Diapause in Aquatic Invertebrates: Theory and Human Use (Monographiae Biologicae)

DIAPAUSE IN AQUATIC INVERTEBRATES THEORY AND HUMAN USE

MONOGRAPHIAE BIOLOGICAE VOLUME 84

Series Editor

H.J. Dumont Aims and Scope The Monographiae Biologicae provide a forum for top-level, roundedoff monographs dealing with the biogeography of continents or major parts of continents, and the ecology of well individualized ecosystems such as islands, island groups, mountains or mountain chains. Aquatic ecosystems may include marine environments such as coastal ecosystems (mangroves, coral reefs) but also pelagic, abyssal and benthic ecosystems, and freshwater environments such as major river basins, lakes, and groups of lakes. In-depth, state-of-the-art taxonomic treatments of major groups of animals (including protists), plants and fungi are also eligible for publication, as well as studies on the comparative ecology of major biomes. Volumes in the series may include singleauthor monographs, but also multi-author, edited volumes.

The titles published in this series are listed at the end of this volume.

Diapause in Aquatic Invertebrates Theory and Human Use VICTOR R. ALEKSEEV Zoological Institute of the Russian Academy of Science St. Petersburg, Russia

BART T. DE STASIO Department of Biology Lawrence University Appleton, WI, USA

and

JOHN J. GILBERT Department of Biological Sciences Dartmouth College Hanover, NH, USA

A C.I.P. Catalogue record for this book is available from the Library of Congress

ISBN-10 1-4020-5679-6 (HB) ISBN-13 978-1-4020-5679-6 (HB) ISBN-10 1-4020-5680-X (e-book) ISBN-13 978-1-4020-5680-2 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com

Printed on acid-free paper

Cover illustration: Daphnia pulicaria, photo by Victor R. Alekseev All Rights Reserved

© 2007 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.

We dedicate this book to Professor Alexander Danilevsky

TABLE OF CONTENTS

Preface

xiii

PART I: STRATEGIES AND MECHANISMS OF DIAPAUSE IN AQUATIC INVERTEBRATES 1. Introduction to Diapause Victor R. Alekseev, Oscar Ravera, and Bart T. De Stasio 1.1 Diagnosis of diapause 1.2 Ecological causes of diapause in aquatic organisms 1.3 Terminology on dormancy 2. Timing of Diapause in Monogonont Rotifers: Mechanisms and Strategies John J. Gilbert 2.1 Introduction 2.2 Female types and the fertilized resting egg 2.3 The timing of sex: environmental controls 2.3.1 Preface 2.3.2 Crowding 2.3.3 Dietary tocopherol 2.3.4 Photoperiod 2.3.5 General comments 2.4 The timing of sex: endogenous controls 2.4.1 Mixis delay 2.4.2 Mictic stem females 2.5 General mechanistic models for the control of mixis 2.6 Theoretical models for maximizing resting-egg production 2.7 Diapausing parthenogenetic eggs 2.7.1 Preface 2.7.2 The pseudosexual egg 2.7.3 The diapausing amictic egg of Synchaeta pectinata Acknowledgments

vii

3 3 5 8 11 11 12 16 16 17 18 19 20 21 21 23 23 24 25 25 25 26 27

viii

TABLE OF CONTENTS

3. Diapause in Crustaceans: Peculiarities of Induction Victor R. Alekseev 3.1 Introduction 3.2 Diapause in crustacean life cycles 3.2.1 Monocyclic species 3.2.2 Bicyclic and polycyclic species 3.2.3 Species with complicated life cycles 3.2.4 Species with life cycle without diapause 3.3 Presence of diapause among crustaceans 3.3.1 Embryonal diapause 3.3.2 Larval diapause 3.3.3 Adult diapause 3.4 Evolution of points of view on inducing factors 3.4.1 Embryonal diapause 3.4.2 Larval and adult diapause 3.5 Diapause as a photoperiodic response 3.5.1 Developmental stages in crustaceans responsible for perception of photoperiodic signal 3.6 Light as the source of information about the season 3.6.1 Peculiarities of crustaceans’ perception of photoperiodic signals 3.6.2 Role of photoperiod gradient in diapause induction 3.6.3 Geographical variability of photoperiodic reactions 3.7 Role of temperature and photoperiod in diapause induction 3.7.1 Embryonal diapause 3.7.2 Larval diapause 3.7.3 Adult diapause 3.8 Population density and manifestations of photoperiodic reactions 3.9 Food quality and diapause induction in the crustacea 3.10 Population polymorphism and inheritance of photoperiodic responses 3.10.1 Intrapopulation dimorphism for photoperiodic responses 3.10.2 Population polymorphism for photoperiodic responses 3.11 Heredity of photoperiodic responses Acknowledgments 4. Reactivation of Diapausing Crustaceans Victor R. Alekseev 4.1 Introduction 4.2 Patterns of reactivation processes for different types of diapause 4.2.1 Embryonal diapause

29 29 29 29 30 31 32 32 34 35 38 39 39 40 41 44 44 46 47 48 49 50 50 51 52 55 56 57 59 61 63 65 65 65 66

TABLE OF CONTENTS

4.3 4.4 4.5 4.6

4.2.2 Larval diapause 4.2.3 Adult diapause Endogenous phase of diapause Reactivation action of oxygen Participation of carbon dioxide in reactivation Hormonal basis of diapause Acknowledgments

5. Diapause in Aquatic Insects, with Emphasis on Mosquitoes Elena B. Vinogradova 5.1 Introduction 5.2 Mosquitoes (Culicidae) 5.2.1 Egg diapause 5.2.1.1 Diapause and quiescence 5.2.1.2 Hatching stimuli 5.2.1.3 Viability, drought, and cold hardiness 5.2.1.4 Photoperiodic and temperature induction of egg diapause 5.2.1.5 Diapause termination 5.2.2 Larval diapause 5.2.2.1 Syndrome of larval diapause 5.2.2.2 Photoperiodic and temperature induction and termination of larval diapause 5.2.3 Adult diapause 5.2.3.1 Syndrome of adult diapause 5.2.3.2 Photoperiod and temperature induction of adult diapause 5.2.3.3 Adult diapause termination 5.3 Other groups of aquatic insects 5.3.1 Chironomids (Chironomidae) 5.3.2 Biting midges (Ceratopogonidae) 5.3.3 Dragonflies (Odonata) 5.3.4 Heteroptera 5.3.5 Ephemeroptera Acknowledgments 6. A Brief Perspective on Molecular Mechanisms of Diapause in Aquatic Invertebrates Victor R. Alekseev 6.1 Introduction 6.2 Molecular mechanism of diapause in the nematode Caenorhabditis elegans Acknowledgments

ix 67 69 70 75 76 77 82 83 83 85 85 85 86 87 88 94 96 96 97 103 103 105 109 110 110 111 111 112 113 113 115 115 116 118

x

TABLE OF CONTENTS

PART II: THE ROLE OF DIAPAUSE IN SCIENCE AND HUMAN USES 7. Egg Bank Formation by Aquatic Invertebrates: A Bridge Across Disciplinary Boundaries Bart T. De Stasio 7.1 Introduction 7.2 Dormancy processes 7.2.1 Dormancy initiation 7.2.2 Release from dormancy 7.2.2.1 Additional emergence data 7.2.3 Predation and infection of dormant stages 7.2.4 Deep burial of dormant stages 7.2.5 Senescence and egg viability 7.3 Egg bank size and dynamics 7.4 Creating an egg bank 7.5 Conclusions Acknowledgments 8. Use of Cladoceran Resting Eggs to Trace Climate-driven and Anthropogenic Changes in Aquatic Ecosystems Susanne L. Amsinck, Erik Jeppesen, and Dirk Verschuren 8.1 Introduction 8.2 Tracing acidification 8.3 Tracing eutrophication 8.4 Tracing fish introductions and biomanipulation 8.5 Tracing heavy metal pollution 8.6 Tracing climate change 8.7 Discussion and conclusion: limitations, concerns and future potentials Acknowledgments 9. Reconstructing Microevolutionary Dynamics from Layered Egg Banks Luc De Meester, Joachim Mergeay, Helen Michels, and Ellen Decaestecker 9.1 Introduction: dormant stages and the study of microevolution 9.2 A short survey of recent success stories 9.3 Pitfalls 9.4 Conclusions and future directions Acknowledgments 10. Does Timing of Emergence within a Season Affect the Evolution of Post-diapause Traits? Post-diapause and Directly Developing Phenotypes of Daphnia Kestutis Arbacˇauskas 10.1 Introduction 10.2 Daphnia life cycle

121 121 121 121 123 123 129 129 130 131 131 133 133 135 135 137 141 146 149 150 153 157 159

159 160 162 164 166

167 167 168

TABLE OF CONTENTS

10.3 10.4 10.5 10.6 10.7

Neonates: biochemical quality and body size Physiology: respiration and starvation resistance Life-history: growth, allocation, and relative fitness Descendants of post-diapause and directly developing females Conclusions Acknowledgments

11. Diapause and its Consequences in the Daphnia galeata – cucullata – hyalina Species Complex Piet Spaak and Barbara Keller 11.1 Introduction 11.2 Hybridization in Daphnia 11.3 Genetic markers to identify parental and hybrid taxa within the D. galeata – cucullata – hyalina complex 11.4 Factors that determine sexual reproduction of parental Daphnia species 11.5 Are hybrids still produced? 11.5.1 Are hybrid diapausing eggs present in the sediment? 11.5.2 Do males and sexual females of hybridizing species temporally and spatially co-occur? 11.6 Taxon distribution of asexual and sexual daphnids as well as from their offspring 11.7 Can the sediment tell us something about past hybridization events? 11.8 Conclusions Acknowledgments 12. Role of Diapause in Dispersal of Aquatic Invertebrates Vadim E. Panov and Carla Caceres 12.1 Introduction 12.2 Mechanisms and vectors of dispersal of diapausing invertebrates 12.2.1 Natural vectors of dispersal 12.2.2 Human-mediated dispersal 12.3 Conclusions: generalized model of dispersal of aquatic invertebrates with prolonged diapause Acknowledgments 13. The Role of within Trophic Level Chemical Interactions in Diapause Induction: Basic and Applied Aspects Egor S. Zadereev 13.1 Introduction 13.2 The effect of chemical interactions on diapause induction at the individual level 13.3 The effect of chemical interactions on diapause induction at population and ecosystem levels

xi 168 169 170 173 173 173 177 177 177 178 180 181 182 182 183 184 185 185 187 187 188 188 189 193 195 197 197 198 202

xii

TABLE OF CONTENTS

13.4 Biotechnological applications 13.5 Conclusions Acknowledgments 14. Studying the Phenomenon of Dormancy: Why it is Important for Space Exploration Victor R. Alekseev, Vladimir N. Sychev, and Natalia D. Novikova 14.1 Introduction 14.2 Study of dormancy from the perspective of its integration into ecological life support systems 14.3 Planetary and interplanetary quarantine 14.4 Microbiological safety of space flight 14.5 Diapause and adaptation of higher vertebrates to extended body metabolism 14.6 Search for extraterrestrial life 14.7 The first results and perspectives 14.7.1 Effect of space flight conditions on survivorship and life cycle parameters in resting stages of some crustaceans 14.7.2 Dormancy-based resistance of bacteria and fungi to extreme space environments 14.8 Conclusions Acknowledgments

204 205 206 207 207 208 208 209 209 210 210

212 213 214 214

References

215

Index

255

PREFACE

Dormancy is a suspension of the vital functions in an organism for a certain, sometimes very long, time period to overcome harsh environmental conditions. It is a widespread adaptation in many phyla of animals and plants, from bacteria to vertebrates (Weismann 1876; Winberg 1936; Rogick 1938; Werner 1955; Steele 1965; Holmes 1966; Pourriot & Clement 1973; Luning 1980; Alekseev 1990; Hirche 1996; Hairston et al. 2001). The best-defined types of dormancy according to the cause of the arrest in development are diapause and quiescence. Quiescence is driven directly by the dynamics of environmental factors, whereas diapause is determined by a predictive mechanism combining environmental signals and an internal biological clock. The adaptive function of diapause applies both to biorhythms and to defensive aspects. Biorhythms determine synchronization of the life cycle with environmental seasonal rhythms. This function is mainly based on reactions involving photoperiod, food dynamics, and temperatures as signal factors. Such reactions are discussed below. Defensive traits allow an organism to endure the actions of the complex set of suppressive or even lethal factors occurring during an unfavorable period. The defensive function of diapause, which provides unspecific resistance of an organism to a wide complex of unfavorable actions, is based on reduced metabolic rate (in the case of anabiosis such reduction is close to zero), which also appears as a response to signal factors. In addition, a variety of protective structures like various resting forms (e.g. gemmules, resistant eggs, cysts, cocoons, statoblasts, and ephippia) are of a size and have morphological and physiological characteristics to maintain a good viability level until the end of the harsh environmental conditions. Some diapausing eggs can even maintain their viability after passing through the digestive system of a predator, whereas the resting eggs of other species are digested (Hairston & Olds 1984; Marcus 1984a,b; Hairston & Cáceres 1996). This difference is due to the material and permeability of the protective covering of the resting animals. Diapausing eggs are different from parthenogenetic eggs because they are usually enclosed within a resistant protective capsule, and because they contain stored nutritive substances that are abundant in relation to their depressed metabolism and the duration of the diapause (Zaffagnini 1987). Consequently, the ex-ephippial generation may have a different life history from that of the parthenogenetic generation. Arbacˇauskas and Gasiu¯naitè (1996) and Arbacˇauskas (1998, 2001) carried out detailed studies on this subject. The most important results were as follows. The ex-ephippial individuals of daphnidae grow faster, mature earlier, and have greater fecundity than the individuals from parthenogenetic females. In addition, the intrinsic rate of natural increase of the ex-ephippial generation is higher than that resulting from parthenogenetic eggs in environments with invertebrate predation or moderate xiii

xiv

PREFACE

fish predation pressures. The daphnids hatching from diapausing eggs are adapted to optimal environmental conditions, while those from parthenogenetic eggs are better adapted to an unpredictable environment. Both generations have developed strategies to maximize long-term fitness (see Chapter 10). In larval and adult diapauses, along with a reduction of the rate of oxygen consumption, other functions (nutrition, mobility, and reproductive activity) are also suppressed (Alekseev 1998; Chapter 3, this volume). Offensive functions of these diapauses are realized with the help of behavioral reactions that include vertical and horizontal migrations as well as searching for shelters. The ratio among behavioral, constitutional, and metabolic elements is usually species- or even population-specific, but there are many common features in the range of each type of diapause. The combined effects of genetic characteristics, the result of selective mechanisms, and environmental and physiological constraints determine the duration of diapause. The relative importance of each of these factors varies according to the situation. For example, Gilbert (1998) demonstrated that heritable characteristics are not essential when the response to environmental change is phenotypical (Chapter 2). There is a need for more research on the physiological and ecological mortality of animals during diapause. For example, the diapausing eggs of some species of zooplankton are viable for several decades under optimal environmental conditions, whereas under adverse conditions their life span becomes very short. Optimal conditions for the physiological functioning of a species can be identified by laboratory experiments; physiological mortality can then be estimated. Conversely, field studies can produce data on ecological mortality. It follows that it is a risky undertaking to extend the conclusions drawn from laboratory experiments to the natural field without due caution; this is true for all demographic variables. The most important effect of diapause is that during the period of adverse environmental conditions it preserves an adequate number of viable individuals from a species population to assure its permanence in the community. To achieve this, the duration of the diapause cannot be shorter than the period of harsh conditions. Neither can the duration of diapause be too long, because the longer the diapausing forms remain dormant, the longer they are exposed to various causes of mortality such as predation pressure, and bacterial and fungal infections (De Stasio 1990; Marcus et al. 1994; Decaestecker et al. 2004). Diapause may exert a great influence at the population and community levels. Unfortunately, few studies on aquatic populations and communities take diapause and its consequences into account because it is so difficult to quantify the order of magnitude of the individuals emerging from diapause (Marcus 1984a, b; Hairston & De Stasio 1988; De Stasio 1989; Wolf & Carvalho 1989). This difficulty increases when individuals of the same species go into diapause, emerging at different times and producing cohort overlapping. Hairston and Cáceres (1996) studied in two shallow lakes the ways in which zooplankton resting eggs may be related to the benthic community. Resting eggs (or other resting stages such as larvae and adult) accumulated in the surface sediments represent a “bank” of zooplankton species that assures their persistence in a community, in spite of periodic harsh conditions (Herzig 1985; Carvalho & Wolf 1989; De Stasio 1990; Chapter 7, this volume).

PREFACE

xv

This strategy is of fundamental importance in aquatic environments such as temporary ponds that become terrestrial ecosystems during the hot, dry season. The community of these environments is exclusively composed of species selected to overcome the dry season by making resting forms. Cousyn and De Meester (1998) estimated the size and distribution of the egg bank of Daphnia magna populations from five fish ponds, calculating the density of resting eggs in the first 20 cm of sediment, and observed that ~65% of the resting eggs present in the first 10–12 cm of the sediment core were viable; in the deeper sediments, hatching success decreased with depth (Chapter 9). A number of studies demonstrate that diapause provides the advantage of promoting the colonization of new environments facilitating the passive transport of the resting stages, e.g. the ephippia (Alekseev & Starobogatov 1996; Hairston & Cáceres 1996; Cáceres 1997). Enlarging the distribution area of the species, colonization of new environments is regarded as a safeguard against its extinction. Furthermore, because diapausing animals are selected to overcome adverse periods, they have the advantage of retaining this characteristic in new environments as well, which gives them some measure of protection from invasion antagonistic species (competitors and predators) that have no resting stages (Chapter 12). Studies on the vertical distribution of resting eggs in sediment cores yield important information, for example, on the variations over time of the environment and of the populations producing resting eggs, and the genetic differences between recent and past populations of the same species (Weider et al. 1997). Copepod resting eggs are particularly useful for reconstructing the history of their environment, since no information can be obtained from the exoskeleton, which is easily decomposed in contact with the sediments. Conversely, the variations over time of cladocera populations reflect ephippia as well as carapace and postabdomen remains, which are highly resistant to decomposition (see Chapter 8). Cultivation of live food like rotifers, Daphnia, Artemia, or marine copepods is an expanding application of practical use of diapause in modern aquaculture (Marcus & Murray 2001). Better understanding of the timing mechanism of dormancy in these species and many others promises to increase efficiency of these biotechnologies (see Chapter 13). Sexual reproduction in cyclical parthenogenetic Daphnia might lead to the production of interspecific hybrids. Since sexual reproduction in Daphnia is coupled with diapause, the study of diapause and diapausing eggs can clarify questions related to the frequency of hybrid production, the occurrence and strength of mating barriers, the abundance of hybrids in egg banks, and the likelihood that hybrids will colonize new habitats through diapausing egg dispersal with subsequent hatching (Chapter 11). Creation of artificial ecosystems outside the earth’s biosphere using plants and animals in dormancy becomes an actual possibility with the plans of humans to colonize our nearest planets (Alekseev & Sychev 2006). Resting stages provide at least two properties for such ecosystems. They can be easily transported in space for a long time without special care as compared with an ecosystem in an active state. In addition, storage of seeds and diapausing animals will provide a reserve in case of an unpredictable destruction of the active part of an ecosystem, e.g. a meteorite attack. Another aspect of this problem concerning space biologists is to avoid unpredictable contamination of other planets where life is possible with earth’s organisms

xvi

PREFACE

transported on space-research apparatus and human-led expeditions. Very strong stability of resting stages of primitive organisms, in particular, produces a real danger for such interplanetary penetration of alien species from earth. This ability of resting stages of many organisms to survive in open space is under practical testing now (Chapter 14). Finally, one of the central problems of biology, the origins of life on our planet, to some extent can be connected with studies of the dormancy phenomenon. The presence of different kinds of dormancy, including diapause, in all groups of living organisms studied so far, allows speculation about cryptobiosis as a common characteristic of life itself. Together with the extremely long period of viability found in some dormant organisms frozen in ice of the Antarctic Shield reaching many thousands of years, this creates a possible link to explain how life could have once disseminated on our planet (Rothschild & Mancinelli 2001; Alekseev & Sychev 2006). Early studies demonstrated the importance of the research on diapause, and the development of quantitative approaches for estimating seasonal rhythms has allowed the formation of new fields in areas such as botany, entomology, ornithology, and mammalogy (Farner 1964; Marcovich 1923; Rowan 1926; Shull 1928; Chapter 5, this volume). Remarkable progress in theory and practice was achieved in these fields. However, the data on formation, progression, and significance of diapause in the life cycle of aquatic invertebrates were obtained only from research on a narrow range of systematic groups (Stross 1965; Strempel 1976; Grice & Marcus 1981). Only recently the results in this field of study were generalized in a preliminary fashion for crustaceans, as part of several special conferences focusing on diapause in the crustaceans and then at a workshop on diapause in aquatic invertebrates in Pallanza, Italy (Alekseev et al. 2004). Many authors of this book were participants at the workshop in Pallanza, during which they had the opportunity to discuss and share their current findings on diapause in aquatic invertebrates. The book consists of two major parts indicated in its title: phenomenology of diapause and significance of this profound and widespread adaptation in scientific and practical uses. This division is highly conditional as authors of the chapters concentrate on both theoretical and applied aspects of diapause nearly in equal proportion. Last but not least, we should express our respect for the previous generations of diapause researchers, whose results and knowledge gained through numerous experimental and field studies became a scientific basis of this book. Victor R. Alekseev, Bart T. De Stasio, John J. Gilbert, and Oscar Ravera

LIST OF CONTRIBUTORS

VICTOR R. ALEKSEEV Department of Taxonomy and Systematics Zoological Institute of the Russian Academy of Science University emb., 1 199034 St. Petersburg Russian Federation [email protected] SUSANNE L. AMSINCK Department of Freshwater Ecology National Environmental Institute Vejlsoevej 26 8600 Silkeborg Denmark [email protected] KE˛STUTIS ARBACˇ IAUSKAS Institute of Ecology Vilnius University, Vilnius Lithuania [email protected] CARLA CÁCERES Department of Animal Biology University of Illinois at Urbana-Champaign 477 Morrill Hall 505 South Goodwin Ave. Urbana, Illinois 61801 USA [email protected]

ELLEN DECAESTECKER Laboratory of Aquatic Ecology Katholieke Universiteit Leuven Ch. De Bériotstraat 32 3000 Leuven Belgium [email protected] JOHN J. GILBERT Department of Biological Sciences Dartmouth College, Hanover New Hampshire 03755 USA [email protected] ERIK JEPPESEN Department of Freshwater Ecology National Environmental Institute Vejlsoevej 26 8600 Silkeborg Denmark [email protected] BARBARA KELLER Eawag, Swiss Federal Institute of Aquatic Science and Technology 8600 Dübendorf, Switzerland, and Institute of Integrative Biology ETH Zurich 8092 Zurich Switzerland [email protected]

xvii

xviii

LIST OF CONTRIBUTORS

LUC DE MEESTER Laboratory of Aquatic Ecology Katholieke Universiteit Leuven Ch. De Bériotstraat 32 3000 Leuven Belgium [email protected] JOACHIM MERGEAY Laboratory of Aquatic Ecology Katholieke Universiteit Leuven Ch. De Bériotstraat 32 3000 Leuven Belgium [email protected] HELEN MICHELS Laboratory of Aquatic Ecology Katholieke Universiteit Leuven Ch. De Bériotstraat 32 3000 Leuven Belgium [email protected] NATALIA D. NOVIKOVA Laboratory of Microbiology of the Environment & Antimicrobial Protection The State Scientific Center of the Russian Federation Institute for Biomedical Problems of the Russian Academy of Science Khoroshovskoe shosse, 76a 123007 Moscow Russian Federation [email protected] VADIM PANOV Zoological Institute of the Russian Academy of Science University emb., 1 199034 St. Petersburg Russian Federation [email protected]

OSCAR RAVERA CNR Institute of Ecosystem Study Largo Tonolli 50 28922 Verbania Pallanza Italy [email protected] PIET SPAAK Eawag, Swiss Federal Institute of Aquatic Science and Technology 8600 Dübendorf Switzerland [email protected] BART T. DE STASIO Department of Biology Lawrence University Appleton, WI 54912 USA [email protected] VLADIMIR N. SYCHEV Laboratory of Biological Life Support Systems The State Scientific Center of the Russian Federation Institute for Biomedical Problems of the Russian Academy of Science Khoroshovskoe shosse, 76a 123007 Moscow Russian Federation [email protected] DIRK VERSCHUREN Department of Biology, Research Group Limnology Ghent University Ledeganckstraat 35 B-9000 Ghent Belgium [email protected]

LIST OF CONTRIBUTORS

ELENA B. VINOGRADOVA Laboratory of Experimental Entomology and Biocontrol Zoological Institute of the Russian Academy of Science University emb., 1 199034 St. Petersburg Russian Federation [email protected]

EGOR S. ZADEREEV Institute of Biophysics SB RAS Akademgorodok 660036 Krasnoyarsk Russian Federation [email protected]

xix

This page intentionally blank

PART I

STRATEGIES AND MECHANISMS OF DIAPAUSE IN AQUATIC INVERTEBRATES

VICTOR R. ALEKSEEV, OSCAR RAVERA, AND BART T. DE STASIO

1. INTRODUCTION TO DIAPAUSE

1.1 DIAGNOSIS OF DIAPAUSE

At high and moderate latitudes, and sometimes in the tropics, organisms from fresh and brackish waters inhabit environments that are characterized by heterogeneity. This heterogeneity has different degrees of manifestation and is mainly determined by the annual cycle of solar radiation. Organisms adapt through adjustments of their life cycle to periodical fluctuations of biotic and abiotic factors. This is seen in shifts in the tuning of periods of reproduction and population growth to specific periods of the year, and in a delay of these processes during unfavorable periods. The life cycle of an organism consists of periods of active reproduction and population growth, alternating with periods during which these processes are delayed or essentially stop. This state of physiological rest has traditionally been called dormancy (Keilin 1959). But such a broad definition includes several closely related depressive states such as quiescence, diapause, and even sleep. To further our understanding of this phenomenon it is helpful to restrict usage of the term diapause to describe an adaptation to seasonal heterogeneity of environments. The first part of this adaptation is connected with the duration of the diapause. It is known that some physiological and biochemical processes (e.g. mobility, nutrition, synthesis of proteins) have short-term depressions, often commensurate with day length. Sleep in mammals is a well-known example of this sort of feature. Diapause, in contrast, is an adaptation that occupies a period approaching or exceeding the duration of the active part of the organism’s lifetime. It may last from weeks to years, as in the case for populations living in temporary basins in arid zones, to decades and hundreds of years in other environments such as lake sediments (Danilevsky 1961; Hairston et al. 1999a). The second part of this life history feature concerns the intensity of “depth” of diapause, which is caused by changes in underlying systems, involving neurohumoral changes, intracellular transformations, and gene expression (Crag & Denlinger 2000; Gerisch & Antebi 2004; Zhang et al. 1992). The secretion of hormones and accumulation of energy reserves requires time to accomplish. Thus, it is evident that diapause does not occur as an immediate response to a simple worsening of conditions, but rather precedes such worsening. Also, the rapid improvement of environmental conditions cannot initiate the termination of diapause. That is the main difference between diapause and quiescence, which begins or ends practically immediately after a change in the depressive factor, e.g. temperature. The necessity of clearly distinguishing diapause from sleep and quiescence should not exclude an understanding of the relationship and genetic similarity of these 3 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 3–10. © 2007 Springer.

4

V. ALEKSEEV ET AL.

phenomena. Adaptations to daily heterogeneity of conditions evidently precede adaptations to seasonal heterogeneity. This is reflected in the mechanisms of induction and termination of diapause, which are based on circadian rhythms (Tyshenko 1977). In addition, quiescence appears to be an undeveloped form of seasonal adaptation in which induction and reactivation are realized without the participation of neurohumoral systems (Zaslavsky 1988). This similarity between diapause and quiescence is supported by the ability to reverse diapause processes during the initial period of its development, and to the interruption of diapause under the influence of temperature or other factors. The reversibility of diapause, particularly during its initial stages, partly connects it with quiescence. At the same time, diapause is clearly a unique form of adaptation to seasonal heterogeneity of the environment, and as such can undoubtedly be observed and studied as an independent phenomenon. There are a number of hypotheses on the origin and evolution of diapause, ranging from those that suggest a phylogenetic explanation (e.g. Corbet 1980; Hairston & Cáceres 1996) to those that see environmental stresses as the primary driving force (Fryer 1996). According to Alekseev and Starobogatov (1996), diapause originated only once in the animal kingdom, thereafter evolving into various forms (i.e. monophyletic origin); in contrast, other authors favor the hypothesis of a polyphyletic origin based on studies of various arthropod species (Danilevsky 1961; Tauber et al. 1986). There are at least two lines of evidence supporting the dependent, evolutionarily transmitted origin of dormancy and diapause as the main manifestation of photoperiodic response. The first line includes the historical stability of temporal and spatial heterogeneities of external conditions such as temperature and light, which are caused by the earth revolving around the sun. The first inhabitants of water and land should have benefited from adaptation to periodic changes of these very essential factors. The most successful sets of traits were fixed genetically and supported by stabilizing natural selection. These adaptations might have served as the genetic basis for development and specialization in different phylogenetic groups. Differences among the forms of diapause in their ecological and physiological significance were likely obtained as the result of such evolutionary development and specialization. The second line of evidence, necessary for a correct evaluation of the first one, is based on the assumption that many links of organization of the system are similar in different phylogenetic groups of animals, especially those that regulate the seasonal cycles and can be labeled as neurohormonal mechanisms (Carlisle 1957; Otsu 1963; Novak 1966; Quackenbush 1986; Zaslavsky 1988). The ubiquity and economy of hormonal material in diapause pathways is reminiscent of that found in hemoglobin. Studies clearly indicate that a common phylogenetic origin for certain traits does not automatically mean commonness of separate properties, especially those related to adaptations of organisms from different phylogenetic groups to concrete environmental conditions. Research on similarities and differences of properties and characteristics in organisms of different systematic association is one of the central problems in

INTRODUCTION

5

biology. It is clear that these problems, which are related to the origin, spread, and formation of diapause, should be solved separately for different taxonomic groups before conclusions about their similarity or distinctiveness are reached. We begin with an account of the causes of the appearance of diapause in the life cycle of crustaceans and other aquatic organisms. An examination of such causes is necessary for understanding the adaptive function of diapause as a resting phase in the life cycle of invertebrates. 1.2 ECOLOGICAL CAUSES OF DIAPAUSE IN AQUATIC ORGANISMS

The factors that determine the appearance and preservation of diapause in the life cycle of an organism must respond to at least two demands: they should be both important for life and occur in a periodic fashion. Factors that do not correspond to both conditions, e.g. occasional exposure to toxic materials (important for survivorship, but not periodic in nature) or seasonal rains in spring (periodic but not important for survival), cannot drive an adaptation such as diapause. There is no benefit to evolving diapause in response to such factors, and natural selection will not favor organisms with this life history feature under these conditions (Timofeev-Resovskij et al. 1977). The seasonal rhythms of meteorological conditions, observed at most latitudes on the earth, lead to cyclical pulses of the main factors important for life in continental and oceanic waters. The warm season of the year is the most favorable for the overwhelming majority of aquatic organisms, and at high latitudes it is often the only period when organisms are active. Temperature optima and tolerance limits are generally similar and rather narrow for different groups of crustaceans developing during summer (Ivleva 1981). Appearance of favorable temperatures is determined by one of the most stable processes, the movement of the earth around the sun. So seasonal fluctuations of temperature should be regarded as one of the fundamental factors creating the need for diapause among aquatic invertebrates. In addition, this factor is important for organisms living in cold environments, where development of favorable conditions follows alternative rhythms. Trophic conditions are the other periodic factor essential for aquatic animals. The periodic food supply for filter feeders and their predators at latitudes with temperate and arctic climates is a result of the greater amount of solar energy in summer and the action of correlated processes such as the income of bio-organic material from deeper layers due to mixing by wind or thermal dynamics. The significance of trophic factors in the induction of diapause has been proven experimentally many times (Woltereck 1911a; Stuart & Banta 1931; Makrushin 1968; Stross 1969a). Declining oxygen concentrations in the deepest part of highly eutrophic basins in winter and summer may be regarded as one of the causes of the interruption of active development by diapause in many crustaceans, especially those that live near the bottom (Marcus 1996; Alekseev et al., 1999). The pressure of predators, especially juvenile fish, is also a seasonal and periodic factor. Some authors consider it as the main cause of the appearance of diapause (Nilssen 1978; Gliwicz & Rowan 1984). In invertebrates, two main strategies involving

6

V. ALEKSEEV ET AL.

diapause have been shown to be important for avoiding fish predation. One is migration into oxygen-depleted zones to find a refuge, as was shown for cyclopoid copepods (Alekseev 1990). Because of diapause the cyclopoids reduce basic metabolism and are able to remain in the refuge for months until juvenile fish leave the plankton for near-shore areas. Another strategy is to form resting eggs that are stable enough to survive passage through fish digestive systems and can still be reactivated later, perhaps even in the beginning of the next season, so that at least part of the population will survive. This adaptation has been found in clones of large-sized Daphnia magna living in shallow ponds (Slusarczyk 1995; Pijanowska & Stolpe 1996). An important aspect of these strategies is that the decrease in the intensity of metabolism, which occurs during the period of diapause, decreases the susceptibility of crustaceans to predators because their reaction and speed are gradually reduced along with their metabolism. It is not accidental that the proportion of copepods in the diet of benthivorous fishes increases in autumn when the majority of warm-season copepods begin diapause. As a result, diapause in some water basins may lead to a decrease of crustacean survival and hence to the extinction of the species from lake ecosystems (Alekseev 1990). According to Callaghan (1998), emergence from diapause should occur when the probability of mortality while dormant exceeds the probability after emergence. This means that the diapausing form must receive one or more hatching cues informing it that the environmental conditions (temperature, light, oxygen concentration) are acceptable. Although in some cases the nature of the cue cannot be identified, there is general agreement on the importance of the cue; according to Gilbert (1974), diapausing eggs that sink to deep sediments are lost because they do not receive any hatching cues such as temperature or light. A cue that leads some species to break diapause may not be effective for others. For example, the photoperiodism, which influences a number of Daphnia spp. (Stross 1965) and some rotifer species (Pourriot & Snell 1983), is not a cue for the genus Bythotrephes, while temperature is a positive cue for the diapausing eggs of the species B. longimanus from Lake Mondsee, Austria (Herzig 1985) and B. cederstroemi from Lake Michigan, USA (Yurista 1997). In the Laurentian Great Lakes, USA, a thermal bar develops at the end of winter, separating warmer shallow waters from colder pelagic areas. As the lake continues to warm, the bar moves outward toward deeper waters. The mixed water column offshore continues to warm and the thermal bar disappears. Yurista (1997) considers the thermal bar as a cue for the resting eggs of Bythotrephes to break the diapause. Hatching starts in the inshore sediments, which are the first to be warmed, after which, as the thermal bar migrates toward deeper waters, the eggs located at greater depths hatch. The result is a progressive increase in size of the Bythotrephes population as it receives individuals from ever-increasing depths. There are different consequences caused by the developmental delay in species with subitaneous (immediately hatching) eggs and in those with diapausing eggs. The life span of an animal species can be divided into three physiological periods: prereproductive, reproductive, and post-reproductive. The post-reproductive period in invertebrates is generally short or absent. The duration of each period and the ratio

