Zooplankton: Sensory Ecology and Physiology [1 ed.] 9056990225, 9789056990220

Citation preview

Zooplankton: sensory ecology and physiology

The material in this book was presented at the Symposium on the Sensory Ecology and Physiology of Zooplankton Held at the Ala Moana Hotel Honolulu, Hawaii January 8-12,1995 The meeting was organized by Petra H. Lenz, University of Hawaii Daniel K. Hartline, University of Hawaii Jennifer E. Purcell, University of Maryland System The meeting was sponsored by US Office of Naval Research International Brain Research Organization Pacific Biomedical Research Center, University of Hawaii at Manoa University of Hawaii Sea Grant College Program University of Maryland System, Center for Environmental Studies Horn Point Environmental Laboratory

(Photograph courtesy of Mr. T. Oyama, Tokai University)

Sergia lucens showing lens bearing dermal photopohores (red spots) on lateral and ventral surfaces as well as thoracic legs (See Figure I, Page 176).

Zooplankton: sensory ecology and physiology edited by Petra H. Lenz Bekesy Laboratory of Neurobiology Pacific Biomedical Research Center University of Hawaii at Manoa, USA .

Daniel K. Hartline Bekesy Laboratory of Neurobiology Pacific Biomedical Research Center University of Hawaii at Manoa, USA

Jennifer E. Purcell Horn Point Environmental Laboratory Center for Environmental and Estuarine Studies University of Maryland System, USA

David L. Macmillan Department of Zoology University of Melbourne, Australia

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 1996 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

Contents

Introduction

ix

General T. H. BULLOCK Neuroethology of zooplankton

WM.HAMNER Predation, cover, and convergent evolution in epipelagic oceans

17

R. K. ZIMMER-FAUST, M. N. TAMBURRI and A. W DECHO Chemosensory ecology of oyster larvae: benthic-pelagic coupling

37

AN. POPPER The teleost octavolateralis system: structure and function

51

H. BLECKMANN, J. MOGDANS and A. FLECK Integration ofhydrodynamic information in the hindbrain of fishes

67

D.MELLONJr Dynamic response properties ofbroad spectrum olfactory interneurons in the crayfish midbrain 85 M.A. R. KOEHL Small-scale fluid dynamics of olfactory antennae·

101

T. BREITHAUPTand J. AYERS Visualization and quantitative analysis ofbiological flow fields using suspended particles

117

Micronekton B. U. BUDELMANN Active marine predators: the sensory world of cephalopods

131

D.-E. NILSSON Eye design, vision and invisibility in planktonic invertebrates

149

M.I.LATZ Physiological mechanisms in the control of bioluminescent countershading in a midwater shrimp 163 M. OMORI, M. I. LATZ, H. FUKAMI and M. WADA New observations on the bioluminescence of the pelagic shrimp, Sergia lucens (Hansen, 1922)

v

175

T. M. FRANK and E. A.WIDDER UV light in the deep sea: in situ measurements of downwelling irradiance in relation to the visual threshold sensitivity ofUV-sensitive crustaceans

185

E. R. LOEW, F. A. MCALARYand W N. MCFARLAND Ultraviolet visual sensitivity in the larvae oftwo species of marine atherinid fishes

195

Meroplankton and Rotifers T.W CRONIN, N. I MARSHALL, R. L. CALDWELL and D. PALES Compound eyes and ocular pigments of crustacean larvae (Stomatopoda and Decapoda, Brachyura)

211

R. B. FORWARD JR., R. A. TANKERSLEY, M. C. DE VRIES and D. RITTSCHOF Sensory physiology and behaviour ofblue crab (Callinectes sapidus) postlarvae during horizontal transport

225

M. S. LAVERACK, D. L. MACMILLAN and S. L. SANDOW Neural development in the planktonic and early benthic stages of the palinurid lobster Jasus edwardsii 241 P. L. STARKWEATHER Sensory potential and feeding in rotifers: structural and behavioral aspects of diet selection in ciliated zooplankton

255

T.W SNELL and R. RICO-MARTINEZ Characteristics of mate-recognition pheromone in Brachionus plicatilis (Rotifera)

267

Copepods D. I LONSDALE, T.W SNELL and M.A. FREY Lectin binding to surface glycoproteins on Coullana spp. (Copepoda: Harpacticoida) can inhibit mate guarding 277 I W AMBLER, S. A. BROADWATER, E. I BUSKEYand I 0. PETERSON Mating behaviour of Dioithona oculata in swarms

287

E. I BUSKEY, I 0. PETERSON and IW AMBLER The role of photoreception in the swarming behavior of the copepod Dioithona oculata

301

N.M.BUTLER Effects of sediment loading on food perception and ingestion by freshwater copepods

315

D. M. FIELDS and I YEN The escape behavior of Pleuromamma xiphias in response to a quantifiable fluid mechanical disturbance

323

D. K. HARTLINE, P. H. LENZ and C. M. HERREN Physiological and behavioral studies of escape responses in calanoid copepods

341

P. H. LENZ, T. M. WEATHERBY, WWEBER and K. K. WONG Sensory specialization along the first antenna of a calanoid copepod, Pleuromamma xiphias (Crustacea)

355

Cladocerans S.I.DODSON Optimal swimming behavior of zooplankton

365

P. LARSSON and 0. T. KLEIVEN Food search and swimming speed in Daphnia

375 vi

J. RINGELBERG and E.VAN GOOL Migrating Daphnia have a memory for fish kairomones

389

E.VAN GOOL and J. RINGELBERG Swimming of Daphnia galeata x hyalina in response to changing light intensities: influence of food availability and predator kairomone

399

L. DE MEESTER and J. PIJANOWSKA On the trait-specificity of the response of Daphnia genotypes to the chemical presence of a predator

407

J. E. GOKCEN and D. C. MCNAUGHT A behavioral bioassay employing Daphnia for detection of sublethal effects: response to polarized light

419

M. C. BREWER and J. N. COUGHLIN Virtual plankton: a novel approach to the investigation of aquatic predator-prey interactions

425

Gelatinous Zooplankton G.O.MACKIE Defensive strategies in planktonic coelenterates

435

J. E. PURCELL and P. A.V. ANDERSON Electrical responses to water-soluble components offish mucus recorded from the cnidocytes of a fish predator, Physalia physalis

447

N. GRIGORIEV and A. N. SPENCER A mechanism for fatigue of epithelial action potentials in the hydromedusa, Polyorchis penicillatus: a case of non-neuronal habituation

461

WM.HAMNER Sensory ecology of scyphomedusae

475

L. J. HANSSON and K. KULTIMA Behavioural response of the scyphozoanjellyfish Aurelia aurita (L.) upon contact with the predatory jellyfish Cyanea cap illata (L.)

493

G. I. MATSUMOTO Observations on the anatomy and behaviour of the cubozoan Carybdea rastonii Haacke

501

S. E. STEWART Field behavior of Tripedalia cystophora (Class Cubozoa)

511

G.O.MACKIE Unconventional signalling in tunicates

525

R. A. SATTERLIE, M. LAGRO, M. TITUS, S. JORDAN and K. ROBERTSON Morphology of wing mechanoreceptors involved in the wing retraction reflex in the pteropod mollusc Clione limacina

535

L. P. MADIN Sensory ecology of salps (Tunicata, Thaliacea): more questions than answers

547

R.L. MILLER Chemosensory phenomena during sexual interactions in gelatinous zooplankton

569

Index

581

vii

Introduction

Planktonic organisms constitute a major portion of the planet's biomass, biodiversity and food base. Traditionally the fields of biological oceanography and limnology have been concerned with planktonic production and the transfer of energy from one trophic level to another, where the focus is on the group rather than the individual. As the quantitative relations among different parts of pelagic food webs are determined with increasing precision, difficulties in accounting for observed interactions make it clear that more detailed information is needed about particular species and the interactions among members of plankton communities. Sensory systems provide animals with information about their environment that is necessary for them to survive and reproduce, and hence maintain their position in the community. Oceanographers and limnologists have long been aware of the special sensory challenges facing animals with a planktonic lifestyle. Complex interactions between them, extensive migrations and orientation in water masses and currents are just some of the behaviours of zooplankton that have intrigued researchers. Because they are generally small and fragile, few physiological studies have been made on zooplankton but those few suggest that sensory systems among planktonic organisms differ in many significant ways from those of benthic forms. A substantial body of information has accumulated over the past two decades on the behaviour of zooplankton in both field and laboratory. The interpretation of these behavioural studies has been hampered, however, by a limited knowledge of the underlying sensory mechanisms. The frustration this has generated in the scientific community is apparent from the repeated calls for additional information on sensory systems in planktonic organisms (e.g. Marine Zooplankton Colloquium I., "Future marine zooplankton research- a perspective", Mar. Ecol. Progr. Ser. 55: 197-206, 1989).The focus of sensory neurobiologists, however, has generally been on particular physiological processes, which are studied in accessible animals that

ix

serve as convenient models, adult benthic molluscs and crustaceans and adult fishes, for example. Recent technological advances have opened the way for the application of neurobiological techniques to small planktonic organisms. Concurrently, new designs in video recording and analysis have expanded behavioural studies in both the laboratory and field. Unfortunately, neuroscientists, marine biologists, biological oceanographers and limnologists do not often attend the same scientific meetings nor do they publish in the same journals. An improved flow of information among these disciplines is likely to elucidate many important aspects of the sensory biology of zooplankton, ranging from morphology and physiology to ecology and evolution. In order to encourage interdisciplinary investigations of these aspects of biological processes in the marine pelagic environment, we convened a diverse group of scientists in Honolulu in January, 1995. Many of the presentations at the meeting were subsequently prepared for publication. All submitted papers were subjected to full peer review*. Rather than the more traditional grouping of sensory studies according to sensory modality, we have used taxonomic groupings which bring into proximity papers concerned with a particular group's role in the community structure and thereby emphasize the fact that several sensory modalities must work in concert to permit organisms to survive in their ecological niches. The broad neurobiological principles which serve as a starting point for an understanding of zooplankton sensory systems at the cellular level are laid out in several papers on non-planktonic organisms.

*All chapters in this volume were accepted following the full peer review procedure of the Journal Marine and Freshwater Behaviour and Physiology, in which versions of some of these and other contributions were published.

X

Comparisons of sensory inputs to pelagic and benthic organisms are highlighted in papers on sensory characteristics and by experimental approaches used in wellstudied vertebrate and invertebrate models. Hopefully, this will encourage the application of similar approaches to zooplankton in the future. In the absence of direct studies on zooplankton, such examples may give us the best direction on how to approach further studies on zooplankton. The contributed papers give details of a diversity of behavioural adaptations that are critically dependent on sensory mechanisms, as well as information on some of the underlying structures and mechanisms. The challenge to all of us interested in zooplankton sensory systems is to combine what is known and what is continuing to be discovered about zooplankton behaviour and ecology with information derived from neurobiological approaches, to better understand the sensory world of small pelagic animals. As emphasized by Ted Bullock in his plenary paper, there are many unexplored leads and opportunities that point to the way neuroethological approaches can contribute to our understanding of zooplankton behaviour. On a trial basis, for a indeterminant time, we will maintain a site on the World Wide Web dedicated to disseminating information on zooplankton sensory systems: http://www.pbrc.hawaii.edu/-lucifer We welcome contributions and comments. PetraLenz Dan Hartline Jenny Purcell David Macmillan

xi

NEUROETHOLOGYOFZOOPLANKTON THEODORE HOLMES BULLOCK Department ofNeurosciences and Neurobiology Unit, Scripps Institution of Oceanography University of California, San Diego La Jolla, CA 92093 email: tbullock@ucsd edu · Application of a neurophysiologist's techniques to the study of zooplankton sensory systems need not be ruled out on the basis of small size or slipperiness. Successes in this endeavor are accumulating steadily. Zooplankton offer a rich source of opportunities for interesting and different sensory systems for neuroethological investigation. Behavioral studies already suggest a number of fruitful areas for sensory physiological investigation. In other cases, sensory structures have been found which invite both physiological and behavioral study. Sensory capabilities found in non-planktonic groups are likely to appear also in planktonic forms, but with their own unique characteristics reflecting the planktonic life style. Following the main discussion is a bibliography of recent papers, and a selection of earlier ones, related to the themes discussed.

INTRODUCTION Whereas the neural analysis of behavior of planktonic species and stages has been relatively neglected, we have many clues that it is going to be rich, diverse and interesting. The aims of this contribution are to defend that statement, with selected examples, and to suggest that neural analysis, particularly sensory physiology, has great explanatory power of ecologically significant behavior. I have to begin with a personal note about plankton, recalling the lasting impression made long ago by a film on invertebrates in the Arctic where scyphomedusan jellyfish were pulsing at a rate well within the range familiar in summer temperate waters, warmer by 20°C. I must have been influenced by this observation and my own experiences in a study of the neural basis of fluctuations in the rate of pulsation of medusae (Bullock 1943), some of which was made in December 1941 in Pensacola, where my wife and I collected Rhopilema cruising at random in the Sound, stopped now and then by Army bridge guards concerned about saboteurs in that first fortnight after Pearl Harbor. At any rate, by the early fifties about half of my laboratory group was devoted to the physiological ecology of temperature acclimation in marine invertebrates. That field, which I left in the early sixties, still offers a challenge in the ecologically fundamental question of why some species are able to acclimate much more than others. The proposal I made in 1955, that different rates in the same organism acclimate to different degrees, resulting in greater disharmony in some species than others, may still be viable and most likely applies to rate processes in sensory and central nervous functions, among others. Medusae are large animals, relatively, although generally treated as planktonic. The first reaction from most workers when neurophysiology of plankton is mentioned concerns their small size or gelatinous nature. The first message I bring is not new but also not widely appreciated.

2

THEODORE HOLMES BULLOCK

SMALL SIZE AND SLIPPERINESS ARE NO EXCUSE Technics have been successfully developed to record nerve impulses from single neurons in zebra fish larvae (Eaton and Kimmel, 1980). (Only one or a few sample citations are given, here and in the following, often representing a substantia/literature). Many papers deal with electrical recording from small flies (Ogmen and Garnier, 1994), even Drosophila (Wyman et al., 1984; Elkins and Ganetsky, 1990; Engel and Wu, 1992; Trimarchi and Schneiderman, 1993) and mosquitoes, and from copepods (Yen et a/., 1992). Single unit action potentials have been recorded from scyphomedusans (Horridge 1953), Clione (Arshavsky et al, 1988, 1991, 1992; Huang and Satterlie, 1990; Satterlie, 1993; Norekian, 1989, 1993; Norekian and Satterlie, 1993b), Melibe (Trimarchi and Watson, 1992), a number of other opisthobranchs, and cephalopods (Maturana and Sperling, 1963; Laverack, 1980; Boyle et al., 1983; Bullock and Budelmann, 1991) and larval fish (Eaton and Nissanov, 1985 mentions a predatory protozoan causing escape · responses in larval zebrafish; Eaton and DiDomenico, 1986). Many studies have been done upon unanesthetized animals, free to move and behave, within limits. BEHAVIOR TURNS UP NEW SENSES AND FORMS OF RECOGNITION TO BE ACCOUNTED FOR The chief source of clues to sensory biology and interesting neuroethology of zooplankters is the close observation of their behavior and responses. I will cite some examples that present opportunities for new analysis of their neural bases. Responses to pure hydrostatic pressure stimuli have been reported in a number of species (Morgan, 1984; Forward, this symposium). Following earlier suggestions from much greater stimulus intensities, Knight-Jones and Qasim (1955) Baylor and Smith (1957) and Enright (1961, 1962, 1963, 1967) found responses to changes in pressure as low as 10 em of water or 10 millibars (mb). Knight-Jones and Qasim provided no details about their experimental methods but reported up-swimming to increases and down-swimming to decreases of 10 mb in Carcinides and Galathea megalops larvae. Baylor and Smith found lower thresholds in pteropods and copepods (Ponte/la, Temora) and, like the previous authors, thresholds of < 1 atmosphere in annelids, hydromedusae, chaetognaths, ctenophores, copepods and others. According to my memory of their verbal presentation, though not in the brief printed version, Baylor and Smith, in one type of experiment, watched individual zooplankters in a vertical glass cylinder, swimming up and down within a range of a few centimeters. They then raised or lowered a leveling bulb connected to the cylinder - whose top was closed - by a flexible U-tube, following each vertical movement of the animal. Keeping the pressure constant at the level of the animal, it increased its vertical excursion markedly, as though lacking the normal change-ofpressure feedback. This behavior implies a non-drifting, absolute pressure sense. Baylor and Smith report that a 15 em stimulus causes a 15 em response in the compensatory direction. They emphasized that reliable responses require animals brought in with extreme care to avoid pressure or pH or other shocks. Depth compensating responses by planktonic animals had been reported in some species to persist with undiminished intensity for several hours (Hardy and Bainbridge, 1951, who used stimuli in the 500 mb range), and have therefore been interpreted as

NEUROETHOLOGYOFZOOPLANKTON

3

indicating a sort of "barostat" by which the animals might maintain constant depth, to within a few meters in the sea. In contrast to the sustained type of response, which requires a tonic receptor, Enright observed transient but intense behavioral responses in Synchelidium sp., an intertidal, predominantly benthic amphipod. He did not exclude their having some sustained sense but showed the importance of the phasic component. He started with many amphipods in a closed jar of sea water, most of them resting on the bottom. Raising the leveling bulb or adding small weights to a piston, both of which were done in the next room, in double blind experiments, caused many animals to swim about vigorously for a few seconds. Decreasing the pressure caused transient reduction of ongoing spontaneous activity. He later found similar responses in the anomuran Emerita, particularly the megalops larval stages, immediately after they have settled following a prolonged planktonic development. He excluded the possibility of small gas bubbles, and we have no precedent for other structures with appreciable compressibility different from aqueous tissues or with piezoelectric properties immersed in aqueous tissues. Digby (1961) proposed a mechanism involving a monolayer of gas but his evidence has not been generally accepted. Even in the best studied species, Emerita, we do not know where the reception occurs. Enright has observed responses in some individuals after removal of all periopods or all four antennae or both eyes (pers. comm,). I believe a hitherto unknown sense organ, indeed a new class of sense organs is awaiting discovery. Only after localizing and identifying the organ can we expect to deduce the detection principle it employs. Mechanoreceptors for water movement and vibration are probably among the most amenable to new physiological study (Newbury, 1972 in chaetognaths; Wiese and Marschall, 1990 in euphausids). Lateral line-like sense organs have been reported in penaeid shrimps (Denton and Gray, 1985) and in cuttlefish (Budelmann this volume; Budelmann and Bleckmann, 1988; Budelmann, 1989; Budelmann et a/., 1991; Bleckmann eta/., 1991a; Bleckmann, 1994). Also awaiting discovery in zooplankton are temperature receptors, especially those with a non-adapting, thermometer-like response that can explain the known behavioral response manifested by a consistent thermopreferendum. It has been repeatedly pointed out that at least teleosts and probably many taxa and stages of development show an ability to stay in a layer of water of a preferred temperature, to within a fraction of a degree, i.e. to close to an isotherm. It seems unlikely that this is explainable in the general case by a parallel isodensity (Forward, 1989b and this symposium) or other clue. If it is, then yet another non-adapting sensory receptor is to be sought. Thermometer organs are well known in mammals, including those that are excited by increases in temperature in the normal living range ("warm receptors") and those that are excited by decreases in temperature in this range ("cold receptors"). They are believed to occur in fish and other exothermic taxa but a convincing demonstration of specific temperature sense organs is an outstanding opportunity (Spath, 1978), particularly in zooplankters. The adapting aspect of sense organ response, i.e. a temperature rate-of-change sensibility is likely to be found first, either in separate receptors or as an initial part of the response of thermometer-like receptors (Forward, 1990b, and this symposium). But, the experience with the ampullae of Lorenzini of elasmobranchs warns us to be cautious; they were first thought to be very sensitive temperature change receptors (Sand, 1938) and only later was this modality shown not to be the normal adequate stimulus (Bullock, 1974).

4

THEODORE HOLMES BULLOCK

Chemoreceptors are indicated by many behavioral observations which, moreover, point to high specificity and sensitivity (Lazzaretto et al., 1990; DeMott and Watson, 1991; Kassimon and Hufnagel, 1992; Snell and Morris, 1993; Bollens et al., 1994, to cite only some of the more recent reports). Some kinds of behavior have yet to be found, but I believe, will be found in zooplankters and will then trigger the search for the organs mediating them. An example is magnetic orientation, known in bacteria (Kalmijn, 1978; Blakemore et al., 1980), where the behavior is apparently adequately accounted for by a known organelle. The behavior has often been claimed for birds and insects but no sense organ or identified transducer has been convincingly shown as yet. Lohmann et al., (1991) report a particular pair of cells in the gastropod, Tritonia, that alter their firing in response to changes in earth-strength magnetic fields; the same field changes do not influence any of 50 other cells. Magnetic orientation as an indirect consequence of highly sensitive electrosensory organs and central pathways and processors able to extract this information and use it in normal behavior, has been shown in elasmobranchs (Kalmijn, 1988). A similar sense exists in some quite small teleosts, marine (/!lotosus, Arius, Siluriformes), as well as freshwater (Siluriformes, Gymnotiformes, Mormyriformes and one subfamily of Osteoglossiformes). At present, however, none but the elasmobranchs are known to have an adequately high sensitivity to make use of the currents induced by motion in the earth's magnetic field. Still, neither the sensitivity in these groups, already known to be electrosensitive, nor the existence of this sense in other taxa, both vertebrate and invertebrate, can be categorically ruled out. I anticipate new findings of both electrosense and direct magnetic sense. Other forms of behavior are quite familiar and yet the sensory modalities involved are only partly or little known. Schooling in teleosts and other groups, including members of the zooplankton, seems likely to depend on more than one sense in different species and conditions (Partridge and Pitcher, 1980; Wiese, this symposium). I doubt that our present understanding, based on a few species, is a representative picture of this widespread class of behaviors. To mention one example, Kalmijn (personal communication) has suggested that the dense schools of the marine catfish, Plotosus, may under some conditions use their electrosense to keep together. Other familiar behaviors whose sensory bases are rarely or little known are predator avoidance, prey detection, conspecific communication and mate recognition. Data on fish kairomones that influence the avoidance or swarming behavior of Daphnia are presented elsewhere in this symposium (Larsson; Ringelberg and van Gool; DeMeester). Finicky settlement of barnacle, polychaete and other larvae upon substrates according to its "taste" or texture is well known (Johhson and Strathmann, 1989; Harvey, 1993; Dineen and Hines, 1994; Forward, Zimmer-Faust, this symposium). Long ago I reviewed the literature on predator recognition by invertebrates and found few examples at that time, apart from scallops and freshwater snails (/!lanorbis); I reported observations of limpets and abalones fleeing from a starfish tubefoot (Bullock, 1953b). Specific chemical signals must be much more common than we then appreciated. The swimming escape response triggered by a brief contact with starfish tubefeet in the sea anemone, Stomphia (Wilson, 1959) and the nudibranch, Tritonia, (Willows and Hoyle, 1969) have been studied physiologically; in the latter case a network of neurons was identified that makes the decision whether to trigger the prolonged behavior.

