heterogeneity. Am.
Nat. 110: 351-369.

Oceanogr. 35: 680-696.
Munk, W., and G. A. Riley. 1952. Absorption of nutrients by aquatic plants.
J. Mar. Res. 11: 215-240.
Myers, A. 1977. Sediment processing in a marine subtidal sandy bottom
community. II. Biological consequences. J. Mar. Res. 35: 633-647.
Neilson, B. J., and L. E. Cronin (Editors). 1981. Estuaries and nutrients.
Humana Press, Clifton, N. J.
Nichols, F. H. 1985. Increased benthic grazing: an alternative explanation for
low phytoplankton biomass in northern San Francisco Bay during the
1976-1977 drought. Est. Coast. Shelf Sci. 21: 379-388.
Nichols, F. H., J. E. Cloern, S. N. Luoma, and D. H. Peterson. 1986. The
modification of an estuary. Science 231: 525-648.
Odum, E. P. 1961. The role of tidal marshes in estuarine production.
Conservationist (New York State Conservation Department, Albany,
N. Y.) 15: 1215.
1971. Fundamentals of ecology. 3rd ed. W. B. Saunders, Philadelphia.
1980. The status of three ecosystem level hypotheses regarding salt marsh
estuaries: tidal subsidy, outwelling and detritus-based food chains, p. 485-
495. In V. S. Kennedy (ed.) Estuarine perspectives. Academic Press, New
York.
Odum, E. P., J. B. Birch, and J. L. Cooley. 1983. Comparison of giant cutgrass
productivity in tidal and impounded marshes with special reference to
tidal subsidy and waste assimilation. Estuaries 6: 88-94.
Odum, E. P., J. T. Finn, and E. H. Franz. 1979. Perturbation theory and the
subsidy-stress gradient. Bioscience 29: 349-352.
Officer, C. B., T. J. Smayda, and R. Mann. 1982. Benthic filter feeding: a
natural eutrophication control. Mar. Ecol. Prog. Ser. 9: 203-210.
Orth, R. J. 1977. The importance of sediment stability in seagrass
communities, p. 281-300. InB. C. Coull (ed.) Ecology of marine benthos.
University of South Carolina Press, Columbia.
Orth, R. J., K. L. Heck, and J. V. van Montfrans. 1984. Faunal communities in
seagrass beds: a review of the influence of plant structure and prey
characteristics on predator-prey relationships. Estuaries 7: 339-350.
Paine, R. T. 1979. Disaster, catastrophe and local persistence of the sea palm
Postelsia palmaeformis. Science 205: 685-687.
Patriquin, D. G. 1975. Migration of blowouts in seagrass beds at Barbados and
Carracou, West Indies and its ecological and geological implications.
Aquat. Bot. 1: 163-189.
Peterson, C. H. 1982. Clam predation by whelks (Busycon sp.): experimental
tests of the importance of prey size, prey density and seagrass cover.
Mar. Biol. 66: 159-170.
Peterson, C. H., and B. F. Beal. 1989. Bivalve growth and higher order
interactions: importance of density, site and time. Ecology 70: 1390-
1404.
Peterson, C. H., E. A. Irlandi, and R. Black. 1994. The crash in suspension-
feeding bivalve populations (Katelysia spp.) in Princess Royal Harbour:
an unexpected consequence of eutrophication. J. Exp. Mar. Biol. Ecol.
176: 39-52.
Peterson, C. H., and M. L. Quammen. 1982. Siphon nipping: its importance to
small fishes and its impact on growth of the bivalve Protothaca staminea
(Conrad). J. Exp. Mar. Biol. Ecol. 63: 249-268.
Peterson, C. H., H. C. Summerson, and P. B. Duncan. 1984. The influence of
seagrass cover on population structure and individual growth rate of a
suspension-feeding bivalve, Mercenaria mercenaria. J. Mar. Res. 42: 123-
138.
Pohle, D. G., V. M. Bricelj, and Z. Garcia-Esquivel. 1991. The eelgrass
canopy: an above-bottom refuge from benthic predators for juvenile bay
scallops Argopecten irradians. Mar. Ecol. Prog. Ser. 74: 47-59.
Prins, T. C, N. Dankers, and A. C. Smaal. 1994. Seasonal variation in the
filtration rates of a semi-natural mussel bed in relation to seston
composition. J. Exp. Mar. Biol. Ecol. 176: 69-88.
Prins, T. C, and A. C. Smaal. 1990. Benthic-pelagic coupling: the release of
inorganic nutrients by an intertidal bed of Mytilus edulis. p. 89-103. In M.
Barnes and R. N. Gibson (eds). 24th European Mar. Biol. Symp.
Aberdeen University Press.
Raupach, M. R., and A. S. Thorn. 1981. Turbulence in and above plant
canopies. Ann. Rev. Fluid. Mech. 13: 97-130.
Reusch, T. B. H., A. R. O. Chapman, and J. P. Groger. 1994. Blue mussels
Mytilus edulis do not interfere with eelgrass Zostera marina but fertilize
shoot growth through biodeposition. Mar. Ecol. Prog. Ser. 108: 205-282.
Rowe, G. T. 1971. Benthic biomass and surface productivity, p. 441-454. In
J. D. Costlow (ed.) Fertility in the sea. Gordon & Breach, New York.
Rowe, G. T., C. H. Clifford, K. L. Smith, and P. C. Hamilton. 1975. Benthic
nutrient regeneration and its coupling to primary productivity in coastal
waters. Nature 255: 215-217.
Schlichting, H. 1968. Boundary-layer theory. 6th ed. J. Kestin (trans.)
McGraw-Hill, New York.
Scoffin, T. P. 1970. The trapping and binding of sub tidal carbonate sediments
by marine vegetation in Bimini Lagoon, Bahamas. J. Sed. Petrol. 40: 249-
273.
Shieh, W. Y., U. Simidu, and Y. Maruyama. 1989. Enumeration of nitrogen-
fixing bacteria in an eelgrass {Zostera marina L.) bed. Microb. Ecol. 18:
249-259.
Smaal, A. C, and T. C. Prins. 1993. The uptake of organic matter and the
release of inorganic nutrients by bivalve suspension feeder beds, p. 271-
298. In R. F. Dame (ed.) Bivalve feeders in estuarine and coastal
ecosystem processes. NATO ASI Ser. Vol. G33. Springer-Verlag, Berlin.
Sorokin, Y. I. 1993. Coral reef ecology. Springer-Verlag, Berlin.
Steever, E. Z., R. S. Warren, and W. A. Niering. 1976. Tidal energy subsidy
and standing crop production of Spartina alterniflora. Est. Coast. Mar. Sci.
4: 473-478.
Thayer, G. W., S. M. Adams, and M. W. LaCroix. 1975. Structural and
functional aspects of a recently established Zostera marina community. In
L. E. Cronin (ed.) Estuarine research. Vol. 1. p. 518-540. Academic Press,
New York.
Tracey, G. A. 1988. Feeding reduction, reproductive failure, and mortality in
Mytilus edulis during the 1985 "brown tide" in Narragansett Bay, Rhode
Island. Mar. Ecol. Prog. Ser. 50: 73-81.
Valiela, L, J. M. Teal, S. Volkman, D. Shafer, and E. J. Carpenter. 1978.
Nutrient and particulate fluxes in a salt marsh ecosystem: tidal exchanges
and inputs by precipitation and ground water. Limnol. Oceanogr. 23: 798-
812.
Wheeler, W. N. 1980. Effect of boundary layer transport on the fixation of
carbon by the giant kelp Macrocystis pyrifera. Mar. Biol. 56: 103-110.
Wickens, P. A., and J. G. Field. 1986. The effect of water transport on
nitrogen flow through a kelp bed community. S. Afr. J. Mar. Sci. 4: 79-92.
Wolaver, T. G., J. C. Zieman, R. Wetzel, and K. L. Webb. 1983. Tidal
exchange of nitrogen and phosphorus between a mesohaline vegetated
marsh and the surrounding estuary in the lower Chesapeake Bay. Est.
Coast. Shelf Sci. 16: 321-332.
Wulff, F. V., and J. G. Field. 1983. Importance of different trophic pathways in
a nearshore benthic community under upwelling and downwelling
conditions. Mar. Ecol. Prog. Ser. 12: 217-228.
9
Future directions
To get an interesting answer, one needs to pose a pertinent question.

In this final chapter, our aim is two-fold. Firstly, we refine a method for
choosing pertinent scientific questions. Secondly, we present some sug-
gestions for interdisciplinary benthic biology/hydrodynamic questions
which occur to us. If the latter questions prove to be of no, or less than
riveting, interest to those active in benthic biology in the future, then we
may console ourselves with the knowledge that our readers should be
able to choose pertinent questions for themselves. Wefirmlybelieve that
the creative natural scientist is rooted in extraordinary powers of obser-
vation so that she/he can
Stare in wonder
At this world of beauty
and, by this, be motivated enough to determine how it works.
To recapitulate what is presented in the preceding six chapters (Table
9.1): Each of the subjects reviewed is variable in the attention or success
of the studies that are focused on it. This is clearly shown by the rank
order of the number of hypotheses associated with each chapter (Table
9.1). Also shown is the associated dominant theory for each subject.
This may also be the best point at which to summarize our scientific
judgement on hypotheses dealt with in Chapters 3 through 8 (Table 9.2).
Such information is of help in choosing novel questions for further
scientific study. Obviously, this will involve hypotheses which do not
appear in Table 9.2.
Selecting pertinent scientific questions

The general aspects of asking pertinent scientific questions were consid-
ered in Chapter 1. There we learned the difference between proximate
362
Future Directions 363
Table 9.1. Rank order of the number of hypotheses concerning flow

and suspension feeders, populations, or ecosystems.
Number of
hypotheses Subject Chapter Associated theory
21 Ecosystems Suspension-feeding benthic
communities are important in
pelagic-benthic materials
cycling
15 Larval biology Passive-active theory of
factors influencing the early
life history recruitment of
benthic suspension feeders
Benthic populations Theory of benthic limitation
byflow:trophic group
mutual exclusion; benthic
impoverishment by water
movement
5 Initial suspension- 4 Unimodal theory of
feeding responses suspension feeding to flow,
and bivalve pump theory
4 Behavioral 6 Aggregation theory
responses to flow
0 Filtration physiology 5 Particle collection theory
and ultimate, and, at least until now, our questions were limited to the
former. In this chapter, perhaps because all we are doing is asking and
need not answer them here, we have the luxury of posing both proximate
and ultimate questions.
We believe that there are three important additional steps which have
to be answered affirmatively in selecting a pertinent scientific question:
Is the question of personal interest to the scientist?
Is the question posed testable by scientific methods?
Will the information gained in the answer be relevant?
The first two points seem self-evident. Because of the considerable
amount of work required to complete a research project, an individual
with sufficient motivation and commitment is required to complete it.
We include here the private, moral responsibility of an individual scien-
tist to ascertain, before it is begun, that the research contemplated is not
contrary to the common good. There will always be some questions that
Table 9.2. Judgements on whether alternate hypotheses are upheld by the experimental evidence presented
in Chapters 3-8."
Further
Table No. Hypothesis Yes No tests
3.3 1 Estuarine retention of larvae X X
2 Passive shoreward transport by wind X X
3 Passive shoreward transport by internal waves X X
4 Passive shoreward transport by internal tidal bores X X
5 Passive shoreward transport by upwelling X X
6 Plant canopies entrain larvae X X
3.6 1 Post-larval dispersal supply X X

2 Post-settlement mortality is flow related X
3.7 1 Passive larval settlement - soft X

2 Active larval settlement - soft X
3 Passive larval recruitment near small protuberances X
3.8 1 Passive larval settlement - hard X

2 Active larval settlement to surface roughness X
3 Active larval settlement to chemical inducers X
4 Benthic density dominated by supply side ecology X X
4.5 1 Filtration of sieving suspension feeders a unimodal function of flow X

2 Combined passive-active pumping enhancing filtration X
3 Energetic costs of filtration low and independent of pump output X X
4 Filtration rate and orientation to flow unrelated in scallops and brachiopods X X
6.8 1 Local flow conditions enhancing group living X X

2 Group living allowing protection from flow X X
3 Group living resulting in improved reproductive success X X
4 Group living resulting in better predation protection X X
7.2 1 Sestonic food depletion limiting suspension feeding populations X

2 Suspension feeder production positively related to flow X X
3 Energetic water movement impoverishment X X
4 Deposit feeder production inversely related to flow X X
5 Sestonic food depletion controlled by suspension feeder density X
6 Tidal excursions controlling reef path lengths of suspension feeders X X
8.1 1 Current velocity inhibited by kelp X

2 Sedimentation greater within a kelp bed X
3 Key predators controlling kelp forests X X
4 Enhanced sedimentation in kelp forests enhancing local production X X
5 Most organic production locally consumed within the kelp forest X X
8.2 1 Enhanced wave action increasing gas or nutrient fluxes X X

2 Wave stirring increasing light and therefore photosynthesis of kelp X
3 Wave exposure resulting in shearing stresses which remove kelp plants X
4 Energetic wave action inhibiting grazing by herbivores X X
5 Strong wave action enhancing kelp competitiveness X X
8.4 1 Current velocity inhibited by seagrasses X

2 Seagrass primary production a unimodal function of flow X
3 Sediment trapping in seagrass beds enhancing nitrogen availability X
4 Suspension feeder production enhanced in seagrass meadows by flux rates X X X
5 Suspension feeder production enhanced in seagrass meadows by POM and DOM X X
6 Energetic costs for shore maintenance dependent on velocity X X
8.8 1 Benthic plant or bivalve productivity enhanced by water movement X X