INTRODUCTION

7

between them vary with the taxon. For example, the pre-reproductive period in Lepidoptera is long compared with its reproductive period; conversely, in the bivalved Unionidae, the reproductive period is much longer than the pre-reproductive period. Differences in duration of life stages are especially evident where temperature greatly affects development time. In contrast to homeothermic species, in which development time is independent of environmental conditions (e.g. temperature), poikilothermic species have a development time that is strongly influenced by environmental factors such as temperature and food supply. Consequently, the duration of each of the three physiological periods varies in relation to the season in which the animals were born. The longer duration of one or more stages due to low temperature results in an increased risk of mortality and a decrease in population production. In contrast, the longer duration of a diapausing stage is the only strategy for overcoming an adverse period of time and for permitting sensitive species to persist in a seasonally adverse environment. A clear example is provided by the cyclical parthenogenetic reproduction of cladocerans. B. cederstroemi is a predatory species that has invaded the Laurentian Great Lakes. During autumn it produces gametogenic diapausing eggs, which overwinter and hatch in spring. Because Bythotrephes is present in the lake during winter only in the egg diapausing stage, its population development each year is the combined result of the hatching of diapausing eggs and the subsequent parthenogenetic reproduction by the emergent animals during the period from spring to autumn (Yurista 1997). A lack of appreciation of the importance of diapause has been evident in ecological studies. For example, studies on the dynamics of zooplankton species are commonly based on the variations in abundance of the active individuals in the water column. Anomalous variations are related to the influence of external (e.g. hydrological load, temperature) and/or internal (e.g. competition, predation, primary production) variables; the influence of diapause is rarely mentioned. The drying of temporary waters is a special factor that selects organisms for their ability to form diapausing stages. Organisms that inhabit these widespread basins are exposed to the periodic action of a complex suite of unfavorable factors such as high levels of UV radiation, high temperatures, acute fluctuations of salinity, pH, and oxygen, and finally even the disappearance of the aquatic environment. Only those species that are able to modify their development and produce resting stages can survive and wait for new water from rains or floods. Dormancy in the different orders of crustaceans inhabiting temporary waters was repeatedly demonstrated by experimental analyses (Champeau 1970, 1979; Alekseev 1980, 1984b; Monchenko 2003). Although dormancy seems to be common in temperate and polar areas as a strategy for surviving low temperatures, this does not prove that it originated in these regions. Some forms of seasonality are peculiar to tropical regions such as monsoon rainfall (Williams-Howze 1997). Consequently, some authors have posited a tropical origin of the phenomenon, since dormancy has been observed in some tropical insects despite stable environmental conditions (Denlinger 1974). Similarly, interesting data concerning oceanic calanoid copepods were obtained from tropical zones of upwelling (Thirot 1978; Owen 1981; Herman et al. 1981). Here, the periodic lack of biogenic imports from deeper layers is similar to the situation in basins that dry

8

V. ALEKSEEV ET AL.

seasonally. Because the gradient of trophic conditions near algal blooms and regions outside of the upwelling zone is huge, the crustaceans, which occur outside of the bloom or experience the end of a bloom, feel as though their environment had “disappeared.” Some calanoid copepods, e.g. Calanus hyperboreus, in such situations demonstrate the ability to form diapausing stages. These stages sink to a depth of 400 m where low oxygen concentrations occur, and stay at this depth until they are lifted to the surface by upwelling. When this happens they start a new generation in a temporary flowering “oasis” of food. It was postulated that the ability of invertebrates to form diapausing stages decreases from high latitudes to the equator due to the reduction of seasonality in hydro-meteorological factors (Danilevsky 1961). However, more recent investigations indicate that this idea should be revised. It is now evident that diapause appears wherever annual or seasonal rhythmic fluctuations of factors important for life take place. 1.3 TERMINOLOGY ON DORMANCY

As mentioned earlier, strategies for survival in fluctuating environments usually include life cycle adaptations, like evolution of a sequence of active and resting periods in major vital activities such as growth and breeding. Resting phases of the life cycle are quite similar in a wide range of organisms. In spite of extensive differences in the terminology of resting stages, all of them can be organized into a few groups. Keilin (1959), in a major review of phenomena of latent life in invertebrates, suggested a classification in which he used the metabolic rate of an animal as the basic criterion. A problem with his classification arose because he (and many others afterwards) recognized a complete gradation in the same organism, from the active metabolic state through various degrees of dormancy to cryptobiosis or anabiosis. Laudien (1973) proposed an alternative system. In his scheme, all resting stages were grouped under the category of dormancy and the cause of the arrest in development was the fundamental criterion used for separating two major categories: diapause and quiescence. Quiescence, a simpler adaptation, is induced directly by unfavorable conditions, and development is resumed after the return to favorable conditions. Diapause, in contrast, includes a complicated neurohormonal mechanism that requires time for initiating and finishing its action. As a result, diapause starts before really harsh conditions develop and cannot be immediately broken, even if favorable conditions return. Evans and Perry (1976) modified Laudien’s system and divided diapause and quiescence into facultative and obligate dormancy. Most of Keilin’s specific terms were retained in this new classification as different types and degrees of facultative quiescence. Finally, latent life stages such as cryptobiosis, anabiosis, and even abiosis were classified as extremes of the diapause and quiescence categories. Hereafter, we mainly follow the Laudien (1973) scheme, with some changes (Table 1.1). We exclude the separation of dormancy into obligate and facultative types, as

9

ACTIVE LIFE Normal metabolism, development, growth, and breeding

Sleep, physiological stress, thermal shock, some illness

BOUNDARY STATES Short-time metabolism declining

DORMANCY or HYPOBIOSIS Suppressed metabolism, with long-time cessation of development, growth, and breeding QUIESCENCE DIAPAUSE (hibernation and/or aestivation) driven driven by hormones, induced by by unfavorable/favorable conditions both signal and vital factors for life Latent life Suppressed life Cryptobiosis Anabiosis Abiosis Superpause Mesopause Oligopause more than 1 3–12 months less than 3 year months

TABLE 1.1. Physiological States of Organisms Based on Metabolic Level and on the Cause of the Arrestation of Vital Functions

10

V. ALEKSEEV ET AL.

often in the same species it depends on environmental conditions and latitude (or altitude) (Wereshagin 1912; Danilevsky 1961). As there are no real differences among cryptobiosis, anabiosis, and abiosis, we use them as synonyms of deep, usually multiyear quiescence. Some boundary states like sleep and shocks are also included. Some information on variation in diapause duration should also be noted. Danilevsky (1961) predicted that for Insecta there should be a correlation between duration of diapause and its depth when measured as suppression of metabolic activity. Later, the same was confirmed for Crustacea (Alekseev & Starobogatov 1996). As soon as diapause duration is connected with physiological peculiarities of an organism in a state of dormancy, we suggest oligopause (less then 3 months), mesopause, or diapause itself (3–12 months), and superpause (longer than 1 year) as terms for these phenomena. The difference among these is clear if we compare the duration of dormancy and time for ontogenesis (Entomostraca, most insects), or the active period of the year (Decapoda, long-lived insects). This leads to a suspension of vital activities for an interval shorter than the duration of ontogenesis of the species (or season length for long-lived animals). Oligopause suppresses vital functions atleast half as much as diapause itself (Alekseev & Starobogatov 1996). An example of oligopause may be observed in some small cyclopoid copepod species in rain pools and other temporary waters of arid regions (Alekseev 1978). Superpause enables an organism to survive unfavorable conditions for years; therefore the duration of superpause exceeds the length of ontogenesis of the species. Most examples of superpause are of the embryonal type and are found among the most ancient crustaceans – Branchiopods and Cladocerans (Alekseev 1990). In some respects, superpause is close to the cryptobiotic state of hibernation (metabolism suppression, stability to negative factors), and perhaps they should be considered synonyms. It is interesting that large species of crustaceans have no superpause at all. In regard to all parameters, diapause occupies an intermediate place in this scheme and its duration is similar to that of ontogenesis (short-lived animals) or to season length (Decapoda). Viewed in another way, all of the different forms of diapause observed thus far in various orders of aquatic invertebrates can be easily molded to fit the logical pattern advanced by Danilevsky (1961) for insects. The foundation of his classification system is the instar at which an animal goes into dormancy. The largest variety of diapause is known in insects in which embryonic, larval, pupil, and adult (sometime named imaginal) types of diapause occur. Crustaceans have fewer diapausing stages, including embryonal, larval, and adult diapauses. In monogonont rotifers, only embryonic diapause exists, while Bdelloid rotifers spend dormancy in the adult stage (Gilbert 2004a,b). In this book, the properties of diapause are discussed in the three best-studied groups of aquatic invertebrates: crustaceans, rotifers, and insects. This will allow us to distinguish common features and individual peculiarities of this ancient and well-developed phenomenon in evolutionarily distant taxa like crustaceans, rotifers, and insects.

JOHN J. GILBERT

2. TIMING OF DIAPAUSE IN MONOGONONT ROTIFERS Mechanisms and Strategies

2.1 INTRODUCTION

Monogonont rotifers can have high birth rates and short generation times, and they often live in water bodies where environmental factors restrict population growth to several weeks or months. Adverse conditions include evaporation of water in temporary habitats, a variety of crustacean and insect predators, cladoceran interference competitors, unfavorable temperatures and perhaps especially for specialist species, food limitation. Consequently, rotifer populations may develop rapidly, fluctuate markedly in size over short time periods and quickly disappear from plankton or littoral communities. In environments unfavorable for population growth, populations typically survive as diapausing eggs in sediment egg banks. Dormancy in monogonont rotifers seems to rarely involve quiescence, where greatly reduced activity in a life-cycle stage is directly caused by an unfavorable condition and lasts only as long as that condition persists (Ricci 2001). Quiescence is the only mechanism of dormancy in bdelloid rotifers (Gilbert 1974; Ricci 2001). Eggs, embryos, juveniles or adults of some bdelloids can remain viable for long periods when dried (anhydrobiosis) or at very low temperatures (cryobiosis) and then quickly recover when rehydrated or warmed (Ricci 1998; Ricci & Caprioli 1998; Ricci et al. 1987, Örstan 1998). Monogononts can have greatly extended survival and generation times when cultured at low temperatures (Halbach 1970; Pourriot & Deluzarches 1971; Hirayama & Kusano 1972; Pourriot & Rougier 1975; Walz 1983, 1987). Furthermore, Brachionus plicatilis can survive storage at low temperatures for many days with or without food. Survival of fed rotifers at 4°C varied from 30% to 65% after 31 days, but both rotifer and food (yeast) densities in these experiments were very high (Lubzens et al. 1990). Maximal survival of rotifers previously fed on Isochrysis and then starved at –1°C was ~50% after 12 days, but again rotifer densities were very high (Lubzens et al. 1995). Further research under more ecological conditions is needed to determine the ability of monogonont rotifers to reproduce or persist, perhaps in a quiescent state, at temperatures too low for positive population growth. An ability to do this might allow a rotifer population to survive in winter or in hypolimnetic refuge habitats between growing seasons without producing diapausing eggs. This chapter considers mechanisms and strategies for the production of diapausing eggs in monogonont rotifers, and shows that some of the generally accepted rules about the life cycle may be broken. The focus is on different types of diapausing eggs, and on extrinsic and intrinsic factors that control the initiation of diapause. A summary of this information is presented in Table 2.1. Timing of diapause is of considerable ecological interest, because production of diapausing 11 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 11–27. © 2007 Springer.

12

J. GILBERT

TABLE 2.1. Diapausing Eggs Found in Monogonont Rotifers, their Diapause Durations, and Mechanisms Controlling their Induction Type of diapausing egg

Diapause duration

Fertilized

Months to years

Resting egg

Pseudosexual egg

Months to years

Amictic egg

*Indicates

Days or months

Rotifer Brachionus, 4 species Epiphanes, 2 species Rhinoglena frontalis Asplanchna, 4 species Notommata, 2 species Trichocerca rattus Hexarthra from Chihauhuan Desert Keratella hiemalis Notholca squamula Synchaeta pectinata

Mechanism of induction

References

Crowding

Gilbert 2004b*

Dietary αtocopherol Long photoperiod

Gilbert 1992*

Endogenous

Walsh et al., unpublished data Ruttner-Kolisko 1946 Schröder 1999

Unknown Unknown Food limitation

Pourriot 1963; Pourriot & Clément 1975

Gilbert 1995; Fradkin 1997; Gilbert and Schreiber 1998

a review.

eggs occurs at the expense of population growth via female parthenogenesis (Snell 1987; Ciros-Pérez et al. 2002; Gilbert 2002; Serra et al. 2004; Schröder 2005). Many details of the life cycle have been reviewed (Gilbert 1983a, 1992, 1993; Schröder 2005), and factors controlling the survival and hatching of diapausing eggs are considered elsewhere (Gilbert 1974; Gilbert & Schröder 2004; Schröder 2005). 2.2 FEMALE TYPES AND THE FERTILIZED RESTING EGG

In most rotifers with a heterogonic life cycle there are two categories of females (Table 2.2, Fig. 2.1). Amictic females produce diploid eggs (amictic eggs) that develop by ameiotic parthenogenesis into females. These eggs typically are subitaneous, developing without arrest or diapause. Mictic females are morphologically similar to amictic females (Gilbert 1974), but produce haploid (mictic) eggs via classical meiosis (Gilbert 1983a). These develop parthenogenetically into haploid males or, if fertilized, into resting eggs. The fertilized resting egg is the only known diapause stage, and diapausing egg, in most rotifers. Thus, diapause usually is initiated in the life cycle only when mictic females are produced and sexual reproduction, or

TIMING DIAPAUSE IN ROTIFERS

13

TABLE 2.2. Some rules of, and Exceptions to, the Classical Life Cycle of Heterogonic Rotifers (see Fig. 2.1) Rule

Statement

Exceptions

1

Amictic females produce some mictic daughters whenever environmental signal is present

2

Amictic females hatched from resting eggs and subitaneous eggs are similar physiologically and ecologically Resting eggs always hatch into amictic females

(a) Stem females, and amictic females in some subsequent generations, may not respond fully, or at all, to signal (b) Mixis in Hexarthra from rock pools (Chihauhuan Desert) is under endogenous control and not triggered by an environmental signal (a) See 1a (b) Stem females have more lipid reserves

3

4

Fertilized resting eggs are the only diapause stage

Mictic stem females are rare in Brachionus, but common in Hexarthra from Texas huecos (see 1b) (a) Pseudosexual eggs in Keratella hiemalis and Notholca squamula (b) Diapausing amictic eggs in Synchaeta pectinata

mixis, ensues. In some rotifers and habitats, sexual reproduction has not been observed, and may not occur. Absence of sexual reproduction often is associated with relatively stable environments such as large lakes (Wesenberg-Lund 1930; Ruttner-Kolisko 1974). However, in habitats where some species reproduce sexually, others do not (Schröder 2001). Typically, mictic females produce resting eggs only when they are inseminated at a young age (Buchner et al. 1967; Snell & Childress 1987; Gómez & Serra 1996). Males copulate more readily with young than older females, and sperm fail to fertilize mictic eggs when males do inseminate older females. Thus, males do not fertilize the eggs of their own mothers, and populations must contain high densities of males and mictic females to assure a high probability that young mictic females will be encountered, inseminated, and able to produce resting eggs (Snell & Garman 1986). Mictic females usually produce only resting eggs. However, some mictic females can produce both males and resting eggs (Bogoslovsky 1960; Buchner et al. 1967; Haberman & Sudzuki 1998). During periods of sexual reproduction, amictic females usually continue to produce some amictic as well as mictic daughters. In laboratory populations, the percentage of mictic offspring usually does not exceed 40–60% (Pourriot and Clément 1975; Kabay & Gilbert 1977; Snell 1987; Gilbert 2002, 2003a, 2004b). In natural populations, the percentage of mictic females is rarely close to 100%,

14

J. GILBERT

Figure 2.1. The classical heterogonic life cycle of monogonont rotifers. Four long-held rules for this life cycle are indicated; see Table 2.2 for statements of, and exceptions to, the rules. Amphoteric females, known to occur in a few rotifers, are not included in the diagram.

commonly is not more than 30%, and may be as low as 1–10% (Gilbert 1974; Carmona et al. 1995; Miracle & Armengol-Díaz 1995; Schröder 2001, 2005). The partial mixis response in a clonal population is a bet-hedging strategy that balances the trade-off between population growth by female parthenogenesis and sexual reproduction, and can optimize the number of resting eggs produced (Snell 1987; Serra & Carmona 1993; Serra & King 1999; Serra et al. 2004). In at least several rotifers the transition from female parthenogenesis to sexual reproduction is more complex, because there is a third category of female, an

TIMING DIAPAUSE IN ROTIFERS

15

amphoteric female, which produces both amictic and mictic eggs (Gilbert 1983a; Haberman and Sudzuki 1998). Amphoteric females are not included in Fig. 2.1, but have been in the life-cycle diagrams of other authors (King & Snell 1977; Nogrady et al. 1993; Wallace & Snell 2001). The percentage of amphoteric females in a population usually is low. It was 12 h, 16°C 16 h, 25°C or 30°C

LD 15°C and 20°C 24 h, 29°C

Jordan and Bradshaw 1978 Ahmadi et al. 1985 Mori et al. 1985; Galka and Brust 1987b

Love and Whelchel 1955; Vinogradova 1967; Gallaway 1985

98

E. VINOGRADOVA

Figure 5.3. Effect of photoperiod on the larval development in Anopheles plumbeu. (After Vinogradova 1969.) A, LD 16:8, 16°C; B, LD 9:15, 16°C; C, LD 9:15, 20°C.

All immature stages are sensitive to photoperiod, but the key factor is the photoperiodic regime experienced in late larval instars. Continuous photoperiodic sensitivity promotes the effectiveness of photoperiodic reactivation. Under long-day conditions at 15°C all 4th-instar larvae resume development and pupate within 30 days, whereas under short-day conditions (LD 12:12) no reactivation is observed and during 6–7 months one half of the diapausing larvae perish; under LD 13:11 only a portion of the larvae may terminate diapause. At 15°C diapausing larvae may survive up to 13 months, but only 4% pupate within this time frame. The increase of temperature (20–25°C) stimulates pupation, though its rate under some photoperiods is lower when compared with the pupation incidence at 15°C. A similar effect of photoperiod and temperature on the induction, maintenance, and termination of larval diapause was revealed in the spring malaria mosquito, A. claviger from Luga (60°N), Russia (Vinogradova 1963). The nonbiting pitcher-plant mosquito, W. smithii inhabits water-filled leaves of S. purpurea. This plant is found from the Gulf of Mexico to Canada. The mosquito range follows that of its host (30–55°N). Larvae of W. smithii normally diapause as 3rd-instars, but observations in Massachusetts revealed that the 4th-instar diapause phenotype is abundant in the spring after termination of 3rd-instar diapause and in autumn, when a new overwintering generation of 3rd-instar larvae accumulated in the pitcher-plant habitat (Farkas & Brust 1986; Louinibos & Bradshaw 1975). However, 4th-instar larvae did not survive the winter. 4th-instar diapause could be induced in diapausing 3rd-instar larvae by brief exposure to long-day photoperiods followed by short-day or by a long-term exposure in short days at 25°C. Continuous exposure to long days readily terminated 4th-instar diapause. The 3rd-instar diapause of W. smithii is initiated, maintained, and terminated by photoperiod (Bradshaw & Louinibos 1972; Bradshaw & Phillips 1980). The photoperiodic cues are monitored by early instar larvae. Development is also limited by

DIAPAUSE IN AQUATIC INSECTS

99

Figure 5.4. Effect of photoperiod on the induction and termination of the 3rd-instar diapause in pitcher-plant mosquito, Wyeomyia smithii. (Modified from Bradshaw & Louinibos 1972.) The photoperiodic response curve for induction of diapause (on the left) and termination of diapause (on the right). The induction of diapause was studied in groups of the 1st-instar larvae which were reared at 21°C; the termination of diapause (pupation) was studied in the 4th-instar larvae of 25 days old at 21°C.

temperatures below 15°C. Long days avert or terminate, and short days promote or maintain diapause (Bradshaw & Louinibos 1972). Approximately 3 long days are required for the median number of larvae to terminate diapause, although they do not molt to 4th-instar for another 6.5 days. The critical day length is identical for both the initiation and termination of diapause (Fig. 5.4). The interaction between photoperiod, temperature, and chilling was analyzed in diapausing larvae of T. rutilus (Bradshaw & Holzapfel 1977). Chilling of dormant larvae promoted a response to progressively shorter day lengths, thus reducing the critical photoperiod. Chilling also accelerated the response to long day lengths, thereby reducing the depth of diapause; after a prolonged period of exposure to cold, the larvae sometimes eventually terminated diapause directly, and subsequent development was independent of photoperiod. The best chilling temperature for producing these effects was between 4°C and 16°C (probably ~7°C). In connection with photoperiodism the photic environment of W. smithii was studied (Bradshaw & Philips 1980). Both in dawn and in dusk diapausing larvae were photoperiodically most sensitive to blue light (390–450 nm) with a shoulder in response in the blue-green and green (480–540 nm) regions of the spectrum. The photic environment of W. smithii during twilight is rich in yellow-green light, but sufficient light is available at 390–540 nm to trigger a photoperiodic response early during morning civil twilight and to sustain the response until late in evening civil twilight.

100

E. VINOGRADOVA

In O. triseriatus the threshold of light intensity, which may be perceived by larva is very low, about 0.012 lux, and the effective region of the spectrum is 320–720 nm (Wright 1967). In T. rutilus it is even less than 0.002 lux (Jenner & McCrary 1964). In W. smithii larval food is an additional environmental cue, which was most marked in the larval part of the life cycle (Roughgarden et al. 1975). Each larva was programed for pupation or diapause when in the 3rd-instar at 20–30 days of age: the food level widely modified the fraction of the cohort entering diapause (from 15% to 100%) compared with an expected figure of 56% from photoperiod alone. Thus, the most common stage for a larval diapause in mosquitoes is the 3rd- and 4th-instars. This diapause is most commonly induced by short-day photoperiods experienced during larval development; long-day photoperiods promote continuous development resulting in pupation. No maternal effects have been linked to larval diapause. The photoperiodic effect is observed usually within a certain temperature range; high temperatures decrease the frequency of diapause, while low temperatures increase it. Typical photoperiod–temperature interaction for larval diapause induction may be illustrated by data in O. togoi (Galka and Brust 1987b). As Fig. 5.5 shows, larval diapause occurs when immature individuals are maintained at 16°C and lower, at photoperiods shorter than LD 12:12. There are some trends concerning the geographical variation in larval diapause and the environmental cues controlling it. These include the frequency of diapause over a range of distribution, the intensity of diapause and the critical photoperiod inducing 50% response. The spring malaria mosquito, A. claviger, occurs in Europe, North Africa, and Asia (West Siberia, Middle Asia, Minor Asia) (Gutsevich et al. 1970). In the northern part of its range mainly 3rd- and 4th-instar larvae overwinter, whereas in southern and western parts all instars may be observed in winter. Only the 4th-instar larva undergoes diapause. The retarded development of other instars is induced directly by low temperatures and resumes in response to its increase

Figure 5.5. Effect of photoperiod (A) and temperature (B) on the larval development in Ochlerotatus togoi (Modified from Galka & Brust 1987a, b.) A, 16°C; B, LD 10:14.

DIAPAUSE IN AQUATIC INSECTS

101

(Vinogradova 1969). Duration of the 4th-instar hibernation varies over the area of distribution from 6–7 months in the northern regions of Russia to 2–3 months in Uzbekistan and Georgia. From north to south both the diapause intensity and incidence of diapause decrease, and in the southern regions such as Italy, Syria, Morocco, and southern Turkmenia no diapause occurs. In the tree-hole mosquito, O. sierrensis, a widespread species on the Pacific slope that ranges from British Columbia, Idaho, and Montana to southern California, five populations from the USA, from a range of over 10 degrees latitude (33–44°N) were studied (Jordan 1980a, b; Jordan & Bradshaw 1978). Short days elicited 100% of diapause among the 4th-instar larvae from central Oregon and northern California, but the incidence of diapause decreased with latitude so that in southern California only 35% of the sample populations entered diapause. Among that portion of the population capable of responding to photoperiod, the critical photoperiod increased by 1 h for each 4–8 degree increase in latitude. A second tree-hole mosquito, O. geniculatus, occurring in Europe, Africa, and Asia Minor, has both an egg and larval diapause. Two strains of this species were investigated (Sims & Munstermann 1983). Larvae of the English strain have a longer critical photoperiod for diapause induction and stronger diapause intensity as compared with those from Sardinia, Italy. In a third species, W. smithii photoperiodic response was studied for 22 populations collected at different latitudes, longitudes, and altitudes in North America (Hopkins & Bradshaw 1976). It was established that the growing season (the mean number of freeze-free days), which closely correlated with latitude and altitude but not with longitude, was an excellent predictor of critical photoperiod, and that an increase in latitude of 1 degree was equivalent to an increase in altitude of 122 m. The subsequent experiments with exotic light and dark cycles of varying period supported the conclusion that photoperiodic time measurements regulating larval diapause in W. smithii vary in a close relationship with latitude. The critical photoperiod mediating the maintenance and termination of diapause was found to be positively correlated with latitude among populations from southern (30–31°N), intermediate (40°N), and northern (46–49°N) latitudes in the USA and Canada. The geographic variation of larval diapause in O. triseriatus from the USA was analyzed using experimental data for eight local populations distributed from 30–40°N and from 30–1,100 m of altitude (Holzapfel & Bradshaw 1981). Photoperiod was shown to have a significant effect not only on the initiation and maintenance of diapause but also on the rate of postdiapause development, both directly and by modifying response to temperature. The response to temperature was mainly a function of photoperiod, and the Q10 for rate of completion of the 4th-instar was proportional to photoperiod. The critical photoperiod for the induction and maintenance of larval diapause, and for rate of development is shorter than the critical photoperiod controlling egg diapause (Fig. 5.6). This was also confirmed by Sims (1982), who studied O. triseriatus from other locations in the USA from 26°N to 46°N. According to Holzapfel and Bradshaw (1981), the critical photoperiod for rate of larval development increases by 1 h for each increase in latitude of 2.06° or in altitude of 549 m. The adaptive significance of larval diapause may relate to the

102

E. VINOGRADOVA

Figure 5.6. Effect of latitude on the critical photoperiod for the induction of larval and egg diapauses and for the rate of postdiapause development in the tree-hole mosquito, Ochlerotatus triseriatus (Modified from Holzapfel & Bradshaw 1981.)

modification of late winter and spring development rather than overwintering. In southern populations having incomplete egg diapause, progressively milder winter conditions favor larval diapause and continuous development. Larval diapause is mainly a backup or fail-save system for egg diapause. All authors suggest that egg diapause is the primary state in which O. triseriatus enters winter. During winter diapause eggs undergo chilling and may terminate diapause. Developments of larvae, which hatch during winter or spring, then become dependent on temperature and photoperiod. It has been proposed that egg and larval diapauses are not discrete, adaptive developmental strategies, but are part of an integrated finely tuned developmental continuum. Therefore it is suspected that polygenic control of diapause and development with diverse pleiotropy underlies the initiation, maintenance and termination of both egg and larval diapauses (Holzapfel & Bradshaw 1981). In certain mosquito species sex-related differences in diapause were established. Among eight populations of O. triseriatus from the USA from 26°N to 46°N female larvae had a stronger diapause than male larvae (Sims 1982). The differences between the male and female intensities of diapause were most pronounced in four southernmost populations. Similar data are known for larval diapause in a population of O. geniculatus (Sims & Munstermann 1983). English males were less likely to enter either diapause state and had a less intense diapause than females. In a laboratory strain of Culiseta melanura the larval diapause is induced by short day (LD 9:15) at 15°C. After transfer of these diapausing larvae at 5 weeks of age to 23°C, long day (LD 18:6) males began to pupate on days 11–13 post transfer and were succeeded by females on days 40–44 post transfer. The waves of pupation were separated by a prolonged delay with no overlap between the two. In this case the ability to enter diapause was preserved after 18 years (400 generations) of

DIAPAUSE IN AQUATIC INSECTS

103

continuous laboratory rearing in standard conditions (27–29°C and LD 16:8) without larval diapause. 5.2.3 Adult Diapause 5.2.3.1 Syndrome of adult diapause. Adult mosquitoes overwinter in various natural (caves, hollows, holes, burrows, etc.) and artificial (cellars, vegetable store-houses, empty sheds, unheated basements, catacombs, bunkers, etc.) shelters. For instance, in Henan Province, China, hibernating females of Culex pipiens pallens preferred warm (>5°C), moist (RH > 60%), and dim (illumination no more than 5 lux) shelters away from the wind (< 0.25 m/sec) (Su-TianYun et al. 1994). In mosquitoes only females undergo an adult or reproductive (ovarian) diapause. Usually nulliparous inseminated females enter hibernation as has been shown for C. pipiens pipiens (Oda & Kuhlow 1974; Vinogradova 2000), C. pipiens pallens, C. inornata (Hudson 1979), C. tarsalis (Reisen et al. 1986a), A. earlei, C. territans (Hudson 1978), and C. peus (Skultab & Eldridge 1985). Exceptions occur rarely; for instance, in England only 5–9% of hibernating females of C. pipiens pipiens were unfertilized (Onyeka & Boreham 1987). In mosquitoes the adult diapause syndrome involves a set of important characteristics, such as arrest of ovary development, reduced avidity, metabolic changes leading to progressive accumulation of fat body reserves, and an altered behavioral pattern. Mosquitoes, like other blood-sucking Diptera, which periodically ingest blood, are characterized by a strong correlation between blood-digestion and ovarian development. This is a main link in the gonotrophic cycle, which includes host-seeking, blood-feeding with subsequent blood digestion and ovary development, and oviposition (Swellengrebel 1929; Beklemishev 1940; Clements 1963; Washino 1977). Gonotrophic concordance is typical of the summer gonoactive females in which one blood meal is necessary and sufficient for maturation of one batch of eggs. On the contrary, gonotrophic dissociation (failure of ovarian follicles to mature beyond the resting stage following a full blood-meal) is interpreted as an expression of facultative diapause. This term was originally used to denote the cessation of egg production despite the continued taking of blood meals by overwintering anopheline females, but later it was also used similarly for other mosquito genera. The process of egg maturation in mosquitoes was divided into several developmental stages, which are frequently used to evaluate the state of the ovary (Christophers 1911; Mer 1936). The follicle length or follicle to germarium length ratio (F:G) is another criterion for characterizing the ovarian state in mosquitoes before the first blood meal. The relationship between the two above-mentioned classifications was determined in C. peus (Skultab and Eldridge 1985). The mean length of the primary follicle and F:G ratio appear to increase from 0.060 mm (1.4:1) to 0.117 mm (2.4:1), respectively, when the follicles develop from stage N to stage IIb (the resting stage). The main characteristic of reproductive diapause is the inactive state of the ovaries. In females of C. pipiens pipiens, C. pipiens pallens, and C. peus, C. restuans, which enter diapause after adult emergence, the primary follicles are usually small (0.05–0.06 mm) and the F:G ratio is less than 2:1 (Spielman & Wong 1973; Oda &

104

E. VINOGRADOVA

Kuhlow 1974; Skultab & Eldridge 1985; Madder et al. 1983; Wang et al. 1984). However, during hibernation the primary follicles may develop up to stage II as was shown in C. pipiens pipiens (Kupriyanova 1968; Oda & Kuhlow 1973). Mosquito females that enter diapause may have hypertrophied fat body, which contains huge lipid reserves (Vinogradova 1969; Clements 1992). About 23 to 43 fatty acids were found in the fat body of diapausing C. tritaeniorhynchus, but only seven of them were predominant (Xue-RuiDe et al. 1991). In autumnal females of C. pipiens pipiens (St. Petersburg, Russia) fat deposits formed as much as 67% of the dry weight and 37% of the live weight of a mosquito, while in spring it dropped to 49% and 22%, respectively (Vinogradova 1969). In southern England, in October the mean quantity of fat in overwintering mosquitoes was found to be 1.06 mg/mosquito and in March it was 0.2 mg/mosquito (Onyeka & Boreham 1987). In some species, e.g. C. tarsalis, the hibernating females synthesize lipid from plant juices, which they consume in autumn. Thus, the females ingest fructose when entering and terminating diapause. In another species, e.g. A. freeborni in California, USA, females develop extensive lipid reserves whether fed sugar alone or blood alone (Reisen et al. 1986b; Clements 1992). In the majority of mosquito species the reproductive diapause is followed by an abrupt decrease in avidity and a cessation of blood feeding (C. pipiens pipiens, C. tarsalis, C. restuans, A. hyrcanus, A. messeae, and probably C. modestus, C. apicalis, and C. bitaeniorhynchus). Only few mosquitoes take a blood meal periodically during hibernation; e.g. A. superpictus, A. sacharovi, and A. atroparvus (Vinogradova 1969). These mosquitoes usually overwinter in warm hibernation shelters, such as cattle sheds, where they may take occasional blood meals resulting in gonotrophic dissociation. However, the capacity to winter bloodsucking may vary considerably both within the species area of distribution and in the same place in connection with different weather conditions. The behavior patterns of diapausing female mosquitoes include: negative phototaxis forcing them to migrate into hibernation shelters, a reduced locomotor activity, whose degree depends on temperature and illumination, and modified host-seeking behavior. As to locomotor activity in the hibernacula, C. pipiens pipiens females, for example, still retained mobility at temperatures above zero, but at −3°C they acquired the characteristic “hibernational” pose and would not react to external stimuli (Vinogradova 2000). It was shown that ~50% of the diapausing individuals of this species changed their location every 6 days, searching for more favorable temperature conditions in the hibernacula (Onyeka & Boreham 1987). One further component of the mosquito hibernation behavior concerns hostseeking activity. Different aspects of this behavior have been considered in C. pipiens pipiens (Mitchell 1983; Bowen 1990, 1991, 1995). It was established that hostseeking behavior of diapausing females was depressed during the whole reproductive diapause, while in nondiapausing individuals it was absent only within 4 days after emergence. Changes in host-seeking behavior are known to correlate with the responsiveness of mosquito peripheral receptors sensitive to lactic acid, which is

DIAPAUSE IN AQUATIC INSECTS

105

one of the host-seeking attractants. Electrophysiological experiments established that in diapausing mosquitoes the state of some parts of the sensory system may change so that the peripheral receptors lose their sensitivity to lactic acid. But following diapause termination both this sensitivity and host-seeking behavior are commonly restored. In postdiapausing females some highly sensitive lactic acidexcited cells in the antennal basiconic sensilla of A3 type have been identified. 5.2.3.2 Photoperiod and temperature induction of adult diapause. Almost all of the mentioned characters of the reproductive diapause of mosquitoes appear to be controlled by photoperiod and temperature. The effect of photoperiod on bloodfeeding activity is typical for those species, which do not take a blood meal during hibernation. In experiments with C. tritaeniorhynchus from Japan (Eldridge 1963), an average 92% of females fed on blood at 28°C when they were reared under longday conditions (LD 14:10), but not more than 11% of females fed on blood under short day exposure (LD 8:16). Low blood-feeding activity increased after transfer of mosquitoes to long-day conditions, depending on the number of long-day cycles received (Fig. 5.7). Similar results were obtained in C. pipiens pipiens from St. Petersburg: under LD 24:0 and 12:12 50% and 2% of the females fed, respectively (Danilevsky & Glinyanaya 1958). Likewise, A. hyrcanus (Vinogradova 1969) and C. pipiens pallens (Hosoi 1954) responded the same way.