NEUROETHOLOGYOFZOOPLANKTON

5

Complex visual form recognition is indicated by behavior such as the use of dozens of distinct color patterns for as many social signals, shown by Moynihan and Rodaniche (1982) in a Caribbean squid (Sepioteuthis). Startle responses are a fertile field, convenient for sensory analysis because they are relatively stereotyped and hence successive experiments are likely to represent the same behavior. In the wide variety of taxa where they are found, different adequate stimuli are known, from a moving shadow to a tap or an acoustic click (Eaton, 1984). Mackie (1990) has pointed out interesting parallels between the giant fiber jet swimming in squid and jellyfish, a form ofbehavior that is not sterotyped but quite flexible (Otis and Gilly, 1990). Careful experimental ethology may reveal that in some cases several stimuli of different modalities may function in sequence to bring an animal into close proximity with a desirable target or prime it to be more sensitive to some forthcoming event. A POINT OF VIEW MAKES YOU A NEUROETHOLOGIST Ecologically significant behavior opens another dimension when we begin to uncover the mechanisms in sensory physiology, central analysis of sensory input, recognition of species characteristic sign stimuli, plastic modulation by age, state or other sensory inputs, selection of response from the species repertoire and neural control of effectors. With emphasis on the sensory side, I will illustrate with a selection of examples from near-planktonic or related or paradigmatic species that have received some successful study. These are intended to underline the opportunities and needs in further extension to the great range of planktonic taxa. It is important to note that, whereas some started with known behavior to be accounted for, others began as bottom-up or inside-tooutside or anatomy-to-physiology curiosity. Sometimes the relevant behavior has yet to · be defined. A good example is the. discovery of a sensory system in cephalopods - actually in young, planktonic cuttlefish (Sepia), that appears to be an analogue of the lateral line system in aquatic vertebrates (Budelmann and Bleckmann, 1988; Budelmann, 1989; Budelmann et al, 1991; Bleckmann et al., 199la; Bleckmann, 1994). Anatomical suggestions of possibly sensory structures are widespread (Hayashi and Yamane, 1994; Jensen et al, 1994) and led the physiologists in this case to look for responses in the rows of cutaneous organs on the head. They proved to be responsive to disturbances in the water and not to other stimuli- a new modality for molluscs. Quite a different story is represented by a recent finding in a classical sense organ in a squid (Alloteuthis), the statocyst. Here Williamson (1989) surprises us with the demonstration of electrical coupling between secondary hair (sense) cells - a step toward uncovering the cellular mechanisms of reception. Arkett and Mackie (1988) have shown the sensory hair cells for mechanoreception in the planktonic medusan, Aglantha, to be amenable to physiological study. The same must be true for water movement sensors in many other groups (Budelmann, 1989; Bleckmann, 1994). The sharp distinction found in vertebrates between the lateral line and the inner ear reception of disturbance in the aquatic medium outside the animal has yet to be properly compared with an adequate. sampling of invertebrate systems. A large literature exists on the physiology and anatomy of hearing in fish (Atema et al., 1988; Kalmijn, 1988; Popper, this symposium~ including some small enough to be marginally planktonic and including very young sharks (Bullock and Corwin, 1979). A smaller but substantial literature exists on the lateral line (Coombs eta/., 1989). One

6

THEODORE HOLMES BULLOCK

feature of the octavolateral sense organs ofvertebrates which may have deep significance for their function in planktonic stages of fish is that the number of sensory hair cells increases with size dramatically. Presumably this confers greater ability to detect feeble signals and one has to wonder whether larvae and young fish are relatively deaf. Whereas well controlled behavioral experiments on adequately motivated animals are the final arbiter, simple electrophysiological endpoints may often be the first way to study such questions as the upper frequency limit of hearing or the influence upon hearing of the developing swim bladder and, in some species, its later disappearance. We found in young yellow-tail tuna (Thunnus albacares) that brain responses to inner ear reception cut off sharply above the remarkably low frequency of 350 Hz, even lower than an earlier report based on conditioned responses in two specimens (Iversen, 1967; Bullock, Brill and McClune, unpublished experiments). Electroreception has already alluded to. Many species of agnathans, holocephalans, teleosts and others have been shown to be electroreceptive - more readily by physiological than by behavioral responses (Bullock et a/., 1961; Bullock and Heiligenberg, 1986; Fields et al., 1993). Other taxa with this sense modality seem likely to be found, especially among teleosts, but small size confers a serious disadvantage. Sometimes the technic for recording from single receptors is exceedingly simple and requires no surgery (Viancour, 1979; DeWeile, 1983). Photoreception and vision have attracted much attention among invertebrates (Therman 1940; MacNichol and Love, 1960a, b; Wiersma et al, 1961; Waterman and Wiersma, 1963; Gwilliam, 1963; Hartline and Lange, 1974; Lange and Hartline, 1974; Lange et al., 1974; York and Wiersma, 1975; Schiff 1987, 1989; Cronin et al, 1994) including a few studies on planktonic forms (Smith and Macagno, 1990; Frank and Widder, this symposium). I expect many adaptive specializations to be found among zooplankters - for detecting color, moving shadows, dim light and the like. What is taste and what is olfaction? Why are they so distinct in the peripheral and central structures that mediate them in the vertebrates - already in aquatic taxa, long before terrestrial forms evolved? These questions come primarily from the anatomy and physiology but depend on ethology and ecology for essential clues. Taxa differ greatly in the mechanisms of chemoreception and comparative physiology is essential, in parallel with comparative behavior, to understand the dynamic range, degree of specificity, temporal and spatial resolution of these senses. I like to tell how important taste was in the history of the Scripps Institution of Oceanography and of the unique concentration of neuroscientists in La Jolla. Yngve Zotterman, Professor of Physiology at the Royal Veterinary College in Stockholm, was a prominent comparative physiologist of gustation. He started the series of international congresses on Olfaction and Taste, of which the volume edited by Kurihari, Suzuki and Ogawa (1994) is the eleventh. He visited his fellow Scandinavian, Per Scholander, in La Jolla, in 1959. Scholander was a comparative physiologist of respiration, cardiovascular, water and salt functions and got Zotterman to support his idea of a laboratory vessel dedicated to comparative physiology and biochemistry by showing how he would set up to record nerve impulses from taste fibers in teleosts on board one of the smaller Scripps vessels at sea. Yngve didn't succeed on that trip but supported Pete's idea, which led shortly to the RNAlpha Helix. After that, one thing led to another until Pete and others at S.I.O. recruited the first neuroscientist to La Jolla, in 1965 - Susumu Hagiwara, who was followed by a swelling stream oflike ilk, now many hundreds strong, more than a score of them doing marine biology. In spite ofa substantial literature (represented in our reference list by Finger and

NEUROETHOLOGYOFZOOPLANKTON

7

Silver, 1987; Atema, 1994; Atema and Voigt, 1995), the spectrum of taste receptors and even more of olfactory receptors is still only fragmentarily known, even in the most studied arthropods, molluscs and fish. In order to do justice to the ecologically significant neuroethology of zooplankters, we have to look a bit farther along the central nervous pathways from sense organ to behavior. I will mention only a few examples from taxa that include planktonic members deserving study. Giant fiber systems have been evolved again and again, convergently, among the phyla and classes; even orders and families may differ profoundly in the development of these systems (Bullock, 1948a, 1953a), usually associated with startle responses and the first phase of escape (Eaton, 1984). They are quite amenable to study in small forms, as already pointed out for Drosophila (Wyman et al., 1984), even with extracorporeal, noninvasive electrodes (Featherstone and Drewes, 1991). The meaning of the giant fiber diameter cannot always be its greater velocity, for in small animals like Drosophila the absolute saving in time is small. Nevertheless, in some shrimp, adaptations of the giant fiber for high velocity result in phenomenally fast axons- by far the fastest known (Fan et al., 1961; Huang and Yeh, 1963; Hsu et al., 1964, 1975a, b; Hao and Hsu, 1965; Kusano, 1965, 1966, 1971; Hsu, 1982; Terakawa and Hsu, 1991). This is an elegant case where a simple property revealing a remarkable specialization was overlooked for decades, even though shrimp giant fibers had been studied and shown to be unusual (Holmes et al., 1941). Motor output, its patterning, and central and peripheral organization intimately complement the sensory input and are sure to be of interest in zooplankters. Some illustrative studies include that of Spencer (1988) showing non-spiking interneurons in the swimming system of a pteropod and of Arshavsky et al., (1988) who found nonsynaptic interaction - both discoveries being of general neurobiological import. Arshavsky et al., (1991, 1992) and Satterlie (1993) represent a series of studies of the organization of the swimming system in these gastropods. Wilson (1960) studied the nervous control of movement in annelids, and Bowerman and Larimer (1976) that in crustaceans. A number of chapters in Sandeman and Atwood (1982) and Wiese et al., (1990) have relevant recent examples. Moss and Tamm (1993) show that even the delicate movements of ctenophores can be successfully studied with electrophysiological methods. Important effectors other than moving major parts of the body include the chromatophores. Some studies on them and their control underline the opportunities (Cooper et al., 1990; Hanlon et al., 1990; Novicki et al., 1990), especially when cephalopods in good condition can be as readily available as they are now (Hanlon et al., 1978, 1983). Nervous control of luminescence is also interesting and approachable (Nicol, 1960; Baxter and Pickens, 1964; Latz et al., 1990; Bowlby and Case, 1991; Bannister, 1993). ANATOMY HAS MANY NEW TOOLS AND RICH REWARDS Although our symposium emphasizes sensory ecology and physiology, I must call attention to the opportunities for advancing both of them through anatomical investigation. The armamentarium of available new methods, especially those applicable to revealing neural organization has expanded dramatically in recent years and most of the newer procedures have yet to be exploited on zooplankters. Immunocytochemistry, intracellular dyes and markers transported throughout a

8

THEODORE HOLMES BULLOCK

neuron and its processes, even in fixed material, laser confocal microscopy and the use of optical signals of activity, with or without voltage-sensitive dyes are some of the technics now in use for identifYing and tracing nerve cells and their connections, distinguishing among types of neurons and visualizing active neurons. A limited selection of examples concern the retina (Saidell980; Saidel et al, 1983; Cronin et al., 1994; Arikawa and Matsushita 1994; Becerra et al., 1994; Evans et al., 1993; Munz and McFarland 1977). A selection on other sensory and central structures is represented by the reports of Tamm and Tamm 1991; BoHner et al, Bundy and Paffenhofer 1993. Many sense organs have been long known anatomically, or more recently recognized, but are still without any secuie assignment of function. An example is the dorsal organ of many crustaceans, including planktonic taxa (Laverack and Sinclair, 1994).

THE CODA INTEGRATES THE THEMES AND LEITMOTIFS Our organizers have rightly called attention to the great gap in understanding the principal fauna of the bulk of the biosphere. However complete our list of species, our zoogeography, foodwebs, and life histories, we cannot claim understanding before we know a good deal concerning what each zooplankter does about food, enemies, mates and other conspecifics, diurnal and seasonal states, what it can recognize and discriminate, what behavior follows each adequate stimulus, the pattern of succession of the behavioral repertoire - and the sensory and neural apparatus that accomplishes vital tasks. A major development in zoological neurology in the last quarter century has been the discovery that many taxa of invertebrates have a large proportion of their nerve cells unique and identifiable in every specimen, making it possible, bit by bit, to piece together all or nearly all of the circuitry. This should apply at least as much to the zooplankters as to the bulky lobsters and sea slugs. Such encouragement, combined with the message documented above, that small size and slipperiness need not prevent microelectrode recording with controlled stimulation, makes it clear that the time is ripe and the technics available to make real inroads into this massive agenda. The clues and precedents from work already accomplished by pioneers in the area also make it clear that we can expect surprises and major discoveries, not simply smaller versions of familiar neuroethology. Zooplankton, in its marvellous variety, faces a set of problems in everyday living different from those of benthic, littoral and other faunas and not at all uniform or uneventful as we might imagine from our human perspective on their watery world. I look forward to the next convening of this range of specialists since it seems certain that in the interim this meeting will have sparked an abundance of new efforts and fascinating stories. References This first bibliography ofzooplankton neuroethology includes titles cited in the text and, at the request ofthe Editors, others relevant to the subject, selected mainly from recent literature. For the sake of brevity, many important older references are not included.

Arikawa K. and Matsushita A. (1994) Immunogold co-localization of opsin and actin in Drosophila photoreceptors that undergo active rhabdomere morphogenesis. Zoo/. Sci., 11, 391-398.

NEUROETHOLOGYOFZOOPLANKTON

9

Arkett, S. A. and Mackie, G. 0. (1988) Hair cell mechanoreception in the jellyfish Aglantha digitale.1 Exp. Bioi., 135, 329-342. Arshavskii, Y 1., Orlovskii, G. N. and Panchin, YV. (1991) E1ectrophysiological study of serotoninergic neuron Cl in the pteropod mollusk Clione. Neurophysiol., 23, 14-20. Arshavksy,Y I., Deliagina, T. G., Orlovsky, G. N., Panchin,YY. and Popova, L. (1992) Interneurones mediating the escape reaction of the marine mollusc Clione limacina. 1 Exp. Bioi., 164, 307-314. Arshavsky,Y 1., Deliagina,T. G., Gelfand, I. M., Orlovsky, G. N., Panchin,YY., Pavlova, G. A. and Popova, L. B. (1988) Non-synaptic interaction between neurons in molluscs. Comp. Biochem. Physiol., 91C, 199-203. Atema, J. and Voigt, R. (1995) Sensory biology and behavior. In: The Biology of the Lobster, Homarus americanus. J. Factor, ed. Academic Press, New York. Atema, J. (1994) Turbulent odor dispersal, receptor cell filters, and chemotactic behavior. In: Olfaction and Taste XI. K. Kurihara, N. Suzuki and H. Ogawa, eds. Springer-Verlag, New York. Atema, J. Fay, R. R., Popper, A. N. and Tavolga, W N. (1988) Sensory Biology of Aquatic Animals. SpringerVerlag, New York, pp. 923. Atwood, H. L. and Govind, C. K. (1990) Activity-dependent and age-dependent recruitment and regulation of synapses in identified crustacean neurones. 1 Exp. Bioi., 153, 105-127. Atwood, H. L. (1973) An attempt to account for the diversity of crustacean muscles. Am. Zoot., 13, 357-378. Bannister, N. J. (1993) Innervation of luminous glands in the calanoid copepod Euaugaptilus magnus. 1 Mar. Bioi. Assoc. UK., 73,471-423. Baxter, C. and Pickens, P. E. (1964) Control ofluminescence in hemichordates and some properties of a nerve net system. 1 Exp. Bioi., 41, 1-14. Baylor, E. R. and Smith, F. E. (1957) Diurnal migration of plankton organisms. In: Recent Advances in Invertebrate Physiology. B. T. Scheer, T. H. Bullock, L. H. Kleinholz and A. W Martin, eds. Univ Oregon Publications, Eugene, OR, pp. 21-35. Becerra, M., Manso, M. J., Rodriguezmoldes, M. I. and Anadon, R. (1994) The structure and development of dopaminergic interplexiform cells in the retina of the brown trout, Salmo trutta fario - a tyrosine hydroxylase immunocytochemical study. 1 Anat., 185, 377-385. Blakemore, R. R., Frankel, R. B. and Kalmijn, A. J. (1980) South-seeking magnetotactic bacteria in the southern hemisphere. Nature, 286,384-385. Bleckmann, H. (1994) Reception of hydrodynamic stimuli in aquatic and semiaquatic animals. Fortschr. Zoo/., 41, i- ix, 1-115. Bleckmann, H., Budelmann, B. U. and Bullock, T. H. (199la) Peripheral and central nervous responses evoked by small water movements in a cephalopod. 1 Comp. Physiol. A, 168, 247-257. Bleckmann, H., Breithaupt, T., Blickhan, R. and Tautz, J. (1991b) The time course and frequency content of hydrodynamic events caused by moving fish, frogs, and crustaceans. 1 Comp. Physiol A, 168, 749-757. Bollens, S. M., Frost, B. W and Cordell, J. R. (1994) Chemical, mechanical and visual cues in the vertical migration behavior of the marine planktonic copepod Acartia hudsonica. 1 Plank. Res., 16, 555. Bollner, T., Stormmathisen, J. and Otterson, O.P. (1991) GABA-like immunoreactivity in the nervous system of Oikopleura dioica (Appendicularia). Bioi. Bull., 180, 119-124. Bowerman, R. F. and Larimer, J. L. (1976) Command neurons in crustaceans. Comp. Biochem. Physiol., 54A, 1-6. Bowlby, M. R. and Case, J. F. (1991) Ultrastructure and neuronal control ofluminous cells in the copepod Gaussia princeps. Bioi. Bull., 180, 440-446. Boyle, M. B., Cohen, L. B., Macagno, E. R. and Orbach, H. (1983) The number and size of neurons in the CNS of gastropod molluscs and their suitability for optical recording of activity. Brain Res., 266, 305-317. Budelmann, B. U. (1989) Hydrodynamic receptor systems in invertebrates. In: The Mechanosensory Lateral Line: Neurobiology and Evolution. S. Coombs, P. Gomer, H. Munz, eds. Springer-Verlag, New York, pp. 607-631. Budelmann, B. U. and Bleckmann, H. (1988) A lateral line analogue in cephalopods: water waves generate microphonic potentials in the epidermal head lines of Sepia and Lo/liguncula. 1 Comp. Physiol. A, 164,1-5. Budelmann, B. U., Riese, U. and Bleckmann, H. (1991) Structure, function, and biological significance of the cuttlefish "lateral lines". In: The Cuttlefish. E. Boucaud-Camou, ed. Centre de Publications de l'Universite de Caen, pp. 201-209. Bullock, T. H. (1943) Neuromuscular facilitation in Scyphomedusae.1 Cell. Comp. Physiol., 22, 251-272. Bullock, T. H. (1946) A preparation for the physiological study of the unit synapse. Nature, 158, 555-556. Bullock, T. H. (1948a) Physiological mapping of giant nerve fiber systems in polychaete annelids. Physiol. Comp. Oecol., 1, 1-14. Bullock, T. H. (1948b) Properties of a single synapse in the stellate ganglion of squid. 1 {Veurophys., 11, 343-364.