2 Seston transport enhancing bivalve reef production X X
3 Bivalve reefs enhancing benthic-pelagic coupling X X
4 Bivalve reefs retaining plant nutrients X X
5 Bivalve reefs influencing microalgal bloom development in eutrophic estuaries X X
a
If further tests are needed as confirmation, this is indicated in the final column.
simply cannot be answered, either by the inductive methods we have

eschewed for proximate questions or the deductive ones for ultimate
questions (see Chap. 1, the section, Asking the right question). It is
therefore important to determine in which category the question lies and
abandon it if it is unanswerable.
With the last of the steps, relevancy, involved in examining a potential
scientific question, the deliberations become more complex. When we
speak of the relevancy of scientific questions, we mean their relevancy
either to society as a whole (Peters 1991) or to the particular disciplinary
field of science involved - in this case, benthic biology. The former would
be classified as an applied, and the latter as a fundamental, question.
Fundamental research is equated with curiosity-driven questions, and
applied research to questions of obvious practical concern in human
society. We believe that Peters (1991) has glossed over the difficulty of
recognizing a priori which are the societally relevant questions for Homo
sapiens. As an example, consider the ecological question addressed in
the seston depletion hypothesis of Chapter 7. The origin of this work was
clearly curiosity driven and, therefore, was fundamental research. Fur-
ther development of the concept has generated interest in it as an applied
research method to predict the production potential for bivalve
aquaculture. This application was not clearly foreseen when this work
was initiated in the mid-1970s. For many scientific questions, as for this
one, it is often difficult a priori to determine which are applied and which
fundamental in nature.
Beyond the difficulty of determining a priori the societal relevance of
a particular scientific question, scientists also have a responsibility to
their own discipline to maintain high standards in developing a logical
and ordered body of knowledge. This would involve all of the types of
scientific method outlined earlier in Table 1.4, including those that
clearly have no direct societal relevance, e.g. the naming of animals and
plants according to Linnaean principles. We consider that such funda-
mental research in benthic biology is necessary if we aspire to develop
practically useful predictive models.
We agree with Peters (1991) that the production of universally appli-
cable deterministic predictive models is the most important sign of a well
developed discipline. An example is provided by the discipline of hydro-
dynamics, which has a respectable 150-yr-long history developed by
engineers and scientists, both practically and theoretically. Hydrody-
namic theory has accumulated numerous useful models or empirical
methods which help manipulate water resources. By contrast, benthic
biology is a much younger discipline - 86 years old in 1997, if we accept

1911 as its origin (see Chap. 1). It still has little theory or universally
applicable predictive models, which remain to be developed and fully
accepted. In order that this can be done, resolution of fundamental
questions in benthic biology, as well as repeated efforts to apply this
work, must go on hand in hand. We expect that important parts of this
effort will be dependent upon hydrodynamic theory and practice.
We suggest that it is a healthy state of affairs that a certain tension
between societal and scientific disciplinary relevance questions should be
swirling within the mind of each practising scientist. But we believe that
scientists themselves should have the ultimate responsibility for framing
both fundamental and applied scientific questions.
Proximate fundamental questions

Examples of questions which can be answered by currently available
scientific methods are shown in Table 9.3. Here we have selected a
representative question, or questions, applicable in each chapter, viz: 3,
a and b; 4, c; 5, d; 6, e, f, and g; 7, h and i; 8, j .
Questions a and h in Table 9.3 belong within the theory of mutual
exclusion. Wildish (1977) suggested the active selection of certain flow-
sediment habitats by larval colonizers, or post-settlement competition
among the recruits whose outcome was dependent on flow-sediment
interactions. Butman (1987) also hypothesized about this as follows:
"Suspension feeders may not co-occur with deposit feeders because the
two functional groups have larvae with different fall velocities that are
passively deposited in different fluid-dynamic environments." Thus,
questions a and h in Table 9.3 represent a comprehensive coverage
regarding the relative importance of flow-sediment interactions as they
directly or indirectly affect both passive and active settlement and post-
recruitment mortality. As far as we are aware, such questions have not
yet been addressed in the literature.
We present in Table 9.4 some generic hypotheses developed from
questions a and h of Table 9.3. The field experimental testing of hypoth-
eses 1 and 2 in Table 9.4 might be achieved by providing various hard
surfaces and de-faunated soft sediments in core tubes checked daily for
new larval settlers. We believe daily observations to be a satisfactory
compromise to measure settlement rather than recruitment, although, as
pointed out by Olafsson, Peterson, and Ambrose (1994), this measure
cannot avoid some very early post-settlement mortality. Although it is
Table 9.3. Representative fundamental proximate questions in the interdisciplinaryfieldof hydrodynamics/benthic
biology.
Scale
Spatial Temporal
No. (km) (y) Question
0.1-0.01 0.01 Do hydrodynamic-sediment interactions determine the settlement success of deposit versus
suspension feeding larvae so that one or the other is variably excluded from the population by the
physical conditions?
b 0.1-0.01 0.1 Do hydrodynamic-seston interactions determine the recruitment success of suspension feeders at a
particular location?
c 0.1-0.01 0.1 Can we present a model of the internal regulators and external environmental variables which
influence filtering, feeding, and growth of a named suspension feeder?
d What is the mechanism of filtration in active suspension feeders?
e 0.001 0.0 How do tube-living suspension-feeding epifauna respond behaviorally to flow?
f What ethological changes among epifaunal suspension feeders minimize dislodgement by flow?
g 0.001-0.1 >1 What causes epifaunal suspension feeders to aggregate at some sites?
h 0.1 Do hydrodynamic-sediment interactions determine the post-settlement mortality rate of deposit-
versus suspension-feeding populations in such a way that one or the other is variably excluded by
the physical conditions?
i 0.1 >1 What are the mechanisms of water movement-induced inhibition of suspension feeder
production?
i 1-1000 How do physical oceanographic currents and wind-wave effects interact to influence pelagic-
benthic coupling in a macrotidal estuary?
Table 9.4. Generic hypotheses referable to the special theory of

mutual exclusion.
No. Ho/Hi Hypothesis
1 Ho The settlement of deposit-feeding larvae is less than expected

from physical transport calculations
Hx The settlement of suspension-feeding larvae equals that
expected from physical transport calculations
2 Ho The settlement of deposit feeders is greater on soft substrates
than on hard ones
Hi The settlement of suspension feeders is greater on hard
substrates than on soft ones
3 Ho Post-settlement growth of deposit feeders is inversely related
to velocity but unrelated to seston concentration
Hx Post-settlement growth of suspension feeders is unimodally
related to velocity and seston concentration
4 Ho The post-settlement mortality rate of deposit feeders is
inversely related to velocity
Hx The post-settlement mortality rate of suspension feeders is
unimodally related to velocity
5 Ho The post-settlement dispersal of deposit feeders is to lower
mean velocity conditions
Hx The post-settlement dispersal of suspension feeders is to an
optimum velocity range, characteristic for each species
very difficult, it is not impossible, using modern time recording sensors,

to obtain a measure of the velocity field in the benthic boundary layer
with changing tides so that the passive larval transport rates can be
calculated. For hypothesis 2 (Table 9.4), a number of sites with varying
flow characteristics could be selected so that a range of velocities and,
therefore, passive larval transport rates could be compared. It may be
preferable to run both these experiments in laboratory flumes, in which
case the hydrodynamic transport calculations with a well defined unidi-
rectional flow would be much easier. Special requirements of such flume
experiments would be that they were flowthrough and not recirculating
devices and that the incoming seawater was filtered to exclude naturally
occurring larvae. It should then be possible to control the absolute den-
sities and proportions of suspension and deposit feeders present during
the flume experiments. Settlement responses by the larvae would be
indicated by comparing the proportions of the trophic groups which

colonize each type of substrate.
Post-settlement growth and mortality of deposit- versus suspension-
feeding juveniles (numbers 3 and 4, Table 9.4) probably necessitate a
field experimental test. This is because of the great difficulty of simulat-
ing the complex water movement patterns present in the sea for the
weeks to months required for determining mortality rates. A number of
sites with a range of water movement conditions would be required, and
various substrates precolonized by representative juveniles of each
trophic group placed at each, so that growth and mortality could be
assessed. For completeness in Table 9.4, we have included as number 5 a
post-settlement dispersal hypothesis since some, although not all, species
are potentially able to leave an originally settled, but unsuitable, locality.
Almost certainly it would be a field experimental test, perhaps of the
type presented by Emerson and Grant (1991).
Question (g) of Table 9.3 may prove to be a special case of (a) in the
same table. This is because aggregated larval settlement may be
behaviorally or passively controlled by flow. Thus, field or laboratory
experiments similar to those offered for hypotheses 1 and 2 (Table 9.4)
would serve to sort out the behavioral and/or physical mechanisms in-
volved. Larval responses to constant unidirectional flows in flumes do
not, however, bear much resemblance to flow variation in the field,
where tidal (regular, hourly) and wind-wave (stochastic, days) effects
produce complex and unpredictable patterns. The aggregation may also
be induced by behavioral responses to chemical inducers or cues to
settlement, as has been demonstrated for some species (see review by
Pawlik 1992). This question is of central concern in benthic ecology
because of the ubiquitous occurrence of aggregation among suspension
feeders at various spatial scales. We suggest that it requires the use of
fresh approaches in observing sublittoral spatial distributions of benthos,
as is beginning to be done in some field studies discussed in Chapter 6 at
SCUBA diving depths. We also suggest that interdisciplinary studies
involving biologists, physical oceanographers, and recent sedimentary
geologists should be developed to provide accurate observational data
over small spatial scales (Table 9.3), but longer, at least seasonal, tempo-
ral scales regarding benthic boundary layer flow and sediment
movements.
Whether hydrodynamic-seston interactions are important in the suc-
cess of suspension feeders (Table 9.3, b) provides an alternative hypoth-
esis to that of (a) and (h). The concentration of seston which reaches a
particular location, as well as seston quality, will have an effect on active

suspension feeder growth directly in a physiological sense (Chap. 4).
Indirectly, in an ecological sense (Chap. 7), it will affect population
growth by downstream seston depletion effects. Both direct and indirect
effects can result in starvation and death, affecting the post-settlement
mortality and production of suspension feeders. In an experimental test
of this question, the null hypothesis might be that newly settled larval
deaths and growth are unaffected by changes in seston concentration or
quality. Thus, the alternate hypothesis would be that these seston vari-
ables do affect mortality and production. Laboratory experiments are
envisaged in which the flow is kept constant while observing test panels
with a known density of post-settlement larvae. Different runs of the
experiment, preferably in a multiple-channel flume, could be set up with
different seston concentrations or qualities, and regular observation of
deaths and growth made.
For the physiologically based questions of Table 9.3 (c and d), we
suggest that the conceptual model of scallop feeding in Wildish and
Saulnier (1993) could be applied more generally to active suspension
feeders. With the physiologists' interest in classifying environmental fac-
tors (e.g. Fry 1947), this may assist in determining the nature of the
external factors which affect filtration/feeding. Although the Fry para-
digm (Kerr 1990) is applicable in autecology, we do not believe that it is
applicable in synecology for the reasons explained in Chapter 7. Con-
trary views seem to have been expressed by Kerr (1990) and J0rgensen
(1992a). To understand the internal factors which may affect filtration/
feeding in bivalve molluscs - inclusive of valve and mantle opening/
closing, pumping performance through time, causes of gill bypassing,
seston selection at gills or mouth palps, satiation effects, and how growth
is controlled - it would be necessary to link them with nervous or
neurohormone systems. Detailed knowledge of the integrated control
systems involved would, of course, be of potential value in enhancing the
culture of commercially important species.
As we have seen in Chapter 5, question (d) of Table 9.3 is unanswered
for species of active suspension feeders, such as the bivalve molluscs. All
that we can suggest is more intense direct observation of feeding
behavior in general, with a video camera, so that it is continuous, or with
endoscopes (Chap. 2). Neither of these methods would be able to con-
firm a mechanism directly, but the clues it could provide, coupled with
physical and theoretical modelling of the kind proposed by J0rgensen
(1981), might be enough to devise further testable hypotheses.
For the two ethological questions of Table 9.3 (e and f), we believe
that both field and laboratory observations of a wide range of species of
suspension feeder are required. For locations which are deeper than
SCUBA diving depth, underwater, time-lapse photography, from either
an anchored vessel or a remotely operated vehicle, may be adequate. In
the laboratory, a video camera focused through the acrylic plastic
(Plexiglas) walls of the flume may often be suitable. With the latter
equipment, it should be possible to record tube-living suspension feeder
behavior as a function of unidirectional or oscillatory flows - some
results with spionid polychaetes were discussed in Chapter 6. As postu-
lated in Chapter 4, we would expect feeding to be unimodally related to
velocity, and at the upper end of this relationship, to see behavioral
responses to minimize dislodgement by water movement, e.g. halting
feeding and retiring within the tube. Free-living epifauna may also re-
spond in a characteristic way toflowswhich are nearly strong enough to
sweep them away. We would expect that each species could tolerate a
particular velocity range and thus would indicate to the experienced
benthic biologist which ambient flows were available at the site. The
bottom line is that more detailed observations of suspension feeder
behavior must be completed before hypotheses and experimentation can
be initiated to determine the mechanisms involved.
In some ways, question (i) of Table 9.3 is a continuation of (f), al-
though it is cast at the synecological level. Numerous studies have shown
the drastic influence that a violent storm can have on nearshore benthic
communities (see Chap. 7). Even if it were possible to simulate wave
erosion events in the flume, it would be necessary to provide natural
groups of organisms, since it has been established (Chap. 6) that the
erosion threshold may be influenced by the presence of other individuals
of the same, or different, species. Almost certainly there will be energetic
costs associated with maintaining position in flows, whether behavioral,
e.g. in the ability of sand dollars to burrow prior to storms and scallops to
recess into sediments (Chap. 6), or structural, e.g. in skeletal stiffening by
spicule formation in sponges, deposition of calcium salts in corals, or
secretion of adhesives by barnacles (Chap. 6). As far as we are aware, the
comparative energy costs associated with this effort have yet to be deter-
mined. Such costs would be expected to increase with the water move-
ment experienced by populations at each locality examined.
At subcritical erosion thresholds in either unidirectional or oscillating
flows, there is an inhibitory effect on feeding - the c stage of Chapter 4
- due to increasing flows. We believe that behavioral observations, per-

