AQUATIC INSECTS

Challenges to Populations
This page intentionally left blank
AQUATIC INSECTS
Challenges to Populations
Proceedings of the Royal
Entomological Society’s 24th
Symposium

Edited by

Jill Lancaster
Institute of Evolutionary Biology
University of Edinburgh
Edinburgh, UK
and

Robert A. Briers
School of Life Sciences
Napier University
Edinburgh, UK
 CABI is a trading name of CAB International

CABI Head Office CABI North American Office

Nosworthy Way 875 Massachusetts Avenue
Wallingford 7th Floor
Oxfordshire OX10 8DE Cambridge, MA 02139
UK USA
Tel: +44 (0)1491 832111 Tel: +1 617 395 4056
Fax: +44 (0)1491 833508 Fax: +1 617 354 6875
E-mail: cabi@[Link] E-mail: cabi-nao@[Link]
Website: [Link]
 CAB International 2008. All rights reserved. No part of this publication may be
reproduced in any form or by any means, electronically, mechanically, by
photocopying, recording or otherwise, without the prior permission of the
copyright owners.
A catalogue record for this book is available from the British Library, London, UK.
Library of Congress Cataloging-in-Publication Data
Royal Entomological Society of London. Symposium (24th : 2007 : University of
Edinburgh)
Aquatic insects : challenges to populations : proceedings of the Royal
Entomological Society’s 24th symposium / edited by Jill Lancaster, Rob A. Briers.
p. cm.
Includes bibliographical references and index.
ISBN 978-1-84593-396-8 (alk. paper)
1. Aquatic insects--Congresses. I. Lancaster, Jill. II. Briers, Rob A. III. Title.

QL472.R69 2007
595.7176--dc22
2008000626
ISBN: 978 1 84593 396 8

Typeset by AMA Dataset, Preston, UK

Printed and bound in the UK by Cromwell Press, Trowbridge

The paper used for the text pages in this book is FSC certified. The FSC (Forest
Stewardship Council) is an international network to promote responsible
management of the world’s forests.
 Contents

Contributors vii

Preface ix

1 Aquatic Insect Adaptations to Winter Cold and Ice 1

Hugh V. Danks

2 Saline-water Insects: Ecology, Physiology and Evolution 20

Timothy J. Bradley

3 Larval Cannibalism and Population Dynamics of Dragonflies 36

Frank Johansson and Philip H. Crowley

4 The Ecology of Host–Parasite Interactions in Aquatic Insects 55

Steven L. Kohler

5 Effects of Drought on Stream Insects and its Ecological

Consequences 81
Andrew J. Boulton and P. Sam Lake

6 The Effect of Floods on Aquatic Invertebrate Communities 103

Russell G. Death

7 Life-history and Behavioural Adaptations to Flow Regime

in Aquatic Insects 122
David A. Lytle

8 Movement and Dispersion of Insects in Stream Channels:

What Role does Flow Play? 139
Jill Lancaster

v
vi Contents

9 Population Responses of Drifting Stream Invertebrates

to Spatial Environmental Variability: an Emerging
Conceptual Framework 158
Sebastian Diehl, Kurt E. Anderson and Roger M. Nisbet

10 What is the Spatial Structure of Stream Insect Populations?

Dispersal Behaviour at Different Life-history Stages 184
Barbara J. Downes and Paul Reich

11 Polarization Vision in Aquatic Insects and Ecological

Traps for Polarotactic Insects 204
Gábor Horváth and György Kriska

12 Evolution and Physiology of Flight in Aquatic Insects 230

James H. Marden

13 Evolutionary Drivers and the Ecological Traits of Adult

Aquatic Insects 250
Simon S.C. Harrison and Michael Dobson

14 Population Genetic Structure in Stream Insects: What

Have We Learned? 268
Jane M. Hughes, Daniel J. Schmidt, Alison McLean and
Arlene Wheatley

15 Habitat Constraints and the Generation of Diversity

in Freshwater Macroinvertebrates 289
Ignacio Ribera

Index 313
 Contributors

Kurt E. Anderson, Department of Biology, University of California, Riverside,

CA 92521, USA
Andrew J. Boulton, Ecosystem Management, University of New England, Armidale,
New South Wales, Australia
Timothy J. Bradley, Department of Ecology and Evolutionary Biology, University
of California, Irvine, CA 92697-2525, USA
Philip H. Crowley, Department of Biology, University of Kentucky, Lexington,
Kentucky, USA
Hugh V. Danks, Biological Survey of Canada (Terrestrial Arthropods), Canadian
Museum of Nature, Ottawa, Ontario, Canada
Russell G. Death, Institute of Natural Resources – Ecology, Massey University,
Private Bag 11-222, Palmerston North, New Zealand
Sebastian Diehl, Department Biologie II, Aquatische Ökologie, Ludwig-Maximilians-
Universität München, Großhaderner Straße 2, D-82152 Planegg-Martinsried,
Germany
Michael Dobson, Freshwater Biological Association, Far Sawrey, Windermere,
UK
Barbara J. Downes, Department of Resource Management and Geography, 221
Bouverie St, The University of Melbourne, Victoria 3010, Australia
Simon S.C. Harrison, Department of Zoology, Ecology and Plant Sciences,
University College Cork, Cork, Ireland
Gábor Horváth, Biooptics Laboratory, Department of Biological Physics, Physical
Institute, Eötvös University H-1117 Budapest, Pázmány sétány 1, Hungary
Jane M. Hughes, Australian Rivers Institute, Griffith University, Nathan, Queens-
land 4111, Australia
Frank Johansson, Department of Ecology and Environmental Science, Umeå
University, SE - 901 87, Umeå, Sweden
Steven L. Kohler, Department of Biological Sciences and Environmental Studies
Program, Western Michigan University, Kalamazoo, MI 49008, USA

vii
viii Contributors

György Kriska, Group for Methodology in Biology Teaching, Biological Institute,

Eötvös University, H-1117 Budapest, Pázmány sétány 1, Hungary
P. Sam Lake, School of Biological Sciences, Monash University, Clayton, Victoria,
Australia
Jill Lancaster, Institute of Evolutionary Biology, University of Edinburgh, Ashworth
Labs, West Mains Road, Edinburgh EH9 3JT, UK
David A. Lytle, Department of Zoology, Oregon State University, 3029 Cordley
Hall, Corvallis, OR 97333, USA
James H. Marden, Department of Biology, Penn State University, 208 Mueller
Lab, University Park, PA 16802, USA
Alison McLean, Australian Rivers Institute, Griffith University, Nathan, Queens-
land, 4111, Australia
Roger M. Nisbet, Department of Ecology, Evolution, and Marine Biology,
University of California, Santa Barbara, CA 93106, USA
Paul Reich, School of Biological Sciences, Monash University, Clayton, Victoria
3800, Australia
Ignacio Ribera, Departamento de Biodiversidad y Biología Evolutiva, Museo
Nacional de Ciencias Naturales (CSIC), José Gutiérrez Abascal 2, 28006
Madrid, Spain
Daniel J. Schmidt, Australian Rivers Institute, Griffith University, Nathan, Queens-
land 4111, Australia
Arlene Wheatley, Australian Rivers Institute, Griffith University, Nathan, Queens-
land 4111, Australia
 Preface

It is customary in the preface or introduction of a book to stress the importance

of the subject matter, in order to justify the publication and to encourage readers
on to the main text. One of the problems with this is the overuse, and consequent
devaluation of ‘importance’ (Brooker et al., 2005). As far as aquatic insects
are concerned, we feel that little justification for having an interest in this group
is necessary. Their persistent presence in the Earth’s fauna (>300 million years
since insects invaded aquatic habitats), their numerical dominance in contempo-
rary freshwaters (and some saline waters), and the diversity of roles they perform
in aquatic ecosystems speak for the importance of the group. In an academic
context, aquatic insects have served as model systems for the development of
understanding of many aspects of insect physiology, behaviour, ecology and
evolution, so there are many aficionados of this group.
The papers in this book were presented at the 24th International Sympo-
sium of the Royal Entomological Society, which was held at the University of
Edinburgh, UK, in July 2007. The title of the symposium and this book, Aquatic
Insects: Challenges to Populations, reflects an unashamed bias of the editors
towards processes acting at the population level, but the book’s content takes a
very broad view of populations. In planning the symposium, our aim was to
bring together key workers in diverse fields, to take an integrated view of the
challenges facing aquatic insects and to foster a broad appreciation of the links
between subject areas that can lead to deeper understanding. It is easy and com-
forting to stay within the relatively narrow confines of one’s own discipline or
area of interest, but the greatest prospect for progress and new ideas often comes
through discovering connections between apparently unrelated ideas. Thus, this
book considers some of the potential influences on individuals and populations
(e.g. environmental stresses, parasites, cannibalism, dispersal limitations), the
‘cunning tricks’ used by aquatic insects to overcome challenges (e.g. polarization
vision, life-history strategies, osmoregulation, cold hardiness) and the consequences
of those challenges at different levels of organization (e.g. distribution patterns,

ix
x Preface

population structure, population genetics, evolution). Similarly, the set of papers

encompass many taxa and all life stages of aquatic insects, not just the ‘aquatic’
stages. It is impossible, in one book, to provide exhaustive cover of all the chal-
lenges to populations, and though this set of topics may seem eclectic, we hope
it provides a thought-provoking sample. During the symposium, we hoped that
the presentations would spark new ideas and synergisms; they certainly sparked
conversation! Similarly, we hope that readers will take the opportunity to read
multiple chapters and find inspiration for their own work, or simply a wider
appreciation for the wonders of aquatic insects.
It is also customary in the preface of a book to acknowledge the people who
often work behind the scenes but, without whom, the book never would have
existed. We would like to acknowledge our co-convenor of the symposium, Craig
Macadam, who was instrumental in getting us involved in the first place, and
contributed equally to the success of the meeting. We were also heartened by the
eagerness with which the speakers (and subsequent authors) accepted our invi-
tations and by their promptness in producing the required manuscripts. All those
who attended the meeting also contributed to its success, and to much lively
debate and discussion. Those who agreed to review chapters are also acknowl-
edged for their insightful and constructive comments, which improved the final
volume immeasurably.
We would also like to thank the Royal Entomological Society who provided
the generous funding, which enabled the meeting of a truly international, and
very high quality, set of speakers. Bill Blakemore, Registrar of the Society, and its
other staff and officers provided invaluable assistance with the organization of
the meeting and the flow of funds. Last, but most certainly not least, we thank
Graham Stone (University of Edinburgh) and his team who did so much to
ensure that the symposium ran smoothly.
Jill Lancaster
Robert A. Briers

References

Brooker, R., Kikvidze, Z., Pugnaire, F.I., Callaway, R.M., Choler, P., Lortie, C.J. and
Michalet, R. (2005) The importance of importance. Oikos 109, 63–70.
1 Aquatic Insect Adaptations to
Winter Cold and Ice
HUGH V. DANKS
Biological Survey of Canada (Terrestrial Arthropods), Canadian Museum of
Nature, Ottawa, Ontario, Canada
E-mail: hdanksca@[Link]

Abstract
Most aquatic habitats in cold regions are frozen or covered by ice during winter. Aquatic
insects survive these cold and icy conditions by a variety of adaptations. Different species
move seasonally between habitats, choose particular overwintering sites, modify local
conditions by constructing shelters, and withstand the effects of ice break-up and flood-
ing during the spring thaw. Limited data indicate that several kinds of aquatic insects
withstand subfreezing temperatures while surrounded by ice. Some of these species
supercool and others tolerate freezing. A review of the available information in the con-
text of recent discoveries in insect cold hardiness suggests fruitful avenues for research,
including the roles of dehydration and of antifreeze proteins. Also required is study of the
cold hardiness of individual species in relation to habitat conditions, seasonal move-
ments and mechanical protection by overwintering structures. We need even basic infor-
mation about the production and interaction of cryoprotectants, which are scarcely
known compared with those of terrestrial insects. Therefore, detailed biological informa-
tion must be collected about individual species of aquatic insects (in contrast to some
current preferences for generalized statistics about aquatic systems) and put into a wide
life-cycle context.

Introduction

Many aquatic insects live in regions that are cold enough for their habitats to be
ice-covered or frozen during winter. Questions to be answered in understanding
insect survival in these habitats include: where do aquatic insects go for the win-
ter? What do they do there? And how do they cope with the low temperatures
and the ice that it creates in many of the available habitats? I provide ecological
context by outlining the patterns of habitat choice and modification, and ways of
coping with the fact that ice seals the water surface during winter and greatly
disturbs lotic habitats during spring break-up, but focus here on the effects of
subfreezing temperatures (Oswood et al., 1991; Irons et al., 1993; Frisbie and
 CAB International 2008. Aquatic Insects: Challenges to Populations
(eds J. Lancaster and R.A. Briers) 1
2 H.V. Danks

Lee, 1997). Surprisingly little detailed biological and physiological knowledge

about these effects has been obtained in aquatic systems. Therefore, I emphasize
future research directions that are suggested by information on both aquatic
insects and insect cold hardiness in general.

Habitat Features
Winter conditions differ widely among different habitats (Table 1.1). Small bodies
of water are not well buffered against winter cold, and after formation of surface
ice, may steadily freeze to the bottom, especially above arctic permafrost. How-
ever, in cool-temperate regions, conditions are ameliorated because overlying
snow provides insulation from cold air temperatures and the ice may not get
very thick even on small ponds. Provided snowfall is moderate or heavy, large
areas of the bottom sediments remain unfrozen even where mid-winter tem-
peratures (December–February) average −8°C for 3 months (Danks, 1971a).
In large lentic habitats the sediments also remain unfrozen, typically at 4°C, the
temperature at which water is heaviest and stratifies beneath the surface ice.
In lotic waters, flow restricts ice formation, although rivers can freeze into the
substrate if flow is low during cold winters (Clifford, 1969; Irons et al., 1989).
When flow is maintained, river temperatures remain near 0°C and instead of
surface ice, frazil ice – suspended crystals that cannot consolidate because of
turbulence – is formed (Shen, 2003). The crystals may attach to the bottom and
to each other to form anchor ice, but even so the temperature stays close to 0°C.
Although winter conditions are not especially harsh in lotic waters of adequate
size, the melting of accumulated snow and ice produces major disturbances in
the spring, including ice scour, high flows and flooding (Prowse and Culp,
2003).
Winter conditions in aquatic habitats, therefore, depend on how the size,
geometry and flow of each habitat, together with the cold, ice and snow in a
given region and year, combine to govern winter severity (such as the minimum
winter temperature and the duration of ice cover), timing (such as the dates of
freeze up and thaw), spring disturbance and other features. Therefore, different
species show different habitat choices, habitat modifications, cold hardiness and
ways of withstanding other effects of ice.

Choice of Habitat

A major component of insect winter survival is the choice of a place to spend the
winter. Aquatic insects can spend the winter in their summer habitats even if
these might freeze. Alternatively they can move to different subhabitats within
the same water body, to different water bodies and even on to land. The winter
habitat dictates requirements for coping with adversity, especially levels of cold
hardiness.
Habitats that do not freeze completely, such as deep lakes and groundwater-
influenced streams, even allow activity during winter in species adapted for
 Aquatic Insect Adaptations to Winter
Table 1.1. Synopsis of conditions created by winter cold and ice in different aquatic habitats in cold regions.

Lentic habitats Lotic habitats

Small and Deep, permanent

Feature shallow ponds* ponds Large lakes Small creeks Medium rivers Large rivers

Start of permanent Early Early Late if at all Early Late Very late
ice cover
Exposure of Common Varies according to Infrequent Common Occasional; also Infrequent
organisms to habitat and weather frazil and anchor
adjacent ice ice
Time of thaw Early Varies according to Late, especially Varies according Varies Varies
size, morphometry, if ice is thick to flow, weather,
exposure, etc. etc.
Bottom condition Frozen Unfrozen Unfrozen Frozen Usually unfrozen, Unfrozen
depending on flow
Bottom temperature 0°C to much Below 4°C 4°C 0°C to much Typically close to Typically close
colder colder 0°C to 0°C
Winter modifiers Snow insulation Winter weather Winter weather Snow cover, etc. Winter weather, Winter
above warmer water supply weather
ground
Degree of ice None None Some None Often moderate High
scour in spring to high
Other common Anoxia beneath
factors ice
Peak flow None None Little Briefly high Usually high High
Flooding Usually yes Yes Usually no Often yes Yes Yes

*Including temporary ponds, container habitats, edges of larger water bodies, etc.

3
4 H.V. Danks

growth and reproduction at low temperatures. Indeed, several species complete

significant growth in winter, including some black flies, stoneflies, mayflies, cad-
disflies and chironomids; some emerge in late winter, as in winter stoneflies and
some diamesine and orthocladiine chironomids (sample references in Danks,
1991, p. 239).
Relatively few species remain in their summer habitat when it will freeze.
Many that do so are chironomid midges from northern ponds, lakes and streams,
which survive in habitats experiencing winter temperatures well below freezing
(Scholander et al., 1953; Danks, 1971b; Andrews and Rigler, 1985; Irons et al.,
1993; Bouchard et al., 2006; Lencioni, 2004). Other species that remain in
potentially frozen habitats include some dragonflies from prairie ponds (Daborn,
1971), empidids from Alaskan streams (Irons et al., 1993), caddisflies from tundra
pools (Wiggins and Winchester, 1984), midges from pitcher plants (Paterson,
1971) and mosquitoes from tree holes (Copeland and Craig, 1990).
Many insects move within the habitat from summer locations to areas less
likely to freeze, for example, from the edge to the centre of streams and ponds
(Gibbs, 1979; Olsson, 1982; Irons et al., 1993). Other species move deeper into
the substrate for winter, even into the hyporheic zone deep in the bed of streams.
However, there are many interspecific differences and it is not possible to con-
clude that penetration into the substrate is necessarily or commonly associated
with protection from cold. The hyporheic zone is preferred for summer growth
by some stream species, such as stonefly and mayfly larvae; and some lake and
pond species, including chironomid larvae, penetrate the substrate and become
dormant in summer when oxygen levels are low. A few species adjust position
even during the low temperatures of winter. For example, larvae of Leptophlebia
mayflies stay close to the ice and move with the ice-front as it changes (Olsson,
1983).
Some species move for the winter to deeper habitats of the same type that
are less likely to freeze (Pajunen and Jansson, 1969, for corixids). Several species
of beetles and bugs move from the temporary ponds in which they develop to
permanent ponds for overwintering, recolonizing temporary ponds in the spring
(Williams, 1997), although this strategy is not confined to regions with cold win-
ters. Species of gerrids, mosquitoes, limnephilid caddisflies and other taxa spend
the winter as adults in terrestrial habitats, typically sheltered in cavities or litter
beneath the snow (e.g. Hudson, 1978; Berté and Pritchard, 1983; Spence and
Andersen, 1994).
Species that move to different habitats for the winter return to the original
habitats in spring. However, some seasonal movements appear to reflect adapta-
tions against spring disturbance rather than against winter cold (see Withstanding
other seasonal effects of ice).

Modification of Winter Conditions

Some aquatic insects modify the habitat to improve winter survival, as opposed
to simply choosing the most potentially favourable sites. Many species that
penetrate the substrate build burrows lined with silk or mucilage. The roles of
Aquatic Insect Adaptations to Winter 5

most such structures in cold hardiness have not been demonstrated, although
several unrelated adaptive values have (Danks, 2002), such as the protection
against predation afforded by the burrows of chironomids (Van de Bund and
Groenenkijk, 1994).
Other species overwinter in cocoons or shelters. Often these structures are
more robust or of a different form than the summer equivalents. For example,
the distinctive winter cocoons of chironomids enhance survival during experi-
mental freezing (Danks, 1971b). Inside these cocoons the larvae normally fold
under at least the posterior segments, presumably to protect the body and espe-
cially the anal processes from damage caused by surrounding ice (Danks,
1971b). Cases of some other aquatic insects are attached to the substrate (Hauer
and Stanford, 1982) or sealed (Olsson, 1981) for the winter period. Of course,
similar adaptations are seen in summer (Hauer and Stanford, 1982) and the
shelters protect against predation and other challenges (Dillon, 1985; Danks,
2002).
Finally, females of some species govern the specific winter conditions for their
eggs or progeny by selecting and modifying oviposition sites. Eggs of Lestes drag-
onflies that overwinter encased in ice in prairie ponds are inserted into plant stems,
which may provide some mechanical protection (Sawchyn and Gillott, 1974).
Limnephilid caddisflies deposit their eggs in temporary ponds in a gelatinous
matrix that imparts drying and freezing resistance to the eggs (Wiggins, 1973).

Cold Hardiness

Cold hardiness in insects in general has been studied in some detail. For recent
reviews and updates see Ramløv (2000), Bale (2002) and Danks (2005).
Low temperatures above freezing are injurious to many insects (‘chilling
injury’). However, larvae of some stoneflies, mayflies and chironomids from
northern regions and from cold habitats such as mountain streams are adapted
for activity at temperatures as low as 0.1°C (Brittain and Nagell, 1981). Cold-
stream species that are killed by freezing survive winter even when water tem-
peratures are very close to 0°C (e.g. Oswood et al., 1991). Temperatures in most
deep lakes are stable at 4°C during winter. Therefore, although there is little
information for aquatic species, susceptibility to chilling injury may be less wide-
spread in aquatic than in terrestrial species from cool regions.
Most terrestrial species remain unfrozen in winter, even at temperatures
below 0°C, by supercooling. Ice forms in solution by aggregating around a
nucleus that encourages crystal formation, so that if nucleators can be masked or
eliminated water will supercool to about –40°C, the temperature at which the
molecules move so slowly that their aggregation leads to crystallization at the
homogeneous or spontaneous freezing point (Vali, 1995).
Nucleators can include internal molecules and surfaces, gut contents and ice
crystals that inoculate the body fluids through the cuticle (‘inoculative freezing’).
Inoculative freezing would be expected in aquatic habitats in which overwinter-
ing insects are surrounded by ice (but see below). Of course, the true freezing
points of typical insects, depressed by the presence of cell constituents, are up to
6 H.V. Danks

a few degrees below the 0°C freezing point of pure water. Some aquatic insects
freeze in the presence of ice at relatively high subfreezing temperatures (table 3
in Frisbie and Lee, 1997). However, the supercooling points of other aquatic
species are still lower even when wet, suggesting that they would tend to remain
unfrozen in the aquatic habitat too. For example, supercooling points of empidid
larvae overwintering in an arctic stream were between −5 and −10°C (Oswood
et al., 1991). Dragonflies from shallow ponds survive only if unfrozen, but have
supercooling points of about −4 to −8°C (Moore and Lee, 1991). Larvae of the
pitcher plant mosquito Wyeomyia smithii, which are killed by freezing, supercool
to −5°C (Evans and Brust, 1972). Insulation from still lower air temperatures by
ice and snow is important for the winter survival of such species (Danks, 1971a;
Farkas and Brust, 1986).
A few species remain unfrozen when surrounded by ice at much lower
temperatures. Gehrken and Sømme (1987) showed that eggs of the stonefly
Arcynopteryx compacta frozen to −29°C in water maintained a supercooled
state, apparently through dehydration.
Other aquatic insects tolerate freezing itself. Most of them belong to the Diptera,
especially the family Chironomidae. Species living in arctic ponds and lake edges
that freeze to −18 or −20°C in winter withstand comparable experimental expo-
sures (Scholander et al., 1953; Danks, 1971b; Andrews and Rigler, 1985), an
ability enhanced by the mechanical protection given by cocoons.
Typical terrestrial species that are freezing tolerant manufacture ice-nucleating
proteins to initiate freezing at relatively high subfreezing temperatures, preventing
the rapid and injurious ice formation in the body that would occur after extensive
supercooling (review by Duman, 2001). It is unlikely that such ice-nucleating
substances are needed in freezing-tolerant aquatic larvae surrounded by ice. Ter-
restrial species that are freezing tolerant even though they do not experience very
cold temperatures tend to live in very moist places where inoculative freezing is
probable (Ramløv, 2000, pp. 32–33), providing a possible parallel with aquatic
species.
Supercooling ability and protection against freezing depend partly on cryo-
protectants. Very common in terrestrial insects are antifreeze compounds of low
molecular weight, chiefly polyhydric alcohols and sugars (Storey and Storey,
1991). These solutes occur mainly in the haemolymph and have several roles,
including colligative depression of the supercooling point, binding of water (see
below) and control of osmolalities. Some compounds of low molecular weight
also stabilize and protect protein molecules or membranes directly (Storey and
Storey, 1992). Even in freezing-tolerant species they stabilize molecules (includ-
ing cell membranes) and control ice formation. Such compounds have not yet
been isolated from fully aquatic species, apparently chiefly because no one has
looked for them. Adaptive changes in membrane composition also contribute to
cold hardiness (e.g. Šlachta et al., 2002; Michaud and Denlinger, 2006): changes
in molecular species and increasing levels of unsaturation in fatty acids enhance
membrane fluidity at low temperatures by keeping membranes in the liquid crys-
talline rather than gel state.
A second class of cryoprotective compounds is antifreeze proteins (AFPs),
which depress the freezing point especially by inhibiting ice-crystal growth at the
Aquatic Insect Adaptations to Winter 7

ice–water interface (review by Duman, 2001). Various organisms have AFPs of

different kinds, but all are characterized by molecular binding to ice (Doxey et al.,
2006). AFPs have been reported in aquatic insects only from eggs of a northern
stonefly and from a northern adult gerrid (Gehrken and Sømme, 1987; Duman
et al., 2004), but undoubtedly are widespread in other species. In terrestrial
species, these compounds appear to play multiple roles in controlling or limit-
ing nucleation (including masking nucleation sites) and protecting membranes.
They also prevent inoculative freezing through the cuticle by ice on its surface
(e.g. Olsen et al., 1998 for beetles; Zettel, 2000 for springtails). Some ice-active
proteins inhibit ice recrystallization, a potentially injurious phenomenon whereby
large ice crystals grow at the expense of small ones as frozen tissues melt (Knight
and Duman, 1986; Knight et al., 1995; Wharton, 2003). Moreover, in some roles
AFPs interact with each other in the presence of smaller cryoprotectants such as
glycerol (Wang and Duman, 2005).
Involved in all of the biochemical relationships is the status of internal water.
Cold hardiness associates some of the internal water with various molecules to
make it unavailable for the freezing process (Storey and Storey, 1988; Danks,
2000; see also Block, 2002). Indeed, adaptations for cold hardiness, including an
increase of cryoprotectants, habitat selection, and adjustments of water content,
parallel those for withstanding dehydration, which share so-called cryoprotectants,
notably trehalose (Ring and Danks, 1994, 1998). Also, the need to resist or sur-
vive intracellular water loss during dehydration parallels needs when intracellular
water is lost to freeze on to extracellular ice. This pathway prevents intracellular
freezing, which is almost invariably fatal even in freezing-tolerant species.
Water loss is the key to a different method of cold hardiness, dehydration to
ice outside the body that keeps the concentration of body solutes high enough
that freezing cannot occur (Holmstrup and Zachariassen, 1996 for earthworm
egg cases; Holmstrup et al., 2002). This mechanism relies on the fact that the
vapour pressure of ice is less than that of water at the same temperature (cf.
Holmstrup and Westh, 1994), favouring dehydration in unfrozen individuals sur-
rounded by ice. Aquatic insects that supercool extensively in the presence of ice
might use a version of this strategy (see the section Key research topics).
Finally, overwintering terrestrial adults of aquatic species may require resis-
tance to dehydration. For example, diapause females of the mosquito Culex
pipiens lose water more slowly than non-diapause adults, because of their larger
size, metabolic adjustments and greater cuticular waterproofing (Benoit and
Denlinger, 2007).

Withstanding Other Seasonal Effects of Ice

A complete layer of surface ice confines the water beneath it and prevents con-
tact with the air, and also freezes out and concentrates solutes particularly in
lentic habitats. Therefore, solute concentrations can be high and oxygen levels
low in winter especially in small ponds (Daborn and Clifford, 1974). Some insect
larvae, including mayflies and chironomids, can withstand anoxia for a long time
(Nagell and Brittain, 1977; Redecker and Zebe, 1988). Some species move
8 H.V. Danks

seasonally in response to anoxia (Brittain and Nagell, 1981). High solute con-
centrations in water remaining in some ponds as surface ice finally freezes into
the substrate appear to favour dehydration and hence cold hardiness (cf. Danks,
2000, pp. 842–843).
A great challenge to lotic insects is the enormous destructive power of ice in
high flows during the melt, which adds to the considerable disturbances from
meltwater alone. Abrasive broken-up ice is driven by these flows, scouring insect
habitats. Moreover, in larger rivers, broken-up ice tends to lock together into
dams, resulting in greater flooding than would be caused by the increased flow
alone, and typically followed by a final catastrophic flow event. The presence of
ice also delays the temperature rise in spring (Prowse and Culp, 2003).
Some species tolerate the high flows (Malmqvist, 1999). Others move away
from the main channel to tributaries as break up starts (Clifford et al., 1979),
avoiding peak flow in the main channel. Stream species in habitats that are not
severely frozen may overwinter at the stream edge (Messner et al., 1983). Some
stream species overwinter in headwater lakes or spring habitats and thus escape
disturbances in the outlet channels during spring (Mendl and Müller, 1978).
Others oviposit outside the habitat. For example, eggs of the black fly Simulium
pictipes are deposited on the stream bank (Kurtak, 1974) where they avoid ice
formation and spring spates but can drop into accessible habitats upon hatch.
The staggered egg hatch of some mayflies and stoneflies, which has often been
interpreted as insurance against unpredictable water supply (Danks, 1983),
might also insure against loss of the whole population during spates (e.g. Wise,
1980).

Key Research Topics

This brief review shows that winter-related adaptations have been little studied in
aquatic insects. Therefore, I recommend research areas (Table 1.2) where more
information is necessary to allow the range of adaptations and their interactions
to be understood.

Cold hardiness and habitat conditions

A major need for many years has been to associate the winter conditions expe-
rienced by insects in nature with mechanisms of cold hardiness (e.g. Danks,
1978, 1991, 1996, 2005), rather than studying the physiological adaptations in
isolation. This conclusion applies with particular force to the topic of aquatic habitats,
where many questions revolve around habitat temperatures and their variations
in relation to the proximity of ice. Compact data loggers that were not available
in past years can now record microhabitat temperatures in detail for a whole year
(e.g. Bennett et al., 2003), but have seldom been deployed in aquatic habitats.
A key element of overwintering habitats is the insulation provided by snow,
a result of the air trapped between the snow flakes. Even in cold-temperate
regions small container, pond and creek habitats remain partly unfrozen because
Aquatic Insect Adaptations to Winter 9

Table 1.2. Key research topics about cold hardiness and winter survival in aquatic insects.

Topic Key themes Key research needs

Habitat Actual conditions experienced Simultaneous measurement of

influence the need for and microhabitat conditions and cold
nature of cold hardiness hardiness
Mechanical Substrates, burrows and shelters Assess conditions and the protection
protection may protect against ice afforded inside burrows and cocoons
Dehydration Water and cold hardiness Measure pre-winter water content,
are closely linked clarify occurrence and influence of
water loss to surrounding solute-rich
water; clarify water relationships
with inoculative freezing; study
supercooling through dehydration
when surrounded by ice
Time Characteristically slow Conduct cold-hardiness experiments
temperature changes in aquatic with suitably slow cooling rates
habitats might favour particular
cold-hardiness adaptations
Cryoprotectants Cryoprotectants of low molecular Examine a range of aquatic species
weight are very little known for polyhydric alcohols, sugars, etc.
Antifreeze The occurrence and roles of Examine a range of aquatic species
proteins AFPs are very little known for AFPs. Test potential multiple roles
including inhibition of inoculation and
recrystallization
Seasonal Habitat choice is a key feature Careful seasonal studies on movement
movements of overwintering of individual species, and in relation
to cold hardiness
Overwintering Many adults of aquatic species Integrate information about aquatic and
on land overwinter on land terrestrial stages
Winter strategies Origin of different overwintering Synthetic analysis of phylogeny, life
strategies cycles, elements of cold hardiness,
etc.

AFPs, antifreeze proteins.

the warmth of deep substrate layers ameliorates conditions beneath the insulat-
ing snow (Danks, 1971a). Precise logging of temperature data would show
whether conditions in these places are really as severe as has often been sup-
posed. Indeed, as already noted, it is insulation by snow that permits overwinter-
ing survival in plant pitchers of the freezing-susceptible mosquito Wyeomyia
smithii (Farkas and Brust, 1986).
Conditions in wet habitats are distinctive, because the high specific heat of
water and the latent heat of ice slow temperature changes, especially near the
freezing point. On one hand, this buffering would ameliorate temperature expo-
sures (as in the wet stumps monitored by Baust, 1976), but on the other it might
slow the melting of ice in freezing-tolerant organisms, allowing potentially injurious
10 H.V. Danks

recrystallizations that appear to be mitigated by certain proteins (see above).

Because the temperature of aquatic habitats changes slowly, several features of
terrestrial insects that have attracted particular recent attention are less relevant
for aquatic stages, including rapid cold hardening (see Danks, 2005) and repair
of chilling injury during milder intervals (e.g. Colinet et al., 2006).
In summary, data about winter survival by aquatic species are difficult to
interpret because we lack detailed information on habitat temperatures, including
patterns of change as well as winter extremes. Furthermore, little actual winter
sampling and experiment have been done, as opposed to later recovery of living
animals from habitats observed to have been frozen. For example, although
some dragonfly larvae survive in frozen prairie ponds (Daborn, 1971), we do not
know if they supercool or freeze (compare the discussions of Moore and Lee,
1991; Frisbie and Lee, 1997; Danks, 2007). In addition, responses related to
cold-hardiness may differ strikingly within the same terrestrial species according
to the conditions experienced (Bennett et al., 2005 for differences in winter dia-
pause and dehydration). By analogy, winter adaptations of chironomid larvae
probably differ between arctic and temperate latitudes not only in degree but
also in character; however, at present we can only generalize because information
is so limited.

Mechanical protection against ice

Adaptations of aquatic insects appear to provide protection against the mechan-

ical effects of external ice. However, both observational and experimental
research about the range of possibilities is limited.
Mechanical protection can take several forms. Overwintering in deep layers
protects animals beneath a layer of ice above them. Burrowing even to a shallow
depth would be expected to mitigate the spread of injurious ice crystals when the
substrate freezes (Danks, 2007). This protective effect would be increased in
insects that build a substrate burrow. Many aquatic species build cases or shelters
in or on the substrate for summer activity, and in winter these would serve as
mechanical barriers against ice and against mechanical damage caused by sur-
rounding substrates. Without such protection, most aquatic insects frozen in
water are killed by mechanical stresses, as shown by experiments on chironomid
larvae and other taxa (Scholander et al., 1953; Danks, 1971b).
It will be difficult to separate any specific functions of these shelters in winter
from their general roles in feeding, attachment, and protection against physical
factors, natural enemies, toxic substances and so on (Danks, 2002, 2004b), but
investigation would be worthwhile especially in the few aquatic species that build
special structures for the winter. As noted above, some chironomid larvae make
winter cocoons that differ strikingly from the summer feeding cases and are
applied tightly to the body. Larvae build and leave the cases as the water cools
or rises above temperatures close to freezing (Danks, 1971b; Danks and Jones,
1978). Experiments on case construction in relation to temperature would be
instructive. Case building enhances survival during freezing and might also be
linked with life-cycle timing such as diapause.
Aquatic Insect Adaptations to Winter 11

Cold hardiness and dehydration

Water content and water status are closely associated with cold hardiness, so
investigations into aquatic insects should test potential similarities with terres-
trial species. The internal status of water in relation to inoculative freezing
would be of particular interest. Aquatic species that supercool in the presence
of ice (Gehrken and Sømme, 1987) might do so either by resisting inoculation,
or by losing water to external ice through a mechanism similar to that used by
certain soil taxa (see above). That mechanism relies on permeable cuticles that
allow the escape of water. It also requires changes slow enough to prevent
freezing before vapour–pressure equilibrium is established. Permeable cuticles
are normal in the aquatic stages of insects (but not in most terrestrial insects)
and relatively slow changes are especially characteristic of freezing aquatic
systems, potentially allowing equilibrium to be established between inside
and outside. Of course, internal solutes must occur in sufficient concentration
that little supercooling is required as dehydration proceeds, so that inoculative
freezing does not occur. Assessments of cryoprotectant profiles, as well as
experiments on properly conditioned winter aquatic larvae cooled slowly
under suitable conditions in the presence of ice, would help to show
whether the aquatic insects known to survive in frozen ponds and streams
are cold hardy by dehydration, by resisting inoculation, or by freezing
tolerance.

Cold hardiness and time

The slow pace of autumn cooling in most aquatic habitats correlates with a num-
ber of cold-hardiness traits of aquatic insects. For example, the stonefly eggs just
referred to that supercool in the presence of ice can do so only if the cooling is
relatively slow, and they die if cooled rapidly (Gehrken, 1989). Chironomid win-
ter cocoons take some time to build, even though this activity does not start until
water temperatures fall very close to freezing (Danks, 1971b). Time is required
for habitat selection, even when movement takes place just before temperatures
become cold (Danks, 1991). Preparations for diapause need time, including the
reception of photoperiodic or other inductive signals, changes in developmental
pathways and energy storage (Danks, 1987). Many dormant insects cease
feeding but then take some time to clear the gut, which is a common source of
nucleators.
These findings suggest that ‘conditioning’ plays an important part in sea-
sonal adaptations of aquatic insects, even though temperature changes in
aquatic habitats are normally slower than in terrestrial ones. Cold hardiness is
not simply a static ‘hardened’ winter state (Danks, 2006), but a component
of development, so that ‘acclimation’ is expected. Consequently, detailed
studies of habitat conditions and life cycles in the context of the patterns of
temperature change will be required to understand the acquisition of cold
hardiness. Simple assessments of survival under winter conditions are by no
means sufficient.
12 H.V. Danks

Occurrence of low molecular weight cryoprotectants

We know very little about the cryoprotectants of aquatic insects. Table 2.1 in Lee
(1991) and table 2 in Ramløv (2000) list 28 arthropod species in which cryopro-
tectants have been characterized, but none of these species is aquatic. Glycerol
in particular is known from many other terrestrial insects (Storey and Storey,
1992). Analyses would be especially useful in aquatic species that survive freez-
ing, including some chironomid midges, as well as in species that supercool in
the presence of ice. By analogy with terrestrial insects, common compounds that
might be expected in addition to glycerol are sorbitol and other polyhydric alco-
hols, trehalose, glucose and other sugars, and a few other substances, including
proline and alanine. Indeed, glycerol was found in the Antarctic (but typically
terrestrial) chironomid Belgica antarctica, together with sorbitol, trehalose, glu-
cose and fructose (Baust, 1980). The desiccation-resistant African chironomid
Polypedilum vanderplanki contains high levels of trehalose (Watanabe et al.,
2005).

Occurrence of antifreeze proteins

Antifreeze proteins (AFPs) have so far been reported from only two aquatic species
(see Cold hardiness above), but the known properties and multiple roles of these
compounds suggest that they would be especially worthy subjects for investiga-
tion in aquatic insects. In addition to depression of the freezing point, AFPs help
to prevent inoculative freezing in terrestrial species (see above). Duman (2001)
pointed out that AFPs might be able to prevent inoculation through cuticular
pores provided the pores are very small, because AFPs are most effective when
the potential inoculating ice crystals are very small (Zachariassen and Husby,
1982) – and the way in which AFPs inhibit ice-crystal growth depends on surface
geometry. AFPs and cuticular structure might therefore combine to prevent
inoculative freezing in aquatic insects surrounded by ice. This potential rela-
tionship has not been investigated.
A role for smaller proteins as recrystallization inhibitors (compare Ramløv
et al., 1996 and Wharton, 2003 for a nematode) also cannot be ruled out. Such
substances might be characteristic of arctic chironomid species frozen into pond
or stream substrates that take many days to warm, from a few degrees below
freezing to the thaw.

Seasonal movements

Interesting anecdotes, but few specific studies, suggest how seasonal movements
help aquatic insects to survive the winter and escape spring flows and ice.
Detailed long-term sampling for individual species in carefully chosen multiple
locations in one habitat is needed to track seasonal movements. Such studies will
be especially complex because although the movements of aquatic larvae as
they grow could indeed reflect responses to winter cold or spring disturbances,
Aquatic Insect Adaptations to Winter 13

they might equally be influenced by summer warmth, food availability and pre-
dation risk. Tracking the movement of individuals fitted with miniature radio
transmitters (cf. Hayashi, 1994 for corydalid larvae) might be a way to gather
information efficiently.

Features of aquatic species overwintering on land

The fact that adults of some aquatic species overwinter on land presumably
reflects not only ancestral life cycles according to phylogeny, but also the fact that
winter conditions in aquatic habitats may be more severe than conditions in
sheltered sites beneath the snow in adjacent terrestrial ones. Parallel study of
these conditions and the cold hardiness of aquatic species sheltering there would
be valuable, but so far we have information only for a few adult mosquitoes and
gerrids. Recent molecular studies of the genetic control of cold-hardiness traits in
mosquitoes (Kim et al., 2006; Rinehart et al., 2006) are confined to the terrestrial
adult stage.
Cold hardiness and timing in these terrestrial stages resemble those of fully
terrestrial species, as might be expected. Because temperature in many terrestrial
habitats is less well buffered than in most aquatic ones, such species must man-
age energy resources well. For example, autumn (after entry into the overwinter-
ing site) and spring (before feeding or reproduction is possible) are warm enough
to consume substantial energy reserves, even though energy use is reduced dur-
ing the winter itself (Irwin and Lee, 2003 for a terrestrial species). These require-
ments are in addition to any metabolic requirements for the manufacture of
cryoprotectants. Therefore, studies of the seasonal energetics of aquatic species
that consider energy budgets for both larval feeding in water and adult over-
wintering on land would be particularly helpful.

Evolution of winter strategies

More synthetic questions can be asked once detailed information about cryopro-
tectants and other aspects of cold hardiness in aquatic insects is known. At the
moment, we do not understand how cold hardiness and winter movements
correlate with phylogeny, lifestyle, habitat, trophic status, winter temperatures,
spring conditions and so on. We do know that the marked cold hardiness of
chironomid midges relative to some other taxa stems from their original evolu-
tion in cold habitats (Brundin, 1966; Danks, 1971b), for example, but we know
relatively little about their diapause (Goddeeris, 2004), winter cocoons (Danks,
1971b), cryoprotectants or other essential components of that cold hardiness.

Climate change

Discussion about the effects of future climate change on insects has focused on
terrestrial species (for an introduction to these ideas see Hodkinson et al., 1998;
14 H.V. Danks

Danks, 2004a, 2006). Most work is concerned with the effect of increased sum-
mer temperatures but, as regional climates become less severe, higher winter
survival, higher supercooling points, less prevalent freezing tolerance and less
complex but potentially more variable cryoprotectant profiles would be expected
(Danks, 1992).
Work on aquatic insects has particular value because change alters aquatic
habitats in characteristic ways, including the extent and timing of ice cover and
spring flow. The relatively limited information on the cold hardiness of individual
species hinders biological experiments, but assessing relevant habitat conditions
in detail in different climatic regimes, including latitudinal comparisons, would be
feasible and instructive.

Conclusions

In habitats likely to freeze, most aquatic insects move to less severe sites for winter
(in the same habitat or elsewhere) and only a few, notably chironomid larvae, do
not. Other movements in lotic species help to avoid spring disturbances caused
by high flows and ice. Most species are dormant for winter, often in burrows or
constructed shelters, but a few remain active in cold but unfrozen habitats. How
species cope with cold and ice depends on where they spend the winter, but
some species can survive body freezing and others can remain unfrozen even
when surrounded by ice.
These generalizations are based on surprisingly few data compared with the
wealth of detailed evidence about the cold hardiness of terrestrial species. As a
result, needed research (summarized in Table 1.2) dwarfs the existing informa-
tion. Missing in particular is information on the physiological and biochemical
basis of cold hardiness, including cryoprotectant production. But concurrent
information about ecology and behaviour is also required, because of linkages
with habitat conditions, microhabitat choice, seasonal movements, energy use
and other requirements to survive the winter.
This conclusion suggests that two particular approaches are most valuable.
First, we need detailed biological information about individual species to inter-
pret the linkages between ecology and physiology – despite a tendency in recent
studies of aquatic insects to generalize, for example, by inventories made at only
a low level of taxonomic resolution – or by analysis only through summary sta-
tistics such as biomass and productivity. Second, a wider perspective is needed
to provide context for adaptations to cold and ice, giving due weight not only to
cold hardiness but also to such features as habitat variability, bet-hedging, food
range, dormancy and life-cycle timing (Danks, 2007).