10

THEODORE HOLMES BULLOCK

Bullock, T. H. (1953a) Properties of some natural and quasi-artificial synapses in polychaetes. J Comp. Neurol., 98,37-68. Bullock, T. H. (1953b) Predator recognition and escape responses of some intertidal gastropods in presence of starfish. Behaviour, 5, 130--140. Bullock, T. H. (1955) Compensation for temperature in the metabolism and activity of poikilotherms. Bioi. Rev., 30,311-341. Bullock, T. H. (1974) An essay on the discovery of sensory receptors and the assignment of their functions together with an introduction to electroreceptors. In: Handbook of Sensory Physiology, JfJ/. III/3: Electroreceptors and Other Specialized Receptors in Lower Tkrtebrates. A Fessard, ed. Springer Verlag, Berlin, pp. 1-12. Bullock, T. H. and Budelmann, B. U (1991) Sensory evoked potentials in unanesthetized, unrestrained cuttlefish: a new preparation for brain physiology in cephalopods. J Comp. Physiol. A, 168, 141-150. Bullock, T. H. and Corwin, J. T. (1979) Acoustic evoked activity in the brain in sharks. l Comp. Physiol. 129, 223-234. Bullock, T. H. and Hagiwara, S. (1957) Intracellular recording from the giant synapse of the squid. l Gen. Physiol., 40,565-577. Bullock, T. H. and Heiligenberg,W F. (1986) Electroreception. New York, John Wiley. Bullock, T. H., Hagiwara, S., Kusano, K. and Negishi, K. (1961) Evidence for a category of electroreceptors in the lateral line of gymnotid fishes. Science, 134, 1426-1427. Bundy, M. H. and Paffenhofer, G. A. (1993) Innervation of copepod antennules investigated using laser scanning confocal microscopy. Mar. Ecol. Prog. Ser., 102, 1-14. Cohen, M. J., Katsuki, Y and Bullock, T. H. (1953) Oscillographic analysis of equilibrium receptors in crustacea. Experientia, 9, 434-435. Coombs, S., Gomer, P. and Miinz, H. (1989) The Mechanosensory Lateral Line: Neurobiology and Evolution. Springer-Verlag, New York. Cooper, K. M., Hanlon, R. T. and Budelmann, B. U (1990) Physiological color change in squid iridophores. II. Ultrastructural mechanisms in Lolliguncula brevis. Cell. Tiss. Res., 259, 15-24. Corning,W C., Dyal, J. A. and Willows, A. 0. D. (1973) Invertebrate Learning. JfJ/1, Protozoans through Annelids. Plenum Press, New York. Costello, J. H., Strickler, J. R. Marrase, C., Trager, G., Zeller, R. and Freise, A. J. (1990) Grazing in a turbulent environment - behavioral response of a calanoid copepod, Centropages hamatus. Proc. Nat!. Acad. Sci. USA, 87,1648-1652. Cronin, T.W, Marshall, N.J., Caldwell, R. L. and Shashar, N. (1994) Specialization of retinal function in the compound eyes of mantis shrimps. Vision Res., 34, 2639-2656. DeMott,W R. and Watson, M. D. (1991) Remote detection of algae by copepods- responses to algal size, odors and motility. l Plank. Res., 13, 1203-1222. Denton, E. J. and Gray, J. A. (1985) Lateral-line-like antennae of certain of the Penaeidae (Crustacea, Decapoda, Natantia). Proc. Roy. Soc. Lond. B, 226, 249-261. DeWeile, J. (1983) Electrosensory information processing by lateral-line lobe neurons of catfish investigated by means of white noise cross-correlation. Comp. Biochem. Physiol., 74A, 677-680. Digby, P. S. B. (1961) Mechanism of sensitivity to hydrostatic pressure in the prawn, Palaemonetes varians Leach. Nature, 191,366-368. Dineen, J. F. and Hines, A. H. (1994) Larval settlement of the polyhaline barnacle Balanus eburneus (Gould)cue interactions and comparisons with two estuarine congeners. J Exp. Mar. Bioi. Ecol., 179, 223-234. Eaton, R. C. (1984) Neural Mechanisms of Startle Behavior. Plenum, New York, pp. 377. Eaton, R. C. and DiDomenico, R. (1986) Role of the teleost escape response during development. Trans. Am. Fish. Soc., 115, 128-142. Eaton, R. C. and Nissanov, J. (1985) A review of Mauthner-initiated escape behavior and its possible role in hatching in the immature zebrafish, Brachydanio rerio. Env. Bioi. Fish. 12, 265-279. Eaton, R. C. and Kimmel, C. B. (1980) Directional sensitivity of the Mauthner cell system to vibrational stimulation in zebrafish larvae.! Comp. Physiol. A, 140,337-342. Elkins, T., Ganetzky, B. (1990) Conduction in the giant nerve fiber pathway in temperature-sensitive paralytic mutants of Drosophila. J. Neurogen. 6, 207-219. Engel, J. E. and Wu, C. F. (1992) Interactions of membrane excitability mutations affecting potassium and sodium currents in the flight and giant fiber escape systems of Drosophila. l Comp. Physiol. A, 17193-104. Enright, J. T. (1962) Responses of an amphipod to pressure changes. Comp. Biochem. Physio/., 7,131-145. Enright, J. T. (1961) Pressure sensitivity of an amphipod. Science, 133, 758-760. Enright, J. T. (1963) Estimates ofthe compressibility of some marine crustaceans. Limnol. Oceanogr., 8, 382-387. Enright, J. T. (1967) Temperature compensation in short-duration time-measurement by an intertidal amphipod. Science, 156, 1510-1512.

NEUROETHOLOGYOFZOOPLANKTON

11

Evans, B. I., Harosi, F. I and Fernald, R. D. (1993) Photoreceptor spectral absorbance in larval and adult winter flounder (fseudopleuronectes americanus). Visual Neurosci., 10, 1065-71. Fan, S-F, Hsu, K. and Chen, F-S. (1961) On the high conduction velocity of the giant nerve fiber of shrimp Penaeus orienta/is. Kexue Tongbao, 4, 51-52. Featherstone, D. and Drewes, C. D. (1991) Non-invasive detection of electrical events during the startle response in larval medaka. J. exp. Bioi., 158, 583-589. Fields, R. D., Bullock, T. H. and Lange, G. D. (1993) Ampullary sense organs, peripheral, central and behavioral electroreception in chimeras (Hydrolagus, Holocephali, Chondrichthyes). Brain Behav. Evol., 41, 269-289. Finger, T. E. and Silver,W L. (1987) Neurobiology of Taste and Smell. John Wiley & Sons, New York. Forward, R. B. (1989a) Depth regulation of larval marine decapod crustaceans - test of an hypothesis. Mar. Bioi., 102,195-201. Forward, R. B. (1989b) Behavioral responses of crustacean larvae to rates of salinity change. Bioi. Bull., 176, 229-238. Forward, R. B. (1990a) Responses of crustacean larvae to hydrostatic pressure - behavioral basis of high barokinesis. Marine. Behav. Physiol., 17, 223-232. Forward, R. B. (1990b) Behavioral responses of crustacean larvae to rates of temperature change. Bioi. Bull., 178, 195-204. Forward, R. B. and Buswell, C. U. (1989) A comparative study of behavioral responses of larval decapod crustaceans to light and pressure. Mar. Behav. Physiol., 16,43-56. Foster, B. A., Cargill, J. M. and Montgomery, J. C. (1987) Planktivory in Pagothenia borchgrevinki (Pisces: Nototheniidae) in McMurdo Sound, Antarctica. Polar Bioi., 8, 49-54. Gibson, R. M. (1983) Visual abilities and foraging behavior of predatory fish. Trends Neurosci., 6, 197-199. Gilmer, R. W (1990) In situ observations of feeding behavior ofthecosome pteropod molluscs. Am. Malacolog. Bull., 8, 53-59. Goto, T. and Yoshida, M. (1988) Histochemical demonstration of a rhodospin-like substance in the eye of the arrowworm Spadella schizoptera Chaetognatha. Exp. Bioi. (Berl), 48, 1-4. Goto, T. and Yoshida, M. (1983) The role of the eye and central nervous system components in phototaxis of the arrowworm Sagitta crassa. Bioi. Bull., 164, 82-92. Goto, T., Takasu, N. and Yoshida, M. (1984) A unique photo receptive structure in the arrowworms Sagitta crassa and Spadella schizoptera Chaetognatha. Cell Tiss. Res., 253, 471-478. Greve,W (1975) Tkrhaltensweisen der Rippenquallen Pleurobrachia pileus (Ctenophora). Inst Wissenschaftlichen Film.Wissenchaftlicher Film C 118111975. Giittingen. Gwilliam, G. F. (1963) The mechanism of the shadow reflex in Cirripedia I. Bioi. Bull., 125, 470-485. Hanlon, R. T., Hixon, R. F. and Hulet,W H. (1978) Laboratory maintenance ofwild-caught loliginid squids. In: Proceedings of the l#Jrkshop on the Squid I /lex i/lecebrosus. N. Balch, T. Amaratunga, R. K. O'Dor, eds. Fisheries and Marine Service, Halifax, Nova Scotia. Hanlon, R. T., Hixon, R. F. and Hulet, W H. (1983) Survival, growth, and behavior of the loliginid squid Loligo plei, Loligo pealei, and Lolliguncula brevis (Mollusca: Cephalopoda) in closed sea water systems. Bioi. Bull., 165, 637-685. Hanlon, R. T., Cooper, K. M., Budelmann, B. U. and Pappas, T. C. (1990) Physiological color change in squid iridophores. I. Behavior, morphology and pharmacology in Lo/liguncula brevis. Cell Tiss. Res., 259, 3-14. Hao, B. and Hsu, K. (1965) The birefringence properties of the myelin sheath of shrimp nerve fiber. Acta Physiol. Sin., 28, 373-377. Hardy, A. C. and Bainbridge, R. (1951) Effect of pressure on the behavior of decapod larvae (Crustacea). Nature, 167, 354-355. Hartline, P. H. and Lange, G. D. (1974) Optic nerve responses to visual stimuli in squid. J. Comp. Physiol., 93, 37-54. Harvey, A. W (1993) Larval settlement and metamorphosis in the sand crab Emerita ta/poida (Crustacea, Decapoda, Anomura). Mar. Bioi., 117, 575-581. Hayashi, I. and Yamane, S. (1994) On a probable sense organ newly found in some eunicid polychaetes. J. Mar. Bioi. Ass. UK, 74, 765. Head, E. I. H. and Harris, L. R. (1994) Feeding selectivity by copepods grazing on natural mixtures of phytoplankton determined by HPLC analysis of pigments. Mar. Ecol. Prog. Ser. 110,75-83. Holmes,W Pumphrey, R. J. and Young, J. Z. (1941) The structure and conduction velocity of the medullated nerve fibres of prawns. J. Exp. Bioi., 18, 50-54. Horridge, G. A. (1953) An action potential from the motor nerves of the jellyfish, Aurellia aurita Lamarck. Nature (LondJ, 171,400. Hough, A. R. and Naylor, E. (1992) Endogenous rhythms of circatidal swimming activity in the estuarine copepod Eurytemora affinis (Poppe). J. Exp. Mar. Bioi. Ecol., 161, 27-32.

12

THEODORE HOLMES BULLOCK

Hsu, K. (1982) Structural and functional characteristics of nerve fibers of shrimp Penaeus orienta/is. IBRO News, 10, 10-11. Hsu, K., Tan, T-P. and Chen, F-S. (1964) On the excitation and saltatory conduction of giant fiber of shrimp (fenaeus orienta/is). In: Theses ofthe 14th National Congress of Chinese Association ofPhysiologists. Hsu, K., Tan, T-P. and Chen, F-S. (1975a) Saltatory conduction in the myelinated giant fibre of the shrimp (fenaeus orienta/is). Kexue Tongbao, 20, 380-382. Hsu, K., Yang, Q-Z. and Tsou, S-H. (1975b) On the apparent lack of resting membrane potential in the shrimp giant nerve fibre. Kexue Tongbao, 20,383-386. Huang, S-K.,Yeh,Y and Hsu, K. (1963) A microscopic and electron microscopic investigation of the myelin sheath of the nerve fiber of Penaeus orienta/is. Acta Physiol. Sin., 26, 39-42. Huang, Z. and Satterlie, R. A. (1990) Neuronal mechanisms underlying behavioral switching in a pteropod mollusc. J Comp. Physiol. A., 166, 875-887. Irisawa, H., Irisawa, A. F., Matsubayashi, T. and Kobayashi, M. (1962) The nervous control of the intracellular action potential of the Squilla heart. J Cell. Comp. Physiol., 59, 55-60. Ishii, H. (1990) In situ feeding rhythms of herbivorous copepods, and the effect of starvation. Mar. Bioi., 105, 91-98. Iversen, R. T. B. (1967) Response of yellowfin tuna (Thunnus albacares) to underwater sound. In: Marine Bio-acoustics, J-012. W. M. Tavolga, ed. Pergamon Press, Oxford. Jensen, P. G., Moyse, J., Hoeg, J. and Alyahya, H. (1994) Comparative SEM studies oflattice organs- putative sensory structures on the carapace oflarvae from Ascothoracida and Cirripedia (Crustacea Maxillopoda Thecostraca). Acta Zoot., 75, 125-142. Johnson, L. E. and Strathmann, R. R. (1989) Settling barnacle larvae avoid substrata previously occupied by a mobile predator. J Exp. Mar. Bioi. Ecol., 128, 87-103. Jonsson, P. R. and Tiselius, P. (1990) Feeding behaviour, prey detection and capture efficiency of the copepod Acartia tonsa feeding on planktonic ciliates. Mar. Ecol Prog. Ser., 60, 35-44. Jordan, C. E. (1992) A model of rapid-start swimming at intermediate Reynolds number - undulatory locomotion in the chaetognath Sagitta elegans. J Exp. Bioi., 163, 119-137. Kabotyanski, E. A. and Sakharov, D. A. (1988) Monoamine-dependent behavioral states in the pteropod mollusc Clione limacina. Symp. Bioi. Hung., 36, 463-477. Kabotyanskii, E. A. and Sakharov, D. A. (1989) Catecholaminergic neurons of the pteropod mollusc Clione limacina. J Evol. Biochem. and Physiol., 25, 198-207. Kalmijn, A. J. (1988a) Hydrodynamic and acoustic field detection. In: Sensory Biology of Aquatic Animals. J. Atema, R. R. Fay, A. N. Popper and W. N. Tavolga, eds. Springer Verlag, New York, pp. 83-130. Kalmijn, A. J. (1988b) Detection of weak electric fields. In: Sensory Biology of Aquatic Animals. J. Atema, R. R. Fay, A. N. Popper and W. N. Tavolga, eds. Springer Verlag, New York, pp. 151-186. Kalmijn, A. J. and Blakemore, R. P. (1978) The magnetic behavior of mud bacteria. In: Animal Migration, Navigation and Homing. L. Schmidt-Koenig and W.T. Keeton, eds. Springer Verlag, New York, pp. 354-355. Kassimon, G. and Hufnagel, L. A. (1992) Suspected chemoreceptors in coelenterates and ctenophores. Micros. Res. Tech., 22, 265-284. Kawabata, K. (1991) Ontogenetic changes in copepod behavior- an ambush cyclopoid predator and a calanoid prey. J Plank. Res., 13,27-34. Kerfoot, W. C. and Kirk, K. L. (1991) Degree oftaste discrimination among suspension-feeding cladocerans and copepods- implications for detritivory and herbivory. Limnol. Oceanog., 36, 1107-1123. King, D. G. (1976) Organization of crustacean neuropil. I. Patterns of synaptic connections in lobster • stomatogastric ganglion. J Neurocytol., 5, 207-237. Knight-Jones, E. W. and Qasim, S. Z. (1955) Responses of some marine plankton animals to changes in hydrostatic pressure. Nature, 175, 941-942. Kotrschal, K., Adam, H., Brandstatter, R., Junger, H., Zaunreiter, M. and Goldschmid, A. (1990) Larval size constraints determine directional ontogenetic shifts in the visual system of teleosts. A mini-review. Z Zool Syst. Evolutionsforsch., 28, 166-182. Krishnan, S. N., Frei, E., Swain, G. P. and Wyman, R. J. (1993) Passover - a gene required for synaptic connectivity in the giant fiber system of Drosophila. Cell, 73, 967-977. Kurihara, K., Suzuki, N. and Ogawa, H. (1994) Olfaction and Taste XL Springer Verlag, New York, pp. 700. Kusano, K. (1965) Electrical characteristics and fine structure of the Kuruma-shrimp nerve fibres (fenaeusjaponicus). Proc. Jpn. Acad., 41,952-957. Kusano, K. (1966) Electrical activity and structural correlates of giant fibers in Kuruma shrimp (fenaeus japonicus). J Cell Physiol., 68, 361-384. Kusano, K. (1971) Impulse conduction in the shrimp medullated giant fiber with special reference to the structure of functionally excitable areas. J Comp. Neurol., 142, 481-494. Lange, G. D. and Hartline, P. H. (1974) Retinal responses in squid and octopus. J Comp. Physiol., 93, 19-36.

NEUROETHOLOGYOFZOOPLANKTON

13

Lange, G. D. and Hartline, P. H. and Hurley, A. C. (1979) Retinal responses in Nautilus. J Comp. Physiol., 134, 281-285. Larson, R. J. (1986) Studies on the fauna of Curacao and other Caribbean Islands No. 213. Observations on the light-inhibited activity cycle and feeding behavior of the hydromedusa Olindias tenuis. Uitg. Natuurwet. Studiekring Suriname Ned. Anti/len, 118, 191-199. Latz, M. 1., Bowlby, M. R. and Case, J. F. (1990) Recovery and stimulation of copepod bioluminescence. J Exp. Mar. Bioi. Ecol., 136,1-22. Laverack, M. S. (1980) Electrophysiology of the isolated central nervous system of the northern octopus Eledone cirrhosa. Mar. Behav. Physiol., 7, 155-169. Laverack, M. S. and Sinclair, A. (1994) Innervation of the dorsal organ of the shrimp Macrobrachium intermedium (Decapoda, Natantia). J Crust. Bioi., 14, 1-5. Lazzaretto, 1., Salvato, B. and Libertini, A. (1990) Evidence of chemical signalling in Tigriopus fulvus (Copepoda, Harpacticoida). Crustaceana, 59,171-179. Lenz, P. H. and Yen, J. (1993) Distal setal mechanoreceptors of the 1st antennae of marine copepods. Bull. Mar. Sci., 3, 170--179. Lohmann, K. 1., Willows, A. 0. D. and Pinter, R. B. (1991) An identifiable molluscan neuron responds to changes in earth-strength magnetic fields. J Exp. Bioi., 161, 1-24. Mackie, G. 0. (1984) Fast pathways and escape behavior in Cnidaria. In: Neural Mechanisms ofStartle Behavior, R. C. Eaton, ed. Plenum, New York, pp. 15-42. Mackie, G. 0. (1990) Giant axons and control of jetting in the squid Loligo and the jellyfish Aglantha. Can. l Zoo/., 68, 799-805. MacNichol, E. F. Jr. and Love,W E. (1960a) Electrical responses of the retinal nerve and optic ganglion of the squid. Science, 132,737-738. MacNichol, E. F. Jr. and Love,W E. (1960b) Impulse discharges from retinal nerve and optic ganglion of the squid. In: Symposium on the Visual System. R. Jung, J. Kornhuber, eds. Springer, Heidelberg, pp. 97-103. Marsden, J. R. (1990) Light responses of the planktotrophic larva of the serpulid polychaet Spirobranchus polycerus. Mar. Ecol. Prog. Ser., 58,225-234. Matsumoto, G. I. and Harbison, G. R. (1993) In situ observations of foraging, feeding, and escape behavior in three orders of oceanic ctenophores- Lobata, Cestida, and Beroida. Mar. Bioi., 117, 279-287. Maturana, H. R. and Sperling, S. (1963) Unidirectional response to angular acceleration recorded from the middle crista! nerve in the statocyst of Octopus vulgaris. Nature (Lond), 197, 815-816. McClintock, J. B. and Janssen, J. (1990) Pteropod abduction as a chemical defence in a pelagic antarctic amphi pod. Nature, 346, 462-464. Montgomery, J. C., Macdonald, J. A and Housley, G. D. (1988) Lateral line function in an antarctic fish related to the signals produced by planktonic prey. J Camp. Physiol. A, 163, 827-833. Morgan, E. (1984) The pressure-responses of marine invertebrates: a psychophysical perspective. Zoo/. J Linn. Soc., 80,209-230. Moss, A. G. and Tamm, S. L. (1993) Patterns of electrical activity in comb plates of feeding Pleurobrachia (Ctenophora). Phil. Trans. Roy. Soc. Lond. B, 339,1-16. Moynihan, M. and Rodaniche, A. F (1982) The Behavior and Natural History of the Caribbean Reef Squid Sepioteuthis sepioidea.Verlag Paul Parey, Berlin. Mpitsos, G. J. and Lukowiak, K. (1985) Learning in gastropod molluscs. In: The Mollusca, JiJ/8, Neurobiology and Behavior, Part 1. AOD Willows, ed. Academic Press, New York, pp. 95-267. Munz, S. W and McFarland, W N. (1977) Evolutionary adaptations of fishes to the photic environment. In: Handbook of Sensory Physiology, H. J. Autrum, ed., vol.VII/5 Visual System in Vertebrates. F. Crescitelli, ed. Springer-Verlag, New York pp. 193-274. Newbury, T. K. (1972) Vibration perception by chaetognaths. Nature, 236, 459-460. Nicol, J. A. C. (1960) The regulation oflight emission in animals. Bioi. Rev., 35, 1-42. Nilsson, D-E. (1983) Evolutionary links between apposition and superposition optics in crustacean eyes. Nature (Lond), 302, 818-821. Norekian, T. P. (1989) Neurons determining passive defensive response in the pteropod mollusk Clione limacina. Neurophysiol., 21, 495-502. Norekian, T. P. (1993) Cerebral neurons underlying prey capture movements in the pteropod mollusc, Clione limacina. 2. Afterdischarges. l Comp. Physiol. A, 172, 171-181. Norekian, T. P. and Satterlie, R. A. (l993a) Co-activation of antagonistic motoneurons as a mechanism ofhighspeed hydraulic inflation of prey capture appendages in the pteropod mollusk Clione limacina. Biol Bull., 185, 240--247. Norekian, T. P. and Satterlie, R. A. (1993b) Cerebral neurons underlying prey capture movements in the pteropod mollusc, Clione limacina.l. Physiology, morphology. J Camp. Physiol. A, 172, 153-169.