Figure 5.7. The activity of blood-feeding of females in Culex tritaeniorhynchus. (After Eldridge 1963.) 1, females were kept at 28°C and LD 8:16 from the time of pupation; 2, females were kept at 28°C and LD 8:16 from the time of pupation to 10 days after adult emergence and were transferred to LD 14:10.

106

E. VINOGRADOVA

Photoperiod and temperature may have effects both on the follicle size in unfed females and on ovarian development in blood-fed females. Thus, in C. peus a low temperature of 15°C caused retardation of follicle growth at the preresting stage or earlier (stage N, 0.06 mm follicle length) regardless of photoperiod in non-blood-fed females. Only under the influence of both low temperature and short photoperiod did ovaries remain in this condition for as long as 21 days (Fig. 5.8). At 25°C follicles of both photoperiod groups developed well beyond the preresting stage. The critical photoperiod was LD 13:11 (Skultab & Eldridge 1985). Similar results were observed earlier for C. pipiens pipiens from the USA (Eldridge 1968; Sanburg & Larsen 1973). At 22°C and short photoperiods (LD 12:12 and less) the follicle length was small (0.054 mm), but at long photoperiods (LD 13:11 and more) it was longer (0.070 mm). In C. restuans the females maintained at 15°C and long- or short-day conditions differed also in their follicle length and stage of follicles (Eldridge et al. 1976). Females of C. tarsalis responded to short photoperiod (LD 8:16) at 22°C by minimal ovary length and by increased size of the fat body (Harwood & Halfill, 1964) The photoperiodic response curve for induction of winter reproductive diapause in blood-fed females was established for C. pipiens pipiens from St. Petersburg (60°N), Russia (Vinogradova 1961). At 23°C the incidence of diapausing individuals with gonotrophic dissociation was only 3–6% under long-day conditions (LD 20:4, 18:6) and increased to 86% under short-day treatment (LD 12:12). In a more southern strain from Azerbajan, Russia (40°N), at 24°C the photoperiodic response was weak and the incidence of diapause was low; only 11% of the females entered diapauses at LD 12:12. Effects of both photoperiod and temperature on ovarian development in blood-fed mosquitoes were investigated in detail in C. pipiens pipiens from Indiana, USA (Eldridge 1968). As Fig. 5.9 shows, at 20°C and 25°C almost all

Figure 5.8. Effect of photoperiod and temperature on the ovarian follicle size of Culex peus from Oregon, USA (Modified from Skultab & Eldridge 1985.) 1, LD 8:16, 15°C; 2, LD 16:8, 15°C; 3, LD 8:16, 25°C; 4, LD 16:8, 25°C.

DIAPAUSE IN AQUATIC INSECTS

107

Figure 5.9. Photoperiodic response curve for the induction of the adult diapause in Culex pipiens from Indiana, USA. (After Eldridge 1966.)

individuals had matured ovaries, irrespective of photoperiod length. Gonotrophic dissociation occurred in response to the same conditions that inhibited bloodfeeding, i.e. a combination of low temperature and short photoperiod. Besides photoperiod and temperature, larval crowding can also influence the incidence of adult diapause in C. pipiens pipiens, as reported for a Canadian strain from southern Ontario (Madder et al. 1983). The combined effect of two photoperiods, three temperatures, and four larval population densities (50, 100, 250, and 500 larvae per 700 ml water) was investigated to show that, with a decrease in temperature and day length, and rising larval density, the incidence of diapausing females increased. The photoperiodic sensitivity during ontogeny in C. pipiens pipiens was studied using different combinations of temperature and photoperiod: larvae were reared at 25°C, pupae were held at 15°C, and adults at 10°C (Eldridge 1965). Ovarian development was reduced by ~50% in mosquitoes subjected to a short photoperiod (LD 12:12) during at least two of the three developmental stages, regardless of the order of treatment. Conversely, when two or more stages were subjected to a long photoperiod (LD 16:8) ovarian development occurred in no less than 83% of the cases. Sanburg and Larsen (1973) confirmed that in determination of the adult diapause the two last stages (pupa and adult) were most important. In another species, C. restuans from Washington, USA, both temperature and photoperiod, to which females were subjected during the pupal stage, and for up 6–8 days afterwards influenced the development of the ovaries; the combination of 15°C and LD 8:16 resulted in the ovaries remaining in a diapause state (Eldridge et al. 1976). Interesting data on the induction of diapause by a decrease in day length were obtained for C. inornata (Hudson 1977). Females were reared and maintained at constant LD 16:8 and 20°C until larval–pupal ecdysis and then were transferred to regimes of LD 16:8 or LD 12:12 at 10, 15, or 20°C. At 14–15 days after adult emergence, follicles of the females at short day were as small as those of diapausing

108

E. VINOGRADOVA

females. The proportion of females 7–8 days after emergence that fed on a man and the proportion of blood-fed females that matured eggs, increased both with day length and temperature, the lowest rates being seen in females at short day length and 10°C (20% fed and none matured eggs) and the highest in females at long day length and 15°C (72% fed and 96% matured eggs). Thus, in this species the final regime determined both the blood-feeding activity and ovarian development. The photoperiodic response was also studied in some anopheline mosquito species (Vinogradova 1958, 1960). The malaria mosquito, A. messeae from St. Petersburg, Russia, is a typical “long-day” insect. Only 11% of the females that were subjected to a long photoperiod (LD 22:2) entered diapause, while short photoperiods (LD 15:9 or less) induced diapause in nearly all individuals. A comparison of the photoperiodic response of two strains from St. Petersburg (60°N) and Astrakhan’ (47°N) confirmed the geographical variation of the critical day length in this species. A. hyrcanus (probably A. hyrcanus complex), a group widely distributed in southern Europe and Asia (to 50°N) undergoes adult diapause only in the northern and middle parts of the area – Central Asia, China, Japan. The photoperiodic induction of diapause was experimentally established in two strains (Vinogradova 1969). In the strain from Astrakhan, at 23°C under long day length (LD 18:6, 16:8, 15:9), no diapause was observed, but under short day length (LD14:10, 13:11, 12:12) 31%, 46%, and 68% of females, respectively, entered diapause. In another strain from southern Tadjikistan (37°N) only 37–56% of the females underwent diapause under short day length. The hypertrophied fat body, which is typical for diapause of mosquitoes, may have different origins. Diapausing females may accumulate their fat reserves both through sugar or blood-meal and through the utilization of larval reserves (Vinogradova 1969). It was shown experimentally that in C. bergrothi, C. alaskaensis, and Allotheobaldia longiareolata the extent of fat body in females of 10–12 h age was dependent on the photoperiod and temperature conditions during their larval development. Low temperature and short day exposure during immature stages promoted the accumulation of adult fat reserves. This effect was more highly expressed under daily variations of temperature from 6°C to 29°C as compared to a constant temperature of 18°C. The role of carbohydrate feeding by adults on fat body formation has been extensively examined in C. pipiens pipiens (Mitchell & Briegel 1989). Sugarfed mosquitoes accumulated significantly more lipids under short-day conditions (LD 9:15) than under long-day (LD 15:9) conditions, the values being 11.4 and 6.64 cal., respectively. The extent of the fat reserves also depended on the temperature at which the mosquitoes were maintained: the fat reserves of females fed on sugar at 25°C and 15°C were 1.610 mg (33%) and 2.09 mg (37%), respectively (Tekle 1960). A photoperiodic effect also is apparent in mosquito species, which periodically take a blood meal during diapause. It was established experimentally that in A. atroparvus, A. messease, A. superpictus, and A.hyrcanus the percentage of the blood-fed females with hypertrophied fat body was greater in short-day conditions than in long-day conditions (Vinogradova 1969). It is interesting to record that a

DIAPAUSE IN AQUATIC INSECTS

109

similar situation was observed in southern populations of C. inornata where females aestivate during the hot summer month and general activity occurs primarily in autumn and winter. In California, USA, at 33–34°C the blood- and sugar-feeding females developed increasingly hyperthrophied fat bodies between April and June before ceasing activity. It has been shown experimentally that both parous and nulliparous females synthesized very substantial lipid reserves when reared from egg to adult under long photoperiod (Barnard & Mulla 1977). 5.2.3.3 Adult diapause termination. The termination of adult diapause in mosquitoes has not been adequately studied. The majority of data concerns C. pipiens pipiens. Experimental evidence for the photoperiodic stimulation of blood-feeding in field hibernating mosquitoes in England was obtained long ago (Tate & Vincent 1936). These females were subjected to two light regimes at 17–20°C: under continuous illumination and in the dark 70–85% and 13% of mosquitoes took blood meals, respectively. Diapausing females of two other populations, from Boston, USA, and from Hamburg, Germany, exposed to a long photoperiod also renewed ovarian development (Spielman & Wong 1973; Oda & Kuhlow 1974). Experiments with an English population testify to a dynamic character of diapause (Onyeka & Boreham 1987). The minimum photoperiodic exposure (LD 24:0, 21°C), which restored blood-feeding activity in diapausing mosquitoes decreased as the hibernation period increased. This exposure was 12 days in August–October (26% females took a blood meal), 6 days in February–March (25–47%), and 3 days in April (53%). A similar picture was observed for diapausing mosquitoes near Moscow (Kupriyanova 1968). The most intensive diapause occurred in August–November, when in the laboratory females fed on blood reluctantly, and after the first blood-feeding only 36% of the females had mature ovaries; in January all blood-fed individuals became gravid, i.e. during field hibernation diapause ends gradually and spontaneously during early or late winter. In C. tarsalis diapause termination is a function of increased day length (Mitchell 1981). Exposure to a long day (LD 15:9) results in diapause termination in essentially all mosquitoes by day 7 at 25°C, and the host-seeking behavior is restored in such females. In another mosquito, C. inornata, females entering adult diapause in Canada during September–October had small ovarian follicles and did not take a blood-meal in nature (Hudson 1979). In some females that were kept for 7 days at 20°C and LD 16:8 follicle growth occurred and blood-feeding was followed by egg maturation. When females were kept for 2–3 months at 15°C and short day lengths, follicle development also occurred and egg maturation was observed in all blood-fed females. Thus, during hibernation diapausing mosquitoes are undergoing what is commonly called “diapause development,” or reactivation. Though adult diapause can be terminated by subjecting them to long photoperiod and high temperature, in nature individuals usually overwinter in dark shelters, where photoperiodic reactivation hardly occurs, therefore low temperature reactivation is likely to be more relevant.

110

E. VINOGRADOVA 5.3 OTHER GROUPS OF AQUATIC INSECTS

5.3.1 Chironomids (Chironomidae) Chironomidae are an important part of the benthos and provide food for many fish species. In chironomids the diapausing stage(s) are species-specific as well and even in the same genus diapause may occur in the egg stage or in instars 2, 3, or 4, depending on the species (Danks 1971; Goddeeris et al. 2001). In the voluminous genus Chironomus diapause most often takes place in mature larvae of the 3rd- or 4th-instars. Diapausing larvae display important differences in their metabolism. As was shown in Chironomus plumosus, the oxygen consumption of the diapausing larvae drops ~30% as compared to that observed in nondiapausing larvae (Adamek & Fischer 1985). In this species the diapausing larvae are still active and take up food (Ineichen et al. 1979), whereas inactive diapausing larvae not feeding and completely enclosed in characteristic cocoons have been observed in many other species (Danks 1971). Freezing tolerance is widely distributed in the genera of most subfamilies except Tanypodinae, and is probably widespread in the Orthocladinae. It was found in some temperate and all arctic larvae and was temperature- and timedependent. Chironomidae can probably be considered as preadapted to a rigorous winter environment (Danks 1971). For the first time in C. tentans it was experimentally established that short photoperiod induced and maintained the larval diapause, while its termination was favored by long day (Engelmann & Shapirrio 1965). The photoperiodic induction of diapause was later described in C. plumosus and C. nuditarsis (Fischer 1974). The larvae of Clunio marina and C. decorus enter diapause under short photoperiod and low temperature (15°C), whereas for C. staegeri and Endochironomus nigricans short photoperiod is effective both at 15°C and 20°C (Danks 1978; Neumann & Kruger 1985). C. plumosus develops continuously at 15°C and long day, and enters diapause under short day; the duration of diapause may vary from one to several months (Ineichen et al. 1979). In C. riparius both 3rdand 4th-instars may enter diapause in response to short-day and temperature (15°C and lower) induction. Geographical variation in the duration of diapause in 4th-instar of this species is known: the diapause of a Belgian strain was shorter (by at least several weeks) as compared to a synantropic strain from St. Petersburg basements, Russia, (3–9 months) (Goddeeris et al. 2001; Vinogradova & Petrova 2004). Shilova (1976) divided chironomids from Borok, Russia, into three groups depending on the environmental cues responsible for their diapause: 1. Diapause is induced by short photoperiod and comes to end after long-day treatment (C. plumosus, Polypedium nubeculosum, C. tentans, Stictochironomus crassiforceps). 2. Diapause is initiated by short photoperiod and terminates in response to long day, low temperature, or freezing (C. pulicornis, Psectrotanypus varius). 3. The induction of diapause is the same, but the reactivation occurs only after freezing (Procladius choreus, Anatopynia plumipes). These examples show that in principle the environmental control of larval diapause in chironomids is similar to that of mosquitoes.

DIAPAUSE IN AQUATIC INSECTS

111

5.3.2 Biting Midges (Ceratopogonidae) Diapause of ceratopogonids was actively studied in the USSR (Isaev 1975, 1976a, b, 1978, 1997; Glukhova 1989). Hibernation during the stage of the 3rd- to 4th-instar larva is typical for the blood-sucking midges in the genus Culicoides in the temperate zone. Such a diapause was recorded for 17 species from the Ivanovo region. In the field the inclination for larval diapause increased from the first to the second generation: for instance, in C. circumscriptus, C. pulicaris, C. salinarius, and C. nubeculosus the mean incidence of diapausing larvae was 0% and 100%, respectively. The cold reactivation of diapausing larvae occurred gradually during the winter. In 10 among 25 species studied, the tendency to monovoltinism dominated, the remaining species were multivoltine. In experiments with C. odibilis a role for photoperiodism and temperature in the induction and termination of larval diapause was shown. Besides larval diapause some species have an obligate or facultative egg diapause (Isaev 1960, 1976b). In C. punctatus the termination of egg diapause was elicited by both high and low temperatures, and the degree of the synchronous hatching of larvae increased as the temperature exposure increased. Differences in the inclination to diapause were recorded in two geographical populations of C. punctatus from Ivanovo and the Far East. In 20 species of nonblood-feeding midges in the genera Spheromias, Probezzia, Mallohelea, Palpomyia, Bezzia, Phaenobezzia, Alluaudomyia, Stilobezzia, Dasyhelea, and Forcipomyia a 4th-instar larval diapause has been observed: larvae collected in July–August did not pupate in laboratory conditions (16–20°C and natural day length) during 8–10 months. In all studied species the 1st–2nd-instar larvae also may be observed in winter but their development is delayed exclusively by low winter temperature. The reactivation of diapausing larvae occurs in winter, and diapausing 4thinstar larvae collected in the field in February–March pupate in the laboratory usually 2–3 months after, whereas the 1st–2nd instars develop without delay. Intra-as well as interpopulation variation in the incidence of diapause and its duration appears to be typical for ceratopogonids. 5.3.3 Dragonflies (Odonata) In Odonata both egg and larval diapauses occur in different instars (Corbet, 1980). Embryonic diapause occurs in certain temperate species, notably Aeshna, Sympetrum, and Lestes. For instance, Lestes congener oviposits in dry stems, the eggs undergo a bit of embryogenesis in autumn and then enter diapause in winter, at which time they are resistant to both low temperature and desiccation. Hatching of larvae is observed only after wetting and exposure to temperatures of 5°C and higher (Sawchin & Gillott, 1974). Such a response may be augmented in some other species of Lestes by sensitivity to photoperiod. Larval diapause is the most common diapausing stage for dragonflies in the temperate zone. Larval growth rate is controlled by the interaction of responses to temperature and photoperiod such that morphological development within and between certain instars is arrested or accelerated at different times of year (Corbet, 1980). A relatively simple example of the mechanism of

112

E. VINOGRADOVA

environmental regulation is provided by L. eurinus from North Carolina, USA, where the populations overwinter in three larval instars preceding the final one (Lutz, 1968). Over a wide temperature range larvae of these instars develop more rapidly under summer than under winter photoperiods. Such a response magnifies the seasonal change in growth rate due to temperature. More complex responses to temperature and photoperiod exist among certain other species from North Carolina, South Ontario, Sweden, and England. Their common feature is that one or more late instars became unresponsive to a long photoperiod stimulus in late summer or early autumn and thus enter diapause. The larvae of some dragonflies may also diapause in a dried (anhydrobiotic) state (Van Damme & Dumont, 1999). In Brazil, one larva of Pantala flavescens survived drought at least a few months and after flooding successfully completed metamorphosis. It is argued that early larval tolerance to drought may be common in Pantala contributing its success in semiarid environments; possible other species in which a similar phenomenon occurs are also listed by Van Damme and Dumont (1999). In Enallagma hageni it has been shown experimentally (Ingram 1975) that termination of diapause can be caused by exposure to a low temperature, regardless of photoperiod, or to short photoperiod at a permissive temperature. A critical element in the seasonal regulation of many dragonflies at higher latitudes is the annual reversal of response to photoperiod among one or more late instars at, or sometimes before, the autumnal equinox. This reversal can induce the population to molt synchronously at that time and can also establish a latent sensitivity to spring photoperiod (Lutz 1974). In European Leucorrhinia dubia, which spends its last winter mainly in the final instar, analogous differential responses to photoperiod operate within the final instar and thus enhance the responses to photoperiod and the degree to which each of several developmental phases is synchronized within the larval population (Norling 1976). Such responses prevent autumnal emergence and reduce temporal variation among overwintering larvae that are due to emerge the next summer. 5.3.4 Heteroptera Heteroptera, divided into Gerromorpha and Nepamorpha, include approximately 20 families. Gerromorpha are semiaquatic whereas Nepamorpha are aquatic. Their overwintering stages and the environmental cues regulating diapause have been inadequately studied. Among the species studied, both adult and embryonic diapause has been recorded. Adult diapause has been reported more frequently than embryonic diapause. Both obligate (Pelocoris femoratus, Iliocoris cimicoides) and facultative diapauses are known. It was shown experimentally that environmental control of adult diapause is similar to that of mosquitoes and other insects. Adult and certain preceding stages are sensitive to photoperiod in Notonecta undulata (Vanderlin & Streams 1977) and Gerris odontogaster (Vepsalainen 1978). Short photoperiod induction of adult diapause is correlated with wing length and alary dimorphism in Aquarius paludum (Harada & Numata 1993) and Gerris odontogaster (Vepsalainen 1971). In the latter species, a gradually changing photoperiod induced the appearance of diapausing macropters. Therefore in South Finland the 2nd generations was predominantly macropterous (Vepsalainen 1978).

This page intentionally blank

DIAPAUSE IN AQUATIC INSECTS

113

5.3.5 Ephemeroptera Immature stages of these insects inhabit the littoral zone and develop over a long time, sometimes up to 3 years. Data on their diapause is fragmentary. Probably both embryonic and larval diapauses occur. Embryonic diapause was shown in Ephemerella ignata, and it was terminated within 12 months at temperatures from 1°C to 16°C (Bohle 1972). In Baetis rhodani and B. vernus embryonic diapause began at 0°C, as well as 20°C, and continued for 1 and 12 months, respectively (Bohle 1969). Larval hibernation was recorded in Cloeon dipterum in Sweden and England (Nagell 1981). Low temperature, but not decreasing photoperiod, was decisive for the final induction of its hibernation. In Sweden this occurs at 4–5°C. Swedish larvae initiated hibernation more rapidly and were more resistant to starvation than English larvae of the same species. Acknowledgments. I thank Dr. D.L. Denlinger (Entomology Deptartment, Ohio State University) and Dr. Bart De Stasio (Department of Biology, Lawrence University, Wisconsin) for critical reading of the manuscript and language corrections and Dr. S.Y. Reznik (Zoological Institute, RAS) for help with the preparation of the text.

VICTOR R. ALEKSEEV

6. A BRIEF PERSPECTIVE ON MOLECULAR MECHANISMS OF DIAPAUSE IN AQUATIC INVERTEBRATES

6.1 INTRODUCTION

As mentioned in the introductory chapter, diapause or dormancy states similar to it, aimed to overcome harsh environmental conditions, are known in an evolutionarily wide range of organisms (Table 6.1). It is interesting that practically in all phyla studied, photoperiod plays a leading role in the seasonal biological clock of organisms, from plants to vertebrates. Bunning (1936) was the first to propose a photoperiodic response (PPR) as part of a biological clock common to all living creatures. Danilevsky, the founder of an international school of photoperiodism in invertebrates argued that the mechanism of PPR is based on a common principle in all organisms and called this one of the most intriguing problems of physiology and ecology (Danilevsky 1961). His ideas were supported by a comparison of PPR in the induction of insect and plant diapause (Tyshenko 1977). The central idea in photoperiodism is that (some part of an) organism is capable of sensing and adequately interpreting diurnal periodicity of environmental light (or the absence of it). The three main groups of aquatic invertebrates discussed in this book show many similarities, not only in common principles but also in the details of diapause induction, termination and life cycle organization (Figs. 3.4, 3.7, 5.2; Tables 6.1 and 6.2). This suggests a monophyletic origin of this ancient adaptation and a similarity in the molecular basis and genetics of diapause mechanisms among these organisms. As far as photosensitivity goes, it is now well known that when anhydrobiotic embryos of crustaceans are rehydrated, metabolic reactivation is hampered by an absence of light. This was first shown in Daphnia by Stross (1965, 1969a, b), then in Artemia by Sorgeloos et al (1977), and later confirmed in a variety of other species, in which the light stimulus could even be quantified (Alekseev 1990; Murugan & Dumont 1995). Van der Linden (1985) attempted to identify the molecular basis of this light sensitivity, and thought it might be the heme molecule. This, however, has not been confirmed, and it remains possible that other, conserved photosensitive molecules like rhodopsin are involved. Likewise, some insights in the restarting of the development program, arrested at a stage of 4,000+ cells have been obtained by Dumont et al. (1992) in fairy shrimp cysts. Here, it appears that opening calcium gates in the cell membranes of the cysts upon rehydration causes a rapid inflow of Ca++ to the encysted gastrulas, and this inflow reactivates the calmodulin molecule, which in its turn reactivates a variety of dormant (phosphorylated) enzymes. The result is a rapid return from zero (Clegg 1997) to full-speed metabolism, differentiation of the 4,000+ embryonic cells into a functional nauplius, and eclosion. How the initial deactivation of the development 115 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 115–118. © 2007 Springer.

116

V. ALEKSEEV TABLE 6.1. Examples of Diapause Among Aquatic Animals (Modified from Alekseev 1990)

Types, classes

Species

Diapausing stages

Suppressed function

Diatomea Spongia Coelenterata Turbellaria Nemertini Rotatoria Polychaeta Oligochaeta Crustacea Insecta Mollusca Bryozoa Echinodermata Pisces

Coicinodiscus coicunus Halichondria panices Aurelia aurita Hemaniella retunuova Prostoma graescense Notommata copeus Dinophilus teaniatus Aelosoma hemprichii Daphnia pulex Culex pipiens Sepia officinalis Lophopodella carteri Stichopus japonicus Nothobranchius gardneri

Aukospora Gemmules Plannules Eggs Cysts Eggs Cysts Cysts Eggs Eggs Larvae Statoblasts Adults Eggs

Development Development Embryogenesis Embryogenesis Growth Embryogenesis Growth Growth Embryogenesis Embryogenesis Maturation Development Growth and breeding Hatching

TABLE 6.2. Properties of Diapausing Organisms in Some Aquatic Animals Diapause peculiarities

Insects

Crustaceans

Rotifers

Metabolic decline, % of basal metabolism Food consumptiona, % of active animals Digestive enzyme activitya, % of active animals Metabolic enzyme activitya, % of active animals Photoperiod participation in life cycle regulation Steroid (pro)hormone participation in life-cycle regulation Maximal fat accumulation, % of dry body mass Resistance to toxic substances Minimal termination time, months Average termination time, months Transformation of the critical PPR threshold by temperature, h/°C

8–35b 0–35b 10–30b 8–50b + b,e + b,e 15b +c 1.5–3 b 6.3 + 1.2 e 2.9 + 0.8 c

12–30c,d 0–33c 5–20d 15–35d +c +c 18 d +c 2–3 c 7.2 + 0.8 c 1.43 + 0.6 c

– – – – +f +f

aFor

+c – – –

larval and adult diapause; bDanilevsky 1961; cAlekseev 1990; dAlekseev 1998; eTushenko 1972; 1980a.

fGilbert

program works, however, remains completely unknown, except that it is reasonably certain that the daf gene pathways discussed hereafter in the nematode Caenorhabditis are somehow involved. 6.2

MOLECULAR MECHANISM OF DIAPAUSE IN NEMATODE CAENORHABDITIS ELEGANS

To date, few studies on the molecular processes involved in diapause as well as on genes coding diapause induction and termination have been performed in eutelic aquatic animals (but see Tunnecliffe 2005 for the role of a particular class of late

MOLECULAR GENETIC MECHANISM

117

embryogenesis abundant (LEA) proteins in chaperoning other proteins during anhydrobiosis). Terrestrial organisms like silk worm, flesh fly, and especially, the classic model of molecular studies, the soil nematode C. elegans, have been much better studied. Some review of recent studies on this latter species may be useful as a guideline for future diapause investigations in aquatic organisms. C. elegans, despite its apparent simplicity and body composed of only about 1,000 cells, possesses a well-organized endocrine system that regulates an early dichotomy of development: a choice for a larval diapause (dauer form) or for continuous, reproductive growth (Cassada & Russell 1975). The complicated chain of events begins with signals from the environment, which are somehow transduced to set in motion a genetic cascade. Environmental cues favoring diapause in C. elegans include high levels of a crowding pheromone, low food, and high temperatures (Golden and Riddle, 1984). Diapausing larvae arrested before sexual maturity shift metabolism to a lower level to maximize survival in harsh conditions. After diapause termination they resume development and start reproduction. Recent studies have suggested that in C. elegans gene expression changes between the third larval stage diapause and reproductive development. This choice is mediated by a cascade of genes and their gene products: daf-9, a cytochrome P450 gene related to steroidogenic hydroxylases and daf-12, a nuclear receptor gene encoding for lipophilic hormones that control the physiological status of an organism (Gerisch & Antebi 2004). A simple model is that the daf-9 gene produces a hormone regulating daf-12; its gene product bypasses diapause, promotes reproductive development and, perhaps, shortens life span. This hormone might be a sterol (Gerisch et al. 2001). Expressed in potential endocrine tissues daf-9 appears to control developmental decisions for the entire organism. Recent findings implicate daf-9 as a central point of developmental control, producing hormonal signals that regulate C. elegans life history (Gerisch & Antebi 2004). A choice between development and diapause is also regulated by insulin, some specific peptides (transforming growth factor b [TGFb]) and serotonergic signaling that control programs throughout the body (Finch & Ruvkun, 2001). Insulin and TGFb peptides are synthesized in response to stimuli, mainly from sensory neurons (Li et al. 2003). Alternatively, the complex of the daf-3 and daf-5 genes effects a shift to diapause in adverse environments when TGFb inactivate daf-3 and daf-5, allowing reproductive development (Georgi et al. 1990; da Graca et al. 2003). Along with controlling diapause, insulin also regulates somatic endurance and longevity across taxa (Tatar et al. 2003). Insulin-like agonists stimulate the daf-2, insulin-like receptor, initiating a kinase cascade that phosphorylates daf-16 forkhead transcription factor (Morris et al. 1996; Pierce et al. 2001; Gerisch & Antebi 2004). This results in cytoplasmic sequestration of daf-16, and as a consequence animals undergo reproductive growth and live short lives. In adverse environments, daf-16 enters the nucleus, promoting stress resistance, diapause, and longevity (Henderson and Johnson 2001; Lee et al. 2001; Lin et al. 2001; Gerisch & Antebi 2004).

This page intentionally blank

118

V. ALEKSEEV

There is evidence that both insulin and TGFb receptors convey signals through downstream secondary endocrines and that daf-2 regulates diapause and life span by systemic signals (Apfeld and Kenyon 1998; Wolkow et al. 2000). This confirms physiological observations in crustaceans and insects, as well as driving of diapause in these organisms by a two-step hormonal mechanism (Otsu 1963; Zaslavsky 1988). Gene daf-9, a cytochrome P450, resembles steroidogenic and fatty acid hydroxylases, as well as xenobiotic detoxifying enzymes (Gerisch et al. 2001; Jia et al. 2002). It probably produces a hormone for daf-12, a nuclear receptor transcription factor related to vitamin D, which is well known as a factor responsible for diapause in rotifers (Gilbert & Thompson 1968; Antebi et al. 2000). Much of this information has been collected by experiments with experimentally elicited mutants, some of which had lost the possibility to activate cytochrome P450 (daf-9) or, alternatively, the nuclear receptor responsible for lipophilic hormone production (daf-12). Gene daf-9 null mutants form diapausing larvae constitutively and have a life span ~25% longer than normal ones. Somewhat opposite, daf-12 null mutants fail to form diapausing larvae and live short lives (Gerisch et al. 2001). These observations establish a link between diapause and aging. To my mind, fundamental new insights into how genes and environment influence the metabolism, the development and the diapause of a variety of other metazoans will soon emerge from genetic studies similar to C. elegans or species-specific genes recently found in bdelloid rotifers (Tunnecliffe 2005). These studies will illuminate how endocrine networks integrate environmental cues and transduce them into adaptive life history choices. Acknowledgments. Prof. Henri Dumont is much appreciated for encouraging me to add this brief chapter as well as for editing it.

PART II

THE ROLE OF DIAPAUSE IN SCIENCE AND HUMAN USES

BART T. DE STASIO

7. EGG BANK FORMATION BY AQUATIC INVERTEBRATES A Bridge Across Disciplinary Boundaries

7.1 INTRODUCTION

Aquatic organisms live in variable environments. Changes in factors that affect survival and reproduction of individuals occur on various temporal and spatial scales, and can have both short- and long-term consequences (e.g. Brendonck et al. 1998). One life history adaptation to lessen the impact of such variability is to undergo dormancy, a period of suppressed development, during part of a lifetime (Tauber et al. 1986; Danks 1987; Hairston 1998). Similar to the seed banks of plants, the accumulation of dormant invertebrates in the aquatic environment has been termed the “egg bank” of the organism, or more generally, a biotic reservoir (De Stasio 1989; Hairston 1996; Brendonck & De Meester 2003). When dormancy persists over multiple seasons or generations, the egg bank is considered a mixed persistent egg bank, whereas shorter durations of dormancy result in transient egg banks that may persist less than a single year (Brendonck & De Meester 2003). Contributions of individuals from the egg bank to the active population will vary depending on local conditions and dynamics. Typically, only the top few centimeters of sediment contain individuals that hatch and contribute to the active population. This zone has been labeled the active egg bank (Cáceres & Hairston 1998), and the factors that determine the depth of the active egg bank can be important features of ecosystems. Recent extensive reviews (Brendonck & De Meester 2003; Gyllström & Hansson 2004) provide comprehensive overviews of the structure and dynamics of egg banks of freshwater organisms. This chapter provides an update of those reviews with respect to egg bank formation, and also includes information on egg banks in marine systems. As one of the common features of all aquatic environments, the study of egg banks provides a bridge across disciplinary boundaries and encourages a focus on processes that apply to all habitats. 7.2 DORMANCY PROCESSES

7.2.1 Dormancy Initiation Egg banks are the result of the balance of two general processes: the production of dormant individuals that persist through time and the removal of individuals from the dormant stage (Fig. 7.1). Dormancy may result from a wide range of mechanisms, with diapause and quiescence defining the extremes (Danks 1987). True diapause involves internal neurohormonal regulation of dormancy, with prediapause, refractory, and postdiapause phases (Marcus 1996). In many aquatic species dormant individuals result exclusively from the production of diapausing embryos. This type 121 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 121–133. © 2007 Springer.