References

Andrews, D. and Rigler, F.H. (1985) The effects of an Arctic winter on benthic inverte-
brates in the littoral zone of Char Lake, Northwest Territories. Canadian Journal of
Zoology 63, 2825–2834.
Aquatic Insect Adaptations to Winter 15

Bale, J.S. (2002) Insects and low temperatures: from molecular biology to distributions
and abundance. Philosophical Transactions of the Royal Society of London B
Biological Sciences 357, 849–862.
Baust, J.G. (1976) Temperature buffering in an arctic microhabitat. Annals of the Ento-
mological Society of America 69, 117–119.
Baust, J.G. (1980) Low temperature tolerance in an Antarctic insect: a relict adaptation?
CryoLetters 1, 360–371.
Bennett, V.A., Lee, R.E. Jr, Nauman, J.S. and Kukal, O. (2003) Selection of overwintering
microhabitats used by the Arctic woollybear caterpillar, Gynaephora groenlandica.
CryoLetters 24, 191–200.
Bennett, V.A., Sformo, T., Walters, K., Toien, O., Jeannet, K., Hochstrasser, R., Pan, Q.,
Serianni, A.S., Barnes, B.M. and Duman, J.G. (2005) Comparative overwintering
physiology of Alaska and Indiana populations of the beetle Cucujus clavipes (Fabricius):
roles of antifreeze proteins, polyols, dehydration and diapause. Journal of Experimental
Biology 208, 4467–4477.
Benoit, J.B. and Denlinger, D.L. (2007) Suppression of water loss during adult diapause
in the northern house mosquito, Culex pipiens. Journal of Experimental Biology
210, 217–226.
Berté, S.B. and Pritchard, G. (1983) The life history of Glyphopsyche irrorata (Trichoptera,
Limnephilidae): a caddisfly that overwinters as an adult. Holarctic Ecology 6, 69–73.
Block, W. (2002) Interactions of water, ice nucleators and desiccation in invertebrate cold
survival. European Journal of Entomology 99, 259–266.
Bouchard, R.W. Jr, Carillo, M.A., Kells, S.A. and Ferrington, L.C. Jr. (2006) Freeze tolerance
in larvae of the winter-active Diamesa mendotae Muttkowski (Diptera: Chironomi-
dae): a contrast to adult strategy for survival at low temperatures. Hydrobiologia 568,
403–416.
Brittain, J.E. and Nagell, B. (1981) Overwintering at low oxygen concentrations in the
mayfly Leptophlebia vespertina. Oikos 36, 45–50.
Brundin, L. (1966) Transantarctic relationships and their significance as evidenced by
chironomid midges with a monograph of the subfamilies Podonominae and Aphro-
teniinae and the austral Heptagyiae. Kongliga svenska Vetenskaps-akademiens
Handlingar 11, 7–472.
Clifford, H.F. (1969) Limnological features of a northern brown-water stream, with special
reference to the life-histories of the aquatic insects. American Midland Naturalist 82,
578–597.
Clifford, H.F., Hamilton, H. and Killins, B.A. (1979) Biology of the mayfly Leptophlebia
cupida (Say) (Ephemeroptera: Leptophlebiidae). Canadian Journal of Zoology 57,
1026–1045.
Colinet, H., Renault, D., Hance, T. and Vernon, P. (2006) The impact of fluctuating ther-
mal regimes on the survival of a cold-exposed parasitic wasp, Aphidius colemani.
Physiological Entomology 31, 234–240.
Copeland, R.S. and Craig, G.B. Jr (1990) Cold hardiness of tree-hole mosquitoes in the Great
Lakes region of the United States. Canadian Journal of Zoology 68, 1307–1314.
Daborn, G.R. (1971) Survival and mortality of coenagrionid nymphs (Odonata: Zygoptera)
from the ice of an aestival pond. Canadian Journal of Zoology 49, 569–571.
Daborn, G.R. and Clifford, H.F. (1974) Physical and chemical features of an aestival pond
in western Canada. Hydrobiologia 44, 43–59.
Danks, H.V. (1971a) Overwintering of some north temperate and arctic Chironomidae. I.
The winter environment. Canadian Entomologist 103, 589–604.
Danks, H.V. (1971b) Overwintering of some north temperate and arctic Chironomidae.
II. Chironomid biology. Canadian Entomologist 103, 1875–1910.
16 H.V. Danks

Danks, H.V. (1978) Modes of seasonal adaptation in the insects. I. Winter survival. Canadian
Entomologist 110, 1167–1205.
Danks, H.V. (1983) Extreme individuals in natural populations. Bulletin of the Entomo-
logical Society of America 29, 41–46.
Danks, H.V. (1987) Insect Dormancy: an Ecological Perspective. Biological Survey of
Canada (Terrestrial Arthropods), Ottawa.
Danks, H.V. (1991) Winter habitats and ecological adaptations for winter survival. In:
Lee, R.E. Jr and Denlinger, D.L. (eds) Insects at Low Temperature. Chapman and
Hall, New York, pp. 231–259.
Danks, H.V. (1992) Arctic insects as indicators of environmental change. Arctic 45, 159–
166.
Danks, H.V. (1996) The wider integration of studies on insect cold-hardiness. European
Journal of Entomology 93, 383–403.
Danks, H.V. (2000) Dehydration in dormant insects. Journal of Insect Physiology 46,
837–852.
Danks, H.V. (2002) Modification of adverse conditions by insects. Oikos 99, 10–24.
Danks, H.V. (2004a) Seasonal adaptations in arctic insects. Integrative and Comparative
Biology 44, 85–94.
Danks, H.V. (2004b) The roles of insect cocoons in cold conditions. European Journal of
Entomology 101, 433–437.
Danks, H.V. (2005) Key themes in the study of seasonal adaptations in insects I. Patterns
of cold hardiness. Applied Entomology and Zoology 40, 199–211.
Danks, H.V. (2006) Insect adaptations to cold and changing environments. Canadian
Entomologist 138, 1–23.
Danks, H.V. (2007) How aquatic insects live in cold climates. Canadian Entomologist
139, 443–471.
Danks, H.V. and Jones, J.W. (1978) Further observations on winter cocoons in Chirono-
midae (Diptera). Canadian Entomologist 110, 667–669.
Dillon, P.M. (1985) Chironomid larval size and case presence influence capture success
achieved by dragonfly larvae. Freshwater Invertebrate Biology 4, 22–29.
Doxey, A.C., Yaish, M.W., Griffith, M. and McConkey, B.J. (2006) Ordered surface carbons
distinguish antifreeze proteins and their ice-binding regions. Nature Biotechnology
24, 852–855.
Duman, J.G. (2001) Antifreeze and ice nucleator proteins in terrestrial arthropods. Annual
Review of Physiology 63, 327–357.
Duman, J.G., Bennett, V., Sformo, T., Hochstrasser, R. and Barnes, B.M. (2004) Anti-
freeze proteins in Alaskan insects and spiders. Journal of Insect Physiology 50,
259–266.
Evans, K.W. and Brust, R.A. (1972) Induction and termination of diapause in Wyeomyia
smithii (Diptera: Culicidae), and larval survival studies at low and subzero tempera-
tures. Canadian Entomologist 104, 1937–1950.
Farkas, M.J. and Brust, R.A. (1986) Phenology of the mosquito Wyeomyia smithii in
Manitoba and Ontario. Canadian Journal of Zoology 64, 285–290.
Frisbie, M.P. and Lee, R.E. Jr. (1997) Inoculative freezing and the problem of winter sur-
vival for freshwater macroinvertebrates. Journal of the North American Benthological
Society 16, 635–650.
Gehrken, U. (1989) Diapause termination in eggs of the stonefly Arcynopteryx compacta
(McLachland) in relation to dehydration and cold hardiness. Journal of Insect Physi-
ology 35, 377–385.
Gehrken, U. and Sømme, L. (1987) Increased cold hardiness in eggs of Arcynopteryx
compacta (Plecoptera) by dehydration. Journal of Insect Physiology 33, 987–991.
Aquatic Insect Adaptations to Winter 17

Gibbs, K.E. (1979) Ovoviviparity and nymphal seasonal movements of Callibaetis spp.
(Ephemeroptera: Baetidae) in a pond in southwestern Quebec. Canadian Entomolo-
gist 111, 927–931.
Goddeeris, B. (2004) Diapause and ecological strategies in Chironomidae (Diptera).
Journal of Limnology 63 Supplement 1, 92.
Hauer, F.R. and Stanford, J.A. (1982) Bionomics of Dicosmoecus gilvipes (Trichoptera:
Limnephilidae) in a large western montane river. American Midland Naturalist 108,
81–87.
Hayashi, F. (1994) Life-history patterns in 15 populations of Protohermes (Megaloptera:
Corydalidae): Effects of prey size and temperature. In: Danks, H.V. (ed.) Insect Life-
cycle Polymorphism: Theory, Evolution and Ecological Consequences for Seasonality
and Diapause Control. Kluwer Academic Publishers, Dordrecht, pp. 227–243.
Hodkinson, I.D., Webb, N.R., Bale, J.S., Block, W., Coulson, S.J. and Strathdee, A.T. (1998)
Global change and arctic ecosystems: conclusions and predictions from experiments
with terrestrial invertebrates on Spitsbergen. Arctic and Alpine Research 30, 306–313.
Holmstrup, M. and Westh, P. (1994) Dehydration of earthworm cocoons exposed to cold:
a novel cold hardiness mechanism. Journal of Comparative Physiology B: Biochemical
Systemic and Environmental Physiology 164, 312–315.
Holmstrup, M. and Zachariassen, K.E. (1996) Physiology of cold hardiness in earthworms.
Comparative Biochemistry and Physiology A: Comparative Physiology 115, 91–101.
Holmstrup, M., Bayley, M. and Ramløv, H. (2002) Supercool or dehydrate? An experi-
mental analysis of overwintering strategies in small permeable arctic invertebrates.
Proceedings of the National Academy of Sciences of the United States of America 99,
5716–5720.
Hudson, J.E. (1978) Overwintering sites and ovarian development of some mosquitoes
in central Alberta, Canada. Mosquito News 38, 570–579.
Irons, J.G. III, Ray, S.R., Miller, L.K. and Oswood, M.W. (1989) Spatial and seasonal patterns
of streambed water temperatures in an Alaskan subarctic stream. In: Woessner, W.W.
and Potts, D.F. (eds) Proceedings of the Symposium on Headwaters Hydrology, June
1989, Missoula, Montana. American Water Resources Association, Bethesda, Mary-
land, pp. 381–390.
Irons, J.G. III, Miller, L.K. and Oswood, M.W. (1993) Ecological adaptations of aquatic
macroinvertebrates to overwintering in interior Alaska (USA.) subarctic streams.
Canadian Journal of Zoology 71, 98–108.
Irwin, J.T. and Lee, R.E. Jr. (2003) Cold winter microenvironments conserve energy and
improve overwintering survival and potential fecundity of the goldenrod gall fly,
Eurosta solidaginis. Oikos 100, 71–78.
Kim, M., Robich, R.M., Rinehart, J.P. and Denlinger, D.L. (2006) Upregulation of two
actin genes and redistribution of actin during diapause and cold stress in the northern
house mosquito, Culex pipiens. Journal of Insect Physiology 52, 1226–1233.
Knight, C.A. and Duman, J.G. (1986) Inhibition of recrystallization of ice by insect thermal
hysteresis proteins: a possible cryoprotective role. Cryobiology 23, 256–262.
Knight, C.A., Wen, D. and Lauersen, R.A. (1995) Nonequilibrium antifreeze peptides and
the recrystallization of ice. Cryobiology 32, 23–34.
Kurtak, D. (1974) Overwintering of Simulium pictipes Hagen (Diptera: Simuliidae) as
eggs. Journal of Medical Entomology 11, 383–384.
Lee, R.E. Jr. (1991) Principles of insect low temperature tolerance. In: Lee, R.E. Jr and
Denlinger, D.L. (eds) Insects at Low Temperature. Chapman and Hall, New York, pp.
17–46.
Lencioni, V. (2004) Survival strategies of freshwater insects in cold environments. Journal
of Limnology 63 Supplement 1, 45–55.
18 H.V. Danks

Malmqvist, B. (1999) Life history of Metacnephia lyra, a blackfly highly characteristic of

large Swedish river rapids at the time of maximum discharge (Diptera: Simuliidae).
Aquatic Insects 21, 89–98.
Mendl, H. and Müller, K. (1978) The colonization cycle of Amphinemura standfussi Ris
(Ins.: Plecoptera) in the Abisko area. Hydrobiologia 60, 109–111.
Messner, B., Groth, I. and Taschenberger, D. (1983) Zum jahreszeitlichen Wanderverhalten
der Grundwanze Aphelocheirus aestivalis. Zoologische Jahrbücher, Abteilung für
Systematik Oekologie und Geographie der Tiere 110, 323–331.
Michaud, M.R. and Denlinger, D.L. (2006) Oleic acid is elevated in cell membranes during
rapid cold-hardening and pupal diapause in the flesh fly, Sarcophaga crassipalpis.
Journal of Insect Physiology 52, 1073–1082.
Moore, M.V. and Lee, R.E., Jr (1991) Surviving the big chill: overwintering strategies of
aquatic and terrestrial insects. American Entomologist 37, 111–118.
Nagell, B. and Brittain, J.E. (1977) Winter anoxia—a general feature of ponds in cold
temperate regions. Internationale Revue der Gesamten Hydrobiologie 62, 821–824.
Olsen, T.M., Sass, S.J., Li, N. and Duman, J.G. (1998) Factors contributing to seasonal
increases in inoculative freezing resistance in overwintering fire-colored beetle larvae
Dendroides canadensis (Pyrochroidae). Journal of Experimental Biology 201, 1585–
1594.
Olsson, T.I. (1981) Overwintering of benthic macroinvertebrates in ice and frozen sedi-
ment in a North Swedish river. Holarctic Ecology 4, 161–166.
Olsson, T.I. (1982) Overwintering sites and freezing tolerance of benthic invertebrates in
a north Swedish river. CryoLetters 3, 297–298.
Olsson, T.I. (1983) Seasonal variation in the lateral distribution of mayfly nymphs in a
boreal river. Holarctic Ecology 6, 333–339.
Oswood, M.W., Miller, L.K. and Irons, J.G. III. (1991) Overwintering of freshwater benthic
macroinvertebrates. In: Lee, R.E. Jr and Denlinger, D.L. (eds) Insects at Low
Temperature. Chapman and Hall, New York, pp. 360–375.
Pajunen, V.I. and Jansson, A. (1969) Dispersal of the rock pool corixids Arctocorisa carinata
(Sahlb.) and Callicorixa producta (Reut.) (Heteroptera: Corixidae). Annales Zoologici
Fennici 6, 391–427.
Paterson, C.G. (1971) Overwintering ecology of the aquatic fauna associated with the
pitcher plant Sarracenia purpurea L. Canadian Journal of Zoology 49, 1455–1459.
Prowse, T.D. and Culp, J.M. (2003) Ice breakup: a neglected factor in river ecology.
Canadian Journal of Civil Engineering 30, 128–144.
Ramløv, H. (2000) Aspects of natural cold tolerance in ectothermic animals. Human
Reproduction 15, 26–46.
Ramløv, H., Wharton, D.A. and Wilson, P.W. (1996) Recrystallization in a freezing tolerant
Antarctic nematode, Panagrolaimus davidi, and an alpine weta, Hemideina maori
(Orthoptera; Stenopelmatidae). Cryobiology 33, 607–613.
Redecker, B. and Zebe, E. (1988) Anaerobic metabolism in aquatic insect larvae: studies
on Chironomus thummi and Culex pipiens. Journal of Comparative Physiology B
158, 307–315.
Rinehart, J.P., Robich, R.M. and Denlinger, D.L. (2006) Enhanced cold and desiccation
tolerance in diapausing adults of Culex pipiens, and a role for hsp70 in response to
cold shock but not as a component of the diapause program. Journal of Medical
Entomology 43, 713–722.
Ring, R.A. and Danks, H.V. (1994) Desiccation and cryoprotection: overlapping adaptations.
CryoLetters 15, 181–190.
Ring, R.A. and Danks, H.V. (1998) The role of trehalose in cold-hardiness and desiccation.
CryoLetters 19, 275–282.
Aquatic Insect Adaptations to Winter 19

Sawchyn, W.W. and Gillott, C. (1974) Life history of three species of Lestes (Odonata:
Zygoptera) in Saskatchewan. Canadian Entomologist 106, 1283–1293.
Scholander, P.F., Flagg, W., Hock, R.J. and Irving, L. (1953) Studies on the physiology of
frozen plants and animals in the arctic. Journal of Cellular and Comparative Physiology
42 Supplement 1, 56 pp.
Shen, H.T. (2003) Research on river ice processes: progress and missing links. Journal of
Cold Regions Engineering 17, 135–142.
Šlachta, M., Berkova, P, Vambera, J. and Košt’ál, V. (2002) Physiology of cold-acclimation
in non-diapausing adults of Pyrrhocoris apterus (Heteroptera). European Journal of
Entomology 99, 181–187.
Spence, J.R. and Andersen, N.M. (1994) Biology of water striders: interactions between
systematics and ecology. Annual Review of Entomology 39, 101–128.
Storey, K.B. and Storey, J.M. (1988) Freeze tolerance in animals. Physiological Reviews
68, 27–84.
Storey, K.B. and Storey, J.M. (1991) Biochemistry of cryoprotectants. In: Lee, R.E. Jr and
Denlinger, D.L. (eds) Insects at Low Temperature. Chapman and Hall, New York, pp.
64–93.
Storey, K.B. and Storey, J.M. (1992) Biochemical adaptations for winter survival in
insects. Advances in Low-temperature Biology 1, 101–140. JAI Press, London.
Vali, G. (1995) Principles of ice nucleation. In: Lee R.E. Jr, Warren, G.J. and Gusta, L.V.
(eds) Biological Ice Nucleation and its Applications. American Phytopathological
Society, St Paul, Minnesota, pp. 1–28.
Van de Bund, W.J. and Groenenkijk, D. (1994) Seasonal dynamics and burrowing of
littoral chironomid larvae in relation to competition and predation. Archiv für
Hydrobiologie 132, 213–225.
Wang, L. and Duman, J.G. (2005) Antifreeze proteins of the beetle Dendroides canadensis
enhance one another’s activities. Biochemistry 44, 10305–10312.
Watanabe, M., Kikawada, T., Fujita, A. and Okuda, T. (2005) Induction of anhydrobiosis
in fat body tissue from an insect. Journal of Insect Physiology 51, 727–731.
Wharton, D.A. (2003) The environmental physiology of Antarctic terrestrial nematodes:
a review. Journal of Comparative Physiology B 173, 621–628.
Wiggins, G.B. (1973) A contribution to the biology of caddisflies (Trichoptera) in temporary
pools. Life Sciences Contributions, Royal Ontario Museum 88, 1–28.
Wiggins, G.B. and Winchester, N.N. (1984) A remarkable new caddisfly genus from
northwestern North America (Trichoptera, Limnephilidae, Limnephilinae). Canadian
Journal of Zoology 62, 1853–1858.
Williams, D.D. (1997) Temporary ponds and their invertebrate communities. Aquatic
Conservation: Marine and Freshwater Ecosystems 7, 105–117.
Wise, E.J. (1980) Seasonal distribution and life histories of Ephemeroptera in a Northumbrian
river. Freshwater Biology 10, 101–111.
Zachariassen, K.E. and Husby, J.A. (1982) Antifreeze effect of thermal hysteresis agents
protects highly supercooled insects. Nature 298, 865–867.
Zettel, J. (2000) Alpine Collembola: adaptations and strategies for survival in harsh envi-
ronments. Zoology 102, 73–89.
2 Saline-water Insects: Ecology,
Physiology and Evolution
TIMOTHY J. BRADLEY
Department of Ecology and Evolutionary Biology, University of California,
Irvine, CA 92697-2525, USA
E-mail: tbradley@[Link]

Abstract
This chapter presents an overview of the numerous genera of insects capable of surviving
in saline water, including information on their distribution, ecology and physiology.
Detailed description is provided of the two families of insects that have been most
intensively studied, namely the mosquitoes (Culicidae) and the brine flies (Ephydridae).
The life histories of these groups are well known and the mechanisms of salt and water
balance have been examined in considerable detail. Evidence is presented that the
primitive condition in mosquitoes is restriction to freshwater. Invasion of saline waters has
occurred frequently (at least six times) in the family Culicidae, but the number of species
capable of surviving in saline water is small (about 5% of all known mosquito species).
The physiological modifications supporting these invasions are discussed.

Introduction

My goal in this chapter is to review interesting aspects of the ecology, physiology

and evolution of insects residing in saline waters. Salinity, and more specifically
certain ions such as sodium and sulphate, present unusual challenges for insects.
Insects are an extremely successful clade; the precise number of species is
unknown, but most estimates are for numbers well over 10 million. Insects arose
in the terrestrial environment and the current thinking is that ancestral arthro-
pods, similar and related to branchiopods, moved on to land. As such, insects
arose in an environment in which plants and detritus were the principal sources
of food. Some species eventually evolved forms that could take advantage of
freshwater habitats, particularly in the larval stages. The evolutionary advan-
tages of flight for mating and dispersal led most species to retain a terrestrial or
aerial adult stage. Whether the food is plant and detrital materials, or the habitat
is freshwater, the common feature is a relative paucity of sodium. Most plants
do not require sodium and can be grown in environments in which sodium
is entirely absent. Animals, however, require sodium for some physiological
 CAB International 2008. Aquatic Insects: Challenges to Populations
20 (eds J. Lancaster and R.A. Briers)
Saline-water Insects 21

functions, for example, nervous activity. This creates a substantial physiological

problem for freshwater and terrestrial animals. The insects got around this prob-
lem by reducing the amount of sodium in the body to an essential minimum.
Indeed, in some insect species, sodium is restricted to tiny compartments sur-
rounding the nerves. Other osmotically active osmolytes occur in the blood of
many insects.
Having evolved as a group supremely adapted to terrestrial and freshwater
habitats, the insects have neural and excretory systems geared to obtaining and
retaining sodium. As will become clear in text to follow, evolving excretory sys-
tems designed not to obtain sodium, but rather to rid the body of sodium, has
proven challenging for insects. None the less, some insects have managed to
exploit sodium-rich habitats.
Insects in saline waters are subject not only to osmotically stressful condi-
tions, but also to high levels of sodium. Roughly 70% of the Earth’s surface is
covered by saline water, representing numerous potential habitats for insects. In
addition to the oceans, there are saline marshes, large salt lakes, salt ponds and
desert pools. Most of these habitats have been invaded and exploited by insects,
with the exception of the open oceans. Given the incredible ecological success of
insects in most environments on the planet, and indeed their success in places
where other groups cannot survive, it is remarkable that the oceans remain
largely insect-free.

Marine Insects

Very few insects are found in the open ocean and those few are generally associ-
ated with the benthos near the shore. Some insect larvae inhabit coral outcrops,
feeding on algae and bacteria on these substrates. The insects with a worldwide
oceanic distribution are the marine water striders. These semi-aquatic insects do
not live in the water, but walk instead on the ocean surface.
I will describe briefly the most abundant and widely studied groups of saline-
tolerant insects. A short review cannot do justice to all of these insect groups and
the topic has been reviewed thoroughly by Cheng (1976). Although many papers
have been published since that volume appeared, it remains the best source of
information on saline-tolerant insects. I will not attempt to provide the level of
review contained in a lengthy book but, instead, I will mention the major insect
groups that are saline tolerant in order to illustrate that this capacity is not
taxonomically limited. I discuss in greater depth two families of insects, the salt-
tolerant mosquitoes and the brine flies, that have been the subject of intense
study, particularly to determine the physiological mechanisms that permit them
to survive and thrive in saline waters.

Collembola

Collembolans are among the most primitive of the insect orders with a fossil
record going back 400 million years. The group is thought to be terrestrial in
22 T.J. Bradley

origin, although they are most abundant and diverse in moist habitats. Collembolans
tend to accumulate on water surfaces, being largely non-wettable, and they can
walk and hop across bodies of water with ease. Collembolans are generally either
predacious on other invertebrates and/or feed by scavenging. They are found in
all types of substrates, including sea wrack, sandy beaches and rocky shores.
Collembolans can survive immersion indefinitely, suggesting that they can obtain
adequate oxygen across the integument. They have survived in the laboratory
for 14 days with access only to seawater, indicating that they do not require
access to freshwater (Joose, 1976).

Thysanura

Marine thysanurans are another abundant and highly successful group of aptery-
gotes able to exploit the ocean shore. These insects are of course ametabolous.
All life stages feed on algae on rocky or sandy substrates (Joose, 1976). Weig-
mann (1973) found that some thysanuran species could tolerate a wide range
of salinities – they were osmoregulators in dilute media, but osmoconformers
in more concentrated media. The physiological mechanisms associated
with this type of osmoregulatory behaviour are described for mosquito larvae
below.

Trichoptera

Caddisflies, at least of the species Philanisus plebeius, may be among the most
fully marine of all insects (Leader, 1972, 1976). The larvae of several species
have been investigated with regard to their salinity tolerance and osmoregulatory
capacities. Limnephilus affinis, which occurs in marine marshes in Europe, can
survive in salinities up to 75% seawater. Sutcliffe (1961) found that the larvae
osmoregulate in dilute media, but are osmoconformers in more concentrated
media. By contrast, Philanisus plebius, a species found in New Zealand, was a
strict osmoregulator and able to survive indefinitely in full-strength seawater. Phi-
lanisus plebius spends its entire larval period immersed in seawater, feeds on
coralline algae, and the larval silk cases are decorated with sand grains or the
coralline algae on which it feeds (Leader, 1976).

Corixidae

Water boatmen can be abundant in coastal marshes, rocky tide pools and
coastal marshes. Corixids are often thought to be herbivorous, but the forms in
saline waters feed on nematodes, aquatic crustaceans, chironomids and other
insects. Several species can osmoregulate in varying osmolarities, including
full-strength seawater, and Trichocorixa is the only strictly marine genus (Scudder,
1976).
Saline-water Insects 23

Ephydridae

Shore flies inhabit saline lakes and the ocean shore around the world. In the
marine habitat, they tend to be excluded from open oceans, but are abundant in
sea wrack, tidepools and marshy sites (Simpson, 1976). In saline lakes, both
larvae and adults can be found on sandy and rocky substrates, feeding on algae
and bacteria (Bradley and Herbst, 1994). The section below on brine flies
describes their ecology and physiology in more detail.

Gerridae

The insects with the widest distribution on the oceans are the water striders in the
genus Halobates. Of the 1700 species of water striders, 10% are marine. Both
the immature and adult stages of marine water striders live atop the ocean waters
and, in this sense, their adaptations are similar to those of terrestrial insects. The
species that are strictly marine in distribution tend to be wingless, even as adults.
Cheng (1976) reviews her many articles examining the ecology and taxonomy
of marine water striders.

Ceratopogonidae and Chironomidae

Biting midges (Ceratopogonidae) are a bane to humans and wildlife. Salt marshes
and estuaries can be sites of intense production of biting midges and their pres-
ence makes many coastal areas distinctly unpleasant for humans. The larvae are
fully aquatic and feed on detritus, bacteria and algae. The larvae are fully saline-
tolerant, at least up to full strength seawater, and appear to be osmoregulators,
but they are never found in the open ocean (Linley, 1976).
Some species of chironomids (non-biting midges) are also fully saline-
tolerant. The larvae of many species feed on detritus, algae and bacteria, but
some of the saline-water species are carnivorous, a trait attributed to the increased
availability of prey in productive saline waters (Hashimoto, 1976). The larvae
are osmoregulators and may be the closest thing to a truly marine insect. They
have been found on the substrate well below the lowest level of the lowest tide,
in some cases in water several metres deep (Hashimoto, 1976).

Other coastal insects

Shore bugs (Saldidae) are abundant on most marine coasts. Although generally
less abundant than shore bugs, the beetles (Coleoptera) found in marine coastal
habitats are more speciose. Both shore bugs and beetles can withstand pro-
longed submersion, yet, in terms of morphological and physiological adapta-
tions, they are similar to strictly terrestrial species. Although abundant and
ecologically important, they have few adaptations to a saline habitat (Cheng,
1976). Many insect species are found in salt marshes but, as they do not have
24 T.J. Bradley

aquatic life stages, these insects are more accurately considered to be terrestrial
than aquatic saline-adapted.

Two Intensely Studied Insects

As the above section suggests, a variety of insects have life stages that are saline-
tolerant. Many species are found in coastal marshes and desert pools where the
salinity can often exceed that in the open ocean. When insects occur in the
ocean, they are generally associated with the intertidal region, with the exception
of chironomids that can occur in deep waters, Trichoptera that occur just off-
shore in benthic algal mats, and water striders that occur on the ocean surface.
Two insects, saline-tolerant mosquitoes and brine flies, have been studied in
considerable detail. The former because of their nuisance and medical impor-
tance, the latter because of their ecological significance. I will describe what is
known about these two insect groups in more detail, thus providing a better under-
standing of the processes employed by insects to deal with saline waters.

Brine flies

Adult brine flies feed largely on algae, including filamentous algae and diatoms.
In many species, the adults are rather long lived and algal feeding is required to
support repeated reproductive bouts.
Adult brine flies have a waxy layer on their exterior cuticle and, when they
walk along the substrate feeding on algae or ovipositing underwater, they are sur-
rounded by a thin bubble of air. This serves as a source of oxygen and, judging
from the length of time they can remain underwater, also serves as a surface for
the exchange of respiratory gases. Females lay eggs that are more dense than the
surrounding water in rock interstices or in algal mats. The larvae hatch and imme-
diately begin feeding on algae. Bradley and Herbst (1994) demonstrated that
several species of algae could support complete larval development. Concretions
of calcium carbonate in the Malpighian tubules make the larvae more dense than
the medium and allow them to stay on the benthos. The larvae also have hooks
on the prolegs, which assist in holding on to the bottom and in locomotion.
Development in ephydrids involves three larval instars. At the end of the last
larval instar, the concretions from the lime glands are evacuated through the
anus and the larva metamorphoses into a pupa. The pupa, which must remain
submerged to survive, is now less dense than the medium and a holdfast near
the anus is used for attachment to twigs, rocks or algal mats. The adults emerge
directly from the pupal case and, being less dense than water, they float to the
surface and emerge from the water.

Physiology

The larvae of ephydrids are astounding for their capacity to survive in highly
saline waters. Brine flies from the Great Salt in Utah, thrive by the billions in
Saline-water Insects 25

NaCl-rich water with a salinity up to 10 times that of the open ocean. At Mono
Lake in California, USA, the larvae of Ephydra hians also thrive in water three
times the concentration of the ocean, but these waters are rich in sulphate and
bicarbonate, with a pH of 10!
Given the extraordinary salt concentrations that the larvae experience, the
first issue to address is the size of the gradients across the cuticle. All the species
of brine fly examined to date are osmoregulators (Simpson, 1976). Table 2.1
shows the concentration of ions measured in the larval haemolymph of E. hians
reared in water from Mono Lake, as well as the ionic concentrations in the exter-
nal water (Shimizu and Bradley, 1994). The haemolymph had a total osmotic
concentration of about 300 mOsm under these circumstances; the osmotic con-
centration of the lake was well over 1500 mOsm. It follows that the larvae must
lose water across the cuticle due to the large osmotic forces. Studies in our labo-
ratory have shown the cuticle to be permeable, although a quantitative estimate
is not available (personal observation).
To replace the water lost through osmosis, the larvae drink the external
medium. This replaces the volume they have lost, but poses the additional prob-
lem of forcing the larvae to rid themselves of the ingested ions. The very high
concentration of sodium, carbonate and sulphate concentrations in Mono Lake
water obviously pose serious problems in terms of ionic and pH homeostasis.
Figure 2.1 illustrates the internal organs associated with the gut and with
excretion in the larvae of Ephydra hians. The fluids and food that are ingested
pass into the midgut. In all insects examined to date, the fluid in the midgut rapidly
comes into osmotic equilibrium with the haemolymph (Bradley, 2002). In E. hians,
the ions in the midgut are transported into the haemolymph and water follows
osmotically. The Malpighian tubules produce the urine by transporting ions from
the haemolymph into the tubule lumen, and water follows this transport osmotically
also. The fluid in the Malpighian tubules is, however, isosmotic with haemolymph
and therefore does not contribute directly to osmoregulation.
In many insects, crystals precipitate in the lumen of the Malpighian tubules.
This is due to shifts in the pH and to the transport of ions, which at low concen-
trations produce insoluble compounds (e.g. divalent cations and carbonate, or
potassium and urate). In E. hians, crystals form in the Malpighian tubules and are
accumulated in an expanded region of the Malpighian tubules termed the lime

Table 2.1. Ionic concentrations (mm/l) in the larval haemolymph of Ephydra hians
and in the Mono Lake water in which they were reared.

Concentration in larval Concentration in Mono

Ion haemolymph Lake water

Sodium 135.7 + 1.0 1224

Potassium 6.9 + 0.4 18
Calcium 5.6 + 0.1 <1
Magnesium 13.2 + 0.5 <1
Chloride 120.0 + 1.4 627
Sulphate 0.6 + 0.1 151
26 T.J. Bradley

Midgut
Unmodified tubule
(posterior)
Rectum
Hindgut

Ureter

Proximal tubule of
lime gland
Storage region of lime gland

Distal tubule of lime gland

Fig. 2.1. Internal organs of the brine fly Ephydra hians. The insect has two pairs of
Malpighian tubules, one pair is modified into the lime glands. (Redrawn from Herbst
and Bradley, 1989).

Fig. 2.2. Photograph of two, third-instar larvae of Ephydra hians. Note the prominent,
white lime glands (arrows) visible through the cuticle of the larvae. – equals 1 mm
(from Herbst and Bradley, 1989).

gland. The accumulations of concretions are large enough to form a large white
sack in the larva that is visible through the cuticle (Fig. 2.2), and the crystals are
composed almost entirely of calcium carbonate (Herbst and Bradley, 1989).
The lime crystals serve as ballast, permitting the larva to remain submerged in highly
Saline-water Insects 27

saline water, which is more dense than most biological materials. The precipitation
of carbonate and bicarbonate contributes to pH regulation in the larvae.
Sutcliffe (1960) showed that brine flies produce a urine more concentrated
than the medium in which they reside. Shimizu and Bradley (1994) demonstrated
that the rectum was not the site of hyperosmotic urine formation in E. hians. The
rectum has a very thin epithelium and the cells do not exhibit the ultrastructural
specializations (highly folded membranes and numerous mitochondria) expected
in tissues active in ion transport. Instead, the hindgut contains a colon with thick
walls and active cells. The authors found that the entire colon was capable of pro-
ducing a hyperosmotic fluid. The colon was also the site of active transport of
sulphate, but this capacity was restricted to the anterior half of the organ.
Survival in highly saline media requires, therefore, transport functions in the
Malpighian tubules, particularly for the production of dense concretions. The
production of a hyperosmotic and ion-rich urine for osmoregulation is a function
of the hindgut and, more specifically, the cells in the colon. In Ephydra, hyperos-
motic fluid secretion occurs throughout, while sulphate excretion is restricted to
the anterior end of the organ (Shimizu and Bradley, 1994).

Saline-tolerant mosquitoes

Saline waters are often extremely biologically productive. Salt marshes along the
coast receive nutrients from the sea and also in runoff from the land. In particu-
lar, estuaries and coastal swamps are rich in nutrients, benefit from stirring and
oxygenation through the shifting tides and exhibit high primary productivity.
Although the constantly shifting osmotic concentration can be difficult for ani-
mals, these habitats can be extraordinarily productive for mosquitoes. Mosqui-
toes are unquestionably the most intensively studied of all saline-tolerant insects
with regard to their ecology, physiology and evolutionary history. The reasons
for this are obvious. Mosquitoes are not only a nuisance to humans, but they also
spread a variety of diseases, including the most deadly of all infectious diseases,
malaria. Several researchers have examined the ecological distribution of mos-
quitoes and found that about 5% of all species known have larvae that are capa-
ble of surviving in hyperosmotic media, i.e. media more concentrated than about
30% seawater. The degree of salinity that mosquito larvae can withstand varies
substantially among species. The reasons for this will be discussed in the section
below dealing with physiology. Many species from many genera, however, live
in marine waters in rocky tide pools, in salt marshes, which are often much more
concentrated than seawater, and in desert pools, which can reach very high
salinities and may possess ionic ratios highly different from seawater.

Ecology

Mosquitoes fall broadly into two groups based on their oviposition behaviour. In the
first group, which contains the most genera, females lay their eggs directly on the
water surface. Species in the genera Culex and Culiseta form the eggs into an egg
28 T.J. Bradley

raft, whereas species in other genera, such as Anopheles, lay the eggs singly. In all
cases, the eggs float on the surface until hatching. As the females come into intimate
contact with the aquatic medium, it is clear that they test the medium (perhaps by
taste) to determine if oviposition should proceed. It has been demonstrated repeat-
edly, and in many species, that females show marked preferences in oviposition
sites with regard to salinity, pH, presence of microorganisms, etc. (O’Meara, 1976).
Although the larvae can survive in both fresh and saline waters, salinity preference
may be adaptive due to the reduction in competition and predation.
Mosquitoes species in the genera Aedes and Opifex are referred to as floodwater
mosquitoes. These species lay their eggs on the soil adjacent to a body of water, or
even at sites that currently contain no water, but will be flooded at some time in the
future. The eggs remain in a form of diapause until covered by rising water. This may
occur due to rain in inland pools, or due to storm surge or high tide in marine habi-
tats. Oviposition site selection involves selecting a site that will flood during the next
rise in water. Both freshwater and saline-water species have an uncanny ability to
predict sites that will flood with the next rise in waters, but how the females select
these sites is poorly understood. What is clear is that females partial to saline sites
often must lay their eggs amidst salt crystals left behind by the receding, drying waters.
The larvae of mosquitoes feed on algae, detritus and microorganisms, either
by filter feeding using bristles around the mouth, or by grazing on algae and
bacterial mats using the mandibles. Saline-water species benefit from the high
productivity of brackish and saline sites. Mosquito larvae are very vulnerable to
predation by fish, notonectids, dragonfly larvae, etc. but saline habitats, and par-
ticularly those subject to rapid changes in salinity, tend to have few of these pred-
ators. As a result, saline-water species of mosquitoes tend to occur in high-nutrient,
productive waters with few predators. Saline habitats are famous for their extra-
ordinarily high densities of mosquitoes: in Florida, more than 250,000 mosquitoes
were trapped in one location in one night, and egg densities of 10,000/square foot
(>110,000/m2) have been recorded (Provost, 1952; Nayar, 1985).
The pupae of mosquitoes are mobile, but lack a mouth and do not feed or
drink. The cuticle appears to be quite impermeable as Sheplay and Bradley
(1982) showed that pupae could complete their 2-day metamorphosis even in
extraordinarily concentrated media. No site of active salt uptake or excretion is
known from pupae.
The adults emerge from the water surface and begin a terrestrial and aerial
existence. The adults of saline-water forms are indistinguishable from freshwater
forms with regard to behaviour. The males feed only on nectar, the females on
nectar and blood. Some species are autogenous, meaning that they can produce
at least one batch of eggs without feeding, using nutrients carried over from the
larval stages. The only behaviour known to vary in the saline-water forms is the
preference for oviposition sites.

Physiology of osmoregulators

Biologists have long been fascinated by the capacity of mosquito larvae to sur-
vive in saline environments that are much too toxic for any aquatic vertebrate.
Saline-water Insects 29

Beadle (1939) was one of the first to investigate the physiology of this group and
his studies focused on Aedes detritus, a species that resides in coastal marshes in
northern Europe. He demonstrated that the insects ionoregulate and that organs
in the posterior portion of the animal were responsible for that regulation.
Ramsay (1950) demonstrated that the rectum was the site of ion concentration
and a hyperosmotic urine was excreted via the anus. Similar results were obtained
in the species Opifex fuscus, a species endemic to New Zealand (Nicolson and
Leader, 1974).
On the basis of these studies, it was clear that the species of Aedes that had
been examined up until that time were osmoregulators, meaning that they regulate
the osmotic concentration of their haemolymph at about 300 mOsm, regardless of
the concentration of the external medium. As these larvae often are found in
saline waters with osmotic concentrations above that of seawater (1000 mOsm),
they clearly face large osmotic gradients driving water out of the larva. Bradley
and Phillips (1975, 1977b) conducted a series of studies to determine how the
larvae were able to maintain osmotic and volume regulation under these circum-
stances. They found that the saline-water larvae drink the medium at a rapid
rate, equal to about their body volume every 10 h. Interestingly, this occurs in
freshwater larvae as well, although the rate may be lower. Drinking the external
medium therefore occurs in all larvae examined to date, and may be an additional
mechanism for obtaining nutrients in the form of dissolved organic material. In
the case of saline-water forms, drinking replaces the water lost across the integu-
ment by osmosis, as well as any water lost in the urine. This solves the problem
of volume regulation, but the high salinity of the water exacerbates the problem
of ion regulation. The rectum, therefore, must deal with these ionic and osmotic
loads.
An important feature found in the saline-water forms of Aedes and Opifex is
a two-part rectum, in contrast to freshwater forms that have a rectum with only
one segment. Bradley and Phillips (1975, 1977a,b,c) conducted several studies
examining the process of hyperosmotic urine production in the larvae of Aedes
taeniorhynchus and Aedes campestris. They examined the function of the rectum
in detail and found that the anterior rectal segment, that portion resembling the
rectum of freshwater forms in its histology, served a similar function in saline-
water larvae, i.e. as the site of potassium and chloride uptake from the urine (Fig.
2.3). The posterior segment, namely the portion that is histologically distinct
from the freshwater rectum, is unique to saline-water forms and serves as a salt
gland. It is the site of active ion transport into the rectal lumen. Ion transport is
rapid and, as it is unaccompanied by equivalent amounts of water, the secretion
is hyperosmotic. Bradley and Phillips (1975, 1977a,c) found that the concentra-
tion of the rectal fluid was always hyperosmotic to the medium in all waters in
which the larvae could survive.
The types of ions and rates of ion transport are influenced by the water in
which larvae are reared (Bradley and Phillips, 1977a). The larvae can survive in
very dilute waters if they hatch in such media. Under these circumstances, the
recta presumably function in a manner identical to those of freshwater species. In
concentrated media, the recta adjusted the ions transported to match the ions
taken in during drinking. In water rich in sodium and bicarbonate, the rectal fluid
30 T.J. Bradley

2+
Mg
2–
SO4
+
– Na +
CI K
+
K 2+
Mg
Anterior Posterior
Midgut
rectum rectum

–
lleum HCO3
–
CI

Malpighian tubule

Fig. 2.3. Internal organs of the larvae of the saltwater mosquito Aedes
taeniorhynchus. The directions of active ion transport are depicted. The rectum is
divided into an anterior and a posterior segment, which have highly differentiated
roles in ion transport (Redrawn from Bradley, 1985).

was also rich in sodium but low in chloride. When the water was rich in sodium
and magnesium sulphate, the rectal fluid was also rich in sodium and magne-
sium. The only ion which it seems the larval rectum could not transport was
sulphate. This large, divalent anion was never transported in substantial amounts
by the rectal epithelium. Instead, the Malpighian tubules are the site of excretion
of this ion. The Malpighian tubules produce a fluid isosmotic with the haemo-
lymph. As the haemolymph is regulated at a concentration of about 300 mOsm,
the amount of sulphate the larvae can eliminate through these organs limits their
capacity to survive in sulphate-rich waters.
One remaining question is how larvae obtain physiologically necessary ions
from the medium if these are in low supply. For example, in some inland saline
waters, the medium can be highly saline, yet be very low in chloride, an ion
required for normal physiological function. In dilute media, freshwater species
take up sodium and chloride from the external medium using the anal papillae.
These are protruberances near the anal opening at which active ion uptake
occurs (Patrick et al., 2006, Donini et al., 2006). When saline water species are
grown in dilute media, they have enlarged anal papillae that are also likely to be
involved in ion uptake. In concentrated media, the anal papillae may be involved
in ion uptake if certain ions occur at low concentrations in the media (Phillips
and Meredith, 1969).