THEODORE HOLMES BULLOCK

14

Norekian, T. P. and Satterlie, R. A. (1994) Small cardioactive peptide B increases the responsiveness of the neural system underlying prey capture reactions in the pteropod mollusc, Clione limacina. J Exp. Zoo!., 270, 136-147. Novicki, A., Budelmann, B. U and Hanlon, R. T. (1990) Brain pathways of the chromatophore system in the squid. Brain Res., 519, 315~ 323. Ogmen, H. and Garnier, L. (1994) Quantitative studies of fly visual sustained neurons. lnternat. J Bio-Med. Computing, 36, 299~310. Ogura,T. and Obara, S. (1992) The membrane properties and Ca-currents of the trigeminal root ganglion cells in primary culture of the marine catfish, Plotosus, studied with whole-cell recordings. Brain Res., 597, 84--91. Olsson, R., Holmberg, K. and Lilliemarck, Y (1990) Fine structure of the brain and brain nerves of Oikopleura dioica (Urochordata, Appendicularia). Zoomorphol., 110, I~7. Otis, T. S. and Gilly,W F. (1990) Jet-propelled escape in the squid Loligo opa/escens: concerted control by giant and non-giant motor axon pathways. Proc. Nat/. Acad. Sci. USA, 87, 2911 ~2915 Paffenhiifer, G. A. and Lewis, K. D. (1990) Perceptive performance and feeding behavior of calanoid copepods. J Plank. Res., 12, 933~946. Page, L. R. (1993) Development of behavior in juveniles of Melibe leonina (Gastropoda, Nudibranchia). Mar. Behav. Physio/., 22, 141~161. Partridge, B. L. and Pitcher, T. J. (1980) The sensory basis of fish schools: relative roles oflateralline and vision. J Comp. Physiol. A, 135, 315~325. Paul, D. H. (198la) Homologies between body movements and muscular contractions in the locomotion of two decapods of different families. J Exp. Bioi., 94, 159~ 168. Paul, D. H. (1981 b) Homologies between neuromuscular systems serving different functions in two decapods of different families. J Exp. Bioi., 94, 169~ 187. Paul, D. H. (1989) A neurophylogenist's view of decapod crustacea. Bull. Mar. Sci., 45, 487~504. Paulus, U, Kortje, K. H. and Rahmann, H. (1993) Effects of development and altered gravity conditions on cytochrome oxidase activity in a vestibular nucleus of the larval teleost brain ~ a quantitative electronmicroscopical study. J Neurobiol., 24, 1131 ~ 1141. Poulin, R., Curtis, M. A. and Rau, M. E. (1990) Responses of the fish ectoparasite Sa/mineola edwardsii (Copepoda) to stimulation, and their implication for host-finding. Parasitol., 100, 417-421. Prokopy, R., Bergweiler, C., Galarza, J. and Schwerin, J. (1994) Prior experience affects the visual ability of Rhagoletis pomonella flies (Dipotera, Tephritidae) to find host fruit. J Insect Behav., 7 663~677. Przysiezniak, J. and Spencer, A. N. (1989) Primary culture of identified neurons from a cnidarian. J Exp. Bioi., 142, 97~113. Ryland, J. S. (1990) A circadian rhythm in the tropical ascidian Diplosoma virens (Ascidiacea, Didemnidae) J Exp. Mar. Bioi. Ecol., 138, 217~225. Saidel, W M. (1980) Orthogonal polarization sensitivities of squid photoreceptors: implication for a retinal design. Bioi. Bull., 158, 490. Saidel, W M., Lettvin, J. Y and MacNichol, E. F. Jr. (1983) Processing of polarized light by squid photoreceptors. Nature, 304,534--536. Sand, A. (1938) The function of the ampullae of Lorenzini, with some observations on the effect of temperature of sensory rhythms. Proc. Roy. Soc. Lond. B, 125, 524--533. Sandeman, D. C. and Atwood, H. L. (1982) Neural integration and behavior. The Biology of Crustacea, Vol. 4. D. E. Bliss, ed. Academic Press, New York, pp. 327. (Several relevant chapters} Satterlie, R. A. (1993) Neuromuscular organization in the swimming system of the pteropod mollusc Clione limacina. J Exp. Bioi., 181, 119~ 140. Saunders, A. J. and Montgomery, J. C. (1985) Field and laboratory studies of the feeding behavior of piper Hyporhamphus ihi with reference to the role of the lateral-line in feeding. Proc. Roy. Soc. Lond. B, 224, 209~222.

Schiff, H. (1987) Optical and neural pooling in visual processing in Crustacea. Comp. Biochem. Physiol., 88A, 1~13.

Schiff, H. (1989) Visual input patterns correlated to behavior and habitat of the mantis shrimp Gonodactylus. Comp. Biochem, Physiol., 94A, 75~87. Sejnowski, T. J., Reingold, S.C., Kelley, D. B. and Gelperin, A. (1980) Localization of[ 3 H)-2-deoxyglucose in single molluscan neurones. Nature, 287, 449-451. Sensenbaugh, T. and Franzen, A. (1987) Fine structural observation of the apical organ in the larva of Polygordius Annelida Polychaeta. Scanning Microsc., 1, 181 ~ 190. Sherman, R. G. and Atwood, H. L. (1972) Correlated electrophysiological and ultrastructural studies of a crustacean motor unit. J Gen. Physiol., 59, 586-615. Smith, K. C. and Macagno, E. R. (1990) UV photoreceptors in the compound eye of Daphnia magna (Crustacea, Branchiopoda). A fourth spectral class in single ommatidia. J Comp. Physiol. A, 166, 597~606.

NEUROETHOLOGYOFZOOPLANKTON

15

Snell, T. W and Morris, P. D. (1993) Sexual communication in copepods and rotifers. Hydrobiologia, 255, 109-116. Spath, M. (1978) First central processing of temperature in teleost fish. U?rh. Deutsch.Zool. Ges., 1978, 210. Spencer, A. N. (1988) Non-spiking interneurons in the pedal ganglia of a swimming mollusc. J. Exp. Bioi., 134,442-450. Sumida, B. H. and Case, J. F. (1983) Food recognition by Chaetopterus variopedatus. Synergy of mechanical and chemical stimulation. Mar. Behav. Physiol., 9, 249-274. Sweatt, A. J. (1985) Rod outer segment-like structure of an invertebrate photoreceptor, Phylum Chaetognatha. J. Cell Bioi., 101, 222A. Sweatt, A. J. and Forward, R. B. Jr. (1985) Spectral sensitivity of the chaetognath Sagitta hispida. Bioi. Bull., 168,32-38. T?mm, S. and Tamm, S. L. (1991) Actin pegs and ultrastructure of presumed sensory receptors of Beroe (Ctenophora). Cell. Tiss. Res. 264,151-159. Terakawa, S. and Hsu, K. (1991) Ionic currents of the nodal membrane underlying the fastest saltatory conduction in myelinated giant nerve fibers of the shrimp Penaeus japonicus. J. Neurobiol., 22, 342-352. Theisen, B., Zeiske, E., Silver,W L. Marui, T. and Caprio, J. (1991) Morphological and physiological studies on the olfactory organ of the striped eel catfish, Plotosus lineatus. Mar. Bioi., 110, 127-135. Therman, P. 0. (1940) The action potentials of the squid eye. Am. J. Physiol., 130, 239-248. Trapido-Rosenthal, H. G. and Morse, D. E. (1986) Availability of chemosensory receptors is down-regulated by habituation oflarvae to a morphogenetic signal. Proc. Nat!. Acad Sci. USA, 83, 7658-7662. Trimarchi, J. R. and Schneiderman, A. M. (1993) Giant fiber activation of an intrinsic muscle in the mesothoracic leg of Drosophila melanogaster. J. Exp. Biol, 177, 149-167. Trimarchi, J. and Watson, W H. (1992) The role of the Melibe buccal ganglia in feeding behavior. Mar. Behav. Physiol, 19,195-209. Viancour, T. A. (1979) Peripheral electrosense physiology: a review of recent findings. J. Physiol. (Paris), 75, 321-333. Waterman, T. H. and Wiersma, C. A. G. (1963) Electrical responses in decapod crustacean visual systems. J. Cell. Comp. Physiol., 61, 1-16. Watson,W H. and Chester, C. M. (1993) The influence of olfactory and tactile stimuli on the feeding behavior of Melibe leonina (Gould, 1852) (Opisthobranchia, Dendronotacea). U?liger, 36, 311-316. Watson, W H. and Trimarchi, J. (1992) A quantitative description of Melibe feeding behavior and its modification by prey density. Mar. Behav. Physiol., 19, 183-194. Welsford, I. G., Meyers, R. A.,Wilson, D. S. and Satterlie, R. A. (1991) Neuromuscular organization for wing control in a mollusc (Clione limacina) and a bird (Columba Iivia)- parallels in design. Am. Zoo!., 31,670-679. Wiese, K. and Marschall, H-P. (1990) Sensitivity to vibration and turbulence of water in context with schooling in Antarctic krill, Euphausia superba. In: Frontiers in Crustacean Neurobiology. K. Wiese, W-D, Krenz, J. Tautz, H. Reichert and B. Mulloney, eds. Birkhiiuser, Boston, pp. 121-130. Wiese, K., Krenz, W-D., Tautz, J., Reichert, H. and Mulloney, B. (1990) Frontiers in Crustacean Neurobiology. Birkhiiuser, Boston. (Several relevant chapters.). Wiersma, C. A. G. and Hirsh, R. (1974) Memory evoked optomotor responses in crustaceans. J. Neurobiol., 5, 213-230. Wiersma, C. A. G. and Van Harreveld, A. (1938) A comparative study of the double motor innervation in marine crustaceans. J. Exp. Bioi., 15,18-31. Wiersma, C. A. G.,Waterman,T. H. and Bush, B. M. H. (1961) The impulse traffic in the optic nerve of decapod crustacea. Science, 134, 1435. Williamson, R. (1989) Electrical coupling between secondary hair cells in the statocyst of the squid Alloteuthis subulata. Brain Res., 486, 67-72. Willows, A. 0. D. and Hoyle, G. (1969) Neuronal network triggering a fixed action pattern. Science, 166, 1549-1551. Wilson, D. M. (1959) Long-term facilitation in a swimming sea anemone. J. Exp. Bioi., 36,526--532. Wilson, D. M. (1960) Nervous control of movement in annelids. J. Exp. Bioi., 37, 46--56. Wilson, L. J. and Paul, D. H. (1987) Tailflipping of Munida quadrispina (Galatheidae): conservation ofbehavior and underlying musculature with loss of anterior contralateral flexor motoneurons and motor giant. J. Comp. Physiol. A, 161,881-890. Wyman, R. 1., Thomas, J. B., Salkoff. Land King, D. G. (1984) The Drosophila giant fiber system. In: Neural Mechanisms ofStartle Behavior. R. C. Eaton, ed. Plenum, New York, pp. 133-162. Yen, J., Lenz, P. H., Gassie, D. V. and Hartline, D. K. (1992) Mechanoreception in marine copepods electrophysiological studies on the first antennae. J. Plank. Res., 14,495-512. York, B. and Wiersma, C. A. G. (1975) Visual processing in the rock lobster (Crustacea). Prog. Neurobiol., 5,127-166.

16

THEODORE HOLMES BULLOCK

Zakharov, I. S. and Ierusa1imsky, V. N. (1992) The neuroanatomica1 basis of feeding behavior in the pteropod mollusc, Clione limacina (Phipps). J Comp. Physiol. A, 170 525-532.

PREDATION, COVER, AND CONVERGENT EVOLUTION IN EPIPELAGIC OCEANS WILLIAM M. HAMNER

Department ofBiology, University of California, Box 951606, Los Angeles, CA 90095-1606 USA The factor of most importance to the structure of epipelagic oceanic communities is the absence of cover and the inability to hide from predators in surface waters during the day (Elton, 1939).Visual predation in an environment devoid of cover has resulted in convergent evolution into only six modal adaptive patterns. Large, fast, visual predators roam the water, I) alone or in 2) schools, and they eat anything of appropriate size that they see. Prey escape only by dint of 3) very small size, 4) invisibility due to tissue transparency, 5) diurnal vertical migration, or by 6) exploitation of the sea surface. The sensory ecology and physiology of zooplankton are different from that of all other animal categories in all other habitats. Epipelagic zooplankton are either extremely small animals, with small and structurally simple sense organs, or they are large, with gelatinous, transparent bodies which often lack sense organs.

INTRODUCTION The sensory ecology and physiology of zooplankton are different from that of all other animal categories in all other habitats, yet the reasons for these differences are not immediately obvious. Light, sound, and chemical stimuli are transmitted readily through water, although with different quantitative properties than through the medium of air, and the phyletic distribution of animals that comprise the plankton is similar to, indeed richer than, the categories of animals that live on land. One might logically suppose that those planktonic organisms with evolutionary histories similar to their terrestrial relatives would process environmental stimuli via similar sorts of sense organs, yet this is not the case. In the epipelagic zone of the sea, zooplankton are either extremely small, with commensurately small and structurally simple sense organs, or they are large, with gelatinous, transparent bodies which often lack obvious sense organs. The reasons for the simplicity or absence of sense organs among zooplankton appear to be related to the importance of convergent evolution and visual predation in an epipelagic environment devoid of cover. Plankton is a generic term of Greek derivation that means drifting or wandering (Hardy, 1965). It is a term that implies behavioral passivity and an inability to move freely through the water, even though it is now clear that even the least mobile plant and animal cells are capable of routine movements within the water column (Omori and Hamner 1982; Hamner, 1988). The term plankton is somewhat ambiguous and difficult to define qualitatively (Aleyev, 1977), but this term is of immense importance because it refers to a complex assemblage of organisms, not present on land, that directly or indirectly mediates all aspects of ecology in the sea. Zooplankton can be defined qualitatively and also quantitatively in biomechanical terms (Aleyev, 1977). I begin with qualitative biological considerations because a strictly quantitative engineering approach can generate conclusions that are biologically unacceptable (Hamner, 1979), and these incorrect conclusions can be corrected best by discussing broader issues first. 17

18

WM.HAMNER

------Figure 1 Flotsam accumulating at an ocean front provides cover at the air-sea interface. Redrawn from Hamner (1988).

The sensory ecology and physiology of zooplankton must be considered within the context of the planktonic environment, but many of the animals to be discussed during this symposium are not planktonic at all, and not all aspects of sensory physiology are rele-

EVOLUTION IN THE EPIPELAGIC OCEAN

19

vant to survival in the pelagic realm. For example, no one would contend that adult lobsters, goldfish, or squid are zooplankton, but these large animals are easier to investigate physiologically than are their planktonic larvae. There is no compelling reason why large sensory organs on large animals should not function similarly to identical but smaller organs on smaller species that are perhaps planktonic. It is nonetheless important to consider those organisms which really are planktonic in order to generalize about their sensory biology. Although some of the material presented here will be familiar to a few, most physiologists have a laboratory orientation, many have never been to sea, and almost none have personally viewed the empty, blue vastness of the open ocean while swimming or diving at sea. ON THE NATURE OF THE EPIPELAGIC OCEAN The blue water of the open sea presents a set of strange sensory stimuli totally foreign to us as terrestrial animals. For an observer swimming or diving in the open sea in daytime, there is usually nothing to see but clear blue water and gelatinous zooplankton (Hamner, 1974, 1975). The diver is neutrally buoyant, gravity does not pull on the body, and normal proprioceptive sensations are absent. Sounds other than one's own breathing can not be heard. One can not smell or taste, and wrapped within a wet suit, one's sense of touch is impaired. Were it not, there still would be nothing to touch. Yet, in spite of acute sensory deprivation, all divers at sea, when surrounded by sharks, share the powerful sensation of absolute vulnerability. In blue oceanic water there is no place to hide. Many authors have discussed differences between oceanic and terrestrial ecosystems (Hamner eta/., 1975; Parsons, 1976; Aleyev, 1977; Isaacs, 1977; Dayton, 1984; Steele, 1985; Smetacek and Pollehne, 1986; Hamner, 1988; Denny, 1990; McFall-Ngai, 1990; Nybakken, 1993), and although the most obvious difference is the presence or absence of water itself (Smetacek and Pollehne, 1986) and its associated physical properties (Vogel, 1981; Dejours, 1987; Denny, 1990, 1993; McFall-Ngai, 1990), the factor of most importance to the structure of epipelagic communities is the nature of cover (Figure 1), more properly, the absence of cover and the absence of places to hide from predators at sea (Elton, 1939, 1966; Williams, 1964; Hamilton, 1971; Hamner, 1974, 1977, 1985; Hamner, eta/., 1975; Zaret, 1975; Isaacs, 1977; McFall-Ngai, 1990). Charles Elton (1939) appears to have been the first ecologist to address explicitly the importance of environmental cover as sites of physical refuge from predators in terrestrial ecosystems, ideas developed while watching two cats practicing the game of predator and prey, stalking each other in " ... search, pursuit, escape, and refuge, and the influence of cover on them:' Elton considered " ... that cover is only one of the components of a wider system ... the protective system of the prey-protection here being primarily from predatory enemies ... The influence of the physical qualities and behavior of the prey, such as size, speed, color, and quickness of sense ... and ... nocturnal habits ... profoundly affect the kind of predator-prey relations that can occur:' Later, Elton (1966) added insight into his considerations of the system of cover to include a" ... class of animals [that] come out in the dark ... [and a] ... third class of animals ... that are comparatively free from enemies, through size and strength, either of the individual or the herd, or else have managed to be quite inedible or to live in some place where there just are no effective enemies at all:' Elton (1966) stated unequivocally, "In all except a few communities cover is paramount in the life of animals, ... [and] ... when we look at the patterned landscape we are looking at a very complicated structure of cover, within and

20

W.M.HAMNER

to a lesser extent outside and around which are living a very large number of animals ... highly dependent on the kind of structure that the cover provides?' Elton thus noted that I) absolute size, 2) speed and quickness, 3) color, and 4) nocturnality are central considerations for the predator-prey system in terrestrial communities with abundant cover. Elton's thoughts on the nature of cover were never explicitly applied to the open water of lakes or ponds, even though the water bodies in Wytham Woods were discussed in some detail (Elton, 1966). Later, Hamner et al. (1975) and Zaret (1975) independently considered strategies for survival of epipelagic animals in the homogeneous environments of the sea and fresh water, although the four adaptive strategies listed in these two articles were not co-equal. Near the surface of the open sea during the day, the ocean is clear and blue and in the absence of cover there has been convergent evolution due to the overriding importance of visual predation, resulting in a small number of modal adaptive strategies for survival at sea (Hamner 1974, 1977, 1985; Hamner et al., 1975). These modal survival patterns for the open sea are almost identical to those classes of prey noted by Elton for terrestrial habitats, even though the nature of cover in these contrasting environments is diametrically different. Thus at sea, large, fast, visual predators roam the water, 1) alone or in 2) schools, and anything edible and of appropriate size that they see, they attack and devour. Potential prey escape from these predators only by dint of 3) very small size, 4) invisibility due to tissue transparency, 5) diurnal vertical migration, or by 6) exploitation of the sea surface. Among these six categories are those animals that are too small to be seen and those which are larger but gelatinous and transparent, and these are the two groups of organisms in the sea that we traditionally categorize as epipelagic zooplankton. Each of these categories of oceanic animal has convergent representatives from many different taxonomic groups, and convergent evolution ofform and function dominates the epipelagic ocean. There is absolutely no question about the reality and the importance of these modal convergent categories of animals in the epipelagic zone of the world's oceans. However, there are epipelagic waters where this scheme needs modification and there are animals that seem anomalous, and both need to be identified so that these inductive generalizations are not prematurely dismissed. Four caveats need consideration; specific exceptions to the rules will be noted separately. The most obvious deviation from these categories of convergent pelagic organisms occurs in pelagic freshwater habitats. With the exception of two genera of extremely rare hydromedusae, there are no freshwater gelatinous zooplankton. In fresh water there are no ctenophores, appendicularians, salps, doliolids, heteropods or pteropods, and almost no medusae. It is not clear why this is so, and the absence ofgelatinous zooplankton in fresh water is most peculiar. Because there are occasional hydromedusae in fresh water, the answer may not be a purely physiological constraint involving osmotic incompatibility. Meroplankton, the planktonic larval stages of benthic invertebrates, are also poorly represented in fresh water and insects and rotifers exploit fresh and brackish waters but are not important members of oceanic communities; the reasons for these anomalies are not clear either. Most freshwater habitats have a quite limited number of pelagic fish species (by accident, recent evolution, or extinction) and freshwater epipelagic predators generally are much smaller than those that dominate the sea. For example, there are many lakes and ponds that contain no fish at all, and in these environments various invertebrates or other vertebrates (e.g. salamanders) become fearsome top predators. This never happens in the open sea. Only in odd saltwater environments such as the marine lakes in Palau (Hamner and Hauri, 1981; Hamner et al., 1983) can one find marine communities

EVOLUTION IN THE EPIPELAGIC OCEAN

21

that contain no large, fast swimming predators. In some bodies offresh water there are obvious large nektonic animals (e.g. sharks in Lake Nicaragua, 5 species of freshwater dolphins, seals in Lake Baikal, arawanas, Nile perch, etc), and in these habitats these large predators have a profound impact on the epipelagic community, yet as a generalization it is fair to state that the nature of nekton in fresh water is quite different from that in the sea. The nature of micronekton also is different in freshwater and oceanic environments. At sea diurnal vertical migration is a ubiquitous behavior of many members of the midwater micronektonic community. However, there are almost no freshwater micronekton. The ecological equivalents of lantern fish do no occur in fresh water nor are there any ecological equivalents ofholopelagic, migratory shrimp, such as euphausiids. This is certainly partly because of different taxonomic distributions between different biomes, but it is more clearly related to the fact that in most freshwater habitats there simply is no midwater realm. Most freshwater habitats are very shallow, and the freshwater animals that engage in diurnal vertical migration are either very small or they are benthic during the day, emerging from hiding in the vegetation or mud on the bottom to exploit the resources in the water column at night. A last admonition is that in polar oceans in summer, diurnal vertical migration is not necessarily advantageous. An example drawn from our observations of Antarctic krill illustrates this issue (Hamner et al., 1983; Hamner, 1984). In polar oceans during summer, diurnal vertical migration to avoid surface predators during the long daylight hours may be disadvantageous for an herbivore such as Euphausia superba because all phytoplankton production occurs quite near the surface and pelagic herbivores either feed there or they starve. Rather than swimming down into deep water to evade predators, Euphausia superba engages in schooling behavior in the epipelagic zone, evolutionary electing a modal survival strategy that is quite different from the diurnal vertical migratory solution so effective for tropical euphausiids. Thus, when the temporal periodicity of the 24-hour light environment changes, so must the nature of the convergent adaptive solutions change to permit survival in an illuminated three dimensional environment devoid of cover.