haps with the video camera over a sufficiently wide range of velocities,
will indicate characteristic changes in behavior which are velocity de-
pendent in both passive and active suspension feeders. One important
ecological implication, applicable only to active suspension feeders, is
that if they are filtering at less than optimum carrying capacity as a result
of flow inhibition, they will effectively share out the available seston
along a greater length of the downstream path among a greater number
of individuals (Wildish and Kristmanson 1993). This is an example of an
"emergent property" perceived at a lower hierarchical level and pre-
dicted to occur at the population level. If this prediction is supported by
experimental tests, it may help explain why current ecological modelling
in this respect is so inaccurate.
The final question, (j), in Table 9.3 is clearly cast at the ecosystem
level. We presented in Chapter 8 sufficient evidence to show that the
original ecosystem model of pelagic-benthic coupling which assumed a
direct spatial link between the pelagos and benthos was often contra-
dicted by the observations. We believe that a new predictive model is
required to assist in answering this question. An ecological simulation
model of the type required would combine physical oceanographic vari-
ables with flows of energy or nutrients. Models, or submodels, of this
general type, based on the work described in Chapters 7 and 8, have been
designed for the practical purpose of determining carrying capacity and
optimal seeding density of commercially important bivalves (e.g. Newell
and Shumway 1992; Grant et al. 1992; Heral 1992). A more general
submodel is given by Herman (1992) as a means of determining the role
of suspension feeders in ecosystems, e.g. in pelagic-benthic coupling, in
the context of the framework of holistic estuarine ecosystem models. We
expect that an answer to the question of Table 9.3 will require the
combined efforts of a physical oceanographer and a marine biologist
working at spatial scales determined by the former. The results of such
models are site specific since the degree of transport during pelagic-
benthic coupling will vary; e.g. in the Bay of Fundy it will be higher than
in the nearly tideless Baltic Sea. The variables measured and included
in ecosystem simulation modelling are often complex, e.g. in measur-
ing primary production, and suffer some uncertainty regarding their
accuracy. The reader should also be aware that ecosystem simulation
modelling does not offer the clean kill with Occam's razor between two
simple alternatives common at lower hierarchical levels of science
Table 9.5. Some representative applied benthic biology questions in

which ambient hydrodynamics play a part
Scale
Spatial Temporal
No. (km) (yr) Question
0.001 ~1 What are the environmental variables which
affect the physiology of feeding/growth in
commercially important bivalves?
1.0 ~1 What is the carrying capacity of blue mussels in
suspension culture in a named location?
10-100 1-100 What are the benthic and consequent ecosystem
effects of landfilling and wharf building at the
mouth of an estuary?
1-1000 1-100 What are the near and far field ecological effects
of building a major tidal power dam?
(reductionism). One reason for this is that the model is a means of

deriving hypotheses for testing, but is itself an untested hypothesis (see
J0rgensen 1992b).
Proximate, applied benthic biology questions

If the answers to questions clearly have practical value in, for example,
aquaculture or environmental management (Table 9.5), it is clear that
one is dealing with an applied question. Thus, within Table 9.5, a full
answer to (a) will help a culturist to choose the best locality to site a
bivalve farm and to (b) to set it up to make the maximum use of the
potential available. Ecosystem simulation modelling in relation to (c)
and (d) should ideally precede large-scale civil engineering projects. This
might include wharf or tidal barrage building and the modelling results
used interactively in the debate regarding coastal zone management, but
before construction commences.
As we saw in Chapters 4 and 5, the effects of environmental variables
on bivalve feeding, particularly flow, have not yet been examined in
many species. We suggest that the filtration rate of bivalves needs to be
examined in both siphonate and non-siphonate species with respect to
velocity (including unidirectional and oscillating flows of different perio-
dicity), seston concentration, seston quality, temperature, and salinity.
Commercially important siphonate bivalves include many species of

clams and cockles, while non-siphonate forms include various species of
scallops, oysters, and mussels. Some information about the unimodal
filtration/feeding response of non-siphonate bivalves is presented in
Chapter 4. As far as we are aware, no siphonate bivalves have yet been
tested in this way. Thus, it is premature to think of a universal unimodal
feeding response to velocity applicable to all bivalves.
For siphonate bivalves, the most important part of theflowresponse is
that which is inhibited, (c) in Fig. 4.1. It needs to be de-limited for a much
wider range of species and the mechanism - including the possibility of
velocity-induced valve closing - verified. Research necessary for this
would include real-time observations of siphon opening/closing in con-
trolled flume flows, as well as experiments to understand the environ-
mental factors acting on the control mechanism. The capacity of bivalves
to sense ambientflowpressures, seston concentration, and seston quality
also requires more detailed study.
As indicated in Chapter 4, some bivalves may be stimulated or inhib-
ited to filter feed, suggesting that they respond to chemicals in the envi-
ronment. This work is still in its infancy, yet may provide a mechanism to
explain why filtration is inhibited in the presence of some microbial or
toxic microalgal species.
As a point of practical importance to those thinking of running filtra-
tion/feeding experiments with bivalves, control filtration rates are better
obtained with mixed rather than unialgal cultures. Another unsatisfied
need in this research is a method of measuring seston as an indication of
bivalve food quality.
With regard to (b) in Table 9.5, we mentioned several authors who
have already described ecological methods designed to determine the
carrying capacity of cultured bivalves. The model of Newell and
Shumway (1992) may be described as a physiological approach to a
carrying capacity model of blue mussels, and at a smaller spatial scale
than, for example, the ecosystem simulation model presented by Grant
et al. (1992). The former relies on standard methods of determining
filtration rates of individuals, which certainly do not include velocity as
the controlling variable we believe it to be. Filtration rates are then
linked to scope for growth models in which temperature, sestonic qual-
ity, and bivalve density are input variables. The model predicts produc-
tion for a specified location where turbulent seston flux and uptake by
the mussel bed are simulated. The ecosystem simulation model used by
Grant et al. (1992) requires extensive site-specific effort infieldstudies of
physical and biological variables in order to determine realistic inputs for

the model. This and similar models (e.g. Smaal 1991; Bacher et al. 1991)
cannot be directly extrapolated to other localities.
We feel that this still developing field is close to a stage where simple,
universally applicable benthic carrying capacity predictions, based on
two or three variables, are feasible. Thus, the dependent variable would
be bivalve growth or production of a named species for a particular type
of culture method. The independent variables might beflowrelated (xl)
and food related (x2). Possibilities for xl are yU (Chap. 7), U* (Newell
and Shumway 1992), or a diffusion coefficient based on tidal excursion
length and residence times. Possibilities for xl are mean seston concen-
tration, perhaps modified by seasonal temperature and seston quality
parameters. Further work is required to choose the most efficient predic-
tors, as well as establish the empirical relationship over a wide geo-
graphic area. Simple models like this have the potential to be practically
useful for all bivalve culturists in, for example, choosing between alter-
nate sites to begin a bivalve culture venture. Because the data are rela-
tively simple to obtain and do not depend on the input from continuously
recording electronic sensors, the cost involved would be moderate.
Commonly, a large proportion of the human population of any coun-
try lives and works near the sea, often at the mouth of an estuary. The
seaward part of the interface from the high tide line on land to the edge
of the continental shelf at sea is termed the coastal zone. Humans use the
coastal zone for a great variety of purposes. Currently, these include
commercial fishing; aquaculture; recreational activities inclusive of
whale watching, sportfishing,etc.; source of industrial cooling or process
water; receiving environment for industrial and municipal wastes; min-
eral and hydrocarbon exploration; as well as commercial shipping. It is
not surprising that, because of these multiple uses of the coastal zone,
there are frequent resource-use conflicts.
In general, coastal zone resource conflicts are well documented (e.g.
Wells and Ricketts 1994), although we believe that question (c) in Table
9.5 has received insufficient attention in proportion to the frequency with
which landfilling and wharf building are an integral part of modern
estuarine developments. Many civil engineering projects, over periods
up to 200 yr, accumulate; therefore it is usually impossible to obtain a
before/after picture of the estuarine hydrography, sedimentary, and
benthic biological environments.
During benthic biological studies to determine the effects of dredging
and dumping in the Saint John estuary, New Brunswick, Canada, circum-
stantial evidence suggested that landfilling and wharf building in the

harbour had significant benthic biological effects (Wildish and Thomas
1985). Saint John Harbour serves a small industrialized city of -120,000
people, and the building of a breakwater there resulted in localized
sediment deposition that required maintenance dredging. In sediments
well separated from the dredging and dumping areas, Wildish and
Thomas (1985) noted benthic impoverishment which seemed to be asso-
ciated with sediment-tidal current stresses. Whether these stresses were
indirectly caused by landfilling and wharf building around the harbour as
a result of altered hydrodynamics could not be established during this
study, although it seems a likely possibility.
What is required in a definitive study is a multidisciplinary team -
physical oceanographer, surficial sedimentary geologist, and biologist -
working closely together to document the relevant characteristics of an
estuary. The best field site would be one where the initial work could be
done before any harbour development, involving wharf building, had
been initiated. This would be followed after the harbour had been built
to determine what environmental changes resulted from the civil engi-
neering work. Important in this would be to understand any ecosystem
level changes caused by the development, and not just those associated
with the benthos. These would apply to any possible multiple-use re-
source conflicts within the estuary, e.g. between commercial fishing and
shipping.
The final question, (d), posed in Table 9.5 concerns the ecological
effects of building a major tidal power dam. Because offluctuationsin the
price of oil, many countries have sought alternate sources of power,
inclusive of tidal power generation of electricity. Engineering feasibility
and preliminary environmental studies of the potential effects of such a
development have been made in the 1920s and 1970s in the Bay of Fundy,
Canada. The Bay is actually a large macrotidal estuary with a subtidal
area of -11,149km2. Baker (1984) described the preferred tidal dam
site in Minas Basin in the upper reaches of the Bay. A finite difference
model of the Gulf of Maine/Bay of Fundy was constructed by Greenberg
(1979) to investigate tidal phenomena, both with and without the Minas
Basin tidal barrage in place. Because of resonance effects in the Bay,
it was estimated that tidal differences due to the dam would be felt as
far away as Boston, causing possible loss of terrestrial habitat, greater
storm surges, and changed sedimentary environments. In regard to the
latter, Amos and Greenberg (1980) and Greenberg and Amos (1983)
have predicted potential local, near the tidal barrage, and far-field
sediment effects. Potential secondary production lost by benthic

suspension feeding animals was estimated to be 17% throughout the Bay,
which is a non-significant value (Wildish et al. 1986), largely because of
the very large variances associated with this form of sampling. Further
development of ways to predict changes occurring near the barrage are
required, as well as methods for determining ecosystem level effects.
To date, the price of oil has not yet reached levels which mandate the
use of alternative energy sources, so the predictions presented have not
been tested in a real-world field experiment. Both in Canada and in other
parts of the world, only smaller demonstration tidal power projects have
so far been initiated.
Ultimate questions
The history of life on planet Earth is considered by Lincoln, Boxshall,
and Clark (1982) to have begun some 4500 x 106yr ago. Throughout this
time, tectonic plate movements of all the major land masses and eustatic
changes in seawater levels have repeatedly occurred (Myers 1994). Cli-
matic and ocean current changes separate from or associated with these
events have also occurred (Berggren and Hollister 1977). Thus, the
evidence from geology suggests continuous environmental change
through geological time caused by these universal forcing functions.
Additional discontinuous events such as meteorite impacts (Swinburne
1993) or violent earthquake activity (Officer 1993) may have sudden and
catastrophic effects on animal life on both sea and land. The importance
of violent discontinuous events in relation to extinction events is still
being debated (e.g. Moses 1989; Quinn and Signor 1989; Serjeant 1990).
Because of the adaptational nature of plant and animal life, we should
expect continuous changes - evolution or extinction - of these organisms
to fit, or exclude, them from such changing environments.
All of the theories or hypotheses which can be formulated for the
ultimate questions we have posed in Table 9.6 can only be addressed by
deductive methods. In deduction, the formal scientific propositions must
be made without direct observation of the natural or experimentally
contrived events involved. Because the scientist cannot be present when
the evolutionary events associated with the questions in Table 9.6 occur,
the only evidence for, or against, an hypothesis is circumstantial. A
frequent problem with circumstantial evidence is knowing how much is
required to prove that a particular hypothesis is correct or to choose
between competing hypotheses when the balance of evidence is nearly
Table 9.6. Some ultimate questions related to hydrodynamics and

referable to suspension feeders or the communities in which they live.*
Discipline No. Question