Physiology of osmoconformers

Not all species of mosquitoes with saline-tolerant larvae engage in osmoregula-

tion; larvae in the genera Culex and Culiseta osmoconform in all media more
Saline-water Insects 31

concentrated than 30% seawater. This means that, in media ranging from
about 300 mOsm to 1000 mOsm (the concentration of seawater), the osmotic
concentration of the haemolymph exactly matches that of the external medium
(Garrett and Bradley, 1984).
Analyses of the haemolymph of larvae of Culiseta inornata and Culex tarsa-
lis reared in concentrated media, revealed that the extracellular fluids in these
larvae contain high levels of amino acids, particularly proline and serine (Garrett
and Bradley, 1984). The haemolymph also contained unusually high levels of
trehalose, the common blood sugar of insects. Clearly, these larvae allow high
levels of compatible solutes to accumulate in the haemolymph, thereby raising
the osmotic concentration of the haemolymph to a level equal to that of the
external medium.
Compatible solutes are osmotically active compounds that, although they
raise the osmotic concentration of the water in which they are dissolved, do not
disrupt proteins and membrane structure in the same manner as salts (Yancey,
2005). Ionized salts such as Na+, K+ and Cl– have high charge densities. These
interact with proteins and membranes, often binding to charged sites and dis-
rupting hydrogen bonds that determine the tertiary structure of proteins. Com-
patible solutes are osmotically active, as they are fully solubilized, but they lack
the charge density required to disrupt hydrogen bond and protein structure. As a
result, the osmotic concentration of the haemolymph is elevated, but physiological
functions in the tissues are not disrupted.
The above studies elucidated the osmolytes that occur in the haemolymph
of osmoconforming mosquito larvae. Given that the cells must themselves osmo-
conform with the haemolymph that bathes them, what are the osmolytes in the
intra-cellular compartment? Patrick and Bradley (2000) answered that question
by analysing the organic solutes in the extra- and intra-cellular compartments of
Culex tarsalis larvae acclimatized to 60% seawater. They found that trehalose
was restricted to the haemolymph as an osmolyte, but that the intracellular
compartment was protected by high levels of proline. This was the first demon-
stration that the same compatible osmolyte could be used both intra- and extra-
cellularly in any animal.
Garrett and Bradley (1987) speculated on the physiological advantages of
the osmoconforming (Culiseta and Culex) versus the osmoregulating (Aedes and
Opifex) strategy. Osmoconforming larvae do not experience a large osmotic
gradient across the body wall and, as a result, do not lose water osmotically to
the external medium. This reduces the need for drinking and greatly reduces
the need for active ion transport. The strategy of osmoconforming, therefore,
should result in a substantial energy savings. Some energy is required to syn-
thesize the organic compounds used as compatible osmolytes, but this does
not represent a real energy loss as these compounds are in the body and can
be used later, for example, during the pupal or adult stage as sources for
metabolic energy. Despite the advantages of this system, Garrett and Bradley
(1987) indicate that osmoconforming species have been reported in water
only ranging up to about 60% seawater. Since then, we have collected Culex
tarsalis from the Salton Sea in California, USA, in water that had a concentra-
tion of 110% seawater (personal observations). It is unknown, however, if the
32 T.J. Bradley

larvae were always in such waters or could complete development under these
conditions.
The osmoregulatory strategy, by contrast, rests on the notion that the inter-
nal fluid compartment will be strictly and rigorously regulated with regard to
ionic and osmotic strength. Substantial energy output will be devoted to that
purpose and specialized organs (the posterior rectal segment) must evolve to
meet these demands. The advantage of this mechanism is that the mosquitoes
can survive in any medium. Ion pumps exist for most of the major ions found in
the environment and the larvae can use the rectum and Malpighian tubules to
remove these from the haemolymph, should they enter through ingestion. The
disadvantage, that energy is expended in active ion transport, may be overcome
by the rich nutrient levels in the water and the abundance of food in the form of
algae and bacteria. The support for these notions resides in the fact that the
osmoregulating larvae of Aedes taeniorhynchus, A. sollicitans and A. campestris
are routinely found in salt marshes and saline ponds ranging in salinity up to
200% seawater. Bradley and Phillips (1977b) examined larvae that could com-
plete their development in NaCl-rich water with a concentration equivalent to
300% seawater. On the basis of these observations, it would seem that osmo-
regulation is a viable strategy for invading highly saline media. The energy cost
this strategy engenders can seemingly be overcome if the environment is
sufficiently rich in food for the larvae.

Evolution

In recent years, my collaborators and I have been investigating the evolutionary

pathways by which saline tolerance arose in mosquitoes (Grueber and Bradley,
1994; Albers and Bradley, 2005). The first step was to establish the primitive
condition in the family Culicidae. The first clue is that 95% of all mosquito spe-
cies are restricted to freshwater, i.e. they possess a one-part rectum, can only
produce urine that is iso-osmotic or hypo-osmotic to the haemolymph, and cannot
survive in media more concentrated than about 30% seawater. The rectum of
these species can only transport ions into the haemolymph, not out. In addition,
the two families of insect most closely related to the Culicidae are both strictly
freshwater. Taken together, these observations argue that the freshwater condi-
tion is primitive in the Culicidae. This conclusion is further supported by the
observation that salt tolerance occurs in many genera (nine genera are listed by
O’Meara, 1976) yet, in most genera, saline tolerance is a rare and derived trait.
The most likely scenario, therefore, is that mosquitoes were initially restricted to
freshwater and that saline tolerance arose repeatedly and fairly recently in sev-
eral lineages in the family.
To examine the process in more detail, we chose a group of mosquitoes
formerly in the genus Aedes, but now considered to be in the genus Ochlerota-
tus. Many of the species discussed above (Aedes taeniorhynchus, A. campestris,
A. sollicitans, A. detritus) have now been placed in the genus Ochlerotatus. We
mapped the ecology and physiology of the species in this group on to a phyl-
ogeny provided by Schultz et al. (1986); the results are shown in Fig. 2.4. Note
Saline-water Insects 33

Character
state

A. campestris EO

A. caspius EO
A. dorsalis EO
A. melanimon EO
A. niphadopsis ?
A. canadensis ?
A. detritus EO
A. cantator ?
A. cataphylla ?
A. flavescens ?
A. taeniorhynchus EO
A. sollicitans EO
A. nigromaculis OF
A. aegypti OF

Fig. 2.4. The phylogeny of species of Aedes (now Ochlerotatus). Salt-tolerant

species are indicated by a character state of EO (euryhaline osmoregulators).
Freshwater species are designated by OF (obligate freshwater) (Redrawn from
Grueber and Bradley, 1994). The filled box indicates a change of character state to
salt tolerant; the open box indicates reversion to the obligate freshwater condition.

that the saline-tolerant species are scattered throughout the phylogeny. Assum-
ing that the freshwater condition was ancestral, saline tolerance apparently arose
repeatedly in the group. Alternatively, the entire clade may have been saline-
tolerant with several cases of reversion to the freshwater condition.
The species Aedes melanimon provides an interesting situation. The larvae
of this species have a two-part rectum and can survive in full-strength seawater.
In nature, however, the larvae are found in freshwater habitats in flooded pas-
tures in Southern California. It seems, therefore, that the physiology and ecology
of the species do not always map precisely on each other.
The species Aedes nigromaculis is also an interesting case. Aedes nigromaculis
(a freshwater species with a one-part rectum) is a sister species to A. sollicitans (a
saline-water species), and their closest relative is A. taeniorhynchus, another saline-
water species. This species cluster provides the strongest evidence that reversion
to the freshwater condition has occurred in this clade, with reversion involving
behavioural, physiological and morphological traits.
We are left, therefore with the surprising result of multiple evolutionary steps,
forward and back, as species move in and out of a variety of ecological niches.
The changes involve complex genetic influences on behaviour and physiological
capabilities.
Our understanding of the evolution of enhanced saline tolerance in this
clade will improve as we gain further molecular information regarding the phylo-
genetic relationships. Even now, however, it is evident that the evolutionary pro-
cess by which saline tolerance arises in mosquitoes is complex and governed by
stochastic events as much as by environmental pressures.
34 T.J. Bradley

References

Albers, M. and Bradley, T.J. (2005) A phylogenetic analysis of the evolution of saline
tolerance in mosquito larvae. Integrative and Comparative Biology 45, 1104.
Beadle, L.C. (1939) Regulation of the hemolymph in saline-water mosquito larvae,
Aedes detritus Edw. Journal of Experimental Biology 16, 346–362.
Bradley, T.J. (1985) The excretory system: Structure and physiology. In: Kerkut, G.A. and
Gilbert, L.I. (eds) Comprehensive Insect Physiology, Biochemistry and Pharmacology.
Pergamon Press, Oxford.
Bradley, T.J. (2002) Excretion. In: Encyclopedia of Insects. Academic Press, Amsterdam,
pp. 380–387.
Bradley, T.J. and Herbst, D.B. (1994) Growth and survival of the larvae of Ephydra hians
Say (Diptera: Ephydridae) on unialgal diets. Journal of Environmental Entomology
23, 276–281.
Bradley, T.J. and Phillips, J.E. (1975) The secretion of hyperosmotic fluid by the rectum
of a saline-water mosquito larva, Aedes taeniorhynchus. Journal of Experimental
Biology 63, 331–342.
Bradley, T.J. and Phillips, J.E. (1977a) Regulation of rectal secretion in saline-water mos-
quito larvae living in waters of diverse ionic composition. Journal of Experimental
Biology 66, 83–96.
Bradley, T.J. and Phillips, J.E. (1977b) The effect of external salinity on drinking rate and
rectal secretion in the larvae of the saline-water mosquito, Aedes taeniorhynchus.
Journal of Experimental Biology 66, 97–110.
Bradley, T.J. and Phillips, J.E. (1977c) The location and mechanism of hyperosmotic
fluid secretion in the rectum of the saline-water mosquito larvae, Aedes taenio-
rhynchus. Journal of Experimental Biology 66, 111–126.
Cheng, L. (ed.) (1976) Marine Insects. North Holland Publishing Co., Amsterdam.
Donini, A., Gaidhu, M.P., Strasberg, D.R. and O’Donnell, M.J. (2006) Changing salinity
induces alterations in hymolymph ion concentrations and Na+ and Cl– transport
kinetics of the anal papillae in the larval mosquito, Aedes aegypti. Journal of
Experimental Biology 210, 983–992.
Garrett, M.A. and Bradley, T.J. (1984) The pattern of osmotic regulation in larvae of the
mosquito Culiseta inornata. Journal of Experimental Biology 113, 133–141.
Garrett, M.A. and Bradley, T.J. (1987) Accumulation of proline, serine and trehalose in
the hemolymph of osmoconforming brackish-water mosquitoes. Journal of Experi-
mental Biology 129, 231–238.
Grueber, W.B. and Bradley, T.J. (1994) The evolution of increased salinity tolerance in larvae
of Aedes mosquitoes: a phylogenetic analysis. Physiological Zoology 67, 566–579.
Hashimoto, H. (1976) Non-biting midges of marine habitats (Diptera: Chironomidae). In:
Cheng, L. (ed.) Marine Insects. North Holland Publishing Co., Amsterdam, pp. 377–
414.
Herbst, D.B. and Bradley, T.J. (1989) A Malpighian tubule lime gland in an insect inhabiting
alkaline salt lakes. Journal of Experimental Biology 145, 63–78.
Joose, E.N.G. (1976) Littoral apterygotes (Collembola and Thysanura). In: Cheng, L. (ed.)
Marine Insects. North Holland Publishing Co., Amsterdam, pp. 151–186.
Leader, J.P. (1972) Osmoregulation in larvae of the marine caddis fly, Philanisus plebeius
(Walk.) (Trichoptera). Journal of Experimental Biology 57, 821–838.
Leader, J.P. (1976) Marine caddis flies (Trichptera: Philanisidae). In: Cheng, L. (ed.) Marine
Insects. North Holland Publishing Co., Amsterdam, pp. 291–302.
Saline-water Insects 35

Linley, J.R. (1976) Biting midges of mangrove swamps and salt-marshes (Diptera: Ceratopo-
gonidae). In: Cheng, L. (ed.) Marine Insects. North Holland Publishing Co., Amsterdam,
pp. 335–376.
Nayar, J.K. (ed.) (1985) Bionomics and Physiology of Aedes taeniorhyncus and Aedes
sollicitans, the Salt Marsh Mosquitoes of Florida. Florida Agriculture Experiment
Stations, University of Florida, Gainesville, Florida.
Nicolson, S.W. and Leader, J.P. (1974) The permeability to water of the cuticle of the
larva of Opifex fuscus (Hutton) (Diptera: Culicidae). Journal of Experimental Biology
60, 593–604.
O’Meara, G.F. (1976) Saltmarsh mosquitoes (Diptera: Culicidae). In: Cheng, L. (ed.)
Marine Insects. North Holland Publishing Co., Amsterdam, pp. 303–333.
Patrick, M.L. and Bradley, T.J. (2000) The physiology of salinity tolerance in larvae of two
species of Culex mosquitoes: the role of compatible solutes. Journal of Experimental
Biology 203, 821–830.
Patrick, M.L., Aimanova, K., Sanders, H.R. and Gill, S.S. (2006) P-type Na+/K+ ATPase
and V-type H+-ATPase expression patterns in the osmoregulatory organs of larval
and adult mosquitoes, Aedes aegypti. Journal of Experimental Biology 209, 4638–
4651.
Phillips, J.E. and Meredith, J. (1969) Active sodium and chloride transport by anal papillae
of a salt-water mosquito larva (Aedes campestris). Nature 222, 168–169.
Provost, M.W. (1952) The dispersal of Aedes taeniorhynchus. I. Preliminary studies.
Mosquito News 12, 174–190.
Ramsay, J.A. (1950) Osmotic regulation in mosquito larvae. Journal of Experimental
Biology 27, 145–157.
Schultz, J.H., Meier, P.G. and Newsom, H.D. (1986) Evolutionary relationships among
the salt marsh Aedes (Diptera: Culicidae). Mosquito Systematics 18, 145–180.
Scudder, G.G.E. (1976) Water-boatmen of saline waters (Hemiptera: Corixidae). In: Cheng,
L. (ed.) Marine Insects. North Holland Publishing Co., Amsterdam, pp. 263–289.
Sheplay, A.W. and Bradley, T.J. (1982) A comparative study of MgSO4 tolerance in
saline-water mosquito larvae. Journal of Insect Physiology 28, 641–646.
Shimizu, M and Bradley, T.J. (1994) Concentrated excretion of sulfate by the anterior
colon of the brine fly, Ephydra hians. Physiological Zoology 67, 54–67.
Simpson, K.W. (1976) Shore flies and brine flies (Diptera: Ephydridae). In: Cheng, L.
(ed.) Marine Insects. North Holland Publishing Co., Amsterdam, pp. 465–495.
Sutcliffe, D.W. (1960) Osmotic regulation in the larvae of some euryhaline Diptera.
Nature 187, 331–332.
Sutcliffe, D.W. (1961) Studies on salt and water balance in caddis larvae (Trichoptera). I
Osmotic and ionic regulation of body fluids in Limnephilis affinis. Curtis. Journal of
Experimental Biology 38, 501–519.
Weigmann, G. (1973) Zur phaenologic von Collembolen in Einflussbereich des meeres.
Faun.-Oekol, Mitt, 4, 185–196.
Yancey, P.H. (2005) Organic osmolytes as ‘compatible, metabolic’ and counteracting
cytoprotectants in high osmolarity and other stresses. Journal of Experimental
Biology 15, 2819–2828.
3 Larval Cannibalism and
Population Dynamics of
Dragonflies
FRANK JOHANSSON1 AND PHILIP H. CROWLEY2
1Department of Ecology and Environmental Science, Umeå University,
SE-901 87, Umeå, Sweden
2Department of Biology, University of Kentucky, Lexington, Kentucky, USA

E-mail: [Link]@[Link]

Abstract
Cannibalism, the process of killing and eating conspecifics, is common among odonate
larvae and is believed to influence odonate population dynamics. Here we attempt to
summarize and consider interactions among key factors linked to cannibalism at both the
individual and population levels. Through cannibalism, odonate larvae receive energy
directly from the consumption of the conspecific, and indirectly from reduced exploitation
competition because the per capita food supply may increase. Cannibalism might,
however, also incur costs such as risk of death and pathogen infections. Alternative food
availability, population density and size structure of the cannibalistic population, and
habitat structure are environmental factors that affect cannibalism rate in odonate larvae
on a short-term basis. Theoretical models predict that cannibalism reduces size variation
under most cannibalism intensities and life histories. The models also show that cannibalism
can – but will not always – stabilize population dynamics. Unfortunately few long-term
studies examining the population dynamics on dragonfly larvae have been performed,
and we urge more such studies.

Introduction

Cannibalism is the process of killing and eating a conspecific individual. It was

not until the influential reviews by Fox (1975) and Polis (1981) that ecologists
started to recognize cannibalism as a ubiquitous phenomenon in natural sys-
tems. The two reviews also sparked a lot of interest in the relationship between
cannibalism and such phenomena as behaviour, sexual selection, life history,
competition, and population dynamics and structure. In this chapter we show
that studies of dragonflies have widely contributed to the increased understand-
ing of these relationships. In cases where no evidence of cannibalism has been
found in dragonflies, we will turn to examples from other organisms.

 CAB International 2008. Aquatic Insects: Challenges to Populations

36 (eds J. Lancaster and R.A. Briers)
Larval Cannibalism of Dragonflies 37

Our focus is on the ecologically important larval stage of dragonflies, but we

would like to encourage the study of cannibalism in adult dragonflies as well.
Such studies are to our knowledge non-existent, though we both have personal
observations of this phenomenon. A well studied aspect of cannibalism that we
will not address here is sexual cannibalism (Elgar, 1992), as we are aware of no
examples in dragonflies. We start our chapter by evaluating the costs and bene-
fits of cannibalism. After that, we look at mechanisms that affect cannibalism
rate, and then we focus on implications of cannibalism for population dynamics
and structure, and, finally, we provide some suggestions for future research on
the topic.

Cannibalism Evidence from Diet Data

Diet data show that predation by odonate larvae on other odonate larvae is
frequent in nature, suggesting a potential for cannibalism to affect population
dynamics. A literature survey covering six studies showed that the highest per-
centage occurrence of any odonate species in the diet was 18, 5, 8, 10, 38 and
0% in studies by Pritchard (1965), Thompson (1978), Folsom and Collins (1984),
Merrill and Johnson (1984), Blois (1985), Dudgeon and Wat (1986), respec-
tively. Cannibalism intensity also varies within species. Van Buskirk (1992) found
that the frequency of cannibalism changed during development in natural envi-
ronments. His study found that cannibalism ranged from 0 to 14.7%, being low
in small larvae and higher in large larvae (Van Buskirk, 1992). Laboratory stud-
ies and short-term semi-field experiments that have explored the various aspects
of cannibalism are common, but more studies that estimate cannibalism of odonate
larvae across species in the wild are clearly needed to identify factors influencing
the frequency and impact of cannibalism.

Benefits and Costs of Cannibalism

By cannibalizing other individuals, odonate larvae could accelerate development

and increase growth rate. Accelerated development leads to earlier emergence,
which has important fitness implications (Anholt, 1991; De Block and Stoks, 2005).
A higher growth rate (all else equal) would result in a larger size at emergence,
another important fitness factor in odonates (Banks and Thompson, 1987; Anholt,
1991; De Block and Stoks, 2005). By shortening the dangerous aquatic larval
stage, early emergence also reduces risk of larvae being cannibalized themselves.

Energy gain

The most obvious benefit of cannibalism is the extra energy and nutrients gained
from consuming a conspecific, yet few experimental studies have actually shown
an increase in growth as a result of the extra energy gain from the consumption
of conspecifics. Cannibalism may also result in increased feeding rate with the
38 F. Johansson and P.H. Crowley

removal of a competitor – either because of reduced distraction or risk from

neighbours (Crowley et al., 1988), or because of increased per-capita food avail-
ability (De Block and Stoks, 2004). De Block and Stoks (2004) managed to
separate these two effects in a manipulative laboratory study (Fig. 3.1). They did
this by manipulating the food availability of the cannibals. In Fig. 3.1, per capita
food availability was corrected after cannibalism occurred, so that cannibals and
non cannibals received the same per capita food level after cannibalism. In the
case of no food correction, the cannibal received double the amount of food,
since the food of the victim was available to the cannibal. They found that can-
nibalism increased growth rate, increased mass at emergence and decreased
time to emergence. Figure 3.1 shows that mass at emergence increases in the
case of food correction (so that cannibals and non-cannibals received the same
amount of food after cannibalism to permit estimation of the direct gain from
cannibalism). Interestingly, De Block and Stoks (2004) also found that cannibal-
ism early during ontogeny had a large impact on size at emergence (Fig. 3.1).
The increased per capita food availability is an indirect benefit of cannibalism
because it may reduce competition by removal of a potential competitor, poten-
tially increasing feeding rate. This indirect benefit is a special case of the indirect
food web phenomenon ‘thinning’ (Abrams et al., 1996; Van Buskirk and Yure-
wicz, 1998). De Block and Stoks (2004) explicitly examined the effect on growth
and development when a competitor was eliminated by cannibalism. Their study
found that cannibalism increased growth rate and mass at emergence and
decreased time to emergence. Figure 3.1 shows that cannibals grew to a larger
size at emergence by eating the food that would have gone to the cannibalized
competitor. This result shows that thinning benefits are possible, but tells us little
about their potential importance in nature.
Other studies on odonate larvae have also shown that growth and develop-
ment increase via cannibalism. These include laboratory experiments (Hopper
et al., 1996; Johansson, 1996) and semi-field experiments (Anholt, 1994). In the
tree-hole-living damselfly Megaloprepus coerulatus, Fincke (1994) found no dif-
ference in growth between individuals that cannibalized one smaller individual
compared with a control, indicating no advantage gained from the cannibalism.
None of these studies separated the direct effect of eating the victim from the
indirect effect of reduced exploitation competition, but both factors may have
contributed to the life-history responses observed.

Injury or death

One potential cost of cannibalism is that engaging in a cannibalistic interaction

might result in an injury and, in the extreme case, the potential cannibal might
end up being the prey. There exists no direct evidence of this cost in the odonate
literature. However, indirect evidence comes from the fact that similar sized indi-
viduals often engage in staring contests against each other, resulting in an apparent
war of attrition (Crowley et al., 1988). These contests usually end with one of the
contestants withdrawing. One interpretation of the withdrawal could be that it
is too costly and risky to attack. In contrast, cannibalism between individuals
Larval Cannibalism of Dragonflies 39

8 8

Mass at emergence (mg)

7 7

6 6

5 5

4 4

No Yes No Yes No Yes No Yes

Period 1 Period 2 Period 1 Period 2
Food correction No food correction
Cannibalism

Fig. 3.1. The effect of cannibalism on mass at emergence of female larval Lestes
viridis. Circles denote high food level and triangles low food level. White symbols
indicate early photoperiod (low time stress) and dark symbols late photoperiod
(time stress). Period 1 denotes if cannibalism occurred during the first half of the
developmental period (before 50 days), and Period 2 if cannibalism occurred during
the second half of the developmental period (day 51–104). See text for explanation
of food correction. Modified from De Block and Stoks (2004).

differing in size is common even in the case where staring contests are going on
initially (Baker and Dixon, 1986), suggesting that risk for the large individual is
less in interactions. It could be argued that individuals do not have the physical
capacity to prey on similar-sized individuals, but one of us (F. Johansson) has
observed final instar larvae of Aeschna juncea cannibalizing each other under
starvation. This observation suggests that there exists a negative relationship
between aggressiveness and condition, and theory predicts that animals should
be more risk prone when in poor condition (Houston et al., 1988).

Pathogens and parasites

As individuals might be more likely to acquire pathogens from conspecifics than

from heterospecifics (Freeland, 1983), we can assume that cannibals incur a
higher risk of pathogen infection. Though no studies of cannibalistic transmission
are available, there are examples from other organisms showing transmission of
pathogens via cannibalism (references in Pfennig et al., 1998). Such transmis-
sion might cause death or reduced fitness. For example, Pfennig (2000) elegantly
demonstrated a reduced growth rate by tadpoles fed infected conspecifics in
an experiment. However, recent theoretical modelling suggests that pathogens
40 F. Johansson and P.H. Crowley

transmitted by cannibalism might be rare except under certain circumstances

(Rudolf and Antonovics, 2007).

Loss of fitness

Another potential cost of cannibalism is that it might reduce inclusive fitness

because of the risk of eating close relatives. As eggs of odonates are often attached
together or deposited close together, there is a high risk of siblicide. On the other
hand, cannibalizing a sibling reduces future resource competition, and hence
sibling cannibalism might be beneficial in certain circumstances (Mock and
Parker, 1997). From a theoretical point of view, we should thus expect cannibal-
ism on relatives to be dependent on resources (Mock and Parker, 1998). No
studies have been undertaken with odonates to determine whether they can
discriminate among relatives with respect to cannibalism. From other inverte-
brates we know that such discrimination is possible (Michaud, 2003), but in
many cases kin may tend to be similar in size and therefore cannibalism is less
probable, even in the absence of sib discrimination (Wise, 2006).
In summary, we have good evidence that cannibalism provides benefits in
dragonflies. Besides the costs of injury and death, we have little or no evidence
of potential costs of cannibalism in dragonflies, though such costs occur in many
other organisms. Certainly, more research is needed concerning these other costs.

Factors Affecting Cannibalism

Intraspecific density

If intraspecific density is high then we should expect frequent encounters between

individuals, which could increase cannibalism. Such encounters are determined
by the movement activity of the larvae. Support for higher rates of cannibalism
at higher densities in odonate larvae has been found in simple laboratory studies
(Fischer, 1961; Van Buskirk, 1989; Hopper et al., 1996; Claus-Walker et al.,
1997; De Block and Stoks, 2004). In Fig. 3.2a, for example, it can be seen that
cannibalism is more frequent at high densities of Epitheca cynosura under labo-
ratory conditions. Field experiments, field enclosure experiments and mesocosm
experiments suggest that the pattern observed at the laboratory level is found at
larger-scale experiments as well (Johnson et al., 1985; Van Buskirk, 1989, 1993;
Anholt, 1994; Fincke, 1994). For example, survival rate of Epitheca cynosura
was dependent on intraspecific density, and this effect was not attributed to prey
density as food was not limited (Johnson et al., 1985). A laboratory study by
Crowley and Martin (1989) showed that per capita cannibalism rate went down
with increasing density of potential cannibals. This suggests that similar sized and
large cannibals engage in interactions that reduce cannibalism, though not
enough to eliminate or reverse the density-dependent effect at the population
level. But, a study by Van Buskirk (1992) did not find any density-dependent
effects of cannibalism in field experiments run for 10–15 days.
Larval Cannibalism of Dragonflies 41

10,000 10,000
(a) High density (b) High density
Low density Low density
Density (number/m2)

1000 1000

100 100

0 0
0 20 40 60 80 0 20 40 60 80
Time (days) Time (days)

Fig. 3.2. Survival of high and low density cohorts of larval Epitheca cynosura. In (a), larvae
were hatched asynchronous, and in (b) larvae were hatched synchronous. Dashed lines
denote the span of the hatching period. Modified from Hopper et al. (1996).

In summary, density-dependent cannibalism is common in odonate larvae,

but exceptions can occur. We suggest that the exceptions may arise from factors
such as size distribution of larvae, habitat structure and behaviours, which will be
discussed below.

Alternative prey

We define alternative prey as other species available for consumption. Alternative

prey are likely to affect cannibalism through hunger level. If we assume that canni-
balism is risky, then cannibalism should increase with decreasing alternative prey
densities. The reason is that as density of alternative prey goes down, consumption
of those prey goes down as well, increasing hunger. An increase in hunger usually
triggers active and riskier foraging (Johansson, 1991, 1993; Werner and Anholt,
1993), which results in more encounters among conspecifics and more canni-
balism. Support for this scenario has been found in odonates (Johansson, 1992;
Hopper et al., 1996). Though hunger increases cannibalism, starvation seems
unlikely in the field, as odonates can withstand long periods of starvation (Lawton
et al., 1980; Wissinger, 1988). The incidence of cannibalism suggests that the bene-
fits of a fast development and growth rate can outweigh the associated risk. Even if
the risk of cannibalism is low, hunger level should increase cannibalism as encoun-
ters with conspecifics will be more common than encounters with alternative prey.

Habitat structure

Prey capture success declines with increasing habitat complexity, primarily because
predator mobility and detection of prey are hindered by structure (Diehl, 1988).
42 F. Johansson and P.H. Crowley

Similarly, cannibalism rate in odonate larvae may go down as complexity increases.

We have no direct evidence of this, but we provide examples from odonate systems
where predation on alternative prey is reduced by habitat structure. First, Folsom and
Collins (1984) found that predation by Anax junius on amphipods decreased with
increasing habitat complexity. Second, predation rate on the damselfly Coenagrion
puella by the backswimmer Notonecta glauca was lower at high habitat complexity
compared with low complexity (Thompson, 1987). Thus, if predation success by
odonate larvae goes down and their susceptibility as prey goes down as complexity
increases, this suggests less cannibalism in complex habitats. Anholt (1994) found
support for this prediction: survival of small larvae of the damselfly Enallagma bore-
ale in the presence of large larvae was 10 times higher in a high complexity habitat
compared with a low complexity habitat. But greater habitat structural complexity
can sometimes reduce prey numbers (Warfe and Barmuta, 2006). Similarly, we
suggest that cannibalism could increase with greater habitat complexity under
certain circumstances. For example, for sit-and-wait cannibals, a higher habitat
complexity might allow for a higher access to prey swimming in the water column.

Size structure

Cannibalism generally increases with relative size difference between the canni-
bal and the victim. Predators prefer to prey on prey smaller than themselves,
presumably because smaller prey are less dangerous and easier to subdue. This
preference is supported in studies on odonate larval cannibalism (Crowley et al.,
1987a; Wissinger, 1988; Fincke, 1994). Hopper et al. (1996) showed that can-
nibalism was low if larvae did not differ in instar, intermediate if larvae differed
by one instar and high if the difference was two instars. A similar increase in can-
nibalism as an effect of instar difference between larvae was found by Wissinger
(1988). His study showed a positive relationship between number of small larvae
cannibalized and the size of cannibal in two species of dragonflies (Fig. 3.3). This
pattern suggests that asynchronous hatching of eggs should increase the rate of
cannibalism. In a laboratory experiment, Hopper et al. (1996) showed that asyn-
chronous hatching (resulting in a broad size distribution) of larvae resulted in
higher cannibalism compared with synchronous hatching. The results of their
study can be seen in Fig. 3.2 where per capita death rate (the slope of the
curve) is higher at high densities and asynchronous hatching. A high cannibalism
intensity on small instar larvae can alter the size structure and developmental
synchrony of the population (Crowley and Hopper, 1994).

Time constraints

Most odonates live in seasonal environments that impose time constraints, as larval
development must cease with the onset of sufficiently severe conditions. Spe-
cies that must emerge before winter should speed up their development and
growth rates to keep size at emergence as large as possible (Rowe and Ludwig,
1991). One way of speeding up growth rate under time constraint is to increase
Larval Cannibalism of Dragonflies 43

12 Libellula
lydia

8
No. of F6 larvae cannibalized

F6 F4 F2 F1

12 Libellula
luctuosa

Fig. 3.3. Number of F-6 instar larvae

4 cannibalized by Libellula lydia and
L. luctuosa larvae in different instars.
F denotes final instar larvae, F1 first instar
from the last, F2 second from the last and
F6 F4 F2 so on. Vertical bars represent 1 SD. Modified
Predator instar from Wissinger (1988).

foraging activity. However, an increase in foraging activity usually results in a

higher encounter rate with predators and conspecifics (Brodin and Johansson,
2004). We should therefore expect cannibalism rate to increase under time stress.
For example, larvae that are late in their development with regard to the onset of the
winter, should forage more actively and hence experience higher cannibalism.
Time constraints can be imposed by rearing larvae under a photoperiod
simulating late season light conditions to which damselflies react by speeding up
development. Johansson and Rowe (1999) found that cannibalism was higher
among time-constrained than non-time-constrained larvae of Lestes congener.
The larvae under time constraint in that study had a higher foraging activity, sug-
gesting that a high encounter rate between individuals was the cause of the
increased cannibalism. Moreover, time-constrained larvae may have been more
willing to accept the risk of attacking a conspecific, possibly subjecting the attacker
to counter attacks. Cannibalism was also more frequent under time constraint in
larvae of Lestes viridis; in individuals with a high rate of cannibalism, develop-
ment and growth rates both increased (De Block and Stoks, 2004).

Other factors

There are many other factors that can also affect cannibalism intensity. We will
mention two additional ones. First, dragonflies are not always the top predators,
44 F. Johansson and P.H. Crowley

and it is well known that top predators reduce activity of other predators that are
potential prey via intraguild predation (e.g. Schmitz et al., 1997). Such reduction
in activity should result in less cannibalism in the presence of fish. Claus-Walker
et al. (1997) found a trend indicating reduced cannibalism in Epitheca cynosura
in the presence of fish scent in a semi-field experiment. Behavioural observation
in the laboratory supported the observed trend, as larvae of E. cynosura reduced
activity in the presence of fish.
Second, odonate larvae do differ in their ‘lifestyle’, sensu Sih (1987). A slow
lifestyle is characterized by low activity, restricted microhabitat use, low feeding
rate and low growth rate, whereas a fast lifestyle implies the opposite. In a suite
of six species of odonate larvae, Johansson (2000) showed a strong difference in
activity and microhabitat use. The study found a continuum from relatively inac-
tive species exploiting the bottom as a microhabitat to relatively active species
using the water surface as their microhabitat. The fast lifestyle odonate species
were associated with fishless lakes while the slow lifestyle species were associated
with lakes containing fish (Johansson, 2000). We should expect cannibalism to
be reduced in the slow lifestyle species with reduced activity living on the bottom.
No empirical studies have examined how such lifestyle differences among spe-
cies relate to cannibalism rates. However, a study on the effect of intraguild pre-
dation among species found differences between odonate species in intraguild
predation rate (Johansson, 1993). The difference in intraguild vulnerability in
that study yielded equivocal results on the relationship between lifestyle charac-
teristics and intraguild predation.
The primary challenge for ecologists is to determine which of the above
factors and others are the most important and how they affect population
structure, population dynamics and distribution of species (Crowley et al.,
1987b). Figure 3.4 provides an overview on how cannibalism might influence
these population effects. The basic life cycle moves early instar (small) larvae by
growth and development into the later instars (large larvae), contingent on sur-
vival, and the surviving late-instar individuals mature and reproduce, yielding
the next generation after a time-lag of days to months. Prey availability (an alter-
native to eating conspecifics) generally speeds these vital rates with little effect on
survival; but under food scarcity, survival can be reduced by risky foraging
(shown as positive effects of growth and development or maturation and repro-
duction on survival). In general, however, survival and developmental processes
trade off, depending especially on seasonal constraints and the influence of top
predators. To the extent that larvae of different sizes maintain similar diets (per-
haps more typical of slow than fast lifestyles), there may be competition for food,
though evidence for exploitation competition in odonate larvae is limited.
Cannibalism influences these processes in important ways, strongly increasing
in intensity with density of potential cannibals (larger larvae) and victims (smaller
larvae), producing density dependence that can help stabilize odonate population
dynamics. Because cannibalism is often risky, exposing the potential cannibal to
counter-attack or to other predators or parasites, the frequency of cannibalism is
generally expected to be inversely related to the abundance of alternative prey.
By evaluating the relative importance of these processes and effects on different
species, especially with respect to slow and fast lifestyles, seasonal constraints,
Larval Cannibalism of Dragonflies 45

–
ALTERNATIVE
PREY
AVAILABILITY
EXPLOITATION –

+ EXPLOITATION
+ +
+ + +
–
DENSITY OF + DENSITY OF +
GROWTH & MATURATION &
SMALL DEVELOPMENT LARGE REPRODUCTION
+
LARVAE LARVAE
– – +
–
+ + + +
TRADE TRADE
OFF OFF
SURVIVAL – – SURVIVAL –
+
– CANNIBALISM
–/+
+

Fig. 3.4. An overview of how cannibalism influences densities and individual-level processes
in larval odonates, based on literature reviewed in this chapter. The primary patterns and
processes are indicated within elipses; environmental factors and state variables are shown
in rectangles.

and predation regime, different dynamic patterns produced by different species

can be better understood.

Population Size Structure and Dynamics

Models

Cannibalism can affect the size structure and dynamics of a population. A

comprehensive review of general theoretical models on cannibalism and the
population dynamics associated with different models is given in Claessen et al.
(2004). Many of the studies they address indicate that cannibalism has a strong
regulatory effect by strengthening density dependence. However, regulation of
population density depends on the number and size of victims a cannibal can eat
and, if saturation of the functional response is too abrupt, then cannibalism fails
to control and can even destabilize population dynamics (Claessen et al., 2004).
Claessen et al. (2004) conclude that when population cycles are induced by
density-dependent effects other than cannibalism (e.g. size-dependent competition
for a common resource) cannibalism is more likely to act as a stabilizing mechanism.
In contrast, age-dependent cannibalism with a time delay may cause population
cycles.
46 F. Johansson and P.H. Crowley

Cannibalism could also serve as a ‘lifeboat mechanism’ (van den Bosch

et al., 1988). Under this scenario a cannibalistic population is able to persist
under severely reduced food availability, whereas it would have gone extinct if
the population was non-cannibalistic. In this case there are two stable states:
extinct and persistent. The lifeboat mechanism allows the population to survive
at low resource densities through a positive feedback in which the cost of can-
nibalism (mortality of victims) is smaller than the benefits gained from consum-
ing victims (cannibalism). Hence the population survives a bottleneck situation
of low food availability and thereafter recovers again as resource density goes
back up. Claessen et al. (2004) end their review by concluding that: (i) canni-
balism tends to be destabilizing with little or no competition for resources between
cannibals and victims, and (ii) if competition is present, cannibalism is stabilizing
when the juvenile victims are the stronger competitors for resources.
We are aware of only one model that considers the size structure and dynamics
of dragonfly populations. Crowley and Hopper (1994) developed a density-
dependent dynamic game model to explore the population consequences of canni-
balism using parameter values from the cannibalistic dragonfly Epitheca cynosura.
Their model showed that cannibalism decreased size variation within the popula-
tion at low to moderate cannibalism intensities if cannibalism increased with relative
size of the cannibal. Cannibalism increased the size variation when cannibalism was
very intense and intensity depended on the absolute size of the cannibal. The model
also showed that, at high alternative prey availability and low to intermediate fecun-
dity, stable population dynamics were generated. But at low alternative prey avail-
ability and high fecundity, the population dynamics were more prone to be irregular.
As odonate systems can be viewed as metapopulations, cannibalism seems
likely to dominate population dynamics only in rare circumstances, such as when
the focal species is an abundant top predator with limited alternative prey available
(see Van Buskirk, 1992 and below). Yet cannibalism (and the threat of cannibalism)
may often influence phenological patterns, especially for species with plastic life
histories and high population densities (Fig. 3.5). This influence hinges to a
considerable extent on the feedback loop from cannibalism, to intra-cohort devel-
opmental synchrony, to instantaneous size variation and back to cannibalism –
potentially facilitated by the effect of synchrony on cohort splitting (Fig. 3.5).
When cannibalism desynchronizes the cohort by speeding the growth of the
large cannibals relative to smaller at-risk individuals, this becomes a positive
feedback loop that could help generate between-year variation. The typically
strong damping effects of dispersal influx, other sources of density dependence
(e.g. predators), and physiologically mediated seasonal forcing (Gurney et al., 1992,
1994), however, should generally prevent major year-to-year fluctuations.
Nevertheless, odonates with circum annual life cycles capable of surviving
the high stress season (winter or dry season) as larvae may cycle through com-
plex seasonal patterns across generations under seasonal forcing (Gurney et al.,
1992, 1994). The result will be multimodal size distributions within years that will
repeat each year, but with individual lineages reaching different modes in differ-
ent years. These life histories usually result in overlapping stable limit cycles,
though not of the period-doubling type. Multimodality is especially susceptible to
shaping by cannibalism resulting from size asymmetries.
Larval Cannibalism of Dragonflies 47

OTHER
ENVIRONMENTAL DISPERSAL SOURCES OF SEASONAL
VARIATION INFLUX DENSITY FORCING
DEPENDENCE

– –
+ – +

+ + STAGGERED
INSTANTANEOUS BETWEEN-YEAR
SIZE VARIATION VARIATION COHORT
+ + INITIATION
–
+ – +
–

INTRA-COHORT COHORT
DEVELOPMENTAL VOLTINISM
– SPLITTING –
SYNCHRONY

+ + +
+
+/–

INTRA-COHORT – LIFE HISTORY LIFE STYLE

CANNIBALISM HATCHING
+ PLASTICITY SPEED
SYNCHRONY

– +
+

POPULATION ALTERNATIVE
DENSITY PREY
DENSITY

Fig. 3.5. Cannibalism and its implications at the population level. The primary patterns and
processes are indicated within elipses; environmental factors and state variables are shown
in rectangles. All influences are primarily unidirectional except for the bi-directional link
between instantaneous size variation and staggered cohort initiation. Voltinism is the
number of generations per year, which may be fractional.

Empirical evidence for models

Few studies have focused on the population dynamics of cannibalism in dragonflies,

but see Van Buskirk (1992, 1993) for an exception. Most studies have addressed
one generation or less, and those studies have focused on size structure and
density dependence of cannibalism.
Support for the model prediction that cannibalistic species should feature
low size variation within a cohort has been found in several studies on dragonfly
larvae (Wissinger, 1988; Van Buskirk, 1989; Fincke, 1994; Hopper et al., 1996).
However, this pattern might simply reflect hatching synchronization and initial
size distribution of larvae. If hatching is asynchronous, cannibalism might spread
48 F. Johansson and P.H. Crowley

the size distribution out, probably as a result of a combined effect of competition

and cannibalism (Hopper et al., 1996).
Support for the model prediction that density-dependent survival might syn-
chronize population dynamics has also been found in dragonfly larvae. Larvae
of Epitheca cynosura showed stabilizing density dependence when hatching of
larvae was asynchronous (Hopper et al., 1996). Seventy-eight days after the first
larvae had hatched, the initial high and low densities of larvae had converged to
similar densities (Fig. 3.2). Interestingly, mortality from cannibalism was present
in larvae when hatching was synchronous, but cannibalism did not result in con-
vergence of densities (Fig. 3.2), suggesting that relative size distribution between
larvae is an important factor that affects dynamics (Crowley and Hopper, 1994).
Density convergence of larvae has also been found in field experiments (Johnson
et al., 1985).
Cohort splitting, where for example one part of a cohort may be on a 1-year
life cycle track and another part on a 2-year life cycle track, is a common phe-
nomenon in dragonflies (Norling, 1984; see also the models of Gurney et al.,
1992, 1994). Cohort splitting could result from competition among larvae or
hatching asynchrony (references in Claus-Walker et al., 1997). As cannibalism is
common in odonate larvae and has the potential to act in a density-dependent
fashion, we suggest that without cannibalism odonate larval density in natural
habitats could show much larger variation among years than generally observed
(Crowley and Johnson, 1992), but effects of cannibalism on within-year and
within-cohort synchrony are more ambiguous. Cohort splitting may result in part
from a behavioural response of the smaller larvae. In the presence of large lar-
vae, small larvae reduce their activity and increase their hiding behaviour as an
antipredator response (Van Buskirk, 1992; Johansson, 1993; Crowley and Hop-
per, 1994). With enough shelter and habitat complexity, this behaviour would
slow down growth and development, resulting in a longer development time
compared with the cannibals. A semi-field experiment by Claus-Walker et al.
(1997) found that larvae from initially high densities with a high degree of
cannibalism varied more in head width compared with larvae at initially low
densities where cannibalism was less. An additional behavioural experiment in
their study showed that the small larvae reduce their activity in the presence of
large larvae. This finding supports the possibility that an antipredator response
could increase size variation.
An intriguing population dynamical pattern in odonate larvae that might be
driven by cannibalism comes from a field study by Van Buskirk (1992). The
study showed a striking 2-year cycle in the age structure of Aeschna juncea
populations (Fig. 3.6). Van Buskirk suggested that a large final (third-year)
cohort suppresses the second-year cohort, thus releasing the first-year cohort.
Once a large third-year cohort emerges, the eggs that are laid by them will
comprise a subordinate cohort suppressed by the immediately preceding
cohort. This 2-year cycle in age structure appears to be a period-doubling
instability attributable to intense cannibalistic interactions between annually
successive cohorts. The synchrony of the 2-year cycle may be maintained, in
part, because the population studied is on an isolated island with presumably
little impact from migrants. This can be visualized in Fig. 3.5 as a case with
Larval Cannibalism of Dragonflies 49

1.0

Proportion of larvae in 3-year age class 0.8

0.6

0.4

0.2

0.0
1986 1987 1988 1989 1990 1991
Year

Fig. 3.6. Two-year cycle in the age structure of Aeschna juncea at Isle Royal. Each
line represents a single pool. Modified from Van Buskirk (1992).

negligible dispersal influx or other sources of density dependence, but low

voltinism and thus high size variation. The size variation results here in between-
year variation but within-cohort synchrony maintained by intense relative, size-
dependent cannibalism.
In summary, empirical results over a short timespan suggest that cannibalism
reduces size variation within a cohort. However, the long-term dynamics of size
distribution are less known. There are also important consequences for commu-
nity structure resulting from cannibalism, and some of these are reviewed in
Claessen et al. (2004) and Crumrine et al. (2008).