NEKTON Convergent evolution in the epipelagic environment is expressed most dramatically by large, nektonic predators (Figure 2), best described both qualitatively and quantitatively by Aleyev (1977). Aleyev stated that '~daptation to active propulsion in water ... had the most fundamental influence on ... convergence of animals from different, and at times very remote, systematic groups ... Life in the pelagic zone has many special features. Here, as nowhere else, one feels the 'three-dimensionality' of the region. There is no firm support for organisms, no shelter and hardly anything visible by which to take one's bearing.... Among the factors ... determining ... nekton as a specific ecomorphological type, two are decisive: 1. the fact that animals are always ... without ... solid substrate, and 2. the absence in the pelagic zone of any shelters.... For animals which are sufficiently large, as are most nektonic forms, the absence of any shelters in the pelagic zone makes it necessary to escape from predators by fleeing ... The three-dimensionality of the environment as such is conducive to greater mobility of pelagic animals.... A permanently 'exposed' existence compels improvement of camouflage and ways of escap-

22

W.M.HAMNER

Figure 2 Convergence of body form in a) shark, the porbeagle (Lamia nasus), and b) bluefin tuna (Thunnus thynus). Redrawn from Parin (1970).

ing from predators by fleeing .... The evolution of nekton exemplifies how deep convergence may be and how important convergent processes are in the evolution of the animal kingdom:' In order to swim fast, one must be long, narrow and streamlined, without protuberances that generate drag. Aleyev (1977) developed these ideas quantitatively. He noted, "Both planktonic and nektonic animals exhibit a certain variety of forms oflocomotion and of general body structure; however, in plankton this variety is incomparably greater than in nekton, where uniformity of functional-morphological organization, the degree of convergence, attains its maximum .... As the absolute speed oflocomotion increases, hydrodynamic resistance grows very rapidly, under an exponential law, and minimizing this drag becomes the organism's most crucial task. In this connection, with rising Re (particularly for Re > 10 6) one observes a steeply increasing convergence of the functional-morphological spectrum, which results in ever growing uniformity of the general morphological organization of nektonic animals belonging to various systematic groups:'

23

EVOLUTION IN THE EPIPELAGIC OCEAN

u 1.0

II

!II]] EP

II

§EN

• Movement energetically wasteful

0.8 0.6

•

0.2

1::.

XN

Increasing Cop

• 10 5

• EP • EN

....•• ____.; ; f

A

~

10 8

f

Increasing Cot

10 9

1010 Re

Figure 3 The relationship of U, the slenderness ratio, where 1.0 represents a round object and 0.2 an object that is long (5X) and thin (IX), toRe, Reynolds number. A, average values for eunekton; a, range of optimal hull elongation for submarines; EP, euplankton; EN, eunekton; XN, xeronekton. Cvp, form drag; Cv1, frictional drag. Redrawn from Aleyev (1977).

"Nekton describes collectively all swimming animals that are free to choose their path ... [and it is possible] ... to create a new biohydrodynamic conception oflife forms of the pelagic zone which, apart being helpful in a quantitative assessment of the dividing line between plankton and nekton, made possible the establishment of the basic factors, determining the development of all essential nektonic adaptations:' In this regard Aleyev's slenderness ratio, U, is particularly helpful. U = D/Lc, where Lc =compact length of the body without thread-like or leaf-like appendages, and D = diameter of a circle equal to the maximum cross-sectional area of the animal's body. U is 1.0 when the organism is round and less than 1.0 when it is long and thin. Thus, (Figure 3), zooplankton vary from round to long and narrow, whereas nektonic animals, including the category called "xeronekton" (aquatic birds, mammals, and reptiles), are all quite slender, with U < 0.4 and converging on a ratio of0.2 as Re increases commensurate with increased size and absolute waterline length. Line A in Figure 3 represents the average slenderness ratio for all nekton, whereas zone a represents the region of optimal hull elongation for submarines. Below this optimal hull ratio propulsion becomes energetically wasteful owing to frictional drag (CDf) and above this ratio form drag (Cop) becomes costly. As these curves converge, just so do body shapes of the largest nektonic animals converge to resemble submarines, with torpedo-like bodies and almost no protuberances except those fins necessary to control pitch, yaw, and roll (Figure 2). Nekton is a readily recognizable category of oceanic animals, yet one must not rely exclusively on quantitative categories when these are obviously biologically misleading. For example, Aleyev (1977) incorrectly considered the Chaetognatha to be one of his 12

24

W.M.HAMNER

categories of nektonic animals simply because they have an appropriate slenderness ratio. He ranked "sperm whales ... directly below sticklebacks and directly above dogfish ... " (Hamner, 1979) in a tabulation of nektonic species by an index of tail-fin shape, and goldfish just above blue whales. Such a biologically nonsensical tabulation does not help refine our understanding of nekton as an adaptive category in open water and Aleyev mixed up his animals and his habitats badly, yet his approach still has enormous merit because it clarifies quantitatively why nekton converge with submarines in terms of their slenderness ratio, U. Thus, "Nekton describes an ecomorphological type ofbiont (animal) which is suspended in the mass of water all or most of the time, which is capable of sustained active propulsion in the horizontal direction and swimming in the regimen ... Re > 103 ... [and which have] ... adaptations functionally associated with decreasing hydrodynamic resistance and increasing the capacity for active propulsion in the mass of water with the minimum expenditure of energy" (Aleyev, 1977). A satisfactory qualitative treatment of oceanic nekton emerges from N. V. Parin's Ichthyofauna ofthe Epipelagic Zone (1970). According to Parin, oceanic nekton fall into two categories: large, mostly solitary, fast swimming predators (mostly > 30 em) and smaller, slower, schooling species (15-30 em). Parin's large, fast swimming species survive in the epipelagic ocean because they can evade predators and can outswim and devour their prey. In terms of their sensory equipment, all nekton have large eyes and are highly visual, many with overlapping fields of view and, presumably, with excellent depth perception. This is true for most categories of oceanic nekton, the sharks, bony fish, squid, seals, and cetaceans, animals that have shaped the nature of the surface waters of the open sea. Anything alone, small, and visible in the sunlit, 3-dimensional coverless expanse of blue water is quickly captured and eaten. Prey escape fast, visual predators by vertical migration into deep water, by schooling, by transparency, by growing too large or reducing their size to below the visual acuity of the predators, or by exploiting the special properties of the sea's surface. Some species do not fit into any of these six modal, convergent groups of epipelagic animals. For example, oceanic triggerfish and filefish, sunfish, pelagic sting rays, manta rays, oceanic puffers and pelagic pipefish all swim quite slowly with undulations of their dorsal and/or ventral fins, and Parin would except these from the nekton, based on their swimming abilities. Yet one wonders how they manage to survive in the epipelagic if they cannot outswim their pelagic predators. For sunfish, triggerfish, puffers and sea turtles the answer must relate at least in part to their extremely tough skin. Pelagic pipefish hide in or mimic floating debris (personal observations). Mature manta rays may not have any predators because of their size. MICRONEKTON AND DIURNAL VERTICAL MIGRATION Parin (1970) did not constrain his survey to include only those fishes that are exclusively nektonic. He noted a third group of animals, the micronekton, intermediate in size between nekton and plankton, (5-15 em in length), which can swim rapidly for short distances but which cannot sustain long periods of active horizontal swimming. "The small planktonic fishes (e.g. the myctophid and gonostomid types) form within the plankton a single group together with the euphausiids, shrimps, and small squids, to which they are very similar in swimming performance and in their position in the food chain. . .. American authors call this group of plankton (which is of great importance in the diet of tuna) 'forage organisms' and regard it as separate from the net zooplankton:' 'Macro-

EVOLUTION IN THE EPIPELAGIC OCEAN

25

plankton' was the older Russian equivalent of the American term 'forage organisms' and 'micronekton' is the current term in the literature for those small epipelagic species that do not swim extensively in the horizontal but which do move hundreds of meters vertically every day, the micronektonic diurnal vertical migrators. Micronekton are captured at depth during the day by large trawls, but at night many of these species can be caught quite near the surface because of their diurnal vertical migration patterns (Pearcy, 1983). Micronekton of all taxa are rather small creatures, about 2 to 10 em long (Omori and Ikeda, 1984), often highly pigmented, black and red, with large eyes and a full complement of those sense organs typical of that particular taxonomic group. Micronekton are seldom seen alive by humans, except from submersibles or remote vehicles. A few divers have filmed these migrators at night in the open sea, and occasionally images of them appear in popular nature magazines. Many micronekton (e.g. myctophid fishes, euphausiid and sergestid crustaceans) are highly bioluminescent. They live at the lowest levels oflight penetration during the day and at night they are often near the surface, which then is illuminated by the stars and the moon. These small animals obviously are difficult prey in dim light or at night for diurnal, visual, nektonic predators, and during the day diurnal vertical migration is a necessary antipredatory behavior for these relatively slow swimming, bite-sized animals. Although they are slow swimmers, they easily swim up to a thousand meters round trip each day to avoid surface predators. At about 10 em/sec a euphausiid need spend only three hours a day in directed swimming to reach the surface and to return to depth; those that don't get back to deep water by dawn are eaten. An enormous amount has been written about diurnal vertical migration (see references in other articles from this symposium), but most of this literature deals with freshwater animals that are not micronektonic. Limnologists, perhaps properly, consider hypotheses other than predation to explain the vertical movement patterns that they observe, but at sea these alternatives are of limited applicability because at sea predation drives most, if not all, diurnal vertical migratory behavior. No biologist who has ever dived in blue water would question this dogmatic assertion. Biologists and lost euphausiids are both absolutely exposed to predators near the surface of the sea during the day. The absence of cover in the epipelagic ocean is without question the single most important issue for pelagic ecology and evolution. SCHOOLING SPECIES At sea and in modest to larger bodies offresh water, nektonic animals are present both as free swimming individuals and as members of schools. In contrast to the larger nektonic animals that capture prey and avoid predators by their ability to swim at high speed, schooling species achieve protection from predators within the confines ofthe structure of the school. Fish, squids and crustaceans that form schools typically aggregate in monospecific schools of an exceedingly narrow modal size range, and schooling individuals invariably are morphologically and behaviorally similar (Packard, 1972). The lower size range of animals that form schools is smaller than that of nekton which survive as lone individuals in open water. This is because schooling provides means of protection other than sheer speed (Pitcher and Parrish, 1993). Schoolers are partly protected by the very presence of other individuals within the school. Much has been written about these issues, but the importance of group membership is unquestionable when many vulnerable prey are in the presence of a predator that eats only one individual at a time. Our

26

W.M.HAMNER

own perceptions confirm this. SCUBA divers in the open sea are far more comfortable in the presence of sharks when other divers are close by, because one's probability of being eaten decreases linearly with an increase in the number of colleagues in the water (the "buddy system"). Schoolers achieve protection also because the predator becomes optically confused when all of the prey look alike. Most predators attack only one prey at a time and, if visually distracted by crowd confusion, the predator cannot catch the individual initially selected (Pitcher and Parrish, 1993). Animals in schools engage in complex coordinated patterns of movement when under attack, and they collectively perform behaviors while in schools that they could not display alone. The best known collective behavior is called "balling," wherein the animals in the school move closer together so that they almost touch one another and the school itself becomes more regular, rounded and ball-like. Schools may cruise, split, form hourglass shapes, exhibit vacuoles of clear water within the school itself, join other schools, "fountain" around a predator so that the school regroups behind the predator, engage in group avoidance or evasion, or, in extreme cases, where the risk of remaining in the school becomes greater than the risk of isolated escape behavior, engage in "flash expansion", where all of the individuals suddenly scatter randomly and at high speed (Pitcher and Parrish, 1993). The antipredatory effectiveness of these collective maneuvers increases as the size of the school increases. This has been shown for cephalopods, pike and perch, sticklebacks, piranhas and mink, and seems to be widely applicable. When prey are separated from a group they are far more likely to be eaten by a predator than if they had remained within the group (Pitcher and Parrish, 1993). The two most recent reviews on schooling in the literature are by Pitcher (1986) and Pitcher and Parrish (1993), both of which explicitly discuss the concept of cover but which dismiss it as an important consideration for schooling, claiming that the earlier arguments of Williams (1964) and Hamilton (1971) regarding the importance of cover were wrong. Pitcher (1986) devoted considerable attention to an attempt to dismiss the importance of cover as a factor in the evolution of schooling, but in Pitcher and Parrish (1993), an update of the 1986 review, although they claim in the introduction that " ... the arguments presented in this chapter do not support the view that shoaling is primarily a matter of cover-seeking ... ",there are actually none of Pitcher's prior arguments related to cover-seeking in the chapter that follows. Indeed the term "cover" is not mentioned thereafter at all. Thus the idea of cover-seeking is baldly dismissed by Pitcher and Parrish (1993) without explanation, and one suspects that the authors in the second review did not in retrospect find the previous presentation (Pitcher, 1986) any more compelling than did I. There is no question in my mind that schooling behavior does provide protective cover to the individual squid or fish or shrimp. There is also no question that, when there is sufficient inanimate cover in the vicinity, aquatic prey obtain protection from predators in precisely the same way that Elton's cats used cover when engaged in play behavior. If there are any objects behind which or amid which one can seek shelter, then it is more difficult for a predator to capture prey. Although it is even more difficult for a predatory fish to capture prey in the midst of a school because of the collective behavior of the members of that school, the physical presence of the objects themselves, whether animate fish or inanimate rocks, nonetheless provides cover and a clear measure of protection for the prey (Williams, 1964; Hamilton, 1971). Many fishes school when away from cover but flee individually to cover at approach of a predator if refuges are available (Pitcher and Parrish, 1993). I have perhaps belabored the obvious in this consideration

EVOLUTION IN THE EPIPELAGIC OCEAN

27

of cover as it relates to the selective advantages of schooling as a modal adaptive strategy in the open sea, but it is important scholastically to evaluate one's arguments with appropriate reference to previous contributions on the subject. As noted by Williams (1964), the prevalence of schooling is most dramatic in habitats devoid of cover, and I concur with Williams (1964) and Hamilton (1971) rather than with Pitcher (1986) regarding the central importance of cover in the evolution of schooling behavior. The essence of the school as an antipredatory device is nicely stated in Pitcher and Parrish (1993), where the predator in an environment with limited visibility and in the presence of a visually confusing crowd of prey finds it far easier to capture individual stragglers than to focus on one animal within the school and to successfully effect its capture. They note that "Under these circumstances, grouping pays; for example, an individualleaving a group of 100 reduces its chance of encounter from one to a half, but its chance of death given an attack goes from one-hundredth to one. Put another way, this is simply that it will be a good idea to be a member of a group if a predator comes across you, predators being unpredictable and nasty things;' Pitcher and Parrish (1993) also nicely present the concept that prey, on both an immediate behavioral and long term evolutionary basis, must evaluate the relative advantages of remaining within a school versus the advantages of leaving. In the absence of predation, it may be generally disadvantageous to stay within the school because of competition for food or because of potential deterioration of the water within which they swim due to oxygen depletion and the build up of metabolic wastes if the school is very large. This must happen evolutionarily also, an example from our experience with Antarctic krill (f:uphausia superba) serving as an example. As noted earlier, Antarctic krill (after metamorphosis from furcilia VI, Hamner et al., 1989) are always found within the structure of highly organized schools, and they exhibit a biology that is similar to that of temperate zone clupeid fishes (Hamner, 1984). Baleen whales are important predators of krill in the Antarctic (certainly the most conspicuous), and the aggregated patterns of krill appear in this context to be disadvantageous because many of these great whales (certainly the rorquals which feed via envelopment with an expandable throat pouch, as opposed to the right whales that power-filter large volumes of water with their jaws agape) cannot feed on krill unless they are tightly aggregated. Why then do krill school? Obviously, their dominant predators must feed on individual krill one at a time. Penguins and fish are the obvious candidates, although there are almost no epipelagic bony fish in the Antarctic except Pleurogramma antarctica; midwater fishes such as myctophids actually may well be their most important predators (Robison, personal communication). Whales may be quite recent predators relative to the evolution of schooling of krill. The schooling behavior of krill thus reflects the evolutionary sum of the advantages of avoiding predators that eat one prey at time versus the disadvantages of being swallowed along with a crowd of colleagues by a whale. Most nekton, whether schooling or not, begin their lives as immobile, transparent fertilized eggs, as part of the net plankton, and they graduate into tiny transparent fish larvae that achieve protection through invisibility. Later they metamorphose into small schoolers with the sole responsibility of growing as fast as conceivably possible to that final nektonic size at which absolute swimming speed or protection within the school permits the luxury of diverting resources to reproduction. Nekton thus employ four different modal antipredatory strategies during development and they must successfully make the transitions from net zooplankton (eggs) to zooplankton (transparent larvae), to schooling juveniles, and finally to the nektonic adults.

28

W.M.HAMNER

NET ZOOPLANKTON When one places a plankton net of almost any design into the sea during the day one captures myriads of small zooplankton, all ofwhich are almost transparent and which are generally smaller than about 2 mm. The taxonomic diversity of the creatures captured from almost any body of salt water is bewildering and marvelous. Many of us have had our entire lives changed after our first look through the microscope at this stunning array of tiny herbivores and micro-monsters. Certainly these are members in good standing of the zooplankton, drifters with the ocean currents, described so elegantly in Alister Hardy's (1965) The Open Sea, my first textbook of biology. Other textbooks and monographs elaborate on Hardy's rather simple definition, and some (e.g. Omori and Ikeda, 1984, modified after Dussart, 1965) even subdivide the plankton by metric size, using Latin prefixes to create categories ofultrananoplankton ( < 2 ttm), nanoplankton (2-20 ttm), microplankton (20-200 ttm), mesoplankton (200 t-tm-2 mm), macroplankton (2-20 mm), micronekton (20-200 mm) and megaloplankton (gelatinous plankton > 20 mm). Other schemes of categorization emphasize habitats instead of size [Haeckel's (1890) "ploteric" and "spanipelagic"; Aleyev's (1977) "euplankton", "planktonekton" and "nektoplankton"]. I advocate use of the term "net zooplankton" instead of 'mesoplankton' to connote those creatures that are captured by an approximately 200 ttm mesh zooplankton net, a definition somewhat devoid of imagination, yet a term that is becoming increasingly widely used because of its functional simplicity (Omori and Ikeda, 1984; Landry eta/., 1994). Although I have no personal insight into the biological validity of the categories of plankton smaller than about 200 Jtm, those categories for plankton above this size represent functional or real biological entities 1• As noted by Omori and Ikeda (1984), "The smallest size of net plankton, namely those sampled by nets is 200 Jtm. Plankton smaller than that are difficult to sample quantitatively even in the oligotrophic open ocean because of clogging of nets:' The lower end of the size range for net zooplankton (200 ttm) is thus defined operationally by the mesh size of an effectively deployed net. The upper end of the size range for net zooplankton (2 mm) is defined by the absence of any zooplankters larger than this in surface waters during the day, because larger plankton would be seen and eaten by visual predators. The term net zooplankton thus permits us to distinguish this array of tiny creatures from the generally, but not always, much larger gelatinous megaloplankton, to be discussed later in more detail. Gelatinous zooplankton are rarely captured in small zooplankton nets because these animals often are widely dispersed and uncommon. When gelatinous animals are abundant, they are often terribly abundant, and when oceanographers fill their nets with a gelatinous mass of damaged tissues, they usually throw out the catch. When one views the upper layers of the sea through a face mask, the window of a submersible or a camera lens, one sees only blue water and occasionally gelatinous zooplankton. Net zooplankton are extremely abundant in these apparently empty, blue waters, but they are too small to be resolved by the human eye, and we can't see them. Similarly they are too small to be seen by most predatory fish. Net zooplankton, therefore, utilize cover in epipelagic waters in the wider sense of Charles Elton's "protective system of the prey" by avoiding visual predators through overall reduction in size. As overall size of the animals decreases, so does the surface area to volume ratio increase. 1 Below 200 J.Lm, of course is a wealth of tiny plants, animals, protozoans, bacteria and viruses. Their importance in the economy of the sea is enormous, but they are not pertinent to this discussion except to note that the food webs of the open sea are even more complex than we envisioned before we were aware oftheir significance.

EVOLUTION IN THE EPIPELAGIC OCEAN

Figure 4

29

Calocalanus pavo. Redrawn from Davis (1955).