Physiology 1 Why did active suspension feeders evolve from passive
forms (or vice versa)?
2 Why did a ciliary pump evolve?
Ethology 3 Can the fact that some sessile suspension feeders have
short, and others long, periods of planktonic life be
explained hydrodynamically?
4 Why do hydrodynamic factors cause epifaunal suspension
feeders to aggregate?
5 What adaptive imperatives resulted in swimming bivalves?
Ecology 6 How did larvae with specific responses to environmental
variables, e.g. light, gravity, and flow, evolve?
7 What causes a mixed benthic assemblage dominated by
suspension feeders to evolve?
8 Why do ecosystems evolve over geological time?
a
From Table 1.2.
equal between them. Thus, deductive methods lack the power of deci-
sion available to those answering proximate questions using inductive
methods.
As background to our consideration of ultimate questions (Table 9.6),
we discuss three common types of circumstantial evidence which might
be available. The first involves the geological record and could include
direct preservation of parts of the organism, which gives clues to the
species evolutionary history. Alternatively, the geological record could
provide inferences about past climates and environment from the types
of organisms preserved in ancient sediments, e.g. the palaeoecology of
marine sediments (Molfino 1994). Surficial geologists study modern
sediments and the benthic flora and fauna they contain to improve their
view of the factors which may have operated in the geological past (e.g.
Rhoads and Young 1970; Thorns and Berg 1985). Surficial geologists also
have independent means of measuring the age of sediments by isotopic
decay rates (Molfino 1994). A second general approach is through
phylogeny. Modern taxonomy requires determining a wide range of,
largely, morphological characters in each species, and this has led to the
construction of phylogenetic trees which are thought to indicate the

degree of relatedness of each taxon in a common group. Increasingly
more important and independent views of genetic relatedness are pro-
vided by modern biochemical methods, notably deoxyribonucleic acid
(DNA) sequencing of mitochondria or other genetic material (e.g. Field
et al. 1988; Powers, Allendorf, and Chen 1990). The third general ap-
proach which could provide circumstantial evidence for ultimate ques-
tions is biogeographic study of patterns of the distribution of animals and
plants. This method allows the palaeoecologist to see how spatial mecha-
nisms in modern environments affect speciation, extinction, and patterns
of species richness (Myers 1994), thus suggesting how it could have
happened in the past.
To return now to the ultimate questions of Table 9.6: For the first two
physiological questions, we propose that active suspension feeders
evolved from passive ones to enable the former to colonize a wider range
of flow environments than those permissible to passive suspension feed-
ers. Thus, for the passive suspension feeding gorgonian coral S. suberosa
the range was 7-9 cm -s"1; the unimodal peak relating velocity and
filtration was sharper, and proportionately more of the (a) part of the
response was present (Chap. 4, the section, Background). For scallops as
an example of an active suspension feeder, the range is 2 to >45 cm s"1,
the unimodal peak being broader with less (a) and more (c) in the
filtration response (Chap. 4, the section, Background). One test of this
hypothesis would be to compare the unimodal responses of a wide range
of passive versus active suspension feeders over a sufficiently wide range
of ambient velocities. We do not mean to imply that the possession of a
ciliary pump insulates the active suspension feeder from a need to be in
the flow. This is because active suspension feeders commonly form beds
of near-monoculture composition which require energetic turbulent
fluxes to supply the necessarily large amounts of sestonic food that such
populations need to avoid seston depletion effects.
A further aspect of the possession of a ciliary pump is the concomitant
energetic costs involved. The following are two aspects of such costs:
The ontogenetic development of a trophicfluidtransport system with an in-
line ciliary pump
The physiological cost of operating a ciliary pump relative to the other
physiological activities of a suspension feeder
As far as we are aware, the first of these has not yet been considered.
With regard to the latter, J0rgensen, Larsen, and Riisgard (1986) have
reviewed the total volumetric flow through active filter feeders in rela-
tion to concomitant oxygen consumption. Two methods of determining

the relative energy costs of active suspension feeding were considered by
J0rgensen et al. (1986). The first involved comparing the product of the
operating pressure of the pump and volume of water filtered, from which
an estimate of the work done could be obtained. This estimate in relation
to the total of all energy expended by various active suspension feeders
amounted to <1%. A second method involved estimating the energy
expenditure of the ciliary pump. In blue mussels, the ciliary beating work
done at the gills amounted to 0.7% of the total energy budget. Thus, at
a normal operating efficiency of the mussel pump, the energetic cost
would be a small percentage of the total metabolic cost (J0rgensen et al.
1986).
The preceding considerations are consistent with the view that the low
pressure ciliary pump in bivalves operates at one rate and may be either
on (valves open) or off (valves closed). Differences in volumetric flow
through the bivalve trophic fluid transport system are controlled by
compression or collapse of the system by valve closure, as indicated by
the relative size of the exhalant opening. Environmental control of the
degree of compression has been shown to include ambient velocity in
scallops (Wildish and Saulnier 1993). The large volumetric flow through
the trophic fluid system of blue mussels is primarily adapted to the
filtration function (the need for large throughputs of ambient seawater
from which the gills can obtain sufficient sestonic food to grow). Famme
and Kofoed (1980) showed by perfusion experiments through the trophic
fluid transport system of mussels that there was no change in oxygen
uptake after gill removal and that this process did not occur there but
across the general mantle cavity walls. J0rgensen et al. (1986) point out
that the diffusive boundary layer in the mantle cavity is -0.1mm thick
and, because laminar flow is present, only a small fraction of pumped
seawater is available for oxygen exchange. This explains the finding
(J0rgensen et al. 1986) that oxygen consumption in mussels is an asymp-
totic function of the volumetric flow rates within the trophic fluid trans-
port system. Maximum oxygen consumption rates are reached at rela-
tively low volumetric flow rates through the mussel, at flows equal to
approximately one-fifth of the maximum pump capacity. The limits here
are controlled by the diffusion rates across the mantle cavity wall, which
vary with the thickness of the boundary layer and, therefore, flow rate.
Another view considered by J0rgensen et al. (1986) was that the
curvilinear increase of oxygen uptake as a function of pumping rate was
a direct indication of the energetic cost of ciliary pumping (e.g. Verduin
1969; Bayne, Thompson, and Widdows 1976). This interpretation implies

that ciliary pumping is a major energetic cost and is at odds with the low
energetic costs estimated by J0rgensen et al. (1986), as mentioned. Di-
rect energy measurements of ciliary beating in isolated mussel gills by
Clemmesen and J0rgensen (1987) also suggest that costs are low. The
explanation for the experimental results obtained by Bayne et al. (1976)
could be that the full volumetric capacity of the mussel pump remained
untested because the experimental conditions offered resulted in inhib-
ited pumping. Thus, the results were all obtained in the lower range of
volumetric pump output, where the thickness of the diffusive boundary
layer and mantle cavity flow velocities are the determinants of oxygen
uptake.
Bougrier et al. (1995) examined the allometric relationships among
temperature, clearance, and oxygen uptake rates for non-reproductive
oysters Crassostrea gigas. These researchers showed that the latter was a
unimodal function of temperature, with a peak of 19C that clearly was
not related to oxygen consumption over the same temperature range.
Clearly, further work is required in this area to determine for a wide
range of active suspension feeders what the relative energy costs are,
inclusive of ciliary pumping. If the concept of J0rgensen et al. (1986) is
upheld, then we can see the low pressure ciliary pump as an energetically
efficient organ adapted to suspension feeding in a dilute suspension of
seston in seawater. Active suspension feeders should occur in a much
wider range of ecological niches than passive forms if the theory of
expansion to a wider range of niches is correct.
Turning now to question 3 (Table 9.6), we know that the observations
incorporated in it are consistent with the types of larvae associated with
benthic suspension feeders (Chap. 3). Previous authors have considered
what has caused the larval stages of ontogeny to evolve in such different
directions, one to lecithotrophy, the other to planktotrophy (Table 9.7).
Vance (1973) proposed models in which natural selection chose among
individual parents, specifically those who maximized successful settle-
ment per unit of reproductive effort. Chia (1974) proposed that
lecithotrophy developed from planktotrophy as a means of minimizing
reproductive energy expenditure. Thus, the few larger eggs of
lecithotrophy allowed a lower total reproductive expenditure because far
fewer eggs were required to achieve the same settlement rate. Strathman
(1974) proposed a concept in which spatially and temporally varying
environments were seen as the driving force of selection and, hence,
evolution of divergent strategies to lecithotrophy or planktotrophy. By
Table 9.7. Comparison of two larval strategies in benthic marine

suspension feeders.
Characteristic Lecithotrophic larva Planktotrophic larva

Larval type Larger, non-feeding, Smaller, feeding in
food store plankton, no food store
Larval numbers Few Many
Larval period Short Long
Larval dispersal distance Short Long
(maximum)
Inbreeding potential High Low
Biogeographic dispersal Slow Fast
rate
Biogeographic range Endemic Pandemic (cosmopolitan)
Geological history of Short Long
taxa
Speciation rate High Low
After Vance (1973); Strathman (1974); Chia (1974); Scheltma (1986).
this proposal, individual suspension feeders that spread sibling larvae

evenly over a wide area, inclusive of temporally favorable and un-
favorable niches, had higher fitness over many generations. By contrast,
lecithotrophic forms produced recruits that varied in settlement only
according to the temporal changes that occurred at the parental niche. A
review of current ideas relating to the evolutionary ecology of larval types
has been presented by Havenhand (1995).
We suggest that the environmental grain postulated in theoretical
studies of Strathman (1974) could be an hydrodynamic one. Then, spe-
cies with lecithotrophic larvae are betting that the best hydrodynamic
environment for benthic life is provided close to the parental population.
Species with planktotrophic larvae, on the other hand, are betting that
there is a better flow environment at some distance from the parental
population. Scheltema (1986) has summarized the evidence in the geo-
logical record pertaining to the larval strategies of Table 9.7, including
biogeographic range, occurrence of extinctions, and speciation rate as
indicated by the geological record. Nor is it impossible that rapid larval
evolution could occur and be observed directly during the working life of
a scientist. As an example, the barnacle Balanus amphitrite was intro-
duced to a newly created inland sea in California, where many environ-
mental variables differed from those of the donor coastal populations

(Raimondi 1992). The latter author was able to observe the larval
phenotypic differences between the inland sea and coastal populations
and establish, by field and lab, transplant experiments that the
phenotypic divergence was probably due to natural selection.
Fewer lecithotrophic larvae are produced by benthic suspension feed-
ers. They have a shorter larval period and local settlement with dispersal
limited to a small ambit near the parental population (Table 9.7). The
deleterious consequences frequently associated with inbreeding do not
seem to have occurred in some taxa possessing lecithotrophic larvae (e.g.
Grosberg 1987). In colonial ascidians, although fertilization barriers
which prevent members of the same colony from mating are present,
breeding experiments undertaken by Grosberg (1987) suggest that
outbreeding is very costly in terms of fitness and that the degree of
inbreeding within a local population is high. Grosberg and Quinn (1986)
have shown that in at least one colonial ascidian the mechanism of
maintaining closely related individuals in a population involves kin rec-
ognition at the larval stage (see also Chap. 6, the section, Epifauna and
aggregation). In some species of Bryozoa, Keough (1984) showed that
kin recognition by larvae enhanced the formation of colonial
aggregations consisting of closely related individuals (see also Chap. 3,
the section, Hard substrates).
Suspension feeders frequently do form aggregated populations of high
density (Chap. 6). The hydrodynamic aggregation theory is that because
of localized flow conditions, there are some niches that provide good,
and some bad, opportunities for the later benthic stages of life. In some
cases, the larval settlement cues are also hydrodynamic ones (see Chap.
3, the section, Hard substrates). Thus, the presence of a living suspension
feeder - either larva or adult - on a particular substrate can indicate to a
potentially settling larva that it is a suitable niche which could support
growth. However, there may be other adaptive advantages associated
with aggregation in suspension feeders, besides flow-related ones (see
Table 6.8, the section, Epifauna and aggregation). Planktotrophic larval
dispersal has also been suggested to be a mechanism carrying the larvae
away from the crowding associated with aggregated distributions
(Strathman 1974).
The bivalves capable of swimming were discussed in Chapter 6 (the
section, Swimming). Among the best swimmers are the scallops, of which
-7000 species of both extant and extinct forms of Pectinidae are known
(Waller 1991). The morphological criteria employed in taxonomy have
also been used to construct a proposed phylogenetic tree to indicate a

geological history beginning in the Oligocene. The available evidence in
the fossil record reviewed by Waller (1991) gives little assistance in
determining which environmental conditions led to the establishment of
swimming in the scallops. Possible factors are predation, active migra-
tion, and post-settlement dispersal associated with either improved
trophic conditions or more effective reproduction. Some modern scal-
lops do elicit escape swimming responses when tissue extracts of poten-
tial predators are introduced (e.g. Stanley 1970). Morton (1980), how-
ever, considers that the fast and efficient swimming of Amusium
pleuronectes is adapted for seasonal migration and reproduction. It is
tempting to speculate that swimming in this scallop has evolved from a
predator escape mechanism, but any evidence to confirm this suggestion
is absent. Some studies of the shell form of extant bivalves have proved
to be helpful to geologists in interpreting how fossil bivalves lived, and by
inference, the type of environment in which they were found (Stanley
1970).
The evolution of photo-, geo-, and rheotaxes in larvae (number 6,
Table 9.6) can best be understood in an ecological context. The use of
light or gravity as environmental cues to direct the swimming of
planktotrophic larvae could have occurred anywhere on the continental
shelf. Planktotrophic larvae originating from sessile benthic suspension
feeders all need to swim up to the surface layers of seawater, where
microalgal food is available for them to grow. After a period of plank-
tonic development (Chap. 3), the competent larvae need to return again
to the bottom, where they find a suitable substrate and complete the
settlement process. The benthic stage is the major growth part of the life
cycle, and post-settlement success depends critically on wise larval
choices made during settlement, particularly for those species lacking a
post-settlement dispersal mechanism. The simple behavioral responses,
positively phototactic or negatively geotactic, reverse towards the end of
larval life and are thus "pre-adapted" to estuarine retention. This mecha-
nism allows competent larvae to be returned to the parental habitat. It is
difficult to see how any hypotheses generated from the preceding can be
tested by the deductive method. Perhaps number 6 in Table 9.6 qualifies
as an unanswerable question - at least by today's standards.
The last two questions of Table 9.6 are both ecological ones and will be
dealt with together. For the background information applicable to ques-
tion 8, a description of the universal forces which result in environmental
changes throughout geological time is required (see also the section,
Ultimate questions). As an example of how these forces might operate,