Future Directions

In this chapter we have shown that cannibalism is common in dragonfly larvae,

and that cannibalism has several advantages for successful cannibals, such as
faster growth and development. One of the potential costs of cannibalism is that
it can reduce inclusive fitness. Many species of dragonflies deposit their eggs in
clumps or other aggregates, and hence there is a high probability that cannibalism
occurs among siblings upon hatching of eggs. We should therefore expect selec-
tion to act on traits to reduce kinship cannibalism. Many studies have shown that
potential cannibals can recognize kin and therefore minimize kinship cannibalism
(Pfennig, 1997); however, no such studies have been performed in dragonfly
larvae. Two basic types of kin recognition have been identified (Pfennig, 1997).
In the first type, ‘phenotypic recognition’, the cues are produced by the organism
itself. Such cues may often be chemical. In the second type, ‘non-phenotypic recog-
nition’, the cues are based on time and place, which influence cannibalism rate.
For example, eggs within an egg clutch might hatch synchronously and this
50 F. Johansson and P.H. Crowley

timing might decrease cannibalism, as cannibalism is reduced for synchronously

hatching larvae compared with when hatching is asynchronous (Hopper et al.,
1996). Laboratory experiments with manipulation of egg clutches and their
hatching synchrony should be able to explore this issue in dragonfly larvae.
Some organisms feature polyphenism, such that one morph is cannibalistic
while another is non-cannibalistic (e.g. Pfennig, 1992). Such differences are usu-
ally induced by the environment and thus express phenotypic plasticity. Genetic
differences with regard to cannibalism do also occur in some organisms (Baur,
1994; Stevens, 1994; Wagner et al., 1999). Some of the genetic differences in
cannibalism have been attributed to environments. For example, heritable differ-
ences in cannibalism intensity were expressed at low food levels but not at high
food levels in the ladybird beetle Harmonia axyridis (Wagner et al., 1999).We
are not aware of any studies on genetic difference in cannibalism rate within or
between populations in dragonfly larvae. As dragonflies disperse between habi-
tats, such genetic patterns could be evident in dragonflies as well. For example,
in years of harsh conditions with few resources and rapid pool drying, cannibal-
istic genetic lines would be at a selective advantage. But, in years with abundant
resources and little risk of pool drying, cannibalistic genetic lines would be less
favoured. However, genotypes could also respond to resource and pool drying
among years by adjustments in phenotypic plasticity. Whether a fixed or a flex-
ible genotype is favoured depends on several factors, such as predictability of
the habitat quality and cost of plasticity (e.g. Sultan and Spencer, 2002). Two
interesting questions would be: (i) is there genetic variation in the intensity of
cannibalism? (ii) Is there an interaction with environment? These questions
could be studied at the level of differences between populations as well as
between species.
Using a combination of modelling and empirical studies, more work needs
to address the way that cannibalism shapes size distributions and the timing of
life cycles in natural populations. From field patterns and laboratory analyses of
size-related intensities of cannibalism and magnitudes of behavioural shifts,
models can begin to reveal the precise contributions of cannibalism to the dynam-
ics of odonate populations.
Life-history dynamics also need to be addressed from a game-theoretic per-
spective: What is the role of cannibalism in making one life cycle pattern invasi-
ble by another within populations? And, how are these patterns influenced by
the metapopulation structure and seasonal forcing typical of most odonate
populations?
Finally, we emphasize the need to study cannibalism in the adult stage. For
example, cannibalism on egg-depositing females should have a major impact on
the dynamics, as the death of a female eradicates the whole clutch of offspring
that potentially could be deposited into the system.

Acknowledgements

Thanks to Patrick Crumrine, Karin Nilsson and Tim Sesterhenn for comments on
earlier versions of this chapter. FJ was supported by the Swedish Research Council.
Larval Cannibalism of Dragonflies 51

References

Abrams, P.A., Leimar, O., Nylin, S. and Wiklund, C. (1996) The effect of flexible growth
rates on optimal sizes and development times in a seasonal environment. American
Naturalist 147, 381–395.
Anholt, B.R. (1991) Measuring selection on a population of damselflies with a mani-
pulated phenotype. Evolution 45, 1091–1106.
Anholt, B.R. (1994) Cannibalism and early instar survival in a larval damselfly. Oecologia
99, 60–65.
Baker, R.L. and Dixon, S.M. (1986) Wounding as an index of aggressive interactions in
larval Zygoptera (Odonata). Canadian Journal of Zoology 64, 893–897.
Banks, M.J. and Thompson, D.J. (1987) Lifetime reproductive success of females of the
damselfly Coenagrion puella. Journal of Animal Ecology 56, 815–832.
Baur, B. (1994). Inter-population difference in propensity for egg cannibalism in hatch-
lings of the land snail Arianta arbustorum. Animal Behaviour 48, 851–860.
Blois, C. (1985) The larval diet of three anisopteran (Odonata) species. Freshwater Biology
15, 505–514.
Brodin, T. and Johansson, F. (2004) Conflicting selection pressures on the growth/predation
trade-off in a damselfly larvae. Ecology 85, 2927–2932.
Claessen, D., de Roos, A.M. and Persson, L. (2004) Population dynamic theory of size-
dependent cannibalism. Proceedings of the Royal Society B 271, 333–340.
Claus-Walker, D.B., Crowley, P.H. and Johansson, F. (1997) Fish predation, cannibalism
and larval development in the dragonfly Epitheca cynosura. Canadian Journal of
Zoology 75, 687–696.
Crowley, P.H. and Hopper, K.R. (1994) How to behave around cannibals: a density
dependent dynamic game. American Naturalist 143, 117–154.
Crowley, P.H. and Johnson, D.M. (1992) Variability and stability of a dragonfly assem-
blage. Oecologia 90, 260–269.
Crowley, P.H. and Martin, E.K. (1989) Functional responses and interference within and
between year classes of a dragonfly population. Journal of the North American
Benthological Society 8, 211–221.
Crowley, P.H., Dillon, P.M., Johnson, D.M. and Watson. C.N. (1987a) Intraspecific in-
terference among larvae in a semivoltine dragonfly population. Oecologia, 71,
447–456.
Crowley, P.H., Nisbet, R.M., Gurney, W.S.C. and Lawton, J.H. (1987b) Population regu-
lation in animals with complex life-histories: formulation and analysis of a damselfly
model. Advances in Ecological Research 17, 1–59.
Crowley, P.H., Gillet, S. and Lawton, J.H. (1988) Contests between larval damselflies:
empirical steps toward a better ESS model. Animal Behaviour 36, 1496–1510.
Crumrine, P.W., Switzer, P.V. and Crowley, P.H. (2008) Structure and dynamics of
odonate communities: accessing habitat, responding to risk and enabling repro-
duction. In: Córdoba-Aguilar, A. (ed.) Dragonflies: Model Organisms for Eco-
logical and Evolutionary Research. Oxford University Press, Oxford (in press).
De Block, M. and Stoks, R. (2004) Cannibalism-mediated life history plasticity to com-
bined time and food stress. Oikos 106, 587–597.
De Block, M. and Stoks, R. (2005) Fitness effects from egg to reproduction: bridging the
life history transition. Ecology 86, 185–197.
van den Bosch, F., DeRoss, A. and Gabriel, W. (1988) Cannibalism as a life boat mechanism.
Journal of Mathematical Biology 26, 619–633.
Diehl, S. (1988) Foraging efficiency in three freshwater fishes: effects of structural complexity
and light. Oikos 53, 207–214.
52 F. Johansson and P.H. Crowley

Dudgeon, D. and Wat, C.Y.M. (1986) Life cycle and diet of Zygonox iris insignis (Insecta:
Odonata: Anisoptera) in Hong Kong running waters. Journal of Tropical Ecology 2,
73–85.
Elgar, M.A. (1992) Sexual cannibalism in spiders and other invertebrates. In: Elgar, M.A.
and Crespi, B.J. (eds) Cannibalism: Ecology and Evolution Among Diverse Taxa.
Oxford University Press, New York, pp. 128–155.
Fincke, O. (1994) Population regulation of a tropical damselfly in the larval stage by food
limitation, cannibalism, intraguild predation and habitat drying. Oecologia 100,
118–127.
Fischer, Z. (1961) Cannibalism among the larvae of the dragonfly Lestes nympha Selys.
Ekologia Polska Seria B 7, 33–39.
Folsom, T.C. and Collins, N.C. (1984) The diet and foraging behaviour of the larval dragon-
fly Anax junius (Aeschnidae), with an assessment of the role of refuges and prey
activity. Oikos 42, 105–113.
Fox, L.R. (1975) Cannibalism in natural populations. Annual Review of Ecology and
Systematics 6, 87–106.
Freeland, W.J. (1983). Parasites and the coexistence of animal host species. American
Naturalist 121, 223–236.
Gurney, W.S.C., Crowley, P.H. and Nisbet, R.M. (1992) Locking life-cycles onto seasons:
circle-map models of population dynamics and local adaptation. Journal of Mathe-
matical Biology 30, 251–279.
Gurney, W.S.C., Crowley, P.H. and Nisbet, R.M. (1994) Stage-specific quiescence as a
mechanism for synchronizing life cycles to seasons. Theoretical Population Biology
46, 319–343.
Hopper, K.R., Crowley, P.H. and Kielman, D. (1996) Density dependence, hatching asynchrony,
and within-cohort cannibalism in young dragonfly larvae. Ecology 77, 191–200.
Houston, A.I., Clark, C.W., McNamara, J.M. and Mangel, M. (1988) Dynamic models in
behavioural and evolutionary ecology. Nature 332, 29–34.
Johansson, F. (1991) Foraging modes in an assemblage of odonate larvae – effects of
prey and interference. Hydrobiologia 209, 79–87.
Johansson, F. (1992) Effects of zooplankton availability and foraging mode on cannibalism
in three dragonfly larvae. Oecologia 91, 179–183.
Johansson, F. (1993) Intraguild predation and cannibalism in odonate larvae – effects of
foraging behaviour and zooplankton availability. Oikos 66, 80–87.
Johansson, F. (1996) The influence of cannibalism and prey density on growth in the
damselfly Coenagrion hastulatum. Archiv für Hydrobiologie 137, 523–535.
Johansson, F. (2000) The slow–fast life style characteristics in a suite of six species of
odonate larvae. Freshwater Biology 43, 149–159.
Johansson, F. and Rowe, L. (1999) Life history and behavioral responses to time
constraints in a damselfly. Ecology 80, 1242–1252.
Johnson, D.M., Crowley, P.H., Bohanan, R.E., Watson, C.N. and Martin, T.H. (1985)
Competition among larval dragonflies: a field enclosure experiment. Ecology 66,
119–128.
Lawton, J.H., Thompson, B.A. and Thompson, D.J. (1980) The effects of prey density on
survival and growth of damselfly larvae. Ecological Entomology 5, 39–51.
Merrill, R.J. and Johnson, D.M. (1984) Dietary niche overlap and mutual predation
among coexisting larval anisoptera. Odonatologica 13, 387–406.
Michaud, J.P. (2003) A comparative study of larval cannibalism in three species of lady-
bird. Ecological Entomology 28, 92–101.
Mock, D.W. and Parker, G.A. (1997) The Evolution of Sibling Rivalry. Oxford University
Press, Oxford.
Larval Cannibalism of Dragonflies 53

Mock, D.W. and Parker, G.A. (1998) Siblicide, family conflict and the evolutionary limits
of selfishness. Animal Behaviour 56, 1–10.
Norling, U. (1984) Life history patterns in the northern expansion of dragonflies.
Advances in Odonotalogy 2, 127–156.
Pfennig, D.W. (1992) Polyphenism in spadefoot toad tadpoles as a locally adjusted
evolutionarily stable strategy. Evolution 46, 1408–1420.
Pfennig, D.W. (1997) Kinship and cannibalism. Bioscience 47, 667–675.
Pfennig, D.W. (2000) Effect of predator–prey phylogentic similarity on the fitness conse-
quences of predation: a trade-off between nutrition and disease? American Naturalist,
155, 335–345.
Pfennig, D.W., Ho, S.G. and Hoffman, E.A. (1998) Pathogen transmission as a selective
force against cannibalism. Animal Behaviour 55, 1255–1261.
Polis, G.A. (1981) The evolution and dynamics of intraspecific predation. Annual Review
of Ecology and Systematics, 12, 225–251.
Pritchard, G. (1965) Prey capture by dragonfly larvae (Odonata: Anisoptera). Canadian
Journal of Zoology 43, 271–289.
Rowe, L. and Ludwig, D. (1991) Size and timing of metamorphosis in complex life cycles:
time constraints and variation. Ecology 72, 413–427.
Rudolf, V.H.W. and Antonovics, J. (2007) Disease transmission by cannibalism: rare event
or common occurrence? Proceedings of the Royal Society B 274, 1205–1210.
Schmitz, O.J., Beckerman, A.P. and O’Brien, K.M. (1997) Behaviorally mediated trophic
cascades: effects of predation risk in food web interactions. Ecology 78, 1388–
1399.
Sih, A. (1987) Predators and prey life style. In: Kerfoot, W.C. and Sih, A. (eds) Preda-
tion: Direct and Indirect Impacts on Aquatic Communities. University Press of New
England, Hanover, New Hampshire, pp. 203–224.
Stevens, L. (1994) Genetic analysis of cannibalism behavior in Tribolium beetles. In:
Boake, C.R.B. (ed.) Cannibalism: Ecology and Evolution Among Diverse Taxa.
University of Chicago Press, Chicago, pp. 206–227.
Sultan, S.E. and Spencer, H.G. (2002) Metapopulation structure favors plasticity over
local adaptation. American Naturalist 160, 271–283.
Thompson, D.J. (1978) The natural prey of the damselfly, Ischnura elegans (Odonata:
Zygoptera). Freshwater Biology 8, 377–384.
Thompson, D.J. (1987) Regulation of damselfly populations: the effects of weed density
on larval mortality due to predation. Freshwater Biology 17, 367–371.
Van Buskirk, J. (1989) Density-dependence cannibalism in larval dragonflies. Ecology 70,
1442–1449.
Van Buskirk, J. (1992) Competition, cannibalism, and size class dominance in a dragonfly.
Oikos 65, 455–464.
Van Buskirk, J. (1993) Population consequences of larval crowding in the dragonfly
Aeschna juncea. Ecology 74, 1950–1958.
Van Buskirk, J. and Yurewicz, K.L. (1998) Effects of predators on prey growth rate: relative
contributions of thinning and reduced activity. Oikos 82, 20–28.
Wagner, J.D., Glover, M.D., Moseley J.B. and Moore, A.J. (1999) Heritability and fitness
consequences of cannibalism in Harmonia axyridis. Evolutionary Ecology Research
1, 375–388.
Warfe, D.M. and Barmuta, L.A. (2006) Habitat structural complexity mediates food web
dynamics in a freshwater macrophyte community. Oecologia 150, 141–154.
Werner, E.E. and Anholt, B.R. (1993) Ecological consequences of the trade-off between
growth and mortality rates mediated by foraging activity. American Naturalist 142,
242–272.
54 F. Johansson and P.H. Crowley

Wise, D.H. (2006) Cannibalism, food limitations, intraspecific competition, and the regu-
lation of spider populations. Annual Review of Entomology 51, 441–465.
Wissinger, S.A. (1988) Effects of food availability on larval development and inter-instar
predation among larvae of Libellula lydia and Libellula luctuosa (Odonata:
Anisoptera). Canadian Journal of Zoology 66, 543–549.
4 The Ecology of Host–Parasite
Interactions in Aquatic Insects
STEVEN L. KOHLER
Department of Biological Sciences and Environmental Studies Program,
Western Michigan University, Kalamazoo, MI 49008, USA
E-mail: [Link]@[Link]

Abstract
The importance of parasites in affecting characteristics at the individual, population, com-
munity and ecosystem levels in natural systems has been increasingly recognized and
documented. A literature survey of the interactions between parasites and aquatic
insects revealed strong taxonomic biases with respect to both hosts and parasites stud-
ied. Some relatively speciose and ecologically important groups (e.g. aquatic Coleoptera,
Ephemeroptera, Plecoptera, Trichoptera) were poorly represented while others (e.g. Odo-
nata, Culicidae) dominated the literature. Similarly, some groups of aquatic insects appear
to be parasitized by relatively few parasite groups, and some parasites were studied from
a narrow range of hosts. Studies on the ecological effects of parasites were dominated by
those addressing effects on host fitness, but utilized a narrow range of hosts (largely
odonates and mosquitoes) and parasites (largely gregarines, microsporidia, nematodes
and water mites). Less than 5% of the 551 papers surveyed addressed population- or
community-level effects of parasites. Because all host–parasite interactions take place
within the framework of a larger set of species interactions occurring in a community, I
discuss examples that address the context-dependent effects of parasites, and their effects
at the population and community levels.

Introduction

It is an exciting time to be interested in the ecology of host–parasite interac-

tions. Since the early 1990s, there has been a tremendous increase in interest
in the effects of parasites on hosts, effects of hosts on parasites and interactions
between parasites. Following the appearance of several highly influential
models of host–parasite population dynamics that showed that, under realis-
tic conditions, parasites should commonly have strong effects on host popu-
lations (Anderson and May, 1978, 1981), there has been a growing body of
theory and empirical examples demonstrating the importance of parasites at
the individual and population levels (Hudson, 2005; Hudson et al., 2006).
 CAB International 2008. Aquatic Insects: Challenges to Populations
(eds J. Lancaster and R.A. Briers) 55
56 S.L. Kohler

Recent reviews (Thomas et al., 2005b; Collinge and Ray, 2006) suggest that
the study of parasites is even becoming part of mainstream community and
ecosystem ecology.
By contrast, general reviews of aquatic insect ecology (Resh and Rosenberg,
1984; Ward, 1992; Williams and Feltmate, 1992; Allan, 1995; Hershey and
Lamberti, 2001) either largely ignore host–parasite interactions or discuss para-
sitism in the context of aquatic insects that are themselves parasites or parasitoids
(e.g. the aquatic Hymenoptera). Thus, although several reviews focused on par-
ticular parasite groups or host–parasite systems are available (e.g. Hominick and
Welch, 1980; Molloy, 1981, 1988; Beier and Craig, 1985; Smith, 1988; Rolff,
2001; Smith et al., 2001), it seemed appropriate to take a broader view of the
literature here. In this chapter, I first survey the literature to identify general pat-
terns in the nature of research conducted on host–parasite interactions in aquatic
insects. I then highlight examples of research focusing on effects of parasites at the
individual and, especially, population and community levels. I conclude with some
recommendations for future studies.

Literature Survey

The Web of Science was used to survey the literature on host–parasite interac-
tions in aquatic insects. Initially, I used a search strategy such as: (ephemeroptera
OR mayfl*) AND (parasit* OR diseas* OR infect*). This was followed by a more
focused search on each of the major parasite groups encountered. These searches
were supplemented by examining references cited in many papers, and papers
that cited those works. These searches were performed for each of the wholly
aquatic orders (Ephemeroptera, Plecoptera, Odonata, Trichoptera and Mega-
loptera), and for the largely aquatic families in the partially aquatic orders
(Hemiptera, Coleoptera, Diptera, Lepidoptera and Neuroptera). I excluded
papers from the voluminous literature dealing with aquatic insect vectors of
human or wildlife diseases, or that dealt with the use of pathogens, such as Bacil-
lus thuringiensis in biocontrol programmes. I categorized the resulting papers by
host group (order or family), parasite group (e.g. microsporidia) and by the
nature of the study. Study type categories included: description, prevalence, sev-
eral categories dealing with effects of the parasite(s) on host fitness (growth,
condition, survivorship, fecundity, morphology, physiology, behaviour, mating),
host range, host immune response, context-dependent effects, population-level
effects, community-level effects, parasite behaviour, parasite fitness and coevolu-
tion. Most papers were classified as representing two or more of these categories.
Studies classified as ‘descriptive’ often described one or more new parasite spe-
cies and presented information on the parasite’s life cycle or nature of the host–
parasite relationship. Studies classified as ‘prevalence’ presented information on
the proportion of the host population infected by a parasite. To be classified as
presenting ‘population-level effects’ the paper needed to present information,
often time-series data, on both parasite prevalence and host population size.
A total of 551 papers was considered to satisfy the search criteria and was
included in the survey. However, this list of papers is far from exhaustive. In
Host–Parasite Interactions 57

particular, a number of papers cited in World Health Organization summaries of

pathogens of biting flies (e.g. Clark, 1977; Federici, 1977; Hazard and Chapman,
1977; Strand et al., 1977) did not appear in my searches. None the less, most of
those papers present descriptive information and little, if any, information on
ecological effects on host individuals or populations.
There was appreciable variation among aquatic insect orders/families in the
number of papers reporting information on host–parasite interactions (Table 4.1).
For example, only 9.4% of the papers reported on interactions in Ephemeroptera,
Plecoptera and Trichoptera, three orders that are especially important in the
structure and function of running water ecosystems (Allan, 1995), figure promi-
nently in measures of stream ecosystem health (Wallace et al., 1996) and repre-
sent about 25% of aquatic insect species considered in this survey. The three
stonefly studies reported on somewhat atypical parasites: chironomid (Diptera)
larvae (Giberson et al., 1996; Doucett et al., 1999) and gordian worms (Nemato-
morpha) (Winterbourn, 2005). Even more striking is the dearth of studies on a
very speciose group, the Coleoptera. Sprague (1977) reported microsporidia
from only one stonefly and none from aquatic Coleoptera, but over 100 mos-
quito (Diptera: Culicidae) hosts, which suggests that my survey results for these
group may not be strongly biased. It is unclear whether the relative paucity of
reports on these groups indicates they are seldom parasitized, or that they have
been poorly studied with respect to host–parasite interactions. Given that most
workers expect that every species is host to at least one parasite (Dobson et al.,
2006), lack of study seems the most likely explanation. None of the papers
encountered in this survey reported the results of surveys to determine the
incidence of parasites in a broad range of potential hosts.
Similarly, some groups of aquatic insects appear to be parasitized by rela-
tively few parasite groups (Table 4.1). For example, 85% of the papers on odo-
nates reported on interactions with gregarines (phylum Apicomplexa) or water
mites (Arachnida), and in nearly all of those papers the host was a damselfly. In
another well-studied group, the black flies (Diptera: Simuliidae), a high propor-
tion of studies involved either nematodes, especially the family Mermithidae or
microsporidia. Finally, some parasites were studied from a narrow range of
hosts. Most studies on gregarines come from odonates or mosquitoes, and most
studies on viruses are from Diptera. By contrast, studies of microsporidia, nem-
atodes and water mites are reasonably well represented in most aquatic insect
groups.
There was considerable variation among parasites in the nature of studies
that were reported. Overall, 59% of the studies presented only descriptive infor-
mation or information on disease prevalence, but that varied substantially among
parasites (Table 4.2). A high proportion of studies on microsporidia fell into this
category, and that held true for nearly all host groups except the Culicidae. It
would be a mistake to dismiss such studies as not reporting information of eco-
logical interest, as several report substantial information on disease prevalence
(e.g. Andreadis, 1999, 2002) that can be useful in determining when and under
what conditions epizootics occur.
One hundred and sixty studies (27.5%) presented information on parasite
effects on host fitness (Table 4.3). 70% of these studies came from just four
 58
Table 4.1. The number of studies satisfying the literature search criteria by host and parasite taxonomic groups.

Parasite

World Total Trypanos- Micro- Nema- Water Chirono-

Taxon species papers Virus Bacteria Gregarine Ciliate omatid Fungus sporidia tode Helminth mite midae Other

Ephemeroptera 3,100 28 1 1 2 1 15 6 2
Plecoptera 2,000 3 1 2 1
Odonata 5,500 89 1 41 5 1 4 35 5
Hemiptera 2,600 27 1 1 4 1 2 18 1
Trichoptera 11,000 21 1 13 2 1 3
Megaloptera 300 5 1 1 2 1
Coleoptera 9,500 4 2 2

Diptera
Ceratopogonidae 5,300 13 3 3 2 2 6
Chironomidae 5,000 68 7 3 7 15 22 15
Culicidae 3,500 216 16 9 30 13 17 78 33 11 13
Simuliidae 1,800 62 12 2 9 17 33 1 1
Other Diptera >15,000 20 1 4 1 2 2 4 3 4

S.L. Kohler
 Host–Parasite Interactions
Table 4.2. The per cent of studies reporting only descriptive information or information on disease prevalence. The number of studies in each category is in
parentheses. The total number of studies (582) is greater than the total number of papers in Table 4.1 because some papers reported on >1 parasite group.

Parasite

Trypanos- Micro- Water Chirono-

Taxon Virus Bacteria Gregarine Ciliate omatid Fungus sporidia Nematode Helminth mite midae Other

Ephemeroptera 100 (1) 100 (1) 0 (2) 100 (1) 40 (15) 83 (6) 100 (2)
Plecoptera 100 (2) 100 (1)
Odonata 100 (1) 63 (41) 100 (5) 100 (1) 75 (4) 14 (35) 0 (5)
Hemiptera 100 (1) 0 (1) 25 (4) 100 (1) 100 (2) 44 (18) 100 (1)
Trichoptera 100 (1) 100 (1) 77 (13) 100 (2) 100 (1) 100 (3)
Megaloptera 0 (1) 100 (1) 50 (2) 100 (1)
Coleoptera 50 (2) 100 (2)
Diptera
Ceratopogonidae 67 (3) 100 (3) 0 (2) 100 (2) 67 (6)
Chironomidae 86 (7) 33 (3) 86 (7) 73 (15) 77 (22) 47 (15)
Culicidae 44 (16) 56 (9) 43 (30) 23 (13) 71 (17) 65 (78) 32 (33) 18 (11) 15 (13)
Simuliidae 83 (12) 100 (2) 56 (9) 88 (17) 79 (33) 0 (1) 100 (1)
Other Diptera 100 (1) 100 (4) 100 (1) 50 (2) 100 (2) 100 (4) 100 (3) 100 (4)

59
60 S.L. Kohler

parasite groups: water mites, nematodes, microsporidia and gregarines. Further-

more, nearly all (96%) of the studies addressing gregarine effects on host fitness
came from two host groups, odonates and mosquitoes, and 78% of the studies
on microsporidia are from mosquitoes. Mosquitoes account for 52% of the stud-
ies addressing affects of parasites on host fitness and odonates account for an
additional 19%. Twenty-six studies (4.5%) addressed the context dependence of
host–parasite interactions (see below), and nearly 60% of those studies were
conducted using mosquito hosts. Finally, <4% of studies addressed population-
or community-level effects of parasites.
This survey indicates that considerable effort in the study of host–parasite
interactions in aquatic insects has been directed at the effects of parasites on host
fitness and, to a substantially lesser degree, effects of the host on parasite fitness.
However, all host–parasite interactions take place within the framework of a
larger set of species interactions occurring in a community. Because it is clear
that both the strength and outcome of interactions can change as a function of
their biotic and abiotic context (Agrawal et al., 2007), it is important to address
host–parasite interactions in a broader context to better understand their impor-
tance for populations and communities. I next consider studies that have begun
to do this in aquatic insect–parasite systems.

Context-dependent Effects of Parasites

Most models of host–parasite population dynamics assume that parasite viru-

lence (the rate of parasite-induced mortality, often represented as a) is a constant
that does not vary with prevailing environmental conditions (Anderson and May,
1981, but see their model E where this assumption is relaxed). Currently there is
considerable interest in whether the response of hosts to parasite infections var-
ies with environmental conditions, such as host food availability, conditions in
the habitat, such as temperature, and host sex (Sheridan et al., 2000), and this
interest is reflected in several recent studies involving aquatic insects. Most of
these studies have examined whether food availability for hosts influences effects
of a parasite on one or more fitness-related attributes of hosts, such as survivor-
ship and adult size. The studies manipulate host food availability (or host density,
which should affect per capita food availability) and parasite infection level in
factorial designs, and test for a significant interaction between food and infec-
tion. These studies, most of which have been conducted in the laboratory, can
be placed in three groups: those that find condition-dependent effects for all
response variables, those that find condition-dependent effects for some but not
all host attributes, and those that did not detect any condition-dependent effects.
Agnew and Koella (1999a,b) provide good examples of the first group. Both
studies examined effects of the microsporidian Edhazardia aedis on the mos-
quito Aedes aegypti in laboratory experiments. In both cases they found that the
magnitude of parasite effects increased as food availability decreased. Several
other studies have found consistent condition-dependent effects of food availability
(Arnqvist and Maki, 1990; Braune and Rolff, 2001; Nguyen et al., 2002). More
commonly, studies find condition-dependent effects for some response variables,
 Host–Parasite Interactions
Table 4.3. The number of studies reporting results in one or more study categories (see text for a description of the categories). The total
number of studies (582) is greater than the total number of papers in Table 4.1 because some papers reported on >1 parasite group.

Host Host Immune Context Parasite Parasite

Parasite fitness range response dependence Population Community behaviour fitness

Virus 8 1 3
Bacteria 5 1
Gregarine 24 3 5 7 1
Ciliate 12 1 1 1
Trypanosomatid 2 2
Fungus 8 1 3
Microsporidia 23 4 1 6 7 1 4
Nematode 26 5 4 3 3 3
Helminth 10 2 2 1 1
Water mite 39 1 8 3 2 4 6
Other 3 4 1

61
62 S.L. Kohler

but not all. For example, Comiskey et al. (1999) found that a gregarine parasite
significantly reduced survivorship of the mosquito Aedes albopictus only when
mosquitoes experienced low food availability. All other host attributes (emer-
gence time, adult size, female fecundity) were significantly affected by food avail-
ability and infection status, but there were no interactive effects. Studies reporting
similar results include: Lanciani (1975), Walker et al. (1987), Bedhomme et al.
(2004) and Tseng (2004). Finally, only three studies reported condition-independent
effects of food availability on host responses to parasites: Nasci et al. (1992),
Klingenberg et al. (1997) and Agnew et al. (1999). Klingenberg et al. (1997)
found no effects of gregarine load on body size or developmental time in a water
strider, even when reared with low food availability. The other studies in this lat-
ter group found significant effects of both food level and parasite infection, but
no interactive effects. Studies addressing the effects of abiotic conditions on host–
parasite interactions are less common (Petersen and Willis, 1970; Mercer and
Anderson, 1994; Taylor and Merriam, 1996; Van Rhein et al., 2000; Gurski and
Ebbert, 2003). Only Mercer and Anderson (1994) employed an experimental
design allowing condition-dependent effects to be detected but, unfortunately,
only two individuals became infected in their field experiment, which prevented
meaningful tests of condition-dependent effects. Collectively, these studies suggest
that much greater attention should be paid to the ecological context of host–
parasite interactions involving aquatic insects.

Population-level Effects

Strong effects of a parasite on the dynamics of its host population imply, by

definition, that the parasite exerts strong effects on the fitness of individuals in
the population. However, even if strong effects of a parasite on the fitness of host
individuals can be demonstrated, it does not necessarily follow that the parasite
will have demonstrable effects on host population dynamics. The reason, of
course, is that the host will nearly always be embedded in a food web and, as a
consequence, direct and indirect interactions with other community members
may play a more important part in shaping the dynamics of the host than inter-
actions with the parasite (Washburn et al., 1991; Washburn, 1995).
The processes influencing host population dynamics can be complex, even
in seemingly simple communities that support few species. For example, the
population dynamics of mosquitoes whose larvae develop in containers (e.g.
treeholes and other phytotelmata, tyres) are influenced by several processes
affecting larvae directly and indirectly, such as intra- and interspecific competi-
tion (which is influenced by larval density and food availability) and the presence
of natural enemies (predators, parasites). Long-term dynamics are also influ-
enced by periodic recruitment of new individuals into the population, which
depends on female oviposition into the habitat. This process is also complex, as
it is influenced by larval density, e.g. females oviposit preferentially in habitats
containing an intermediate density of conspecific larvae (Zahiri and Rau, 1998),
and by whether an oviposition repellent is being produced in the habitat, either
as a result of larvae being stressed (by competition) or infected by parasites
Host–Parasite Interactions 63

(Lowenberger and Rau, 1994; Zahiri et al., 1997; Zahiri and Rau, 1998; but see
Reeves, 2004).
Schwab et al. (2003) performed a large-scale laboratory experiment in
which several of these factors were manipulated simultaneously. They followed
populations of Aedes aegypti for 14 days where adult females had the choice of
ovipositing in containers that either contained larvae or did not. Containers
with larvae were exposed to no, low or high levels of cercariae from a digenean
trematode parasite, and all such containers received high levels of food to min-
imize intraspecific competition. They found strong effects of the parasite on
Aedes population dynamics, largely achieved through much higher pupal mor-
tality in the parasite treatments relative to the control, which resulted in substan-
tially lower production of adults in the parasite treatments. Interestingly, females
oviposited preferentially in habitats containing larvae in the final days of the
experiment but, in contrast to other studies, did not discriminate among the
parasite treatments.
Observations on the population dynamics of hosts in natural settings are
made under much less controlled conditions than can be achieved in the labo-
ratory, which affects what inferences can be drawn regarding the role of para-
sites in host population dynamics. Observations of host population collapses in
the absence of accompanying data on parasite prevalence in the population
(e.g. Hunter, 1968) or of disease epizootics without accompanying data on
changes in host population density (e.g. Glenn and Chapman, 1978), while
suggestive, do not demonstrate causation. Patterns in changes in host popula-
tion density over time along with changes in parasite prevalence in the popula-
tion has been used repeatedly to argue that parasites regulate host population
dynamics (e.g. Molloy, 1981; Majori et al., 1986) but, as discussed below, such
data must also be interpreted with caution. Because the majority of population-
level data on host–parasite interactions in aquatic insects are of this form, I will
briefly consider patterns that have been observed, and then explore more gen-
erally the problem of inferring the role of parasites in the dynamics of host
populations.
Micieli et al. (2001) reported on dynamics in an interesting system in which
the parasite (the microsporidian Amblyorspora albifasciati) has a complex life
cycle with two hosts, requiring that both host populations and disease prevalence
in both hosts be monitored. Amblyospora is transmitted vertically from infected
adult females to their progeny. These larvae develop lethal infections, releasing
thousands of meiospores into the water. The meiospores are infective to the
copepod Mesocyclops annulatus and produce lethal infections in adult females,
which also release spores into the environment upon their death. These spores
are infective to mosquito larvae, producing benign infections targeting female
gonadal tissue. Micieli et al. monitored the dynamics of this system in an ephem-
eral floodwater habitat for one year, during which eight mosquito broods were
produced as a result of periodic drying and filling of the habitat. They observed
disease epizootics in the copepod, but not in the mosquito. Transovarial trans-
mission was detected in five of eight broods, but disease prevalence in vertically
infected larvae never exceeded 7%. Despite the occurrence of disease epizootics
in the copepods, horizontal transmission from copepods to mosquitoes was
64 S.L. Kohler

detected only twice and appears to be the main factor limiting the impact of the
disease on the mosquito population (see also Larkin et al., 1995). The authors
concluded that the disease may significantly affect the copepod population, but
not the mosquito. Copepods were largely absent from the system from Septem-
ber through May, following a period (May–September) where several major epi-
zootics occurred. Circumstantial evidence supporting this conclusion was that, in
50 other pools in the area and that were monitored over the same time period,
no infections in Mesocyclops were observed and copepods were abundant
throughout the year.
Especially striking dynamics were observed in a somewhat atypical system:
the cholera bacterium (Vibrio cholerae) associated with egg masses of Chironomus
sp. (Diptera: Chironomidae) (Halpern et al., 2006). Vibrio cholerae occurs natu-
rally as a free-living organism in aquatic habitats and has been found associated
with zooplankton, especially copepods. An additional natural reservoir for V.
cholerae is chironomid egg masses (Halpern et al., 2004) and a haemagglutinin/
protease enzyme secreted by the bacteria can degrade the egg masses, prevent-
ing the eggs from hatching (Halpern et al., 2003). Halpern et al. (2006) moni-
tored chironomid egg mass density and V. cholerae in the egg masses at two
sites, a river and a waste stabilization pond. A striking pattern was observed at
both sites: increases in chironomid egg mass density were followed by increases
in V. cholerae density in the egg masses and a subsequent reduction in egg mass
density (Fig. 4.1), suggesting that V. cholerae was driving the dynamics of midge
egg masses and, hence, the dynamics of the midge populations.
Similar dynamics, although not always quite so striking, have been observed
in a wide variety of systems and a diverse array of hosts and parasites: an ento-
mopoxivirus in chironomid larvae in flood-control channels (Harkrider and Hall,
1978); nematodes in black fly larvae in streams (Phelps and DeFoliart, 1964;
Takaoka, 1981); a fungus in ceratopogonid larvae (Diptera: Ceratopogonidae)
in canal beaches (Wright and Easton, 1996); a microsporidian in mosquito and
copepod hosts in bromeliad leaf axils (Micieli et al., 2007); an iridescent virus in
black fly larvae in streams (Hernandez et al., 2000); and a baculovirus in mosqui-
toes in ponds (Becnel et al., 2001). All of these cases are similar to the examples
described above in greater detail in that they are, at best, suggestive that the
parasite plays a major role in the dynamics of its host population(s). I will now con-
sider approaches that can provide stronger evidence that a parasite is responsible
for driving the dynamics of its host.
If a parasite plays a major role in regulating the population dynamics of its
host, we should observe: (i) strong density dependence in the dynamics of the
host, and (ii) strong evidence of density-dependent parasitism (Murdoch, 1994).
Kohler and Hoiland (2001) monitored the population dynamics of the univoltine
caddisfly Brachycentrus americanus (Trichoptera: Brachycentridae) and the preva-
lence of a microsporidian disease in the population for 15 years. The microspo-
ridian is an undescribed species, tentatively placed in Microsporidium, which,
based on its SSU rDNA gene sequence, appears to be more closely related to
Cougourdella (Microsporidia: Cougourdellidae) than any other microsporidia.
The disease attacks the fat body and eventually spreads to other tissues, resulting
in host death, generally before pupation. The Brachycentrus population was
Host–Parasite Interactions 65

1000 (a) June, 17 600

900 Sep., 8
500
800
700 400

CFU per EM
EM per plate

600
500 Nov., 6 300
400
200
300
200 100
100
0 0
0 5 10 15 20 25 30 35 40 45
Time (weeks)

250 (b) 900

800
June, 17
200 700
600
EM per plate

CFU per EM
150
500
400
100
300

50 200
100
0 0
0 5 10 15 20 25 30 35 40 45
Time (weeks)

Egg masses per plate [Link] (CFU) per egg mass

Fig. 4.1. Chironomus sp. population dynamics, as the number of egg masses (EM)
per Styrofoam plate, and Vibrio cholerae population dynamic, expressed as colony
forming units (CFU) per Chironomus sp. egg mass, in (a) a waste stabilization pond,
and (b) the Kishon River, Israel between April 2003 and January 2004 (week 41).
From Halpern et al. (2006).

censused annually near the time of peak pupation (pupation and adult emer-
gence are both fairly synchronous), while disease prevalence and Brachycentrus
density were also measured frequently in each generation to determine when
epizootics occurred and to obtain weighted-average estimates of prevalence in
each generation. Both the host and its pathogen appeared to exhibit cyclical
dynamics over the 15-year period (Fig. 4.2), and statistical analyses of both time
series supported this conclusion. For example, the autocorrelation function for
the detrended time series of (ln-transformed) Brachycentrus population density
exhibited a significant negative spike at a lag of two generations and a positive
spike at a four-generation lag, suggesting that the population exhibited cyclic
66 S.L. Kohler

80
500
Density
Prevalence
Density (no./m2) 400 60

Prevalence (%)
300
40
200

100 20

0
0
1984 1986 1988 1990 1992 1994 1996 1998 2000
Generation

Fig. 4.2. Brachycentrus americanus population density (mean ± 95% confidence

interval) in 15 consecutive generations (years) and the mean prevalence of
Microsporidium in Brachycentrus larvae and pupae within each generation.
Prevalence is the mean over all sampling dates in a generation, weighted by
Brachycentrus population density on each date. Error bars are 95% confidence
limits. Figure redrawn from Kohler and Hoiland (2001).

oscillations in density with a period of about four generations. The autocorrela-

tion function for Microsporidium disease prevalence behaved similarly. A set of
well-accepted statistical procedures provided strong evidence for density depen-
dence in Brachycentrus population growth from the annual census data. While
all tests provided evidence for significant direct density dependence, even stron-
ger support was provided for the hypothesis of delayed density dependence with
a lag of one generation (i.e. density in the previous generation significantly
affected per capita population growth rate in the next generation). Finally, Kohler
and Hoiland (2001) found no relationship between disease prevalence in a gen-
eration and Brachycentrus population density at the start of the generation, but
disease prevalence in a generation increased markedly with increasing popula-
tion size at the start of the previous generation (Fig. 4.3). Thus they observed
strong evidence of delayed density-dependent parasitism with a lag of one gen-
eration, which is consistent with the pattern of density dependence in the host.
Overall, their observations provided strong support for the hypothesis that the
microsporidian disease is largely responsible for driving observed Brachycentrus
population dynamics in this system.
Ultimately, the strongest tests of the role of parasites in affecting the dynam-
ics of host populations will come through controlled manipulations of parasites
(Hudson et al., 1998; Tompkins and Begon, 1999; Redpath et al., 2006) but, in
many host–parasite systems, such experiments will be difficult if not impossible
to perform, except in the laboratory. In lieu of controlled experimental manipula-
tions in field settings, natural experiments, in which dynamics in systems with
Host–Parasite Interactions 67

70
(a)

Prevalence in generation t
60

50

40

30

20

10
0 500 1000 1500 2000 2500
70
(b)
Prevalence in generation t + 1

60

50

40

30

20

10
0 500 1000 1500 2000 2500
Initial density in generation t

Fig. 4.3. The relationship between disease prevalence in a generation and

Brachycentrus population density (a) at the start of the generation, or (b) at the
start of the previous generation. Figure redrawn from Kohler and Hoiland (2001).

and without the parasite are compared, could be used much more frequently to
gain insight regarding the importance of parasites in host population dynamics.
Although the limitations of natural experiments are well known, their use in com-
bination with other approaches to causal analysis, such as structural equation
modelling, could provide a powerful alternative to controlled experiments when
manipulations cannot be performed.
An example of the use of natural experiments is provided by the studies of
the caddisfly Glossosoma nigrior (Trichoptera: Glossosomatidae) infected by the
microsporidian Cougourdella sp. in coldwater streams (Kohler and Wiley, 1992).
Glossosoma dynamics in ecosystems where Cougourdella is present are strik-
ingly different from those in systems where Cougourdella infections have not
been detected (Fig. 4.4). In systems where Cougourdella appears to be absent
(or Glossosoma is highly resistant to infection), Glossosoma populations are
relatively stable and do not exhibit marked fluctuations in size. In systems where
68 S.L. Kohler

Cougourdella infections have been detected, Glossosoma populations appear to

occur in either of two conditions or states: (i) a high density state where Cou-
gourdella prevalence is consistently low, and (ii) a low density state where the
population is maintained at low density by recurrent Cougourdella epizootics
(Fig. 4.4a). These striking differences in patterns of dynamics between systems
with and without the pathogen strongly suggest that the pathogen can play a
major role in driving Glossosoma population dynamics.
The time-series data in Fig. 4.4a are not long enough to address the interest-
ing question of whether, in the long term, Glossosoma populations in systems
supporting the pathogen will continue to oscillate between high and low density
states. To begin addressing this question, Kohler and Wiley (unpublished results)

summer generation
8000 100
(a)
Density

Cougourdella prevalence (%)

Glossosoma density (no./m2)

Prevalence 80
6000

60
4000
winter
40
generation

2000
20

0 0

(b)
Glossosoma density (no./m2)

4000

3000

2000

1000

0
1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Fig. 4.4. (a) Glossoma nigrior population density at the end of overwintering and
summer generations, and the prevalence of Cougourdella sp. in Glossosoma larvae
and pupae in Seven Mile Creek, Michigan. Error bars are 95% confidence limits.
(b) Glossosoma nigrior population density at the end of overwintering and summer
generations in two nearby streams where Cougourdella infections have not been
detected. Data are unavailable for these streams prior to 1992.
Host–Parasite Interactions 69

have applied a replicated, space-for-time substitution approach. They selected

62 streams at random from the >200 streams supporting Glossosoma in the
lower peninsula of Michigan and, beginning in 1991, sampled the streams at
generally 3-year intervals in the spring (at the end of Glossosoma’s overwinter-
ing generation), measuring Glossosoma population density and disease preva-
lence. In agreement with Fig. 4.4a, they found that none of the streams maintained
Glossosoma at high density (>1000/m2) between 1991 and 2004, and that <5%
of the populations remained at low density throughout the period. In all 3-year
periods, a high proportion of the populations either collapsed (declined in size by
at least an order of magnitude) or recovered (increased by at least an order of
magnitude). However, on average, in any given 3-year period, the proportion of
populations that collapsed was nearly balanced by the proportion recovering.
These patterns suggest that: (i) at the regional scale Glossosoma populations are
relatively stable (although they clearly fluctuate markedly at the scale of individ-
ual streams), and (ii) the long-term trajectory in any given stream may be one of
regular oscillations in state (i.e. from high density, low disease prevalence to
maintenance at low density by recurring pathogen epizootics). Because Cou-
gourdella’s complete life cycle is unknown (Heilveil et al., 2001), it is not possible
to test hypotheses regarding mechanisms responsible for state transitions (i.e. are
state transitions driven by coevolutionary dynamics?).

Community-level Effects

While a rich literature addresses effects of parasites at the individual and, to a

lesser extent, population levels, our understanding of how parasites affect com-
munity structure and ecosystem function is in its infancy (Marcogliese and Cone,
1997; Hudson, 2005). Dobson et al. (2006) suggest that ‘developing a deeper
understanding of the role that parasites and pathogens play in food webs is per-
haps the greatest empirical challenge in food web biology.’ Examples have
appeared only recently demonstrating the important roles of parasites in natural
communities, both in terms of food web topology (Thompson et al., 2005; Lafferty
et al., 2006a,b) and community structure (Kohler and Wiley, 1997; de Castro
and Bolker, 2005; Mouritsen and Poulin, 2005b; Mouritsen and Polml, 2006;
Wood et al., 2007). In this section, I consider examples of both of these approaches
(i.e. food web topology, direct and indirect interactions between species) of
addressing the roles of parasites at the community level.