Frictional drag increases, swimming speeds decrease, and net zooplankton become more like an idealized passive particle, supported and adrift in the sea. Aleyev (1977) defines plankton as "an ecomorphological type of biont (plant or animal) which is suspended in the mass of water all or most of the time ... and which swims in the regimen ... Re < or = 5.0 x 105 ... and whose general body structure is determined ... by ... adaptations associated with hydrodynamic resistance ... and passive hovering ... with minimal expenditure of energY.'At low Reynolds numbers the total drag on a body moving in water and air alike depends very little on its shape, so for net zooplankton there is a wide variation in Aleyev's slenderness ratio U < 0.10 to > 1.0. Referring again to Figure 2, we see where zooplankton lie on Aleyev's plot of slenderness ratio versus Reynolds number, and in Figure 4 a drawing of Calanocalanus pavo illustrates just how extreme the surface area to volume ratios can become in the realm of the plankton. Net zooplankton in the surface waters of the blue ocean are invisible because they are tiny. As sunlight attenuates with depth, so also does visual acuity; therefore the absolute size of net plankton should, and does, increase with depth because in darker waters somewhat larger animals are still impossible to see. Of course, in water several hundreds of meters deep, downwelling light is very dim indeed, and it is here that the net zooplankton community finally gives way to the midwater assemblage of micronekton, both migrators and also non-migrators. Larger net zooplankton move from the surface into dimmer waters during the day, even though they do not migrate to the extreme depths exhibited by the micronektonic champion migrators. In a recent publication on net plankton migration Hays eta/. (1994) compared the temporal distributions of 24 species of copepods from some 13,000 net plankton samples collected over a 44 year period in the northeast Atlantic, using a continuous plankton recorder with 270 p,m mesh nets from ships of opportunity. Hays et a/. (1994) examined the presence versus absence of species in samples taken in the day and in the night near the surface (6.5 m) in order to

30

WM.HAMNER

assess the actual size of those copepods that stayed at the surface during the day as opposed to those that swam down, perhaps only 20-30 m, during the day. They found a clean behavioral demarcation between migrators and non-migrators at a prosome width of 1.0 mm. Thus, even at the very surface there are quite subtle changes in depth due to diurnal vertical migration ofeven the smallest members of the net plankton community, migratory patterns that are associated directly with the degree of visibility and relative vulnerability during daylight hours to visual nektonic predators. Net zooplankton are all small. Consequently their sense organs must be commensurately tiny. Net zooplankton mostly do not have large and complicated sense organs, although some copepod eyes are fairly complex. Although there must be structural versus size constraints, there are also perceptual constraints. Tiny animals that spend much of their lives within a viscous boundary layer may have perceptual fields that are limited to distances of only about one body length for most external stimuli (Haury and Yamazaki, 1995). For those that respond to light, and most do in some way, photoreceptors need only indicate direction and intensity; image forming eyes would seem to have little utility for an animal only 2 mm long. The fancy eyes and elaborate photophores that characterize micronektonic crustaceans, fish, and squid have no equivalents amid the net zooplankton in the well illuminated surface waters of the open sea. Simple olfactory and tactile sensors, widely distributed about the body, and a simple directional light detection system probably provide all of the sensory information necessary for most of the members of the net zooplankton. GELATINOUS ZOOPLANKTON When divers float suspended in the open sea, the only animals that usually can be seen are gelatinous zooplankton (Hamner, 1974, 1977; Hamner et al., 1975), although large nektonic animals do sometimes visit divers at sea. Underwater we occasionally see sharks, tunas, bill fish, turtles, ocean sunfish, marine mammals, sea birds, sometimes pelagic sea snakes, and, less often, schools of fish, krill and squid, although these tend . to avoid divers in open ocean. Relatively large gelatinous zooplankton, ca. 2 to 20 em, are almost always present, however, although they are difficult to see. John Lythgoe (1979) commented on the issue of invisibility at sea. He wrote, "Logically we do not know if there are any perfectly camouflaged animals because we could not see them if there were.... Amongst the zooplankton it is really true that in the open sea an alert and normally sighted diver simply does not see the plankters that surround him and may have difficulty in detecting one several centimeters in diameter at a comfortable reading distance even when its position is indicated by the pointed finger of his companion. In the case of zooplankters, it is the gonads and nerves that remain opaque, or at least translucent. But they tend to be more dispersed within the tissues than in the higher animals and thus defeat the resolution of the observer's eye. Extreme transparency is recorded in many unrelated groups including the Hydrozoa, siphonophores and true jellyfish, comb jellies, Heteropoda and Pteropoda, salps, larvaceans, and arrow worms:' In the course of our own work many of us have taken brilliant, beautiful photographs of these poorly known animals (e.g. see Hamner, 1974), but it is important to remember that almost without exception these photographs were taken with artificial illumination. In ambient light it is possible to see the animal of interest only when it is backlit by the downwelling light that penetrates Snell's window.

EVOLUTION IN THE EPIPELAGIC OCEAN

31

The most thoughtful discussion ofthe issue of transparency in epipelagic animals is by McFall-Ngai (1990), who noted that the factors which affect the quality oflight vary enormously in different environments. For example, the only factor that affects the light environment in terrestrial habitats is the vegetation, which shades the substrate in extremely complex ways. In benthic aquatic and nearshore pelagic environments, both biotic elements, such as bacteria, plants and animals, and abiotic elements, such as the medium itself, dissolved materials and suspended particles, all modify the quality of light. In contrast, the oceanic pelagic realm contains no sediments, phytoplankton blooms are rare, and only the medium of sea water has a moderating influence on the nature of the downwelling light. Here the light environment is so simple, constant and predictable that the adaptive responses of pelagic organisms are also relatively simple, constant, and predictable. Indeed, "The light environment of the pelagic zones has selected for the convergent evolution of transparency ... in a wide variety of pelagic phyla.... As a mode of crypsis, transparency differs from most others by involving most, if not all, of the anatomy and morphology of the animal" (McFall-Ngai, 1990). Convergent evolution of transparency has occurred in quite unrelated phyla, and one might expect the physiological mechanisms that are exploited to achieve invisibility might be somewhat different in different phyla. Among the mechanisms involved are reduced chromatophore content (all groups), high water content and low tissue complexity (medusae, ctenophores, salps), bodies thin in at least one dimension (fish larvae, many crustaceans, phyllosome larvae), and regular arrangement and physical alignment of cellular components (chaetognaths; Chapman, 1976a, b). The effectiveness of transparency is particularly obvious when one member of a group of transparent animals is somehow unwell. Tissue transparency, whether in gelatinous zooplankton or in the human eye, is maintained only in healthy tissues. An eye lens with a cataract and a dying jellyfish both exhibit cloudy and whitish tissues. Occasionally one sees salp chains at sea in which only one individual in the chain is chalky white and all the rest are perfectly clear, except, of course, for the gut. No animal can produce transparent feces or invisible visual pigments. We have also seen many individual whitish krill within dense krill schools while diving in the Antartic. When these were captured individually by hand in separate jars and examined within the hour through a microscope, 100% of these krill were found to be parasitized, improperly molted, or physically damaged. All of the whitish animals captured this way died within the next day, whereas transparent, apparently healthy animals captured similarly as controls lived for many weeks in the collection jars (Hamner, 1984). Whole body transparency, like other cryptic mechanisms, although perhaps not as energetically costly as diurnal vertical migration, is under active metabolic control. It is worth reminding the non-specialist that most of the tiny net zooplankton that live in surface waters in both freshwater and marine habitats are also transparent, and indeed much more has been written about this issue for freshwater forms (see many references in Kerfoot, 1980). It is important to remember that net zooplankton are almost always < 2 mm in size, and these small species readily achieve transparency by the simple expediency of elimination of pigments. It is astonishing how vulnerable freshwater net zooplankton become if their eye spots are slightly too large (Zaret and Kerfoot, 1975). Similar studies of the relative vulnerability of marine net zooplankton or of gelatinous zooplankton have not been conducted. Gelatinous zooplankton, as noted above, includes a spectrum of unrelated taxa, yet almost all of these animals are without structurally complex sense organs. For example, most medusae that do have sense organs only have tiny eye spots and statocysts; cteno-

32

W.M.HAMNER

phores have extremely small, simple apical organs. Appendicularians have no visible sensory equipment, and chaetognaths seem to react only to vibrations, having only rudimentary pigment spots that apparently perceive light. Even large pteropod mollusks have no eyes, and salps, doliolids, and pyrosomes seem positively without concern for their surroundings, cruising through the water column without any driver at the wheel (although see Madin, this conference). Only heteropod mollusks have complex sensory equipment, using complex eyes to hunt tasty prey such as salps. NEUSTON (SWIMMERS) AND PLEUSTON (FLOATERS) The surface of the sea provides a unique environment in the pelagic realm, promoting the evolution of a host of strange and marvelous creatures found nowhere else on the Earth. These organisms include Halobates (Insecta), Physalia, Velella and Porpita (Cnidaria), Ianthina, Fiona and Glaucus (Gastropoda), Lepa (Cirripedia), and many species of larval and adult fishes (Vertebrata) (Zaitsev, 1970; Cheng, 1975; Moser, 1981). The sea surface has unique properties which these species exploit in unusual ways. The sea surface provides a barrier between the air and water which is variously penetrable to different species. From below, for most species, this interface is the ceiling to their vertical migration. At night the sea surface provides the barrier against which most diurnal vertical migrants accumulate. Just as a stoplight on a long country road accumulates vehicles because they must slow down or stop or turn, so does the sea surface slow down, stop, turn about, and accumulate the diurnal vertical migrators each night. The sea-air interface is the most reliable site in the ocean for aggregating. Predators can reliably find prey at the sea surface at night in high concentration. In the dark it is difficult for visual predators to see their prey; nonetheless much nocturnal feeding by these species does occur at the surface at night. Tentaculate predators do not experience this disadvantage and collectively the carnage at night at the sea surface is extraordinary. At night, oceanic plankton sometimes are packed against the surface in layers so thick that human vision is reduced to less than a meter. With plankton this thick it is difficult to imagine how any given species manages to swim, much less perceive and capture prey or evade predators. Perhaps feeding at night at the surface is somewhat random with predators swallowing whatever they happen to bite. No one has ever investigated this problem; it will not prove easy to study. The sea surface is somewhat penetrable from below, as illustrated by the effective use of this barrier by flying fishes, which dart into the air to avoid predators, but many other species of fish and particularly larval fish (Moser, 1981), exploit the sea surface not only because it concentrates food, but also because the shimmering, rippling surface produces an optically complex habitat that somehow protects small, transparent, or schooling fishes. No one has conducted experiments to learn just how this complex interface provides protection from visual predators, but all authors writing about this habitat concede that this must occur because the sea surface is a nursery for a huge variety oflarval fish. The sea surface protects the animals that float above or depend below this interface, but it also exposes them both to predators and to harmful radiation from sunlight. Seabirds glean prey from the sea and although many dive below the surface, others pluck prey from the surface while on the wing. Ultraviolet radiation strikes the sea surface directly and pleuston and neuston must somehow deal with this source of potential damage (Moser, 1981). The brilliant blue pigment with absorption maxima 630-660 nm

EVOLUTION IN THE EPIPELAGIC OCEAN

33

which characterizes many pleustonic species may help in this regard and also may provide protective coloration that shields them from aerial visual predators. This pervasive blue pigment in all phyletic groups is clear evidence, once again, of the importance of convergent evolution in the epipelagic realm. CONCLUSIONS Each of the modal categories discussed above cannot be construed as a sufficient descriptor for all of the species which are members of that class. Convergent selection has shaped the evolution of epipelagic oceanic animals perhaps more dramatically than in any other habitat on the planet, excepting perhaps the bizarre convergence in body form shown by earthworms, caecilians (apodan amphibians), and legless lizards (4.nguis fargilis, the slow-worm) in fossorial habitats. Although there are a huge number of epipelagic similarities due to homoplasy within convergent assemblages (Eldredge, 1989), each species still must be able to function uniquely within an ecosystem seemingly devoid of discrete ecological niches and in the face of intense competition from other species, all of which superficially do much the same things. This issue of competition in a seemingly niche-free habitat has not been adequately addressed by marine ecologists or by evolutionary biologists (Hutchinson, 1961). With regard to the convergent evolution of nektonic fishes, those ecologists who study oceanic fish are trained traditionally in Fisheries Biology and their job descriptions primarily relate to fish production, whereas those ecologists who occasionally do discuss convergence are mostly terrestrial biologists (e.g. Cody and Mooney, 1978). Modern evolutionary biologists treat convergence either not at all or as an impediment to cladistic analysis (Futuyma, 1983). All of this aside, all those species within a convergent guild must manage somehow to partition resources, even though we lack sufficient insight into how this occurs. The animals within the modal groups that characterize epipelagic oceans, of course, could not persist over evolutionary time either as perfect predators or as infinitely safe prey. Perfect predators would decimate their prey and unattainable prey would overwhelm their resources. Clearly there must be trophic connections between these modal groups, and these linkages are often marvelously strange. A simple linkage connects the tentaculate carnivores (medusae and ctenophores) in the guild of transparent, gelatinous zooplankton to their prey (copepods and other small animals) in the guild of net zooplankton. During the day net zooplankton are relatively safe from one category of predator (large nektonic fish) yet vulnerable to another (gelatinous zooplankton). Small hydromedusae and ctenophores are themselves eaten by large scyphomedusae, which are eaten by sea turtles and ocean sunfish (Mola mola). Adult Mola mola are relatively free of predators, but ocean sunfish are heavily parasitized, and it is true that there really is no free lunch, even at sea. Another example of trophic interactions between guilds is worth comment. Yellowfin tuna in the equatorial Pacific feed during the day at the surface on juvenile squid, the adults of which normally reside during the day at depths of hundreds of meters in dimly illuminated waters where they cannot easily be caught by visual hunters. Schools ofyellowfin tuna, however, drive the juvenile squid upward until they are bunched against the surface and highly visible and then the tuna hurtle into the aggregated mass of squid to feed. The squid try to escape the tuna by jetting out of the water, but when they do they are captured by seabirds attracted to the feeding frenzy. This interguild trophic transfer was first de-

34

WM.HAMNER

scribed by ornithologists (Ashmole and Ashmole, 1968), perplexed as to how some of the seabirds that nest on tropical islands manage to feed their young with fresh midwater squid and myctophids in the middle of the day, a time when these animals should be at depth, far below the diving range of surface feeding birds. This marvelous interrelationship of squid, fish, and seabirds illustrates that modal predatory and antipredatory patterns are not foolproof, but the particular and totally unexpected nature ofsome of these interguild interactions also illustrates the effectiveness of modal antipredatory patterns and the complexity of the behaviors necessary to circumvent them in a three dimensional medium devoid of cover. Acknowledgements

I thank John Hunter, Steve Strand, Larry Madin, Peggy Hamner, Richard Young, and an anonymous reviewer for thoughtful suggestions which improved this review. References Aleyev, Yu. G. (1977) Nekton. Dr. W Junk, The Hague. vi + 435 p.f. 120. Ashmole, M. J. and Ashmole, N. P. (1968) The use of food samples from sea birds in the study of seasonal variation in the surface fauna of tropical oceanic areas. Pac. Sci. 22, 1-10. Chapman, G. (1976a) Reflections on transparency. In: Coelenterate Ecology and Behavior, G. 0. Mackie, ed. Plenum Press, New York, pp. 491-498. Chapman, G. (1976b) Transparency in organisms. Experientia 15, 123-125. Cheng, L. (1975) Marine pleuston- animals at the sea-air interface. Oceanogr. Mar. Bioi. Ann. Rev.13, 181-212. Cody, M. L. and Mooney, H. A. (1978) Convergence versus nonconvergence in Mediterranean-climate ecosystems. Ann. Rev. Ecol. Syst. 9, 265-321. Dayton, P. K. (1984) Processes structuring some marine communities: Are they general? In: Ecological Communities, D. R. Strong, Jr., D. Simberloff, L. G. Abele, and A. B. Thistle, eds., Princeton Univ. Press, Princeton, New Jersey, pp. 181-197. Dejours, P. (1987) Water and air: Physical characteristics and their physiological consequences. In: Comparative Physiology: Lifo in ffilter and on Land, P. Dejours, L. Bolis, C. R. Taylor, and E. R. Weibel, eds., FIDIA Res. Ser. vol. 9. Springer-Verlag, New York, pp. 3-11. Denny, M. W (1990) Terrestrial versus aquatic biology: the medium and the message. Amer. Zoo/. 30, 111-121. Denny, M.W (1993) Air and Jfilter. Princeton Univ. Press, Princeton. 341 pp. Dussart, B. H. (1965) Les differentes categories de plancton. Hydrobiologia 26, 72-74. Eldredge, N. (1989) Macroevolutionary Dynamics. McGraw-HilJ, New York. 226 pp. Elton, C. (1939) On the nature of cover. J. Wildlife Management 3, 332-338. Elton, C. (1966) The Pattern of Animal Communities. Wiley, New York. 432 pp. Futuyma, D. J. (1983). Evolutionary interactions among herbivorous insects and plants. In: Coevolution, D. J. Futuyma and M. Slatkin, eds. Sinauer, Sunderland, pp. 207-231. Haeckel, E. (1890) Planktonic studies: A comparative investigation of the importance and constitution of the pelagic fauna and flora. Jena. Z. 25 (English trans!. 1893, Rep US. Commnr. Fish and Fisheries, 1889-91, pp. 565-641. Hamilton,W D. (1971) Geometry for the selfish herd. J. Theor. Bioi. 31, 295-311. Hamner,W M. (1974) Blue-water plankton. Nat/. Geogr. Mag.146, 530-545. Hamner, W M. (1975) Underwater observations of blue-water plankton: Logistics, techniques and safety procedures for divers at sea. Limnol. Oceanogr. 20, 1045-1051. Hamner,W M. (1977) Observations at sea oflive tropical zooplankton. Proc. Symp. Jfilrm ffilter Zoopl. Special Pub!. UNESCO/NIO: 284-296. Hamner, W M. (1979) Review of Nekton by Yu. G. Aleyev. Limnol. Oceanogr. 24, 1173-1175. Hamner,W M. (1984) Aspects of schooling in Euphausia superba. J. Crust. Bioi. 4 (Spec. No.1), 67-74. Hamner, W M. (1985) The importance of ethology for investigations of marine zooplankton. Bull Mar. Sci. 37, 414-424. Hamner,W M. (1988) The 'lost year' of the sea turtle. TREE 3, 116-118. Hamner, W M., Gilmer, R. W, and Hamner, P. P. (1982) The physical, chemical, and biological characteristics of a stratified, saline, sulfide lake in Palau. Limnol. Oceanogr. 27, 896-909.

EVOLUTION IN THE EPIPELAGIC OCEAN

35

Hamner, W. M., Hamner, P. P., Obst, B. S. and Carleton, J. H. (1989) Field observations on the ontogeny of schooling of Euphausia superba furciliae and their relationship to ice in Antarctic waters. Limnol. Oceanogr. 34: 451-456. Hamner, W. M., Hamner, P. P., Strand, S. W., and Gilmer, R. W. (1983) Behavior of Antarctic krill, Euphausia superba: Chemoreception, feeding, schooling, and molting. Science 220, 433-435. Hamner,W. M. and Hauri, I. R. (1981) Long-distance horizontal migrations of zooplankton (Scyphomedusae: Mastigias). Limnol. Oceanogr. 26,414-423. Hamner,W. M., Madin, L. P., Alldredge, A. L., Gilmer, R.W., and Hamner, P. P. (1975) Underwater observations of gelatinous zooplankton: Sampling problems, feeding biology, and behavior. Limnol. Oceanogr. 20, 907-917. Hardy, A. (1965) The Open Sea: Its Natural History. Part L The UVrld ofPlankton. Houghton Mifflin, Boston, 335 pp. Haury, L. R. and Yamazaki, H. (1995) The dichotomy of scales in the perception and aggregation behavior of zooplankton. J Plankton Res. 17, 191-197. Hays, G. C., Proctor, C. A., John, A.W. G., and Warner, A. J. (1994) Interspecific differences in the die! vertical migration of marine copepods: The implications of size, color, and morphology. Limnol. Oceanogr. 39, 1621-1629. Hutchinson, G. E. (1961) The paradox of the plankton. Amer. Nat. 95, 137-145. Isaacs, J. D. (1977) The life of the open sea. Nature 261, 778-780. Kerfoot, C. (1980) Commentary: Transparency, body size, and prey conspicuousness. In: Evolution and Ecology of Zooplankton Communities, C. Kerfoot, ed. Univ. Press New England, Hanover, New Hampshire. pp. 609-617. Landry, M. R., Peterson,W. K., and Fagerness,V L. (1994) Mesozooplankton grazing in the Southern California Bight. I. Population abundance and gut pigment contents. Mar. Ecol. Progr. Ser. 115, 55-71. Lythgoe, J. (1979) The Ecology of Vision. Clarendon Press, Oxford. 244 pp. McFall-N gai, M. J. (1990) Crypsis in the pelagic environment. Amer. Zoo/. 30, 175-188. Moser, H. G. (1981) Morphological and functional aspects of marine fish larvae. In: Marine Fish Larvae, R. Lasker, ed.,Washington Sea Grant Program, Univ.Washington Press, Seattle. Nybakken, J. W. (1993) Marine Biology: an Ecological Approach, 3rd edn. Harper Collins College Pub!., New York, 462 pp. Omori, M., and Hamner, W. M. (1982) Patchy distribution of zooplankton: Behavior, population assessment and sampling problems. Mar. Bioi. 72, 193-200. Omori, M. and Ikeda, T. (1984) Methods in Marine Zooplankton Ecology. Wiley and Sons, New York. 332 pp. Packard, A. (1972) Cephalopods and fish: limits of convergence. Bioi. Rev. 47, 241-307. Parin, N.V (1970) Ichthyofauna ofthe Epipelagic Zone. Israel Prog. Sci. Trans!., Jerusalem, for U.S. Dept. Interior,206pp. Parsons, T. R. (1976) The structure oflife in the sea. In: The Ecology ofthe Seas, D. H. Cushing and J. I Walsh, eds. Blackwell Sci., Oxford, pp. 81-97. Pearcy, W. G. (1983) Quantitative assessment of the vertical distribution of micronektonic fishes with opening/ closing midwater trawls. Bioi. Oceanogr. 2, 289-310. Pitcher, T. J. (1986) Functions of shoaling behavior in teleosts. In: The Behavior of Teleost Fishes, T. J. Pitcher, ed. Johns Hopkins Univ. Press, Baltimore, Maryland, pp. 294-337. Pitcher, T. J. and Parrish, J. K. (1993) Functions of shoaling behavior in teleosts. In: Behavior of Teleost Fishes, 2nd edn. T. J. Pitcher, ed. Chapman and Hall, New York, pp. 363-439. Smetacek, V and Pollehne, F. (1986) Nutrient cycling in pelagic systems: A reappraisal of the conceptual framework. Ophelia 26, 401-328. Steele, J. H. (1985) A comparison of terrestrial and marine ecological systems. Nature 313, 355-358. Vogel, S. (1981) Lifo in Moving Fluids. Willard Grant Press, Boston. Williams, G. C. (1964) Measurement of consociation among fishes and comments on the evolution of schooling. Pub/. Mus. Michigan State Univ. Bioi. Ser. 2, 351-383. Zaret, T. M. (1975) Strategies for existence of zooplankton prey in homogeneous environments. U?rh. Internal. U?rein. Limnol. 19, 1484-1489. Zaret, T. M. and Kerfoot,W. C. (1975) Fish predation on Bosnia longirostris: Body-size selection versus visibility selection. Ecology 56, 232-237. Zaitsev, Yu. P. (1970) Marine Neustonologie. Nauk Dumka, 264 pp. (in Russian, trans!. Israel Programme for Scientific Translations).