122

B. DE STASIO

Figure 7.1. Egg bank general process model. The egg bank results from the balance of the main processes contributing to, and removing, dormant individuals from the environment.

of dormancy occurs in monogonant rotifers (Gilbert 1974; see also Chapter 2, this volume), branchiopods (Brendonck & De Meester 2003), and in the majority of both marine and freshwater calanoid copepods (Grice & Marcus 1981; Marcus 1996; Hairston & Cáceres 1996). Cyclopoid copepods most often enter diapause as copepodites (Elgmork 1985; Dahms 1995; Santer 1998; Frisch & Santer 2004), while some cyclopoid, harpacticoid, and calanoid copepods, as well as bdelloid rotifers, may enter dormancy as adults (Næss & Nilssen 1991; Williams-Howze 1996). For organisms that enter a true diapause, initiation of the developmental delay typically occurs in response to reliable external cues that occur in advance of harsh environmental conditions (Tauber et al. 1986; Danks 1987; Alekseev & Starobogatov 1996; Brendonck & De Meester 2003). Cues such as temperature, photoperiod, food concentration, and high concentration of metabolites released by conspecifics have all been shown to induce the production of dormant stages (Mortimer 1935; Marcus 1986; Hairston & Olds 1987; Stross 1987; Hairston et al. 1990; Kleiven et al. 1992; Hairston & Kearns 1995; Gilbert 2004a, b). Recent work has also shown that chemical cues released by fish may cause production of diapausing eggs by Daphnia (Slusarczyk 1995; 2004; Pijanowska & Stolpe 1996). Quiescence includes arrested development in response to external conditions, where development can resume quickly when the environment changes to more favorable conditions (Danks 1987; Brendonck 1996; Marcus 1996). A wide variety of mechanisms result in quiescence, leading to delays in development. This type of dormancy has been demonstrated for branchiopods (Brendonck 1996; Clegg & Jackson 1998), cyclopoids (Elgmork 1980), and marine calanoid copepods (Grice & Marcus 1981; Marcus 1996; Ohman et al. 1998; Katajisto 2004). Some species are also known to produce stages that cannot be classified as either quiescent or truly diapausing. In some species embryos will delay hatching longer than subitaneous (i.e. immediately hatching) eggs, but shorter than most diapausing eggs (Marcus 1996; Chen & Marcus 1997). In cyclopoid copepods and larvae of the phantom midge fly Chaoborus some individuals may enter an “active diapause” in which animals delay development but will still feed if prey are available (Bradshaw 1973b; Elgmork 1980). Other species have individuals that delay development as adults or employ a “reproductive-resting strategy” (Lonsdale et al. 1993; Ohman et al. 1998; Niehoff & Hirche 2005). Variation in intensity and duration of dormancy has been noted for a

EGG BANK FORMATION

123

range of organisms (Hairston & Cáceres 1996; Elgmork 1996), but direct tests of whether individuals are in a true diapause are not always included in such studies. 7.2.2 Release from Dormancy In comparison to additions to the dormant egg bank, losses from the biotic reservoir can occur by a wider range of processes (Fig. 7.1). The majority of investigations have focused on the processes whereby individuals are released from dormancy and contribute to the active population (Brendonck et al. 1998; Brendonck & De Meester 2003). Following the early fieldwork by Herzig (1985), De Stasio (1989; 1990), and Wolf and Carvalho (1989), a number of studies have documented release from dormancy in the field. Some have employed emergence traps placed on the surface of the sediment, but removal of sediment containing dormant stages followed by incubation in the laboratory has also been used to measure emergence (Marcus 1996; Brendonck & De Meester 2003; Gyllström & Hansson 2004). The majority of these studies have examined emergence of a single species or small group of species from dormancy. Notable exceptions are the investigations by De Stasio (1990) and Hairston et al. (2000). In these studies multiple species of diverse taxa were investigated. Emergence timing and rates were compared to population dynamics and used to determine the importance of dormancy in driving seasonal fluctuations of populations. In a multiple-year study of four crustacean zooplankton, De Stasio (1990) found that emergence dynamics were important for determining the first appearance in the plankton for one cladoceran (Ceriodaphnia reticulata) and two calanoid copepods (Diaptomus sanguineus and Epischura nordenskioldi). Emergence of another cladoceran (Eubosmina longispina) was sporadic during the year and did not correlate closely with plankton dynamics. Hairston et al. (2000) sampled traps during the spring and summer of 1 year in Oneida Lake, New York, and found that three cladoceran and three calanoid copepod species exhibited predominately spring emergence. For the rotifers, cladocerans, and calanoid copepods there was very little correspondence between emergence and seasonal dynamics. Emergence of cyclopoid copepods occurred at various times during the period sampled (May through August) and was highly correlated with plankton abundance and timing. As pointed out by Hairston et al., these few studies of dormancy dynamics and zooplankton communities suggest that emergence patterns may be similar in a variety of environments, but further field investigations are needed to test these ideas. 7.2.2.1 Additional emergence data. To test the ideas on generality of emergence dynamics, emergence patterns of additional taxa were determined in Bullhead Pond, Rhode Island, USA (De Stasio 1989; 1990). Detailed information on Bullhead Pond and sampling methodology has been provided previously (Hairston et al. 1983; Hairston & De Stasio 1988; De Stasio 1990). Emergence patterns of zooplankton from the pond sediments were monitored with plastic, inverted funnel traps placed at each of eight locations situated along two transects encompassing the full range of depths in the pond throughout the year (De Stasio 1989; 1990). A total of 19 different microcrustacean and one rotifer taxa were identified in the emergence

124

B. DE STASIO

samples. Of these taxa, data on hatching patterns of D. sanguineus, E. nordenskioldi, C. reticulata, and E. longispina have been presented elsewhere (De Stasio 1989; 1990). Here I focus on ten additonal groups: six additional cladoceran taxa (Chydorus, Diaphanosoma, Latona, Scapholeberis, Sida, and Simocephalus); two groups of copepods including the harpacticoids (all species combined) and cyclopoids (three species combined, but Diacyclops thomasii primarily); ostracods (at least two species combined); and one rotifer (Keratella). Data presented are mean daily emergence rates for the entire pond for each taxon over the course of the study so that differences across depths and sample locations are included in the analysis. To obtain total pond emergence rates, emergence estimates for each depth region were weighted by the area of sediment at that depth and then summed across depth regions to obtain total daily emergence rates. Hatching for most of the cladoceran taxa was generally restricted to only a particular period of the year. Simocephalus exhibited a bimodal pattern, emerging primarily during June and then from late August to early September (Fig. 7.2). It was only recovered from traps placed in the 3 m depth areas, whereas all the other taxa were collected from each of the four depth regions. Diaphanosoma was collected more widely across the basin, but only during a very short time period in late May

Figure 7.2. Entire pond emergence rates in Bullhead Pond, Rhode Island, USA, during 1984–1985 for cladocerans with a restricted pattern of hatching. Shaded bars indicate periods when traps were not sampled, but remained in place. Note changes in vertical axis scales.

EGG BANK FORMATION

125

Figure 7.3. Entire pond emergence rates in Bullhead Pond, Rhode Island, USA, during 1984–1985 for ostracods and cladocerans with prolonged hatching. Shaded bars indicate periods when traps were not sampled, but remained in place. Note changes in vertical axis scales.

and early June. Hatching for Sida was more polymodal, beginning in April and extending through October 1985 (Fig. 7.2). Both Scapholeberis and Latona exhibited a polymodal pattern of emergence, beginning in April and continuing into December (Fig. 7.3). Timing of peak emergence rates was approximately the same for each of these two groups, with highest hatching occurring in June, August, and November. Of the five cladoceran taxa, emergence rates for Scapholeberis were 2–20 times greater than those estimated for the other groups with peaks of more than 0.2 × 106 animals emerging into the pond per day. The ostracod hatching pattern was also polymodal but, in contrast to most of the cladocerans, hatching was restricted exclusively to winter and spring (Fig. 7.4). In addition, peak emergence rates were up to ten times greater than those of the cladocerans (except for Chydorus). In contrast to the other five cladoceran taxa, Chydorus exhibited a pattern of continuous hatching throughout the year (Fig. 7.4). Highest emergence rates occurred in May and October, with peak rates of nearly 2.0 × 106 animals per day. Harpacticoid copepods emerged only during the period from winter through early summer (midJune), with highest emergence rates in spring. In contrast to this pattern, cyclopoids

126

B. DE STASIO

Figure 7.4. Entire pond emergence rates in Bullhead Pond, Rhode Island, USA, during 1984–1985 for harpacticoid and cyclopoid copepods (top two panels) and for Chydorus (bottom panel). Note changes in vertical axis scales.

hatched more or less continuously throughout the study period (Fig. 7.4). The greatest emergence of cyclopoids occurred during the first 3 weeks of the study (December 1984), and the next greatest rates occurred in spring. Keratella, the one rotifer group monitored, showed emergence primarily in late April and May, with only a few individuals hatching in November. Emergence data for the nine taxa presented here show that the cladocerans exhibit a diversity of temporal hatching patterns, ranging from a short pulse of hatching by Diaphanosoma during May and June to continuous hatching by Chydorus. All of the groups, except Simocephalus, were found to hatch from all possible depths. Simocephalus only emerged from the 3 m depth area of the pond. Data on hatching rates of two other cladocerans in this pond also demonstrate the diversity of patterns observed for cladocerans. De Stasio (1990) showed that E. longispina continuously hatched in a sporadic manner, while C. reticulata hatched from the sediments in a bimodal pattern primarily in spring and fall, but with additional emergence occurring at other times. A similar pattern for Ceriodaphnia was found by Arnott and

EGG BANK FORMATION

127

Yan (2002). In contrast to these patterns, studies of Daphnia dormancy have demonstrated that hatching occurs almost exclusively in spring (Wolf & Carvalho 1989; Cáceres 1998; Hairston et al. 2000; Gyllström 2004). Other studies employing a wider variety of techniques have also shown that noncopepod zooplankton exhibit a diverse array of hatching patterns. In a review of the data for branchiopods, Brendonck (1996) indicates that a wide variety of hatching responses occur, with many population-specific cues involved in determining the actual hatching pattern. Wiggins et al. (1980) and Taylor and Mahoney (1990) have clearly shown differences among cladocerans in hatching phenology, with important consequences for the seasonal succession of zooplankton and community ecology of ponds. For ostracods, less is generally known about hatching characteristics, but the available data indicate similarly diverse patterns. McLay (1978) has demonstrated variability in population ecology and hatching patterns among populations of ostracods, while Rossi et al. (1991, 1996) have also shown within-population differences among clones for life history strategies, including emergence from dormancy. Available data for copepods indicate that hatching patterns are generally more synchronized and seasonal than those obtained for noncopepod groups. Data on the harpacticoids (Fig. 7.3) and for the calanoids in Bullhead Pond (De Stasio 1989, 1990) demonstrate that hatching occurs over a 5- to 6-month period, with highest emergence rates in spring (March through May). A similar pattern has been reported in numerous other studies, although the timing of peak emergence varies by location usually. For freshwater calanoids, studies by Cooley (1971), Taylor and Mahoney (1990), Taylor et al. (1990), and Walton (1985) have all documented highly restricted times of emergence. The few studies on marine calanoids also show similar patterns of termination of dormancy (e.g. Katajisto 1996; Marcus 1996). Data for harpacticoids are sparse, but for both freshwater (Sarvala 1979b) and marine species (Williams-Howze 1996) hatching probably occurs over a relatively short time period. The harpacticoid copepods in Bullhead Pond exhibited hatching only during part of the year, with highest emergence in spring. For cyclopoids, a variety of studies have shown variable timing of emergence from dormancy, but usually all show hatching occurring over a relatively short time period (e.g. Elgmork 1980; Maier 1989; Taylor et al. 1990; Hansen & Jeppesen 1992; Hairston et al. 2000). Given these findings, it is probably reasonable to assume that the continuous emergence of cyclopoids in Bullhead Pond resulted from combining the emergence dynamics of multiple species, since at least three species were collected in trap samples, but were not enumerated separately. As shown by Hairston et al. (2000) cyclopoid emergence from dormancy occurred at various times of the year. Combining emergence data for all cyclopoids in Oneida Lake results in nearly continuous hatching from May through November, a pattern similar to that found in Bullhead Pond (Fig. 7.4). Several other studies have examined hatching in the laboratory in response to environmental stimuli. Many investigations have demonstrated the importance of photoperiod and temperature as external factors affecting the release from diapause by both freshwater and marine crustaceans (e.g. Stross 1966; Marcus 1986; Pfrender &

128

B. DE STASIO

Deng 1998). However, some studies have shown that individuals will terminate diapause without any obvious external cues (Hirche 1983; Elgmork 1973; Williams-Howze & Coull 1992; Alekseev 1998; Elgmork and Lie 1998; De Stasio 2004). Work on Daphnia and fairy shrimp indicates strong maternal effects on the release from diapause. For instance, De Meester et al. (1998) found an important maternal effect on hatching of Daphnia magna diapausing eggs, and De Meester and De Jager (1993a, b) also found family-dependent effects on hatching rates of this same species. Similarly, Van Dooren and Brendonck (1998) found a significant maternal effect related to the age of the mother on hatching success of cysts of the fairy shrimp Branchipodopsis wolfi. However, De Stasio (2004) demonstrated high within-clutch variability in release from diapause for the obligately sexually reproducing calanoid copepod Onychodiaptomus sanguineus, indicating a lack of maternal effects. Studies in which annual addition of dormant stages to the sediments does not occur or where it is prohibited by experimental techniques indicate that the relative importance of contributions to the egg bank varies among taxa. Hairston and De Stasio (1988) demonstrated that the calanoid copepod D. sanguineus continued to emerge from the egg bank for 3 years even following multiple years of failed production of diapausing eggs by the species. De Stasio (1989) maintained inverted funnel traps in place for more than 3 years and observed continued emergence of D. sanguineus nauplii during the entire study. He estimated that the active egg bank contained sufficient diapausing eggs to fuel continued emergence for more than 10 years without new contributions to the egg bank. In the same study, similar results were observed for the calanoid copepod E. nordenskioldi and the cladoceran E. longispina (De Stasio 1990). In contrast to this pattern, emergence by the cladoceran C. reticulata declined and ceased near the end of the study. A similar pattern was observed for ostracod emergence (see above). All other taxa emerged during each of the 3 years of the study, providing evidence that each taxa maintained a mixed persistent egg bank and that annual additions were not required to permit continued emergence in the future. 7.2.3 Predation and Infection of Dormant Stages Predation has been shown to have important consequences for egg banks, but compared to release from dormancy, very few studies have focused on losses from the egg bank due to predation or infection. Parker et al. (1996) demonstrated that predation by the amphipod Gammarus can have important consequences for the egg bank for the calanoid copepod Hesperodiaptomus arcticus. Similarly, Cáceres and Hairston (1998) tested predation on Daphnia ephippia from Oneida Lake by a variety of natural invertebrate benthic predators. The amphipod Gammarus consumed ephippia in laboratory experiments and diet analysis of individuals from the field confirmed predation in the lake sediments. In the same study there was no predation by the gastropod Physa, insect Chironomus, turbellarian Dugesia, or zebra mussel Dreissena polymorpha. A laboratory study of simulated cores of marine sediments did not exhibit any predation by polychaete worms on copepod diapausing eggs (Marcus & Schmidt-Gengenbach 1986). There do not seem to be any studies focusing on the importance of bacterial or fungal infections of dormant stages. Infection of

EGG BANK FORMATION

129

diapausing eggs in the laboratory is common (De Stasio 1989; 2004, personal observation), but in situ estimates are necessary to determine the relative importance of this loss process. Although these studies indicate the potential importance of predation and infection in affected egg banks, no studies have estimated the relative loss of eggs by these processes compared to release from dormancy, burial, or senescence. 7.2.4 Deep Burial of Dormant Stages Another loss process for dormant stages is the burial of individuals in locations beyond which the newly emerged individual can swim into the water column (Cáceres & Hairston 1998; Gyllström & Hansson 2004). This will normally be caused by deep burial, but may also include transport to locations where conditions do not permit release from dormancy (Grice & Marcus 1981). In lakes the process of sediment focusing moves many dormant eggs to deeper locations with less intense environmental cues such as temperature and photoperiod (De Stasio 1989; Arnott & Yan 2002; Hairston & Kearns 2002; Cáceres & Tessier 2003; Gyllström & Hansson 2004). Eggs in marine systems also may be deposited or moved to areas where environmental conditions suppress hatching for both subitaneous eggs and diapausing eggs in the refractory phase (Marcus 1996). Conditions such as low oxygen and low temperature are known to delay development and inhibit hatching of eggs (Johnson 1967; Kasahara et al. 1975a; Lutz et al. 1992; Katajisto 2004). Studies using eggs and colored polystyrene beads in both freshwater and marine systems have shown that burial and sediment focusing are important determinants of dormant egg distribution. In marine sediments benthic invertebrates such as polychaete worms cause bioturbation, which will redistribute resting eggs (Marcus & Schmidt-Gengenbach 1986). Sediment resuspension and the size and density of particles relative to size and density of egg will be important determinants of egg depth distribution (Marcus & Taulbee 1992). Similar processes act on lake sediments, but in freshwater systems the actions of wind-driven resuspension in shallow lakes and nearshore habitats, and bioturbation of sediments by fish feeding and nesting may be most important (Cáceres & Hairston 1998; Hairston & Kearns 2002; Kerfoot 2004). Studies examining the abundance of dormant stages in core samples taken from lakes and coastal areas of the oceans demonstrate nonuniform distributions of eggs in the sediments. Dormant egg abundance generally decreases with depth in cores from freshwater lakes and ponds, suggesting that as eggs are buried they lose viability and die (Brendonck & De Meester 2003). In marine environments, abundance of dormant eggs has often been shown to increase with depth before decreasing in deeper layers (Viitasalo 1992; Marcus et al. 1994; Marcus 1998). It has been hypothesized that many marine sediments are a 3-tier system: surface sediments have low abundance due to hatching of dormant eggs; the layer beneath has higher abundance because conditions inhibit hatching but are sufficient for survival of eggs; and the lower depths have decreased abundance because of dormant egg mortality due to adverse conditions (Viitasalo 1992). A similar pattern of a mid-depth peak in dormant egg abundance has also been found in some lake sediments, but in those cases the reasons have presumably been due to changes over time in production of dormant

130

B. DE STASIO

eggs as environmental conditions such as predation pressure or salinity changed in the lakes (Cousyn & De Meester 1998; Hairston et al. 1999b). 7.2.5 Senescence and Egg Viability Natural senescence is a difficult process to quantify in the field. Incubation of eggs in mesh bags can give estimates of senescence but rates of senescence are usually not reported because the focus is often on hatching rates, not mortality rates (Cáceres & Tessier 2003). Senescence in laboratory experiments has been estimated to range from 50% to 100%. However, the artificial conditions in laboratory experiments raise concern about the applicability of such estimates to natural situations (Cáceres & Schwalbach 2001; De Stasio 2004). Regression estimates of changes in egg age with depth in sediment cores indicate a field mortality rate of 1.1–1.5% per year for the calanoid copepod D. sanguineus (Hairston et al. 1995). Another indicator of senescence can be obtained from viability of dormant stages in sediments. Analyses of viability of eggs in sediments reveal that typically eggs found in deeper sediments are less viable than those in shallower layers, indicating that deep burial is an important loss process from the egg bank (Brendonck & De Meester 2003; Gyllström & Hansson 2004). However, many species have diapausing eggs with the potential to survive for extremely long periods of time. Diapausing eggs of a freshwater calanoid copepod have been recovered and hatched from undisturbed sediments that were 330 years old (Hairston et al. 1995). Reliable estimates of ages of cladoceran ephippia are typically 50 years or less, but Cáceres (1998) successfully hatched Daphnia ephippia from sediments that were dated as 125 years old. Marine copepod eggs collected from sediments typically survive in dormancy for 1–2 years, but studies in nearshore habitats have demonstrated hatching from sediments as old as 40 years (Marcus et al. 1994; Viitasalo & Katajisto 1994; Katajisto 1996). 7.3 EGG BANK SIZE AND DYNAMICS

Resting stage densities in lake and marine sediments typically range from 103 to 105 individuals per square metre, but densities as high as 106–107/m2 are not uncommon (Nipkow 1961; Marcus 1996; Hairston 1996; Brendonck & De Meester 2003). The top 2 cm of lake sediments are often considered the active egg bank, but eggs in the top 10 cm of many lakes should probably be considered possible recruits to the plankton (Cáceres & Hairston 1998; Hairston et al. 2000). In marine habitats eggs below 2 cm do not typically hatch (Kasahara et al. 1975b), but eggs in deeper layers are viable, and mixing processes and bioturbation can move eggs nearer the surface (Marcus & Schmidt-Gengenbach 1986; Marcus et al. 1994). Changes in egg bank abundance over time have been monitored in a number of lake and coastal habitats. Early work in both marine and freshwater systems demonstrated that changes in dormant stage densities in the sediments reflected fluctuations in the abundance of the planktonic population and production of dormant individuals (Elgmork 1980; 1985; Kasahara et al. 1975b; Kankaala 1983; Uye 1983). By definition, egg banks contain dormant individuals that persist through time, but

EGG BANK FORMATION

131

some studies have shown wide fluctuations in the abundance of dormant individuals over time. Sediment abundance of resting eggs of the marine cladoceran Bosmina longispina maritima fluctuated by a factor of approximately 6 over the course of 1 year in the Baltic Sea (Kankaala 1983), while densities in the sediments of the Inland Sea of Japan of calanoid copepod eggs varied by more than a factor of 10 (Kasahara et al. 1975b). Brown and Branstrator (2005) found that resting eggs of the freshwater cladoceran Bythotrephes fluctuated over an annual cycle by a factor of 6–10. Ephippia of the freshwater cladoceran Daphnia found in sediments of Lake Winnebago, Wisconsin, USA, also vary widely over time (Fig. 7.5). Summer densities are lowest, averaging approximately 1–2 × 104/m2, while winter abundance can reach as high as 20–25 × 104/m2. Lake Winnebago is a large, shallow lake which mixes completely all summer (surface area: 557 km2, maximum depth: 6 m, mean depth: 4.7 m). The top 10–20 cm of the sediment is apparently mixed due to winddriven turbulence and feeding activities of fish such as freshwater drum (Aplodinotus grunniens) and lake sturgeon (Acipenser fulvescens) (Gustin 1995). The vertical profile of intact ephippia in the sediments exhibits a steplike decrease in ephippia density below 15–20 cm (Fig. 7.6). This suggests that the active egg bank in this lake likely includes at least the top 15 cm, but ephippia also persist in deeper layers. Seasonal fluctuations such as those documented here suggest that caution should be taken when interpreting egg bank abundance estimates obtained at only one time of the year.

7.4 CREATING AN EGG BANK

The colonization of habitats in both time and space has been suggested as one of the possible selective forces that drives the origin of dormancy as a successful adaptation (Alekseev & Starobogatov 1996; Hairston 1998). The establishment of an egg

Figure 7.5. Abundance of Daphnia ephippia in the sediments of Lake Winnebago, Wisconsin, USA. Samples were collected with an Ekman grab sampler from the surface sediments. Error bars indicate 1 standard error of the mean.

132

B. DE STASIO

Figure 7.6. Abundance of Daphnia ephippia in sediment cores from Lake Winnebago, Wisconsin, USA, on June 22, 1993. Abundance represents mean values at three locations along the north–south axis of the lake. Error bars indicate ±1 standard error of the mean.

bank in a habitat should therefore be an important event for organisms facing an uncertain, variable, or new environment. The creation of a new egg bank has been documented in only a few studies to date. Coulas et al. (1998) found a rapid increase of diapausing eggs of Bythotrephes in sediments of Harp Lake, Ontario, Canada, during the first few years following invasion. Similarly, in a study of new and isolated ponds Vandekerkhove et al. (2005e) determined that dormant egg banks were established in the first year of existence for all but one of 24 pools studied. Mean densities of 104 eggs/m2 were observed in the ponds, mainly from cladocerans like Daphnia, Chydorus, and Simocephalus. The ability to reproduce asexually should permit organisms like cladocerans to create new egg banks more quickly than is possible for sexually obligate species such as calanoid copepods. This difference helps to explain the differential abilities among these groups to invade new areas as well as success in recolonization of habitats following local extinctions of the planktonic populations (De Stasio 1990; Panov et al. 2004; Sarnelle & Knapp 2004). Strong selective pressure to establish a new egg bank rapidly should exist for organisms that disperse widely and invade new habitats. Increased production of resting eggs by aquatic invasive species in new locations compared with their native ranges has been documented for cladocerans such as Bythotrephes and Cercopagis (MacIsaac et al. 1999; Panov et al. 2004). This extension of the duration of resting egg production should increase the chances of successful colonization by seeding the sediments with dormant stages quickly. However, the relatively short duration of diapause in some invasive species like Bythotrephes can lead to failed invasions (Brown and Branstrator 2005; D. Branstrator, 2006 personal communication).

This page intentionally blank

EGG BANK FORMATION

133

7.5 CONCLUSIONS

The majority of studies to date indicate that the formation of a persistent egg bank is a key factor in the success of many aquatic invertebrate taxa. Differences in the creation and persistence of an egg bank appear to occur between organisms with asexual or cyclic and organisms with obligately sexual reproduction (De Stasio 1990; Hairston et al. 2000; Sarnelle & Knapp 2004). Regardless of these differences, taxa with egg banks have been shown to survive through an amazing array of harsh environmental conditions such as complete drying of habitats (Hairston & Olds 1987; Brendonck et al. 1998), local extirpation by acidification (Arnott & Yan 2002), elimination by predation or competition (Parker et al. 1996; Cáceres 1997), changes in salinity (Hairston et al. 1999b), and ecological changes due to eutrophication (Hairston et al. 2001). Egg banks also have important consequences for the evolutionary dynamics of aquatic systems (Hairston & De Stasio 1988; Hairston 1998). In addition, studies of egg banks have led to new approaches to investigating the environment, such as the new field of “resurrection ecology” (Jeppesen et al. 2001a, b; Kerfoot & Weider 2004). Clearly, future investigations of dormancy and egg banks will broaden and deepen our understanding of the ecology and evolution of organisms living in aquatic environments. Acknowledgments. My thanks go to Elizabeth De Stasio for helpful comments on the manuscript, and to Victor Alekseev for assistance during various stages of this project. Funding was provided through a grant from the Lois Almon Fund of the Wisconsin Academy of Science, Arts and Letters, and by the Excellence in Science Fund from Lawrence University.

SUSANNE L. AMSINCK, ERIK JEPPESEN, AND DIRK VERSCHUREN

8. USE OF CLADOCERAN RESTING EGGS TO TRACE CLIMATE-DRIVEN AND ANTHROPOGENIC CHANGES IN AQUATIC ECOSYSTEMS

8.1 INTRODUCTION

Analyses of cladoceran fossils preserved in lake sediments have proven powerful to elucidate natural and anthropogenic changes in lake ecosystems, such as climate variability, ecological succession, acidification, eutrophication, fish stocking, and toxin pollution. So far, most such studies have been conducted in European and North American freshwater lakes (e.g. Whiteside 1970; Kitchell & Kitchell 1980; Hann et al. 1994; Hofmann 1996; Jeppesen et al. 1996; Verschuren & Marnell 1997; Kerfoot et al. 1999; Pollard et al. 2003), whereas brackish and saline north temperate lakes (Bos et al. 1996, 1999; Amsinck et al. 2003b, 2005a,b) and lakes in tropical and subtropical regions (Goulden 1966; Verschuren et al. 1999a,b, 2000) have been studied less intensively. Recent reviews of conducted studies and their results have been elaborated by Jeppesen et al. (2001b) and Korhola and Rautio (2001). As in other fields of biological paleolimnology, early studies mostly employed a qualitative indicator-species approach or focused on patterns of species diversity through time. Yet, in the last decade, application of multivariate statistical techniques (Birks 1998) to large modern reference data sets of cladoceran remains extracted from surface sediments has produced various mathematical transfer functions to quantitatively infer past changes in temperature (e.g. Lotter et al. 1997), chemical conditions (e.g. total phosphorous [TP]: Brodersen et al. 1998; salinity: Bos et al. 1999; Bos & Cumming 2003), and important biological controlling parameters (e.g. planktivorous fish abundance: Jeppesen et al. 1996) from fossil cladoceran assemblages. This has considerably strengthened cladoceran paleoecological research and opened new avenues of investigation. The majority of studies conducted so far have been based on morphological identification and counting of various exoskeleton remains (e.g. carapaces, head shields, mandibles, postabdomens, and postabdominal claws). Resting eggs and ephippia (the resistant structure encasing the resting eggs; Fig. 8.1), mainly from daphnid Cladocera, have also been studied (see review by Brendonck & De Meester 2003), but not as extensively. This is partly due to incomplete knowledge of resting egg and ephippium morphology at the species level and the fact that their concentration in lake sediments is generally lower than that of the other exoskeleton remains. In addition, past cladoceran community structure may be more accurately determined from a complete analysis of exoskeleton remains instead of solely fossil ephippia, since not all cladocerans produce ephippia, and ephippia production varies among species (Jankowski & Straile 2003) and habitat types (e.g. low production in warm lakes: Jeppesen et al. 2003a,b). Comprehensive identification keys for the ephippia of 135 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 135–157. © 2007 Springer.

136

S. AMSINCK ET AL. Characteristics: • 1 egg • 0.27–0.38 mm • spine at lower side of caudal margin Daphnia magna (22)

Bosmina longirostris (12)

Characteristics: • 1 egg • 0.45 mm • spine at upper side of caudal margin

Characteristics: • 2 eggs • 1.03–1.59 mm • eggs more or less horizontal • spine at anterior and posterior side Characteristics: • 2 eggs • 0.45–1.37 mm • eggs vertical • narrowing sharply at posteroventral side

Daphnia pulex (11)

Bosmina coregoni (1)

Characteristics: • 2 eggs • 0.36–0.67 mm • eggs vertical • more or less symmetrical

Characteristics: • 1 egg • 0.30–0.45 mm • egg in dorsal part • often with floating cells • symmetrical Ceriodaphnia reticulata (6)

Characteristics: • 1 egg • 0.46–0.60 mm • transparant • narrowing sharply at posteroventral side

Daphnia parvula (70)

Characteristics: • 1 egg • 0.66–0.93 mm • height < 1:2

Camptocercus rectirostris (2)

Acropherus harpae (24)

Characteristics: • 1 egg • 0.37–0.51 mm • egg in dorsal part

Oxyurella tenuicaudis (13)

Characteristics: • 1 egg • 0.66–0.89 mm • narrowing sharply at posteroventral side Simocephalus vetulus (3)

Figure 8.1. Variation in morphology of cladoceran ephippia, namely the ephippium size and shape; the curvature of the dorsal ridge; the number, position, and orientation of the enclosed resting eggs; and the number and position of dorsal spines. (Photographs reproduced from Vandekerkhove et al. 2004a. With permission.)

Central European (e.g. Flössner 2000) and East African (Mergeay et al. 2005a) Cladocera, supported by controlled rearing and direct identification of the cladoceran species (Vandekerkhove et al. 2004a,b, 2005a–d), have increased the potential of using resting eggs and ephippia in paleoecological studies. Further, in recent years also molecular genetic methods (e.g. microsatellites and polymerase chain reaction (PCR)-based DNA recovery of mitochondrial DNA) have been applied successfully

TRACING OF ENVIRONMENTAL CHANGES

137

to fossil egg banks and thus permitted tracking of long-term shifts in the genetic structure (Reid et al. 2000; Limburg & Weider 2002; Pollard et al. 2003) and behavioral traits (Hairston et al. 1999a; Cousyn et al. 2001) of old zooplankton populations. These methods provide a new exciting potential to decipher the links between shifts in the genetic structure of past cladoceran communities and past changes in their aquatic habitat triggered by either anthropogenic or natural impacts. 8.2 TRACING ACIDIFICATION

Starting with the Industrial Revolution and during much of the 20th century, many north temperate lakes experienced severe acidification, caused by the acid rain falling downwind from industrial emissions of sulfur dioxide (SO2) and nitrogen oxides (NOx). This acidification led to general ecological deterioration, typically characterized by loss of species diversity and important keystone taxa such as fish in large zooplankton. Since the 1980s industrial sulfur emission to the atmosphere has been reduced markedly, and to speed up the recovery of acidified lakes liming has often been conducted directly in the lake or the surrounding catchments. Paleolimnological studies based chiefly on the stratigraphy of pH-sensitive diatoms and chrysophytes have been instrumental to first demonstrate the link between widespread lake acidification and acid rain (e.g. Whitehead et al. 1990; Battarbee et al. 1990; Cumming et al. 1992) and next to monitor the recovery of acidified lakes (e.g. Dixit et al. 1992; reviewed in Smol 2002). In these studies, Cladocera have proven to be a useful alternative to diatoms and chrysophytes as independent proxy indicators of aquatic ecosystem response to acidification. This is because cladoceran species also have distinct physiological tolerances to pH and are highly sensitive to acidification-driven changes in fish and invertebrate predation. Krause-Dellin and Steinberg (1986) examined cladoceran remains in the surface sediments of 26 soft-water lakes in Germany, and they noted significant reduction in species number and diversity associated with low surface-water pH. Many species showed distinct peak (relative) abundance at certain pH levels, which allowed their separation into distinct pH classes for development of pH inference models. The resulting models were then applied to 137Cs-dated sediment cores from three lakes (Pinnsee, Herrenwiesersee, and Grosser Arbersee). The fossil chydorid assemblages in Pinnsee and Herrenwiesersee showed marked and steady pH declines in the 20th century, corresponding well with diatom-inferred pH trends. In Grosser Arbersee, chydorid assemblages detected only a weak and very recent acidification trend, whereas fossil diatoms indicated that pH had been declining since the 1950s. Krause-Dellin and Steinberg (1986) do not specify which exoskeleton remains were counted, hence it remains uncertain if ephippia were included or not. Still, this study is an early demonstration of the potential of using Cladocera as biological pH indicators. Pollard et al. (2003) used qualitative paleoecological inferences based on the stratigraphic distribution of fossil Daphnia ephippia, in part identified by genetic techniques (PCR, single-strand conformation polymorphism [SSCP], and sequencing), to

138

S. AMSINCK ET AL.

8.0

40000 35000 30000

Mean ephippial deposition

7.5

Diatom inferred pH

7.0

25000

6.5

20000

6.0

15000

5.5

10000

5.0

5000

4.5

0 1300

pH

Ephippial accumulation rate (no. yr −1 m−2)

reconstruct five centuries of Daphnia community structure in Hannah Lake, Ontario, Canada. Located close (4.3 km) to the metal smelter industry in Copper Cliff near Sudbury, this was one of the lakes in the Sudbury region most extremely impacted by anthropogenic acidification. Following large area reductions in sulfur emissions and liming of its watershed, Hannah Lake has now begun to recover, although heavymetal concentrations remain high compared to unpolluted lakes. From about 1400 AD to the late 19th century, the resting egg record showed a stable Daphnia community with numerous species of which D. pulicaria was the most prominent (Figs. 8.2 and 8.3). In the late 1800s, the accumulation of Daphnia resting eggs began to decline, reaching its lowest value around 1950, concurrent with the peak in Sudbury smelter activities. Simultaneously, diatom-inferred pH declined almost 2 pH units (Fig. 8.2). This indicated that conditions exceeded Daphnia’s physiological tolerance to acidity; moreover, they had become too harsh to allow local survival of Daphnia through diapausing stages. When liming caused lake water pH to increase above preindustrial levels, egg accumulation rates also began to slowly increase. Concurrently, the species composition of the Daphnia assemblage shifted markedly towards dominance of D. mendotae, which had been absent in the sediment record for the previous 250 years. As daphnid diapausing eggs seldom remain viable for more than 100 years (Weider et al. 1997; Cáceres 1998), this modern D. mendotae population is most likely the result of dispersal from other lakes nearby. Additional genetic analyses indeed showed that the clonal richness and allozyme allelic diversity of the current acid-sensitive D. mendotae population in Hannah Lake are similar to those in reference lakes located beyond the influence of Sudbury smelters. Sarmaja-Korjonen (2002, 2003) used changes in the fossil stratigraphy of parthenogenetic and gametogenetic chydorid populations (the former represented by headshields,

4.0

1500

1700

1900

2100

Date (years) Figure 8.2. Historical mean accumulation rate of sedimentary Daphnia ephippia and diatom-inferred pH during the last approximately six centuries in Hannah Lake. (Reproduced from Pollard et al. 2003. With permission.)