Food web topology

Topological food webs depict the trophic relationships (who eats whom) among
consumers, producers and non-living material in ecosystems, while bioenergetic
webs focus on the flow of energy among species, and interaction webs identify
species and trophic links that most strongly influence community structure (Wine-
miller and Polis, 1996). Several statistics are used to describe topological webs (e.g.
link density, connectance, chain length, predator to prey ratio) and considerable
70 S.L. Kohler

effort has been directed to determine if general patterns in these metrics occur
among food webs. Parasites have only recently been incorporated into topologi-
cal webs (Huxham et al., 1995; Thompson et al., 2005; Lafferty et al., 2006a)
and no published examples include systems with parasite–aquatic insect links.
Recent work suggests that food webs are more sensitive, in terms of secondary
extinctions, to the loss of highly connected species, and that increased con-
nectance (the number of observed links relative to the number of possible links)
helps buffer systems against secondary loss of species from either random or
selective removal of species (Dunne et al., 2002). Lafferty et al. (2006a) found
that inclusion of parasites increased connectance, but the parasites in the largely
marine systems they analysed were dominated by helminths, which commonly
have complex life cycles involving several hosts (Lafferty et al., 2006b). It will be
interesting to see if this pattern holds in systems having greater insect representa-
tion (e.g. rivers, wetlands), as the life cycle characteristics of parasites in such
systems may be markedly different (Thompson et al., 2005).

Species interactions and community structure

In the topology approach to food webs, parasites are, like other species, nodes
in a food web. In the interaction web approach for addressing the effects of
parasites at the community level, a stronger emphasis is placed on the nature of
the interactions between parasites and their hosts and how those interactions can
affect other members of the community. The effects of a parasite on other com-
munity members will probably be least important when the dynamics of the
parasite are driven by its host (i.e. there are strong bottom-up effects of the host
on the parasite population, but not strong top-down effects of the parasite on the
host population; see Briggs et al. (1995) for several examples from terrestrial
insects where the host appears to drive dynamics of its pathogen, but not vice
versa). Effects of parasites on other community members should be strongest in
either of two situations. First, if the parasite has strong top-down effects on the
abundance of their hosts, even regulating host population dynamics (Anderson,
1979; Anderson and May, 1981; Hudson et al., 1998; Kohler and Hoiland,
2001), effects of parasites on their host can extend indirectly to other community
members. In such situations, parasites can assume roles in communities similar
to those of predators in density-mediated interactions, such as trophic cascades,
apparent competition and keystone parasitism (Thomas et al., 2005a; Holt and
Dobson, 2006). Second, by influencing traits of their hosts such as behaviour
(Mouritsen and Poulin, 2005a; Wood et al., 2007) and morphology (Mouritsen
and Poulin, 2005b), parasites may have significant trait-mediated indirect inter-
actions with other community members (Werner and Peacor, 2003).

Trait-mediated indirect interactions

One common form of trait-mediated indirect interaction occurs when two nat-
ural enemies share a resource (Werner and Peacor, 2003). By affecting traits
Host–Parasite Interactions 71

(e.g. behaviour) of the resource, one natural enemy can affect the other, even
without changing the abundance of the resource. Of course, this sort of interac-
tion is common with parasites, as parasites often influence the behaviour of their
hosts to facilitate their transmission to predators that serve as subsequent hosts
(Dobson, 1988). Two interesting cases involving aquatic insects that do not fit
the typical parasite transmission model bear mentioning.
Vance and Peckarsky (1997) studied the impact of a mermithid nematode
on the vulnerability of its host, the mayfly Baetis bicaudatus (Ephemeroptera:
Baetidae), to two invertebrate predators, a stonefly (Kogotus modestus) and a
caddisfly (Rhyacophila hyalinata). In streamside artificial streams, the stonefly
fed preferentially on parasitized Baetis larvae, while the caddisfly did not dis-
criminate between parasitized and uninfected prey. Parasitized Baetis were less
likely to evade an approaching stonefly by entering the water column and drift-
ing downstream, and thus were captured more frequently by the stonefly than
unparasitized larvae. Thus, the mermithid should have a positive indirect effect
on stoneflies in this system by increasing Baetis vulnerability through its effects
on Baetis behaviour. By contrast, the prevalence of nematode infections in the
mayfly Deleatidium was much greater in drifting individuals than in animals on
the stream bottom in a New Zealand stream, suggesting that nematode infection
increased the tendency of Deleatidium to drift (Williams et al., 2001). Because
drifting individuals are more vulnerable to drift-feeding predators such as trout
(Allan, 1978), the mermithid should facilitate trout by increasing the vulnerability
of its prey. However, in both systems, because the predator facilitated by the
nematode is not used by the nematode as a host, the changes in prey behaviour
induced by the nematode appear to have negative fitness consequences for the
nematode (because the nematode is killed along with the mayfly when con-
sumed by a stonefly or trout). While drifting, parasitized larvae are more vulner-
able than uninfected larvae to trout (Vance, 1996) and, in general, trout account
for much greater mayfly mortality than do stoneflies (Kerans et al., 1995). There-
fore, from the parasite’s perspective, parasite-induced changes in prey behaviour
may reflect a trade-off between the risks of being eaten by stoneflies or trout
(Vance and Peckarsky, 1997).

Apparent competition

Apparent competition is frequently described as an indirect interaction between

prey species that share a predator, in which increased abundance of one prey
species enhances predation on the second species, and vice versa. Thus the
prey species appear to be competitors, even though they do not interact, either
directly or through shared resources (Holt, 1977). Of course, such indirect inter-
actions can also occur between hosts that share a generalist parasite (Holt and
Pickering, 1985) and the potential importance of such interactions in shaping
host community structure has received increasing interest (Hudson and Green-
man, 1998; Poulin, 1999; Hudson et al., 2006). None the less, to date only two
experimental studies of apparent competition in animal host–pathogen (exclud-
ing parasitoids) systems have been reported (Chaneton and Bonsall, 2000) and
72 S.L. Kohler

one of these involves aquatic insects (Copeland and Craig, 1992; see also Grosholz,
1992).
Copeland and Craig (1992) studied two closely related treehole mosquitoes,
Aedes triseriatus and A. hendersoni, that exhibit relatively strong niche segrega-
tion: A. hendersoni avoids ovipositing in sites near the ground and each species
specializes in the types of treeholes occupied. Each species is infected by a different
gregarine species, but cross-infections are known to occur with both parasites.
Copeland and Craig conducted laboratory experiments to determine if ongoing
interspecific competition between the mosquitoes accounted for the distribution
patterns and whether the outcome of competition was influenced by a shared
parasite. The competition part of the experiment followed a replacement series
design in which the total density of larvae in experimental chambers was held
constant. Food availability per larva was also held constant in their experiments
at a relatively high level. All of the Aedes species combinations were observed in
both the presence and absence of parasitism. The parasite used was Ascogre-
garina barretti, which occurs naturally in A. triseriatus. In the experiment, both
mosquito species were infected by the gregarine, but infection intensity (number
of trophozoites per larva) was much higher in A. triseriatus. However, gregarine
infection had no effect on A. triseriatus survivorship, the size of emerging adults
or female developmental rate, but there were highly significant effects on A.
hendersoni. Moreover, the competitive effects of A. triseriatus on A. hendersoni
were greater with the parasite present than in its absence. In other words, there
were significant interactive effects between competition and infection, especially
for female developmental rate and female size. By contrast, the presence of the
parasite did not alter the competitive effects of A. hendersoni on A. triseriatus.
These results illustrate complex indirect interactions between the mosquitoes.
When food availability was relatively high, A. triseriatus had strong competitive
effects on A. hendersoni, but only when the shared gregarine parasite was pres-
ent. Thus the species exhibit apparent competition but the interaction is highly
asymmetrical: only A. hendersoni was affected by the presence of the alternate
host species. Asymmetrical apparent competition has been observed more fre-
quently than symmetrical indirect interactions, in which each ‘prey’ species is
affected by the other species (Chaneton and Bonsall, 2000). The mechanisms
producing these effects in Copeland and Craig’s (1992) study are unclear. In
Grosholz (1992), apparent competition appeared to result from greater disease
prevalence in the presence of an alternate host (and potential competitor), but
such data were not reported by Copeland and Craig (1992).

Trophic cascades and keystone parasitism

Parasites may have keystone effects in a community if they strongly affect the
dynamics of other, strongly interacting species in the food web. Such effects have
been observed at large temporal and spatial scales in Michigan (USA) trout
streams. In these systems, the herbivorous caddisfly Glossosoma nigrior is a
particularly strong interactor, because foraging activity of its larvae maintains
the biomass (and productivity) of attached algae at very low levels throughout
Host–Parasite Interactions 73

the year, which results in Glossosoma having strong competitive effects on most
other primary consumers (Kohler, 1992). As described above, Glossosoma is
also host to a host-specific microsporidian microparasite, Cougourdella. Recur-
rent outbreaks of Cougourdella have resulted in whole-stream reductions in
Glossosoma population size by one to two orders of magnitude and maintained
Glossosoma density at low levels for years (Fig. 4a). Fortuitously, such outbreaks
have been observed in many streams that were already the subject of long-term
monitoring, allowing community-level effects of the pathogen to be assessed at
large scales. Pathogen-induced reductions in Glossosoma abundance resulted in
increased abundance of Glossosoma’s food resource (attached algae), and
increased population sizes of most other algal consumers, including grazers and
filter-feeders (Kohler and Wiley, 1997). Whole-stream reduction of Glossosoma
by the pathogen revealed competitive effects of Glossosoma that could not have
been anticipated from the results of more typical smaller-scale experiments (e.g.
Kohler, 1992). For example, several algal grazers (primarily other caddisfly spe-
cies) that had been extremely rare or absent, increased markedly following Glos-
sosoma’s decline, indicating that they had been competitively excluded from
these systems. Cougourdella’s effects in the community extend beyond the pri-
mary producer and primary consumer trophic levels. Because pathogen-induced
reduction in Glossosoma results in increased algal productivity and increased
abundances of relatively vulnerable prey (Glossosoma itself is relatively invulner-
able to invertebrate and vertebrate predators), presence of the pathogen should
facilitate top-level carnivores and perhaps intermediate-level carnivores as well.
Population dynamics of predacious invertebrates (intermediate-level carnivores)
suggests that their populations are often strongly resource-limited. Populations of
predacious caddisflies (Rhyacophila) and stoneflies (Paragnetina, Isoperla) have
increased over twofold following the collapse of Glossosoma populations in sev-
eral streams. Thus Cougourdella plays a keystone role in this system because it
strongly affects the abundance of a dominant competitor, Glossosoma. As a con-
sequence, the long-term dynamics of the community should be strongly tied to
the long-term dynamics of the host–parasite interaction, and will ultimately depend
on the, as yet, poorly understood mechanisms driving long-term patterns in the
interaction.

Conclusions

Hudson (2005) asks a series of insightful questions that should help frame future
research in parasite ecology in a broad range of systems: how might removal of
parasites affect community structure and the flow of energy and materials in an
ecosystem? How do parasites influence biodiversity in a system? How do para-
sites interact with other natural enemies and what are the consequences of these
interactions? How do reductions in biodiversity influence host–parasite interac-
tions? Addressing these questions in systems where aquatic insects play important
structural and functional roles (e.g. streams, wetlands, lake benthos and littoral
communities) will require attention to a few key areas. First, studies to date have
been highly skewed in terms of the taxonomic distribution of hosts and parasites
74 S.L. Kohler

studied. Greater attention to species that are suspected of interacting strongly

with other community members, or who have a relatively large number of links
with other species, would be beneficial. Second, more studies are needed that go
beyond asking whether a parasite affects host fitness and explore more broadly
the context dependence of host–parasite interactions, especially in field settings.
Third, our understanding of the population-level effects of parasites on aquatic
insects is in its infancy and remains a major challenge. Because controlled
manipulations are difficult to perform, progress in this area will often require a
well-designed combination of field observations and mathematical or statistical
modelling (e.g. structural equations). Finally, questions regarding the roles of
parasites in structuring communities and affecting the function of ecosystems
dominated by aquatic insects and other invertebrates are largely unexplored, but
observations in other systems (Mouritsen and Poulin, 2002, 2005a) suggest it
will be well worth the effort.

Acknowledgements

Beth Caplis helped with the literature review and Julie Ryan provided many
helpful comments on the manuscript. Mike Wiley has contributed immeasurably
to the work and ideas presented here. We have benefited tremendously from the
generous support of the National Science Foundation.

References

Agnew, P. and Koella, J.C. (1999a) Constraints on the reproductive value of vertical trans-
mission for a microsporidian parasite and its female-killing behaviour. Journal of
Animal Ecology 68, 1010–1019.
Agnew, P. and Koella, J.C. (1999b) Life history interactions with environmental conditions
in a host–parasite relationship and the parasite’s mode of transmission. Evolutionary
Ecology 13, 67–89.
Agnew, P., Bedhomme, S., Haussy, C. and Michalakis, Y. (1999) Age and size at maturity
of the mosquito Culex pipiens infected by the microsporidian parasite Vavraia culicis.
Proceedings of the Royal Society B 266, 947–952.
Agrawal, A.A., Ackerly, D.D., Adler, F., Arnold, A.E., Caceres, C., Doak, D.F., Post, E.,
Hudson, P.J., Maron, J., Mooney, K.A., Power, M., Schemske, D., Stachowicz, J.,
Strauss, S., Turner, M.G. and Werner, E. (2007) Filling key gaps in population and
community ecology. Frontiers in Ecology and the Environment 5, 145–152.
Allan, J.D. (1978) Trout predation and the size composition of stream drift. Limnology
and Oceanography 23, 1231–1237.
Allan, J.D. (1995) Stream Ecology: Structure and Function of Running Waters. Chapman
and Hall, New York.
Anderson, R.M. (1979) Parasite pathogenicity and the depression of host population
equilibria. Nature 279, 150–152.
Anderson, R.M. and May, R.M. (1978) Regulation and stability of host–parasite population
interactions: I. Regulatory processes. Journal of Animal Ecology 47, 219–247.
Anderson, R.M. and May, R.M. (1981) The population dynamics of microparasites and their
invertebrate hosts. Philosophical Transactions of the Royal Society B 291, 451–524.
Host–Parasite Interactions 75

Andreadis, T.G. (1999) Epizootiology of Amblyosopra stimuli (Microsporidiida: Ambly-

osporidae) infections in field populations of a univoltine mosquito, Aedes stimulans
(Diptera: Culicidae), inhabiting a temporary vernal pool. Journal of Invertebrate
Pathology 74, 198–205.
Andreadis, T.G. (2002) Epizootiology of Hyalinocysta chapmani (Microsporidia: Thelo-
haniidae) infections in field populations of Culiseta melanura (Diptera: Culicidae)
and Orthocyclops modestus (Copepoda: Cyclopidae): a three-year investigation.
Journal of Invertebrate Pathology 81, 114–121.
Arnqvist, G. and Maki, M. (1990) Infection rates and pathogenicity of trypanosomatid gut
parasites in the water strider Gerris odontogaster (Zett.) (Heteroptera: Gerridae).
Oecologia 84, 194–198.
Becnel, J.J., White, S.E., Moser, B.A., Fukuda, T., Rotstein, M.J., Undeen, A.H. and
Cockburn, A. (2001) Epizootiology and transmission of a newly discovered baculo-
virus from the mosquitoes Culex nigripalpus and C. quinquefasciatus. Journal of
General Virology 82, 275–282.
Bedhomme, S., Agnew, P., Sidobre, C. and Michalakis, Y. (2004) Virulence reaction
norms across a food gradient. Proceedings of the Royal Society B 271, 739–744.
Beier, J.C. and Craig, G.B. Jr. (1985) Gregarine parasites of mosquitoes. In: Laird, M.
and Miles, J.W. (eds) Integrated Mosquito Control Methodologies: Volume 2, Bio-
control and other innovative components, and future directions. Academic Press,
Orlando, Florida, pp. 167–184.
Braune, P. and Rolff, J. (2001) Parasitism and survival in a damselfly: does host sex matter?
Proceedings of the Royal Society B 268, 1133–1137.
Briggs, C.J., Hails, R.S., Barlow, N.D. and Godfray, H.C.J. (1995) The dynamics of
insect–pathogen interactions. In: Grenfell, B.T. and Dobson, A.P. (eds) Ecology of
Infectious Diseases in Natural Populations. Cambridge University Press, Cambridge,
pp. 295–326.
de Castro, F. and Bolker, B.M. (2005) Parasite establishment and host extinction in model
communities. Oikos 111, 501–513.
Chaneton, E.J. and Bonsall, M.B. (2000) Enemy-mediated apparent competition: empirical
patterns and the evidence. Oikos 88, 380–394.
Clark, T.B. (1977) Non-microsporidian protozoan pathogens of Culicidae (mosquitoes).
Bulletin of the World Health Organization 55 (Suppl. 1), 109–121.
Collinge, S.K. and Ray, C. (eds) (2006) Disease Ecology: Community Structure and
Pathogen Dynamics. Oxford University Press, New York.
Comiskey, N.M., Lowrie, R.C. and Wesson, D.M. (1999) Role of habitat components on
the dynamics of Aedes albopictus (Diptera: Culicidae) from New Orleans. Journal of
Medical Entomology 36, 313–320.
Copeland, R.S. and Craig, G.B. (1992) Interspecific competition, parasitism, and preda-
tion affect development of Aedes hendersoni and A. triseriatus (Diptera: Culicidae)
in artificial treeholes. Annals of the Entomological Society of America 85, 154–
163.
Dobson, A.P. (1988) The population biology of parasite-induced changes in host behavior.
Quarterly Review of Biology 63, 139–165.
Dobson, A.P., Lafferty, K.D. and Kuris, A.M. (2006) Parasites and food webs. In: Pascual,
M. and Dunne, J.A. (eds) Ecological Networks: Linking Structure to Dynamics in
Food Webs. Oxford University Press, New York, pp. 119–135.
Doucett, R.R., Giberson, D.J. and Power, G. (1999) Parasitic association of Nanocladius
(Diptera: Chironomidae) and Pteronarcys biloba (Plecoptera: Pteronarcyidae): insights
from stable-isotope analysis. Journal of the North American Benthological Society
18, 514–523.
76 S.L. Kohler

Dunne, J.A., Williams, R.J. and Martinez, N.D. (2002) Network structure and biodiversity
loss in food webs: robustness increases with connectance. Ecology Letters 5, 558–567.
Federici, B.A. (1977) Virus pathogens of Culicidae (mosquitoes). Bulletin of the World
Health Organization 55 (Suppl. 1), 25–46.
Giberson, D.J., MacInnis, A.J. and Blanchard, M. (1996) Seasonal frequency and position-
ing of parasitic midges (Chironomidae) on Pteronarcys biloba nymphs (Plecoptera:
Pteronarcyidae). Journal of the North American Benthological Society 15, 529–536.
Glenn, F.E. Jr and Chapman, H.C. (1978) A natural epizootic of the aquatic fungus
Lagenidium giganteum in the mosquito Culex territans. Mosquito News 38, 522–524.
Grosholz, E.D. (1992) Interactions of intraspecific, interspecific, and apparent competition
with host–pathogen population dynamics. Ecology 73, 507–514.
Gurski, K.C. and Ebbert, M.A. (2003) Host age, but not host location within a stream, is
correlated with the prevalence of gut parasites in water striders. Journal of Parasitology
89, 529–534.
Halpern, M., Gancz, H., Broza, M. and Kashi, Y. (2003) Vibrio cholerae hemagglutinin/
protease degrades chironomid egg masses. Applied and Environmental Microbiology
69, 4200–4204.
Halpern, M., Broza, Y.B., Mittler, S., Arakawa, E. and Broza, M. (2004) Chironomid egg
masses as a natural reservoir of Vibrio cholerae non-O1 and non-O139 in freshwater
habitats. Microbial Ecology 47, 341–349.
Halpern, M., Raats, D., Lavion, R. and Mittler, S. (2006) Dependent population dynamics
between chironomids (nonbiting midges) and Vibrio cholerae. FEMS Microbiology
Ecology 55, 98–104.
Harkrider, J.R. and Hall, I.M. (1978) The dynamics of an entomopoxvirus in a field
population of larval midges of the Chironomus decorus complex. Environmental
Entomology 7, 858–862.
Hazard, E.I. and Chapman, H.C. (1977) Microsporidian pathogens of Culicidae (mosqui-
toes). Bulletin of the World Health Organization 55 (Suppl. 1), 63–77.
Heilveil, J.S., Kohler, S.L. and Solter, L.F. (2001) Studies on the life cycle and transmission
of Cougourdella sp., a microsporidian parasite of Glossosoma nigrior (Trichoptera:
Glossosomatidae). Great Lakes Entomologist 34, 9–15.
Hernandez, O., Maldonado, G. and Williams, T. (2000) An epizootic of patent iridescent
virus disease in multiple species of blackflies in Chiapas, Mexico. Medical and
Veterinary Entomology 14, 458–462.
Hershey, A.E. and Lamberti, G.A. (2001) Aquatic insect ecology. In: Thorp, J.H. and
Covich, A.P. (eds) Ecology and Classification of North American Freshwater Inverte-
brates, 2nd edn. Academic Press, San Diego, pp. 733–775.
Holt, R.D. (1977) Predation, apparent competition, and the structure of prey communities.
Theoretical Population Biology 12, 197–229.
Holt, R.D. and Dobson, A.P. (2006) Extending the principles of community ecology to
address the epidemiology of host–pathogen systems. In: Collinge, S.K. and Ray, C.
(eds) Disease Ecology: Community Structure and Pathogen Dynamics. Oxford
University Press, New York, pp. 6–27.
Holt, R.D. and Pickering, J. (1985) Infectious disease and species coexistence: a model of
Lotka-Volterra form. American Naturalist 126, 196–211.
Hominick, W.M. and Welch, H.E. (1980) Mermithids (Nematoda) and mayflies (Ephe-
meroptera). In: Flannagan, J.F. and Marshall, K.E. (eds) Advances in Ephemeroptera
Biology. Plenum Press, New York, pp. 491–502.
Hudson, P. (2005) Parasites, diversity, and the ecosystem. In: Thomas, F., Renaud, F. and
Guégan, J.-F. (eds) Parasitism and Ecosystems. Oxford University Press, New York,
pp. 1–12.
Host–Parasite Interactions 77

Hudson, P. and Greenman, J. (1998) Competition mediated by parasites: biological and

theoretical progress. Trends in Ecology & Evolution 13, 387–390.
Hudson, P.J., Dobson, A.P. and Newborn, D. (1998) Prevention of population cycles by
parasite removal. Science 282, 2256–2258.
Hudson, P.J., Dobson, A.P. and Lafferty, K.D. (2006) Is a healthy ecosystem one that is
rich in parasites? Trends in Ecology & Evolution 21, 381–385.
Hunter, D.K. (1968) Response of populations of Chironomus californicus to a micro-
sporidian (Gurleya sp.). Journal of Invertebrate Pathololgy 10, 387–389.
Huxham, M., Raffaelli, D. and Pike, A. (1995) Parasites and food web patterns. Journal
of Animal Ecology 64, 168–176.
Kerans, B.L., Peckarsky, B.L. and Anderson, C. (1995) Estimates of mayfly mortality: is
stonefly predation a significant source? Oikos 74, 315–323.
Klingenberg, C.P., Barrington Leigh, R.H., Keddie, B.A. and Spence, J.R. (1997) Influence
of gut parasites on growth performance in the water strider Gerris buenoi (Hemiptera:
Gerridae). Ecography 20, 29–36.
Kohler, S.L. (1992) Competition and the structure of a benthic stream community. Ecological
Monographs 62, 165–188.
Kohler, S.L. and Hoiland, W.K. (2001) Population regulation in an aquatic insect: the role
of disease. Ecology 82, 2294–2305.
Kohler, S.L. and Wiley, M.J. (1992) Parasite-induced collapse of populations of a dominant
grazer in Michigan streams. Oikos 65, 443–449.
Kohler, S.L. and Wiley, M.J. (1997) Pathogen outbreaks reveal large-scale effects of
competition in stream communities. Ecology 78, 2164–2176.
Lafferty, K.D., Dobson, A.P. and Kuris, A.M. (2006a) Parasites dominate food
web links. Proceedings of the National Academy of Sciences of the USA 103,
11211–11216.
Lafferty, K.D., Hechinger, R.F., Shaw, J.C., Whitney, K. and Kuris, A.M. (2006b) Food
webs and parasites in a salt march ecosystem. In: Collinge, S.K. and Ray, C. (eds)
Disease Ecology: Community Structure and Pathogen Dynamics. Oxford University
Press, New York, pp. 119–134.
Lanciani, C.A. (1975) Parasite-induced alterations in host reproduction and survival.
Ecology 56, 689–695.
Larkin, T.S., Sweeney, A.W. and Carruthers, R.I. (1995) Simulation of the dynamics of a
microsporidian pathogen of mosquitoes. Ecological Modelling 77, 143–165.
Lowenberger, C.A. and Rau, M.E. (1994) Selective oviposition by Aedes aegypti
(Diptera: Culicidae) in response to a larval parasite, Plagiorchis elegans (Trematoda:
Plagiorchiidae). Environmental Entomology 23, 1269–1276.
Majori, G., Ali, A., Donelli, G., Tangucci, F. and Harkrider, R.J. (1986) The occurrence of
a virus of the Pox group in a field population of Chironomus salinarius Kieffer
(Diptera: Chironomidae) in Italy. Florida Entomologist 69, 418–421.
Marcogliese, D.J. and Cone, D.K. (1997) Food webs: a plea for parasites. Trends in Ecology
& Evolution 12, 320–325.
Mercer, D.R. and Anderson, J.R. (1994) Tannins in treehole habitats and their effects
on Aedes sierrensis (Diptera: Culicidae) production and parasitism by Lam-
bornella clarki (Ciliophora: Tetrahymenidae). Journal of Medical Entomology 31,
159–167.
Micieli, M.V., Garcia, J.J. and Andreadis, T.G. (2001) Epizootiological studies of Ambly-
ospora albifasciati (Microsporidiida: Amblyosporidae) in natural populations of
Aedes albifasciatus (Diptera: Culicidae) and Mesocyclops annulatus (Copepoda:
Cyclopidae) in a transient floodwater habitat. Journal of Invertebrate Pathology 77,
68–74.
78 S.L. Kohler

Micieli, M.V., Marti, G.A., Garcia, J.J., Tranchida, M.C. and Becnel, J.J. (2007) Epizooti-
ological studies of Amblyospora camposi (Microsporidia: Amblyosporidae) in Culex
renatoi (Diptera: Culicidae) and Paracyclops fimbriatus fimbriatus (Copepoda:
Cyclopidae) in a bromeliad habitat. Journal of Invertebrate Pathology 94, 31–37.
Molloy, D.P. (1981) Mermithid parasitism of black flies (Diptera: Simuliidae). Journal of
Nematology 13, 250–256.
Molloy, D.P. (1988) The ecology of black fly parasites. In: Kim, K.C. and Merritt, R.W.
(eds) Black Flies: Ecology, Population Management, and Annotated World List.
Pennsylvania State University Press, University Park, Pennsylvania, pp. 315–326.
Mouritsen, K.N. and Polml, R. (2006) A parasite indirectly impacts both abundance of
primary producers and biomass of secondary producers in an intertidal benthic
community. Journal of the Marine Biological Association of the UK 86, 221–226.
Mouritsen, K.N. and Poulin, R. (2002) Parasitism, community structure and biodiversity
in intertidal ecosystems. Parasitology 124, S101–S117.
Mouritsen, K.N. and Poulin, R. (2005a) Parasites boosts biodiversity and changes animal
community structure by trait-mediated indirect effects. Oikos 108, 344–350.
Mouritsen, K.N. and Poulin, R. (2005b) Parasitism can influence the intertidal zonation of
non-host organisms. Marine Biology 148, 1–11.
Murdoch, W.W. (1994) Population regulation in theory and practice. Ecology 75, 271–287.
Nasci, R.S., Tang, K.H., Becnel, J.J. and Fukuda, T. (1992) Effect of per os Edhazardia
aedis (Microsporida: Amblyosporidae) infection on Aedes aegypti mortality and
body size. Journal of the American Mosquito Control Association 8, 131–136.
Nguyen, D., Dutilleul, P. and Rau, M.E. (2002) Impact of nutrition and exposure to the
parasite Plagiorchis elegans (Trematoda: Plagiorchiidae) on the development of Aedes
aegypti (Diptera: Culicidae): analysis by time-dependent transition probabilities.
Environmental Entomology 31, 54–64.
Petersen, J.J. and Willis, O.R. (1970) Some factors affecting parasitism by mermithid
nematodes in Southern House mosquito larvae. Journal of Economic Entomology
63, 175–178.
Phelps, R.J. and DeFoliart, G.R. (1964) Nematode Parasitism of Simuliidae. Research
Bulletin 245, Agricultural Experiment Station, University of Wisconsin, Madison,
Wisconsin.
Poulin, R. (1999) The functional importance of parasites in animal communities: many
roles at many levels? International Journal for Parasitology 29, 903–914.
Redpath, S.M., Mougeot, F., Leckie, F.M., Elston, D.A. and Hudson, P.J. (2006) Testing
the role of parasites in driving the cyclic population dynamics of a gamebird. Ecology
Letters 9, 410–418.
Reeves, W.K. (2004) Oviposition by Aedes aegypti (Diptera: Culicidae) in relation to
conspecific larvae infected with internal symbiotes. Journal of Vector Ecology 29,
159–163.
Resh, V.H. and Rosenberg, D.M. (eds) (1984) The Ecology of Aquatic Insects. Praeger,
New York.
Rolff, J. (2001) Invited review – Evolutionary ecology of water mite–insect interactions: a
critical appraisal. Archiv für Hydrobiologie 152, 353–368.
Schwab, A.E., Lewis, D.J. and Rau, M.E. (2003) The impact of selective oviposition and
infection with Plagiorchis elegans on Aedes aegypti pre-imago population dynamics
at optimal food availability. Journal of Medical Entomology 40, 830–840.
Sheridan, L.A.D., Poulin, R., Ward, D.F. and Zuk, M. (2000) Sex differences in parasitic
infections among arthropod hosts: is there a male bias? Oikos 88, 327–334.
Smith, B.P. (1988) Host–parasite interaction and impact of larval water mites on insects.
Annual Review of Entomology 33, 487–507.
Host–Parasite Interactions 79

Smith, I.M., Smith, B.P. and Cook, D.R. (2001) Water mites (Hydrachnida) and other arachnids.
In: Thorp, J.H. and Covich, A.P. (eds) Ecology and Classification of North American
Freshwater Invertebrates, 2nd edn. Academic Press, San Diego, pp. 551–659.
Sprague, V. (1977) Systematics of the Microsporidia. In: Bulla, L.A., Jr and Cheng, T.C.
(eds) Comparative Pathobiology, Volume 2. Plenum Press, New York, pp. 1–510.
Strand, M.A., Bailey, C.H. and Laird, M. (1977) Pathogens of Simuliidae (blackflies). Bulletin
of the World Health Organization 55 (Suppl. 1), 213–237.
Takaoka, H. (1981) Further studies of pathogens of blackfly larvae in Guatemala: their
influence on natural populations of three species of Onchocerciasis vectors. In: Laird,
M. (ed.) Biocontrol of Medical and Veterinary Pests. Praeger, New York, pp. 78–104.
Taylor, P.D. and Merriam, G. (1996) Habitat fragmentation and parasitism of a forest
damselfly. Landscape Ecology 11, 181–189.
Thomas, F., Bonsall, M.B. and Dobson, A.P. (2005a) Parasitism, biodiversity, and conserva-
tion. In: Thomas, F., Renaud, F. and Guégan, J.-F. (eds) Parasitism and Ecosystems.
Oxford University Press, New York, pp. 124–139.
Thomas, F., Renaud, F. and Guégan, J.-F. (eds) (2005b) Parasitism and Ecosystems. Oxford
University Press, New York.
Thompson, R.M., Mouritsen, K.N. and Poulin, R. (2005) Importance of parasites and
their life cycle characteristics in determining the structure of a large marine food web.
Journal of Animal Ecology 74, 77–85.
Tompkins, D.M. and Begon, M. (1999) Parasites can regulate wildlife populations. Para-
sitology Today 15, 311–313.
Tseng, M. (2004) Sex-specific response of a mosquito to parasites and crowding. Proceedings
of the Royal Society B 271, S186–S188.
Van Rhein, S.L., Flanary, B.E. and Juliano, S.A. (2000) Effects of habitat type and drying
on Ascogregarina barretti (Eugregarinida: Lecudinidae) infection in Aedes triseritatus
(Diptera: Culicidae). Journal of Medical Entomology 37, 950–956.
Vance, S.A. (1996) The effect of the mermithid parasite Gasteromermis sp. (Nematoda:
Mermithidae) on the drift behaviour of its mayfly host, Baetis bicaudatus
(Ephemeroptera: Baetidae): a trade-off between avoiding predators and locating
food. Canadian Journal of Zoology 74, 1907–1913.
Vance, S.A. and Peckarsky, B.L. (1997) The effect of mermithid parasitism on predation of
nymphal Baetis bicaudatus (Ephemeroptera) by invertebrates. Oecologia 110, 147–152.
Walker, E.D., Poirier, S.J. and Veldman, W.T. (1987) Effects of Ascogregarina barretti
(Eugregarinida: Lecudinidae) infection on emergence success, development time,
and size of Aedes triseriatus (Diptera: Culicidae) in microcosms and tires. Journal of
Medical Entomology 24, 303–309.
Wallace, J.B., Grubaugh, J.W. and Whiles, M.R. (1996) Biotic indices and stream ecosystem
processes: results from an experimental study. Ecological Applications 6, 140–151.
Ward, J.V. (1992) Aquatic Insect Ecology. Wiley, New York.
Washburn, J.O. (1995) Regulatory factors affecting larval mosquito populations in con-
tainer and pool habitats: implications for biological control. Journal of the American
Mosquito Control Association 11, 279–283.
Washburn, J.O., Mercer, D.R. and Anderson, J.R. (1991) Regulatory role of parasites:
impact on host population shifts with resource availability. Science 253, 185–188.
Werner, E.E. and Peacor, S.D. (2003) A review of trait-mediated indirect interactions in
ecological communities. Ecology 84, 1083–1100.
Williams, D.D. and Feltmate, B.W. (1992) Aquatic Insects. CAB International, Wallingford.
Williams, J.K., Townsend, C.R. and Poulin, R. (2001) Mermithid nematode infections and
drift in the mayfly Deleatidium spp. (Ephemeroptera). Journal of Parasitology 87,
1225–1227.
80 S.L. Kohler

Winemiller, K.O. and Polis, G.A. (1996) Food webs: what do they tell us about the world?
In: Polis, G.A. and Winemiller, K.O. (eds) Food Webs: Integration of Patterns and
Dynamics. Chapman & Hall, New York, pp. 1–22.
Winterbourn, M.J. (2005) Dispersal, feeding and parasitism of adult stoneflies (Plecoptera)
at a New Zealand forest stream. Aquatic Insects 27, 155–166.
Wood, C.L., Byers, J.E., Cottingham, K.L., Altman, I., Donahue, M.J. and Blakeslee,
A.M.H. (2007) Parasites alter community structure. Proceedings of the National
Academy of Sciences of the USA 104, 9335–9339.
Wright, P.J. and Easton, C.S. (1996) Natural incidence of Lagenidium giganteum Couch
(Oomycetes: Lagenidiales) infecting the biting midge Culicoides molestus (Skuse)
(Diptera: Ceratopogonidae). Australian Journal of Entomology 35, 131–134.
Zahiri, N. and Rau, M.E. (1998) Oviposition attraction and repellency of Aedes aegypti
(Diptera: Culicidae) to waters from conspecific larvae subjected to crowding, con-
finement, starvation, or infection. Journal of Medical Entomology 35, 782–787.
Zahiri, N., Rau, M.E. and Lewis, D.J. (1997) Oviposition responses of Aedes aegypti and
Ae. atropalpus (Diptera: Culicidae) females to waters from conspecific and hetero-
specific normal larvae and from larvae infected with Plagiorchis elegans (Trematoda:
Plagiorchiidae). Journal of Medical Entomology 34, 565–568.
5 Effects of Drought on Stream
Insects and its Ecological
Consequences
ANDREW J. BOULTON1 AND P. SAM LAKE2
1Ecosystem Management, University of New England, Armidale, New South
Wales, Australia
2School of Biological Sciences, Monash University, Clayton, Victoria, Australia

E-mail: aboulton@[Link]

Abstract
Global climate change scenarios predict more frequent and extended droughts, espe-
cially in the mid-latitudes. For many stream insects, these increasing periods of
drought are likely to cause local if not regional extinctions, with potentially severe
consequences for stream ecosystem function. Drought is a steadily-intensifying ‘ramp’
disturbance that disrupts lateral, longitudinal and even vertical hydrological connec-
tivity, depending on the spatial patterns of drying and drought severity. These hydro-
logical changes affect many stream insects via impacts on refuges and water quality
interacting with changes wrought by human activities. Selective extirpation of specific
groups of insects during drought may alter benthic food web linkages or fundamental
ecosystem processes such as organic matter breakdown. Management of flowing
waters must consider these impacts, seeking to protect natural resilience (e.g. refuges)
and explicitly acknowledging how drought affects aquatic insect community composi-
tion and function. However, there is much still to learn about drought and its impacts
on stream insects. For example, we lack data from long-term studies that illustrate lag
effects or other repercussions of droughts. Factors affecting rates and trajectories of
post-drought recovery by stream insect communities are also poorly understood,
hampering assessments of resistance and resilience to this increasingly common
disturbance.

Introduction

Stream insects play central ecological roles in virtually all running waters. They
underpin in-stream, riparian and flood plain food webs, processing organic mat-
ter and transporting energy along stream channels, laterally to the flood plain
and even vertically into the stream bed (Hynes, 1970; Malmqvist, 2002). For
example, in many forested streams, aquatic insects mediate leaf litter breakdown,
supplying nutrients, carbon and energy to the stream and adjacent ecosystems
 CAB International 2008. Aquatic Insects: Challenges to Populations
(eds J. Lancaster and R.A. Briers) 81
82 A.J. Boulton and P.S. Lake

(Cummins and Klug, 1979; Williams and Feltmate, 1992). Their activities can
alter water quality, patch dynamics and flow patterns across multiple scales
(O’Connor and Lake, 1994; Wiens, 2002) while their biological interactions
often have significant repercussions on community structure (Williams and Feltmate,
1992; Williams, 2006). So ubiquitous and fundamental to riverine processes are
aquatic insects that their assemblage composition is routinely assessed as an
indicator of the ‘health’ of running waters (reviews in Wright et al., 2000; Bonada
et al., 2006a).
Natural disturbances such as flooding and drought also affect virtually all
running waters, impacting on aquatic biota and altering water quality (Resh
et al., 1988; Lake, 2000). Although the effects of floods on riverine fauna are
relatively well-documented (Lytle, 2000; Lake et al., 2006), far less is known
about the impacts of drought, despite the fact that prolonged declines in flow
or loss of water have severe consequences for freshwater ecosystems and
aquatic insects. Part of this ignorance stems from the practical difficulties of
studying ecological responses to drought; most studies to date have been ser-
endipitous, phenomenological assessments hampered by limited replication
and inappropriate design (Lake, 2003, 2006). It is difficult to predict when a
drought will commence, and the realization that it is occurring usually only
emerges some time after the disturbance has started. Droughts are also com-
plex ‘ramp’ disturbances (defined below) whose intensity, history, variability
and extent across multiple catchments confound attempts to generalize about
their impacts or to dissociate these from concomitant changes in catchment
condition and anthropogenic activities during the drought (Boulton, 2003;
Lake, 2003).

Forms of Drought and their Perturbation Dynamics

Given the plethora of definitions of drought (review in Wilhite, 2000), we define

it here as ‘an extended period (seasons to years) of deficit precipitation com-
pared to the statistical long-term average in a region’ (cf. Druyan, 1996). This
precipitation deficit is termed a meteorological drought when rainfall in a region
declines below some threshold determined from long-term records. Inevitably,
meteorological drought leads to hydrological drought, where the reduced pre-
cipitation fails to sustain normal volumes in standing waters or flows in rivers and
streams, typically defined at the basin or regional scale (Svoboda et al., 2002).
Hydrological drought can also arise from water deficits due to land use, and
water abstraction and storage. Groundwater drought occurs when groundwater
storages or hydraulic heads decline relative to those normally expected (van
Lanen and Peters, 2000) and reflects reduced groundwater recharge, also frequently
impacted by human activities.
Droughts start with meteorological drought followed by hydrological and
then groundwater droughts. This sequence depends on the duration of below-
average precipitation and losses through evapotranspiration, human extraction,
and other causes of water deficits. When a meteorological drought breaks, it
invariably takes longer before the hydrological drought breaks and much longer
Effects of Drought on Stream Insects 83

before the groundwater drought breaks, as there is a lag time required for recharge
of groundwater aquifers and recovery of flows (Lake, 2006).
A disturbance (cause) and the subsequent ecological response (effect)
together constitute a perturbation. In aquatic ecosystems, spates are seen as
pulse disturbances where the strength of the disturbing force (in this case, flow)
increases and declines as a pulse. In contrast, press disturbances arise sharply
and then plateau at a constant level, whereas ramp disturbances (Lake, 2000)
occur when the strength of disturbance increases steadily over time and space.
Droughts are a prime example of a ramp disturbance, intensifying over an
increasing area until broken by adequate precipitation. As droughts progress,
variations in rainfall and water use cause irregularities in the trajectory of the
ramp disturbance, some of which may coincide with the onset of hydrological
and groundwater drought. Ramp responses may also be irregular or even
‘stepped’ when certain critical thresholds (e.g. flow cessation, pools drying) are
crossed (Boulton, 2003), resulting in sharp changes in variables such as species
richness or the abundance of certain taxa. For example, mean numbers of taxa
of aquatic invertebrates from sites along two intermittent streams in Victoria,
Australia, declined abruptly when flow ceased and again when pools dried dur-
ing a severe drought (Boulton and Lake, 1992b). A similar stepped response in
invertebrate taxon richness was observed between the same phases during a
drought in a Spanish stream (Acuña et al., 2005).
Droughts can be classified as ‘seasonal’ or ‘supra-seasonal’ depending on
whether the drying is a reasonably predictable, periodic event versus longer, less
predictable and more destructive disturbances (Lake, 2003). Seasonal droughts
typify streams in regions such as the wet–dry tropics (e.g. Rincon and Cressa,
2000; Douglas et al., 2003) and Mediterranean climes (e.g. Gasith and Resh,
1999; Bonada et al., 2006b). In these areas, many streams are intermittent and
cease flow predictably (Gasith and Resh, 1999), selecting for adaptations by
the stream biota to survive the predictable waterless period. In contrast, supra-
seasonal droughts occur virtually anywhere, may persist over multiple seasons,
and their unpredictability militates against evolution of adaptive life-history
responses to drying; impacts on stream biota tend to be more severe during
supra-seasonal than seasonal drought (Boulton, 2003; Lake, 2003; but see Lind
et al., 2006).

Impacts of Drought on Stream Insects

Sequential losses of hydrological connectivity

The ramp disturbance of drought causes a sequence of changes in hydrological

connectivity and wetted habitat in streams that affects the aquatic (and poten-
tially terrestrial) stages of many stream insects. Depending on the severity and
intensity of the drought as well as the river’s size and shape, there will be a pro-
gressive decline in flow and volume that first severs lateral and then longitudinal
hydrological connectivity, ultimately resulting in complete drying of the stream-
bed (Fig. 5.1). During drying, water quality typically deteriorates, especially in
84 A.J. Boulton and P.S. Lake

Hydrological Habitat Biological

changes changes responses

Loss of Overall increases in density as

Decrease in
fast-flowing wetted habitat declines;
flow & volume
habitats loss of specialist torrenticoles

Declines in taxa that use stream-

Loss of lateral Isolate streamside side vegetation for emergence,
connectivity to vegetation/substrate food, shelter, case-material;
stream-edge habitats and floodplain loss of floodplain specialists

Sharp decline in taxa richness

Loss of longitudinal
Loss of due to loss of most rheophiles;
connectivity;
riffles colonization by some lentic
flow stops, pools form
Diptera, Coleoptera & Hemiptera

Progressive decline in intolerant

Pools shrink; Leaf litter builds up; lentic groups, e.g. some Odonata;
water quality DO, pH decline; predation intensifies;
deteriorates EC, temperature rice dominance by few tolerant taxa
Diptera, Coleoptera & Hemiptera

No free water; Loss of virtually all taxa

Pools dry; except some Diptera & Coleoptera;
slows leaf litter
sediments moist replacement by semi-aquatic taxa
breakdown

Sediments dry; Loss of all aquatic insects;

No saturated spaces
groundwater terrestrial insects, e.g. ants
near streambed surface
table declines occupy streambed

Fig. 5.1. Changes in stream habitat and biological responses during the progress-
ive hydrological stages of drought. DO = dissolved oxygen; EC = electrical conductiv-
ity. Table 5.1 lists representative Australian aquatic insect families impacted by these
stages of drought.

remnant pools along the streambed. Leaf litter accumulating in these deposi-
tional habitats accelerates declines in pH and dissolved oxygen (Towns, 1985;
Caruso, 2002), rendering the pools uninhabitable except by stream insects capa-
ble of respiring air directly (e.g. dytiscid beetles and corixid bugs; Gooderham
and Tsyrlin, 2002) or with physiological adaptations for enhanced uptake of
oxygen (e.g. haemoglobin used by some chironomid ‘bloodworms’; Hynes,
1970) (Table 5.1). When surface water completely disappears, a few stream
insects can refuge in moist sediments (see below) but, during especially severe
Effects of Drought on Stream Insects 85

Table 5.1. Occurrence of common families of Australian aquatic insects, listed by

sequential stages of habitat change as drought progresses (Fig. 5.1). The pool stage
is divided into early and final stages to reflect differing tolerances by insect families
to the deteriorating water quality; families present in the final stages are usually
present in early stages of drying too.