CHEMOSENSORY ECOLOGY OF OYSTER LARVAE: BENTHIC-PELAGIC COUPLING RICHARD K. ZIMMER-FAUST, 1 MARION. TAMBURRI 2 '3 '4 and ALAN W DECH0 3,4 1 Department of Biological Sciences, 2 Marine Science Program, 3 Belle W Baruch Institute for Marine Biology and Coastal Research, and 4 Department ofEnvironmental Health Sciences, University of South Carolina, Columbia, South Carolina 29208, USA

Habitat colonization by planktonic larvae is a critical factor regulating population dynamics of marine benthic invertebrates. The chemical properties of marine environments provide important cues used by larvae to select settlement sites. Our results demonstrate a clear association between presence of a dissolved chemical stimulus and rapid behavioral response by oyster larvae. Dissolved substances released by adult conspecifics cause downward-directed swimming in the water column and attachment to substratum by larval oysters (collectively defined herein as "settlement"). As indicated by natural products chemistry and laboratory behavioral assays performed in still water and flume flow, oyster larval settlement inducers are low molecular weight (LMW) peptides with arginine at the C-terminus. Settlement by oyster larvae in response to seawater collected at field sites correlates positively with the concentration of LMW arginine-peptides. Preliminary evidence further suggests that the peptides evoking oyster larval settlement are those also eliciting metamorphosis. We are currently testing the hypothesis that adsorption of LMW arginine-peptides to exopolymers in bacterial biofilms is a key agent regulating larval metamorphosis. Chemical induction of either settlement or metamorphosis might thus be determined by the availability of LMW arginine-peptides in either dissolved or particulate form.

INTRODUCTION Most marine animals possess early larval stages that are carried by ocean currents and function as agents of dispersal for parental stocks. In benthic organisms, planktonic development continues until larvae are competent to metamorphose. The larvae then leave the water column, settle on the seabed and attach to substrata before differentiating into juvenile forms. During the past three decades, it has been widely believed that the structure of benthic populations and communities is largely regulated by mortality among established individuals due to predation, competition, and physical stress rather than larval availability (Connell, 1961; Dayton, 1971; Paine, 1974; Menge and Sutherland, 1976). Recently, however, the importance of larval supply to benthic habitats has been re-emphasized in light of new data. Larval supply has re-emerged as a key regulatory agent in structuring communities (Strathmann et al., 1981; Grosberg, 1982; Connell, 1985; Gaines et al., 1985; Roughgarden et al., 1988; Underwood and Fairweather, 1989). Accordingly, investigators have devoted substantial effort in trying to determine processes that govern larval colonization of benthic substrata (see reviews by Woodin, 1986, 1987; Butman, 1987), but important mechanisms central to these processes still remain unclear. 37

38

RICHARD K. ZIMMER-FAUST ET AL.

The colonization of benthic environments by marine larvae requires both their settlement and metamorphosis (for recent reviews, see Morse, 1990; Pawlik, 1992; Rodriguez eta/., 1993). Settlement is potentially a reversible process delivering larvae from the water column to the seabed and includes substrate exploration by larvae before metamorphosis. Metamorphosis is an irreversible developmental process resulting in biochemical, physiological, and morphological transformations of an individual between biphasic (larva/juvenile) life forms. Settlement and metamorphosis therefore differ and need not be controlled by the same factors. When suspended in the water column, larvae are dependent on passive transport by fluid motion to approach the substratum (Eckman, 1983, 1990; Butman, 1986, 1987; Grosset a/., 1992). However, active larval behavioral responses can affect settlement site selection (see reviews by Crisp, 1974, 1984; Pawlik, 1992). Larvae of benthic animals commonly appear to discriminate between settlement sites. These larvae not only choose between substrata in laboratory experiments, but they cluster on preferred substrata in the field (Crisp, 1974; Keough, 1984; Raimondi, 1988). The chemical properties of marine benthic substrata are known to provide valuable cues for larvae actively colonizing settlement sites (Morse et al., 1979; Johnson and Strathmann, 1989; Jensen and Morse, 1990; Morse and Morse, 1991; Pawlik and Butman, 1993). The induction of substrate exploration is now thought to be principally stimulated by substances adsorbed to the benthic substratum. Such adsorbed cues have been found to be associated with overall organic enrichment (Butman et al., 1988), specific sediment geochemistry (Cuomo, 1985), the sequestration of plants or bacteria encrusting surfaces (Kirchman eta!., 1982; Morse and Morse, 1984; LeTourneux and Bourget, 1988; Maki eta/., 1990), or associations with juvenile or adult conspecifics (see reviews by Burke, 1986; Bonar eta/., 1990; Pawlik, 1992). Presumably, the larvae can only detect these substances after they contact the bottom (e.g., Butman and Grassle, 1992). By contrast, the role of specific waterborne compounds in evoking active larval settlement to the substrate has been only rarely explored in detailed investigations. With the exception of studies by Hadfield (e.g., Hadfield and Scheuer, 1985) and by Rittschof (e.g., Rittschof, 1985), we are unaware of other investigations considering the effects of waterborne larval settlement cues whose molecular structures are at least partially identified (but see Scheltema, 1961; Chia and Koss, 1988; Lambert and Todd, 1994; Stoner, 1994). Furthermore, no study previously has quantified behavioral responses by larvae to either waterborne or substrate-adsorbed chemical cues in real time. There is currently a paucity of information supporting active habitat selection by larvae in response to waterborne chemical cues under natural hydrodynamic conditions (Butman, 1989). It has been suggested that the induction of settlement by waterborne substances is especially unlikely for weakly swimming larvae. Horizontal swimming speeds in these larvae are typically much less than the horizontal flow speeds in benthic boundary layers, even for low energy environments at distances less than one larval body length away from the bed (Butman, 1986). However, circumstances exist where larval responses to chemical cues in the water column might effectively enhance settlement onto the substratum. A recent model indicates that hydrodynamics will control the number of larvae encountering a substratum, and indirectly control larval settlement, as long as larvae have a high probability of accepting the substratum once they come into contact with it (Grosset a!., 1992). By responding quickly to a waterborne cue, larvae might change their

SENSORY ECOLOGY OF OYSTER LARVAE

39

vertical distribution (making more larvae available to be swept near the substratum by fluid flow) and thus dramatically alter their patterns of settlement (Eckman et al., 1994). There is clearly need for further research to examine the role of waterborne cues in controlling larval settlement. Experimental evidence demonstrating that settlement cues are effective even in the water column would have important implications for larval delivery to benthic substrata. OYSTER LARVAE AS A MODEL FOR INVESTIGATING WATERBORNE SETTLEMENT CUES The American oyster, Crassostrea virginica, represents a sensitive tool for investigating and modeling the roles of waterborne cues in the induction of larval settlement. There are compelling reasons to believe that oyster larvae use such cues under natural conditions. Significantly, adult and juvenile oysters are sessile and form large aggregations (commonly extending 10 000 m 2 , or more) at high densities (15 000 individuals/m2) in low-energy estuarine environments (Bahr, 1981). These aggregations are thought to be maintained by the gregarious settlement of conspecific larvae, mediated in part by substances released from either juveniles and adults or bacteria on oyster shell surfaces (see review by Bonar et al., 1990). Waterborne compounds released by bacteria and postmetamorphic oysters will be diffusely distributed over large areas, rather than confined to isolated point sources. Furthermore, water flow velocities measured at 2-10 em above oyster reefs rarely exceed 7-10 cm/s (M. Palmer and D. Breitburg, unpubl. data) meaning oyster larvae drifting near the bottom will likely pass over recently cued substratum for several minutes. This exposure duration is long enough to invoke larval settlement response (see below). We asked if either passive sinking or active swimming by oyster larvae in response to a waterborne chemical cue might deliver these "particles" from the water column to the sea bed. A Rouse number is a dimensionless ratio calculated as the magnitude of forces driving particle transport vertically downward, as opposed to upward in the water column. Net downward transport is indicated by a Rouse number greater than 1. Using data of Wright et al. (1990) and Eckman et al. (1994), we calculated Rouse numbers to describe the net vertical transport of oyster larvae in the benthic boundary layer above an oyster reef at varying flow speeds. These data provided values for oyster larval gravitational fall velocity and swimming speed, as well as the ratio of shear velocity to free stream velocity (u.lurs) for water flowing above these reefs. Our calculations indicate Rouse numbers greater than 1 where free stream speeds are less than 5 cm/s. Such conditions are commonly met in benthic boundary layers above oyster reefs. Since Cole and Knight-Jones (1939) first described gregarious settlement and metamorphosis in oysters, there has been debate concerning the existence and sources of chemical morphogens and settlement-inducing compounds. One school of thought points to juvenile and adult oysters, whereas another postulates that bacterial biofilms on oyster shell surfaces are the sources of the inducer molecules (Bonar eta/., 1990; Paw-. lik, 1992). Significantly, the effects of waterborne compounds released by adult oysters and bacteria biofilms have never been separately assayed in the same study. Coon et al. (1985) proposed that the bacterium, Alteromonas colwelliana, produces L-dihydroxyphenylalanine (L-DOPA), a melanin precursor, and stimulates oyster larval settlement. However, subsequent research revealed that L-DOPA is converted to the neurotransmit-

40

RICHARD K. ZIMMER-FAUST ET AL.

ter dopamine inside the larva, and fails to stimulate settlement under natural conditions (Coon and Bonar, 1987; Fitt eta/., 1989). Ammonia (un-ionized ammonium) has also been proposed as a settlement cue for oyster larvae (Bonar et al., 1990; Coon et al., 1990; Fitt and Coon, 1992). Behavioral reaction is evoked, but only at very high concentrations ( [NH 3] > 100 p,M) where ammonia acts as a poison, even causing death to a wide range of aquatic organisms (e.g., Alabaster and Lloyd, 1980; Chen et al., 1990). Reaction to high concentrations of ammonia, therefore, might not reflect natural settlement activities. RECENT ADVANCES IN UNDERSTANDING OYSTER LARVAL SETTLEMENT RESPONSES TO WATERBORNE CHEMICAL CUES The small size of oyster larvae (250 to 350 J.Lm), and their sensitivity to handling, make them difficult subjects for experimental analysis of settlement behavior. We have overcome these difficulties by using a video image analysis system (Model VP 110 and Expert Vision software, Motion Analysis Corp.) interfaced with a SPARC computer work station (Sun Microsystems, Inc.) to non-invasively track the paths made by free-swimming individual larvae (Weissburg and Zimmer-Faust, 1991; Tamburri et al., 1992).We find this technology indispensable for the study oflarval behavior because (1) it becomes possible to use experimental chambers having large enough volumes to avoid impeding locomotory behavior, (2) locomotory paths can be tracked for many individual larvae simultaneously, and (3) data can be collected rapidly and in sufficient quantities while holding larvae at low densities (1 or 2/ml), thereby avoiding interactions between larvae and density-dependent effects. The proportion (k) of larvae contacting one another during an experimental trial of time interval, t, can be estimated using the expression: k

=

1 - e-f3Ct

(1)

where, (J

= 4nU?

(2)

and, Cis larval density; U is larval swimming speed; and, r is cross-sectional radius of the larval body (Koopman, 1980). These equations assume that larvae are spherical, uniform size, swim at constant speed, and move independently and at random relative to each other. Effects of experimental chamber walls on a larva can be safely ignored, when:

Y /L > 20/Re

(3)

where, Y is distance to the wall; L is larval characteristic length; and, Re is Reynolds number for a larva as it swims (Vogel, 1981). In all our investigations, we carefully maintained test conditions to minimize k (f. e., k < 0.10) and to eliminate wall effects on paths swum by larvae in the water column. Swimming behavior was assayed in infrared light given oyster larvae exhibit phototaxis in visible light (Smith and Chanley, 1975). Our results indicate chemically-mediated changes in swimming behavior which demonstrate a clear association between a dissolved chemical stimulus and behavioral response by oyster larvae. Larval responses differed markedly to seawater (control) and to dissolved compounds released, in separate treatments, both by adult conspecifics bathed in seawater (hereafter referred to as "oyster bath water" OBW) and by

SENSORY ECOLOGY OF OYSTER LARVAE

41

microalgal prey (!sochrysis ga/bana) (Figure 1.). Herein, we provide only a qualitative summary of our results because more rigorous quantitative analyses are either already published (Tamburri eta/., 1992; Zimmer-Faust and Tamburri, 1994; Turner eta/., 1994) or in preparation. In seawater, oyster larvae swam long looping paths at speeds of 1.0 mm/s in the horizontal plane (Figure lC and F). There was no tendency for the larvae to swim vertically either downward or upward in the water column. Except for turning more frequently, oyster larvae swam identically in algal exudates as in seawater (Figure 1B and E). By contrast, oyster larvae exposed to OBW swam vertically downward in the water column, slowed horizontal swimming speed, and increased rate of turning (especially near bottom) (Figure lA and D). The rate at which larvae contacted the bottom and attached with the foot was significantly elevated in response

Figure 1 Representative horizontal paths swum by larvae in response to either oyster bath water (4. and D.), algal exudates (!J. and E.), or artificial seawater medium (C. and F). Larvae were held at 2/ml in a Plexiglas microcosm (5 x 5 x 5 em; 125 ml capacity) and their movements recorded under infrared light. Eight replicate trials (3 min) were performed and over 50 paths analyzed for larvae swimming in each solution. A., B., and C. Paths swum by larvae at 2.5 em away from any wall. D., E., and F., Paths swum by larvae at ~ I mm above the bottom. Relative to artificial seawater, larvae decreased swimming speed and increased turning in response to oyster bath water. This effect was especially pronounced near the bottom of the microcosm. By comparison, larvae increased turning but did not change swimming speed in response to algal exudates. Oyster bath water was prepared by rinsing two oysters (7 g wet tissue mass) for 2 h in 8 liters of artificial seawater medium. Algal exudates were composed of water drawn from cultures oflog-phase cells (fsochrysis galbana) maintained at a density of 2 x 104 cells/mi. All solutions were membrane filtered (0.22 JLm pore size) prior to use. Scale bar is I mm.

RICHARD K. ZIMMER-FAUST ET AL.

42

to OBW, relative to both algal exudates and seawater (control) (Zimmer-Faust and Tamburri, in prep.). Chemoreceptive behavior of larvae indicating settlement was therefore exhibited only in response to OBW The downward swimming response by oyster larvae to OBW (as described above) was exhibited even in the absence of a chemical concentration gradient. Furthermore, all assays were performed in test chambers filled with a homogeneous solution of OBW This last finding eliminates a substantial theoretical argument against settlement induction by waterborne substances in the field: that stable chemical concentration gradients are not maintained above substrata in turbulent benthic boundary layer flows. Additional trials were conducted to determine ifbehavioral responses varied between oyster larvae drawn from different geographical regions and oyster populations. Two separate batches oflarvae, one spawned from adult brood stock collected in Chesapeake Bay, Virginia, and the other from Galveston Bay, Texas, were each assayed for their responses to OBWand seawater (control). We tested two different OBW solutions, each prepared using only adult oysters drawn from either Chesapeake Bay or Galveston Bay populations, respectively. Behavior oflarvae collected from each site was indistinguishable in response to the two different bath water solutions. The rates at which larvae settled on the bottom, as well as larval swimming kinematics, changed dramatically in response to both OBWs relative to seawater (control) (Zimmer-Faust and Tamburri, in prep.), and results closely paralleled findings described in preceding paragraphs. Oysters populations in Galveston Bay and Chesapeake Bay are different physiological strains (Groue and Lester, 1982; Buroker, 1983). Therefore, our findings suggest that larval chemoreceptive behavior mediating settlement is highly stereotyped and genetically conserved. To determine the sources of settlement inducing molecules, we developed a method for removing bacteria from external oyster shell surfaces by applying mechanical agitation and chemical oxidation (Tamburri et al., 1992). This technique removed> 99% of the viable bacteria in the biofilm from shell external surfaces without disrupting the normal production of metabolites by oysters (measured as the weight-specific production of ammonium and dissolved organic carbon). We eliminated only external biofilms because the total number of bacteria in oyster tissues, on inner shell surfaces, and released into bath solutions during incubations of oysters were < 0.1% of the total on the external shell surfaces (Tamburri eta/., 1992). Biofilms attached to aged shell material without the living oyster served as a source of bacterial metabolites (prepared by bathing the material as above). Significantly, we found that waterborne substances released by both oysters and biofilms stimulated settlement (Tamburri et al., 1992). In both cases,> 50% of the larvae assayed actively swam downward in the water column and attached to the bottom within 3 min of initial exposure. We conducted experiments to estimate the exposure time required for OBW to induce oyster larval settlement response. Larvae were tested only when they swam within 2 mm of the bottom. By following this procedure, video imaging was reduced from a 3- dimensional to a 2-dimensional problem greatly simplifYing quantitation of larval behavior. An automated micro-injection system (Picospritzer II, General Valve Corp.) was employed to emit tiny "patches" of chemical stimuli near individual free-swimming larvae. Either seawater (control), seawater with dilute fluorescein dye (control), or OBW with fluorescein dye (test) was injected. Each injection created a single patch between 4 and 5 mm diameter, through which a larva swam. A minimum of 15 trials was performed for each solution, with dilute 3 gil) providing a visual marker enabling us to precisely determine when fluorescein dye a larva entered the injected p!j.tch.Video recordings were limited to the last 5 s before, and to

oo-

-

SENSORY ECOLOGY OF OYSTER LARVAE

43

3.0

"C

=2.5 2.0 "> -... 1.5 "i

80

A.

0

CD "C

CD CD

Cl)

d

40

1.0

20

0.5

z 0.0

"C

"i

ASW

1000

Crude

3.0

=0 2.5

-... CD "C

">

Cl)

CD

CD

Cl)

d

80

B.

.--·t

2.0 1.5 1.0

=

0.5

z 0.0

~

ASW

10·10

10-9

0 0

n

~ 0 (I)

=

CD

-a CD :I

40

e ···· GGR

*

3

n·

iD 60 3

0-0BW

~

--... CD :I

0 (I) 3

Molecular weight (Da)

-

(I)

60 CD 3

Cl)

;:

'#.

Field

10-8

20

3

n· n

0

(I)

0 3

Figure 2 Larval settlement in response to chemical solutions. Larvae were held at 2 /ml in a Plexiglas microcosm (3 x 3 x 4 em; 30 ml capacity) and filmed (7.05 x 6.90 mm field) as they swam within I mm (:::; 3 body lengths) of, attached to, then crawled on the bottom. Settlement, defined as the number oflarvae attached to the bottom via their foot after 3 min exposure, was assessed for each trial. The percentage of all larvae tested that settled per trial was calculated by assuming the video field provided a representative sample of the whole microcosm. Eight replicate trials for each solution were performed at each concentration. A. Mean(± SEM) settlement in response to "crude" (unfractionated) oyster bath water (OBW), molecular weight fractions of OBW, and artificial seawater medium (ASW). Significantly more larvae settled in response to the 500--1000 Da fraction than to control ASW (One-way ANOVA followed byTukey's HSD Test: P < 0.001; all data were arcsine transformed prior to statistical analysis). By contrast, larvae settled no more readily in response to the < 500 Da or > 1000 Da fractions than to ASW (!' > 0.20, both comparisons). Larval settlement in response to the 500-1000 Da fraction and to "crude" OBW was not significantly different(!'> 0.50). B. Dose-response curves for settlement in response to the synthetic tri-peptide, glycyl-glycyl-L-arginine (GGR) and the active fraction (500--1000 Da) of OBW The regression lines do not significantly differ (ANCOVA: P > 0.20). Settlement responses to both artificial seawater medium (ASW) and seawater collected above oyster reefs in field habitats (Field) are also plotted. Concentrations are expressed in molar units of peptide arginine.