139

TRACING OF ENVIRONMENTAL CHANGES 1936

D. pulicaria

1909 1886

Unknown

Midpoint sediment date

1862 1835 1807

D. ambigua

1745 D. longiremis

1710 1692 1572 1487 1440

D. mendotae

1391 0

20

40

1,0

06

80

100

Percent species composition Figure 8.3. Changes in Daphnia community composition from the late 14th to mid-20th century in Hannah Lake, Ontario, Canada, inferred by genetic analysis of fossil ephippia. (Reproduced from Pollard et al. 2003. With permission.)

carapaces, and postabdomens; the latter mainly by ephippia, but also male headshields and postabdomens, and the headshields of ephippial females) to describe the environmental history of Lake Kaksoislammi, Finland, during the Holocene. Some 1,700–1,800 years ago, ephippia abundance of all local chydorid species increased sharply, which the author interpreted as gametogenesis triggered by a common environmental stress. There were also notable changes in species community structure, especially among the pelagic cladocerans. The small-bodied Bosmina longirostris almost disappeared, whereas large-bodied Daphnia and the invertebrate predator Chaoborus (represented by its mandibles) appeared in greater abundance. As zooplanktivorous fish usually show a preference for large-bodied Daphnia, and as invertebrate predators prefer small- to medium-sized cladocerans (e.g. Brooks & Dodson 1965; Hanazato 1990), these faunal changes were interpreted to reflect the changed predation risk associated with a decline in planktivorous fish abundance. As revealed by a diatom-inferred pH decline of 1–1.5 units during the same period, it appears that the local fish community suffered from lake acidification. SarmajaKorjonen (2002, 2003) attributed this preindustrial acidification to disturbance of the

140

S. AMSINCK ET AL.

surrounding acidic peatland by Iron Age farmers, an interpretation consistent with fossil pollen evidence for past changes in terrestrial vegetation. This showed, almost simultaneously with the faunal changes, increases in both Cerealia and various herbs known to reflect anthropogenic disturbance (e.g. Plantago lanceolata, Brassicaceae, Centaurea cyanus). However, judging by the fossil diatom record, this catchment disturbance did not markedly increase Lake Kaksoislammi’s trophic status. Prazákova and Fott (1994) analyzed the stratigraphic distribution of ephippia and other cladoceran remains (headshields, carapaces, and postabdomens) in the sediment record of Lake Cerne, Czech Republic, in which documented changes in local cladoceran species have occured. Lake Cerne has experienced recent acidification and high aluminium toxicity. In consequence, it is currently fishless and the pelagic crustacean zooplankton community is characterized by low species diversity (Faustová et al. 2004). The stratigraphy of cladoceran remains in the upper 18 cm of an undated sediment core revealed successive disappearance of Bosmina longispina, Daphnia longispina, and Ceriodaphnia quadrangula. This is in agreement with documentary records of species decline that had been attributed to the recent acidification (2.2–2.7 pH units since 1936; Fott et al. 1994) due to acid rain brought by industrial sulfur emissions. Lack of radiometric dating of the sediment record evidently precluded validation of this presumed link. After 10 years, Faustová et al. (2004) investigated the egg bank of Daphnia gr. longispina in Lake Cerne and three other Czech acidified lakes (Certovo, Plesné, and Prasilsé) to assess whether future amelioration of environmental conditions could result in autochthonous recovery of the Daphnia population. The state of preservation and the taxonomy of the Daphnia resting eggs were examined by hatching experiments and DNA amplification, respectively. However, no viable resting eggs were found in any of the lakes (none of the eggs hatched) and most were partly decomposed. Thus, future recovery of Daphnia populations from the local egg banks will be unlikely. By contrast, Arnott and Yan (2002) found high emergence of crustacean zooplankton species (cladocerans, copepods) from autochthonous sources in Swan Lake, Sudbury, Canada. Long-term zooplankton monitoring indicated that crustacean species richness had temporarily increased in the year following a 2-year drought and subsequent reacidification (due to climate change, not anthropogenic pollution). To test if this unexpected peak in species richness resulted from mass emergence of zooplankton from the egg bank, triggered by the changes in lake conditions after combined drought and reacidification, the authors set up in situ emergence traps under different regimes of desiccation, light, temperature, and oxygen concentration. Emergence of individual zooplankton species among treatments was distinct, with four, six, and three taxa mainly responding to desiccation, temperature, and light, respectively. However, as the emerging zooplankton faced rather inhospitable habitat conditions, many populations failed to persist and zooplankton species richness declined. Consequently, this triggering of mass emergence may have acted to deplete the egg bank and thus reduce the number of future local colonists available to repopulate the lake when conditions would become favorable.

TRACING OF ENVIRONMENTAL CHANGES

141

This study demonstrates the value of zooplankton resting eggs to increase our insight into zooplankton community structure and probability of recovery after local extinction. 8.3 TRACING EUTROPHICATION

Over the past few centuries, and sometimes for a much longer time, the waste production of a growing human population (e.g. raw discharge of sewage and wastewater from households and industrial plants) and the intensification of agriculture (e.g. chemical fertilizers and large-scale livestock holdings) have caused excess nutrient loading to many lakes. In consequence, many lakes have shifted from a clear-water, macrophyte-dominated (if shallow) state with complex food webs and high abundance of piscivorous fish to a turbid, phytoplankton-dominated state with simplified food webs and high abundance of planktivorous fish. As with the problem of lake acidification, also here only limited detailed information often exists on the timing, magnitude, and duration of the biological response to cultural eutrophication, due to lack of long-term monitoring data that include the preimpact period. So far, most cladoceran paleoecological studies have focused on illuminating the biological effects of eutrophication. Fossil cladoceran remains have proven particularly valuable proxy indicators to track shifts in zooplanktivory by either fish or invertebrate predators, and in the local distribution of different types of aquatic habitat (submerged macrophytes, sediments, and the pelagic zone). Most cladoceran paleoecological studies in this research field have employed remains other than resting eggs and ephippia. One notable exception is the comparative study of cultural eutrophication in three Minnesota (USA) lakes conducted by Birks et al. (1976), a classic early example of integrated multiproxy indicator paleolimnology. These authors combined stratigraphic analyses of fossil pollen, aquatic macrophyte remains, diatoms and other algae, molluscs, chydorid cladocera, and Daphnia ephippia to trace the effects of excess nutrient input variously resulting from logging, agricultural runoff, or sewage effluents. Daphnia ephippia abundance in this set of lakes was found to be positively correlated with lake trophic state. Unfortunately, incomplete understanding of the modern functioning of these lakes precluded to identify the proximate cause of the inferred Daphnia population increase in each particular case (e.g. improved feeding conditions or relaxed fish predation). Various recent studies are now demonstrating the full potential of fossil cladoceran resting eggs as eutrophication tracers. Weider et al. (1997) examined long-term genetic shifts in Daphnia egg banks (here of the D. galeata–hyalina complex) in Lake Constance, Germany, where increased nutrient input started to develop eutrophication effects since the 1960s. By applying allozyme electrophoresis to hatched resting eggs of different known age, they found significant changes in the genetic composition through time, evidenced by shifts in the allelic composition of all four polymorphic enzyme loci examined. These shifts indicate that microevolutionary processes (natural selection, genetic drift, and gene flow) have influenced the genetic composition of the Lake Constance Daphnia population over the last 35 years. To elucidate

142

S. AMSINCK ET AL.

whether these genetic shifts were modulated by changes in food quality, Hairston et al. (1999a) devised an experiment to test for changes in Daphnia adaptation to the nutritionally poor or toxic cyanobacteria that were abundant during peak eutrophication. Genetically distinct Daphnia clones were grown from hatched resting eggs extracted from three sediment horizons of known age: 1962–1964, 1969–1971 (before and just after the appearance of cyanobacteria), 1978–1980 (peak eutrophication), and 1992–1994, 1995–1997 (period of recovery). Each clone was subsequently exposed to two different diets: one containing a mixture of a toxic cyanobacterium and a high-quality algal resource, the other containing only a highquality algal resource. The results showed genotypes from both 1978–1980 and the 1990s to exhibit lower growth rate reductions than those from 1962–1964 and 1969–1971 (Fig. 8.4), suggesting that the Lake Constance Daphnia population evolved an increasing ability to cope with a diet containing cyanobacteria. Jeppesen et al. (1996) used a weighted-averaging (WA) transfer function to relate the relative proportion of five pelagic zooplankton taxa preserved in surface sediments to present-day planktivorous fish abundance (expressed as catch per unit effort [CPUE] in multiple mesh size gill nets) in shallow freshwater Danish lakes. Daphnia ephippia, rotifer resting eggs, Leptodora caudal cerci and Bosmina (B. longirostris and B. coregoni) exoskeleton remains were selected on the basis of their contrasting abundances at different planktivorous fish densities, with the small-sized zooplankton dominating under high predation pressures and vice versa. Empirical relationships between CPUE and a number of other lake variables, such as the ratio of planktivorous to piscivorous fish, the maximum depth distribution of submerged macrophytes, the zooplankton to phytoplankton biomass ratio, and water clarity

Sediment age

1995–1997 and 1992–1994

1978–1980

1969–1971 and 1962–1964 0

0.1

0.2

0.3

high resistance

0.4

0.5 low resistance

Growth rate reduction Figure 8.4. Resistance of Daphnia genotypes to the presence of toxic dietary cyanobacteria. Diamonds represent estimates of growth rate reduction for individual clones hatched from resting eggs in different layers of the sediment from Lake Constance. (Reproduced from Hairston et al. 1999a. With permission.)

TRACING OF ENVIRONMENTAL CHANGES

143

imply that this zooplankton-based method to infer planktivorous fish abundance also yielded meaningful information on these covarying habitat characteristics. Jeppesen et al. (unpublished data) established a transfer function to infer the areal coverage of submerged macrophytes in Danish lakes, using the species composition of littoral chydorid species. Its application in numerous Danish lakes together with cladoceran-inferred CPUE and diatom-inferred TP has increased the understanding of the developmental history of Danish lakes, as well as of the biological response to eutrophication in general (Amsinck et al. 2003a). Specifically, these inferences drew attention to the highly variable and site-specific nature of lake development, even in a small country with a long agricultural tradition such as Denmark (Amsinck et al. 2003a). Some lakes (e.g. Søbygaard and Søgaard) appear to have experienced marked eutrophication, including a pronounced decline in submerged macrophyte coverage and increased planktivorous fish abundance during the 20th century. Other lakes (e.g. Esrum and Skanderborg) revealed these symptoms of eutrophication from a much earlier date, at least the last 200 years. In contrast, inferences from Lake Stigsholm showed no sign of cultural eutrophication to the present day. Historical changes in the trophic structure of brackish Danish lakes (Amsinck et al. 2003a,b; Amsinck et al. 2005a) have been detected using a separate CPUE inference model (Amsinck et al. 2005b), again based on a combination of ephippia (Daphnia, Ctenodaphnia, and Ceriodaphnia) and other cladoceran exoskeleton remains. Jeppesen et al. (2002) demonstrated a strong negative relationship between the size (dorsal length) of Daphnia ephippia and CPUE, which was improved further by adding TP as independent variable in a multiple regression. This is because predation risk is markedly related to lake trophic state, being higher in oligotrophic and hypertrophic lakes than in mesotrophic lakes (Jeppesen et al. 2003b). To maximize the Daphnia size range covered, the calibration data set included all Daphnia spp. (of the three subgenera Ctenodaphnia, Hyalodaphnia, and Daphnia spp.) and CPUE data from 52 mainly shallow lakes in Denmark, Greenland, and New Zealand. Given fixed allometric relationships between ephippia size and the size and weight of eggbearing female Daphnia, these authors could infer past changes in both planktivorous fish abundance and mean body weight of summer-time Daphnia from the mean size of fossil Daphnia ephippia in four lakes with, at present, diverging trophic status (Fig. 8.5). In the presently clear Lake Stigsholm, past CPUE was generally low and Daphnia body weight high. Similar results were inferred for the presently hypertrophic Lake Søbygaard, which is known to have been impacted by periodic fish kills. In contrast, inferred CPUE was high and Daphnia body weight low in two presently eutrophic lakes, Søgaard and Lange. At all sites the inferred present-day CPUE values corresponded well with the contemporary CPUE data. Jeppesen et al. (2003a) developed a third method to infer past planktivorous fish abundance, this one being based on the fraction of Daphnia ephippia in the sum of Daphnia and Bosmina ephippia. Fractions of Daphnia and Bosmina were negatively correlated with CPUE in a calibration data set of 135 lakes in Greenland, Denmark, and New Zealand (Fig. 8.6). Adding TP as covariable in the multiple regression

144

S. AMSINCK ET AL.

Date/AD

Lake Stigsholm

Increasing TP

0 Depth (m)

1993 1963 1888

20 40 60 80 100 0

1983

20

1966

40

1943

Depth (m)

1996

Lake Søgård

60 80 100 120 0

1960

20

1900

Depth (m)

1993

Lake Langesø

40 60 80 100 0

1986

5

1970

Depth (m)

1993

Lake Søbygård

10 15 20 0.4 0.8 1.2 1.6 2.0 2.4 Ephippia size (mm) Daphnia size (mm)

0

5

10 15 20 25

Daphnia mean weight (µg)

0

200 400 600 800 CPUE (no net −1)

Figure 8.5. Sedimentary records of mean ephippial size of Daphnia spp., inferred sizes of ephippia-bearing Daphnia spp. during summer, inferred mean weight of Daphnia spp., and inferred planktivorous CPUE for four Danish lakes. The lakes are arranged according to increasing TP. (Reproduced from Jeppesen et al. 2002. With permission.)

again improved the relationship. However, this also makes the resulting CPUE inferences dependent on the output of diatom-based TP inference models, which are not without their share of problems (Anderson & Odgaard 1994). All above-mentioned cladoceran-based studies on the ecological impact of cultural eutrophication made use of fossil resting eggs and ephippia, alone or in combination

145

TRACING OF ENVIRONMENTAL CHANGES Greenland

New Zealand

Denmark

Daphnia in sediment (%)

100 75 50 25 0 0

15

30

45

60

0

Total fish (no net −1)

15

30

45

60

0

Total fish (no net −1)

100

200

300

400

Total fish (no net −1)

G: 11 (6–29) NZ: 21 (5–52)

DK: 193 (27–800)

Daphnia in sediment (%)

All countries 100

100

75

75

50

50

25

25

0

0 0

100

200

300

Total fish (no net −1)

400

0

20

40

60

80 100

Total fish (no net −1)

Figure 8.6. Above: contribution of Daphnia to the sum of Daphnia and Bosmina ephippia in surface sediments vs contemporary fish abundance in Greenland, New Zealand, and Danish lakes. Below: merged data from the three countries (left) and parts of this figure upscaled (right) with mean total phosphorus (µg l−1) concentrations and variations (in brackets) (top). (Reproduced from Jeppesen et al. 2003a. With permission.)

with other parts of the cladoceran exoskeleton. Many other paleoecological studies only made use of exoskeleton remains other than the diapausing stages. Especially the preserved exoskeleton remains of pelagic cladocerans (e.g. Bosmina and Daphnia) have been used extensively as indicators of past predation regimes on zooplankton. Kerfoot (1974) found close correspondence between the historical intensity of fish planktivory in Frains Lake, Michigan, USA, and the relative abundance of Bosmina and Daphnia (headshields, carapaces, postabdomens, and mandibles). Sanford (1993) showed a positive correlation between fish planktivory and occurrence of the cornuta form in B. longirostris, which has stout and strongly curved antennules. Conspecific populations exposed mainly to invertebrate planktivory exhibited relatively long and

146

S. AMSINCK ET AL.

only slightly curved antennules. Hann et al. (1994) used the fossil cladoceran ratio between Daphnia mandibles and Daphnia mandibles + Bosmina carapaces, and morphological changes in Bosmina carapaces, mucrones, and antennules to validate the fossil record of eutrophication-driven planktivory changes in the experimentally manipulated Lake 227, Ontario, Canada. Littoral and benthic Chydoridae have proved useful indicators of past nutrient regimes and submerged macrophyte abundance. A pioneering study in this field was conducted by Whiteside (1970), who examined species assemblages of chydorid remains (headshields, carapaces, and postabdomens) extracted from the surface sediments of 77 Danish lakes in relation to physical and chemical environmental data. The lakes were grouped into (1) clear-water lakes, (2) ponds and bogs, and (3) polluted clear-water lakes, and chydorid species were classified according to their ecological preferences. This classification (see also Whiteside & Swindoll 1988) was later extensively used in cladoceran-based paleoecological studies on eutrophication, for instance by Hofmann (1996) in a comparative study of 13 north German lakes (Plön district) with different trophic status. Thoms et al. (1999) used the ratio of chydorid remains to the sum of chydorid and bosminid remains to infer changes in nutrient influx and macrophyte abundance following 19th-century settlement by ancestral Europeans near the Murray River, Australia. Their inference method was calibrated using a surface-sediment reference data set from 38 billabongs (standing waters formed in cutoff river meanders). Using modern quantitative techniques, Brodersen et al. (1998) established a chydorid-based TP inference model to estimate changes since the mid-1960s in the trophic status of 32 Danish lakes. The currently, eutrophic lakes showed a systematic increase in TP through time, while only minor changes were found in present-day oligotrophic lakes. Bos and Cumming (2003) established a TP inference model based on both littoral and pelagic Cladocera sampled in surface sediments from 53 lakes in central British Columbia, Canada. 8.4 TRACING FISH INTRODUCTIONS AND BIOMANIPULATION

Fish have been introduced to many lakes to promote commercial and recreational fishing. Once made, these introductions are often irreversible, but in most cases their ecological impacts have not been properly documented, because monitoring data from before, during and following fish introduction are lacking. Here again, stratigraphic analyses of cladoceran remains in lake sediments have proved valuable to evaluate the impact of fish stocking on food web interactions and general aquatic ecosystem functioning. They are thus a powerful tool to develop and apply lake-specific fish management strategies. One pioneering study was made by Kitchell and Kitchell (1980), who used Daphnia postabdominal claws and Bosmina carapaces to show both an immediate response and a long-term shift in zooplankton species composition following the introduction of zooplanktivorous rainbow trout (Oncorhynchus mykiss) to Peter Lake, Michigan, USA. Large-bodied Daphnia pulex was replaced first by smaller D. rosea and then Bosmina, and also within each Daphnia sp. size reduction through time was evident. Examples of related studies,

TRACING OF ENVIRONMENTAL CHANGES

147

all of them using cladoceran remains other than resting eggs, are those of Brugam and Speziale (1983), Salo et al. (1989), and Miskimmin et al. (1995). Brugam and Speziale (1983) studied the ecological impact of Northern pike (Esox lucius) management in Lake Harriet, Minnesota, USA, using remains of Bosmina (carapaces, headshields, and antennules) and Daphnia (postabdominal claws and mandibles). Salo et al. (1989) used size-frequency analysis on the fossil carapaces of Bosmina coregoni and Chydorus sphaericus to elucidate the impact of whitefish (Coregonus lavaretus) introduction to Lake Pyhäjärvi, Finland. Miskimmin et al. (1995) used Bosmina and chydorid headshields and carapaces, and Daphnia postabdominal claws to examine the impact of toxaphene application and trout introduction to two lakes (Annette and Chatwin) in western Canada. It is sometimes challenging to recover the smaller fossilized body parts of some predation-sensitive cladocerans, especially Daphnia mandibles and postabdominal claws, from core samples with abundant coarse plant remains or high silt content. Verschuren and Marnell (1997) hence explored the possibility to trace food web effects of past fish introductions in Glacier National Park, Montana, USA, using the stratigraphic distribution and size of fossil Daphnia ephippia. The purpose of this study was to examine whether the genetically pure population of westslope cutthroat trout (WCT) (Oncorhynchus clarki lewisi) currently inhabiting Avalanche Lake, a subalpine headwater lake located above a strongly cascading mountain stream, was truly indigenous or the result of an undocumented introduction. If indigenous, this population would represent a valuable source of fish stock to reintroduce WCT to other lakes in its native region of western North America (Montana, Idaho, British Columbia, and Alberta) where populations have suffered due to land-use practices and widespread introduction of nonnative fishes. The sediment record of Daphnia ephippia (mainly D. middendorfiana) in Avalanche Lake (Fig. 8.7) reveals low and sporadic occurrence for at least a century before the establishment of Glacier National Park in 1910, suggesting an original situation of intense zooplanktivory by indigenous cutthroat trout. A marked increase in the abundance and size of ephippia deposited during the 1930s and early 1940s could, given the relative climatic and hydrological stability of this particular lake habitat, be inferred to represent an episode of reduced zooplanktivory. This episode of high Daphnia abundance coincided with the period between 1915 and 1943 when massive efforts were undertaken to stock Avalanche Lake with Yellowstone cutthrout trout. It thus appeared logical to conclude that the native trout population had undergone intense food competition and collapsed. However, since Yellowstone cutthroat trout (YCT) prefer large glacial lakes at lower elevations, the stocked juvenile trout would have tended to leave Avalanche Lake before reaching reproductive age. Thus, when by 1943 stocking failed to produce the desired result and was ceased, the local WCT population was able to recover as a genetically pure population. Judging from the ephippial record, by the mid-1950s it again exerted intense zooplanktivory on Daphnia. Verschuren and Marnell (1997) validated their use of ephippia as indicator of past Daphnia population abundance by showing agreement between the abundances of fossil ephippia and postabdominal claws in the Avalanche Lake record, and also between the concentrations of Daphnia ephippia deposited in surface sediments

148

S. AMSINCK ET AL.

Figure 8.7. Stratigraphical record of Daphnia ephippia abundance, Daphnia ephippial size, and Daphnia postabdominal claws in Avalanche Lake. The first observation of native westslope cutthroat trout (WCT), the establishment of Glacier National Park and the stocking record of nonnative Yellowstone cutthroat trout (YCT) are indicated to the right. (Reproduced from Verschuren & Marnell 1997. With permission.)

and live Daphnia population abundances in eight Glacier Park lakes, Montana, USA, with or without fish. Jeppesen et al. (2001a) studied the impact of early 1900s pikeperch (Stizostedion lucioperca) introduction in Lake Skanderborg, Denmark, using the Jeppesen et al. (1996) CPUE inference model based on Daphnia and Bosmina ephippia among other cladoceran remains. This study reconstructed conditions of intense zooplanktivory prior to introduction of the piscivore pikeperch in 1903–1904. Inferred CPUE declined slightly after pikeperch introduction and was further depressed following a major reduction in external phosphorus loading in 1974 and zooplanktivorous fish removal in 1983–1985. The inferred historical CPUE values corresponded well with early fishery studies and documentary records from recreational anglers. CPUE values inferred from the fossil zooplankton assemblage in the surface sediment of Lake Skanderborg also agreed well with recent monitoring data. Moreover, the size of fossil Daphnia resting eggs in the sediment record varied inversely with inferred CPUE, as predicted by the theory of size-selective predation. Similar long-term associations between fish-sensitive planktonic cladocerans and their predators were documented by Cousyn and De Meester (1998), who examined

TRACING OF ENVIRONMENTAL CHANGES

149

the abundance and size of fossil Daphnia magna ephippia in the sediments of former fish ponds in Flanders, Belgium. Pb-210 dating of the sediment profile proved problematic, but comparison with documentary fish records was nevertheless possible by anchoring the bottom of the fossil Daphnia records to the date when the fish ponds had been dug out. This revealed inverse relationships between the historical abundance and mean size of D. magna ephippia and the stocked density of zooplanktivorous fish, as predicted by the theory of size-selective visual predation. Cousyn et al. (2001) followed to show that, in one of the ponds, natural selection among local D. magna clones in response to the historical changes in fish predation pressure had driven rapid evolutionary changes in Daphnia behavior. Oud Heverlee Pond was constructed in 1970 and has a well-documented fish stocking history with three distinct periods of variable fish predation pressure: relatively low fish predation during the early 1970s; high fish predation between 1973 and 1982; and again relaxed fish predation from 1982 onwards. The authors isolated Daphnia resting eggs deposited during each of these periods to quantify changes in population abundance, average adult body size, genetic identity, and phototactic behavior through time. The latter was determined by exposing the Daphnia hatched from each resting egg subpopulation to fish kairomones. Besides the expected inverse correlations between ephippial abundance and size on the one hand, and historical fish density on the other, Cousyn et al. (2001) found clear genetic differentiation between the three temporal subpopulations, and a more strongly negative phototactic response in the clones that were hatched from resting eggs deposited during the period of intense fish stocking. Thus, when exposed to intense fish planktivory, Daphnia seems to rapidly adopt a more pronounced diel vertical migration behavior to become less vulnerable to visual predators. 8.5 TRACING HEAVY-METAL POLLUTION

During the 19th- and 20th-century buildup of heavy industry, many industrial wastes, including highly toxic heavy metals, were released into the natural environment and eventually accumulated in aquatic ecosystems. Detailed knowledge of the timing and magnitude of this environmental perturbation, as well as of the timescale of ecosystem recovery following implementation of more stringent pollution legislation, is often inadequate or even lacking. Cladoceran-based paleoenvironmental reconstruction again has made significant contributions in this field, although the subject is relatively new and focused studies are still comparatively few. Manca and Comoli (1995) used a wide range of preserved cladoceran remains (headshields, carapaces, postabdomens, postabdominal claws, and mandibles) to reconstruct the long-term ecological effects of severe copper and ammonium sulfate pollution in Lake Orta, Italy. Untreated discharges from a rayon factory established in 1926 caused progressive acidification and near-complete disappearance of zoo- and phytoplankton communities, and decimation of the fish population. Correspondingly, the sediment record revealed marked changes in the total abundance and species composition of the local Chydoridae, most notably during the period of toxic stress.

150

S. AMSINCK ET AL.

Kerfoot et al. (1999) used descriptive and genetic stratigraphic analyses in combination with toxicity tests on live zooplankton hatched from old but still viable resting eggs (a new approach termed “resurrection ecology”) to elucidate the ecological impacts of former copper mining activities on the Keweenaw Peninsula in Lake Superior, USA. Between 1850 and 1929, the Keweenaw district was the secondlargest copper producer in the world and, hence, a highly appropriate location to study the ecological effects of metal toxicity. Stratigraphic analysis of sediment cores from Portage Lake, Michigan, USA, showed an overall low abundance of Daphnia ephippia and Bosmina carapaces and headshields during the waste disposal era (1856–1947), indicating depressed productivity. Greater abundances of both taxa in more recent sediments suggested partial recovery after mining ceased. Yet, genetic analysis of Daphnia populations hatched (resurrected) from resting eggs of different age revealed a marked genetic change in species community structure. Originally, Portage Lake was dominated by the relatively large-bodied Daphnia rosea (similar to D. dentifera). Following the creation of a waterway to facilitate ship passage, mixture of inland and Lake Superior zooplankton occurred and a hybrid between D. rosea and D. galeata mendotae (D. mendotae) became dominant in Portage Lake. The results imply that further recovery will not lead to the return of a Daphnia community similar to that existing under premining conditions. In a series of toxicity experiments, the authors then exposed Daphnia hatched from resting eggs recovered from premining, mining, and postmining sediment strata to dissolved copper concentrations which resembled the different historical levels of pollution. These tests confirmed that copper concentrations and fluxes during the mining era were highly toxic to Daphnia. Surviving Daphnia recovered rapidly when subsequently exposed to concentrations resembling the postmining period. Further, a progressive decline in copper resistance among resurrected Daphnia clones originally living under conditions of decreasing copper availability indicated inherent differences in copper resistance among the daphnids. This study is a prime example of how complementing traditional paleolimnological analyses with experimental genetics and toxicity tests can improve insight into the biological consequences of anthropogenic impact at the ecosystem level. 8.6 TRACING CLIMATE CHANGE

Lake sediment records have been used extensively to study past climate change and the various ways in which climate can affect the long-term dynamics of lake ecosystems (Battarbee 2000; Smol et al. 2005). Now that anthropogenic global warming is generally acknowledged to be a reality, climatic impacts on lakes are attracting increased interest. Multiple sedimentological, geochemical, and biological proxy indicators have been exploited for climate reconstruction, and fossil cladocerans constitute no exception. In the hydrologically stable lake systems that are common in humid north temperate regions, cladocerans and other lake biota can be exploited as direct or indirect paleotemperature indicators (e.g. Lotter et al. 1997). In the hydrologically fluctuating lake systems that are typical of many tropical and subtropical

151

TRACING OF ENVIRONMENTAL CHANGES

regions but also occur in continental Europe and North America, they are exploited instead as indicators of climate-driven lake level (e.g. Hofmann 1998) or waterchemistry changes (e.g. Bos et al. 1999; Verschuren et al. 1999b). To date, most climate-related cladoceran studies have dealt only with nondiapausing exoskeleton remains. Below we present a selection of these, together with the notable exceptions that focus on ephippia and resting eggs. Sarmaja-Korjonen (2004) developed a method to trace past temperature changes using the ratio between ephippia (or gamogenetic carapaces) and regular carapaces deposited by Chydoridae and other Cladocera. Cladocerans shift from parthenogenetic to sexual reproduction (gamogenesis) when stressed by a particular environmental factor, which depending on location and time can be low temperature, reducing photoperiod in autumn, enhanced predation risk, poor food conditions, and crowding (Stross & Kangas 1969; Frey 1982; Carvalho & Hughes 1983; Larsson 1989; Pijanowska & Stolpe 1996). The overall proportion of parthenogenetic and gamogenetic reproduction through successive generations is clearly dependent on temperature, as indicated by order-of-magnitude differences in the surface-sediment ephippia to carapace ratio between high Arctic and warm temperate lakes, and the strong correlation between this ratio and mean summer (May–September) air temperatures (Fig. 8.8). In stratigraphic studies, Bennike et al. (2004b) found 20–30% ephippia among the Chydoridae of Lake Bølling, Denmark during the middle Allerød and early Younger Dryas periods. Sarmaja-Korjonen (2003, 2004) found 10–15% ephippia among Chydoridae in Finnish Lapland during the late Pleistocene/Holocene transition and in mid-Holocene sediments, and comparatively lower values during the early Holocene warming; present-day values in Finland are 2–5% (Bennike et al. 2004a). To fully realize the potential of such ephippia to carapace ratios, more targeted research is needed on the various confounding factors that

0

5 10 10 1 15

Air temperature (˚C)

Bosmina egg ratio

100

20

0.1

G

F

DK

NZ

Figure 8.8. Box plot of the percentage ratio of ephippia to the sum of ephippia and carapaces of Bosmina in the surface sediments of Greenland, Faroese, Danish, and New Zealand lakes compared to summer (1 May–1 October) mean air temperature. (Reproduced from Jeppesen et al. 2003a. With permission.)