we examine the post-glacial formation of the Bay of Fundy. Reviewing
the results of other geologists, Fader, King, and MacClean (1977) de-
scribed the likely history of the Bay of Fundy after the Laurentide Ice
Sheet retreated northwards some 13.5 x 103yr B.P. By 8 X 103yr B.P., the
Bay was a shallow, nearly tideless estuarine bay where emergent
Georges and Browns banks protected the Gulf of Maine and Bay of
Fundy from the energetic tidalflowsof the northwestern Atlantic. Begin-
ning at 6 x 103yr B.P., and as a result of eustatic sea level rise that ex-
ceeded land mass rebound, increases in depth and tidal flows began
which have steadily increased to the present. Present-day depths at the
mouth of the Bay are >200m, and maximum tidal current velocities
exceed 400 cm-s"1 in the upper Bay. The increased tidalflowswithin the
Bay have resulted in changes of the surficial sediments. Initially, a uni-
form glacial till was present throughout the shallow Bay. In present
conditions, three major sedimentary types have formed: tidally reworked
sands in the upper Bay, original glacial till in the middle part, and
redeposited silt/clay in the deeper parts of the western side of the mouth
of the Bay (Fader et al. 1977). Our example suggests the relative speed
that the geological imperatives can have in controlling environmental
variables such as flow. In this case, the Bay of Fundy went from nearly
tideless to some of the most energetic tides of the world in -6000 years.
We can expect that there will be changes involving increases, and de-
creases, in tidal flows over relatively short geological periods, and that
these changes must act as powerful selective forces, for both resident and
immigrant suspension feeders.
It should be realized that geologists, like biologists interested in ulti-
mate questions, are forced to use deductive methods. An example of the
indecisiveness which such methods can cause is provided in the debate
(see the section, Ultimate questions) regarding the relative importance
of discontinuous events in causing sudden mass faunal extinctions.
Concerning the question about the causes of deposit versus mixed
benthic assemblages (number 7, Table 9.6), the types of benthic commu-
nity association were described in Chapter 7. If we continue our use of
the Bay of Fundy as an example, we can see that there is insufficient time
- approximately 6000 generations if the benthic organism is an annual
species - for evolutionary processes to produce a suspension feeder from
a deposit feeder from scratch. A further point is that the benthic fauna
will arrive as immigrants from adjacent seas by one or more of the four
methods of dispersal discussed in Chapter 3. The immigrants will include
both deposit and suspension feeders, and the environmental conditions

that they experience will determine which survive and continue to
evolve, as well as those which cannot survive in these conditions. A
possible scenario for the Bay of Fundy would be that in the period 13.5
to 6xl0 3 yr B.P., the major selection would be for deposit feeders,
whereas after 6 x 103yr B.P., the progressively increasing tidal range and
flows would favor the establishment of suspension feeders. The present-
day benthic fauna of the Bay of Fundy (Wildish and Peer 1983; Wildish,
Peer, and Greenberg 1986) consists mainly of mixed benthic assem-
blages, where even on net depositional, silt/clay sediments, tube-living
suspension feeders (e.g. the amphipod Haploops fundiensis and
polychaete Potamilla neglecta) are present. Only in the most reworked
sand sediments of the upper Bay are suspension feeders absent. This
sediment is occupied by a specialized - and impoverished - benthic
macrofauna, consisting of deposit feeders either able to withstand very
high tidal velocities by burrowing deeply below the erosion layer or built
massively to avoid being swept away. Because of the brief time available
and the known zoogeographic distribution of the named deposit feeders
in the section, Water-movement impoverishment, Chapter 7, it is un-
likely that they evolved in situ in the Bay of Fundy.
It is clear that approaches to testing some of our hypotheses regarding
ultimate questions should be made in collaboration with physical ocean-
ographers, surficial geologists, and palaeoecologists. One obvious ques-
tion for a palaeoecologist is whether the fossil macrofauna present within
the Bay of Fundy during the last 13.5 x 103yr conforms to the predictions
made.
References
Amos, C. L., and D. A. Greenberg. 1980. The simulation of suspended
participate matter in the Minas Basin, Bay of Fundy: a region of potential
tidal power development, p. 2-20. In Proceedings Canadian Coastal
Conference, Nat. Res. Counc, Ottawa.
Bacher, C, M. Heral, J. M. Deslous-Paoli, and D. Razet. 1991. Modele
energetique unibiote de la croissance des huitres (Crassostrea gigas) dans
le bassin de Marennes-Oleron. Can. J. Fish. Aquat. Sci. 48: 391-404.
Baker, G. C. 1984. Engineering description and physical impacts of the most
probable tidal power prospect(s) under consideration for the upper
reaches of the Bay of Fundy, p. 333-345. In D. C. Gordon and M. J.
Dads well (eds.) Update on the marine environmental consequences of
tidal power development in the upper reaches of the Bay of Fundy.
Can. Tech. Rep. Fish. Aquat. Sci. 1256.
Bayne, B. L., R. J. Thompson, and J. Widdows. 1976. Physiology I, p. 121-206.
In B. L. Bayne (ed.) Marine mussels, their ecology and physiology.

Cambridge University Press, London.
Berggren, W. A., and C. D. Hollister. 1977. Plate tectonics and
paleocirculation: commotion in the ocean. Tectonophysics 38: 11-48.
Bougrier, S., P. Geariron, J. M. Deslous-Paoli, C. Bacher, and G. Jonquieres.
1995. Allometric relationships and effects of temperature on clearance
and oxygen consumption rates of Crassostrea gigas (Thunberg).
Aquaculture 134: 143-154.
Butman, C. A. 1987. Larval settlement of soft sediment invertebrates: the
spatial scales of pattern explained by active habitat selection and the
emerging role of hydrodynamical processes. Oceanogr. Mar. Biol. Annu.
Rev. 25: 113-165.
Chia, F. S. 1974. Classification and adaptive significance of developmental
patterns in marine invertebrates. Thai. Jugosl. 10: 121-130.
Clemmesen, B., and C. B. J0rgensen. 1987. Energetic costs and efficiencies of
ciliary filter feeding. Mar. Biol. 94: 445-449.
Emerson, C. W., and J. Grant. 1991. The control of soft-shell clam (Mya
arenaria) recruitment on intertidal sandflats by bedload sediment
Fader, G. B., L. H. King, and B. MacLean. 1977. Surficial geology of the
eastern Gulf of Maine and Bay of Fundy. Mar. Sci. Pap. 19, Geological
Survey of Canada Paper 76-17.
Famme, P., and L. H. Kofoed. 1980. The ventilatory current and ctenidial
function related to oxygen uptake in declining oxygen tension by the
mussel Mytilus edulis L. Comp. Biochem. Physiol. 66A: 161-171.
Field, K. G., G. J. Olsen, D. J. Lane, S. J. Giovannoni, M. T. Ghiselin, E. C.
Faff, N. R. Pace, and R. A. Raff. 1988. Molecular phylogeny of the animal
kingdom. Science 239: 748-753.
Fry, F. E. J. 1947. Effects of the environment on animal activity. Univ.
Toronto Stud. Biol. 55. Publ. Ont. Fish. Res. Lab. 68: 1-62.
Grant, J., M. Dowd, K. Thompson, C. Emerson, and A. Hatcher. 1992.
Perspectives on field studies and related biological models of bivalve
growth and carrying capacity. NATO ASI Ser. Vol. G33: 371-420.
Greenberg, D. A. 1979. A numerical model investigation of tidal phenomena
in the Bay of Fundy and Gulf of Maine. Mar. Geodesy 2: 161-187.
Greenberg, D. A., and C. L. Amos. 1983. Suspended sediment transport and
deposition modelling in the Bay of Fundy, Nova Scotia: a region of
potential tidal power development. Can. J. Fish. Aquat. Sci. 40: 20-34.
Grosberg, R. K. 1987. Limited dispersal and proximity-dependent mating
success in the colonial ascidian Botryllus schlosseri. Evolution 41: 372-384.
Grosberg, R. K., and T. F. Quinn. 1986. The genetic control and consequences
of kin recognition by the larvae of a colonial marine invertebrate. Nature
322: 456^59.
Havenhand, J. N. 1995. Evolutionary ecology of larval types, p. 79-122. In
L. McEdward (ed.) Ecology of marine invertebrate larvae. CRC Press,
Boca Raton, Fla.
Heral, M. 1992. Why carrying capacity models are useful tools for
management of bivalve mollusc culture. NATO ASI Ser. Vol. G33: 455-
478.
Herman, P. M. J. 1992. A set of models to investigate the role of benthic
suspension feeders in estuarine ecosystems. NATO ASI Ser. Vol. G33:
421-454.
J0rgensen, C B. 1981. A hydrochemical principle for particle retention in

Mytilus edulis and other ciliary suspension feeders. Mar. Biol. 61: 277-282.
1992a. Adaptational, environmental and ecological physiology: a case for
hierarchical thinking, p. 9-17. In S. C. Wood, R. E. Weber, R. R.
Hargens, and R. W. Millard (eds.) Physiological adaptions in vertebrates.
Marcel Dekker, New York.
J0rgensen, C. B., P. S. Larsen, and H. V. Riisgard. 1986. Nature of relation
between ventilation and oxygen consumption in filter feeders. Mar. Ecol.
Prog. Ser. 29: 73-88.
J0rgensen, S. E. 1992b. Integration of ecosystem theories: a pattern. Kluwer
Academic, Dordrecht.
Keough, M. J. 1984. Kin-recognition and the spatial distribution of larvae of
the bryozoan Bugula neritana (L.). Evolution 38: 142-147.
Kerr, S. R. 1990. The Fry paradigm: its significance for contemporary ecology.
Trans. Am. Fish. Soc. 119: 779-785.
Lincoln, R. J., G. A. Boxshall, and P. F. Clark. 1982. A dictionary of ecology,
evolution and systematics. Cambridge University Press, Cambridge.
Molfino, B. 1994. Palaeoecology of marine systems, p. 517-546. In P. S. Giller,
A. C. Hildrew, and D. G. Raffaelli (eds.) Aquatic ecology. Blackwell,
Oxford.
Morton, B. 1980. Swimming in Amusium pleuronectes (Bivalvia: Pectinidae).
J. Zool. London 190: 375-404.
Moses, C. O. 1989. A geochemical perspective on the causes and periodicity of
mass extinctions. Ecology 70: 812-823.
Myers, A. A. 1994. Biogeographic patterns in shallow-water marine systems
and the controlling processes at different scales, p. 547-574. In P. S.
Giller, A. C. Hildrew, and D. G. Raffaelli (eds.) Aquatic ecology.
Blackwell, Oxford.
Newell, C. R., and S. E. Shumway. 1992. Grazing of natural particulates by
bivalve molluscs: a spatial and temporal perspective. NATO ASI Ser. Vol.
G33: 85-148.
Officer, C. 1993. Victims of volcanoes. New Sci. 1861: 34-38.
Olafsson, E. B., C. H. Peterson, and W. G. Ambrose. 1994. Does recruitment
limitation structure populations and communities of macro-invertebrates
in marine soft sediments: the relative significance of pre- and post-
settlement processes. Oceanogr. Mar. Biol. Annu. Rev. 32: 65-109.
Pawlik, J. R. 1992. Chemical ecology of the settlement of benthic marine
invertebrates. Oceanogr. Mar. Biol. Annu. Rev. 30: 273-335.
Peters, R. H. 1991. A critique for ecology. Cambridge University Press,
Cambridge.
Powers, D. A., F. W. Allendorf, and T. Chen. 1990. Application of molecular
techniques to the study of marine recruitment problems, p. 104-121. In
K. Sherman, L. M. Alexander, and B. D. Gold (eds.) Large marine
ecosystems: patterns, processes and yields. Am. Assoc. Adv. Sci.,
Washington, D.C.
Quinn, J., and P. W. Signor. 1989. Death stars, ecology and mass extinction.
Ecology 70: 824-834.
Raimondi, P. T. 1992. Adult plasticity and rapid larval evolution in a recently
isolated barnacle population. Biol. Bull. 182: 210-220.
Rhoads, D. C, and D. K. Young. 1970. The influence of deposit-feeding
organisms on sediment stability and community trophic structure. J. Mar.