Habitat or habitat change: ecological group Family or subfamily

Decrease in flow/volume: Blephariceridae D

specialist torrenticoles or need fast-flowing, Ameletopsidae E
oxygenated water Coloburiscidae E
Siphlonuridae E
Eustheniidae P
Notonemouridae P
Hydrobiosidae T
Philorheithridae T

Loss of lateral connectivity: Hydraenidae C

uses stream-edge habitats or vegetation Thaumaleidae D
Pyralidae L
Lestidae O
Synlestidae O
Hydroptilidae T

Loss of longitudinal connectivity (flow ceases): Psephenidae C

Rheophiles Ptilodactylidae C
Empididae D
Simuliidae D
Baetidae E
Leptophlebiidae E
Corydalidae M
Neurorthidae N
Austroperlidae P
Gripopterygidae P
Notonemouridae P
Calocidae T
Conoesucidae T
Glossosomatidae T
Helicopsychidae T
Hydrobiosidae T
Hydropsychidae T
Leptoceridae T
Philopotamidae T
Philorheithridae T
Polycentropodidae T
Tasimiidae T

Early stages of pools shrinking: Elmidae C

moderately tolerant leutic taxa Hydraenidae C
(or extremely versatile lentic taxa) Hydrochidae C
Athericidae D
Tanypodinae D
(Continued)
86 A.J. Boulton and P.S. Lake

Table 5.1. continued

Habitat or habitat change: ecological group Family or subfamily

Early stages of pools shrinking: Orthocladiinae D

moderately tolerant lentic taxa Tipulidae D
(or extremely versatile lotic taxa) Tabanidae D
Dolichopodidae D
Psychodidae D
Baetidae E
Caenidae E
Leptophlebiidae E
Naucoridae H
Sialidae M
Sisyridae N
Coenagrionidae O
Isostictidae O
Lestidae O
Protoneuridae O
Synlestidae O
Hemicorduliidae O
Libellulidae O
Gomphidae O
Gripopterygidae P
Atriplectididae T
Calocidae T
Conoesucidae T
Ecnomidae T
Hydroptilidae T
Leptoceridae T
Odontoceridae T

Final stages of pools drying: Dytiscidae C

highly tolerant lentic taxa Gyrinidae C
Hydrophilidae C
Scirtidae C
Ceratopogonidae D
Chironominae D
Culicidae D
Ephydridae D
Muscidae D
Stratiomyidae D
Leptophlebiidae E
Belostomatidae H
Corixidae H
Gerridae H
Hydrometridae H
Nepidae H
Notonectidae H
Veliidae H

(Continued)
Effects of Drought on Stream Insects 87

Table 5.1. continued

Habitat or habitat change: ecological group Family or subfamily

Final stages of pools drying: Mesoveliidae H

highly tolerant lentic taxa Calamoceratidae T
Leptoceridae T

Pools dry: taxa with desiccation-resistant stages or Dytiscidae C

able to survive in moist stream bed Ceratopogonidae D
Tabanidae D
Simuliidae D
Gelastocoridae H
Gripopterygidae P

Order abbreviations are as follows: C, Coleoptera; D, Diptera; E, Ephemeroptera;

H, Hemiptera; L, Lepidoptera; M, Megaloptera; N, Neuroptera; O, Odonata; P, Plecoptera;
T, Trichoptera; and groups are arranged alphabetically.

drought, even this habitat desiccates as groundwater tables fall below the reach
of burrowing taxa.
The sequential local extinction of stream insects during drought is largely a
function of their resistance to changing abiotic and biotic conditions, although
other factors such as the initial species pool, whether the stream typically dries
and the extent of human disturbance are also important. The first aquatic insects
to be impacted are those requiring either very fast-flowing water, or at least cool,
well-oxygenated flowing water, as these conditions disappear when volume and
flow declines during the early stages of drought. The combination of reduced
current velocity and decreased oxygenation through loss of physical turbulence
soon eliminates torrenticoles such as blepharicerid dipteran larvae and several
families of mayflies, stoneflies and caddisflies (Table 5.1).
These groups usually inhabit spring-fed streams in montane areas that are
seldom subject to drought (Hynes, 1970) and take a long time to recolonize
when supra-seasonal drought causes local extinction (Williams, 2006). This long
recovery time is presumably because the spring heads are geographically iso-
lated and there has been little evolutionary selection for effective dispersal of
species restricted to this usually reliable water source. In especially severe drought,
there may be widespread drying across many springs (Erman and Erman, 1995),
reducing the pool of potential colonists and increasing the recolonization distances
of springs that had dried.
Aquatic insects of spring-fed systems appear especially vulnerable to ground-
water drought and discharge fluctuations. Erman and Erman (1995) sampled 21
cold springs in Sierra Nevada, California, for several years prior to an extended
drought and then re-sampled the springs after 4 and 6 years of drought. They
found two to 18 species of caddisflies per spring with nine species restricted to
constant-temperature spring sources (36 caddisfly species were collected over-
all). Before the drought, caddisfly species richness was positively correlated with
88 A.J. Boulton and P.S. Lake

spring discharge. However, after 6 years of drought, there was a stronger negative
relationship between species richness and long-term fluctuations in discharge. In
particular, springs that had few caddisfly species dried completely during the
drought, whereas water in the species-rich springs persisted (Erman and Erman,
1995).

Loss of lateral connectivity

Stream margins become exposed as flow and stream volume decline to critical
levels (Stanley et al., 1997), affecting insects that rely heavily on hydrological
linkages to lateral habitats, ranging from flood plain wetlands connected to the
river through to emergent streamside vegetation (Table 5.1). Stream insects that
use macrophytes along the edge of the channel to pupate and emerge, feed or
for shelter (Harrison, 2000) are the species most impacted when water recedes
from these habitats. Even mobile lentic taxa that scurry and hunt on the water’s
surface (e.g. veliid and gerrid water bugs) may be disadvantaged as they poten-
tially become more prone to predation from birds and fish when lateral fringing
vegetation is isolated from the water’s edge during the early stages of drying and
lateral contraction.
Loss of lateral connectivity in larger rivers and contraction of aquatic habitat
away from the stream margins also reduce prey subsidies and arthropod densi-
ties in the riparian zone (Ballinger and Lake, 2006), constraining this important
pathway for energy and nutrients. Conversely, the exposure of streambed sedi-
ments and lateral gravel bars may provide access for terrestrial insects. The ecology
of semi-aquatic insects of this lateral transition zone is poorly understood and
this would be a useful avenue for further research during seasonal and supra-
seasonal drought. Streams that dry regularly have a ‘clean-up’ crew of semi-
aquatic insects (e.g. ants, staphylinid and carabid beetles) that inhabit the dry
stream bed (Williams and Hynes, 1977), but it is unclear whether the composi-
tion of this community overlaps substantially with that of the insect assemblages
along the margins of permanent streams. If so, their populations may be favoured
by drought where increased habitat is available when water shrinks from stream
margins.

Loss of longitudinal connectivity

The most dramatic change occurs when flow ceases and pools form, a process
that may occur over several weeks across different river sections. Many stream
insects have aquatic stages that rely on flow to provide feeding currents, enhance
respiration, enable passive movement, import detritus, deliver prey, aerate water
through turbulence, and sustain biofilms for herbivores (Hynes, 1970). These
rheophilous (‘flow-loving’) groups (Table 5.1) soon disappear, unlikely to find
suitable refuges except in nearby flowing habitats within dispersal reach. Most
studies of drought on aquatic insects in streams have reported marked changes
in community composition when flow ceases (Williams and Hynes, 1977; Miller
Effects of Drought on Stream Insects 89

and Golladay, 1996; Filho and Maltchik, 2000; Acuña et al., 2005), primarily due to
the disappearance of nearly all the rheophiles and declines in many other species.
Boulton and Lake (1992a,b) tracked changes in taxon richness and compo-
sition of stream insect assemblages at four sites on two streams in Victoria, Austra-
lia, during seasonal drought. Although the sites varied markedly in flow duration,
there was a consistent sharp decline in taxon richness when flow ceased. This
was due to the loss of species of filter-feeding midge larvae (Chironomidae) and
net-spinning caddisflies, as well as many free-living caddisflies (e.g. hydrobio-
sids). There were also sharp declines in abundance of other groups of midge and
other dipteran larvae, mayfly and stonefly nymphs, and some odonates (Boulton
and Lake, 1992b). Similar marked declines in stream insect density and taxon
richness in response to seasonal flow cessation were described in a Spanish
stream (Acuña et al., 2005) for many of the same broad taxa.
The break in longitudinal connectivity also severs an important dispersal
pathway in most streams – invertebrate drift (Williams and Hynes, 1976) –
preventing recolonization of even remnant flowing sections. In many gravel-bed
streams, flow will cease in only a few reaches first yet continue elsewhere along
the stream, sustained by hyporheic discharge. None the less, stream-scale move-
ment of aquatic stages is inhibited, with major effects on predators that rely on a
supply of food imported from upstream (McIntosh et al., 2002). Furthermore, the
cessation of flow often promotes the settlement of fine sediment (Wright, 1992;
Wright and Symes, 1999). Siltation stresses many aquatic insects by coating their
respiratory surfaces and smothering their food sources, such as diatoms or
benthic organic matter (Angradi 1999; Wood and Armitage, 1999), slowing
post-drought recovery in some streams (Wood and Armitage, 2004).

Shrinking pools and ‘predator soup’

As the pools along the stream bed shrink, their worsening water quality imposes
a ramp disturbance on the remaining resistant denizens. Soon, the only taxa that
persist are those capable of tolerating extremely poor water quality. This highly-
tolerant group of stream insects that persists in the dwindling pools primarily
comprises beetles, true bugs and dipterans (Table 5.1), although some mayflies,
e.g. Atalophlebia (Leptophlebiidae), and case-building caddisflies, e.g. Lectrides
varians, some Triplectides (Leptoceridae), also have a surprising capacity to per-
sist in such stagnant warm water (Towns, 1983, 1985). Many of these tolerant
taxa are also highly mobile, able to escape from shrinking pools to seek larger
ones for refuge (Williams and Hynes, 1977; Boulton, 1989; Velasco and Millan,
1998).
The remnant pools also become a ‘predator soup’ where the biological inter-
actions of predation and competition intensify (Stehr and Branson, 1938; Hynes,
1975; Stanley et al., 1994), fuelled by diminishing wetted habitat and increasing
densities of stream insects. For example, Extence (1981) reported that contrac-
tion of wetted benthic habitat in a lowland English stream during drought con-
centrated prey densities and led to increases in numbers of predators. In drying
Australian streams, many of these predators apparently invade stream pools
90 A.J. Boulton and P.S. Lake

from nearby lentic sources such as farm dams and ponds (P.S. Lake, unpublished
data). When flow resumes, they are able to fly back to the original water bodies.

Refuge use in the dry stream bed

In the final stages of drying, an array of refuges may be used by stream insects
(Table 5.2). Remaining pools typically harbour the most taxa (Boulton, 1989),
but some stream insects such as midge larvae and small mayflies persist within
the burrow water of crayfish (as ‘pholeteros’ sensu Lake, 1977) or in moist
microhabitats under bark, in leaf litter, or under rocks. Although many stream
insects survive dry periods as terrestrial adults during prolonged supra-seasonal
droughts, even these individuals are likely to die as food resources decline, habitat

Table 5.2. Refuges (from Fig. 9.5 in Boulton and Brock, 1999) and examples of their use by
stream insects during seasonal and supra-seasonal drought. Taxa using remnant pools are
listed in Table 5.1.

Refuge Stream insect taxa Example references

Under bark and Mining insects such as some Boulton, 1989

in rotting wood Elmidae and Chironomidae
(e.g. Stenochironomus), Tipulidae

Moist microhabitats Adult Dytiscidae; larval Larimore et al., 1959;

in leaf litter Psephenidae, Tipulidae, Boumezzough, 1983;
Chironomidae, Leptoceridae Boulton, 1989

Burrow water of Collembola, Chironomidae, Lake, 1977; Boulton, 1989;

crayfish Ceratopogonidae Williams, 2006
(‘pholeteros’)

Hyporhoeic zone Plecoptera, Chironomidae, Tipulidae, Harper and Hynes, 1970;

and buried in moist Tabanidae, Ceratopogonidae, Gray and Fisher, 1981;
sediments Psychodidae Boulton et al., 1992;
Tronstad et al., 2005

Resting stages Eggs: Ephemeroptera, Chironomidae, Towns, 1983; Miller and

in dry sediments Simuliidae, Leptoceridae; Plecoptera Golladay, 1996; Fritz
Larvae/nymphs: Chironomidae, and Dodds, 2004; Tronstad
Plecoptera, Leptoceridae et al., 2005

Aerial adults Trichoptera, Ephemeroptera, Harrison, 1966; Williams

Hemiptera (e.g. Corixidae, and Hynes, 1977;
Notonectidae), Coleoptera Delucchi, 1989;
(e.g. Dytiscidae, Hydrophilidae), Paltridge et al., 1997;
Odonata, Plecoptera, Diptera Velasco and Millan, 1998;
(e.g. Chironomidae, Simuliidae Shivoga, 2001
Ceratopogonidae)
Effects of Drought on Stream Insects 91

is lost, and the waterless period starts to exceed the lifespan of aerial recolonizers
and egg-laying females.
Even after all surface water has disappeared, some aquatic insects can persist
in either moist sediments (e.g. dytiscid beetles, tabanid fly larvae; Gray and Fisher,
1981; Boulton, 1989) or within the fully saturated alluvial aquifer just below the
streambed (Boulton et al., 1992; Clinton et al., 1996; Table 5.2). In a Sonoran
Desert stream, soon after surface water had dried but hyporheic sediments
remained saturated, numerous ceratopogonid midge larvae (Bezzia and Probezzia
spp.) persisted in the interstitial water (Boulton and Stanley, 1995). However, in
a later study by Clinton et al. (1996) in the same stream, ceratopogonid and
chironomid larvae did not move deeper into the sediments during later stages of
drying, implying a limit to the use of even this refuge.

Ecological Consequences of Progressive Loss of Stream Insects

During Drought
Organic matter processing and carbon transfer

As stream insects are the most common and species-rich macroinvertebrate

component in nearly all headwater streams (Hynes, 1970; Williams and Feltmate,
1992), including those that dry regularly (Boulton and Lake, 1992a,b; Acuña
et al., 2005), they play a number of key ecological roles in stream ecosystems.
They are pivotal in the food web as prey for many vertebrates as well as feeding
on micro- and meso-benthos, microbes, algae and diverse biofilms. They also
process detritus and organic matter entering from forested catchments. In many
streams, functional feeding groups of detritivores termed ‘shredders’ comminute
leaf and other coarse particulate organic matter into smaller particles that sustain
‘collectors’ and ‘filterers’.
In forested streams, shredding insects are represented primarily by caddis-
flies, mayflies and stoneflies, most families of which are vulnerable to drought.
However, a few genera, e.g. Triplectides and Lectrides (Leptoceridae), have rep-
resentatives able to survive deteriorating water quality in pools in drying streams
in Australia where they shred accumulated leaves and wood. None the less, loss
of flow or water in most streams can substantially slow the breakdown of coarse
particulate organic matter, partly due to the local extinction of drought-intolerant
insect shredders (e.g. Kirby et al., 1983; Maamri et al., 2005), with major reper-
cussions for higher trophic levels and the conversion of detritus to energy to fuel
in-stream ecosystems (Acuña et al., 2005).
Many filterers rely on currents to feed passively (e.g. hydropsychid caddis-
flies) and are eliminated when stream flow ceases during drought, removing this
pathway of processing and uptake of fine particulate organic matter. However,
some actively filter-feeding stream insects, e.g. mosquito larvae (Culicidae), pro-
liferate in standing waters (Boulton, 1989) and tolerate poor water quality to
take advantage of the food resource provided by fine suspended organic matter
and bacteria in remnant pools. Similarly, some benthic insects that feed by ‘col-
lecting’ fine organic matter from the sediments (e.g. midge larvae in the genus
92 A.J. Boulton and P.S. Lake

Chironomus) thrive in stagnant waters, relatively free of competitors, and can

reach high abundances in drying pools (Boulton and Lake, 1992a; Acuña et al.,
2005). Despite this ability by some stream insects to persist, cessation of flow and
eventual drying markedly reduce the overall diversity of detritivorous insects in
streams during drought (Bogan and Lytle, 2007), restricting this heterotrophic
pathway that is so important in providing energy to the denizens of most forested
running water ecosystems.

Food web dynamics, aquatic–terrestrial linkages and succession

In-stream and riparian food web dynamics also change dramatically during the
sequence of hydrological changes that occur during drought. As streams dry and
pools become isolated, predation pressure intensifies (Stehr and Branson, 1938;
Hynes, 1975; Stanley et al., 1994). The decline in water quality soon eliminates
all but the hardiest fish, while voracious insect predators such as dytiscid and
hydrophilid beetles, dragonflies, and corixid, notonectid, nepid and veliid bugs
proliferate (Boulton and Lake, 1992a; Stanley et al., 1994; Acuña et al., 2005).
Predation by terrestrial and semi-aquatic insects also increases as they feast on
prey stranded by declining water levels; ants and some beetle families (e.g. Car-
abidae, Staphylinidae) are especially common in many drying stream beds
(Stanley et al., 1994; Williams, 2006).
The variable tolerance by aquatic insects to drying alters the succession pro-
cess when flow resumes and is likely to vary from year to year, even when drying
is a predictable, regular event (Gasith and Resh, 1999). Post-drought community
composition reflects the type of drought (seasonal versus supra-seasonal), its
intensity and duration, the availability of refuges (Table 5.2) and the condition of
the catchment and stream channel. In a small Spanish stream, the Fuirosos, subject
to seasonal drought, severe flow reduction and drying led to an increase in worms,
ostracods and Chironomus sp. midges, whereas during a less harsh dry period
when more refuges persisted, baetid mayflies, black flies and several different
chironomid genera dominated invertebrate biomass and density (Acuña et al.,
2005).
Similarly, Boulton and Lake (1992a,b) working in two seasonally-drying
Australian streams during a drought and the following wetter year found that
some genera of stoneflies, dragonflies and free-living caddisflies were either absent
or much less common a year after the drought. This was interpreted as the result
of failed recruitment, producing a ‘lagged’ response that affected the post-drought
recovery phase. However, in a study of invertebrate recovery after a 4-year
drought in six headwater streams in Georgia, Churchel and Batzer (2006) did not
find significant differences between streams where residual water persisted and
those that dried completely, indicating that the presence of refugial pools is not
always a strong determinant of post-drought community composition. These variable
findings across different streams indicate the need for caution when attempting to
draw generalizations regarding the responses of stream insects to drought.
Recovery after seasonal drought in most intermittent streams follows a rather
predictable sequence where early colonists with short life cycles (e.g. chironomids,
Effects of Drought on Stream Insects 93

simuliids) are joined by longer-lived collectors and grazers (mayflies and caddis-
flies) and taxon richness rises steadily as flows establish (Harrison, 1966; Boul-
ton and Lake, 1992a,b; Fritz and Dodds, 2004). Conversely, recovery after
supra-seasonal drought is less predictable because previously common drought-
intolerant taxa may be extinguished (Wood and Armitage, 2004), whereas new
habitat may be created for taxa that were rare before the drought. In a small acid
stream in England, post-drought recovery was dominated by the tanypod chi-
ronomid Zavrelimyia sp. that was previously uncommon, although it became
rare again when the macroinvertebrate community returned to its pre-drought
composition after high flows in winter (Ledger and Hildrew, 2001). Loss or
change in composition of aquatic macrophyte beds in perennial streams during
supra-seasonal drought can also contribute to unpredictable responses (Wood
and Armitage, 2004).

Indirect effects of drought

Drought can affect stream insects and ecosystem processes indirectly. From 2001
to 2004, severe drought in western USA coincided with earlier emergence of
adult Baetis bicaudatus, a montane mayfly (Harper and Peckarsky, 2006).
Increasing water temperatures rather than stream drying were the cue for this
accelerated emergence. Large-scale climate changes with an increasing fre-
quency of supra-seasonal drought were predicted to shift the timing of mayfly
metamorphosis, with negative impacts on populations that play a key role in
stream ecosystems (Harper and Peckarsky, 2006). Mayflies emerging in dry years
also tend to be smaller than in wetter years when emergence is delayed and
nymphs can feed for longer (Peckarsky et al., 2001).
Drought does not have a negative impact on all aquatic insects. A normally
perennial stretch of Waterston Stream, southern England, ceased flowing from
August 1973 to January 1974, virtually eliminating very common taxa such as
the amphipod Gammarus pulex, whereas other taxa such as the mayfly Ephemer-
ella ignita increased in abundance (Ladle and Bass, 1981). This pattern was
interpreted as a response to the timing of the drought and the availability of
recolonizers. In a Welsh mountain stream, two stonefly and a mayfly species
steadily increased in density after drought reduced the numbers of several other
common stream insects (Hynes, 1958). Depending on the ecological roles of
those aquatic insects favoured by drought, some stream ecosystem processes
may be altered, especially if previously dominant taxa are eliminated or if differ-
ent feeding groups are favoured by high or low flows (Bogan and Lytle, 2007).

Human Activities and the Effects of Drought

Humans compete with the other biota of aquatic ecosystems for water, especially
during drought and especially in arid areas. As conditions become drier, human
demands intensify. Aided by ever-increasing technological efficiency, extraction
of surface and groundwater accelerates the onset of, and slows recovery from,
94 A.J. Boulton and P.S. Lake

hydrological and groundwater droughts. For example, in English chalk streams,

maximum groundwater abstraction coincides with natural periods of low flow
(Agnew et al., 2000), prolonging the waterless periods in the channel. These
‘artificial droughts’ (droughts created by human activity, Boulton, 2003) exac-
erbate the stress of natural drought on aquatic insects (Wright and Berrie,
1987) and may cause permanent extinction of some groups in rivers subjected
to heavy extraction of water. Where serious groundwater droughts occur, recov-
ery of the stream community may take several years because of the need for
recovery of flow associated with recharge of the aquifer (Wood and Armitage,
2004).
Economic hardship due to drought prompts social pressure to ‘drought-
proof’ landscapes, usually by engineering schemes to redirect water or dam riv-
ers, and seldom through more appropriate land use or planning for future
droughts. These inter-basin transfers and dams have numerous deleterious
effects on stream insects (reviews in Boulton and Brock, 1999; Allan, 2004) and
are the inevitable by-product of seasonal and supra-seasonal droughts when
humans seek a more reliable surface water supply. Multiple small storages on
feeder streams further disrupt longitudinal connectivity and reduce total runoff,
rendering rivers downstream more drought-prone than normal and leading to
longer periods when flood plains remain dry, with negative impacts on the
aquatic biota (review in Jenkins and Boulton, 2007).
During severe drought, although humans are generally aware of dwindling
surface water supplies, there is seldom the same appreciation of the concurrent
groundwater drought. Excessive groundwater extraction from alluvial aquifers
during drought reduces river baseflows with reach and catchment-scale impacts
on the biota of these groundwater-dependent ecosystems (Wood and Armitage,
2004; Boulton and Hancock, 2006). The stream insects of perennial springs and
groundwater-fed rivers are particularly vulnerable to drought and often lack
mechanisms to recover after loss of surface water (Williams, 2006). In two German
karstic streams, their perennial springbrook sources supported different taxa from
those in the seasonally-dry reaches downstream and contributed substantially to
overall species richness (Meyer et al., 2003).
In streams under human pressure, the condition and availability of refuges
are impaired by direct water extraction during the drought, pollution by anthro-
pogenic toxicants, or excessive sedimentation that fills deeper pools and clogs
bed sediments. In-stream wood removal and bank slumps, coupled with exces-
sive grazing and loss of surface vegetation exacerbated during drought, hasten
sedimentation and create ‘sand-slugs’ that smother the stream bed and fill crucial
refugial pools (Downes et al., 2006). Remnant pools are one of the most impor-
tant refuges during drying, and their loss through sedimentation severely impacts
on post-drought recovery in sand-slugged streams. Excessive siltation in the
perennial section of an English chalk stream sampled during the 1976 drought
constrained macroinvertebrate habitat quality and availability, reducing the
abundance of benthic insects such as black flies and baetid mayflies (Wright and
Berrie, 1987). Ecological effects were considered most severe where the impacts
of siltation on aquatic invertebrates persisted into the autumn of 1977, but recovery
was complete by 1978 (Wright and Symes, 1999). Similarly, Lind et al. (2006)
Effects of Drought on Stream Insects 95

suggested that salinization due to human activity pushed the disturbance posed
by drought beyond the tolerance of many invertebrates in the Wimmera River,
Australia.
Drought may favour the spread of some diseases for which aquatic insects
are the vector. During severe drought, streams and wetlands shrink to isolated
pools where mosquito populations proliferate. Such areas also attract unusually
high aggregations of birds. Drought and this population concentration effect
greatly heightens the likelihood of contact between the mosquitoes and birds,
allowing arboviruses (arthropod-borne viruses), such as St Louis encephalitis
(Shaman et al., 2002) and West Nile virus (Shaman et al., 2005), to be amplified
in the bird hosts. When the drought breaks, the mosquitoes and birds disperse,
promoting virus transmission and often triggering an epidemic of the viruses in
humans (Epstein and DeFilippo, 2001; Shaman et al., 2002). Global warming
and the resultant cycles of more extreme floods and droughts may thus serve to
increase the distribution and severity of these mosquito-borne diseases (Epstein,
2001).

Avenues for Further Research on Drought and Stream Insects

While preparing this chapter, we were struck by the scattered nature and lack of
synthesis of research on the effects of drought on stream insects and their ecol-
ogy. Most of the work we found was phenomenological, opportunistic, restricted
in time and space, and seldom extended beyond a description of some aspect of
response to drought (either the effects of progressive drying or the response after
the drought had broken and flow resumed). This limited data set poorly equips
us for preparing for the likely effects of climate change scenarios that predict
increasing aridity and drought across much of the globe (Dai et al., 2004). Along
with accelerated ‘artificial drought’ as humans extract or pollute surface and
groundwater reserves, droughts will become more frequent disturbances and
deserve more attention and research than they currently receive.
One of the best sources of information about the effects of drought on stream
insects is well-planned, long-term monitoring studies that reveal changes in
stream communities across a multiple spatial scales, have sufficient pre-drought
information to identify trends in different streams, and continue long enough to
detect ‘lagged’ responses and even localized extinctions from especially severe
droughts (Boulton, 2003; Wood and Armitage, 2004). If such studies are com-
bined with monitoring of anthropogenic effects on streams during drought, it
may be possible to tease out the main synergistic processes and manage streams
to reduce environmental impacts at times of severe water deficit. We predict that
the most severe impacts will be on refuges such as remnant pools or ground-
waters that become heavily extracted or polluted as reserves dwindle and are not
protected during the ‘panic’ of water shortages. Research is needed on refuge
use by stream insects and ways to sustain these refuges during drought when
human pressures on them intensify.
Part of effective protection of refuges will hinge on some understanding
of the ecological strategies used by stream insects, and it is striking how few
96 A.J. Boulton and P.S. Lake

autecological data exist for most aquatic invertebrate species. Perhaps some
form of ‘ecological trait’ analysis (e.g. Bonada et al., 2007) will be a fruitful
approach. By assessing the ecological traits of a suite of stream insects at a
catchment or regional scale and comparing those with the suite exhibited by
survivors after a drought, we may identify some characteristic features that
promote survival in various refuges. From these sorts of studies may emerge
generalizations that can be used worldwide to enhance management and pro-
tection of stream insect populations from increasing frequency of natural and
artificial droughts.
More work is needed to understand the full impacts of drought on ecosys-
tem processes, such as organic matter breakdown and transport, which are
severely impaired when streams dry, removing a crucial pathway of energy in
these systems. Longitudinal and lateral hydrological fragmentation also sever
pathways of stream insect colonization and redistribution among habitats. Fish
and terrestrial predators, reliant on these pathways for prey, suffer during
drought, with repercussions for riparian and flood plain ecosystems. Drought
may also weaken, if not eliminate, prey subsidies consisting of aquatic insects
eaten by terrestrial predators and terrestrial insects being prey for stream preda-
tors, such as gerrid and veliid hemipterans and gyrinid beetles (Ballinger and
Lake, 2006).
Recent studies (Chadwick and Huryn, 2007) have demonstrated that drying
severely constrains insect production in some streams, interpreted as selection
against ecological traits such as large body size and longevity (Huryn and Wallace,
2000). In streams in Maine, large-bodied taxa such as the dragonflies Aeshnidae
and Gomphidae and the stoneflies Perlidae and Pteronarcyidae are eliminated
by drying, resulting in extremely low secondary production (Chadwick and
Huryn, 2005). We need more research that integrates ecological traits, refuge
use, and the effects on stream-scale secondary production to predict how the
trophic transfer of energy will be affected by drought in streams.
Predicting the full suite of ecological effects of drought is hampered by our
limited understanding of stream insect responses to this complex ramp distur-
bance, the effects of which are mediated by site-specific features (e.g. location,
size and shape, nature of bed sediments, groundwater reliance). For example,
while drought may dry upland streams completely, lowland rivers may continue
to flow slowly, but lose all lateral connectivity with flood plain and riparian zones.
Pools in rivers with porous beds may be maintained by hyporheic flow (Stanley
et al., 1997) or springs (Dahm et al., 2003), sustaining a refuge that would be
absent in rivers with less pervious sediments. It is critical that future research into
refuge use and the effects of drought on stream insects and their production
acknowledges the importance of patterns of hydrological drying. Our ‘model’ of
sequential changes in insect community composition in response to drying (Fig.
5.1) is a starting point for refinements, and it would be useful to characterize how
stream types and drying patterns influence the severity of impacts of loss of lat-
eral, longitudinal and vertical hydrological connectivity. For example, in upland
streams the loss of longitudinal connectivity via flow may be more severe than
lateral disconnections with inundated stream-side vegetation, whereas in lowland
rivers, lateral connections may be more important.
Effects of Drought on Stream Insects 97

Conclusions

Our review has illustrated that while there may be relatively predictable changes
in aquatic insect assemblages as drying progresses, the ecological consequences
of these changes are less predictable because of the complex interactions with
human impacts and the modifying effects of stream location, morphology and
groundwater linkage. As drought is a natural part of all stream ecosystems, we
must tailor our management strategies accordingly. ‘Hard’ and reactive engi-
neering approaches to resist the effects of drought are less likely to be successful
than proactive strategies that seek to improve our resilience to this ramp distur-
bance. As droughts are always large-scale disturbances and as global climate
change is a very large-scale phenomenon, the synergism between these two
ramp disturbances brings the grim prospect of severe and prolonged ‘mega-
droughts’ with potentially devastating impacts on aquatic insects and associated
ecosystem processes.
We lack autecological data on the responses of most stream insects to drying
and resumption of flow. This makes it difficult to predict changes in organic matter
transfer, secondary production and food web dynamics in response to increasing
frequency and severity of drought. As most stream insects are resilient (rather
than resistant) to drought, refuge use is especially important to their survival.
Effective management and protection of stream ecosystems and their aquatic
insects therefore relies on preserving functional refuges and making difficult deci-
sions about allocations of water for environmental purposes during times of
severe water scarcity.

Acknowledgements

We thank Bruce Chessman, Marcus Finn and two reviewers for comments on
earlier drafts. We are also grateful to the Australian Research Council, and Land
and Water Australia for financial support.

References

Acuña, V., Munoz, I., Giorgi, A., Omella, M., Sabater, F. and Sabater, S. (2005) Drought
and postdrought recovery cycles in an intermittent Mediterranean stream: structural
and functional aspects. Journal of the North American Benthological Society 24,
919–933.
Agnew, C.T., Clifford, N.J. and Haylett, S. (2000) Identifying and alleviating low flows in
regulated rivers: the case of the Rivers Bulbourne and Gade, Hertfordshire, UK.
Regulated Rivers: Research & Management 16, 245–266.
Allan, J.D. (2004) Landscapes and riverscapes: The influence of land use on stream eco-
systems. Annual Review of Ecology Evolution and Systematics 35, 257–284.
Angradi, T.R. (1999) Fine sediment and macroinvertebrate assemblages in Appalachian
streams: a field experiment with biomonitoring applications. Journal of the North
American Benthological Society 18, 43–65.
98 A.J. Boulton and P.S. Lake

Ballinger, A. and Lake, P.S. (2006) Energy and nutrient fluxes from rivers and streams into
terrestrial food webs. Marine and Freshwater Research 57, 15–28.
Bogan, M.T. and Lytle, D.A. (2007) Seasonal flow variation allows ‘time-sharing’ by dis-
parate aquatic insect communities in montane desert streams. Freshwater Biology
52, 290–304.
Bonada, N., Prat, N., Resh, V.H. and Statzner, B. (2006a) Developments in aquatic insect
biomonitoring: a comparative analysis of recent approaches. Annual Review of
Entomology 51, 495–523.
Bonada, N., Rieradevall, M., Prat, N. and Resh, V.H. (2006b) Benthic macroinvertebrate
assemblages and macrohabitat connectivity in Mediterranean-climate streams of
northern California. Journal of the North American Benthological Society 25, 32–43.
Bonada, N., Rieradevall, M. and Prat, N. (2007) Macroinvertebrate community structure
and biological traits related to flow permanence in a Mediterranean river network.
Hydrobiologia 589, 91–106.
Boulton, A.J. (1989) Over-summering refuges of aquatic macroinvertebrates in two inter-
mittent streams in central Victoria. Transactions of the Royal Society of South Australia
113, 23–34.
Boulton, A.J. (2003) Parallels and contrasts in the effects of drought on stream macro-
invertebrate assemblages. Freshwater Biology 48, 1173–1185.
Boulton, A.J. and Brock, M.A. (1999) Australian Freshwater Ecology: Processes and
Management. Gleneagles Publishing, Adelaide.
Boulton, A.J. and Hancock, P.J. (2006) Rivers as groundwater dependent ecosystems: a
review of degrees of dependency, riverine processes, and management implications.
Australian Journal of Botany 54, 133–144.
Boulton, A.J. and Lake, P.S. (1992a) The ecology of two intermittent streams in Victoria,
Australia. II. Comparisons of faunal composition between habitats, rivers, and years.
Freshwater Biology 27, 99–121.
Boulton, A.J. and Lake, P.S. (1992b) The ecology of two intermittent streams in Victoria,
Australia. III. Temporal changes in faunal composition. Freshwater Biology 27,
123–138.
Boulton, A.J. and Stanley, E.H. (1995) Hyporheic processes during flooding and drying
in a Sonoran Desert stream. II. Faunal dynamics. Archiv für Hydrobiologie 134, 27–52.
Boulton, A.J., Valett, H.M. and Fisher, S.G. (1992) Spatial distribution and taxonomic
composition of the hyporheos of several Sonoran Desert streams. Archiv für Hydro-
biologie 125, 37–61.
Boumezzough, A. (1983) Les communautés animals ripicoles du bassin versant de la
Rivière Aille (Variable-France). II. Composition et structure de la faune épigée.
Ecologia Mediterranea 9, 3–33.
Caruso, B.S. (2002) Temporal and spatial patterns of extreme low flows and effects on
stream ecosystems in Otago, New Zealand. Journal of Hydrology 257, 115–133.
Chadwick, M.A. and Huryn, A. D. (2005) Response of stream macroinvertebrate pro-
duction to atmospheric nitrogen deposition and channel drying. Limnology and
Oceanography 50, 228–236.
Chadwick, M.A. and Huryn, A. D. (2007) Role of habitat in determining macroinvertebrate
production in an intermittent-stream system. Freshwater Biology 52, 240–251.
Churchel, M.A. and Batzer, D.P. (2006) Recovery of aquatic macroinvertebrate communities
from drought in Georgia piedmont headwater streams. American Midland Naturalist
156, 259–272.
Clinton, S.M., Grimm, N.B. and Fisher, S.G. (1996) Response of a hyporheic invertebrate
assemblage to drying disturbance in a desert stream. Journal of the North American
Benthological Society 15, 700–712.
Effects of Drought on Stream Insects 99

Cummins, K.W. and Klug, M.J. (1979) Feeding ecology of stream invertebrates. Annual
Review of Ecology and Systematics 10, 147–172.
Dahm, C.N., Baker, M.A., Moore, D.I. and Thibault, J.R. (2003) Coupled biogeochemical
and hydrological responses of streams and rivers to drought. Freshwater Biology 48,
1219–1231.
Dai, A.G., Trenberth, K.E. and Qian, T.T. (2004) A global dataset of Palmer Drought
Severity Index for 1870–2002: Relationship with soil moisture and effects of surface
warming. Journal of Hydrometeorology 5, 1117–1130.
Delucchi, C.M. (1989) Movement patterns of invertebrates in temporary and permanent
streams. Oecologia 78, 199–207.
Douglas, M.M., Townsend, S.A. and Lake, P.S. (2003) Stream dynamics. In: Andersen,
A., Cook, G.D. and Williams, R.J. (eds) Fire in Tropical Savannas: An Australian
Study. Springer, New York, pp. 59–78.
Downes, B.J., Lake, P.S., Glaister, A. and Bond, N.R. (2006) Effects of sand sedimentation
on the macroinvertebrate fauna of lowland streams: are the effects consistent? Fresh-
water Biology 51, 144–160.
Druyan, L.M. (1996). Drought. In: Schneider, S.H. (ed.) Encyclopedia of Climate and
Weather. Volume 1. Oxford University Press, New York, pp. 256–259.
Epstein, P.R. (2001) Climate change and emerging infectious diseases. Microbes and
Infection 3, 747–754.
Epstein, P.R. and DeFilippo, C. (2001) West Nile virus and drought. Global Change and
Human Health 2, 105–107.
Erman, N.A. and Erman, D.C. (1995) Spring permanence, Trichoptera species richness,
and the role of drought. Journal of the Kansas Entomological Society 68, 50–64.
Extence, C.A. (1981) The effect of drought on benthic invertebrate communities in a
lowland river. Hydrobiologia 83, 217–224.
Filho, M.I.S. and Maltchik, L. (2000) Stability of macroinvertebrates to hydrological
disturbance by flood and drought in a Brazilian semi-arid river (NE Brazil). Verhand-
lungen der Internationalen Vereinigung für theoretische und angewandte Limnologie
27, 2461–2466.
Fritz, K.M. and Dodds, W.K. (2004) Resistance and resilience of macroinvertebrate as-
semblages to drying and flood in a tallgrass prairie stream system. Hydrobiologia
527, 99–112.
Gasith, A. and Resh, V.H. (1999) Streams in Mediterranean climate regions: abiotic influ-
ences and biotic responses to predictable seasonal events. Annual Review of Ecology
and Systematics 30, 51–81.
Gooderham, J. and Tsyrlin, E. (2002) The Waterbug Book: A Guide to the Freshwater
Macroinvertebrates of Temperate Australia. CSIRO Publishing, Collingwood, Melbourne,
Victoria.
Gray, L.J. and Fisher, S.G. (1981) Postflood recolonization pathways of macroinvertebrates
in a lowland Sonoran Desert stream. American Midland Naturalist 106, 249–257.
Harper, M.P. and Peckarsky, B.L. (2006) Emergence cues of a mayfly in a high-altitude stream
ecosystem: potential response to climate change. Ecological Applications 16, 612–621.
Harper, P.P. and Hynes, H.B.N. (1970) Diapause in the nymphs of Canadian winter
stoneflies. Ecology 51, 425–427.
Harrison, A.D. (1966) Recolonization of a Rhodesian stream after drought. Archiv für
Hydrobiologie 62, 405–421.
Harrison, S.S.C. (2000) The importance of aquatic margins to invertebrates in English
chalk streams. Archiv für Hydrobiologie 149, 213–240.
Huryn, A.D. and Wallace, J.B. (2000) Life history and production of stream insects.
Annual Review of Entomology 45, 83–110.
100 A.J. Boulton and P.S. Lake

Hynes, H.B.N. (1958) The effect of drought on the fauna of a small mountain stream in
Wales. Verhandlungen der Internationalen Vereinigung für theoretische und angewandte
Limnologie 13, 826–833.
Hynes, H.B.N. (1970) The Ecology of Running Waters. Liverpool University Press,
Liverpool.
Hynes, J.D. (1975) Annual cycles of macro-invertebrates of a river in southern Ghana.
Freshwater Biology 5, 71–83.
Jenkins, K.M. and Boulton, A.J. (2007) Detecting impacts and setting restoration targets
in arid-zone rivers: aquatic microinvertebrate responses to reduced floodplain
inundation. Journal of Applied Ecology 44, 823–832.
Kirby, J.M., Webster, J.R. and Benfield, E.F. (1983) The role of shredders in detrital dynamics
of permanent and temporary streams. In: Fontaine, T.D. and Bartell, S.M. (eds) Dynam-
ics of Lotic Ecosystems. Ann Arbor Science, Ann Arbor, Michigan, pp. 425–435.
Ladle, M. and Bass, J.A.B. (1981) The ecology of a small chalk stream and its responses
to drying during drought conditions. Archiv für Hydrobiologie 90, 448–466.
Lake, P.S. (1977) Pholeteros—the faunal assemblage found in crayfish burrows. Australian
Society for Limnology Newsletter 15, 57–60.
Lake, P.S. (2000) Disturbance, patchiness, and diversity in streams. Journal of the North
American Benthological Society 19, 573–592.
Lake, P.S. (2003) Ecological effects of perturbation by drought in flowing waters. Fresh-
water Biology 48, 1161–1172.
Lake, P.S. (2006) Drought, the ‘creeping disaster’. Effects on aquatic ecosystems. Unpublished
report to Land and Water Australia, Canberra.
Lake, S., Bond, N. and Reich, P. (2006) Floods down rivers: from damaging to replenishing
forces. Advances in Ecological Research 39, 41–62.
van Lanen, H.A.J. and Peters, E. (2000) Definition, effects and assessment of groundwater
droughts. In: Vogt, J.V. and Somma, F. (eds) Drought and Drought Mitigation in Europe.
Kluwer, Dordrecht, pp. 49–61.
Larimore, R.W., Childers, W.F. and Heckrotte, C. (1959) Destruction and re-establishment
of stream fish and invertebrates affected by drought. Transactions of the American
Fisheries Society 88, 261–285.
Ledger, M.E. and Hildrew, A.G. (2001) Recolonization by the benthos of an acid stream
following a drought. Archiv für Hydrobiologie 152, 1–17.
Lind, P.R., Robson, B.J. and Mitchell, B.D. (2006) The influence of reduced flow during
a drought on patterns of variation in macroinvertebrate assemblages across a spatial
hierarchy in two lowland rivers. Freshwater Biology 51, 2282–2295.
Lytle, D.A. (2000) Biotic and abiotic effects of flash flooding in a montane desert stream.
Archiv für Hydrobiologie 150, 85–100.
Maamri, A., Chergui, H. and Pattee, E. (2005) Leaf litter processing in a temporary north-
eastern Moroccan stream. Archiv für Hydrobiologie 140, 513–531.
Malmqvist, B. (2002) Aquatic invertebrates in riverine landscapes. Freshwater Biology
47, 679–694.
McIntosh, M.D., Benbow, M.E. and Burky, A.J. (2002) Effects of stream diversion on riffle
macroinvertebrate communities in a Maui, Hawaii, stream. River Research and
Applications 18, 569–581.
Meyer, A., Meyer, E.I. and Meyer, C. (2003) Lotic communities of two small temporary
karstic stream systems (East Westphalia, Germany) along a longitudinal gradient of
hydrological intermittency. Limnologica 33, 271–279.
Miller, A.M. and Golladay, S.W. (1996). Effects of spates and drying on macroinvertebrate
assemblages of an intermittent and a perennial prairie stream. Journal of the North
American Benthological Society 15, 670–689.
Effects of Drought on Stream Insects 101

O’Connor, N.A. and Lake, P.S. (1994) Long-term and seasonal large-scale disturbances
of a small lowland stream. Australian Journal of Marine and Freshwater Research 45,
243–255.
Paltridge, R.M., Dostine, P.L., Humphrey, C.L. and Boulton, A.J. (1997) Macroinvertebrate
recolonization after re-wetting of a tropical seasonally-flowing stream (Magela Creek,
Northern Territory, Australia). Marine and Freshwater Research 48, 633–645.
Peckarsky, B.L., Taylor, B.W., McIntosh, A.R., McPeek, M.A. and Lytle, D.A. (2001) Variation
in mayfly size at metamorphosis as a developmental response to risk of predation.
Ecology 82, 740–757.
Resh, V.H., Brown, A.V., Covich, A.P., Gurtz, M.E., Li, H.W., Minshall, G.W., Reice, S.R.,
Sheldon, A.L., Wallace, J.B. and Wissmar, R.C. (1988). The role of disturbance in
stream ecology. Journal of the North American Benthological Society 7, 433–455.
Rincon, J. and Cressa, C. (2000) Temporal variability of macroinvertebrate assemblages
in a neotropical intermittent stream in northwestern Venezuela. Archiv für Hydro-
biologie 148, 421–432.
Shaman, J., Day, F.J. and Stieglitz, M. (2002) Drought-induced amplification of Saint
Louis encephalitis virus, Florida. Emerging Infectious Diseases 8, 575–580.
Shaman, J., Day, F.J. and Stieglitz, M. (2005) Drought-induced amplification and epidemic
transmission of West Nile virus in southern Florida. Journal of Medical Entomology
42, 134–141.
Shivoga, W.A. (2001) The influence of hydrology on the structure of invertebrate communi-
ties in two streams flowing into Lake Nakuru, Kenya. Hydrobiologia 458, 121–130.
Stanley, E.H., Buschman, D.L., Boulton, A.J., Grimm, N.B. and Fisher, S.G. (1994)
Invertebrate resistance and resilience to intermittency in a desert stream. American
Midland Naturalist 131, 288–300.
Stanley, E.H., Fisher, S.G. and Grimm, N.B. (1997) Ecosystem expansion and contraction
in streams. BioScience 47, 427–435.
Stehr, W.C. and Branson, J.W. (1938) An ecological study of an intermittent stream. Ecology
19, 294–310.
Svoboda, M., LeComte, D., Hayes, M., Heim, R., Gleason, K., Angel, J., Rippey, B.,
Tinker, R., Palecki, M., Stooksbury, D., Miskus, D. and Stephens, S. (2002) The
Drought Monitor. Bulletin of the American Meteorological Society 83, 1181–1190.
Towns, D.R. (1983) Terrestrial oviposition by two species of caddisfly in South Australia
(Trichoptera: Leptoceridae). Journal of the Australian Entomological Society 22,
113–118.
Towns, D.R. (1985) Limnological characteristics of a South Australian intermittent stream,
Brown Hill Creek. Australian Journal of Marine and Freshwater Research 36, 821–
837.
Tronstad, L.M., Tronstad, B.P. and Benke, A.C. (2005) Invertebrate seedbanks: rehydra-
tion of soil from an unregulated river floodplain in the south-eastern US. Freshwater
Biology 50, 646–655.
Velasco, J. and Millan, A. (1998) Insect dispersal in a drying desert stream: effects of
temperature and water loss. Southwestern Naturalist 43, 80–87.
Wiens, J.A. (2002) Riverine landscapes: taking landscape ecology into the water. Fresh-
water Biology 47, 501–515.
Wilhite, D.A. (Ed.) (2000) Drought: A Global Assessment. Vol. 1, Routledge, London.
Williams, D.D. (2006) The Biology of Temporary Waters. Oxford University Press, New York.
Williams, D.D. and Feltmate, B.W. (1992) Aquatic Insects. CAB International, Wallingford,
Oxon, UK.
Williams, D.D. and Hynes, H.B.N. (1976) The recolonization mechanisms of stream benthos.
Oikos 27, 265–272.
102 A.J. Boulton and P.S. Lake

Williams, D.D. and Hynes, H.B.N. (1977) The ecology of temporary streams. II. General
remarks on temporary streams. Internationalen Revue gesamten Hydrobiologie 62,
53–61.
Wood, P.J. and Armitage, P.D. (1999) Sediment deposition in a small lowland stream—
management implications. Regulated Rivers: Research & Management 15, 199–210.
Wood, P.J. and Armitage, P.D. (2004) The response of the macroinvertebrate community
to low-flow variability and supra-seasonal drought within a groundwater dominated
stream. Archiv für Hydrobiologie 161, 1–20.
Wright, J.F. (1992) Spatial and temporal occurrence of invertebrates in a chalk stream,
Berkshire, England. Hydrobiologia 248, 11–30.
Wright, J.F. and Berrie, A.D. (1987) Ecological effects of groundwater pumping and a
natural drought on the upper reaches of a chalk stream. Regulated Rivers: Research
& Management 1, 145–160.
Wright, J.F. and Symes, K.L. (1999) A nine-year study of the macroinvertebrate fauna of
a chalk stream. Hydrological Processes 13, 371–385.
Wright, J.F., Sutcliffe, D.W. and Furse, M.T. (eds) (2000) Assessing the Biological Quality
of Fresh Waters. Freshwater Biological Association, Ambleside.
6 The Effect of Floods on Aquatic
Invertebrate Communities
RUSSELL G. DEATH
Institute of Natural Resources – Ecology, Massey University, Private Bag
11-222, Palmerston North, New Zealand
E-mail: [Link]@[Link]

Abstract
Almost all rivers experience increases in discharge, although whether this results in a flood
depends on the size of the increase, channel morphology and hydrology. High flows have
many effects on aquatic invertebrates both directly and indirectly through changes to
habitat. Shear stress on the stream bed increases, removing organisms or forcing them to
seek refuge among substrates or in the drift. Substrates can be moved, scouring inverte-
brates and their periphyton food source. Numerous studies have shown that floods reduce
invertebrate abundances and diversity, although the threshold flows at which this occurs
still remain contentious. However, recovery from most floods is relatively rapid and inver-
tebrates persist even in very flood-prone streams. This is a result of many taxa having
morphological, behavioural and life-history strategies to cope with the effects of floods.
Although we have been aware of the community-level effects of floods on aquatic inver-
tebrates for decades, it still remains unclear how many individual taxa respond, what
motivates that response and what drives the recovery process. Understanding these pat-
terns is key to our ability to manage flow variability to maintain ecosystem integrity in the
face of increasing pressure to remove floods from a river’s hydrology regime.