152

S. AMSINCK ET AL.

may corrupt the ratio’s relationship with temperature. Sample size is also a problematic issue, due to a general scarcity of ephippia in warm-water lakes. Other scientists have used the full set of cladoceran remains to develop quantitative inference models for reconstruction of summer temperature. Lotter et al. (1997) calibrated their models with surface sediment data from 68 Swiss, French, and Italian lakes located between 300 and 2,350 m elevation. The statistically most powerful inference model was based on species assemblages of benthic Cladocera only. Duigan and Birks (2000) later adjusted this model to reconstruct late Glacial and early Holocene temperature changes at Kråkenes, Norway. Their cladoceran-based inferences suggested that the Allerød period was only slightly warmer than the Younger Dryas, and that a strong progressive increase in summer temperatures did not happen until the early Holocene. The temporal pattern of reconstructed temperature change was similar to that inferred from fossil chironomid assemblages, except that absolute values were higher, in fact also higher than expected from the pollen record. The authors attributed this discrepancy to nonanalog situations arising from their use of the Swiss calibration data set at a Norwegian site. Korhola (1999) developed a cladoceran-based temperature inference model for sub-Arctic Finnish Lapland. Much older cladoceran remains, including diverse exoskeleton parts of 25 chydorid species and the ephippia of Daphnia spp. and Ceriodaphnia, were recovered by Frey (1962) from a last interglacial section (Eemian; ~125,000 years old) in Jutland, Denmark. Patterns of species succession upward through the section indicated gradual amelioration of climate after the penultimate glaciation, followed by a climatic optimum and then gradual deterioration as the last glaciation set in. Fairly good agreement between the mean rank abundances of Chydoridae spp. during the Eemian and in modern Danish lakes was interpreted by the author to indicate general ecological (and morphological) stability of these biota over this long period of time. While temperature strongly influences the species composition, richness, and relative abundance of cladoceran communities (Harmworth 1968; George & Harris 1985; Frey 1988; Patalas 1990), in many lakes various factors related or unrelated to climate change (eutrophication, acidification, fish stocking, change in lake depth, and stratification) must have affected species assemblages during the period targeted for climate reconstruction. Consequently, cladoceran-based temperature reconstructions should be interpreted with great care. When climate change affects a lake’s hydrological balance, fossil Cladocera can be used to trace the resulting changes in lake level (Hofmann 1998), salinity (Bos et al. 1996, 1999), and the cascading effects of these physical changes on lake biology (e.g. fish, macrophytes, and nutrients: Verschuren et al. 1999a,b, 2000; Amsinck et al. 2003b, 2005a,b). In one early study, Goulden (1966) used the overall ephippia to carapace ratio in fossil cladoceran assemblages to infer hydrological changes in the Aguada de Santa Ana Vieja, a shallow pond near former Maya settlements in Guatemala. In this case, an elevated ephippia to carapace ratio in portions of the sediment record was inferred to reflect episodes of low water level. Goulden’s ratio was based on summed abundances across all local cladoceran species. Hence, it did

TRACING OF ENVIRONMENTAL CHANGES

153

not represent various species in the same proportion. In fact, almost all enumerated carapaces belonged to Chydoridae, while ephippia were derived exclusively from Macrothrix and ( judging from the pen drawing) Moina. Fossil Cladocera may potentially also be used to trace past changes in ultraviolet (UV) radiation. UV-protective pigmentation is common among the Daphnia inhabiting high-altitude and arctic lakes, which are characterized by low concentrations of dissolved organic carbon (DOC) and are subjected to high UV radiation (Hessen et al. 1999; Rautio & Korhola 2002). This pigmentation likely occurs in their ephippia as well. A potential limitation of this method is, however, that in order to isolate the direct effect of UV, DOC would have to be reconstructed independently by other proxies. 8.7 DISCUSSION: METHODOLOGICAL LIMITS, CONCERNS AND FUTURE POTENTIAL

The studies presented above illustrate the application range of fossil ephippia and resting eggs as tracers of both natural (e.g. climate-driven) and anthropogenic changes in lake community structure and trophic dynamics. Established cladoceran paleoecological techniques range from the more traditional approach to reconstruct past abundances of indicator species (e.g. Prazákova & Fott 1994; Verschuren & Marnell 1997; Cousyn & De Meester 1998) to modern multivariate statistical techniques permitting quantitative inference of past environmental variables (e.g. temperature, pH, TP, salinity, fish, and macrophytes: Krause-Dellin & Steinberg 1986; Jeppesen et al. 1996, 2002, 2003a,b; Brodersen et al. 1998; Bos & Cumming 2003), “resurrection ecology” uncovering changes in behavioral traits and pollution tolerance, and molecular genetic techniques tracing the genetic history of cladoceran populations (Weider et al. 1997; Hairston et al. 1999a; Kerfoot et al. 1999; Cousyn et al. 2001; Arnott & Yan 2002; Pollard et al. 2003). Traditional cladoceran paleoecology has also been invigorated by ingenious exploitation of the abundance ratio between parthenogenetic and gamogenetic remains (Sarmaja-Korjonen 2002, 2003, 2004). Further exciting developments not discussed above are the nitrogen isotope analysis of subfossil cladoceran chitin (Stuck et al. 1998) and analysis of cladoceran egg banks to trace the colonization of newly established and isolated pools (Vandekerkhove et al. 2005c). The former technique elucidated past changes in the dietary composition and trophic position of Bosmina in the Baltic Sea, while the latter may broaden our understanding of species migration patterns, species adaptation to new habitats, and the dynamical relationship between local and regional species diversity. Given the general lack of long-term lake-monitoring data and proxy historical information, a strong impetus evidently exists to optimally combine and integrate two or more of the above methods. For any study reconstructing past species (relative) abundance, analysis of the regular exoskeleton remains (e.g. carapaces, headshields, and postabdomens) is important to validate inferences based on ephippia and resting eggs. This is less the case for studies limited to

154

S. AMSINCK ET AL.

reconstructions of body size, stable isotope composition, or population genetic change through time. A satisfactorily complete picture of past environmental change required to fully understand long-term ecosystem dynamics can only be gained by integration (and cross-validation) of cladoceran paleoecological data with information from independent biological (e.g. diatoms, chironomids, Chaoborus, algal pigments, and aquatic macrophyte fossils), lithological and geochemical proxy indicators. Cladoceran paleoecology has several strong points compared with other methods of lake investigation: 1. In many cases paleoecological methods are the only tool available to gain insight into the developmental history of individual lakes and to elucidate past human impact on lake systems. 2. Recent death assemblages of cladoceran remains in surface sediments produce a more accurate and cost-effective tool to assess local species richness than does snapshot sampling of the living community, because they integrate both spatial (microhabitat) and temporal (seasonal and interannual) variation in the abundance and distribution of species (Frey 1960; Jeppesen et al. 2003a,b; Vandekerkhove et al. 2004a, 2005a, d). 3. Several biologically important environmental variables, such as TP, macrophyte cover, and fish abundance, can neither be measured directly in lake sediments nor quantitatively inferred from the often fragmentary sediment records of fish remains (bones and scales; Davidson et al. 2003) and aquatic macrophytes (e.g. seeds, reproductive structures, and tissues; Jeppesen et al. 1996, 2001b). 4. Long-term viability (~100 years) of cladoceran resting eggs allows resurrection (Hairston et al. 1999a; Kerfoot et al. 1999; Cousyn et al. 2001) of former populations genetically and behaviorally adapted to different selection pressures. These studies provide unique insight into the timescale of genetically determined physiological (stress tolerance) and behavioral (predation avoidance) responses to ecosystem change, and the true ecological magnitude of historic environmental perturbation. Clonal populations cultured from resurrected eggs also allow examination of population genetic changes by allozyme electrophoresis analysis (Weider et al. 1997), complementing genetic analysis of the eggs themselves (Reid et al. 2000; Cousyn et al. 2001; Limburg & Weider 2002). All individual paleoecological and paleogenetic techniques have their own particular methodological problems, placing limits on their range of application. These do not invalidate the technique, but underscore the need to integrate inferences with information from independent sources or sedimentary proxy indicators to achieve valid reconstructions. Inferences of past population abundance based exclusively on fossil ephippial counts must be interpreted with great caution, because ephippia production is species-specific (Jankowski & Straile 2003) and dependent on climate regime (Jeppesen et al. 2003a,b). Another prominent limitation is that species-level identification of ephippia is problematic in some groups (e.g. the Chydoridae) due to lack of useful diagnostic characters or high intraspecific morphological variability (Vandekerkhove et al. 2004a). Species-specific taphonomy, i.e. the relationship

TRACING OF ENVIRONMENTAL CHANGES

155

between ephippia production, deposition, and preservation, may also bias the inferred relative abundance of species within a community. For example, some Daphnidae possess higher buoyancy due to longer dorsal spines, a higher lipid content or presence of gas chambers, factors that increase the probability of transport from limnetic to littoral areas before sinking (Jankowski & Straile 2003). Although preburial transport may still return such ephippia back to offshore locations before final deposition, their exposure to more harsh physical conditions than ephippia sinking directly from the pelagial into deep water tends to result in their underrepresentation in a mid-lake sediment record. Depending on the extent of preburial damage incurred, and (particularly in shallow lakes) further wear at the deposition site, ephippia may lose important characters for identification and become unsuitable for size measurements (Limburg & Weider 2002). Chydoridae and Sididae are known to attach their ephippia to substrata at the time of shedding (Fryer 1996), thus compromising the assumed spatial integration of ephippia across lake microhabitats. If past community composition is inferred solely from the stratigraphy of fossil resting eggs, factors such as these are bound to create biased assessments of species composition, abundance and richness. Examination of other cladoceran exoskeleton remains allow one to partly correct for such bias, but is, however labor-intensive and requires detailed morphological analysis. Problematic species-level identification can also be resolved by culturing clonal populations hatched from the resting eggs. Yet, as hatching stimuli vary among taxa, successful culturing requires incubation under a wide range of chemical and physical conditions (Vandekerkhove et al. 2005d). In older samples, reduced egg viability will become an issue influencing the success of this culturesupported identification: even in recently deposited Daphnia resting eggs, hatching rate is less than complete (Weider et al. 1997). PCR restriction fragment length polymorphism (RFLP) analysis on microsatellite and mitochondrial DNA (Reid et al. 2000; Limburg & Weider 2002; Mergeay et al. 2005a,b) now also permits direct genotyping of resting eggs, adding a promising new dimension to the study of longterm genetic changes both at the population and the community level. WA statistical models used to infer environmental variables (CPUE, TP, pH, and macrophytes) through calibration of species–environment relationships should also be used with caution. Some problems are related to the numerical procedures themselves, while others are related to the nature of regional calibration data sets. The first category includes the “edge effect,” a systematic bias of inferred values close to the lower and upper ends of a calibrated environmental gradient that is inherent to the unimodal species-response modeling in WA techniques (Birks 1998). One way to overcome this problem is to define, for each fossil sample and within a large calibration data set, a smaller dynamic calibration data set of a selected number of surface-sediment samples (e.g. from 10 to 20 lakes) that have species assemblages most similar (analogous) to that individual fossil sample, and then apply linear species-response modeling techniques to the short environmental gradient covered by this smaller calibration data set (Simpson 2003). Techniques based on analogmatching of species assemblages often have the advantage of very large calibration data sets (hundreds of lakes), markedly larger than most existing calibration data sets

156

S. AMSINCK ET AL.

of cladoceran species distribution from which WA inference models are developed (e.g. Jeppesen et al. 1996; Brodersen et al. 1998; Bos et al. 1999; Bos and Cumming 2003; Amsinck et al. 2005b). Issues related to the nature of regional calibration data sets include the obvious necessity that in order to develop a robust inference model a distinct relationship must exist between the distribution of cladoceran species among sites and the environmental variable of interest. Specifically concerning cladoceran-based inference models for TP and pH, it remains uncertain whether cladocerans respond to these variables directly or to concomitant shifts in, for example, primary producer communities substrate availability and prey–predation relationships. This common covariation of biologically important variables and processes complicates or hampers the separation of their effects on cladoceran communities and can generate situations in which taxa modeled to reflect macrophyte presence are not confined to macrophyte-covered habitats. For instance, C. sphaericus tends to be abundant in turbid nutrient-rich lakes lacking submerged plants, and Ceriodaphnia spp. may be common in the pelagic zone, of lakes with poor macrophyte coverage (Hann et al. 1994; Pieczynska et al. 1999). Selective fish planktivory on macrophyte-associated and benthic chydorids may also alter their relative abundance (with selective removal of large-sized taxa such as Eurycercus, Simocephalus, and Sida; Jeppesen 1998; Blumenshine et al. 2000). Field estimates of CPUE are subject to various causes of measurement uncertainty (Jeppesen et al. 1996), which may explain in part why cladoceran-based CPUE inferences (Jeppesen et al. 1996; Amsinck et al. 2005b) are less precise than cladoceran-based inferences of TP (Brodersen et al. 1998; Bos & Cumming 2003) and salinity (Bos et al. 1999), or diatom-based inferences of TP and salinity (Hall & Smol 1992; Bennion et al. 1996; Reed 1998; Ryves et al. 2002). Better quantification of fish abundance (biomass and composition) and inclusion of more benthic invertebrates sensitive to size-selective fish predation may meaningfully improve CPUE inferences (Jeppesen et al. 2001b). In conclusion, cladoceran-based paleoecology has developed rapidly over the last two decades. Especially the now routine development of quantitative inference models and the combined application of resurrection ecology and paleogenetic techniques have considerably strengthened the discipline and opened up new avenues of study. Many of these methods do require further improvements to optimize the trustworthiness and accuracy of produced inferences of past aquatic community structure and trophic dynamics. Also, to enhance confidence in the use of fossil cladoceran records in ecological applications, there is a need to better elucidate the factors regulating the formation of fossil cladoceran assemblages from a living cladoceran community (taphonomy), in particular the nature of bottom substrates, preburial transport, spatial integration across habitats within a lake, bioturbation, and physical sediment disturbance. More routine application of multiple proxy indicators will also enhance the strength of reconstructions via opportunities for cross-validation. Hopefully, future research efforts will lead to the development of more accurate and effective inference models suited for detailed studies of past lake history and anthropogenic perturbation and development of meaningful scenarios predicting future changes in lake trophic structure and dynamics.

This page intentionally blank

TRACING OF ENVIRONMENTAL CHANGES

157

Acknowledgments. The lead authors’ own research presented in this chapter was funded by the Danish Natural Science Research Council project CONWOY “Consequences of weather and climate changes for marine and freshwater ecosystems. Conceptual and operational forecasting of the aquatic environment” (SWF: 2052-01-0034), EU-IP project EUROLIMPACS (GOCE-CT-2003-50554), and the Carlsberg Foundation.

LUC DE MEESTER, JOACHIM MERGEAY, HELEN MICHELS AND ELLEN DECAESTECKER

9. RECONSTRUCTING MICROEVOLUTIONARY DYNAMICS FROM LAYERED EGG BANKS

9.1 INTRODUCTION: DORMANT STAGES AND THE STUDY OF MICROEVOLUTION

The documentation of evolutionary processes is notoriously difficult because of the time aspect involved. Although evolutionary responses may happen at a surprisingly fast rate when selection pressures are strong (reviews in Hendry & Kinnison 1999; Hairston et al. 2005), in most cases at least a few to several tens of generations are needed for populations to genetically adapt to new environmental conditions. Whereas such timescales are easily incorporated in experimental evolution studies using bacteria and unicellular organisms (e.g. Rainey & Travisano 1998; Elena & Lenski 2003), this is less easy when one considers organisms with relatively long generation times (Reznick et al. 1997; Grant & Grant 2002). In many studies, one therefore adopts the “space for time” approach, and quantifies the level of genetic differentiation among spatially separated populations, verifying whether the observed pattern is in accordance with predictions under the assumption of local genetic adaptation. The observed pattern is then interpreted as reflecting a process of local adaptation. Although this approach has been successful, there are some caveats that often impede a straightforward interpretation (see also Kawecˇki & Ebert 2004). For instance, other differences among the studied habitats than the one focused upon may influence the observed pattern. Therefore, this approach requires a relatively large number of populations to cope with this problem. In addition, there is no guarantee that the observed genetic differences among populations arose through natural selection acting upon genetic variation present within the local population – the mechanism implied by the definition of local genetic adaptation (Kawecˇki & Ebert 2004). Indeed, the observed genetic differences may be the result of founder effects and genetic drift, or may have resulted from gene flow from other populations rather than from local adaptation within the studied population. Several of the above-mentioned methodological problems can be overcome by combining a paleolimnological study with an evolutionary approach. Sediment cores containing stratified dormant egg banks can be used as an archive to reconstruct the process of local genetic adaptation and microevolutionary changes in one particular habitat through time. The major advantage of this approach is that one documents directly the changes that occurred through time, thus giving a more reliable account on the process of local adaptation. It allows one to document changes that occurred over years to decades without having to monitor the population intensively during that whole period. Importantly, it allows documentation of evolution under natural conditions rather than in the controlled but potentially artificial conditions used in experimental evolution studies. 159 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 159–166. © 2007 Springer.

160

L. DE MEESTER ET AL.

Many aquatic organisms produce dormant stages that may get buried in sediments. These dormant stages can often be used as an archive of past conditions, and may be used to reconstruct the history of ecological and evolutionary interactions. Until recently, dormant stages have mostly been used in paleoecological research (see Chapter 8). In the early 1990s, pioneering plant ecologists explored the potential of seed banks to study evolutionary changes within specific populations (Bennington et al. 1991; McGraw et al. 1991; Vavrek et al. 1991). These studies were inspiring, but were limited because of the coarse time resolution typical of terrestrial soils. Full appreciation that the dormant egg banks of many aquatic organisms have important ecological and evolutionary implications paralleling those of the much better-studied plant seed banks occurred in the second half of the 1990s (e.g. Brendonck et al. 1998), and has been strongly enhanced by the in-depth and continued dormant egg bank research on calanoid copepods and cladocerans of Nelson G. Hairston Jr. and coworkers (Hairston & Munns 1984; Hairston & Olds 1984; Hairston 1987; Hairston & De Stasio 1988; Hairston et al. 1996, 1999a,b, 2001). In the last decade, interest in dormant egg banks of aquatic organisms has increased dramatically, not in the least with respect to its potential to reconstruct microevolutionary responses in natural populations. Standing waters have the advantage over terrestrial systems in that sediment deposition rates are much higher and sediment disturbance is much lower, thus yielding a much better time resolution. As a result, genetic, ecological, and evolutionary traits of dormant stages (or their hatchlings) preserved at different depths in lake sediments are relatively easy to compare, provided that sediment layers are well-preserved. In deep lakes with undisturbed sediments, varved layers actually allow a 1-year resolution. The power and attractiveness of analyzing dormant egg banks is thus very high. Kerfoot et al. (1999) coined the term “resurrection ecology” for studies on quantitative genetic analyses on hatchlings from dormant egg banks, whereas the term “paleogenetics” has been used to refer to studies using molecular markers directly on the dormant eggs (Duffy et al. 2000). In sections 9.2 and 9.3, we give a short account of developments in this field of research, emphasizing recent progress. We also discuss potential pitfalls and the conditions which study cases must fulfill for a proper analysis. We then point to some future directions. 9.2 A SHORT SURVEY OF RECENT SUCCESS STORIES

Wolf and Carvalho (1989) and Carvalho and Wolf (1989) were among the first to explore hatching dormant eggs of zooplankton from lake sediments in an effort to study the impact of these egg banks on the genetic diversity in natural populations of zooplankton. They were, however, not very successful in obtaining large number of hatchlings. Weider et al. (1997) successfully analyzed changes in allele frequencies of allozyme markers through time in Lake Constance, Germany, in association with eutrophication. They observed marked changes in the frequency of particular PGI alleles, which parallel the history of eutrophication in this lake. For this study, the authors had to hatch dormant eggs from the sediments, as allozyme

RECONSTRUCTING MICROEVOLUTION

161

electrophoresis requires more material than present within one single egg. With the application of polymerase chain reaction (PCR)-based DNA markers, notably variable microsatellite loci, this latter problem could be overcome and studies could be conducted directly on eggs, even nonviable ones, isolated from lake sediments. Hairston et al. (1999b) first used DNA sequencing to show a temporary range expansion of Daphnia exilis in Onondaga Lake, New York, following intensive industrial pollution of the lake, and similarly Duffy et al. (2000) demonstrated the pollutionfacilitated invasion by exotic Daphnia curvirostris in the same lake. Reid et al. (2000) gave a first report on the use of microsatellite markers in paleogenetic studies on Daphnia dormant egg banks, while Cousyn et al. (2001) applied microsatellite markers to show that the local population of Daphnia magna in a shallow lake was genetically continuous despite strong changes in habitat conditions, as genetic differentiation for neutral markers through a 30-year transect was very shallow. Limburg and Weider (2002) also used microsatellite markers to compare genetic changes in a natural Daphnia population over a period of about 200 years. The use of neutral markers such as microsatellites yields different information from that obtained by an analysis of ecologically relevant traits. Whereas neutral markers indicate patterns of genetic drift, historical connectedness and gene flow, the analysis of ecologically relevant traits reveals changes due to natural selection. Hairston et al. (1999a, 2001), building upon the work of Weider et al. (1997), documented changes in the resistance of D. galeata to toxic cyanobacteria in Lake Constance. This study correlated the genetic composition of the local Daphnia population with changes in phytoplankton composition and the occurrence of cyanobacterial blooms associated with the historically well-documented eutrophication and subsequent oligotrophication of this large lake. Kerfoot et al. (1999) documented genetic adaptation to changes in exposure to heavy metals (copper) in Daphnia. Cousyn et al. (2001) showed that a D. magna population in a shallow lake showed strong changes in genotypic values for predator-induced defenses in response to historical changes in fish stocking in this lake. These changes were in agreement with the pattern expected under the hypothesis of local adaptive evolution. Kerfoot and Weider (2004) reported significant evolutionary responses to changes in (invertebrate) predation pressure in a natural population of Daphnia retrocurva. In all these studies, genetic adaptation occurred in a time frame of less than a decade to up to a few decades. Not all studies show significant evolution, however. Spaak and Keller (2004) found no evidence for adaptive microevolution in the Daphnia population of Lake Greifensee, Switzerland, in response to phosphorus reduction in this lake. As neutral markers and ecologically relevant traits tell different stories, a combined study of both traits has important added value. More specifically, to unequivocally show that genetic changes in ecologically relevant traits are due to natural selection, it is important to show that these changes are not paralleled by similar changes in neutral markers (Spitze 1993). If neutral markers show a similar pattern, this may indicate that gene flow or genetic drift has been an important driver of the observed changes. The study of Cousyn et al. (2001) so far is the only one that combined both approaches. Their analysis pointed out that the observed genetic changes

162

L. DE MEESTER ET AL.

through time for ecologically relevant traits (QST ~0.2; QST quantifies the proportion of the total genetic variation that is due to population subdivision, similar to FST values for neutral markers) in the studied Daphnia population was much higher than that of microsatellite markers (FST ~0.02). This shows that the observed changes for ecologically relevant traits are not a side effect of gene flow or another neutral process, but directly result from evolution through natural selection. 9.3 PITFALLS

Although a number of studies that used dormant egg banks to reconstruct microevolutionary responses have reported revealing data, their success depends on a number of conditions, and many more studies may be initiated than are eventually published. Although dormant egg banks offer a straightforward way to document microevolutionary changes in natural populations, this potential can only be realized when a number of conditions are met. First, one needs to identify a population which has been subjected to the environmental change of choice. This can be a change in pollution level (Kerfoot et al. 1999), eutrophication (Weider et al. 1997; Hairston et al. 1999a), predation pressure (Cousyn et al. 2001; Kerfoot & Weider 2004), or any other quantifiable change (e.g. the invasion of an exotic species and a change in temperature or salinity), but it needs to be historically well-documented. Second, there is a need for a well-preserved stratified sediment record. The sediment record should be undisturbed, and dating of different layers should be straightforward. This is essential for the translation of historical changes into depth layers in the sediment. Deep lakes with varved sediments (e.g. Lake Constance) offer excellent perspectives (e.g. Weider et al. 1997; Hairston et al. 1999a). However, the study by Cousyn et al (2001) shows that productive shallow lakes may also be amenable to analysis. Whereas there will always remain some uncertainty about the degree to which sediment records of shallow lakes tend to be disturbed, these authors did find a strongly structured dormant egg bank, as evidenced by variation in the number as well as size of the dormant eggs, and by pronounced and structured differences in genotype value of clones hatched from eggs isolated from different sediment layers. Although these observations do not exclude the occurrence of some disturbance and mixing of sediment layers, the degree of structure can only be explained by accepting that the sediment layers studied are relatively mildly disturbed at most. Such mild disturbance does not interfere strongly with the interpretation of the results, as it actually makes the estimates of genetic differentiation through time conservative. Thirdly, sufficient numbers of dormant stages have to be present in the egg bank. To the extent that this requires the use of multiple sediment cores, this introduces the additional problem of aligning sediment cores. A lower number of dormant eggs may especially be a problem in relatively nutrient-poor, deep lakes. However, as these lakes often have varved sediments, correlation of different sediment cores is often relatively straightforward. Working with neutral markers to document genetic structure requires highly variable marker loci to have enough power to detect temporal changes in allele frequencies within populations. Ideally, PCR-based DNA markers are used for such analyses.

RECONSTRUCTING MICROEVOLUTION

163

Hypervariable tandem repeat markers such as microsatellites are well-suited for this type of analysis. They require, however, proper primer design, so that currently they can only be applied to a limited set of species. Most paleogenetic studies so far have focused on the cladoceran genus Daphnia, and have been limited to Daphnia magna (Cousyn et al. 2001), species of the D. longispina species complex (Ender et al. 1996; Reid et al. 2000; Fox 2004), D. pulex (Colbourne et al. 2004; Mergeay et al. 2005a,b; Mergeay 2005), and D. barbata (Mergeay 2005). There is good potential to work on some other groups of aquatic organisms, as microsatellite markers have been developed for the bryozoan Cristatella mucedo (Freeland et al. 2000a,b), and the rotifer Brachionus plicatilis (Gomez & Carvalho 2000; Gomez et al. 1998, 2002). To study microevolutionary changes in ecologically relevant traits, one needs to engage in a quantitative genetic study, and this requires that the dormant eggs can be hatched. This may be a major bottleneck. Although hatching of old eggs has been reported (>100 years in Daphnia, Cáceres 1998; >330 years in copepods, Hairston et al. 1995), it is often found that hatching rates decline rapidly with increasing age of the eggs. So far, no study on ecologically relevant traits has gone back further in time than 80 years (Hairston et al. 1999a; Kerfoot et al. 1999; Cousyn et al. 2001), although in terrestrial plant systems this time span was about 200 years (Vavrek et al. 1991). If one of the above conditions is not met (historical record of environmental change, adequate sediment record, sufficient numbers of well-preserved dormant stages, availability of decent genetic markers, and capacity to hatch the eggs), the potential for the study of evolutionary dynamics is strongly limited. In addition to these practical issues, there is a more fundamental question as to what extent the archive in the dormant egg banks reflects the history of the system. First, there is uneasy feeling that the dormant eggs used to reconstruct microevolution are precisely those eggs that did not contribute to the population under study, as they did not hatch in the field. Analyzing these egg banks thus may result in a biased picture because of genotype-dependent hatching characteristics. In our opinion, this bias should not be exaggerated. Whether a dormant egg will hatch in the field largely depends on chance events, more specifically whether the eggs are exposed to hatching stimuli. A significant fraction of the dormant eggs is likely to get buried under too much sediment to hatch, irrespective of its characteristics and genotype. These eggs can thus be viewed as a random sample of the dormant egg bank of the population. This at least holds for neutral markers and most quantitative traits. We would caution against the use of dormant egg banks to specifically reconstruct the evolution of hatching characteristics, as the dormant egg banks may not be a perfectly random sample of the population with respect to hatching characteristics. The egg bank may contain a bias towards eggs that have deficiencies in hatching responses. In more general terms, a larger fraction of the dormant egg bank compared with the active population may suffer from such deficiencies, e.g. those associated with inbreeding depression. In most historical reconstructions, however, one may argue that the level of inbreeding depression is unlikely to vary strongly through time, as it is expected to be largely a function of population and thus habitat size.

164

L. DE MEESTER ET AL.

A second potential problem with the extent to which egg banks reflect the history of the population lies in a genotype-dependent contribution to dormant eggs. Several studies have reported that the production of dormant eggs can indeed be genotypedependent (e.g. Innes 1997). The tendency to produce dormant stages is, however, largely correlated with habitat characteristics such as predictability and permanence of the habitat (Tessier & Cáceres 2004; Cáceres & Tessier 2004b). It is therefore expected that this problem is more likely to interfere with a proper interpretation of data for among-population comparisons than for monitoring changes within one specific habitat or population. Verschuren and Marnell (1997) and Prazákova and Fott (1994) found a good correlation between changes in densities of cladoceran ephippia in sediment cores and densities of other subfossil remains (e.g. postabdominal claws), which are generally believed to accurately reflect actual zooplankton densities (Frey 1964). Jankowski and Straile (2003) found a close match between the number of ephippia in a sediment core of Lake Constance and the historical population densities reported in yearly surveys of the active Daphnia population in this lake. They also showed, however, that there were problems when comparing different taxa of the D. longispina species complex. Overall, for both within-population and within-species comparisons, it may be quite safe to assume congruence between numbers of ephippia and relative frequencies of genotypes on the one hand and historical changes in active populations on the other hand. One should be more careful while making comparisons among habitats or species within habitats (Tessier & Cáceres 2004; Cáceres & Tessier 2004a). However, reconstruction of microevolutionary responses occurs at the within-population and within-species scale. 9.4 CONCLUSIONS AND FUTURE DIRECTIONS

Whereas most studies so far have focused on cladocerans (mainly Daphnia) and copepods, there is potential for similar studies in other groups of aquatic organisms producing dormant stages, such as macrophytes, bryozoans, rotifers, protists, and bacteria. In macrophytes, applications may be limited as many species do not invest heavily in seeds but rather on shorter-lived dormant stages such as tubers. The perspective in bryozoans is different from that of crustacean zooplankton, as the resting stages of bryozoans are asexual. Yet, interesting population genetic patterns have already been reported from this group of organisms (Freeland et al. 2001, 2004), and a historical reconstruction might be very revealing. Rotifers are very promising, as they have even shorter generation times than crustacean zooplankton. Rotifer eggs have already been shown to survive up to 60 years in the sediment (Kotani et al. 2001), while microsatellite markers have been developed for at least one Brachionus species (Gomez et al. 1998). Finally, much can be expected from the analysis of dormant stages of algae and bacteria, provided that methods are optimized to isolate strains from specific depths without risk of contamination. There is a need for studies that increase our capacity to hatch old eggs, as this remains a major bottleneck in “resurrection ecology” studies. Many healthy-looking old eggs seem not to respond to hatching cues, and it is not known why they do not

RECONSTRUCTING MICROEVOLUTION

165

respond. If one could still hatch these eggs, this would increase the scope for historical reconstructions of microevolutionary responses considerably. Another way to increase the time frame over which microevolutionary responses in ecologically relevant traits can be documented is to make use of recent developments in molecular markers associated with ecologically relevant traits. It is conceivable that in the near future, both due to the efforts to obtain the full genome sequence of Daphnia as well as through quantitative trait loci (QTL) studies (sites in the genome that determine a significant proportion of the variation in a quantitative trait. See the activities of the Daphnia Genomics Consortium; http://daphnia.cgb.indiana.edu), we will have markers that will pinpoint ecologically relevant genome sections. Several recent reviews (Luikart et al. 2003; Morin et al. 2004; Vasemägi & Primmer 2005) have highlighted methods to use molecular markers to document adaptive changes in natural populations. Monitoring these markers directly on eggs in layered egg banks may reveal whether significant evolution for specific traits and gene products has occurred in the recent past. If successful, this strategy would potentially allow reconstructing microevolutionary responses over several hundreds of years, bringing the preindustrial period within reach. An interesting field of research is the study of coevolutionary arms races among different taxa that all produce dormant stages. Decaestecker et al. (2004) showed that not only Daphnia but also their parasites (epibionts and microparasites) are able to produce dormant stages, which can stay viable over considerable time periods. Ibelings et al. (2004) showed that the diatom Asterionella formosa and its parasite, the chytrid fungus Zygorhizidium planktonicum, both produce dormant stages. The production of dormant-stage banks of antagonists may permit the reconstruction of historical coevolution in a natural setting. First, cross-infection experiments can be performed in which hosts from different sediment depths can be exposed to parasites that did or did not coexist in time (contemporary vs “past” and “future” parasites). This allows a quantification of the degree to which coevolving populations genetically track each other. This approach opens avenues for the reconstruction of historical coevolutionary dynamics in a very straightforward way (Decaestecker et al., unpublished data). Second, detailed genetic studies on traits involved in host immunity and parasite infectivity, and virulence on resurrected clones and parasites can provide further insight into the mechanisms of host–parasite coevolution. Accompanied by parasite population estimates (e.g. through quantitative PCR on parasite spores in the sediments), this offers the opportunity to also tackle important questions such as the association between parasite prevalence and host genetic diversity. For many studies, the ideal design involves work on many parallel study systems. However, doing a complete quantitative genetic analysis (resurrection ecology approach) on all these systems is difficult. Here we suggest that working with population profiles may be a promising avenue. Rather than obtaining fully independent replicate observations from each clonal lineage started from a random set of hatchlings from each time frame of interest, the idea is to work with the population as the unit of interest, without replication at the clonal level. In this approach, the average genotypic value for the trait of interest of a population is quantified, using a single

166

L. DE MEESTER ET AL.

observation on each clonal lineage obtained from hatching dormant eggs as replicates. This procedure thus quantifies a population profile for a given trait, based on the genotypic values of a random sample of genetic lineages from the population. The fact that there is no need for replicate observations for each clonal lineage simplifies the experimental design considerably and allows larger sample sizes, while the resulting population profiles still allow a comparative analysis of different populations (e.g. different depth layers). In practice, the approach involves incubating sediments from a given depth layer in hatching conditions, isolating a representative number of hatchlings (e.g. 20 clonal lineages), culturing them individually for at least one or two generations to reduce the impact of maternal effects, and then measuring the trait of interest on the second or third generation of animals. Interestingly, the community of hatchlings obtained from sediments may provide insight into key environmental factors such as macrophyte cover and fish predation pressure (see Jeppesen et al. 2001a,b). Vandekerkhove et al. (2005e) provide a detailed account on the potential and methods of hatching zooplankton egg banks for biodiversity studies. Although hatched communities differ in species composition from active communities, it should be possible to generate transfer curves (a transfer function relates a given index to a variable of interest, based on a calibration analysis and validation procedure; see Jeppesen et al. 2001a,b for more information) based on criteria such as the relative abundance of large vs small species (cf. fish predation pressure) and chydorids vs daphnids (which may reflect macrophyte cover; see Jeppesen et al. 2001a,b). Once these transfer functions are established, it would be possible to hatch zooplankton communities from sediment layers, use the community composition to derive key environmental conditions for several depth layers and/or lakes, and then relate these conditions to the genetic traits of a focal species as derived from population profiles. This approach can be powerful if a sufficient number of lakes are considered (e.g. comparing the situation before and after eutrophication, or before and after recovery from eutrophication). Acknowledgments. We thank Victor Alekseev, Bart De Stasio, and John Gilbert for taking the responsibility to edit this book, and for stimulating us to contribute with a small and critical review. We thank all the people and collaborators who have contributed to the development of research on zooplankton dormant egg banks and who, through various discussions and interactions, helped shaping our ideas. We thank John Gilbert for valuable comments on an earlier version of the manuscript.

KE˛STUTIS ARBACˇ IAUSKAS

10. DOES TIMING OF EMERGENCE WITHIN A SEASON AFFECT THE EVOLUTION OF POSTDIAPAUSE TRAITS? POSTDIAPAUSE AND DIRECTLY DEVELOPING PHENOTYPES OF DAPHNIA

10.1 INTRODUCTION

Diapause is an adaptation of a great many organisms inhabiting fluctuating environments, which enables them by programmed arrest of development to survive seasonally predictable adverse conditions as well as aseasonal but in some way predictable harsh periods (Hairston 1998). When a hostile environment is avoided, the next challenge, no less important than diapause induction, is when to break developmental arrest? Obviously, when the environment is suitable for recruitment and activity. Environmental cues affecting diapause termination are known (Tauber et al. 1986; Alekseev 1990); however, they were mostly identified in laboratory experiments, thus their significance in the field is still not well enough understood (Cáceres & Schwalbach 2001). Nevertheless, the unambiguous function of environmental cues seems to be to inform the organism that conditions are suitable for life. Habitats spatially and temporarily vary in their life-supporting quality. The response to proximate cues for diapause breaking might be adjusted by internal control, such as genetic or parental control (De Meester & De Jager 1993a, b; Fox & Mousseau 1998; De Meester et al. 1998). Therefore, different strategies of diapause termination seem to be possible. Optimal dormancy strategies of aquatic invertebrates in long-term perspective have been analyzed elsewhere (Ellner et al. 1998; Easterling & Ellner 2000; Spencer et al. 2001; Cáceres & Tessier 2003). Here we will consider the pattern of diapause termination within a season and its link to postdiapause traits. For univoltine organisms, the most likely strategy is to break diapause within a season as soon as possible, in order to have enough time to finish a life cycle. For multivoltine organisms, however, a few emergence strategies are possible. If environment where exiting the diapausing organisms will occur is varying unpredictably, the most adaptive emergence strategy for a population is to disperse the recruitment over the season in order to spread the risk of occurring in adverse conditions, e.g. periods of food absence or disturbed environment. Such a strategy has been found and shown to be adaptive in a bryozoan population inhabiting an unpredictably disturbed stream environment (Callaghan 1998). If the environment where the exiting diapausing organism will occur can be predicted as favorable, or at least sufficient for survival and reproduction, the synchronization of recruitment might be adaptive. The predictability of environmental conditions during activation further implies that selection for specific postdiapause traits, which increase fitness of recruiting individuals in prospective environment, might be expected. Consequently, differences between postdiapause and directly developing phenotypes may evolve, if 167 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 167–175. © 2007 Springer.