Res. 28: 150-178.
Scheltma, R. S. 1986. On dispersal and planktonic larvae of benthic
invertebrates: an eclectic overview and summary of problems. Bull. Mar.
Sci. 39: 290-322.
Serjeant, W. A. S. 1990. Astrogeological events and mass extinctions: global
crises or scientific chimaerae? Mod. Geol. 15: 101-112.
Smaal, A. C. 1991. The ecology and cultivation of mussels: new advances.
Aquaculture 94: 245-261.
Stanley, S. M. 1970. Relation of shell form to life habits of the Bivalvia
(Mollusca). Geol. Soc. Am. Inc. Memoir 125.
Strathman, R. 1974. The spread of sibling larvae of sedentary marine
invertebrates. Am. Nat. 108: 29-44.
Swinburne, N. 1993. It came from outer space. New Sci. 1861: 28-32.
Thorns, R. E., and T. M. Berg. 1985. Interpretation of bivalve trace fossils in
fluvial beds of the basal Catskill formation (late Devonian), eastern
U.S.A., p. 13-20. In H. H. Curvan (ed.) Biogenic structures: their uses in
interpreting depositional environments. Soc. Econ. Paleontol. Mineral
Spec. Publ. 35.
Vance, R. R. 1973. On reproductive strategies in marine benthic invertebrates.
Am. Nat. 107: 353-361.
Verduin, J. 1969. Hard clam pumping rates: energy requirements. Science 166:
1309-1310.
Waller, T. R. 1991. Evolutionary relationships among commercial scallops
(Mollusca: Bivalvia: Pectinidae), p. 1-72. In S. E. Shumway (ed.) Scallops:
biology, ecology and aquaculture. Elsevier, Amsterdam.
Wells, P. G., and P. J. Ricketts (Editors). 1994. Coastal Zone Canada '94.
Cooperation in the coastal zone: conference proceedings, Halifax, N.S.
Wildish, D. J., and D. D. Kristmanson. 1993. Hydrodynamic control of bivalve
filter feeders: a conceptual view. NATO ASI Ser. Vol. G33: 297-324.
40: 309-321.
production in the Bay of Fundy and possible effects of a tidal power
2410-2417.
Wildish, D. J., and A. M. Saulnier. 1993. Hydrodynamic control of nitration in
Wildish, D. J., and M. L. H. Thomas. 1985. The effects of dredging and
dumping on benthos of Saint John Harbour, Canada. Mar. Environ. Res.
15: 45-57.
Glossary
adhesion number the dimensionless ratio of London attraction and

hydrodynamic forces acting on a sestonic particle approaching a
cylinder in a flowing fluid.
advection the transport of fluid molecules, solutes, or seston by the
bulk motion of the fluid, e.g. by horizontal tidal currents.
aerosol a suspension of solid, or liquid, particles in a gas.
aerosol theory the application of the theory developed for the capture
of aerosol particles to sestonic particle capture in water by suspen-
sion feeders.
aggregation statistics that part of probability theory concerned with
contagious or aggregated distributions in space.
ambient velocity the background flow speed and direction.
anemometer a device for measuring flow speed and sometimes direc-
tion in air or water.
asperities surface roughness on a microscopic particle, e.g. seston.
ATP adenosine triphosphate, an energy containing coenzyme present
in every living cell.
back pressure that pressure operating in a direction opposite to that
being considered, e.g. the difference in external pressure between
the outlet and inlet of a hydromechanical pump.
Baconian experimental test a test involving an alternative and a null
hypothesis posed in such a way that the experimental result supports
or rejects the original proposition.
ball valve a mechanism designed to controlfluidflowbased on a buoy-
ant ball which acts to open and close an opening by its own weight
and fluid pressure.
BBL benthic boundary layer, a discrete layer of flow above a natural
benthic substrate. It consists of a viscous sublayer (q.v.) nearest the
391
392 Glossary
substrate, followed by a log-linear layer and an outer layer, each

with a characteristic velocity profile.
bed load sedimentary particles which are being transported on, or im-
mediately above, the bed.
bed load shear velocity that velocity at which sedimentary particles first
begin to be transported as bed load by the flow. Velocity at which
critical shear stress is exceeded and resuspension is initiated.
Bernoulli's principle that the sum of the pressure and kinetic energy of
a fluid parcel moving along a streamline remains constant. Fluid
friction is ignored in this consideration.
bidirectional flow aflowwhich tends in two directions, as in tidal flows,
where the flood and ebb reverse directions.
bilateral the symmetry of an animal in which only one median or
sagittal section can yield right and left mirror image halves, e.g. as in
a sagittal section of mammals.
bluff body an object which presents a large frontal area to the flow.
BOD biological oxygen demand, a measure of the amount of organic
matter in a sample which can be oxidised by aerobic respiration in a
defined period.
Boltzmann's constant a thermodynamic expression which is equal to
the universal gas constant divided by Avogadro's number.
bottom roughness those projections from the seabed which are suffi-
cient to influence the BBL flow. Usually of sedimentary (e.g. sand
ripples) or biological (e.g. epifaunal tubes) origin.
boundary shear the force exerted by the boundary layer at a solid or
fluid surface, e.g. by fluid flow at the sediment-water interface.
bulk velocity the mean flow speed in the mainstream direction defined
as the volumetric flow rate divided by the cross sectional area occu-
pied by the flow.
buoyant weight a measure of the weight of an object in water, which is
less than in air because of the buoyant force exerted by the fluid
vertically on the body immersed, or floating, in it.
camber the slightly convex shape, e.g. on the upper surface of a stream-
lined body which increases flow there and thus enhances lift.
centrifugal pump a device to pump water by accelerating it outward by
impeller to a surrounding casing.
COD chemical oxygen demand, a chemical measure of the total
amount of oxygen required to oxidise organic matter or reducing
compounds present in a sample.
Glossary 393
collecting elements the specific surfaces of a suspension feeder on

which sestonic food particles are first captured.
collimator a device used in flumes to make steamlines parallel; a flow
straightener, e.g. a system of parallel plates placed upstream of the
working section.
concentration boundary layer that region of any boundary layer where
the concentration of a solute after consumption or discharge at the
wall is less, or greater, than that in the bulk flow because of incom-
plete mixing within the boundary layer.
continental shelf break the underwater topographic point where the
continental shelf ends and the continental slope begins as denoted
by an increase in gradient (>1 in 20).
convective diffusion the transport of heat or mass by flows resulting
from density differences within the fluid due to concentration or
temperature differences.
counterion an ion with a charge that is opposite to that of another ion
that it is brought into contact with.
creep the slow slipping or bending of solids due to critical hydrody-
namic stresses, e.g. skeletal tissue of a coral exposed to a strong flow.
creeping flow flow at very low Reynolds numbers, where inertial forces
are negligible with respect to viscous forces.
critical drag that force generated by a flowing fluid which is just suffi-
cient to overcome the static friction acting on an object lying on the
bottom, which is required to cause slippage or detachment of it.
current meter a field device for measuring ambient flow speed and
sometimes direction.
deposition the process by which organic and inorganic particles sus-
pended in the water column settle onto the sediment.
depth limited boundary layer a benthic boundary layer at a shallow
location in which the boundary layer intersects the free water sur-
face, without the presence of a free-stream flow.
detachment force critical drag (q.v.).
dewater a process in which water is removed from a solid by physical
process, e.g. filtration of a sediment slurry.
diffusive sublayer a region of the laminar or viscous sublayer (q.v.)
where the transport of a solute is dominated by molecular diffusion
and turbulent diffusion is unimportant.
dimensionless index a ratio of two mathematical entities used in calcu-
lating a single number, without defining units, and which predicts
394 Glossary
the importance of specified properties of the system, e.g. SDI

(q.v.).
DNA deoxyribonucleic acid.
DOM dissolved organic matter.
downstream vortex an eddy whirlpool water movement pattern down-
stream of an object in a turbulent flow.
down-welling in physical oceanography, the deflection of seawater
downwards towards the sediment.
drag the rate of removal of momentum from a flowing fluid by an
immersed object, where this rate depends on the frontal area of the
object, the velocity past it (assuming an attached object), and the
density and viscosity of the fluid.
drift bottle a buoyant container used by physical oceanographers to
track surface or subsurface currents visually.
dynamic pressure a pressure in a moving fluid which is equivalent to
the kinetic energy per unit volume at a point defined by Bernoulli's
principle, stagnation pressure (q.v.).
dynamic similarity a term used in hydrodynamic modelling where the
prototype and model are geometrically similar. Requires that ratio
of forces around prototype and model be the same at equivalent
points.
ectocrine a biochemical substance produced externally by a marine
organism and which causes a biological response in another of the
same or different species, e.g. toxic microalgal metabolites which
inhibit bivalve initial feeding, or chemical cues that induce larval
settlement.
eddy a turbulent water (or air) current running in a different direction
than that of the main current; or flow in the wake of a bluff body
different from the mainstream flow lines.
eddy diffusivity the rapid spreading in turbulent flow of heat, mass, or
momentum by random movements of the fluid.
electronic particle counter a device for counting and sizing suspended
particles down to a micrometer or so in diameter, e.g. seston.
electrostatic force in aerosol theory, the force generated by differences
in electrical charges of particle and collecting surfaces.
endoscope a device for microscopic observations in small spaces be-
tween animal tissues. Consists of a thin light tube, laser light, and
magnifying optics.
engineering efficiency the success of a collecting surface in a flow in
capturing particles; the ratio of the rate at which particles are col-
Glossary 395
lected to the rate at which they would pass through the space occu-
pied by the collector if it were not there.
entrainment the physical process of transfer of fluid by friction from
one water mass to another. Usually occurs between currents moving
in different directions.
eustatic the slow, universal changes of sea level which are caused by the
melting of continental ice sheets.
fall velocity the rate at which a body passively falls under the influence
of gravity through a still fluid; same as settling velocity.
Fick's laws of molecular diffusion the first law states that the flux of
molecules is equal to the molecular diffusion coefficient times the
concentration gradient. The second law relates the rate of change of
concentration at a specified time and position to the diffusion coef-
ficient and the second derivative of concentration with respect to
distance.
field flume a flume capable of being used in the field, either in situ, as
by surficial geologists, or by benthic biologists where the flume is
portable but not necessarily used in situ.
first law of thermodynamics the conservation of energy law: energy
may be transformed from one form to another, but cannot be cre-
ated or destroyed.
flow cytometer a device which uses optical analysis based on light
scattering and fluorescence for measuring the size and epi-
fluorescence properties of small particles, e.g. red blood cells or
seston.
flow fluorometer a device which measures the fluorescent radiation of
a sample after exposure to monochromatic radiation. The sample is
continuously brought to the sensor in a pumped flow of seawater,
e.g. continuous seawater sampling of chlorophyll a in living phyto-
plankton cells.
flow straightener a collimator (q.v.).
flow meter a device for measuring the volumetric flow rate of a fluid.
flow vector the direction and speed of aflowat a given point in a fluid.
flowthrough flume a type of flow channel where the water makes only
a single pass through the experimental working section.
fluid a liquid or a gas.
flume boundary layer the sheared layer that develops on the side walls
and bottom of flumes during passage of water.
force the external agency which changes the state of rest or motion of
a body and which is measured in units of dynes in the cm/g/s system.
396 Glossary
form drag that portion of drag on an object in a flow caused by the

uneven distribution of pressure over it.
free-stream velocity the velocity characteristic of the water body above
the BBL; mainstream flow (q.v.).
friction those forces which resist relative motion between surfaces in
contact.
frontal area the projected area of a body in a flow where the projection
is onto a plane normal to the flow vector.
Froude number a dimensionless expression calculated as the ratio of
the square of the relative speed to the product of the acceleration of
gravity and a characteristic length of the body.
gypsumfluxmethod a means of measuring water speed by determining
weight losses from solid gypsum, after flow exposure and dissolu-
tion, which is dependent on exposure time,flowspeed, salinity, and
temperature.
HID height/diameter ratio of epifaunal tubes.
Hamaker constant a mathematical expression used to predict the mag-
nitude of the London-van der Waals force between two surfaces in
an electrolyte.
head tank a water storage container placed above its delivery point. If
it is continuously supplied to maintain a constant head, a constant
discharge can be maintained.
hertz (Hz) a measure of electromagnetic wave frequency: one cycle per
second.
homeorheostat a type of self-regulating system which is capable of
changing its responses with time, e.g. filtration rates in bivalves.
hotfilmprobe a velocimeter which sensesflowspeed by measuring the
temperature change of a thermistor heated at a constant rate and
immersed in the flow.
hydrodynamics that branch of physics dealing with fluid motion.
hydrostatic pressure the force per unit area exerted on a surface in a
fluid as a consequence of the weight of thefluidabove it; also termed
gravitational pressure.
inertia the state of rest, or uniform motion, of a body that exists until it
is acted upon by an external force.
inference a conclusion reached as a result of the use of either inductive
or deductive methods.
internal wave an orbital motion propagating through a water body in
which the motion does not reach the water surface and which is
Glossary 397
generated by a variety of physical means, e.g. tidal currents reflect-

ing off a sloping bottom.
inviscid fluid a theoretical concept of an ideal fluid in which there is no
frictional resistance to motion.
invisible college an informal group of people sharing common interests
in a subfield of study.
isokinetic equal kinetic motion, as in sampling from an ambient flow
with the same velocity in the sampling tube.
kinematic viscosity the absolute viscosity (q.v.) of a fluid divided by its
density.
kinetic energy the energy of a body which results from its motion;
calculated as the product of mass and velocity squared.
laminar flow an orderly state of flow whereby the fluid moves in layers
and any small disturbances are dampened by viscosity.
laminar sublayer that part of the BBL closest to the bottom in which
the velocity profile is a creeping flow; i.e. momentum transfer
is determined by the viscosity of the fluid. Also viscous sublayer
(q.v.).
larval mimic a physical model of a larva designed to have the same
physical characteristics in flow, e.g. size and fall velocity, as the
prototype.
laser Dopplerflowmeter a type of velocimeter which utilizes the Dop-
pler shift in the frequency of laser light scattered from particles in
the fluid.
lift force the supporting force on a hydrofoil (or airfoil) in a flow of
water (or air) which is caused when the pressure on the underside
exceeds that on the upper surface, because the velocity on the latter
is greater.
lift spoilers structures present on the underside of an air- or hydrofoil
which help dissipate the pressure differences between the upper and
lower surfaces, thereby decreasing lift or increasing drag.
macrotidal the locations with a large tidal prism.
mainstream flow the free-stream velocity (q.v.).
marine snow an aggregate suspended in seawater consisting of various
types of sestonic particles held in a mucilaginous matrix.
matched flows isokinetic flows (q.v.).
mucous net a net made from the protein mucin by some suspension
feeders in order to collect sestonic food particles by a sieving
mechanism.
398 Glossary
multiple-channel flume a flume with two or more flow channels which