What is a Flood and Does it Represent a Disturbance to Stream

Invertebrates?
Floods are a unique characteristic of almost all natural rivers and streams (Lake,
2000; Robinson et al., 2002) although the predictability, frequency, magnitude and
duration can differ markedly between rivers, regions and continents (Robinson and
Uehlinger, 2003; Poff et al., 2006). Many stream ecologists see floods as one of the
principal structuring forces on stream invertebrate populations and communities
(Resh et al., 1988; Reice et al., 1990; Lake, 2000). This has led a number of
ecologists to advocate the return of natural flood regimes (the natural flow para-
digm) to regulated rivers as one mechanism for maintaining the ecological integrity
of downstream reaches (Poff et al., 1997). Hydrologists define a flood as a flow
 CAB International 2008. Aquatic Insects: Challenges to Populations
(eds J. Lancaster and R.A. Briers) 103
104 R.G. Death

that overtops stream banks, with the term ‘spate’ associated with high flows that do
not overtop the banks (Gordon et al., 2004). However, whether or not an increased
discharge overtops the banks is unlikely to alter the effects of the increased flows
on invertebrates; therefore, the word flood is used in an ecological sense to mean
any increase in discharge, not just increases that overtop the stream bank.
To what extent these increased discharges represent a disturbance to stream
invertebrate communities has been an area of debate (Resh et al., 1988; Poff,
1992; Death and Winterbourn, 1994). Some people reason that the relative pre-
dictability and lack of lasting effects of many high-discharge events (see below)
indicate that stream invertebrates are ‘adapted’ to these events, which therefore
are not disturbances (Resh et al., 1988). However, the focus now is on quantify-
ing the degree of habitat alteration and concomitant changes in invertebrate
communities, even if short lived, rather than the predictability of the discharge
event (Poff, 1992; Death and Winterbourn, 1994; Angradi, 1997). Nevertheless,
not all increased discharges disturb stream invertebrates; identifying what physi-
cal changes in the environment caused by floods represent disturbances to stream
invertebrates and how to quantify these changes remains a challenge. Most
researchers acknowledge that increases in flow that move substrates comprise
disturbances for stream invertebrates, but identifying such flows for particular
rivers is much more difficult (Clausen and Plew, 2004; Matthaei et al., 2004).

What do Floods do?

Floods result from high rainfall events, snowmelt, dam release, removal or break-
age (Minshall et al., 1983) and lead to large volumes of water moving rapidly
downstream. The volume and speed of the water creates high shear stress on the
stream bed and surrounding channel that scours, abrades or moves substrates,
suspends sediment in the water column, deposits logs and detritus, can alter
channel shape and results in a patchwork of both scoured and infilled stream bed
(Carling, 1987; Lake, 2000; Matthaei and Townsend, 2000b).
These changes in habitat structure have corresponding effects on the biotic
component of rivers and streams. Periphyton, macrophytes and bryophytes are
washed away, scoured or buried (Fisher et al., 1982; Biggs et al., 1999; Suren
and Duncan, 1999; Downes and Street, 2005). Even thin epilithic biofilms can
be scoured by flood-induced substrate movement (Scrimgeour et al., 1988). Sur-
prisingly, the quantity of benthic organic matter seems to be largely unaffected
by floods, possibly because similar volumes of material are deposited as are
washed away (Death and Winterbourn, 1995; Lytle, 2000; Robinson et al., 2003).
However, leaf accumulations in forest streams can be reduced by floods (Angradi,
1997).
Benthic invertebrate density, biomass and diversity are usually reduced
by flood events (e.g. Jones, 1951; Maitland, 1963; Scrimgeour et al., 1988;
Robinson et al., 2003) as individuals shift out of the rising channel, are washed
away, or are crushed by moving substrate and debris. Abundance of animals can
be reduced by 70–95% (Siegfried and Knight, 1977; Brooks and Boulton, 1991;
Giller et al., 1991; Matthaei et al., 1997a). However, floods do not always have
Effect of Floods on Aquatic Invertebrates 105

deleterious effects. Invertebrates in the Fraser River, Canada are unaffected by

flooding because they simply move up the shoreline to more stable substrates as
the water level rises (Rempel et al., 1999). Similarly, invertebrates in streams of
the Indian subcontinent are not affected by monsoonal floods because of syn-
chronized seasonal changes in abundance (Arunachalam et al., 1991; Brewin
et al., 2000).
Increases in invertebrate drift as a result of increased flows, often termed
‘catastrophic’ drift, have been recorded in several studies (e.g. Irvine, 1985; Brit-
tain and Eikeland, 1988; Poff and Ward, 1991). Entrainment of substrates at
flows above critical is one of the main reasons for this increased drift as the sur-
face to which an animal clings is removed into the water column (Giberson and
Caissie, 1998; Bond and Downes, 2003; Gibbins et al., 2005). However, increased
shear stress from high flows can also remove animals into the water column
(Lancaster and Hildrew, 1993; Bond and Downes, 2000, 2003), as can physical
abrasion from entrained fine sediment particles (Downes et al., 1998; Bond
and Downes, 2003). Multifactorial experiments by Bond and Downes (2003)
on Australian invertebrate communities indicated that flood-induced drift
was predominantly a result of shear stress from increased flows rather than fine-
sediment abrasion. Gibbins et al. (2005) studied induced drift of baetid mayflies
and concluded that, while increased shear stress does increase drift at higher
flows, the entrainment of animals along with mobile substrates is much greater.
Bed movement is the principal instigator of catastrophic drift from floods but,
in streams that experience smaller increases in discharge or where substrates
are tightly embedded, shear stress alone can lead to changes in the benthic inver-
tebrate communities (Boulton et al., 1992; Rempel et al., 1999). Disturbance
may also potentially alter the fitness of invertebrate populations, without any
mortality or displacement, by increasing the metabolic costs associated with
activities such as reconstructing food-catching nets (Beveridge and Lancaster,
2007).
There are fewer studies of flood effects on invertebrates in tropical streams
(Turcotte and Harper, 1982; Flecker and Feifarek, 1994; Rosser and Pearson,
1995), although the responses and outcomes seem broadly similar to those in
temperate streams. One exception is subtropical Nepalese streams where inver-
tebrates are relatively unaffected by monsoon floods, possibly because of the
seasonal predictability of the floods (Arunachalam et al., 1991; Brewin et al.,
2000).

How do Invertebrates Respond During Floods?

Despite the morphological and behavioural characteristics of many stream inver-
tebrates to maintain themselves in high flows, invertebrate communities exhibit
low resistance to floods (Grimm and Fisher, 1989; Lake, 1990; Lancaster and
Belyea, 1997). Many invertebrates drift or are washed away by floods. It has
long been hypothesized that invertebrates actively or passively seek refuge before
or during high flow events (Sedell et al., 1990; Townsend and Hildrew, 1994),
but there has been relatively little observational or experimental research to
106 R.G. Death

evaluate whether or how this occurs. Such research is fraught with difficulties:
predicting when floods will occur, being able to work during high flows and
observing what invertebrates do in swollen turbid rivers (Winterbottom et al.,
1997). Furthermore, invertebrates may accumulate in refugia or simply maintain
densities relative to less stable surrounding substrates.
Some species of coleopterans and hemipterans leave desert streams in
response to rain storms to avoid floods (Gray, 1981; Lytle, 2000; Lytle and Poff,
2004). Many invertebrates are washed into or actively move into the flood plain
or river margins during floods (e.g. Bishop, 1973; Scrimgeour et al., 1988; Rem-
pel et al., 1999). Although Matthaei and Townsend (2000a) concluded, from
work in a New Zealand river, that flood plains only provide temporary refuge for
stream invertebrates and many are stranded as the flood plain dries. Whether
flood plains act as refugia depends on surface or hyporheic connectivity being
maintained with the main channel. Refugia in river margins, in contrast, can
prevent or ameliorate the effects of increased shear stress from flooding (Rempel
et al., 1999).
It has been widely held that animals seek refuge in the hyporheic zone during
high flow events (Resh et al., 1988; Poff and Ward, 1990; Sedell et al., 1990).
Few studies have tested whether this occurs. Studies in North America and New
Zealand concluded the hyporheic zone did not act as a refuge during floods
(Giberson and Hall, 1988; Palmer et al., 1992; Olsen and Townsend, 2005). In
contrast, work in the Rhône River found that refuge-seeking in the hyporheic
zone during floods does occur, but is not universal and depended on the geo-
morphic complexity of a stream, substrate size and thickness of the sediments
(Marmonier and Chatelliers, 1991; Dole-Olivier and Marmonier, 1992; Dole-
Olivier et al., 1997).
Invertebrates are also known to accumulate in areas of the stream channel
where hydraulic forces are reduced, ‘dead zones’, although whether they do so
actively or passively is still unclear (Lancaster and Hildrew, 1993; Winterbottom
et al., 1997; Lancaster, 2000). Stable substrates, such as debris dams and embed-
ded stones, also provide refugia during high flows (Robertson et al., 1995; Palmer
et al., 1996; Matthaei et al., 2000). Matthaei et al. (2000) postulated that inver-
tebrates may be able to sense a stone vibrating just before it becomes entrained
and use this cue to move to more stable areas of the stream bed. Brooks (1998,
cited in Lake, 2000) found a glossosomatid caddisfly Agapetus kimminsi moved
from the tops of stones to the sides in response to increased turbidity and veloc-
ity, possibly to avoid being washed away. Several authors have recorded that
seasonal movements of some invertebrate taxa to flow refugia may be a response
to forthcoming seasonal flood events (Hauer and Stanford, 1982; Cobb et al.,
1992). However, it is unclear whether these are flood avoidance mechanisms,
life-history behaviours or both. Microform bed clusters, organized stacks of stone
in the stream bed, have also been predicted to act as stable refugia for inverte-
brates and periphyton during floods (Biggs et al., 1997), but Matthaei and Huber
(2002) found no evidence that they provided flood refugia for invertebrates in a
German river.
The availability of refugia, in turn, determines the effect and subsequent
recovery from flood events (Lake, 2000), although to what extent is unclear
Effect of Floods on Aquatic Invertebrates 107

(Lancaster and Belyea, 1997). Streams with more heterogeneous substrates com-
posed of dead zones, debris dams, flood plains and stable surface stones provide
more refugia and thus more colonists once high flows subside. In contrast, bed-
rock streams or those with homogeneous channels may be more affected by
floods and slower to recover post flood (Fisher et al., 1982; Angradi, 1997; Gjer-
løv et al., 2003). Despite these studies, we still know very little about what hap-
pens to invertebrates immediately before and during floods: how and when the
different refugia are used; what cues, if any, they use to avoid being washed away;
and how taxon-specific responses to floods alter the recovery process.

How Quick is Recovery From Floods?

Although there are numerous potential refugia in many streams, floods generally
result in a marked decline in both density and richness of resident invertebrates.
However, one of the most characteristic phenomena of streams worldwide is the
rapidity with which those communities recover (Resh et al., 1988; Reice et al.,
1990; Lake, 2000). Small-scale, instream experimental disturbance studies have
found invertebrate communities can recover within 4 days to 6 weeks (e.g. Boulton
et al., 1988; Brooks and Boulton, 1991; Death, 1996; Matthaei et al., 1996). The
number of taxa generally recovers more quickly than the density of animals
(Mackay, 1992; Death, 1996; Matthaei et al., 1996; but see Minshall et al., 1983;
Lake, 2000). However, mechanisms of recovery in these small-scale experiments
may differ from those in larger floods (Fisher, 1987; Minshall, 1988; Lake, 1990).
Smaller floods create mosaics of disturbed and undisturbed patches of substrate
(Doeg et al., 1989; Lancaster and Hildrew, 1993; Matthaei et al., 1999) so these
experiments are relevant to recovery from small disturbances, but may not be so
for larger ones. Comparisons between natural floods and experimental distur-
bances are few and the results are dissimilar; Matthaei et al. (1997a) found similar
responses between the two events in a Swiss river, but Brooks and Boulton
(1991) found markedly different outcomes in an Australian temporary stream.
Different characteristics of the community (e.g. diversity, abundance, taxo-
nomic composition) recover at differing rates. Furthermore, few authors have
enough before-flood data, or evaluate recovery rates statistically. Recorded
recovery rates from floods vary from 1 month in a Swiss river after a 200-fold
increase in discharge (≈5-year recurrence interval) (Matthaei et al., 1997a) to
greater than 3 years after a 1 in 50-year rainfall event during summer in Ireland
(Giller et al., 1991) and a dam failure (and 3.5 months dewatering) in the Teton
River, Idaho (Minshall et al., 1983). However, changes in channel morphology
and debris accumulation from severe floods (i.e. 1 in 2000 years) may have
much longer lasting (>5 years) effects on invertebrate community density and
productivity (Lamberti et al., 1991; Snyder and Johnson, 2006). Recovery inter-
vals from most floods seem to be about 2–4 months, although recovery is often
incomplete before the study concludes or the next flood occurs (e.g. Fisher et al.,
1982; Scrimgeour et al., 1988; Angradi, 1997).
Comparing recovery rates from floods in different rivers is extremely diffi-
cult, because no two floods will be of the same magnitude. Even for instream
108 R.G. Death

experimental manipulations there have been very few between-stream compari-

sons of recovery from disturbance (but see Yount and Niemi, 1990; Death, 1996;
Gjerløv et al., 2003; Melo et al., 2003). Consequently, we know very little about
whether, how or why recovery rates differ among streams. A greater rate of
recovery would be expected in streams that experience more floods (Poff and
Ward, 1990; Mackay, 1992). In northern hemisphere streams, flood recovery is
more rapid in disturbed streams, but only when limited refugia are available
(Matthaei et al., 1996; Gjerløv et al., 2003). In southern hemisphere streams,
however, recovery rates are similar at sites that differ in disturbance history
(Marchant et al., 1991; Death, 1996).
Death and Zimmermann (2005) found the effects of floods on invertebrate
diversity were less in the forested reaches of New Zealand mountain streams
than in open reaches approximately 1 km downstream (Fig. 6.1). They argued
that this occurred because the recovery of fauna post-flood in the forest reaches
did not depend on recovery of the periphyton food base that delayed recovery
in the downstream reaches. Therefore, recolonization of the invertebrate fauna
may be more rapid where recovery of the stream food base is more rapid, as in
some forest and high-nutrient streams (Gafner and Robinson, 2007). Recovery
rates may also be influenced by the nature of any remnant populations that sur-
vive the flood (Ledger et al., 2006) and/or whether a stream is perennial or
intermittent (Fritz and Dodds, 2004).
Wallace (1990) argued that recovery from floods in warm desert streams
(Gray and Fisher, 1981; Fisher et al., 1982) was more rapid than in cooler

30

25
Number of taxa (/0.1 m2)

20

15

10

0
0 5 10 15 20 25 30 35 40
Mean annual stone movement (%)

Fig. 6.1. Mean (± 1 SE) number of taxa as a function of stone movement at open
(open symbols) and forested (closed symbols) sites on 10 Taranaki, New Zealand
streams. Taxa number is unaffected by substrate movement at forest sites but
declines with greater movement at open sites (y = 22.92 – 0.16x). Redrawn from
Death and Zimmermann (2005).
Effect of Floods on Aquatic Invertebrates 109

temperate regions (Thorup, 1970; Giller et al., 1991) because of flood-selected

life-history traits in desert-stream invertebrates, such as rapid development, con-
tinuous emergence and diapausing eggs. However, it is difficult to judge the
validity of this hypothesis from such a limited number of studies. Many New
Zealand and Andean stream invertebrates have similar ‘flood-adapted’ traits
(e.g. short life cycles, continuous emergence) and are characterized as being
well-adapted to frequent flooding (Flecker and Feifarek, 1994; Winterbourn,
1997), but they recover at rates (Scrimgeour et al., 1988; Flecker and Feifarek,
1994) similar to those of species in some northern hemisphere streams (Matthaei
et al., 1997a). There is also evidence that, in some regions, recovery rate depends
on season: recovery from a flood in a tributary of the River Araglin, Ireland was
much slower in summer than in winter, when flooding is more typical (Giller
et al., 1991). Again, although we know recovery is relatively rapid in streams
following floods, we still do not have a clear understanding of how disturbance
history, productivity, nature of stream food base and other factors affect the rate
of this recovery.

Why is Recovery Quick?

A number of small-scale colonization and disturbance studies have found drift

and movement over the stream substrate to be the main mechanism of (re)colo-
nization (Williams and Hynes, 1976; Mackay, 1992; Downes et al., 2005),
although most studies have recorded only rates of colonization without examin-
ing colonization pathways (Downes and Keough, 1998). Colonization pathways
also differ among taxa within a community. Downes et al. (2005) found some
species of hydropsychid caddisflies colonized by drifting, whereas others used
walking as well as drift. However, I am aware of only two studies that have exam-
ined colonization pathways following an actual flood: aerial recolonists dominated
recovery of diversity following floods in a Sonoran desert stream, but drift and
upstream movements provided the most individuals (Gray and Fisher 1981).
Matthaei et al. (1997b) compared drift with movement within and over the sub-
strate and found drift to be the principal mechanism of recolonization. Others
have implied colonization from aerial sources (Siegfried and Knight, 1977;
Flecker and Feifarek, 1994; Lytle, 2000) and walking along the substrate (Brooks
and Boulton, 1991), based on the type and size of organisms found post-flood
and knowledge of life-history patterns. Several authors (Wallace, 1990; Mackay,
1992) have also attributed differences in seasonal and regional recovery rates to
the presence of adult insects for oviposition. Although there has been consider-
able speculation and investigation of refugia from flood events, corresponding
studies have not been made of the relative contribution of recolonists from
instream refugia compared with recolonization from more distant sources in
adjoining catchments or upstream reaches. However, the use of molecular mark-
ers by Schmidt et al. (1995), Hughes (2007) and others to trace the ancestral
streams of invertebrates is beginning to illuminate the relative contribution of
those more distant colonization sources. The mobility of stream invertebrates
through drift, walking or flying is a critical factor in the rapid recovery of invertebrate
110 R.G. Death

communities from dramatic losses. It has generally been assumed that drift is the
principal mechanism of flood recovery (Minshall and Petersen, 1985; Townsend,
1989), but aerial (Gray and Fisher, 1981; Hughes, 2007) and within-stream
(Matthaei et al., 2000; Downes et al., 2005) movement are equally important for
some taxa and some streams.

How do Floods Affect Community Composition?

Disturbances from floods can have severe effects on invertebrate communities
and occur during at least part of the year in most streams and rivers worldwide;
it is therefore unsurprising that they have been considered one of the main driv-
ing forces of lotic community structure (Resh et al., 1988; Lake, 1990, 2000).
The effects of high flow events (Holomuzki and Biggs, 1999, 2000) and recovery
from floods (e.g. Scrimgeour et al., 1988; Downes and Lake, 1991; Mackay,
1992) depend greatly on the taxon. Ephemeroptera, Simuliidae and Chironomi-
dae generally seem to characterize faunas soon after floods, whereas Mollusca,
cased Trichoptera, Amphipoda, Platyhelminthes and some Hydropsychidae
seem more susceptible to extirpation from floods (Mackay, 1992; Death, 1996;
Townsend et al., 1997a; Robinson et al., 2003). Therefore, invertebrate faunas
differ markedly between flood-prone and stable streams (Scarsbrook and
Townsend, 1993; Death, 1995; Barquin and Death, 2004, 2006), although these
differences are greater in open than in forested streams (Death, 2003; Death and
Zimmermann, 2005). Taxa in flood-prone streams are also characterized as hav-
ing small size, high adult mobility, generalist habitats, clinger habits, streamlined
or flattened bodies, and two or more life stages outside the stream (Townsend
et al., 1997a).
Small floods create a patchy stream bed that maintains habitat heterogeneity
and thus invertebrate diversity (Robinson et al., 2003; Lepori and Hjerdt, 2006).
However, inter-stream variability and regional diversity can be greater in streams
less prone to floods (Death, 1995). All of the flood-prone streams in the study
region had a remarkably similar fauna dominated by the mayfly Deleatidum,
Simuliidae, Chironomidae and an oligochaete, presumably those taxa most able
to recover from flood disturbances. However, there have been very few between-
stream studies of flood effects on community composition on which to judge the
generality of this finding. Potential changes to global rainfall patterns might lead
to the homogenization of stream faunas with only taxa able to survive floods (or
droughts) remaining abundant in some areas (Fowler and Hennessy, 1995; Lake
et al., 2000).
Changes in the abundance of taxa following substrate disturbance from
floods have been proposed as examples of competitive exclusion along a succes-
sional pathway (Fisher et al., 1982; Hart, 1983; Hemphill and Cooper, 1983).
However, few studies show successional change in stream faunas following
floods (Reice, 1985) and in other studies successional change was not detected
(Scrimgeour et al., 1988; Death, 2006; Ledger et al., 2006). Some taxa domi-
nate colonization processes following floods, but whether they are then replaced
or simply joined by more slowly colonizing taxa remains unclear.
Effect of Floods on Aquatic Invertebrates 111

Diversity–Disturbance Relationships

The relationship between diversity and flood frequency or disturbance has been
much studied in streams (Robinson and Minshall, 1986; Death and Winterbourn,
1995; Townsend et al., 1997b; Death, 2002). From their study of Otago, New
Zealand streams, Townsend et al. (1997b) claimed support for the idea that
diversity is greatest at intermediate levels of disturbance (the intermediate distur-
bance hypothesis: Grime, 1973; Connell, 1978). However, Death and Winter-
bourn (1995) and Death (2002), in studies of other New Zealand streams, found
no support for this intermediate disturbance hypothesis, nor a productivity-
dependent, hump-shaped relationship (the dynamic equilibrium hypothesis:
Huston, 1979). A review of stream diversity patterns by Vinson and Hawkins
(1998) concluded that most studies showed a linear increase in diversity with
decreasing disturbance.
Barquin and Death (2004, 2006) found that disturbance–diversity relation-
ships differed between northern and southern hemisphere streams. Stable spring-
brooks in northern Spain had lower diversity than similar more flood-prone
streams, whereas springbrooks in New Zealand had higher diversity than similar
flood-prone streams (Fig. 6.2). Changes to the food base of a stream can also
alter the relationship between disturbance and diversity. Diversity in open streams
is reduced by floods, but not in forest streams (Death and Zimmermann, 2005),
and low-nutrient alpine streams are affected by disturbance, but not when nutrients
are added (Gafner and Robinson, 2007).
The wider ecological debate on neutral theory and metacommunities, sug-
gesting that community structure is determined by colonization dynamics and
not by niche partitioning (Hubbell, 2001; Leibold et al., 2004), is particularly

30

25
Number of taxa (0.1 m2)

20

15

10

0
Arthurs Pass Nelson Ruapehu Taranaki Spain

Fig. 6.2. Mean (± 1 SE) number of taxa in springbrooks (open bars) and rhithral
springs (closed bars) in four regions of New Zealand and Spain. See Barquin and
Death (2004, 2006) for more details.
112 R.G. Death

appropriate for understanding the above patterns in stream communities (Heino,

2005; Thompson and Townsend, 2006). Colonization of small patches of stream
substrate in several New Zealand streams is consistent with these ideas (Death,
2006).

Management Implications of Floods for Maintaining Ecosystem

Integrity
Riverine habitats are increasingly under threat from flow manipulation (Poff
et al., 2003; Richter et al., 2003; Poff et al., 2007) and land-use intensification
(Allan, 2004). Flow manipulations often involve water abstraction and/or dam
construction that reduce the variability of flow, removing the structuring influence
of floods on invertebrate communities. These changes can lead to dramatic, often
deleterious, shifts in community composition (Poff et al., 1997). For example,
alteration of flow patterns in the Tongariro River, New Zealand have led to the
proliferation of periphyton and a corresponding shift from an Ephemeroptera,
Plecoptera and Trichoptera dominated fauna to one dominated by Chironomi-
dae and Oligochaeta (Tonkin, 2006) (Fig. 6.3).
Adaptive management projects are under way around the world to better
understand what level of flow variability is required to maintain some semblance
of ecosystem integrity downstream of dams (Poff et al., 1997; Patten and Stevens,
2001; Robinson et al., 2003). For example, a regime of flushing flow discharges,

60

50

40

30
%EPT

20

10

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

-2
Chlorophyll a (µg cm )

Fig. 6.3. Mean (± 1 SE) per cent Ephemeroptera, Plecoptera and Trichoptera in
the drift as a function of chlorophyll a at 15 sites along the Tongariro River, New
Zealand. Square symbols = sites above dams; circular symbols = sites below upper
Rangipo dam; triangular symbols = sites below Rangipo dam and Poutu Intake.
Effect of Floods on Aquatic Invertebrates 113

triggered by periphyton cover, has been proposed to alleviate the effects of flow
reduction in the Tongariro River, although the magnitude and frequency of the
discharges required to restore an Ephemeroptera, Plecoptera and Trichoptera (EPT)-
dominated community is still unclear.
One of the major anthropogenic influences on streams is modification of
catchment land use, which tends to alter the flood regime so that floods are
greater in magnitude and more flashy, changing the types of invertebrates pres-
ent (Rowe et al., 1997; Walsh et al., 2005b). Changes in flood regime have been
identified as one of the principal driving forces behind changes in the inverte-
brate communities of urban streams (Roy et al., 2005; Walsh et al., 2005a,b), but
the effects of changes in flood regime that result from other land-use modifica-
tions have not been extensively investigated. Melo et al. (2003) subjected 90-m
stream reaches to a standardized hydrological disturbance and found that recov-
ery of invertebrate communities was similar in both native tussock streams and
exotic pasture streams. However, their experiment did not mimic the changes in
magnitude and flashiness of floods created by alteration to land-use patterns.
Collier and Quinn’s (2003) more realistic evaluation did find that communities in
a pasture stream recovered from a 1 in 28-year flood event more slowly than
those in a nearby forest stream. This finding confirms that major catchment mod-
ification, particularly deforestation, can alter the effects of flood events on stream
invertebrates.

Conclusions

Disturbance from floods is one of the most important forces structuring stream
invertebrate communities worldwide (Resh et al., 1988; Lake, 2000). There have
been numerous phenomenological studies of actual floods and experimental
studies of small-scale substrate disturbances, and a few combining both
approaches. However, we still have much to learn about what invertebrates do
during floods, their relative use of potential refugia, local and regional factors
controlling recovery rates and the relationship between disturbance regimes and
diversity. Furthermore, we need to understand those effects in order to minimize
changes in these ecosystems produced by modification of flow regimes by
humans.

Acknowledgements

Thanks to Fiona Death, Alex James, Rob Briers and two particularly helpful referees
for comments on an early draft of this chapter.

References

Allan, J.D. (2004) Landscapes and riverscapes: the influence of land use on stream eco-
systems. Annual Review of Ecology, Evolution and Systematics 35, 257–284.
114 R.G. Death

Angradi, T.R. (1997) Hydrologic context and macroinvertebrate community response

to floods in an Appalachian headwater stream. American Midland Naturalist 138,
371–386.
Arunachalam, M., Nair, K.C.M., Vijverberg, J., Kortmulder, K. and Suriyanarayanan, H.
(1991) Substrate selection and seasonal variation in densities of invertebrates in
stream pools of a tropical river. Hydrobiologia 213, 141–148.
Barquin, J. and Death, R.G. (2004) Patterns of invertebrate diversity in streams and fresh-
water springs in northern Spain. Archiv für Hydrobiologie 161, 329–349.
Barquin, J. and Death, R.G. (2006) Spatial patterns of macroinvertebrate diversity in
New Zealand springbrooks and rhithral streams. Journal of the North American
Benthological Society 25, 768–786.
Beveridge, O.S. and Lancaster, J. (2007) Sub-lethal effects of disturbance on a predatory
net-spinning caddisfly. Freshwater Biology 52, 491–499.
Biggs, B.J.E., Duncan, M.J., Francoeur, S.N. and Meyer, W.D. (1997) Physical charac-
terisation of microform bed cluster refugia in 12 headwater streams, New Zealand.
New Zealand Journal of Marine and Freshwater Research 31, 413–422.
Biggs, B.J.F., Smith, R.A. and Duncan, M.J. (1999) Velocity and sediment disturbance of
periphyton in headwater streams: biomass and metabolism. Journal of the North
American Benthological Society 18, 222–241.
Bishop, J.E. (1973) Observations on the vertical distribution of the benthos in a Malaysian
stream. Freshwater Biology 3, 147–156.
Bond, N.R. and Downes, B.J. (2000) Flow-related disturbance in streams: an experimental
test of the role of rock movement in reducing macroinvertebrate population densities.
Marine and Freshwater Research 51, 333–337.
Bond, N.R. and Downes, B.J. (2003) The independent and interactive effects of fine
sediment and flow on benthic invertebrate communities characteristic of small upland
streams. Freshwater Biology 48, 455–465.
Boulton, A.J., Spangaro, G.M. and Lake, P.S. (1988) Macroinvertebrate distribution and
recolonization on stones subjected to varying degrees of disturbance: an experimental
approach. Archiv für Hydrobiologie 113, 551–576.
Boulton, A.J., Peterson, C.G., Grimm, N.B. and Fisher, S.G. (1992) Stability of an aquatic
macroinvertebrate community in a multiyear hydrologic disturbance regime. Ecology
73, 2192–2207.
Brewin, P.A., Buckton, S.T. and Ormerod, S.J. (2000) The seasonal dynamics and
persistence of stream macroinvertebrates in Nepal: do monsoon floods represent
disturbance? Freshwater Biology 44, 581–594.
Brittain, J.E. and Eikeland, T.J. (1988) Invertebrate drift – a review. Hydrobiologia 166,
77–93.
Brooks, S.S. (1998) Impacts of flood disturbance on the macroinvertebrate assemblage of
an upland stream. Unpublished PhD thesis, Monash University, Melbourne, Australia.
Brooks, S.S. and Boulton, A.J. (1991) Recolonization dynamics of benthic macroinverte-
brates after artificial and natural disturbances in an Australian temporary stream.
Australian Journal of Marine and Freshwater Research 42, 295–308.
Carling, P.A. (1987) Bed stability in gravel streams, with reference to stream regulation
and ecology. In: Richards, K. (ed.) River Channels: Environment and Process. Blackwell,
Oxford, pp. 321–347.
Clausen, B. and Plew, D. (2004) How high are bed-moving flows in New Zealand rivers?
Journal of Hydrology (New Zealand) 43, 19–37.
Cobb, D.G., Galloway, T.D. and Flannagan, J.F. (1992) Effects of discharge and substrate
stability on density and species composition of stream insects. Canadian Journal of
Fisheries and Aquatic Sciences 49, 1788–1795.
Effect of Floods on Aquatic Invertebrates 115

Collier, K.J. and Quinn, J.M. (2003) Land-use influences macroinvertebrate community
response following a pulse disturbance. Freshwater Biology 48, 1462–1481.
Connell, J.H. (1978) Diversity in tropical rain forests and coral reefs. Science 199, 1302–
1310.
Death, R.G. (1995) Spatial patterns in benthic invertebrate communities: products of
habitat stability or are they habitat specific? Freshwater Biology 33, 455–467.
Death, R.G. (1996) The effect of patch disturbance on stream invertebrate community
structure: the influence of disturbance history. Oecologia 108, 567–576.
Death, R.G. (2002) Predicting invertebrate diversity from disturbance regimes in forest
streams. Oikos 97, 18–30.
Death, R.G. (2003) Spatial patterns in benthic invertebrate community composition: is
substrate disturbance actually important? Canadian Journal of Fisheries and Aquatic
Sciences 60, 603–611.
Death, R.G. (2006) Colonisation in New Zealand streams: predictable patterns or chance
events? New Zealand Natural Sciences 31, 93–112.
Death, R.G. and Winterbourn, M.J. (1994) The measurement of environmental stability
in streams: a multivariate approach. Journal of the North American Benthological
Society 13, 125–139.
Death, R.G. and Winterbourn, M.J. (1995) Diversity patterns in stream benthic inverte-
brate communities: the influence of habitat stability. Ecology 76, 1446–1460.
Death, R.G. and Zimmermann, E.M. (2005) Interaction between disturbance and primary
productivity in determining stream invertebrate diversity. Oikos 111, 392–402.
Doeg, T.J., Lake, P.S. and Marchant, R. (1989) Colonization of experimentally disturbed
patches by stream macroinvertebrates in the Acheron River, Victoria. Australian
Journal of Ecology 14, 207–220.
Dole-Olivier, M.-J. and Marmonier, P. (1992) Effects of spates on the vertical distribution
of the interstitial community. Hydrobiologia 230, 49–61.
Dole-Olivier, M.-J., Marmonier, P. and Beffy, J.-L. (1997) Response of invertebrates to
lotic disturbance: is the hyporheic zone a patchy refugium? Freshwater Biology 37,
257–276.
Downes, B.J. and Keough, M.J. (1998) Scaling of colonization processes in streams: paral-
lels and lessons from marine hard substrata. Australian Journal of Ecology 23, 8–26.
Downes, B.J. and Lake, P.S. (1991) Different colonization patterns of two closely related
stream insects (Austrosimulium spp.) following disturbance. Freshwater Biology 26,
295–306.
Downes, B.J. and Street, J.L. (2005) Regrowth or dispersal? Recovery of a freshwater red
alga following disturbance at the patch scale. Austral Ecology 30, 526–536.
Downes, B.J., Lake, P.S., Glaister, A. and Webb, J.A. (1998) Scales and frequencies of
disturbances: rock size, bed packing and variation among upland streams. Freshwater
Biology 40, 625–639.
Downes, B.J., Bellgrove, A. and Street, J.L. (2005) Drifting or walking? Colonisation
routes used by different instars and species of lotic, macroinvertebrate filter feeders.
Marine and Freshwater Research 56, 815–824.
Fisher, S.G. (1987) Succession, scale, and hypothesis testing in streams. Canadian Journal
of Fisheries and Aquatic Sciences 44, 689.
Fisher, S.G., Gray, L.J., Grimm, N.B. and Busch, D.E. (1982) Temporal succession in a
desert stream ecosystem following flash flooding. Ecological Monographs 52, 93–110.
Flecker, A.S. and Feifarek, B. (1994) Disturbance and the temporal variability of inverte-
brate assemblages in two Andean streams. Freshwater Biology 31, 131–142.
Fowler, A.M. and Hennessy, K.J. (1995) Potential impacts of global warming on the
frequency and magnitude of heavy precipitation. Natural Hazards 11, 283–303.
116 R.G. Death

Fritz, K.M. and Dodds, W.K. (2004) Resistance and resilience of macroinvertebrate
assemblages to drying and flood in a tallgrass prairie stream system. Hydrobiologia
527, 99–112.
Gafner, K. and Robinson, C.T. (2007) Nutrient enrichment influences the responses of
stream macroinvertebrates to disturbance. Journal of the North American Bentho-
logical Society 26, 92–102.
Gibbins, C.N., Scott, E., Soulsby, C. and McEwan, I. (2005) The relationship between
sediment mobilisation and the entry of Baetis mayflies into the water column in a
laboratory flume. Hydrobiologia 533, 115–122.
Giberson, D.J. and Caissie, D. (1998) Stream habitat hydraulics: interannual variability in
three reaches of Catamaran Brook, New Brunswick. Canadian Journal of Fisheries
and Aquatic Sciences 55, 485–494.
Giberson, D.J. and Hall, R.J. (1988) Seasonal-variation in faunal distribution within the
sediments of a Canadian Shield stream, with emphasis on responses to spring floods.
Canadian Journal of Fisheries and Aquatic Sciences 45, 1994–2002.
Giller, P.S., Sangpradub, N. and Twomey, H. (1991) Catastrophic flooding and macroin-
vertebrate community structure. Verhandlungen der Internationalen Vereinigung für
Theoretische und Angewandte Limnologie 24, 1724–1729.
Gjerløv, C., Hildrew, A.G. and Jones, J.I. (2003) Mobility of stream invertebrates in rela-
tion to disturbance and refugia: a test of habitat templet theory. Journal of the North
American Benthological Society 22, 207–223.
Gordon, N.D., McMahon, T.A., Finlayson, B.L., Gippel, C.J. and Nathan, R.J. (2004)
Stream Hydrology: An Introduction for Ecologists, 2nd edn. Wiley, Chichester, UK.
Gray, L.J. (1981) Species composition and life history of aquatic insects in a lowland
Sonoran Desert stream. American Midland Naturalist 106, 229–242.
Gray, L.J. and Fisher, S.G. (1981) Post-flood recolonization pathways of macroinverte-
brates in a lowland Sonoran Desert stream. American Midland Naturalist 106,
249–257.
Grime, J.P. (1973) Control of species density in herbaceous vegetation. Journal of Envi-
ronmental Management 1, 151–167.
Grimm, N.B. and Fisher, S.G. (1989) Stability of periphyton and macroinvertebrates to
disturbance by flash floods in a desert stream. Journal of the North American Bentho-
logical Society 8, 293–307.
Hart, D.D. (1983) The importance of competitive interactions within stream populations
and communities. In: Barnes, J.R. and Minshall, G.W. (eds) Stream Ecology. Applica-
tion and Testing of General Ecological Theory. Plenum Press, New York, pp. 99–136.
Hauer, F.R. and Stanford, J.A. (1982) Ecological responses of hydropsychid caddisflies to
stream regulation. Canadian Journal of Fisheries and Aquatic Sciences 39, 1235–
1242.
Heino, J. (2005) Metacommunity patterns of highly diverse stream midges: gradients,
chequerboards, and nestedness, or is there only randomness? Ecological Entomology
30, 590–599.
Hemphill, N. and Cooper, S.D. (1983) The effect of physical disturbance on the relative
abundances of two filter-feeding insects in a small stream. Oecologia 58, 378–383.
Holomuzki, J.R. and Biggs, B.J.F. (1999) Distributional responses to flow disturbance by
a stream-dwelling snail. Oikos 87, 36–47.
Holomuzki, J.R. and Biggs, B.J.F. (2000) Taxon-specific responses to high-flow distur-
bance in streams: implications for population persistence. Journal of the North
American Benthological Society 19, 670–679.
Hubbell, S.P. (2001) The Unified Neutral Theory of Biodiversity and Biogeography, Princeton
University Press, Princeton, New Jersey.
Effect of Floods on Aquatic Invertebrates 117