K. ARBACˇIAUSKAS

168

they use to live under varying environmental conditions. In this chapter, we will analyze postdiapause and directly developing phenotypes of Daphnia from temperate waters in order to answer if and how much they differ, to what environments those phenotypes are fitted, which emergence strategy do their traits and performance imply, and how to qualify this phenomenon. 10.2

DAPHNIA LIFE CYCLE

The cyclic parthenogen Daphnia propagates mostly by parthenogenesis producing subitaneous eggs. When the environment deteriorates, and that may be temporarily predicted, these animals usually switch to sexual reproduction and produce diapausing eggs encased in an ephippium (Stross & Hill 1965; Carvalho & Hughes 1983). Thus, daphnids as all other cladocerans, possess an embryonic diapause. Field observations have shown that hatching from diapausing eggs in daphnids takes place in the early season during a relatively short period (Wolf & Carvalho 1989; Cáceres 1998; Hairston et al. 2000). In temperate lakes and ponds, where daphnids live, the beginning of growing season with sufficient nutrient and increasing sunlight and temperature results in a massive development of edible algae. Hence, postdiapause daphnids hatching during the early season fall into a predictable high-food environment. In contrast directly developing parthenogenetic offspring, which sustain a population during the remaining part of a season, live in unpredictable and variable conditions in terms of food quality and amounts (Sommer et al. 1986). Therefore, if hatching from diapausing eggs in Daphnia indeed is bound to early season, postdiapause and directly developing offspring occur in differing environments, which properties must select for diverse seasonal phenotypes of offspring. 10.3 NEONATES: BIOCHEMICAL QUALITY AND BODY SIZE

The quality of newborn offspring greatly determines fitness of animals in a particular environment, thus first compare postdiapause and directly developing neonates. Postdiapause offspring hatch from bisexually produced dormant eggs in which development is arrested at early embryogenesis, and resumes only when appropriate hatching cues are received. These diapausing eggs can remain viable for extended time periods under adverse conditions (Hairston et al. 1995). Directly developing offspring hatch from parthenogenetically produced subitaneous eggs. The two types of eggs definitely have a different ecological role, thus, must differ. They substantially vary in morphology and histology (Zaffagnini 1987). Analysis of the fatty acid (FA) profiles of subitaneous and diapausing eggs have shown that the diapausing eggs contain similar amounts of FAs as subitaneous eggs. However, there are several differences with respect to the particular FAs present in diapausing eggs compared to subitaneous eggs. When comparing neonates of differing origin, postdiapause offspring have higher carbon content, which is consistent with the higher content of triglyceride lipids, and contain larger nitrogen amounts indicating higher protein levels. Similarly to eggs, the amount of FAs did not differ significantly between neonate

POSTDIAPAUSE TRAIT EVOLUTION

169

types, but there were qualitative differences in the composition of polyunsaturated FAs. Considerable larger amounts of arachidonic acid (ARA) and eicosapentaenoic acid (EPA) were present in postdiapause neonates, when compared with the quantity of these two essential FAs detected in parthenogenetic neonates (Arbacˇiauskas, Fink and Lampert, unpublished data). ARA and EPA are precursors of prostaglandins and leucotriens, which mediate important processes in invertebrate physiology, and have been considered as growthlimiting factors (Stanley-Samuelson 1994). Postdiapause neonates have greater reserves of ARA and EPA during early ontogenesis (which are retained from diapausing eggs), while directly developing neonates first have to acquire ARA and EPA from their food. Hence, offspring exiting diapause have an advantage over parthenogenetic hatchlings as they can start growing (and metabolizing these essential polyunsaturated FAs) immediately. In accordance with biochemical data, the increase in body size between the first and the second instar was estimated in D. pulex for postdiapause neonates to be 5% larger as compared with parthenogenetic neonates (Arbacˇiauskas 1998). The body size of neonates is related to Daphnia fitness (Tessier & Consolatti 1991; Glazier 1992; Lampert 1993), thus the comparison of size between postdiapause neonates and parthenogenetic neonates is also of interest. The body size of ex-ephippial hatchlings is related to the length of ephippium, hence, depends upon mother’s size (Boersma et al. 2000; Arbacˇiauskas 2004b). However, neonates hatching from diapausing eggs produced by differently sized females in D. magna are relatively similar in size, whereas parthenogenetic neonates, due to maternal effects, span a much wider range of sizes. Postdiapause neonates of D. magna were closest in size to first-clutch directly developing neonates produced under high food level (Arbacˇiauskas 2004b). While, exephippial hatchlings of D. pulex were significantly larger than the first-clutch parthenogenetic neonates and even those from the later clutches produced under rich food conditions. In this species, postdiapause neonates exhibited body sizes close to those measured in the later than the first clutches under limiting food conditions (Arbacˇiauskas 2004b). Thus, D. magna and D. pulex differ in relative investment per diapausing egg. Generally, D. magna lives in environments of higher productivity than D. pulex, and observed the difference may reflect the evolutionary response to varying environment. 10.4 PHYSIOLOGY: RESPIRATION AND STARVATION RESISTANCE

The comparison of respiration between offspring origins of D. magna showed that ex-ephippial females have respiratory rates ~11% higher than those of parthenogenetic females (Fig. 10.1, Arbacˇiauskas & Lampert 2003). Metabolic activity seems to be the basic physiological trait distinguishing postdiapause and directly developing offspring of Daphnia, and affecting their life histories. Higher activity of metabolism in ex-ephippial females must be related to higher feeding rate, and obviously is adaptive during low water temperatures and abundant food, as is the case in the beginning of the season when hatching from diapausing eggs is observed. However, elevated metabolism is disadvantageous during severe food shortage. When subjected to starvation, postdiapause neonates showed significantly lower

170

K. ARBACˇIAUSKAS

Figure 10.1. Mass-specific respiration rates (mean ± 1SD) in differently sized postdiapause (closed circles) and directly developing (open circles) offspring of D. magna. Note logarithmic scale. For neonates included only data which are comparable with respect to body size and hatching environment. (Modified from Arbacˇiauskas & Lampert 2003.)

survival than parthenogenetic neonates. A 50% survival in postdiapause offspring of D. magna was even 39% shorter than that in offspring of parthenogenetic origin (5.1 vs 7.1 days, respectively) (Arbacˇiauskas & Lampert 2003). For survival comparison in this study, the fourth-clutch parthenogenetic neonates were used; thus, they were significantly larger than ex-ephippial hatchlings of D. magna. Although the initial size of neonate is important for starvation resistance (Gliwicz & Guisande 1992), it seems that elevated metabolic activity is primarily responsible for lower survival rates in postdiapause neonates under starvation, as suggested by investigation of D. pulex (Arbacˇiauskas, unpublished data). 10.5 LIFE-HISTORY: GROWTH, ALLOCATION, AND RELATIVE FITNESS

Under high food concentrations, juvenile postdiapause females grow faster and, consequently, mature at larger body sizes than directly developing females (Arbacˇiauskas & Gasiu¯naite˙ 1996). When compared with parthenogenetic neonates of close body size, ex-ephippial offspring in D. pulex matured at ~10% larger body

POSTDIAPAUSE TRAIT EVOLUTION

171

length (Arbacˇiauskas 1998). The same difference in body lengths at maturity was observed also in D. magna, despite in this comparison parthenogenetic offspring initially were twice as heavy. The faster somatic growth of postdiapause offspring, which enabled them to compensate the size difference until maturity, was observed in this species also under limited food quantities (Arbacˇiauskas & Lampert 2003). Thus, when food is not severely limiting, not the initial size of hatchlings, but the biochemical quality of postdiapause neonates and their higher metabolic activity must be responsible for the fast body size increase in postdiapause females during juvenile development. After maturity, at least during early maturity, the somatic growth of postdiapause females under rich food is slower in comparison with that in directly developing females (Arbacˇiauskas 1998). That is because of the different resource allocation pattern, which probably is the basic life-history difference that distinguishes postdiapause and parthenogenetic females. Under rich food environments, postdiapause females show significantly higher reproductive effort during early maturity. When food is limiting, the difference in early reproductive effort between offspring origins decreases, and was undetectable when comparing population samples (Fig. 10.2, Arbacˇiauskas 1998, 2004b). Measurements of reproductive investment and body condition in D. magna suggest that during early maturity the larger allocation to reproduction in comparison to body storage probably persists in postdiapause offspring also across limited, leastwise not severely limited, food environments

Figure 10.2. Reproductive effort (mean ± 1SE) over the first two adult instars in postdiapause (black columns) and directly developing (hatched columns) offspring of D. magna and D. pulex raised under high and limited food conditions. Asterisks denote significant differences between offspring origins (Tukey honest significant difference [HSD] test: D. magna, P = 0.015; D. pulex, P < 0.001).

172

K. ARBACˇIAUSKAS

(Arbacˇiauskas, unpublished data). A greater allocation to reproduction results in larger progeny numbers only under high food concentrations. Postdiapause females of D. pulex produced a first clutch which was more than twice that of parthenogenetic females (Arbacˇiauskas 1998). In D. magna, that difference comprised 1.6 times, while progeny numbers between offspring origins during early maturity were the same under limited food availabilities (Arbacˇiauskas & Lampert 2003). When compared with parthenogenetic females, the time until first reproduction in postdiapause females was estimated to be shorter for D. pulex, whereas it tended to be longer for D. magna (Arbacˇiauskas 1998, unpublished data). The relative fitness of postdiapause offspring of D. pulex under rich food was estimated to be 20% higher than that for directly developing offspring. Also, the significantly stronger negative effect of the decrease of food availability on fitness of ex-ephippial females was determined for this daphnid species (Arbacˇiauskas 1998). At high food concentrations, the larger early fecundity despite slightly longer time to reproduction resulted in higher fitness of ex-ephippial females also in D. magna (Arbacˇiauskas & Lampert 2003). In this species, however, a significant trend for higher rate of population increase was measured in parthenogenetic offspring across limited food concentrations (Fig. 10.3). Those results clearly show that postdiapause offspring in Daphnia are superior to directly developing females only under rich

Figure 10.3. Relative fitness (mean ± 1SD) of postdiapause (closed circles) and directly developing (open circles) offspring of D. magna under different food (Scenedesmus obliquus) concentrations. Across limiting food concentrations a trend for higher fitness of directly developing offspring is significant (twoway analysis of variance (ANOVA): offspring origin effect F1,12 = 5.3, P = 0.04; food effect F2,12 = 542.2, P < 0.001). Fitness under 0.1 mg C L−1 is estimated on the basis of the first clutch. (Arbacˇiauskas & Lampert 2003 and Arbacˇiauskas, unpublished data.)

POSTDIAPAUSE TRAIT EVOLUTION

173

food environments. When food is limiting, they show no advantage, and even may be loosing in fitness to parthenogenetic offspring. It is also noteworthy that across differing food environments directly developing females of D. pulex exhibited a lower than postdiapause offspring variation in reproductive characteristics related to the first adult instar, consequently, lower variability of relative fitness in response to feeding conditions (Arbacˇiauskas 2001). This pattern can be interpreted as an adaptation of parthenogenetic offspring to an unpredictability and lower quality of their environment, i.e. bet-hedging strategy (Roff 1992; Stearns 1992). 10.6 DESCENDANTS OF POSTDIAPAUSE AND DIRECTLY DEVELOPING FEMALES

Postdiapause females of D. pulex in treatments with high food in the first clutch were producing neonates whose individual size was significantly smaller than that for neonates from the first clutch of parthenogenetic females (Arbacˇiauskas 1998). Parental fitness depends upon both the number of progeny and the individual fitness of progeny in the environment where they live, therefore, the life history of descendants of postdiapause and directly developing females were examined in D. magna. Although a small but significant difference in egg characteristics was found for the first clutch produced under high food, it did not translate into size difference of firstclutch neonates and difference in fitness between offspring from ex-ephippial and parthenogenetic mothers. Postdiapause females responded in egg size to differing food in common to parthenogenetic females. Consequently, a significant postdiapause effect on fitness in successive parthenogenetic generations may not be expected in D. magna at the population level (Arbacˇiauskas 2004a). However, such effect may be present in D. pulex living at high food concentrations, and may cause some loss of fitness related to offspring quality in postdiapause females in comparison with directly developing females, as among parthenogenetic offspring of Daphnia neonate size is positively correlated with fitness (Lampert 1993). The pattern of allocation per offspring in daphnids exiting diapause may differ across species and environments, and this aspect warrants investigation. As detected already by Weismann (cit. from Alekseev 1990), the postdiapause females in Daphnia never produce diapausing eggs in the first brood, while that may be induced in the later clutches. Generally, the propensity for sexual reproduction in postdiapause generation of daphnids may be decreased in comparison with successive parthenogenetic generations, as has been shown for rotifers (Gilbert 2002) and aphids (Lees 1960). 10.7 CONCLUSIONS

So far, the comparison of traits of postdiapause and directly developing offspring in Daphnia clearly indicates that life-history patterns substantially differ between offspring origins. Although due to selection under species-specific conditions some traits of the postdiapause phenotype, when compared with phenotype of directly

174

K. ARBACˇIAUSKAS

developing individuals, might differ between Daphnia spp. (e.g. size of ex-ephippial hatchlings), offspring exiting diapause show a life-history strategy, which is distinct from that in parthenogenetic offspring. Different biochemical quality of diapausing eggs and elevated metabolic activity of postdiapause offspring (Fig. 10.1) are those traits, which are responsible for distinction in life-history pattern and relative fitness between offspring origins. Postdiapause and directly developing females show during early maturity different resource allocation strategy (Fig. 10.2). The greater allocation to progeny in postdiapause offspring, however, results in increased relative fitness only under rich food environments. On limited nutrition, when comparing population samples no advantage can be discovered over parthenogenetic females, which even may tend to be superior (Fig. 10.3). As indicated by starvation resistance, the postdiapause offspring would be at a disadvantage under severe food shortage. Thus, postdiapause and directly developing females of Daphnia are adapted to different environments. Postdiapause offspring are adapted to a favorable environment without food limitation, which is expected during early season when emergence from diapause is observed, while parthenogenetic offspring are adapted to unpredictable and highly variable food conditions, which are expected later in the season (Arbacˇiauskas 1998, 2001). This pattern, in turn, implies that within a season selection must favor in Daphnia the synchronous hatching of diapausing eggs in the beginning of the season when the spring algal bloom is to be expected, and that should be considered when analyzing seasonal development of Daphnia populations (Arbacˇiauskas & Lampert 2003). In the field, the disappearance of Daphnia from the water column during summertime and their reappearance later in the season frequently are observed. Also, estimates of “negative mortality” in studies of Daphnia population dynamics are often. The common explanation of both, Daphnia reappearance and negative mortality, is the recruitment from resting eggs bank, however, whether is it true and how much hatching from diapausing eggs after the spring burst can quantitatively affect the seasonal development of a population remain unexplored. The cyclic parthenogen Daphnia has evolved two alternative seasonal phenotypes matching environmental conditions in which they occur. Predictability of environment during emergence from diapause has resulted in different adaptation of postdiapause and directly developing offspring. The seasonal performance of Daphnia inhabiting temporal waters is consistent with the phenomenon of seasonal polyphenism, which predicts the presence of threshold traits affecting life-history pattern and fitness trade-offs across varying environments. The elevated metabolic activity of postdiapause offspring is that threshold trait, which determines distinction in life-history pattern and relative fitness between offspring origins (see Arbacˇiauskas 2004b). The transgenerational effect responsible for different quality and, consequently, different life-history pattern of postdiapause offspring in comparison with their mothers has received a definition of a negative maternal effect, which is to be expected when environment for progeny is predictable in long-term, but irrespective of parental environment (see Fox & Mousseau 1998). Among another animal taxa, the adaptive significance of narrow sense seasonal polyphenism, that polyphenism for which the predictable temporal change is sufficient

This page intentionally blank

POSTDIAPAUSE TRAIT EVOLUTION

175

(see Arbacˇiauskas 2004b), probably is best examined in butterflies. The seasonal change of wing pattern in butterflies is important for thermoregulation under temperate climate, and increases fitness of alternative phenotypes in seasons in which they occur (Brakefield 1996 and references therein). Thus, similarly to Daphnia, seasonal polyphenism of butterflies is related to modification of metabolic activity, which should condition the enhancement of resource acquisition. The adjustment of resource acquisition with respect to resource availability might be a common evolutionary trend selecting for seasonal polyphenism in multivoltine organisms inhabiting seasonal environments. Acknowledgments. I thank Professor Winfried Lampert (Institute of Limnology, Ploen, Germany) and Doctor Patrick Fink (University of Cologne, Germany) for helpful comments on the manuscript.

PIET SPAAK AND BARBARA KELLER

11. DIAPAUSE AND ITS CONSEQUENCES IN THE DAPHNIA GALEATA–CUCULLATA–HYALINA SPECIES COMPLEX

11.1 INTRODUCTION

Sexual reproduction in cyclical parthenogenetic Daphnia (water fleas) might lead to the production of interspecific hybrids. Since sexual reproduction in Daphnia is coupled with diapause, the study of diapause and diapausing eggs can clarify questions related to the frequency of hybrid production, the occurrence and strength of mating barriers, the abundance of hybrids in egg banks, and the likelihood of hybrids to colonize new habitats through diapausing egg dispersal with subsequent hatching. This chapter aims to discuss the processes related to the frequency of hybrid and backcross production within Daphnia hybrid species complexes. Important processes in this regard are the timing and induction of sexual reproduction in parental species as well as the fitness and fate of the produced offspring (i.e. diapausing eggs). The hybrid status of daphnids is difficult to detect without genetic markers, which are therefore essential tools. Furthermore, molecular tools give the possibility to study past populations by examining diapausing eggs in the sediment of lakes. In this chapter we review recent papers that have been published about several aspects of sexual reproduction within the Daphnia galeata–cucullata–hyalina species complex. The combination of field studies, experimental work, and genetic analysis gives new insights in the success of hybrid Daphnia. 11.2 HYBRIDIZATION IN DAPHNIA

Over the last two decades eight species complexes (parental species and their hybrids) have been described among Daphnia spp. from Australia, North America, and Europe (Hebert 1985; Hebert et al. 1993; Taylor & Hebert 1992; Wolf & Mort 1986). In Europe one or more taxa (parental species or hybrid) of the D. galeata– cucullata–hyalina species complex are present in many lakes (Schwenk & Spaak 1995). We will concentrate on this species complex. In theory, hybrids are able to follow several evolutionary pathways. They can persist as diapausing eggs and probably hatch later, persist as a parthenogenetic reproducing clonal lineage, become extinct, or reproduce sexually with parental species or other hybrid taxa. Life history characteristics determine the short-term “ecological success” and the niche breadth of hybrid genotypes, whereas the evolutionary significance of interspecific hybridization depends on the possibility of hybrids to spread genes to other generations (Fn or backcross generations). Most research on interspecific hybridization has concentrated on the D. galeata– cucullata–hyalina complex. This species complex has been investigated with respect 177 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 177–185. © 2007 Springer.

178

P. SPAAK & B. KELLER

to ecological differentiation, such as diel vertical migration (DVM), ecophysiology, population dynamics, and population genetic aspects (for review see Schwenk & Spaak 1995; Schwenk & Spaak 1997). The application of allozyme electrophoresis in combination with morphological investigations revealed that D. galeata, D. hyalina, D. cucullata, and their interspecific hybrids occur syntopically in various European lakes. Parental species and hybrid taxa differ in ecological characteristics and patterns of seasonal abundance (Wolf 1987; Weider & Stich 1992; Spaak & Hoekstra 1997). Based on results of earlier studies, it was believed that recombinant genotypes, such as F2 hybrids or backcrosses are rare (Wolf & Mort 1986) and introgression was suspected for only one population (Spaak 1996). However, since only a limited number of species-specific allozyme markers are available, researchers started to exploit a whole set of different molecular techniques for their use to give a higher resolution to genetic studies of hybridizing Daphnia. 11.3 GENETIC MARKERS TO IDENTIFY PARENTAL AND HYBRID TAXA WITHIN THE D. GALEATA–CUCULLATA–HYALINA COMPLEX

11.3.1 Allozymes For the study of hybrid Daphnia relevant genetic markers are essential. The morphological distinction of F1 hybrids is difficult (Flößner & Kraus 1986; Flößner 1993) and for later generation of hybrids (i.e. Fn hybrids and backcrosses), impossible. Traditionally, allozymes have been used to distinguish between the three species and hybrids of the D. galeata–cucullata–hyalina complex (Wolf 1987; Wolf & Mort 1986; Gießler 1997). Gießler (1997) showed that the Got-1 locus (now called Aat) is a discriminating marker between species and hybrids (cf. Wolf & Mort 1986). Further she demonstrated that other allozyme loci (Ao, Pep-1, amylase) show species-specific alleles, which allow more precise identification of species, hybrids, and backcrosses than with Aat alone. All species are homozygote for different alleles of Aat and Ao, which allows the distinction of various hybrid classes. The advantage of the allozyme technique is that it is fast and relatively cheap. The difficulty is that a certain amount of fresh tissue is needed. Therefore, allozymes are only of limited use for the analysis of diapausing eggs (only the nondiagnostic enzyme Pgi can be analysed), and of no use when ethanol-preserved samples have to be analyzed. The other disadvantage of allozymes is that only two species-specific markers (Aat and Ao) are available that can be used on single animals. Although Gießler et al. (1999) found species-specific alleles for Pep-1 and amylase, these enzymes have the weakness that the tissue of more than one specimen is needed to get readable bands. Thus, animals have to be first cloned to get enough material for analyses and therefore cannot be used for field surveys. 11.3.2 Microsatellites Microsatellites are among the most popular genetic markers used nowadays (for review see Zane et al. 2002). They are so popular because they are codominant and easy to interpret. Further they are based on the polymerase chain reaction (PCR)

DIAPAUSE IN HYBRID DAPHNIA

179

technique, which allows the analysis of tiny amounts of DNA, even single diapausing eggs. The disadvantage of microsatellites is that they need to be newly developed for most species examined for the first time. For Daphnia more and more microsatellites have become available. For the D. galeata–cucullata–hyalina complex, several microsatellite markers have been developed (Ender et al. 1996; Fox 2004; Brede et al. 2006). Some of the microsatellites developed for D. pulex (Palsson 2000; Lynch et al. 1999; Colbourne et al. 2004) also worked for the D. galeata–cucullata– hyalina complex (Limburg & Weider 2002). However, none of these D. pulex microsatellite markers showed diagnostic alleles (Reid 1998). There is more and more evidence that allele frequencies differ within the D. galeata–cucullata–hyalina complex in general (Brede 2003). Such alleles, unique and present in all individuals of a given species (Sandrock 2005), were also found in recently developed primers of Brede et al. (2006). However, more testing has to be done with daphnids from multiple lakes over a broad geographic area to evaluate which loci and alleles can be recommended as species-specific. 11.3.3 Nuclear Restriction Fragment Length Polymorphisms Since microsatellites require a laboratory with either a sequencer or a high-resolution electrophoresis facility there is the need for other PCR-based marker systems that can be run in less-equipped molecular laboratories. Recently, Billiones et al. (2004) developed a diagnostic nuclear restriction fragment length polymorphism (RFLP) analyses of the ITS gene region. Several nuclear loci (ITS1–ITS2, CA14, and GA13) were amplified by PCR and products were subjected to diagnostic restriction enzymes. The application of this approach to several populations across Europe revealed that markers are highly consistent and reproducible (Billiones et al. 2004; Brede 2003). The disadvantage of this technique is that only one species-specific marker is available. If later generation hybrids or backcrosses are present in a population, hybrids cannot be distinguished with this technique. But in combination with allozymes it adds an extra species-specific marker. Recently, Keller (unpublished data) found atypical MWO restriction patterns. According to the key of Billiones et al. (2004) individuals had to be classified as D. cucullata; however, allozyme markers and morphological analyses showed that these individuals were actually D. galeata. Together with colleagues from Norway and the Czech Republic we expect to solve this problem soon (Skage, Petrusek, & Keller unpublished data). 11.3.4 Mitochondrial DNA Markers Mitochondrial DNA markers cannot be used to distinguish between parental species and hybrid classes, since the mtDNA inherits maternally. Diagnostic mtDNA markers can, however, give insights into the directionality of hybridization events (who was the mother species?). Schwenk (1993) established a species-specific marker using restriction patterns of amplified cytochrome b segments. He found evidence for unidirectional hybridization in D. galeata × cucullata as well as in D. galeata × hyalina hybrids (Schwenk 1997). Using allozymes, random amplification of polymorphic DNA (RAPD), and mtDNA analysis Schwenk et al. (1998) found high abundance and clonal diversity

180

P. SPAAK & B. KELLER

of hybrid Daphnia in European lakes. These results made clear that hybridization events must occur regularly. But we do not know how often these events take place and why parental species and hybrids are still distinct. In the rest of this chapter we review recent papers that give insight into the following four questions: ● Which factors control sexual reproduction of hybrid classes and parental species? ● How likely is it that hybrids are produced regularly? ● What is the taxon distribution among asexual and sexual daphnids, and their offspring? ● Can diapausing eggs from sediment records tell us something about hybridization histories in lakes? Answers to those questions will help to understand the evolutionary role of diapause in the D. galeata–cucullata–hyalina complex. 11.4 FACTORS THAT DETERMINE SEXUAL REPRODUCTION OF PARENTAL DAPHNIA SPECIES

Laboratory studies on sex allocation in Daphnia have mainly focused on factors that are necessary to switch from parthenogenetic to sexual reproduction, besides wellknown factors such as temperature, photoperiod, and density (Carvalho & Hughes 1983; Korpelainen 1989; Hobæk & Larsson 1990). Fish kairomones have been found to induce the development of ephippia in D. magna (Slusarczyk 1995; Pijanowska & Stolpe 1996). Compared to the number of studies done on environmental factors inducing sex, relatively little work has been done on genetic variation in traits related to sexual reproduction. Intraspecific hatching variability was studied by De Meester and De Jager (1993a, b); genetic differences with respect to the induction of sex have been studied by Ferrari and Hebert (1982), Larsson (1991), Innes and Dunbrack (1993), as well as Innes and Singleton (1994). All these studies, however, done with D. pulex and D. magna showed that there is wide intraspecific variation in sex allocation. D. pulex and D. magna are typical pond species that often need to produce ephippia to survive dry summer periods or the winter when ponds freeze to the bottom. In contrast, the D. galeata–cucullata–hyalina complex inhabits large permanent lakes in which they can survive the winter without going into diapause (e.g. Keller & Spaak 2004). To investigate whether reproductive variation in the D. galeata–cucullata–hyalina complex can explain the occurrence of hybrids and backcrosses, a total of 43 clones from three north German lakes were tested by Spaak (1995) for allocation to sexual reproduction under equal stress conditions. Six replicates per clone were followed until the seventh adult instar. The following cues were used to promote sexual reproduction: short photoperiod, water from a crowded Daphnia culture, fish-conditioned water, and low food concentration. For each animal, clutch size and clutch sex were recorded. Ephippia, which were empty since the animals were cultured individually, were only produced by D. cucullata and D. cucullata × hyalina (26% and 6% of the broods, respectively), whereas almost all taxa produced males (the range was 2–15%). As for D. pulex and D. magna, mentioned before, intraspecific variation for

DIAPAUSE IN HYBRID DAPHNIA

181

male and ephippial production was found. The fact that the taxa tested showed different responses to the stimuli offered may indicate that there was reproductive isolation between them. This is supported by field data from Tjeukemeer, the Netherlands, which show that D. galeata mostly produces ephippia in spring and D. cucullata in autumn (Spaak 1995). The presence of hybrids with ephippia in both seasons in Tjeukemeer, however, shows that backcrossing is theoretically possible. In the previous experimental study (Spaak 1995) daphnids were exposed to a combination of stress factors. Therefore, it could not be distinguished whether fish kairomones alone can induce sexual reproduction in the D. galeata–cucullata– hyalina complex. Spaak and Boersma (2001) studied the combined effect of fish kairomone and food level on the production of males and sexual females in different clones of five Daphnia taxa from the D. galeata–cucullata–hyalina complex. The study was carried out in two large-scale indoor mesocosms, the “plankton towers” in the Max-Planck Institute in Plön, Germany. Although all of the Daphnia taxa produced sexual females in the course of the experiment, only D. galeata produced a significant number of males. Fish kairomones had a significant negative influence on the production of sexual females in contrast to earlier studies of Slusarczyk (1995) on D. magna. Also in another study on D. magna, Boersma et al. (1998) could not reproduce the findings of Slusarczyk (1995) in a study on 16 different D. magna clones. One of the reasons might be that they did not feed the fish with Daphnia so that no “alarm substances” were present in the water. In a follow-up study, Slusarczyk (1999) showed that this is essential to induce a response in Daphnia. Another explanation might be that reactions to fish kairomones only evolved in lakes where D. magna co-occurs with fish, which is only possible in highly turbid, mostly very eutrophic lakes. In conclusion, these studies show that intraspecific and interspecific variation occurs in the ability to produce males and sexual females in general and specifically in the D. galeata–cucullata–hyalina complex. The fact that different taxa differently react to the same cues, as found by Spaak (1995), is most likely an important reason why those taxa remain apart. 11.5 ARE HYBRIDS STILL PRODUCED?

Since daphnids of the D. galeata–cucullata–hyalina complex may overwinter in permanent lakes as parthenogenetic females there is no need for recurrent hybrid production with subsequent reestablishment from diapausing eggs. It is not even necessary that hybrids originate from the lake in which they are found. Theoretically lakes can be colonized by hybrid diapausing eggs hatching from ephippia. If hatched hybrids have a higher fitness compared with the resident population, they could become abundant. The chances for such an event are, however, very small as was pointed out by De Meester et al. (2002). To investigate the question whether, and how frequently, hybrids are produced several approaches can be applied. If hybrids are regularly produced, it must be possible to find diapausing eggs in the sediment or hybrid hatchlings in hatching traps. A prerequisite for regular hybrid production is, however, that sexual females of one species have to co-occur with males from the

182

P. SPAAK & B. KELLER

other species so that eggs can be fertilized. Several of these possibilities have been studied recently and will be discussed. 11.5.1 Are Hybrid Diapausing Eggs Present in the Sediment? The first to study in situ hatching of Daphnia diapausing eggs from lake sediments were Carvalho and Wolf (1989) and Wolf and Carvalho (1989). From their sampled diapausing eggs very low percentages hatched, and among all hatchlings they found only one D. galeata × hyalina hybrid (Carvalho & Wolf 1989). Although their sample size was low, in an experiment with hatching traps they found one hybrid that survived and could be analyzed with allozymes (Wolf & Carvalho 1989). Unfortunately they could not genotype the diapausing eggs directly, or the neonates that hatched but died before reaching a body size enabling allozyme electrophoresis. In Lake Constance, Germany, where the same species complex (D. galeata–hyalina) occurs, almost no hybrids were found that hatched from sediment cores (Weider et al. 1997; Jankowski 2002). Interestingly no D. hyalina hatched, although they are the native Daphnia spp. in this lake. Again, single hybrids were found, indicating that at least sporadically hybrids are produced in this lake. The same method, allozyme analysis of hatched daphnids from sediment cores, was used by Keller and Spaak (2004) in Greifensee, Switzerland, where also D. galeata, D. hyalina, and their hybrids co-occur. Analogous to the results in Lake Constance we found only D. galeata and backcrossed D. galeata that hatched from the sediment and no D. hyalina. All those studies show that obviously the ability to reproduce sexually is different among members of the D. galeata–cucullata–hyalina complex. In conclusion one can say that hybrid diapausing eggs are produced in the studied lakes. One might hypothesize that the fitness of hybrid hatchlings is on average lower than that of parental hatchlings. This might explain the low amount of hybrids found in hatching studies. On the other hand, the low number of hybrid hatchlings might be a frequency effect, because hybrid diapausing eggs are simply less often produced. Among hybrids that do hatch are obviously clones with a relatively higher fitness (Spaak & Hoekstra 1995), which might be the reason for the observed hybrid superiority in many European lakes. 11.5.2 Do Males and Sexual Females of Hybridizing Species Temporally and Spatially Co-occur? Since Wolf and Mort (1986) proved with allozymes that interspecific hybrids occur in mixed populations, researchers have studied the origin of hybrids. For the successful production of hybrids in general, sexual forms of both parental taxa have to co-occur. Wolf (1987) was the first to show that in two hybridizing Daphnia populations males and sexual females co-occurred. However, he did not genotype the sexual forms. Spaak (1995) found, in a 3-year study in the shallow eutrophic lake Tjeukemeer, that D. galeata produces sexual females in summer, whereas D. cucullata produces sexual females in autumn. Hybrid sexual females were found in both seasons. Males were present in both seasons. The small size of males makes it difficult to find them in life samples and to genotype them. However, Schwenk (1997) isolated

DIAPAUSE IN HYBRID DAPHNIA

183

DNA from ethanol-preserved males and applied the RAPD technique. Surprisingly he found D. cucullata males in spring 1989 and spring 1991, but only once (1990) did he view a male D. galeata in autumn. This shows that hybridization between D. galeata females and D. cucullata males is potentially possible in this lake (Tjeukemeer) in spring. The reciprocal cross in autumn would also be possible but seems less likely. Recently, several other studies (Jankowski & Straile 2004; Keller & Spaak 2004; Spaak et al. 2004) investigated genetic patterns of asexual and sexual daphnids in hybridizing populations. They all found some evidence that however the parental species tend to be separated from each other, possibilities exist for mating between the taxa. Studies on the position of sexual daphnids in the water column are very rare. We know that parthenogenetic females, e.g. D. galeata and D. hyalina, differ in their DVM behavior (Stich & Lampert 1981). Since differences in DVM behavior of males and sexual females were found in an experimental study on D. pulex (Brewer 1998) spatial isolation as a mating barrier between Daphnia spp. seemed plausible. In the only field study we know about, Spaak et al. (2004) could not find differences between the vertical migration pattern of males and sexual females of D. hyalina and the hybrid D. galeata × hyalina. 11.6 TAXON DISTRIBUTION OF ASEXUAL AND SEXUAL DAPHNIDS AS WELL AS FROM THEIR OFFSPRING

To understand the role of diapause in hybridizing Daphnia one would ideally want to know how well each of the parental taxa can enter diapause, which means that one has to understand the processes that lead to the production of sexual forms as well as about actual mating. But the same knowledge is necessary for F1 and F2 hybrids, as well as for backcrosses. Furthermore, one would like to know the clonal variability for the possibility to go into diapause for all these hybrid groups. No study exists that gives all this information at once. However, with the knowledge about all involved factors leading to the production of diapausing eggs one would better understand what the archive of diapausing eggs in the sediment tells about the past situation in the lake. For 6 years we have studied the hybrid Daphnia population of Greifensee (Spaak et al. 2001; Keller et al. 2002; Keller & Spaak 2004). We have tried to answer some of the questions listed above. Our major interest was to test the hypothesis that the taxon distribution (hybrids and parents) in a certain sediment layer represents the taxon distribution in the lake during the time of sedimentation. We monitored the taxon distribution of parthenogenetic females in the lake, as well as of males and sexual females. Furthermore, we collected recently produced ephippia on the lake surface. We then tried to hatch the diapausing eggs (after a dormancy period) and compare the taxon frequency of hatchlings with the pelagic lake population. Since not all ephippia hatch, and some hatchlings die within a few hours or days, we used PCR-based methods to determine the taxon of these fractions, as not enough material is available for allozyme analysis. We found significant differences between the

184

P. SPAAK & B. KELLER

genetic composition and the backcross level of pelagic asexual females, sexual females, males, and hatchlings (Keller & Spaak 2004). The asexual daphnids were dominated by hybrids. In contrast, sexual females, especially Daphnia hatched from diapausing eggs, were dominated by D. galeata. We concluded that hybrid Daphnia have a lower sexual reproductive success than the parental D. galeata. The recent hybrid dominance in the pelagic populations suggests that D. galeata hatched from diapausing eggs are not able to alter the pelagic population. Since the genotypic class composition of diapausing eggs does not reflect the extant pelagic population, our data do not support the hypothesis that the Daphnia diapausing egg banks represent the lake taxa structure. 11.7 CAN THE SEDIMENT TELL US SOMETHING ABOUT PAST HYBRIDIZATION EVENTS?