all share the same water supply.
negative buoyancy a characteristic of a body which is of greater aver-
age density than the fluid which surrounds it when the upward
pressure exerted by the fluid is insufficient to keep it floating.
normal to a characteristic of a body exposed to flow in which a surface
or plane opposes, or is oriented at right angles, to the mainstream
flow.
opposed to a characteristic of a body exposed to flow in which a
surface or plane is oriented so that it is parallel with the mainstream
flow.
oscillating flow the orbital water movement created by wind action on
a free water surface.
P:B production/biomass ratio, or the annual turnover ratio of a
population.
Peclet number a dimensionless group which compares the relative im-
portance of advection and molecular diffusion in the transport of
heat or solutes. It is of the form ULkT1, where U is velocity, L a
length characteristic, and AT1 the molecular diffusivity of heat or
solute.
Pitot tube a device which determines the stagnation pressure of a
flowingfluid,in a tube pointing into the flow and measured with a
pressure gauge, as an indication of its speed.
POM particulate organic matter.
porosity the ratio of the volume of interstitial space to solid, as in a
filter or in soil: the void fraction.
positive displacement pump a pump in which a chamber of fixed vol-
ume is alternately filled and emptied, often by a rotating or recipro-
cating piston.
potential energy the energy content of a fluid due to its position in a
gravity field; its capacity to do work.
pressure the force per unit area exerted by a fluid as a result of its
weight and motion.
pressure head that pressure at a point in aflowwhich is expressed as its
hydrostatic equivalent. The head is equivalent to the height that the
fluid would rise in a vertical tube open to the atmosphere and which
is connected to a probe in the flow.
pump characteristic the relationship between the pressure difference
and the volumetric flow rate, or pump capacity, for a hydro-
mechanical pump.
Glossary 399
raceway an open channel, oval-shaped flume which employs a fixed

volume of recirculated water.
radial the symmetry of an animal, e.g., a starfish, whose similar parts
are arranged in a regular pattern around a central axis and from
which any equal section yields mirror image halves.
recirculating flume a type of flume in which the same, fixed volume of
water is passed repeatedly through the working section.
reef path length the distance along a tidal current excursion occupied
by a suspension feeding reef, e.g. a bivalve bed.
residence time the time spent by a conservative molecule in a tidal
body of water before being passed along to an adjacent body by all
types of water movement.
resonance a natural phenomenon occurring in the sea where an exter-
nal periodic driving force enhances the oscillation amplitude be-
cause the force approaches the natural free oscillation frequency of
a local system, e.g. tides in the Bay of Fundy.
resuspension the process by which previously deposited sedimentary
particles are carried into the water column by water movements or
biogenic activity.
reworked sediment a sediment that is being or has been transported
from a bed either by water movement or by bioturbation activity.
Reynolds number a dimensionless number which expresses the relative
magnitude of inertial and viscous forces in a moving fluid. The
transition from laminar to turbulent flow occurs at a critical value of
this number specific for each flow.
rigid circular cylinder a solid cylinder.
rip current a turbulent, localized counter-current to the ambient flow.
roughness the grain, or flow-induced sculpturing, of a soft sedimentary
bed, or sculpturing of a hard substrate as it affects hydrodynamic
resistance, characterized by Zo.
Rouse number a dimensionless number giving the ratio between the
sedimentary particle settling rate and the boundary shear velocity.
Used to determine whether a particle will resuspend or remain as
bed load in a given flow.
sag curve an empirically derived graphic relationship between a spatial
variable on the ordinate and an environmental variable on the ab-
scissa, which shows a pronounced dip, e.g. dissolved oxygen in a
polluted estuary.
SCUBA self-contained underwater breathing apparatus.
SDI seston depletion index (see page 278).
400 Glossary
second law of thermodynamics the entropy (a measure of the random-

ness) of an isolated system must in time either increase or reach an
equilibrium.
sediment-water interface that part of a bed of deposited sediment at
the interface between sediment and water.
sediment trap a device for collecting settling sedimentary particles
from the water column.
sedimentary geology surficial geology (q.v.).
semi-exposed partly open to wind-wave effects.
seston depletion a process in which the seston concentration within the
BBL becomes reduced, as a result of consumption, as it passes over
a suspension feeding reef.
settling velocity fall velocity (q.v.).
shear stress the force acting on a unit area between two layers of a fluid
which are moving at different velocities and where the force is
applied in the flow direction, with the plane of the unit area being
tangent to the streamlines.
Shields parameter a dimensionless expression relating the shear stress
to the buoyant weight of a particle per unit area of bottom. Used in
determining the shear stress needed to move sedimentary particles
of a given size and density.
sieving a process in which sestonic particles are removed from aflowof
water passed through a mesh, with the minimum particle size
trapped being dependent on the mesh spacing.
skimming flow a flow interacting with a regular array of obstacles
which at a certain density cause the shear zone of the flow to relo-
cate above the obstacles, thereby protecting the local sediments
from erosion.
slip velocity the flow speed at which an epifaunal, free-living organism
first begins to be transported downstream by the flow.
stagnation point that point where the central stream line of a steady
flow has zero velocity at the surface of a solid body.
stagnation pressure dynamic pressure (q.v.).
stall for a swimming body, e.g. a scallop, the result of a decrease in lift
and an increase in drag causing the body to become unstable and fall
out of the flow. In scallops the stall is commonly due to a change in
the critical angle of attack of the body, which results in a change
from a smooth to a separated flow over the body.
static friction coefficient if a solid body is placed on a horizontal plane,
and a force applied parallel to the plane, the body will slide at some
Glossary 401
critical force. For this force the static friction is the difference be-
tween buoyant and gravitational forces on the body, multiplied by
the static friction coefficient.
steady flow a flow field which does not change with time.
stochastic a system model in which the relationships are random or
probabilistic, so that an input results in many possible outcomes
decided by chance. Opposite of deterministic, in which an input
results in one outcome only.
Stokes law the relation of the total drag, F, on a sphere of diameter, D,
moving in a still Newtonian fluid, to its velocity, [/, and the viscosity
of the fluid, (i, in creeping flow conditions. Thus: F=3n[iDU.
strain gauge a device for measuring the force of attachment of
an organism to a hard substrate, by the use of attached pressure
sensors which change in electrical resistance as the sensor is
strained.
stratification the separation of a water body into discrete horizontal
layers of the water column by physical environmental variables, e.g.
salinity or temperature.
streamlines the characteristic flow lines of a steady flow as can be
visualized by observing paths of neutrally buoyant particles. The
streamlines are everywhere tangent to the local velocity vector.
Strouhal number a dimensionless frequency used in analyzing oscillat-
ing fluid flows. The ratio of the frequency of the oscillation to
UD'1, where U is the velocity and D a dimension used to character-
ize the flow.
stuw zone the most inland part of some estuaries where freshwater is
backed up by the tidal stream during a flooding tide, causing a
freshwater tidal rise and fall.
submarine canyon a sharply shelving valley on the seabed.
surf zone that part of the shore between the landward limit of wave
uprush and the most seaward breaker.
surface slick that part of the sea surface where up- or down-welling
occurs and which is often marked by flotsam and jetsam, e.g. where
Langmuir cells meet.
surficial geology the scientific study of unconsolidated sedimentary
deposits on the land or seabed.
suspension feeder-biased sampling simulation of feeding by a named
suspension feeder. Sampling may be isokinetic or non-isokinetic,
but at the same velocity as used naturally by the suspension
feeder.
402 Glossary
tangential cross-flow filter a type of industrial filter which employs

centrifugal force to carry particles to the filter and a cross-flow to
remove and carry away the collected particles, so that the filter is in
continuous use.
tectonic the regional movements, caused by geological forces, of the
large plates of which the Earth's crust is composed.
thermistor a device consisting of a semi-conductor whose resistivity
decreases at higher temperatures and which is used in measurement
systems, e.g. temperature or velocity. Used as a velocimeter by
heating the semi-conductor at a constant rate and then measuring
the increase in resistance due to cooling, which is velocity-
dependent.
throat section that part of aflumewhich restricts theflowfrom a larger
to a smaller cross-sectional area.
tidal barrage a civil engineering structure designed to control
tidal flows. May be used in flood control or capture of potential
energy by passing seawater through turbines to convert it to
electricity.
tidal bore the inability of a flooding tide in a shallow estuary to propa-
gate in shallow water as rapidly as required, which causes a discon-
tinuity in level, resolved by a breaking wave advancing rapidly up
the estuary.
tidal excursion the net horizontal distance travelled by a conservative
particle in a tidal flow during one flood-ebb cycle.
tidal power the electrical power obtained by passing seawater through
turbines placed in a tidal barrage.
tidal prism the volume of seawater in a given water body between high
and low water levels.
torque the moment of force that causes torsion or twisting in a body.
tow tank a fluid container through which model objects can be me-
chanically drawn to simulate flow effects, e.g. for testing ship hull
design.
trophicfluidtransport system an enclosed tubular system for conduct-
ing water to, and from, the seston capturing surfaces of a suspension
feeder.
turbulence the irregular motion observed in fluids at high Reynolds
numbers, characterized by rapid mixing of heat, mass, and
momentum.
turbulent boundary layer a discrete layer near a wall where momentum
is transferred across the flow by turbulence rather than by viscous
stress.
Glossary 403
turning vanes flow straighteners designed to assist flume flows turn

corners, so that the stream lines remain parallel.
two-dimensional flow flow where it is assumed that its velocity and
other properties vary in two of the coordinate directions, but are
constant in the third direction.
van der Waals force the weak attractive forces which occur between
two atoms or nonpolar molecules.
vectored jet a jet of water whose direction can be adjusted, e.g. by
scallops to aid swimming or recessing behaviour.
velocimeter a device for measuring velocity.
velocity calibrator a device for calibrating a velocimeter.
velocity defect layer the region offlownear a solid boundary where the
velocity is reduced, e.g. BBL (q.v.).
velocity gradient the rate of change of a velocity component with
distance.
velocity profile the distribution of velocity along a transect perpendicu-
lar to the flow.
velocity vector the magnitude and direction taken by a flow.
vertical mixing the physical processes by which a water body becomes
mixed in the vertical plane with respect to the seabed.
viscosity (absolute) in a Newtonian fluid the shear stress exerted on a
fluid is proportional to the gradient of velocity times the absolute
viscosity. A measure of the resistance afluidoffers toflowwhen it is
subjected to a shear stress.
viscous dissipation the rate at which turbulent energy is lost as heat as
a result of turbulent stresses working against the viscosity of the
fluid.
viscous drag that portion of drag on an object in a flow caused by
viscous stresses on its surface. Also called skin friction. Total drag in
a steady flow is the sum of form drag and viscous drag.
viscous sublayer laminar sublayer (q.v.).
volumetric flow a measure of fluid motion which defines the volume of
fluid passing through a reference area in unit time.
wake that flow region in the lee of a submerged body which is altered
by the interaction between the flow and the body.
wall effect the property of flowing water constrained by walls to form
boundary layers there; also solid boundary effects on the terminal
settling velocity of a particle in a fluid.
water tunnel a closed or tubular conduit, oval shaped flume, which
employs a fixed volume of recirculated water.
wave an orbital motion induced near the free surface of a fluid as the
404 Glossary
result of an applied force, e.g. wind stress, and which propagates

throughout the water body.
wave exposure a relative measure of the degree to which wind-wave
action occurs at a site.
weir a dam in a flow whose purpose is to control or measure the flow
speed, e.g. an adjustable weir to regulate stream runoff.
wind-wave the effect of wind action on a water surface in producing
orbital wave actions.
Index
Specific plant or animal names and words defined in the glossary are excluded.
adductor muscle, 248-62 carbon budget, 319, 346-7

adhesion coefficient, 172-3, 178, 199 carbon isotope, 321
advection, 79-118, 185-6, 276-81 carrying capacity, 248, 294, 373-6
aerosol, 15, 38, 170-209 chemical inducers, 103, 119, 370
allometry, 245, 256 chlorophyll a, 42-51, 155, 286-93, 339-52
ambit, 234, 384 choanocyte, 159
amensalism hypothesis, 301-2, 305 cilia
aquaculture, 9, 350, 366, 374, 376 frontal, 204-7
asymptote, 45, 137, 161, 191, 208, 288, 325, lateral, 146-7, 203-4
381 laterofrontal, 204-5, 234
atrium, 150 circular cylinder, 173-230
autecology, 371 cirri, 156, 205-209, 226
clearance rate, 47-53, 128, 143-52, 173-95,
bedload transport, 92, 101, 229 277-83
benthic ecosystem tunnel, 21-3, 57 coastal zone management, 374
benthic impoverishment, 299-301, 377 colloid, 188
Bernoulli equation, 35, 243 colony, 134,148-50,201,219-22,247-51,384
bicarbonate, 314, 319 community
biodeposit, 350 deposit, 212-4, 299
biofilm, 114 impoverished, 272-4, 299-300
biogeography, 380, 383 mixed, 272-4, 299
biomass, 6, 58, 88, 250-3, 287, 294-5, 344, continental shelf, 32, 85-7, 152, 376, 385
346, 352 criticism, 8, 145, 271
bioturbation, 56, 275 crossflow distance, 179
bivalve reef, 21, 341-53 ctenidia, 162
Blaz ka respirometer, 236 culture, 42-59, 75, 97, 100, 142, 151, 248,
bloom, 95, 351-4 279, 350, 371-6
blowout, 336-7 cysts, 135, 201, 220-1
boreal, 294, 315, 326
branchial basket, 152 de-coupling, 323, 354
Brownian diffusional deposition, 178-209 decomposer, 314
buoyancy, 178-91, 243 deduction, 378
byssus, 89-91, 217-61 demibranch, 147
denitrification, 335-6, 350
camera deposit feeder, 2, 102, 138-9, 162, 245-50,
video, 30, 36, 53-4, 371-2 271-6, 299-305, 334, 367-9, 386-7
U/W stereo photography, 54, 217 deposit-suspension, 105, 129, 161, 200, 233,
U/W time-lapse photography, 234, 372 246
405
406 Index
detritus, 2, 170, 333, 347 flushing time, 83