Hughes, J.M. (2007) Constraints on recovery: using molecular methods to study connec-
tivity of aquatic biota in rivers and streams. Freshwater Biology 52, 616–631.
Huston, M. (1979) A general hypothesis of species diversity. American Naturalist 113,
81–101.
Irvine, J.R. (1985) Effects of successive flow perturbations on stream invertebrates. Canadian
Journal of Fisheries and Aquatic Sciences 42, 1922–1927.
Jones, J.R.E. (1951) An ecological study of the River Towy. Journal of Animal Ecology
20, 68–86.
Lake, P.S. (1990) Disturbing hard and soft bottom communities: a comparison of marine
and freshwater environments. Australian Journal of Ecology 15, 477–488.
Lake, P.S. (2000) Disturbance, patchiness, and diversity in streams. Journal of the North
American Benthological Society 19, 573–592.
Lake, P.S., Palmer, M.A., Biro, P., Cole, J., Covich, A.P., Dahm, C., Gibert, J., Goedkoop,
W., Martens, K. and Verhoeven, J. (2000) Global change and the biodiversity of
freshwater ecosystems: impacts on linkages between above-sediment and sediment
biota. BioScience 50, 1099–1107.
Lamberti, G.A., Gregory, S.V., Ashkenas, L.R., Wildman, R.C. and Moore, K.M.S. (1991)
Stream ecosystem recovery following a catastrophic debris flow. Canadian Journal of
Fisheries and Aquatic Sciences 48, 196–208.
Lancaster, J. (2000) Geometric scaling of microhabitat patches and their efficacy as refugia
during disturbance. Journal of Animal Ecology 69, 442–457.
Lancaster, J. and Belyea, L.R. (1997) Nested hierarchies and scale-dependence of mech-
anisms of flow refugium use. Journal of the North American Benthological Society
16, 221–238.
Lancaster, J. and Hildrew, A.G. (1993) Flow refugia and the microdistribution of lotic
invertebrates. Journal of the North American Benthological Society 12, 385–393.
Ledger, M., Harris, R., Milner, A. and Armitage, P. (2006) Disturbance, biological legacies
and community development in stream mesocosms. Oecologia 148, 682–691.
Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J.M., Hoopes, M.F.,
Holt, R.D., Shurin, J.B., Law, R., Tilman, D., Loreau, M. and Gonzalez, A. (2004)
The metacommunity concept: a framework for multi-scale community ecology. Ecology
Letters 7, 601–613.
Lepori, F. and Hjerdt, N. (2006) Disturbance and aquatic biodiversity: reconciling
contrasting views. BioScience 56, 809–818.
Lytle, D.A. (2000) Biotic and abiotic effects of flash flooding in a montane desert stream.
Archiv für Hydrobiologie 150, 85–100.
Lytle, D.A. and Poff, L.N. (2004) Adaptation to natural flow regimes. Trends in Ecology &
Evolution 19, 94–100.
Mackay, R.J. (1992) Colonization by lotic macroinvertebrates: a review of processes and
patterns. Canadian Journal of Fisheries and Aquatic Sciences 49, 617–628.
Maitland, P.S. (1963) Quantitative studies on the invertebrate fauna of sandy and stony
substrates in River Endrick Scotland. Proceedings of the Royal Society of Edinburgh
B 68, 277–301.
Marchant, R., Lake, P.S. and Doeg, T.J. (1991) Longitudinal variation in recolonization
rates of macroinvertebrates along an upland river in south-eastern Australia. Fresh-
water Biology 25, 349–356.
Marmonier, P. and Chatelliers, M.C.D. (1991) Effects of spates on interstitial assemblages
of the Rhone river – importance of spatial heterogeneity. Hydrobiologia 210,
243–251.
Matthaei, C.D. and Huber, H. (2002) Microform bed clusters: are they preferred habitats
for invertebrates in a flood-prone stream? Freshwater Biology 47, 2174–2190.
118 R.G. Death

Matthaei, C.D. and Townsend, C.R. (2000a) Inundated floodplain gravels in a stream
with an unstable bed: temporary shelter or true invertebrate refugium? New Zealand
Journal of Marine and Freshwater Research 34, 147–156.
Matthaei, C.D. and Townsend, C.R. (2000b) Long-term effects of local disturbance history
on mobile stream invertebrates. Oecologia 125, 119–126.
Matthaei, C.D., Uehlinger, U., Meyer, E.I. and Frutiger, A. (1996) Recolonization by benthic
invertebrates after experimental disturbance in a Swiss pre-alpine river. Freshwater
Biology 35, 233–248.
Matthaei, C.D., Uehlinger, U. and Frutiger, A. (1997a) Response of benthic invertebrates
to natural versus experimental disturbance in a Swiss pre-alpine river. Freshwater
Biology 37, 61–77.
Matthaei, C.D., Wethmuller, D. and Frutiger, A. (1997b) Invertebrate recovery from a
bed-moving spate – the role of drift versus movements inside or over the substratum.
Archiv für Hydrobiologie 140, 221–235.
Matthaei, C.D., Peacock, K.A. and Townsend, C.R. (1999) Patchy surface stone move-
ment during disturbance in a New Zealand stream and its potential significance for
the fauna. Limnology and Oceanography 44, 1091–1102.
Matthaei, C.D., Arbuckle, C.J. and Townsend, C.R. (2000) Stable surface stones as refugia
for invertebrates during disturbance in a New Zealand stream. Journal of the North
American Benthological Society 19, 82–93.
Matthaei, C.D., Townsend, C.R., Arbuckle, C.A., Peacock, K.A., Guggelberger, C., Küster,
C.E. and Huber, H. (2004) Disturbance, assembly rules and benthic communities in
running waters: a review and some implications for restoration projects. In: Temper-
ton, V.M., Hobbs, R.J., Nuttle, T. and Halle, S. (eds) Assembly Rules and Restoration
Ecology: Bridging the Gap Between Theory and Practice. Island Press, Washington,
DC, pp. 367–388.
Melo, A.S., Niyogi, D.K., Matthaei, C.D. and Townsend, C.R. (2003) Resistance, resilience,
and patchiness of invertebrate assemblages in native tussock and pasture streams in
New Zealand after a hydrological disturbance. Canadian Journal of Fisheries and
Aquatic Sciences 60, 731–739.
Minshall, G.W. (1988) Stream ecosystem theory: a global perspective. Journal of the
North American Benthological Society 7, 263–288.
Minshall, G.W. and Petersen, R.C. (1985) Towards a theory of macroinvertebrate commu-
nity structure in stream ecosystems. Archiv für Hydrobiologie 104, 49–76.
Minshall, G.W., Andrews, D.A. and Manuel-Faler, C.Y. (1983) Application of island bio-
geographic theory to streams: macroinvertebrate recolonization of the Teton River,
Idaho. In: Barnes, J.R. and Minshall, G.W. (eds) Stream Ecology: Application and
Testing of General Ecological Theory. Plenum Press, New York, pp. 279–297.
Olsen, D.A. and Townsend, C.R. (2005) Flood effects on invertebrates, sediments and
particulate organic matter in the hyporheic zone of a gravel-bed stream. Freshwater
Biology 50, 839–853.
Palmer, M.A., Bely, A.E. and Berg, K.E. (1992) Response of invertebrates to lotic disturbance:
a test of the hyporheic refuge hypothesis. Oecologia 89, 182–194.
Palmer, M.A., Arensburger, P., Martin, A.P. and Denman, D.W. (1996) Disturbance and
patch specific responses: the interactive effects of woody debris and floods on lotic
invertebrates. Oecologia 105, 247–257.
Patten, D.T. and Stevens, L.E. (2001) Restoration of the Colorado River ecosystem using
planned flooding. Ecological Applications 11, 633–634.
Poff, N.L. (1992) Why disturbances can be predictable: a perspective on the definition of
disturbance in streams. Journal of the North American Benthological Society 11,
86–92.
Effect of Floods on Aquatic Invertebrates 119

Poff, N.L. and Ward, J.V. (1990) Physical habitat template of lotic systems: recovery in the
context of historical pattern of spatiotemporal heterogeneity. Environmental
Management 14, 629–645.
Poff, N.L. and Ward, J.V. (1991) Drift responses of benthic invertebrates to experimental
streamflow variation in a hydrologically stable stream. Canadian Journal of Fisheries
and Aquatic Sciences 48, 1926–1936.
Poff, L.N., Allan, J.A., Bain, M.B., Karr, J.R., Prestegaard, K.L., Richter, B.D., Sparks,
R.E. and Stromberg, J.C. (1997) The natural flow regime: a paradigm for river
conservation and restoration. Bioscience 47, 769–784.
Poff, N.L., Allan, J.D., Palmer, M.A., Hart, D.D., Richter, B.D., Arthington, A.H., Rogers,
K.H., Meyers, J.L. and Stanford, J.A. (2003) River flows and water wars: emerging
science for environmental decision making. Frontiers in Ecology and the Environment
1, 298–306.
Poff, N.L., Olden, J.D., Pepin, D.M. and Bledsoe, B.P. (2006) Placing global stream flow
variability in geographic and geomorphic contexts. River Research and Applications
22, 149–166.
Poff, N.L., Olden, J.D., Merritt, D.M. and Pepin, D.M. (2007) Homogenization of regional
river dynamics by dams and global biodiversity implications. Proceedings of the
National Academy of Sciences of the USA 104, 5732–5737.
Reice, S.R. (1985) Experimental disturbance and the maintenance of species diversity in
a stream community. Oecologia 67, 90–97.
Reice, S.R., Wissmar, R.C. and Naiman, R.J. (1990) Disturbance regimes, resilience, and
recovery of animal communities and habitats in lotic ecosystems. Environmental
Management 14, 647–659.
Rempel, L.L., Richardson, J.S. and Healey, M.C. (1999) Flow refugia for benthic macro-
invertebrates during flooding of a large river. Journal of the North American Bentho-
logical Society 18, 34–48.
Resh, V.H., Brown, A.V., Covich, A.P., Gurtz, M.E., Li, H.W., Minshall, G.W., Reice, S.R.,
Sheldon, A.L., Wallace, J.B. and Wissmar, R.C. (1988) The role of disturbance in
stream ecology. Journal of the North American Benthological Society 7, 433–455.
Richter, B.D., Mathews, R., Harrison, D.L. and Wigington, R. (2003) Ecologically sustain-
able water management: managing river flows for ecological integrity. Ecological
Applications 13, 206–224.
Robertson, A.L., Lancaster, J. and Hildrew, A.G. (1995) Stream hydraulics and the distri-
bution of microcrustacea: a role for refugia? Freshwater Biology 33, 469–484.
Robinson, C.T. and Minshall, G.W. (1986) Effects of disturbance frequency on stream
benthic community structure in relation to canopy cover and season. Journal of the
North American Benthological Society 5, 237–248.
Robinson, C.T. and Uehlinger, U. (2003) Using artificial floods for restoring river integrity.
Aquatic Sciences 65, 181–182.
Robinson, C.T., Tockner, K. and Ward, J.V. (2002) The fauna of dynamic riverine land-
scapes. Freshwater Biology 47, 661–677.
Robinson, C.T., Uehlinger, U. and Monaghan, M.T. (2003) Effects of a multi-year ex-
perimental flood regime on macroinvertebrates downstream of a reservoir. Aquatic
Sciences 65, 210–222.
Rosser, Z.C. and Pearson, R.G. (1995) Responses of rock fauna to physical disturbance in
two Australian tropical rainforest streams. Journal of the North American Bentho-
logical Society 14, 183–196.
Rowe, L., Fajhey, B., Jackson, R. and Duncan, M. (1997) Effects of land use on floods and low
flows. In: Mosely, M.P. and Pearson, C.P. (eds) Floods and Droughts: the New Zealand
Experience. New Zealand Hydrological Society Inc., Christchurch, pp. 89–102.
120 R.G. Death

Roy, A.H., Freeman, M.C., Freeman, B.J., Wenger, S.J., Ensign, W.E. and Meyer, J.L. (2005)
Investigating hydrologic alteration as a mechanism of fish assemblage shifts in urban-
izing streams. Journal of the North American Benthological Society 24, 656–678.
Scarsbrook, M.R. and Townsend, C.R. (1993) Stream community structure in relation to
spatial and temporal variation: a habitat templet study of two contrasting New Zealand
streams. Freshwater Biology 29, 395–340.
Schmidt, S.K., Hughes, J.M. and Bunn, S.E. (1995) Gene flow among conspecific popu-
lations of Baetis sp. (Ephemeroptera): adult flight and larval drift. Journal of the
North American Benthological Society 14, 147–157.
Scrimgeour, G.J., Davidson, R.J. and Davidson, J.M. (1988) Recovery of benthic macro-
invertebrate and epilithic communities following a large flood, in an unstable braided,
New Zealand river. New Zealand Journal of Marine and Freshwater Research 22,
337–344.
Sedell, J.R., Reeves, G., Hauer, F.R., Stanford, J.A. and Hawkins, B.A. (1990) Role of
refugia in recovery from disturbances: modern fragmented and disconnected river
systems. Environmental Management 14, 711–724.
Siegfried, C.A. and Knight, A.W. (1977) The effects of a washout in a Sierra foothill
stream. American Midland Naturalist 98, 200–207.
Snyder, C.D. and Johnson, Z.B. (2006) Macroinvertebrate assemblage recovery following
a catastrophic flood and debris flows in an Appalachian mountain stream. Journal of
the North American Benthological Society 25, 825–840.
Suren, A.M. and Duncan, M.J. (1999) Rolling stones and mosses: effects of substrate
stability on bryophyte communities in streams. Journal of the North American
Benthological Society 18, 457–467.
Thompson, R. and Townsend, C.R. (2006) A truce with neutral theory: local deterministic
factors, species traits and dispersal limitation together determine patterns of diversity
in stream invertebrates. Journal of Animal Ecology 75, 476–484.
Thorup, J. (1970) Influence of a short-termed flood on a springbrook community. Archiv
für Hydrobiologie 66, 447–457.
Tonkin, J.D. (2006) The effects of flow regulation on macroinvertebrate drift in the
Tongariro River, New Zealand. Unpublished BSc (Hons) thesis, Massey University,
Palmerston North.
Townsend, C.R. (1989) The patch dynamics concept of stream community ecology. Journal
of the North American Benthological Society 8, 36–50.
Townsend, C.R. and Hildrew, A.G. (1994) Species traits in relation to a habitat templet for
river systems. Freshwater Biology 31, 265–275.
Townsend, C.R., Dolédec, S. and Scarsbrook, M.R. (1997a) Species traits in relation to
temporal and spatial heterogeneity in streams: a test of habitat templet theory. Fresh-
water Biology 37, 367–387.
Townsend, C.R., Scarsbrook, M.R. and Dolédec, S. (1997b) The intermediate disturbance
hypothesis, refugia, and biodiversity in streams. Limnology and Oceanography 42,
938–949.
Turcotte, P. and Harper, P.P. (1982) The macroinvertebrate fauna of a small Andean
stream. Freshwater Biology 12, 411–419.
Vinson, M.R. and Hawkins, C.P. (1998) Biodiversity of stream insects: variation at local,
basin, and regional scales. Annual Review of Entomology 43, 271–293.
Wallace, J.B. (1990) Recovery of lotic macroinvertebrate communities from disturbance.
Environmental Management 14, 605–620.
Walsh, C.J., Fletcher, T.D. and Ladson, A.R. (2005a) Stream restoration in urban catch-
ments through redesigning stormwater systems: looking to the catchment to save the
stream. Journal of the North American Benthological Society 24, 690–705.
Effect of Floods on Aquatic Invertebrates 121

Walsh, C.J., Roy, A.H., Feminella, J.W., Cottingham, P.D., Groffman, P.M. and Morgan,
R.P. (2005b) The urban stream syndrome: current knowledge and the search for a
cure. Journal of the North American Benthological Society 24, 706–723.
Williams, D.D. and Hynes, H.B.N. (1976) The recolonization mechanism of stream benthos.
Oikos 27, 265–272.
Winterbottom, J., Orton, S., Hildrew, A.G. and Lancaster, J. (1997) Field experiments on
flow refugia in streams. Freshwater Biology 37, 569–580.
Winterbourn, M.J. (1997) New Zealand mountain stream communities: stable yet dis-
turbed? In: Streit, B., Stadler, T. and Lively, C.M. (eds) Evolutionary Ecology of
Freshwater Animals. Birkhauser Verlag, Basel, Switzerland, pp. 31–54.
Yount, J.D. and Niemi, G.J. (1990) Recovery of lotic communities and ecosystems from
disturbance – a narrative review of case-studies. Environmental Management 14,
547–569.
7 Life-history and Behavioural
Adaptations to Flow Regime in
Aquatic Insects
DAVID A. LYTLE
Department of Zoology, Oregon State University, 3029 Cordley Hall, Corvallis,
OR 97333, USA
E-mail: lytleda@[Link]

Abstract
Evolutionary history determines how aquatic insects cope with extreme flow events, both
natural and anthropogenic. From a macroevolutionary point of view, aquatic insects pos-
sess different modes of adaptation (morphological, behavioural, or life history) to cope
with flow events. For example, some Ephemeroptera, Plecoptera and Trichoptera are
adapted to long-term average cycles of flood and drought via life-history timing, while
other groups with aquatic adults (Hemiptera and Coleoptera) may use rain-triggered
behavioural responses to escape floods on a per-event basis. From a microevolutionary
point of view, populations may also evolve in response to flow regimes at very local
scales. Models of life-history and behavioural evolution illustrate how disturbance regime
characteristics (timing, frequency, predictability of events) and population structure
(course- versus fine-grained, organism lifespan) determine how populations might evolve
in response to different flood or drought regimes. All of these factors have implications for
the structure of natural lotic insect communities, and for the effective management of
dammed rivers and streams using ecological flows.

Introduction

Flood and drought disturbances are a fundamental part of most streams and rivers,
and play a central role in the regulation of populations (Hemphill and Cooper,
1983; Holomuzki and Biggs, 2000), the structuring of communities (Fisher et al.,
1982; Wootton et al., 1996; Townsend et al., 1997), and the functioning of eco-
systems (Grimm and Fisher, 1986; Townsend et al., 1998). Aquatic insects, in
particular, occupy the full spectrum of aquatic disturbance regimes, from the
most stable headwater spring to the most dynamic desert wash. This diversity of
habitats also means that aquatic insects experience a variety of selective pressures,
and as such are a model system for understanding how aquatic organisms evolve
in response to flood and drought disturbance. From a conservation point of view,
as natural flow regimes are altered by human endeavours such as dams and
 CAB International 2008. Aquatic Insects: Challenges to Populations
122 (eds J. Lancaster and R.A. Briers)
Life-history and Adaptation to Flow Regime 123

water diversions (Graf, 1993; Bunn and Arthington, 2002; Nilsson et al., 2005),
we need to understand how aquatic organisms might evolve, or fail to evolve, in
response to novel disturbance regimes. Some questions come immediately to
mind: as flow regimes are changed, will organisms adapt to these novel environ-
ments or simply vanish from the system? What are the population-level charac-
teristics that favour adaptation to local flood and drought regimes? To understand
these conservation-oriented issues, a basic understanding of how aquatic insects
evolve in response to flood and drought is required. This chapter seeks to iden-
tify some of the general biological processes that determine how aquatic insects
evolve in response to flood and drought.
Most populations of organisms are under selective pressure to evolve in new
directions and, at the same time, are constrained by their evolutionary past. Aquatic
insects are typical in this respect. The variety of general body plans represented by
the different insect orders presents an array of evolutionary constraints, while the
varied habitats occupied by aquatic insects present a myriad of evolutionary
challenges. This chapter will first outline how macroevolutionary patterns imposed
by respiratory physiology and general life cycle considerations may favour differ-
ent modes of adaptation (life history, behavioural or morphological) to flood or
drought, with an emphasis on behavioural and life-history adaptations. The focus
will then shift to microevolutionary patterns, with an emphasis on predictions
derived from current models of life-history and behavioural evolution.

Modes of Adaptation

Major body plans can be a constraining factor on the evolution of strategies for
coping with flood, drought, and flow variability in general. For aquatic insects,
some a priori predictions about the mode of adaptation (life-history, behavioural,
or morphological) can be made based on basic physiology and life cycle (Lytle
and Poff, 2004). In general, life-history adaptations are expected in taxa that
cannot respond immediately to disturbance events with behavioural avoidance,
but may be able to anticipate disturbance from seasonal cues. Behavioural adap-
tations involve using environmental cues correlated with disturbance (rainfall, for
example) as a signal to escape disturbance by moving within a stream or leaving
the stream entirely. Movement out of the stream is especially important for
mobile, air-breathing taxa that can respond to disturbances on a per-event basis.
Morphological adaptations to flood and drought include traits such as streamlin-
ing to avoid shear stress during floods and drought-resistant life stages (Townsend
and Hildrew, 1994; Poff et al., 2006). This chapter focuses specifically on life-
history and behavioural adaptations, and how these relate to flow regime com-
ponents such as disturbance timing, frequency and predictability.

Life-history adaptations

Life-history adaptations include emergence strategies that are synchronized to

avoid seasons when disturbances are likely (Gray, 1981; Gray and Fisher, 1981;
124 D.A. Lytle

Lytle, 2002), or the use of proximate cues such as habitat drying to alter growth
and development trajectories (De Block and Stoks, 2004, 2005; Danks, 2006).
Life-history adaptations might be expected in taxa that cannot effectively escape
disturbances on a per-event basis. In particular, taxa that require gill respiration
as juveniles but have an aerial adult stage (Ephemeroptera, Plecoptera, Tricho-
ptera, Odonata, Megaloptera and some Diptera) cannot exit the water immedi-
ately to avoid sudden flood events. However, they may be able to synchronize
emergence into the adult stage to coincide with the long-term mean expected
date of disturbance, thereby avoiding disturbance in most years. Life-history
adapted taxa ‘play the odds’ over long timescales; although the strategy may fail
occasionally if a disturbance is early or late (Lytle, 2003), this strategy maximizes
long-term expected fitness given the constraints imposed by disturbance (Lytle,
2001).
Unlike floods, the onset of drought is often accompanied by proximate cues
such as an increase in temperature and ionic concentrations, and a decrease in
flow rate and water level. Particularly in the Trichoptera and Plecoptera, these
cues may allow facultative entry into a drought-resistant stage (aerial adults or
diapausing juveniles) (Wiggins, 1973; Hynes, 1976; Wiggins et al., 1980; Deluc-
chi and Peckarsky, 1989; Brock et al., 2003), or provide a signal to adjust growth
and development rates accordingly (Shama and Robinson, 2006; Robinson and
Buser, 2007). In some cases, the transition to a drought-resistant stage occurs
constitutively at a particular time, typically in synchrony with the expected date
of drought occurrence (Wiggins et al., 1980; Williams, 1996). However, some
taxa apparently use asynchronous hatching of eggs as a way of ‘bet-hedging’
against the uncertainty of drought timing (Frutiger, 1996; Zwick, 1996). The
presence or absence of these traits is sometimes responsible for determining
species distributions along gradients of disturbance (Wissinger et al., 2003).
Trichoptera may provide the most extreme example of an aquatic insect
group that requires much advance notice that a disturbance will occur. The larvae
of most Trichoptera species obtain dissolved oxygen via gills, and their aquatic
pupal stage may last several weeks, during which time they may be exposed to
mortality from flood or drought (Wiggins, 1973). This commitment to an aquatic
environment precludes most behavioural strategies involving a temporary escape
from flood by leaving the stream and subsequently returning to the water. In this
instance a life-history strategy is favoured where emergence into the adult stage
is synchronized with the expected occurrence date of flood events (Gray, 1981;
Lytle, 2001).

Behavioural adaptations

Behaviours are context-dependent responses to different environmental cues.

Although behaviours are often considered more evolutionarily labile than
morphology or life history, studies have found a strong phylogenetic component
to many insect behaviours such as nest building in Hymenoptera (Wenzel, 1992),
web construction in spiders (Bond and Opell, 1998) and case construction in
Trichoptera (Wiggins, 2004).
Life-history and Adaptation to Flow Regime 125

Many behavioural adaptations for disturbance avoidance rely on a proxi-

mate cue (rainfall, rising hydrograph, etc.) to signal a disturbance, which is then
avoided by moving to safe zones within the stream or leaving the stream entirely.
Behavioural strategies tend to take disturbances on a ‘case by case’ basis, rather
than playing the long-run odds as in life-history strategies. For aquatic insects,
behaviours may be subdivided into those facilitating within-stream survival dur-
ing flood or drought, and those allowing escape from the stream during flood or
drought. Instances of behavioural flood escape are best known in the Hemiptera
and Coleoptera, possibly because their ability to breathe air allows them to per-
sist out of water for relatively long time periods (Lytle and White, 2007). Other
taxa that have the ability to persist out of water may possess similar behaviours
(e.g. some Trichoptera and Megaloptera: Anderson, 1967; Erman, 1981;
Contreras-Ramos, 1998), but more studies are needed to explore this. Within-
stream behaviours for avoiding flood disturbance include the utilization of
hydraulic refugia occurring in the substrate landscape, as well as adjustments to
normal movement patterns that reduce the chance of displacement (Olden et al.,
2004; Lancaster et al., 2006).
Flood escape behaviours include the use of proximate environmental cues
as a signal to abandon streams. This has been well documented in the giant
water bugs (Hemiptera: Belostomatidae), which use ‘rainfall response behav-
iour’ (RRB) to escape flash floods in desert streams. Individuals respond to
some characteristic duration of strong rainfall by crawling out of the stream,
moving perpendicularly away from the active channel over steep (sometimes
vertical) stream banks, and not stopping until they have reached protected
riparian areas (Lytle, 1999). The fitness benefit from RRB is substantial: the
flightless belostomatid Abedus herberti experiences about 15% mortality from
flash floods, while many other desert stream taxa endure mortality in excess of
95% (Grimm and Fisher, 1989; Lytle, 2000). Interestingly, RRB in belostoma-
tids turns out to be an ancient, phylogenetically conserved trait that likely
evolved to facilitate seasonal migration from perennial dry-season habitats to
temporary rainy-season pools (Lytle and Smith, 2004). The co-opted use of
RRB to escape flash floods is a more recent evolutionary event that has occurred
separately in at least two distinct belostomatid lineages (an ‘exaptation’ sensu
Gould and Vrba, 1982). The recent discovery of RRB in other aquatic Hemiptera
suggests that flood escape behaviours may be the rule, rather than the excep-
tion, at least for taxa inhabiting desert streams where flash floods are common-
place (Lytle and White, 2007). Although rainfall cues are important for many
taxa, in principle, other flood-associated cues such as changes in barometric
pressure, turbidity, cloud cover or hydraulic pressure waves could be important
as well.
For some taxa, behavioural escape from drought can be a viable strategy. Riv-
ers and streams can experience sudden drops in flow caused by diel fluctuations in
riparian plant evapotranspiration, lowering of the water table due to low rainfall
inputs, and recession to baseflow following flood events (Stanley et al., 1997).
Aquatic Coleoptera adults seem particularly adept at drought escape, and the
literature contains numerous accounts of beetle adults escaping drying habitats.
For example, the dytiscid beetles Rhantus gutticollis, R. binotatis and Eretes sticticus
126 D.A. Lytle

have been observed taking flight en masse from drying desert ponds, and in
some cases the emigration was preceded by loud, sustained vocalization
(Zimmerman, 1959; Smith, 1973; Kingsley, 1985). Within-stream movement is
also a mechanism for escaping the recession point of rapidly drying streams and
rivers. Lytle et al. (in press) observed thousands of adults of the long-toed water
beetle Postelichus immsi (Coleoptera: Dryopidae) crawling upstream concurrent
with diel recession of surface water in a desert river. Larvae of the grey sand-
dragon Progomphus borealis (Odonata: Gomphidae) were also observed bur-
rowing upstream in high densities (690 larvae per m2). Both taxa moved with
sufficient speed to reach upstream perennial river reaches before being over-
taken by drought. Given the proven ability of many aquatic insects to optimally
relocate within streams in response to local fine-scale differences in flow (Lancaster
and Hildrew, 1993; Lancaster, 1999; Wellnitz et al., 2001), for many taxa, behav-
ioural drought escape may be a special case of a more generalized ability to track
sudden changes in flow.

When do Ecological Conditions Lead to Evolutionary Change?

While macroevolutionary considerations may determine the mode of adaptation

to flood and drought, microevolutionary factors, such as ecological context and
population structure, may influence the degree to which populations evolve in
response to local differences in disturbance regime. From theory of evolution in
fluctuating environments (Cohen, 1966; Iwasa and Levin, 1995; Lytle, 2001),
we expect that selection for disturbance-escape strategies will be strongest when:
(i) per-event mortality is high; (ii) predictability of events is high; (iii) frequency
of events is high; (iv) events are ‘coarse-grained’ (synchronous over larger spatial
scales); and (v) the organism’s lifespan is long relative to the disturbance return
interval. Several more predictions arise from basic population genetic consider-
ations. Selection should be strongest when (vi) population-level additive genetic
variance is high, and (vii) gene flow from differently adapted populations is
small.

Mortality, Frequency and Predictability

Factors (i)–(iii) are fairly intuitive. If a lethal event occurs reliably on the same day
each year, selection will favour life histories that are synchronized to escape this
event. In most models of life-history evolution these three factors are multiplica-
tive, in that a low value of one will cancel out high values of the other two
(see Fig. 2 in Lytle and Poff, 2004). For example, selection may be weak for
high-mortality events that are seasonally predictable but very infrequent, such as
freezing events in tropical regions or hurricanes.
The notion of flood or drought ‘predictability’ deserves some discussion,
as this has been defined variously in the literature (Resh et al., 1988; Poff,
1992, 1996). For the purpose of understanding life-history or behavioural evo-
lution, it is useful to define predictability as the degree of correlation between
Life-history and Adaptation to Flow Regime 127

a disturbance and some environmental cue (sensu Lytle and Poff, 2004). For
many types of flood and drought, it is possible to determine the mean Julian
occurrence date of some flow event exceeding a given magnitude (timing)
using long-term hydrograph data. The degree to which this date correlates
with the actual date of disturbance across years (the variance) can thus be
interpreted as a measure of predictability, in that it describes how reliably that
date will predict an actual disturbance event from year to year. Proximately,
many aquatic insects use day-length to estimate seasonal timing, and the lit-
erature is rich with examples of taxa that adjust growth and development in
response to altered day-length (e.g. Johansson and Rowe, 1999; De Blok and
Stoks, 2004). In certain cases predictability can be defined without respect to
seasonal timing. Some organisms respond to rainfall events as a signal that a
flood may soon occur (or a drought may be ending). Desert stream insects
such as giant water bugs (Hemiptera: Belostomatidae), other Hemiptera and
some Coleoptera use rainfall as a signal to escape flash floods by immediately
abandoning the stream and moving into protected riparian areas (Lytle, 1999;
Lytle and White, 2007). For these organisms, the degree to which rainfall
events (the environmental cue) are correlated with floods (the disturbance)
determines the predictability of the flood event. An example of how this kind
of predictability can drive the evolution of behaviours in local populations is
discussed below.

Coarse-grained versus fine-grained environments

Factor (iv) arises from theory related to measuring fitness in fluctuating environ-
ments. In fine-grained environments where disturbances occur independently
across patches, and the breeding population is pooled from these patches, the
arithmetic mean is the appropriate measure of fitness over time (Cohen, 1966;
Gillespie, 1977). This is analogous to the ‘patch dynamics’ scale in stream ecol-
ogy studies (Pringle et al., 1988; Townsend, 1989; Downes et al., 1998; Lake,
2000). By contrast, in coarse-grained environments where the entire population
experiences the same disturbances in a synchronous manner, the geometric
mean is the correct fitness measure. In fluctuating environments, geometric mean
fitness is always lower than arithmetic mean fitness, especially when the fluctua-
tions are large. The way this affects trait evolution can be understood heuristi-
cally by imagining two different populations of a mayfly, one composed of
multiple subpopulations in adjacent streams that flood independently of each
other, and another occupying streams that always flood simultaneously. In the
former population, each subpopulation experiences the flood regime indepen-
dently, and then the adults meet in a single swarm to mate and disperse. The
population fitness is the arithmetic mean of all subpopulation fitnesses, and the
population as a whole is buffered to some degree against a severe disturbance in
any single stream (i.e. a zero value in the arithmetic mean will still result in a
positive population-wide fitness). In the latter population, a severe flood event
affects the entire breeding population simultaneously, producing a low or zero
value for the geometric mean fitness. Because of these differences, when per-event
128 D.A. Lytle

mortality rates are high, selection is much stronger in coarse-grained than fine-
grained environments (Iwasa and Levin, 1995; Lytle, 2001).
Two points arise from this discussion that are important to our understand-
ing of disturbance in aquatic systems. First, large floods and droughts likely fall
into the coarse-grained category as multiple stream reaches are affected simulta-
neously. Thus from an evolutionary point of view, disturbances that occur at the
level of individual substrate particles or reaches are fundamentally different from
disturbances that affect entire streams or rivers simultaneously – a point that may
affect our interpretation of experiments that focus on disturbance acting at the
level of individual rock particles. In other words, patch-dynamic disturbances
may not scale up to full-stream flood or drought disturbances, because selection
is acting in different ways at these two different scales. Second, the population
genetic structure of aquatic insects may have a profound effect on how popula-
tions adapt to disturbance regimes, because it may determine whether the envi-
ronment is coarse- or fine-grained. If a breeding population is drawn from a
single stream, or groups of streams that experience simultaneous disturbance
events, the coarse-grained fitness measures apply and selection may act more
strongly to produce local adaptation. Although the predictions from theory are
clear, the effect of environmental grain on aquatic insect evolution has yet to be
examined empirically. The genetic structure of aquatic insect populations is an active
area of research that should provide many insights concerning how organisms
evolve in response to disturbance.

Lifespan

The lifespan of an insect relative to disturbance recurrence interval (v) determines

to what degree selection can synchronize a life history with a disturbance regime.
At one extreme, univoltine or semivoltine insects must pass through one or several
disturbance seasons before emerging into the adult stage. For organisms in flood-
prone environments, to use one example, selection might be strong for adult
emergence that is synchronous with the expected date of flood disturbance,
assuming that floods are sufficiently frequent, severe and predictable. For these
taxa, the disturbance season entrains the population by restricting emergence to
a specific pre-flood season. Individuals that emerge late risk being removed by
floods, and individuals that emerge too early risk placing their offspring in harm’s
way. This life-history strategy has been observed in univoltine caddisflies inhab-
iting desert streams where severe flash floods occur during a fairly well-defined
season (Lytle, 2002; see example below). At the other end of the spectrum, multi-
voltine taxa with very short lifespans may or may not face a disturbance season
during their lifetime. This kind of multivoltinism has reached an extreme with the
desert stream mayfly Fallceon quilleri (Ephemeroptera: Baetidae), which matures
from egg to adult in an astounding 9–11 days (Gray, 1981). Although this short
life cycle prevents the population from becoming synchronized with the seasonal
flood regime, it is thought that a rapid life cycle may itself be an adaptation to
flooding as it minimizes the time spent as a larva in a flood-prone habitat (Gray
and Fisher, 1981).
Life-history and Adaptation to Flow Regime 129

Life-history strategies and environmental variability

The trade-offs faced by many aquatic insects inhabiting seasonally disturbed

environments are typified by the sycamore caddis Phylloicus mexicanus (Trichop-
tera: Calamoceratidae). Sycamore caddis inhabit perennial streams in the desert
south-western USA and northern Mexico, which are prone to flash floods during
the July–September flash flood season. The flash flood season places limits on P.
mexicanus emergence; individuals must balance the costs of remaining in the
stream (flood mortality) against the benefits (opportunity to continue feeding as
larvae and increase fecundity, as adults are non-feeding). Life-history models
predict an optimal reaction norm strategy, where individuals are expected to
emerge at a smaller body size (and thus lower fitness) as the risk of a flood
increases (Rowe and Ludwig, 1991; Johannson and Rowe, 1999; Lytle, 2001).
As with many caddisflies, a large proportion of body mass increase in P.
mexicanus occurs during the final instar, but the amount of growth varies
substantially among individuals. Among-individual variation in body size can be
due to genetic differences, environmental factors, or some combination of both.
Figure 7.1 shows that fitness, as measured by the number of eggs produced, can
vary substantially among females, in this case by a factor of 3 or more. These
individuals inhabited a range of food conditions that ranged from scarce to
abundant detritus [see Lytle (2002) for background and details]. The lower

1200

1000
Eggs per female

800

600

400

200

0.010 Fig. 7.1. Fitness

gradient in a population
0.009 of the caddisfly Phylloicus
Mass per egg (mg)

0.008 mexicanus. Fitness is

increased by producing
0.007 more eggs, rather than
0.006 higher quality eggs.
Top panel: larger females
0.005 produce more eggs
than smaller females
0.004
(R 2 = 0.78). Bottom panel:
0.003 egg mass does not vary
10 15 20 25 30 systematically with female
Female body mass (mg) body mass.
130 D.A. Lytle

panel of Fig. 7.1 shows that, for this species, there is no trade-off between egg
quality and quantity, and so egg number is a reasonable predictor of fitness.
This population-level variability in potential fitness represents a steep fitness
gradient, where realized body mass determines the position of an individual
along the gradient. The observed variability, which is typical of many aquatic
insects (Statzner et al., 1997), has several implications for life-history evolution
in flood-prone environments. First, there exists no single ‘one-size-fits-all’ optimal
body size at emergence; rather, optimal body size at emergence is a reaction
norm that changes through time as the costs and benefits of emergence change
(in this case, mostly due to the increasing risk of a flood). Indeed, life-history
models that incorporate a reaction norm of body size versus time perform sub-
stantially better than models that do not (Lytle, 2002). Similarly, the imposition
of a disturbance is expected to increase within-population variability in fitness
(Rowe and Ludwig, 1991; Lytle, 2001). To visualize this, imagine a population
that is free from the time constraint of a disturbance. Even though there might
be among-individual differences in growth rate due to genetics or microhabitat,
individuals are free to grow until a maximum body size is reached, thereby
reducing population-wide variability in egg number. Second, if the among-
individual fitness variability is due mainly to environment and not genetics,
individuals must be experiencing a large range of ecological conditions. Essen-
tially, there is much latitude in the environment for individuals to do really well
or really poorly. These differences could be due to microhabitat differences in
ecological conditions (detritus as a food resource, predators, conspecific densities)
or legacy effects (oviposition date, growth conditions during early development).
If among-individual fitness variability is due to genetics, then fluctuating envi-
ronmental conditions must maintain a large diversity of phenotypes within the
population. In all likelihood, the observed variability is due to some combination
of genetics and the environment.
In summary, the large within-population variability in aquatic insect egg pro-
duction has implications for the evolution of life-history strategies in disturbed
environments. A full understanding of these connections will require further stud-
ies, such as comparisons of populations of the same species inhabiting disturbed
versus non-disturbed streams.

Local Adaptation of Populations

Flood and drought disturbance regimes can vary significantly at multiple spatial
scales. Within a catchment, for example, floods may be less common in the
headwaters due to small catchment area but commonplace in lower reaches that
integrate many smaller drainages. Across catchments, flood and drought fre-
quency, severity and predictability can vary due to differences in catchment area,
underlying geology or even land use practices (Allan, 1995; Tague and Grant,
2004). Furthermore, anthropogenic factors such as dams, water diversions,
revetments and alteration of flood plains can produce sudden changes (with
respect to evolutionary timescales) in flood and drought dynamics (Poff et al.,
1997; Bunn and Arthington, 2002). All of these factors raise the issue of the
Life-history and Adaptation to Flow Regime 131

spatial scale at which aquatic insects can adapt to local differences in flood and
drought disturbance regime.

Phenotypic plasticity versus local adaptation

The ‘ideal’ aquatic insect should be able to alter its life-history strategy or behav-
ioural repertoire to match a range of disturbance regimes. For example, insects
that inhabit streams prone to drought in some years and not others may benefit
from a plastic strategy that allows different trajectories of growth and develop-
ment (Johansson and Rowe, 1999; Johansson et al., 2001; De Block and Stoks,
2004, 2005), assuming that sufficient proximate cues are present to signal the
approach of different environmental outcomes (drought versus no drought, for
example). For the aquatic insect populations that have been studied across dis-
turbance regime types, phenotypic plasticity does appear to be common. Studies
examining the proximate cues associated with the onset of drought (drying of
pools, changes in photoperiod that are correlated with drought onset) have gen-
erally found evidence for plasticity – that is, individuals adjust rates of growth,
development or both to compensate for different environments. In a common
garden study of an alpine caddisfly, Shama and Robinson (2006) found that
plasticity, rather than local adaptation, accounted for most observed differences
between populations obtained from perennial versus temporary streams. Simi-
larly, damselfly larvae from two habitats that differed in permanence showed no
evidence for local adaptation in a common garden experiment, but responded to
different treatments with phenotypic plasticity (De Block and Stoks, 2004).
Not all among-stream differences in disturbance regime are accompanied by
useful proximate cues, however. For example, catchment size may determine the
frequency and magnitude of flood events in some stream types (Gordon et al.,
1992), but there is no obvious way that an aquatic insect could determine catch-
ment size from proximate cues. Local adaptation may be expected under these
ecological conditions. Any population with sufficient heritable trait variation may
evolve in response to local conditions, so long as selection is strong enough to
counteract gene flow from nearby populations adapted to different optima.

Local adaptation to flood regime in giant water bugs

An example of local adaptation to different flood regimes is seen in populations

of the giant water bug Abedus herberti that inhabit isolated, perennial streams in
the arid south-western USA. Distinct populations of this flightless aquatic insect
are confined to catchments that experience a large variety of flood regimes,
which is dictated by catchment size. Floods are rare (<1 flood/year) or absent in
smaller catchments, while larger catchments can experience multiple floods per
year (Lytle, 2003). The predictability of floods from rainfall cues also varies
according to catchment size (Fogel, 1969; Osborn and Laursen, 1973): predict-
ability is highest in intermediate-sized catchments, where the correlation between
rainfall and floods is strongest, but decreases as watersheds become smaller
132 D.A. Lytle

(strong rainfall events occur, but catchment area is too small to capture enough
water for a flood) or larger (local rainfall is absorbed into alluvial fill, and floods
are generated by storms occurring in distant parts of the catchment). The cost of
encountering a flash flood is high because mortality rates of most insects caught
in floods is near 95% (Grimm and Fisher, 1989; Lytle, 2000). As noted previ-
ously, A. herberti use flood escape behaviour to survive these events: individuals
sense periods of heavy rainfall that often precede floods and use this as a cue to
crawl out of the stream to protected riparian areas, then return to the stream
post-flood (Lytle, 1999).
The problem faced by individual A. herberti is this: if it rains hard for, say, 30
min, does this mean that a flood will occur? How about 40 min? 50? It turns out
that this is a classic problem in signal detection theory, a branch of information
theory that is useful for understanding how animal behaviours evolve in response
to noisy signals (in this case, rainfall events of a given duration that do or do not
result in a flood) (Getty and Krebs, 1985; Reeve, 1989; Wiley, 1994; Getty,
2002). Essentially, each A. herberti individual must balance the cost of leaving
the stream (predators in the terrestrial environment, desiccation) against the cost
of remaining in the stream (mortality from a flood), in the context of flood pre-
dictability. Two types of mistakes can be made: an individual might abandon the
stream when no flood occurs (a ‘false alarm’), or an individual might choose to
remain when a flood does occur (an ‘incorrect rejection’). Rainfall cues provide
information about the environment – in general, the longer the rainfall event, the
more likely a flood. However, this information profile differs among catchments,
because rainfall is a reliable predictor of floods in mid-sized catchments
(≈10 km2), but less so in smaller or larger catchments. Signal detection theory
thus predicts that response time (the minutes of rainfall required to trigger RRB)
should be fastest in the predictable mid-sized catchments, assuming that costs for
incorrect rejections and false alarms are similar across catchments.
Behavioural experiments across 15 A. herberti populations inhabiting catch-
ments ranging in size from 4 to >400 km2 revealed local adaptation to different
local flood regimes: individuals from mid-sized catchments did, in fact, require
the shortest durations of rainfall to trigger flood escape behaviour (Lytle et al.,
2008). In an extreme case, individuals from the smallest catchment (4 km2),
where floods rarely if ever occur, ‘refused’ to leave the stream at all – for these
individuals, there is simply no benefit to leaving the stream, only a cost.
In a sense, the A. herberti populations encompass the ideal conditions for
local adaptation to different flood regimes. They are flightless and so local adap-
tation is not counteracted by gene flow from differently adapted populations, i.e.
factor (vi) in the section above is satisfied. Phylogeographic analysis of mito-
chondrial data suggests that at least some of these populations may have been
isolated since the Pleistocene (Finn et al., 2007), allowing plenty of time for
natural selection to adapt populations to local flood regime characteristics. Fur-
thermore, high haplotype diversity in many populations suggests that there may
be, or have recently been, much heritable trait variation on which selection can
act [factor (vii) in the section above]. Flash floods are also a very strong selective
pressure, as the cost of making a mistake approaches 100%. Thus, the Abedus
herberti system provides a baseline against which we can evaluate local adaptation
Life-history and Adaptation to Flow Regime 133

to flood or drought in other aquatic insect taxa. We now know that given sufficient
time, sufficient isolation of populations and sufficiently strong selective pressures, that
aquatic insect populations can adapt locally to a range of disturbance regimes.

Further Research and Management Implications

The examples given above show that aquatic insect populations can evolve in
response to local flow regimes, at least under conditions where isolated populations
are experiencing strong local selection. Next, we need to determine how common
local adaptation is: Do most aquatic insect populations show some degree of local
adaptation to flood or drought regime,