Apart from the studies that analyzed hatched daphnids from sediment cores, there are several studies that used PCR-based DNA techniques on diapausing eggs to trace back the Daphnia population structure in time. Using mtDNA Reid et al. (2000) could show that in a lake that is now dominated by D. galeata, both D. hyalina and D. galeata occurred. Because an mtDNA-based marker was used no hybrids could be distinguished. In Belauersee, northern Germany, with known co-occurrence of all three species from the D. galeata–cucullata–hyalina complex in the asexual population (Spaak 1995; Wolf 1987), Limburg and Weider (2002) used microsatellites on diapausing eggs that dated back 200 years. Due to the lack of diagnostic microsatellite markers, all diapausing eggs were pooled together and analyzed. The authors found strong shifts in genetic diversity in time, which could be an indication for species shifts. Recently, Brede (2003) analyzed diapausing eggs from a sediment core from Lake Constance using nuclear and mitochondrial DNA-based molecular markers. Using nuclear RFLP (Billiones et al. 2004) she could show that up to the 1950s only D. hyalina diapausing eggs were present in the sediment. From the 1960s the diapausing eggs in the sediment consisted mostly of D. galeata, but several hybrids were also found. This more or less represents the historical record of the lake (Muckle & Mucklerottengatter 1976; Einsle 1978), although the results of Brede (2003) suggest that hybridization took place in this lake earlier, as expected. The use of powerful (diagnostic) markers like microsatellites is just at the outset. Since many more markers are now available for the D. galeata–cucullata–hyalina complex (Brede et al. 2006) and since these samples can be analyzed on automatic sequencers with large samples, it will be possible to do comparative studies on Daphnia diapausing eggs from multiple lakes in the coming years. It will also be possible to study multiple sites within a lake, and see how diapausing eggs of different taxa are spatially distributed. Furthermore, these techniques will allow much more detailed analyses of hybrid and backcrossed genotypes. At the moment, we only have rough estimates of the real level of backcrossing in natural Daphnia populations. Using adequate numbers of microsatellite markers simultaneously will undoubtedly unravel these data soon (Boecklen and Howard 1997).

This page intentionally blank

DIAPAUSE IN HYBRID DAPHNIA

185

11.8 CONCLUSIONS

Diapausing eggs are fascinating “time capsules” that allow researchers to study populations from decades to centuries ago. Especially in the case of Daphnia hybridization this is a very valuable tool. Since hybridization was not really recognized before the mid-1980s, no specific samples were stored. DNA from archived diapausing eggs made it possible to reconstruct what happened in the past. By hatching daphnids from these buried diapausing eggs it is even possible to test in the laboratory how past and present hybrid and parental taxa cope with different environmental factors. “Resurrection ecology” as it is called by Kerfoot and Weider (2004) allows direct testing of a whole set of evolutionary hypothesis. We are sure that in the future also Daphnia diapausing eggs will be a favorite model for plankton ecologists as well as evolutionary biologists. Acknowledgments. We thank Victor Alekseev for the invitation to participate in this book. Only because of his encouragement and patience could this chapter be written.

VADIM E. PANOV AND CARLA CACERES

12. ROLE OF DIAPAUSE IN DISPERSAL OF AQUATIC INVERTEBRATES

12.1 INTRODUCTION

Diapause can be a key determinant of dispersal ability in aquatic invertebrates. This is especially true for zooplankton that produce diapausing eggs. For decades, anecdotal evidence has suggested the overland movement of zooplankton dormant eggs by vectors such as wind, water, and vertebrates, and recent experimental and genetic evidence has supported this claim. During the last century, the role of human-mediated dispersal of zooplankton has increased, specifically with regard to shipping. Generally, the role of human-mediated vectors is most important for species dispersal across geographical barriers and into large aquatic ecosystems that experience considerable shipping traffic. Diapausing stages facilitate species survival during movement across geographical barriers under extreme conditions, such as in ballast tanks of ships. Once in their new environment, some cladocera show altered seasonal phenologies, switching to both early and prolonged gamogenetic reproduction, which facilitates invasion success and further dispersal into novel ecosystems by both natural and human-mediated vectors. Most organisms live in habitats that vary in space or time. One way in which organisms respond to this variability is through dispersal, which can play a key role in altering relative fitness and influencing ecological and evolutionary dynamics (Gadgil 1971; Levin et al. 1984; McPeek & Holt 1992; Ricklefs & Schluter 1993). A second way in which many organisms respond to environmental variability is through diapause (Tauber et al. 1986; Brendonck et al. 1998), which can also be thought of as dispersal through time (Venable & Lawlor 1980; Hairston 1998). In aquatic invertebrates, the active stage often cannot survive for long periods out of water. Hence, the production of desiccation-resistant propagules such as eggs, cysts, gemmules, or statoblasts (embryonic diapause sensu Alekseev & Starobogatov 1996) has been frequently considered an adaptation for dispersal (Karlson 1992; Korovchinsky & Boikova 1996; Fell 1998; Hairston 1998; Brendonck & Riddoch 1999; Bilton et al. 2001). Despite a long history examining the production of diapausing stages in aquatic invertebrates, most studies have focused on conditions that promote the occurrence of resting eggs, factors that affect their survival and hatching from sediments, the presence of egg banks in sediments, and the impact of hatchlings from resting eggs on plankton community structure (Marcus 1996; Brendonck et al. 1998). Moreover, because so many aquatic invertebrates produce resistant stages, dispersal and colonization ability has often been assumed to be “rapid and frequent” (Brooks & Dodson 1965; Pennak 1989; Lampert & Sommer 1997). This assumption has been supported 187 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 187–195. © 2007 Springer.

188

V. PANOV & C. CACERES

by decades of anecdotal evidence suggesting the overland dispersal of freshwater aquatic species (Darwin 1859; McAtee 1917; Lansbury 1955; Maguire 1959; Proctor 1964; Proctor & Malone 1965; Swanson 1984; Bohonak & Whiteman 1999). Only recently, however, have studies begun to focus explicitly on importance of resting stages for dispersal of aquatic invertebrates (Bilton et al. 2001; Cáceres & Soluk 2002; Panov et al. 2004; Vandekerkhove et al. 2005e). In this chapter, we focus on impacts of diapause on dispersal of aquatic invertebrates, primarily on planktonic organisms. We review major mechanisms and vectors of dispersal, both natural and human-mediated, of diapausing invertebrates. Also, we discuss the importance of diapausing stages in anthropogenic introductions and invasion success of aquatic invertebrates. 12.2 MECHANISMS AND VECTORS OF DISPERSAL OF DIAPAUSING INVERTEBRATES

Diapausing aquatic invertebrates (including their resting stages) can be dispersed by natural (surface water connections, ocean currents, wind, and animals) and humanmediated vectors. Dispersal by human vectors are usually viewed as “introductions,” either intentional or unintentional, and such vectors have been discussed by many authors (Carlton 1996; Bilton et al. 2001; Minchin & Gollasch 2002; Havel & Shurin 2004). Sections 12.2.1 and 12.2.2 briefly summarize natural and human-mediated vectors most important for dispersal of diapausing aquatic invertebrates. 12.2.1 Natural Vectors of Dispersal Natural mechanisms and vectors of dispersal of aquatic invertebrates have been discussed in detail in recent reviews by Bilton et al. (2001), Bohonak and Jenkins (2003), Havel and Shurin (2004), and Panov et al. (2004). Diapausing eggs or cysts of aquatic invertebrates have been often considered as potential agents of dispersal by natural vectors, and even depicted in the literature as adaptations for dispersal (Maguire 1963; Korovchinsky & Boikova 1996). Several authors have suggested that diapausing eggs may be transported by wind and rain (McAtee 1917; Maguire 1963; Cáceres & Soluk 2002). However, the scale at which this mode of dispersal is most effective remains unresolved. Wind dispersal of anostracan eggs may result only in their short-distance transport (Brendonck & Riddoch 1999). In general, longdistance dispersal of diapausing stages as aerial plankton is unlikely (Jenkins & Underwood 1998; Bilton et al. 2001; Bohonak & Jenkins 2003). In contrast, currents in marine ecosystems and rivers may be important vectors for long-distance dispersal of active and diapausing invertebrates (Minchin & Gollasch 2002; Havel & Shurin 2004). Dispersal of gemmules of sponges by currents (as well as by fish and waterfowl) is discussed by Fell (1998). A midsummer shift toward sexual reproduction was observed by Makrushin (1984) in populations of the marine cladocerans Podon leuckarti and Evadne nordmanni in the Northern Atlantic. He hypothesized that the ability to produce resting eggs throughout most of the summer season was an adaptation to the large-scale dispersal of Podon and Evadne by oceanic currents, which are an important natural dispersal vector for marine

ROLE OF DIAPAUSE IN INVERTEBRATES’ DISPERSAL

189

organisms. Shanks et al. (2003) compiled available information on the dispersal distance by currents of the propagules of benthic marine organisms, and found a significant positive correlation between the duration of propagules in plankton and the dispersal distance, which ranged from minutes to months and meters to thousands of kilometers, respectively. In freshwater, zooplankton may be transported among systems during periods of overflow (Michels et al. 2001). Short-distance transport of resistant diapausing eggs in fish stomachs is also potentially possible (Jarnagin et al. 2000), as well as by terrestrial animals (Maguire 1963), but the significance of these vectors in nature is likely limited. However, fish guts as vector of dispersal of diapausing invertebrates can be important if coupled with human-mediated introductions of fish. For instance, ephippia of Daphnia lumholtzi might have been introduced first to one of the lakes of the southern USA in 1983 with the intentionally introduced Nile perch (Lates niloticus) from Lake Victoria, Kenya (Havel and Hebert 1993). Vertebrates have also long been considered to be a primary vector of dispersal for aquatic invertebrates (Darwin 1859). Transfer of resting eggs by waterfowl can be considered a more effective vector of dispersal of invertebrates in inland waters (Bohonak & Jenkins 2003); however, direct evidence of its importance is also limited (but see Figuerola et al. 2005). Eggs may be dispersed by birds with ingested food (Charalambidou et al. 2003), by sticking to their legs and bills, within the plumage. A recent review by Figuerola and Green (2002) showed that bird-mediated transport of propagules of aquatic invertebrates is a frequent process, but limited to local spatial scales. Low resistance to desiccation of some resting stages (Fell 1998) may limit distance of dispersal of diapausing aquatic invertebrates by waterbirds (Figuerola & Green 2002). However, studies of genetic distributions of some zooplankton species along with analysis of major waterfowl flyways suggests a potentially significant role of birds in long-distance intracontinental dispersal of some cladocerans and bryozoans (Taylor et al. 1998; Freeland et al. 2000a, b). Assumptions of the significant role of dispersal of resting stages of freshwater invertebrates by natural factors were recently criticized in the review by Bohonak and Jenkins (2003), which suggested that genetic and direct experimental studies failed to demonstrate evidence of effective passive dispersal, specifically by wind (Jenkins 1995; Jenkins & Underwood 1998). Clearly, zooplankton are not uniformly “good” dispersers. Rather, species both differ in their dispersal ability and the primary vectors by which they are dispersed (Jenkins 1995; Jenkins & Buikema 1998). 12.2.2 Human-mediated Dispersal Human-mediated dispersal vectors (introductions) are broadly classified into two main categories: deliberate or intentional introductions and unintentional introduction. Regarding the first type of vector, Alekseev (1986) suggested that intentional transfers of crustaceans in latitudinal directions are hindered by genetically fixed differences in timing of diapause, and successful acclimatization is more likely for crustaceans transferred after completion of diapause. In some cases, intentional longdistance transfer of target aquatic organisms for stocking purposes can be coupled

190

V. PANOV & C. CACERES

with unintentional introductions of other organisms possessing diapause (see case study of Daphnia lumholtzi in section 12.2.1). Recreational and commercial boating has been shown to be an important vector for both short- and long-distance dispersals of aquatic invertebrates and their diapausing stages for inland waters of North America (Buchan & Padilla 1999; Havel & Stelzleni-Schwent 2000; Johnson et al. 2001). Some intercontinental transfers of diapausing eggs of aquatic invertebrates and their subsequent invasions of inland waters have been linked to the introduction of commercially useful plants (McKenzie & Moroni 1986), importation of industrial equipment (see review by Havel & Shurin 2004), and even transportation of military amphibian vehicles as has been suggested by Flössner and Kraus (1976) in the case of Daphnia parvula Fordyce (Cladocera: Daphnidae) accidental introduction in European inland waters from North America. However, at present the majority of biological invasions into coastal and even inland waters worldwide can be linked to unintentional introductions via different shipping-related vectors: construction of canals, ship’s hull fouling, and ballast water release. Ballast water of ships is a principal vector of global long-distance transfer of aquatic invertebrates and their resistant resting stages, which readily breaches geographic barriers to dispersal and gene flow (Carlton & Geller 1993). Results of several ballast water studies indicated high biological diversity of aquatic communities within ballast water; these communities were often dominated by crustacean taxa (Carlton & Geller 1993; Gollasch et al. 2000; Gollasch et al. 2002). It is likely that harsh conditions in ship’s ballast tanks (darkness and rapid changes in water temperature) may result in induction of embryonic, larval, or even adult diapause in some crustaceans, and thus facilitate their survival during ship journey (Panov et al. 2004). Resting eggs and cysts of algae and aquatic invertebrates have been frequently reported in ballast tank sediments. These sediments vary from a few centimeters to more than 30 cm depth (Hamer 2002), resembling the upper layers of lake, estuarine or sea-bottom sediments. Bailey et al. (2003) recorded the presence of invertebrate diapausing eggs in residual sediments from transoceanic vessels and experimentally studied viability of the eggs collected from ballast tanks on vessels operating on the North American Great Lakes. In this study, 17 cladoceran, copepod, and rotifer taxa hatched from these sediments have been identified, demonstrating that diapausing eggs in sediments may survive treatment of ballast tanks with oceanic water and could potentially hatch in dark ballast tanks if freshwater were added. It is also likely that the observed trend of increasing role of crustacean invaders from marine to freshwater ecosystems can be a result of high susceptibility of the latter to invasions of cladocerans and copepods, e.g. taxa possessing strong diapause (Panov et al. 2004). This phenomenon may favor the hypothesis that the evolutionary appearance of embryonic diapause in crustaceans may facilitate the penetration of marine crustaceans into inland waters (Hairston & Cáceres 1996; Hairston & Bohonak 1998). In an assessment model of the risk of future introductions of aquatic species with ballast waters into the Great Lakes, considering species’ invasions histories, shipping traffic patterns, and physicochemical factors that constrain species survivorship during ballast-mediated transport, Grigorovich et al. (2003) identified 26 high-risk

ROLE OF DIAPAUSE IN INVERTEBRATES’ DISPERSAL

191

species. Among them 24 species belong to crustaceans, with more than half the species possessing embryonal, larval or adult diapause (seven species of Cladocera and six species of Copepoda). The authors suggested that ability to possess diapause, and/or parthenogenetic reproduction and short generation time in these taxa, already introduced into the Great Lakes with ballast water, have fostered their survival during ballast-mediated transfer and ensured rapid population growth in the recipient ecosystems. The results of the genetic analysis of European and North American lineages of freshwater cladocerans indicate that human-mediated vectors of dispersal may affect extraordinarily rates of intercontinental species dispersal: the current rates of species invasions are nearly 50,000 times higher than historical levels (Hebert & Cristescu 2002). It is important that most of these recent human-mediated intercontinental invasions originated, most likely, from transport of diapausing resting stages in ballast tanks of ships as in the recent case studies for cercopagid cladocerans Bythotrephes longimanus (Cladocera: Cercopagidae) and Cercopagis pengoi (Cladocera: Cercopagidae), both invasive species possessing prolonged embryonic diapause as adaptation to dispersal and invasion success (Panov et al. 2004). These two predatory cladocerans have a complicated invasion history both in European and North American inland waters, mediated by the multiple long- and short-distance dispersal vectors (see Panov et al. 2004 for review). Most probably, Bythotrephes and Cercopagis were transferred from Europe to America with the ballast water of cargo vessels, presumably as diapausing resting eggs (Lehman 1987; MacIsaac et al. 1999). Genetic studies revealed that in both cases the eastern Baltic Sea basin (Lake Ladoga and Neva River estuary) served as a source area of the initial invasion to North American Great Lakes, indicating presence of the invasion corridor between eastern Baltic and Lawrentian Great Lakes (Cristescu et al. 2001; Berg et al. 2002). In North America Bythotrephes is mainly spread by boaters and anglers attached to equipment such as fouled fishing lines, boat anchor lines, downrigger cables, via infected bilge water and live well water, and live minnow bait, which contain females bearing resting eggs (Jarnagin et al. 2000; MacIsaac et al. 2004). MacIsaac et al. (2004) revealed that species spread occurred via a combination of dominant, local diffusion (median distance 12.5 km) and rare, long-distance (>100 km) dispersal. For example, one invaded lake (Muskoka Lake, Ontario) apparently served as an invasion “hub,” resulting in up to 18 additional direct and 17 indirect invasions (MacIsaac et al. 2004). Rapid short-distance dispersal of invasive cladocerans is likely facilitated by changes in their reproductive strategy (Fig. 12.1). As most freshwater Cladocera, in native habitats Bythotrephes usually switches from parthenogenetic to gamogenetic reproduction at the end of a season and even then the densities of males and females with resting eggs are relatively low (Straile & Hälbich 2000). However, in new invaded habitats this species showed prolonged embryonic diapause. For instance, during some years in the eastern basin of Lake Erie, Ontario, Canada, males and ephippial females of Bythotrephes appeared in midsummer (Garton et al. 1993). The most remarkable changes in sexual reproduction of this species were noted in

192

V. PANOV & C. CACERES

Figure 12.1. Heterogonic reproductive cycle in Cladocera: parthenogenetic vs amphigonic reproduction.

Harp Lake, Ontario, Canada, where Bythotrephes appeared in the early 1990s. In 1994 and 1995, females with resting eggs were noted from July, and by the end of August 50–80% of females carried resting eggs (Yan and Pawson 1998; Yan et al. 1992; N. D. Yan and T. W. Pawson, personal communication). Later, in 1998, the Harp Lake population returned to a more typical parthenogenetic mode of reproduction during summer: first females with resting eggs were recorded in late August (Yan et al. 2001). Like invading Bythotrephes in Harp Lake in the first 2 years after its first record, a C. pengoi population newly established in the easternmost Baltic Sea (Neva estuary) also showed this remarkable reproductive strategy, producing a large number of resting eggs during summer months during the first years after the invasion (Krylov & Panov 1998). As for Bythotrephes in Harp Lake, the mean seasonal percentage of both males and gamogenetic females in Cercopagis population in the Neva estuary gradually declined during following years after invasion (Panov et al. 2004).

ROLE OF DIAPAUSE IN INVERTEBRATES’ DISPERSAL

193

It has been suggested that the large pool of resting eggs in the Neva estuary Cercopagis population has enabled this species C. pengoi to achieve fast population growth in new environments, and an increasing risk of C. pengoi being dispersed by ships’ ballast water (Panov et al. 1996; Panov et al. 1999). In summer 1998, soon after its establishment in the eastern Baltic, C. pengoi was first found in the North American Great Lakes (MacIsaac et al. 1999; see also above). Populations of C. pengoi in Lake Ontario during first years after invasion also possessed midsummer sexual reproduction (Grigorovich et al. 2000; Makarewicz et al. 2001), characteristic for the source population in the Baltic Sea (Neva estuary), and for Bythotrephes in Harp Lake (see above). Despite existing potential of waterfowl to transfer resting eggs of Cercopagis, boaters and ballast water of ships are considered as primary vectors of C. pengoi dispersal in the Great Lakes area (Makarewicz et al. 2001). Potential for dispersal with fishing equipment for Cercopagis is even higher, than for Bythotrephes, because of the specific morphological feature of its caudal appendage, which is longer and possesses a terminal loop (this feature is reflected in North American common name of C. pengoi: “fishhook waterflea”). It is important to note that the invasion of C. pengoi to the Laurentian Great Lakes has been taken place after implementation of ballast water management options for the ships entering Great Lakes, namely exchange of ballast water in open ocean, which is considered to be an effective measure to decrease risk of transfer of freshwater organisms. However, as has been shown by Bailey et al. (2003), resting eggs of freshwater invertebrates may hatch from the ballast water sediments, even those that have previously been exposed to salt water. Thus, the example of C. pengoi invasion to North America demonstrates the limited effectiveness of ballast water exchange programs in preventing introductions of aquatic invertebrates producing resting eggs, which may accumulate in sediments of ballast tanks (MacIsaac et al. 1999). Enhanced gamogenetic reproduction (prolonged embryonic diapause) during first years after invasion was also recorded in D. parvula (Riccardi et al. 2004), the North American species most likely introduced in European inland waters as ephippia by such specific long-distance human-mediated dispersal vectors as military amphibian vehicles (Flössner & Kraus 1976). According to Flössner (2000), both natural (waterbirds and surface water connections) and human-mediated vectors (e.g. transport and introduction of fish fingerlings; water transport related to the excavation of artificial basins and canals) are responsible for the quick dispersal of this species from southern Germany to other European locations. 12.3 CONCLUSIONS: GENERALIZED MODEL OF DISPERSAL OF AQUATIC INVERTEBRATES WITH PROLONGED DIAPAUSE

The case studies discussed earlier about successful invaders into inland waters of Europe and North America suggest the possible importance of prolonged (embryonic) diapause for effective short- and long-distance (both inter- and intracontinental) dispersals of aquatic invertebrates. These invasive cladocerans were transferred across geographic barriers (Atlantic Ocean) by different human-mediated vectors possibly by means of their diapausing eggs, exposed to adverse conditions during

194

V. PANOV & C. CACERES

transfer, and survived even possible ballast water management options (in case of C. pengoi) and quarantine treatment (in case of D. lumholtzi). Their following rapid short-distance transfer and, less frequent, long-distance transfer by mainly multiple human-mediated vectors might be attributed to their life-cycle patterns in “infected” ecosystems: rapid development of a large pool of diapausing eggs in populations, and, in some cases, even switching from “normal” pattern with short period of gamogenetic reproduction after a prolonged period of parthenogenetic reproduction, to the early and prolonged gamogenetic reproduction. A generalized conceptual model of dispersal patterns in aquatic invertebrates with such a reproduction strategy, involving a high level of development of resistant diapausing resting eggs, is represented in Fig. 12.2. Our model can be considered a variant of the human-vectored invasion model, initially suggested by MacIsaac et al. (2001) (dispersal is determined by the probability of propagule movement by humans from the source to the recipient site), with incorporation of adaptive reproduction strategies in invasive species, which increases the probability of dispersal and successful establishment in the novel ecosystem. Rapid development of large pools of resting eggs in bottom sediments might facilitate invasion success of these species, and their rapid integration into local plankton communities. Diapausing eggs allow an escape from competition (Cáceres 1997), which is likely severe for most invading species. This refuge from competition in the sediment egg bank likely facilitates the rapid establishment of invasive species (Panov et al. 2004). There is also evidence for evolution of life-history traits of invasive microcrustaceans with respect to diapause: while the initial colonizing population appears to possess early prolonged production of diapausing eggs, this characteristic erodes over time. Although in some cases a midsummer switch toward sexual reproduction in Cladocera can be considered to facilitate dispersal by natural factors in marine ecosystems (see above), early prolonged production of resting eggs recorded in B. longimanus, C. pengoi, and D. parvula during first years after their

Figure 12.2. Generalized model of dispersal of aquatic invertebrates with prolonged diapause. (After Panov et al. 2004.)

This page intentionally blank

ROLE OF DIAPAUSE IN INVERTEBRATES’ DISPERSAL

195

invasion into some lake and/or estuarine ecosystems and following fast “erosion” of this phenomenon in B. longimanus and C. pengoi can be attributed, most likely, to rapid microevolution processes characteristic for these polymorphic species in highly variable environments. As we suggested earlier for C. pengoi in the Neva estuary (Krylov & Panov 1998), switch to prolonged period of sexual reproduction in the introduced population can be attributed to some kind of the effect of “founder population.” The probability of clones with an extended gamogenetic phase being pumped into ships’ ballast tank is much higher than for the strains possessing the more “normal” life cycle with a comparatively short period of sexual reproduction. Fast erosion of this phenomenon in Bythotrephes in Harp Lake and Cercopagis in Neva estuary most likely related to fast evolution of the life cycle of invading species under strong selection. In some cases, this was most likely from invader-selective fish predation: by lake herring Coregonus artedii in Harp Lake (Coulas et al. 1998) and by the Baltic herring Clupea harengus membras in the Gulf of Finland (Antsulevich & Välipakka 2000). Fast dispersal and successful establishment of several exotic zooplankton species in inland waters of Europe and North America demonstrate the high potential for dispersal in aquatic invertebrates that develop resistant propagules. Colonization ability may also be facilitated by the ability of invertebrates with prolonged diapause to “travel in time” for persisting through adverse environmental conditions and to use the “storage effect” of egg banks to avoid competition with native species or clones. Our review suggests that diapause in some taxa of aquatic invertebrates might play a crucial role in their dispersal and colonization success in recipient ecosystems, with human-mediated vectors of dispersal acting as a powerful selective force. Human-related selection factors may facilitate dispersal of species with a high level of gamogenetic reproduction, while natural selection in novel habitats may result in fast erosion of this “founder population” effect toward “normal” life cycle with a prolonged period of parthenogenetic reproduction. Acknowledgments. This study has been supported by the European Commission 6th Framework Programme Integrated Project ALARM (contract No GOCE-CT-2003506675).

EGOR S. ZADEREEV

13. THE ROLE OF WITHIN-TROPHIC-LEVEL CHEMICAL INTERACTIONS IN DIAPAUSE INDUCTION Basic and Applied Aspects

13.1 INTRODUCTION

Many species of zooplankton develop in cycles determined by their ability to enter diapause at a specific stage of life. Embryonic diapause – the arrest of development of an animal at the embryo stage – is common in Cladocera (Alekseev 1990). During the diapausing stage the embryo is encysted as an ephippial (or resting) egg. Most often the production of resting eggs is a result of a change of the mode of reproduction from parthenogenesis to gametogenesis. Parthenogenetic females produce female and male offspring, while during gametogenesis they produce ephippial eggs, which after the resting period hatch into females. Resting eggs of Cladocera are able to survive drying and freezing. Resting eggs often form an egg bank at the bottom of the water body. Due to the ability of resting eggs to hatch after tens and hundreds of years (Hairston et al. 1995), banks of resting eggs serve as a source of genetic diversity and can play an important role in the dynamics of zooplankton communities (Brendonck & De Meester 2003). There is also at least one important applied aspect of diapause in zooplankton. Many zooplankton species are used in aquaculture as food when culturing early larval stages of fish (Vedrasco et al. 2002). As a result, diapausing eggs are an important commercial resource that is required in order to hatch zooplankton juveniles in sufficient amounts under controlled conditions for rearing larval fish. However, the biotechnology of mass production of resting eggs is not well developed yet. Consequently, factors that control the change of the reproductive mode and the production of resting eggs are a subject of continuous research. One of the most intriguing research questions is the role of chemical interactions in the induction of diapause. Several studies have demonstrated that the production of diapausing eggs in Rotifera (e.g. Stelzer & Snell 2003), Cladocera (e.g. Slusarczyk 1995; Zadereev & Gubanov 1996), and Copepoda (e.g. Ban & Minoda 1994), could be induced by the effect of chemicals exuded by conspecifics, competitors, or predators. In this contribution the focus will be placed on the effect of withintrophic-level chemical interaction on diapause induction in zooplankton. Within-trophic-level chemical interactions include the effect of chemicals produced by conspecifics and chemicals produced by other species at the same trophic level (usually competitors). We will briefly review the data and methodologies used to investigate the effect of intraspecific and interspecific chemical interactions on diapause induction, analyze the adaptive value and discuss several applied aspects of this phenomenon. 197 V. R. Alekseev et al. (eds.), Diapause in Aquatic Invertebrates, 197–206. © 2007 Springer.

198

E. ZADEREEV 13.2 THE EFFECT OF CHEMICAL INTERACTIONS ON DIAPAUSE INDUCTION AT THE INDIVIDUAL LEVEL

Banta and Brown (1929a, b, c) performed the first intensive and now classical research on the effect of chemicals exuded by conspecifics on the induction of gametogenesis. Two important results should be highlighted. First, the authors maintained a population of Moina macrocopa for 780 parthenogenetic generations under laboratory conditions, thus disproving the hypothesis of Weismann (1876) that a fixed, heritable internal cycle causes life-cycle changes of the mode of reproduction in Cladocera. Second, they performed a number of experiments using crowded water as a culture medium which were basically similar to the common experimental design employed in this type of research today. After the intensive experiments with M. macrocopa the researchers came to the following conclusions: 1. The main cause of male production is the accumulation of metabolic by-products in the medium proportional to the density of the population. 2. The production of male offspring is closely connected with (or determined by) the reduction of growth rate of the mothers, which is also proportional to the population density. 3. Metabolic by-products, which induce male production are nonvolatile, unstable, and disappear even after the storage of crowded water. 4. Most probably there is no causal relationship between the accumulation of metabolic by-products and production of resting eggs. 5. The possible stimulus for the production of resting eggs is food limitation (Banta & Brown 1929a, b, c). Later, Stross (1987) hypothesized two-factorial control of gametogenesis induction. Reviewing the induction of gametogenesis in Daphnia he concluded that photoperiod is a regulatory factor that stimulates females to produce resting eggs. However, secondarily, some density-dependent factor is also very often required to stimulate gametogenesis. Laboratory experiments demonstrated that the number of ephippial eggs produced was in direct proportion to population density. The author pointed out that this density-related stimulus had not been determined yet and most probably was not associated only with the depletion of food (Stross 1987). Recent results describing the effect of conspecific chemicals on diapause induction confirmed that the detected responses are density-dependent. For example, the rate of mixis in the rotifer Brachionus plicatilis increased with the population density used to prepare crowded water (Stelzer & Snell 2003). Zadereev and Gubanov (1996) demonstrated that the water crowded by conspecifics stimulated the change of the reproduction mode in single females of M. macrocopa. The proportion of gametogenetic females increased with the density of the population used to prepare crowded water. However, an increase in food concentration reduced the strength of the crowded water effect. The number of ephippial females was reduced to zero with increasing food concentration for all tested population densities. It should be mentioned that the higher the population density used to prepare crowded water, the higher the food concentration needed to prevent gametogenesis induction (Zadereev & Gubanov 1996).

CHEMICAL INTERACTION AND DIAPAUSE

199

Working with the third major zooplankton order, copepods, Ban and Minoda (1994) demonstrated that females of the calanoid copepod Eurytemora affinis that were individually reared in water crowded by conspecifics produced diapausing eggs, and Walton (1985) found that crowding induced diapausing egg production in a diaptomid copepod. Even though there is strong experimental evidence that chemicals exuded by conspecifics induce production of resting eggs, the role and nature of the chemicals involved remain to be understood. It is worth analyzing them based on the theoretical assumptions and considerations, as well as for practical reasons. As the aforementioned examples demonstrate, the induction of diapause under the effect of chemicals from conspecifics can be considered as a density-dependent population control factor. Such control has a twofold purpose, important from both ecological and evolutionary points of view. First, populations survive in adverse environments because of diapause, and theory indicates that diapause induction should come one generation before the onset of harsh conditions. This prediction was successfully accounted for in a study of the role and nature of chemicals involved in diapause induction by Hairston and Olds (1987) determining environmental factors inducing diapause in a freshwater copepod. Experimental data on the multifactorial control of gametogenesis induction demonstrated that metabolic by-products do indeed act to induce gametogenesis before the onset of unfavorable conditions. It has also been shown that under the effect of metabolic by-products the threshold for induction of gametogenesis cued by photoperiod (Stross 1987) and food concentration (Zadereev & Gubanov 1996) increase. When favorable environmental conditions are reestablished, the diapausing organisms may be reactivated and the population cycle can repeat again. Under certain conditions, the population with a bigger number of diapausing organisms at the reactivation stage will have a competitive advantage. This should be most pronounced for the inhabitants of temporary ponds, where the egg bank is the main source of the reestablishment of the population. In such a situation the effect of regulating reproductive switching factors should synchronize the development of a population with the change of environmental conditions in order to ensure the production of the maximum number of diapausing eggs. In this case, combinations of regulating factors that maximize the production of diapausing eggs will be selected and stabilized in the process of evolution. Similar considerations have been developed for some social insects using Pontryagin’s maximum principle (Oster & Rocklin 1979). Recently, it has been demonstrated theoretically that the combined density-dependent control of gametogenesis induction by metabolic by-products and food concentration maximizes the number of produced ephippial eggs in Cladocera (Zadereev et al. 2003). Thus, the experimental data and theoretical research conducted to date has demonstrated that metabolic by-products exuded by conspecifics affect the production of resting eggs to ensure survival of offspring, often resulting in a competitive advantage of individuals in the population that are following such a strategy.

200

E. ZADEREEV

Data on the effect of chemical interactions within a trophic level on diapause induction are contradictory. Recently, it has been demonstrated that Daphnia pulex produced ephippia in water crowded either by conspecifics or by congeners of D. cuculluta (Lurling et al. 2003). Experiments with nine species of Daphnia demonstrated that water from dense Daphnia cultures depressed growth rate and lowered body size and clutch size at first reproduction of small-bodied Daphnia (adult body length