diatoms, 56, 95, 155, 251, 284, 339^2 flux
dietary resource overlap, 303 carbon, 313
diffusion factor, 33 energy, 313
direct interception, 178-208 nutrient, 313
dispersal, 14, 69, 73-119, 262, 272, 370, food web, 276, 314, 346
384-6 force coefficient, 239
dissolved oxygen, 18, 23, 56-7,160, 292, fossil record, 385
342-6 fouling, 97, 260, 288
distribution front, 82
contagious, 250 Fry paradigm, 371
random, 250 functional group, 334, 367
regular, 250
gastrovascular cavity, 134
ecology, 1-15, 17-8, 43, 47, 56, 97-120, geological record, 4, 379, 383
174, 270-2, 325, 370-83 gill axis, 147
ecosystem analysis, 58, 353-4 gill bypassing, 47, 371
egg, 305-6, 373 gravitational deposition, 178,182,195,201-2
emergent property, 305-6, 373 groundfish, 6
epifauna, 3, 55, 213-62, 302, 372
epifluorescence, 42 halocline, 81
estuary head, 1&-44,106, 134-54
partially mixed, 79, 81-2 herbivore, 319, 329, 333
salt wedge, 79, 81-2 heterotrophic, 342, 353
well mixed, 81 holism, 8, 373
ethology, 4-14, 17, 43, 260 hydrodynamic baffling, 337
eutrophic, 351-3 hydrodynamic retardation, 201
evolution, 3-13, 164, 378-86 hydrogen bubble method, 34-5
exhalant, 30-55, 105-64, 215-43, 283, 381 hypernutrification, 351-2
extant, 4, 384-5 hypothesis
extinction, 378-86 alternate (=alternative), 7-14, 89-98,
118-20, 161, 237, 261-2, 302, 326^0,
faeces, 51, 302, 323-50 370-1
filter multiple, 7-9, 13, 165, 261
fibre bed, 177, 208 null, 7-15, 82-98, 161, 313-37, 371
mucus net, 150
filtration rate, 25, 43, 128-48, 159-62, 236, inbreeding, 384
277-305, 341-2, 374-5 induction, 159-60
direct, 49 inertial impaction, 178-208
indirect, 48-9 infauna, 99, 161, 213, 301-2, 334
fitness, 383^ ingestion, 44, 130-54, 187
fjord, 81 inhalant, 40-50, 130-64, 204-35, 277, 282,
flick response, 138 305, 342
flow initial feeding, 41-7
field, 18, 23, 163, 225, 349 interfilament canal, 146-7
freestream, 35, 93, 103, 133, 217 internal wave, 79-87, 317
laboratory, 24, 27 isotopic decay, 379
turbulent, 99, 199, 298, 316
unidirectional, 18-29, 100-34, 217-36, joint, 133
281, 316, 325, 337, 369-74
flume kelp forest, 87, 90, 315-333, 353
annular, 29-30, 102-3, 162, 327 overstorey, 316-327
Denny, 25-6 understorey, 90, 316-323
Lofquist, 27 key predator, 90
Saunders-Hubbard, 25-6, 244 kin recognition, 115, 119, 384
Vogel-LaBarbera, 25-6, 28, 101, 137, Kirby-Smith growth apparatus, 18, 288,
143-4 297
Index 407
lacustrine, 314 pallial cavity, 255

Lamb solution, 181, 197 paradox, 305
lamina, 315-24 particle counter, 133, 151, 203
laptop computer, 30 Pearson correlation, 301
larva peduncle, 133, 227
competent, 71-119, 385 pediveliger, 44, 75-103
conspecific, 115, 119 pelagic, 69-72, 177, 314, 319, 346-7
cyprid, 73 pelagic-benthic coupling, 58, 346, 348, 373
eyed veliger, 75 perfusion, 381
lecithotropic, 71-5, 383^4 photography, 23-54, 109-13, 137, 217, 256
megalopa, 118 time-lapse, 36, 234, 372
naupillus, 134 photokinesis, 83
planktotrophic, 71-99, 383-5 phylogeny, 379-85
plantigrade, 75, 91 physiology, 4 ^ 3 , 128, 272, 345-6
sibling, 115, 383 phytoplankton, 39-43, 95-6, 170-95,
tadpole, 74 283-454
trocophore, 75 pinnules, 133. 137,197-205, 234
velichonca, 75 plankton net, 45, 117
veliger, 75, 83 plant canopy, 87-9, 118, 313-26
larval polyp,133-5, 201-2, 220, 241
flux rate, 103 population, 2-20, 56-9, 82-92, 117-30, 161,
orientation, 17 235-50, 270-306, 337-84
Laurentide ice sheet, 386 post-larva, 69-118
ligament (=resilium), 254-5 pre-adapted, 385
lunule, 246 predator, 90, 247-62, 302-40, 385
predator exclusion cage, 116
macrofauna, 1-14, 56, 92-116, 163, 227-41, pressure
270-300, 321-47, 387 difference, 35, 49, 140-2, 158, 176, 223,
macrophyte, 326 246
maintenance dredging, 377 field, 50, 132, 140, 160, 164
mantle wall, 140, 147 loss, 174, 205
mass transfer, 61, 270-9, 314-19 rise, 152, 154
materials cycling, 313-23 transducer, 35, 255
meiofauna, 1, 56, 88-105, 301, 342 primary producer, 174, 313-14, 333-46
metamorphosis, 69-113, 225 production, 9, 56-9, 97, 139, 227, 247,
meteorite, 378 271-385
microfauna, 1 proximate question, 366, 379
monoculture, 59, 380 pseudofaeces, 47, 51, 128, 130, 335, 350
mud box, 46 pulsating flow, 206, 209
multiple-use resource conflict, 377 pump
ciliary, 4, 11, 132, 140-64, 205, 223,
Neo-Darwinian evolution, 13 380-2
net surface charge, 186-7 flagellar, 129, 158
neurohormone, 371 muscular, 3, 38, 129, 209
niche, 73, 98, 250, 300, 382^ positive displacement, 154
nitrogen fixation, 335-6, 350 pump capacity, 381
notopodia, 203 pumping rate, 47-51, 144-62, 279-85, 381
nutrient cycling, 314, 332-5
Q10, 147, 259
Occam's razor, 373
Oligocene, 385 rachis, 133
ontogeny, 81, 118, 256-63, 382 re-mineralized, 350
osculum, 158-9 rebound, 386
ostium, 205 recruitment, 14, 44-7, 69-119, 247, 272-
outbreeding, 384 306, 367
reductionism, 374
palaeoecology, 379-80 reefs, 3-12, 54-9, 73-119, 219-61, 284, 313
408 Index
research standing crop, 250, 287, 294, 329

experimental, 8, 271 static head, 40
respiration, 56, 145, 148, 286 statocyst, 101
retention, 79-84, 118, 173-203, 318, 352, stigmata, 150, 152
385 stopping distance, 183-4, 208
retention efficiency, 47-50, 151-63, 176- Streamflo probe, 35, 37
207 stretch receptors, 255
rheokinesis, 213 sublittoral, 1, 217, 251, 278-9, 301, 320-1,
rheotropism, 213, 260 345, 370
rhizome, 89, 327-39 substrate, 1-4, 21, 46, 69-119, 225-59,
rip current, 249 272-6, 315-24, 370, 384-5
rugophilic, 113 suprabranchial cavity, 146
surface drifter, 86
science suspension feeder
applied, 366 active, 3-4, 40-50, 129-45, 159-75,198-
fundamental, 366 226, 287, 371-82
sea carousel, 30 combined passive-active, 129, 153-8, 200
sea table, 241 deposit-, 88, 129, 161, 200, 233, 246, 270
sea urchin barrens, 319-20 facultative active, 129, 155, 200, 238
seagrass passive, 38, 129-32, 173, 198-226, 251,
meadow, 21, 87-90, 326-254 292, 380
productivity, 329-32 sieving, 150-63, 203
stability, 328-36 swimming, 44-56, 74-118, 171, 214, 235,
secondary production, 294, 301, 347, 378 254-62, 384-5
sediment symmetry
deposition, 377 bilateral, 160
erosion velocity, 277 radial, 135, 214
friction, 300 synecology, 371
sediments
clay, 170, 275, 295, 334, 347, 387 taxes
sand, 57, 132, 217, 291, 303, 387 geotaxis, 83, 85, 101
silt, 275, 295, 387 phototaxis, 81, 83, 85
surficial, 27, 277, 386 rheotaxis, 213-41, 385
seeding density, 373 taxonomy, 130, 379, 384
seston tentacles, 101, 133-62, 198-219, 229-33,
concentration, 9, 20-52, 130-63, 172-3, 241, 260
221, 236, 277-305, 337-9, 371-6 theory
flux, 172, 247-9, 275-99, 337, 375 aerosol (=particle collection), 15, 170
quality, 9, 43, 59-60, 145, 292, 337-40, aggregation, 363, 384
371-6 benthic impoverishment, 363
sampling, 38, 41 bivalve pump, 363
settlement, 14, 44-7, 69-120, 216-50, 276- limiting factor, 270
303, 367-85 particle collection, 363
gregarious, 115-19, 249, 262 passive-active, 363
plate, 46, 73, 100-16 perturbation, 315, 353
sink, 44, 71-104, 171, 182, 208, 254, 259, trophic group mutual exclusion, 270-6,
283, 334 301
siphon unimodal, 363
atrial, 150-2 thermistor bead velocimeter, 35
buccal, 150-2 tidal current, 3, 18, 21, 83, 156, 215, 249,
exhalant, 30, 49, 55, 132-47 300, 332-4, 377, 386
inhalant, 41-2, 162 till, 386
solar radiation, 314 transplant experiments, 294, 354, 384
spatfall, 83, 99 trophic level, 314, 321
spawning, 54, 95-6 tube
speciation rate, 380, 383 complex, 234
species richness, 380 normal to, 227-30
Index 409
opposed to, 227, 232 sampling, 40

spar buoy, 227, 234 vertical eddy diffusion, 281-3
truncated cone, 227,232 virus, 188
tube feet, 54, 137, 247, 251 viscous, 56, 94-101, 152, 181, 199-209, 229,
tube feet flick, 137 244
vortex, 229
ultimate question, 3-6, 363-87
unimodal function, 130-45, 163, 294, 353 wave surge, 152, 219, 226, 242-7
upwelling, 79, 86-7, 323, 361 "whirling" feeders, 232
wind
valve shelter, 301
closure, 144-8, 375, 381 stress, 301
gape, 54, 147, 254 tunnel, 56
velocity
bedload shear, 100 zooplanktivore, 202
mean tidal, 59, 295

978 0883850305

Uploaded by

Copyright:

Available Formats

978 0883850305

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

978 0883850305

Uploaded by

Copyright:

Available Formats

Benthic suspension feeders and flow

Benthic suspension feeders

Cambridge University Press

The Edinburgh Building, Cambridge CB2 2RU, UK

www. Cambridge. org

This publication is in copyright. Subject to statutory exception

First published 1997

This digitally printed first paperback version 2005

A catalogue recordfor this publication is available from the British Library

Library of Congress Cataloguing in Publication data

ISBN-13 978-0-521 -02347-4 paperback

Facultative active suspension feeders 155

Hydrodynamics and benthic biology represent a very broad field encom-

training in chemical engineering and special interest in how flows allow

Many people, in various ways, helped in the completion of this book.

Oceans, in the persons of Mike Sinclair and Wendy Watson-Wright,

Material modified with permission from D. O. Duggins, C. A. Simenstad, &

One objective of benthic ecology is to describe the spatial distribution of

Table 1.1. Arbitrary size divisions of benthic fauna.

Microfauna 1-100 Some smaller Protozoa

Source: Based on Parsons et al. (1977).

The reader should be aware that there is a parallel interest among

pended in and transported by seawater flows. If the seston is carried by

Asking the right question

Discipline No. Proximate Ultimate

Ethology 3 How do competent larvae Why have some sessile

Ecology 6 How does larval How do larvae with specific

for the first physiological ultimate question of Table 1.2, we have no

forms, such deductive methods often do not provide an adequate answer

Table 1.3. Important concepts in ecology derived from a survey

No. Description Examples

Source: Cherrett (1989).

In this book we have excluded the long-term, temporal perspective in

Table 1.4. Major types of scientific method.

Librarian Research Scientist

|lnternal review|-- Reject

Figure 1.1 Outline of the process used in science to conceive, experimentally

alternative and null hypotheses are tested by inferential experimentation

individual creates a new theory and/or new hypothesis referable to an

original abbreviated research articles thought to be of wide general

cations, to contain clear descriptions of most of the ideas of interest to

Table 1.5. Most cited journals in Chapters 2-8.

Aims and structure

Table 1.6. Pioneering modern interdisciplinary studies involving

Subdiscipline Chapter Authors and date

sionals in the interdisciplinary field of hydrodynamics/benthic biology or

(Chap. 4) and capture mechanisms at the filtration surface (Chap. 5). In

Parsons, T. R., M. Takahashi, and B. Hargrave. 1977. Biological

The purpose of this chapter is to present an overview of the methods

and laboratory flumes. We also describe various ways of measuring

Figure 2.1 Kirby-Smith growth tube apparatus as used by Wildish and

calculated by timing collections of the volumetric flow and then dividing

The hydrodynamics of flow in a tube also differs from that in a natural

Multiple-channel flumes have also been used in field conditions so the

instantaneous rate of oxygen uptake, /, in units of g O 2 m 2 h 1, is

where v and c are the velocity and concentration of dissolved oxygen,

Table 2.1. Recirculating flumes designed for measuring physiological or

Volumetric Free stream

Water tunnel, 200 30 Marine plants Charters and Anderson