44

RICHARD K. ZIMMER-FAUST ET AL.

the first 5 s after the patch was initially contacted by larvae. Because horizontal swimming speed decreases dramatically when oyster larvae are stimulated by OBW (see above), this parameter was used to measure behavioral response. The mean speed (0.65 mm/s ± 0.05 mm/s SEM) at which larvae swam before entering a patch did not differ significantly between any of the test or control groups. Following injection, larvae swam at 0.66 mm/s (± 0.07 mm/s SEM) in response to seawater whereas they swam at 0.61 mm/s (± 0.04 mm/s SEM) in response to seawater plus fluorescein, a non-significant difference. This result shows that the addition of fluorescein had little effect on the behavior oflarvae. By contrast, larvae significantly reduced swimming speed to 0.44 mm/s (± 0.05 mm/s SEM) in response to injection of OBW Clearly, oyster larvae behaviorally respond to waterborne settlement cue within seconds after initial contact. There are now several lines of experimental evidence all pointing to the singular identity of the waterborne cues eliciting oyster larval settlement (Zimmer-Faust and Tamburri, 1994). Molecular weight fractionations of seawater used to bathe adult oysters were bioassayed and indicated the presence of waterborne cues between 500 and 1000 daltons (Figure 2A). The inducers were degraded by proteases, but not by carbohydrases or by lipase. Of several proteases we applied, only those cleaving basic amino acids (lysine and arginine) from the C-terminus, and arginase (an enzyme condensing arginine at the C-terminus to ornithine) eliminated settlement-inducing activity. Trypsin hydrolysates of casein (which produce only peptides containing either arginine or lysine at the C-terminus) were found to be significantly more effective in causing larval settlement than products of either acid or pronase hydrolysis of this protein. A synthetic tri-peptide having arginine at the C-terminus, glycyl-glycyl-L-arginine (hereafter referred to as GGR), evoked settlement at a concentration as low as 10-IO M. Dose- response curves for GGR and for the active fraction (500-1000 Da) of oyster bath water were essentially identical (Figure 2B). Finally, tests of 21 free amino acids identified only lysine and arginine as settlement cues. Larval settlers were far more sensitive to arginine than lysine, but significantly less sensitive to arginine than to peptides with arginine at the C-terminus. These combined results are all consistent in identifYing LMW pep tides with arginine at the C-terminus as the natural, water-soluble cues inducing oyster larval settlement. Besides oysters, peptide cues mediating larval settlement or metamorphosis have now been demonstrated for abalones (Morse and Morse, 1984; Morse eta/., 1984), sand dollars (Burke, 1984), and barnacles (Rittschof, 1985; Tegtmeyer and Rittschof, 1989). Agreement between findings for oyster and barnacle larvae is particularly striking, because barnacle larvae (Balanus amphitrite) similarly settle in response to GGR (Tegtmeyer and Rittschof, 1989). In fact, arginine-containing peptides commonly act as signal molecules and cause a diversity of behavioral responses in a wide range of marine animals (see review ofRittschof, 1993). For example, arginine-peptides evoke both shell search by hermit crabs (Rittschof eta/., 1990) and abdomen pumping and larval release by ovigerous female mud crabs (Forward et al., 1987). There is now correlative evidence suggesting that LMW arginine-peptides are released into the natural environment by juvenile and adult oysters, and induce larval settlement. We collected seawater from five sites, each elevated 0.5 m above oyster reefs in the North Inlet estuary, South Carolina (see methods in Zimmer-Faust eta/., 1995). These water samples were collected in June, 1994, during a two-week period of oyster larval recruitment into North Inlet (P. Kenny, unpubl. data). At the same time, seawater was collected from five off-reef sites located at the mouth of this same estuary. Seawater col-

SENSORY ECOLOGY OF OYSTER LARVAE

45

lected from each site was immediately filtered (to 0.22 J-Lm), maintained on dry ice during transport to the laboratory, then held at -76°C until used in laboratory assays oflarval settlement, as previously described. The magnitude oflarval settlement in response to seawater collected at each on-reef site did not significantly differ. However, settlement in response to on-reef seawater was elevated significantly higher than settlement in response to off-reef seawater. Using pooled data, larval settlement in response to on-reef seawater was plotted as a function of LMW arginine-peptide concentration (Figure 2B). This relationship is in excellent agreement with the dose-response curve determined for OBWThe LMWarginine-peptide levels in off-reef seawater were lower than in on-reef seawater, and below the analytical detection limits established by our methods for concentrating, desalting, and chromatographing such molecules. All of our research (cited above) was performed by assaying larval settlement responses in still water. Since oysters are weak swimmers, we conducted experiments to determine if the LMW arginine-peptides could mediate settlement in flowing seawater. These trials were conducted in a raceway flume (5 m long) providing larvae with a single pass over cued substratum (see Turner et al., 1994). Trials were performed at mean flow speeds between 2 and 6 cm/s and shear velocities(~ 0.25 cm/s) typical of estuarine flows. Small target wells (7 em diameter) filled with clean oyster shell were perfused with either a GGR peptide (at w- 7 M) or control (seawater) solution. In a single pass, oyster larvae consistently settled significantly more onto the GGR wells, responding by swimming downward and attaching to substrate. We are currently developing methods for purifYing LMW arginine-pep tides from seawater through cation exchange and reverse-phase HPLC (Melaro, 1994; Melaro eta!., in prep.). These procedures should allow us to continue working productively toward fully identifYing structures mediating oyster larval settlement. To date, our experiments demonstrating peptide induction of larval settlement have been limited to laboratory trials. Field tests are needed to determine whether peptide release in natural estuarine environments increases larval settlement, leading to higher recruitment of juveniles. Larval collectors with controlled peptide release are currently being constructed by adapting the design of the sediment chemistry alteration tubes of Hentschel and Jumars (1994). Such devices will permit the time-release of test peptide at known concentrations into the environment. WATERBORNE VERSUS SUBSTRATE-ADSORBED CHEMICALS AS CUES MEDIATING OYSTER LARVAL SETTLEMENT AND METAMORPHOSIS: CURRENT AND FUTURE DIRECTIONS Compounds that induce settlement and metamorphosis are usually identical for larvae of a given species (Morse, 1990; Pawlik, 1992). Although there is some evidence to suggest these inducers might differ for oyster larvae (e.g., Bonar eta/., 1990), we recently found GGR and the 500-1000 Da fraction of OBW stimulating both metamorphosis and settlement (Figure 3). We report these results with some caution because our experiment was conducted using only a single batch of oyster larvae. Metamorphosis could only be induced in the presence of preferred substrate and gently flowing water in addition to the appropriate chemical cue. A key agent regulating oyster larval metamorphosis may be adsorption of waterborne, LMW arginine-peptides to benthic substrata. Microbial biofilms attached to

RICHARD K. ZIMMER-FAUST ET AL.

46

tn

·u;

50

0 40 .J:

c.

A.

... 30

0

E as 20 '&;

==

eft.

10 0 ASW

1000

Crude

Molecular weight (Da)

tn

·u;

50

0 40 .J: c. ... 30 0

E as '&;

:! eft.

B.

20 10 0 ASW

10-6

10-B

10·10

Crude

Gly- gly- L- arg (M) Figure 3 Larval metamorphosis during 24 h exposure to chemical solutions. Larvae were held at I /ml in Plexiglas microcosms (30 ml capacity) with sterilized oyster shell (3 cm 2) provided in each chamber as substrate. The microcosms were placed on an orbit shaker and rotated at low speed (I revolution/s) to induce gentle fluid motion. Metamorphosis was defined as larval cementation to the substratum, resorbed velum, and significant new shell growth. Eight replicate trials of each solution were performed at each concentration. A. Mean(± SEM) %-metamorphosis in response to "crude" (unfractionated) oyster bath water (OBW), molecular weight fractions of OBW, and artificial seawater medium (ASW). Significantly more larvae metamorphosed in response to the 500-1000 Da fraction than to ASW (One-way ANOVA followed byTukey's HSD Test: P < 0.001; all data were arcsine transformed prior to statistical analysis). By contrast, larvae metamorphosed no more frequently in response to the < 500 Da or > 1000 Da fraction than to control ASW (P > 0.20, both comparisons). Larval metamorphosis in response to the 500-1000 Da fraction and to "crude" OBW was not significantly different (P > 0.50). B. Mean (± SEM) %-metamorphosis in response to the synthetic tri-peptide, glycyl-glycyi-L-arginine (GGR), "crude" (unfractionated) OBW, and ASW Metamorphosis induction by Io-s M GGR and by "crude" OBW do not significantly differ (P > 0.50).

SENSORY ECOLOGY OF OYSTER LARVAE

47

substrata consist of bacteria embedded within a matrix ofhigh-molecular-weight secretions called exopolymers (see review of Decho, 1990). Exopolymers act as an adsorptive sponge to sequester peptides and other low-molecular-weight molecules from the surrounding seawater environment for later assimilation and use in catabolism by bacteria. For example, acidic polysaccharides are components of exopolymers possessing negatively-charged functional groups, most notably carboxyls. These polymers are common constituents of bacterial biofilms, and they exhibit high binding affinities for many positively-charged organic compounds which should include the LMW arginine-peptides cueing oyster settlement (Sutherland, 1977). Unfortunately, we do not yet know if arginine-peptides are adsorbed to microbial biofilms, and if induction of oyster larval metamorphosis is enhanced by pep tides bound to exopolymers. SUMMARY AND PERSPECTIVE Habitat colonization by planktonic larvae is a major factor regulating population dynamics of benthic invertebrates. Current understanding of such populations recognizes the importance oflarval supply, and assumes that larval delivery to the substratum is a passive, hydrodynamically-driven process. Chemical cues mediating larval settlement are commonly believed to be effective only after larvae reach the substratum. In contrast, we propose that waterborne chemical cues near the seabed may be perceived by larvae and induce behavior which significantly influences larval delivery to benthic substrata. We have established that planktonic oyster larvae (Crassostrea virginica) actively swim downward upon contacting waterborne peptides released by adult conspecifics. Swimming downward in response to peptide cues substantially increases larval settlement on substratum, as demonstrated in both still water and flume flow experiments. Chemical adsorption and desorption properties of microbial exopolymers might be important in determining larval habitat colonization. Compounds serving as waterborne settlement cues may best function as morphogens once bound to bacterial exopolymers coating benthic substrata. Continued investigation of waterborne cues is clearly warranted and new discoveries may provide novel insights on processes structuring marine communities through larval delivery to benthic substrata.

Acknowledgements The authors thank E. W. Melaro, N. D. Pentcheff, E. J. Turner, D. W. Schar, Y Ishikawa, M. W. Luckenbach, and M. L. Tamplin, for their valuable collaborations. S.V. Viscido, J. E. Commins, C. C. Gee, and K. A. Browne made comments on earlier drafts, which greatly improved this manuscript. We are especially indebted to P. H. Lenz, D. K. Hartline, and J. E. Purcell, for organizing a splendid symposium and encouraging us to collect our thoughts in writing this short synthesis. Our research was supported by awards from the National Science Foundation (OCE 94-16749), South Carolina Sea Grant Consortium (P-M-2F, P-M-2K, R-92-888), and the Carolina Venture Fund.

References Alabaster, J. S., and R. Lloyd (1980) ffilter Quality Criteria for Freshwater Fish. Butterworths, London. Bahr, L. M. (1981) The ecology of intertidal oyster reefs of the south Atlantic coast: a community profile. U. S. Fish and Wildlife Service FWS/OBS-81/15.

48

RICHARD K. ZIMMER-FAUST ET AL

Bonar, D. B., S. L. Coon, M. Walch, R. M. Weiner, and W Fitt (1990) Control of oyster settlement and metamorphosis by endogenous and exogenous chemical cues. Bull. Mar. Sci., 46, 484-498. Burke, R. D. (1984) Pheromonal control of metamorphosis in the Pacific sand dollar, Dendraster excentricus. Science, 225,442-443. Burke, R. D. (1986) Pheromones and gregarious settlement of marine invertebrate larvae. Bull. Mar. Sci., 39, 323-331. Buroker, N. E. (1983) Population genetics of the American oyster, Crassostrea virginica, along the Atlantic coast and the Gulf of Mexico. Mar. Bioi., 75, 99-112. Butman, C. A. (1986) Larval settlement of soft-sediment invertebrates: some predictions based on analysis of near-bottom velocity profiles. In Marine Interfaces £co hydrodynamics, J. C. J. Nihoul (ed.), Elsevier Science Publications, New York. pp. 487-513. Butman, C. A. (1987) Larval settlement of soft-sediment invertebrates: the spatial scales of pattern explained by habitat selection and the emerging role of hydrodynamic processes. Oceanogr. Mar. Bioi. Annu. Rev., 25, 113-165. Butman, C. A. (1989) Sediment trap experiments on the importance ofhydrodynamical processes in distributing settling invertebrate larvae in near-bottom waters. J Exp. Mar. Bioi. Ecol., 134, 37-88. Butman, C. A., and J. P. Grassle (1992) Active habitat selection by Capitella sp. I larvae: two-choice experiments in still water and flume flows. J Mar. Res., 50, 669-715. Butman , C. A., J. P. Grassle, and C. M. Webb (1988) Substrate choice made by marine larvae settling in still water and in a flume flow. Nature, 333, 771-773. Chen, J. C., P. C. Liu, and S. C. Lei (1990) Toxicities of ammonia and nitrite to Penaeus monodon adolescents. Aquaculture, 89, 127-137. Chia, F. -S., and R. Koss (1988) Induction of settlement and metamorphosis of the veliger larvae of the nudibranch, Onchidoris bilamellata. Int. J Inverte. Reprod. Develop., 14, 53-70. Cole, H. A., and E. W Knight-Jones (1939) Some observations and experiments in the settling behavior of larvae of Ostrea edulis. J Cons. Cons. Int. Explor. Mer., 14, 86-105. Connell, J. H. (1961) The influence of intraspecific competition and other factors on the distribution of the barnacle Chthalamus stellatus. Ecology, 42, 710-723. Connell, J. H. (1985) The consequences of variation in initial settlement vs. post-settlement mortality in rocky intertidal communities. J Exp. Mar. Bioi. Ecol., 93, 11-45. Coon, S. L., and D. B. Bonar (1987) The role of DOPA and dopamine in oyster settlement behavior. Am. Zoo/. 27, 128A (abstract). Coon, S. L., D. B. Bonar, and R. M. Weiner (1985) Induction of settlement and metamorphosis of the Pacific oyster Crassostreagigas (Thunberg) by L-DOPA and catecholamines. J Exp. Mar. Bioi. Ecol., 94, 211-221. Coon, S. L., M. Walch,W K. Fitt, D. B. Bonar, and R. M. Weiner (1990) Ammonia induces settlement behavior in oyster larvae. Bioi. Bull., 179, 297-303. Crisp, D. J. (1974) Factors influencing settlement of marine invertebrate larvae. In Chemoreception in Marine Organisms, P. T. Grant and A.M. Mackie (eds.), Academic Press, London. pp.l77-265. Crisp, D. J. (1984) Overview of research on marine invertebrate larvae, 1940-1980. In Marine Biodeterioration: An Interdisciplinary Study, J. D. Costlow and R. C. Tipper (eds.), Naval Institute Press, Annapolis. pp. 103-126. Cuomo, M. C. (1985) Sulphide as a larval settlement cue for Capitella sp. I. Biogeochemistry, 1, 169-181. Dayton, P. K. (1971) Competition, disturbance, and community organization: the provision and subsequent utilization of space in a rocky intertidal community. Ecol. Mono gr., 41, 351-389. Decho, A. W (1990) Microbial exopolyrner secretions in ocean environments: their role(s) in food webs and marine processes. Oceanogr. Mar. Bioi. Annu. Rev., 28, 73-153. Eckman, J. E. (1983) Hydrodynamic processes affecting benthic recruitment. Limnol. Oceanogr., 28, 241-257. Eckman, J. E. (1990) A model of passive settlement by planktonic larvae onto bottoms of differing roughness. Limnol. Oceanogr., 35, 887-901. Eckman, J. E., F. E. Werner, and T. F. Gross (1994) Modeling some effects of behavior on larval settlement in a turbulent boundary layer. Deep Sea Res., 41, 185-208. Fitt, W K., and S. L. Coon (1992) Evidence for ammonia as a natural cue for recruitment of oyster larvae to oyster beds in a Georgia salt marsh. Bioi. Bull., 182,401-408. Fitt,W K., M.P. Labare,W C. Fuqua, M.Walch, S. L. Coon, D. B. Bonar, R. R. Colwell, and R. M.Weiner (1989) Factors influencing bacterial production of inducers of settlement behavior of larvae of the oyster, Crassostrea gigas. Microb. Ecol., 17, 287-298. Forward, R. B., Jr., D. Rittschof, and M. C. DeVries (1987) Peptide pheromones synchronize crustacean egg hatching and larval release. Chern. Senses, 12, 491-498. Gaines, S., S. Brown, and J. Roughgarden (1985) Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnacle, Balanus glandula. Oecologia, 61, 267-272.

SENSORY ECOLOGY OF OYSTER LARVAE

49

Grosberg, R. (1982) Intertidal zonation of barnacles: the influence of planktonic zonation oflarvae on the vertical distribution of adults. Ecology, 63, 894-899. Gross, T. F., F. E. Werner, and J. E. Eckman (1992) Numerical modeling oflarval settlement in turbulent bottom boundary layers. J. Mar. Res., 50, 611-642. Groue, K. J., and L. T. Lester (1982) A morphological and genetic analysis of geographical variation among oysters in the Gulf of Mexico. Ji!liger, 24,331-335. Hadfield, M. G., and D. Scheuer (1985) Evidence for a soluble metamorphic inducer in Phestilla sibogae. Bull Mar. Sci., 37, 556-566. Hentschel, B. T., and P. A. Jumars (1994) In situ chemical inhibition of benthic diatom growth affects recruitment of competing, permanent and temporary meiofauna. Limnol. Oceanogr., 39, 816-832. Jensen, R. A., and D. E. Morse (1990) Chemically induced metamorphosis of polychaete larvae in both the laboratory and the ocean environment. J. Chern. Ecol., 16, 911-930. Johnson, L. E., and R. R. Strathmann (1989) Settling barnacles avoid substrata previously occupied by a mobile predator. J. Exp. Mar. Bioi. Ecol., 128,87-103. Keough, M. J. (1984) Kin-recognition and the spatial distribution of larvae of the bryozoan Bugula neritina. Evolution, 38, 142-147. Kirchman, D., S. Graham, S. Reish, and R. Mitchell (1982) Bacteria induce settlement and metamorphosis of Janua (Dexiospiral) brasiliensis (Polychaeta: Spirobidae). J. Exp. Mar. Bioi. Ecol., 56, 152-163. Koopman, B. 0. (1980) Search and Screening. Pergamon Press, New York. LeTourneux, F., and E. Bourget (1988) Importance of physical and biological settlement cues used at different spatial scales by the larvae of Semibalanus balanoides. Mar. Bioi., 97, 57-66. Maki, J. S., D. Rittschof, M.-0. Samuelsson, U. Szewzwk, A. B. Yule, S. Kjelleberg, J. D. Costlow, and R. Mitchell (1990) Effects of marine bacteria and their exopolymers on the attachment of barnacle cypris larvae. Bull. Mar. Sci., 46,499-511. Melaro, E. W. (1994) Methods for concentrating, de-salting, and partially purifYing basic peptides from seawater and oyster mantle fluid. Masters Thesis, University of South Carolina, Columbia. Menge, B. A., and J.P. Sutherland (1976) Species diversity gradients: synthesis of the roles of predation, competition, and temporary heterogeneity. Am. Nat., 110, 351-369. Morse, A. N. C., and D. E. Morse (1984) Recruitment and metamorphosis of Haliotis larvae induced by molecules uniquely available at the surfaces of crustose red algae. J. Exp. Mar. Bioi. Ecol., 75, 191-215. Morse, D. E. (1990) Recent progress in larval settlement and metamorphosis: closing the gaps between molecular biology and ecology. Bull Mar. Sci., 46, 465-483. Morse, D. E., N. Hooker, H. Duncan, and L. Jensen (1979) 1-Aminobutyric acid, a neurotransmitter, induces planktonic abalone larvae to settle and begin metamorphosis. Science, 204,407-410. Morse, D. E., and A. N.C. Morse (1991) Enzymatic characterization of the morphogen recognized by Agaricia humilis (scleractinian coral) larvae. Bioi. Bull., 181, 104-122. Paine, R. T. (1974) Intertidal community structure: experimental studies on the relationship between a dominant competitor and its principal predator. Oecologia, 15, 93-120. Pawlik, J. R. (1992) Chemical ecology of the settlement of benthic marine invertebrates. Oceanogr. Mar. Bioi. Annu. Rev., 30, 273-335. Pawlik, J. R., and C. A. Butman (1993) Settlement of a marine tube worn as a function of current velocity: interacting effects of hydrodynamics and behavior. Limnol. Oceanogr., 38, 1730-1740. Raimondi, P. T. (1988) Settlement cues and determination of the vertical limit of an intertidal barnacle. Ecology,

69, 400-407.

Rittschof, D. (1985) Oyster drills and the frontiers of chemical ecology: unsettling ideas. Am. Malacol. Bull. Spec. Ed., 1, 111-116. Rittschof, D. (1993) Body odors and neutral-basic peptide mimics: a review of responses by marine organisms. Am. Zoo!., 33,487-493. Rittschof, D., C. M. Kratt, and A. S. Clare. (1990). Gastropod predation sites: the roles of predator and prey in chemical attraction of the hermit crab Clibanarius vittatus (Bose). J. Mar. Bioi. Assoc. UK., 70, 583-596. Rodriguez, S. R., F. P. Odeja, and N.C. Inestrosa (1993) Settlement ofbenthic marine invertebrates. Mar. Ecol. Prog. Ser., 97, 193-207. Roughgarden, J., S. Gaines, and H. Possingham (1988) Recruitment dynamics in complex life-cycles. Science,

241, 1460--1466.

Scheltema, R. S. (1961) Metamorphosis of the veliger larvae of Nassarius obsoletus (Gastropoda) in response to bottom sediment. Bioi. Bull., 120,92-109. Smith, W. L., and M. H. Chanley (1975) Provisions of light. In Culture of Marine Invertebrate Animals, G. G. Simpson and W. S. Beck (eds.), Plenum Press, New York. pp. 266-267. Strathmann, R. R., E. S. Branscomb, and K.Vedder (1981) Fatal errors in set as a cost of dispersal and the influence of intertidal flora on set of barnacles. Oecologia, 48, 13-18.

50

RICHARD K. ZIMMER-FAUST ET AL.

Sutherland, I. W (1977) Bacterial exopolysaccharides - their nature and production. In Surface Carbohydrates ofthe Procaryote Cell, I.W Sutherland (ed.), Academic Press, New York. pp. 209-245. Tamburri, M. N., R. K. Zimmer-Faust, and M. L. Tamplin (1992) Natural sources and properties of chemical inducers mediating settlement of oyster larvae: a re-examination. Bioi. Bull., 183, 327- 338. Tegtmeyer, K., and D. Rittschof (1989) Synthetic peptide analogs to barnacle settlement pheromone. Peptides, 9, 1403-1406. Turner, E. J., R. K. Zimmer-Faust, M. A. Palmer, M. Luckenbach, and N. D. Pentcheff (1994) Settlement of oyster (Crassostrea virginica) larvae: effects of water flow and a water-soluble chemical cue. Limnol. Oceanogr., 39, 1579-1593. Underwood, A. J., and P. G. Fairweather (1989) Supply-side ecology and benthic marine assemblages. Trends Ecol. Evol., 4, 16-20. Vogel, S. (1981) Life in Moving Fluids. Princeton University Press, Princeton. Weissburg, M. J., and R. K. Zimmer-Faust (1991) Ontogeny versus phylogeny in determining patterns of chemoreception: initial studies with fiddler crabs. Bioi. Bull., 181, 205-215. Woodin, S. A. (1986) Settlement of in fauna: larval choice? Bull. Mar. Sci., 39,401-407. Woodin, S. A. (1987) Recruitment of infauna- positive or negative cues? Am. Zoo!., 31, 797-807. Wright, L. D., R. A. Gannisch, and R. J. Byrne (1990) Hydraulic roughness and mobility of three oyster-bed artificial substrate materials. J Coast. Res., 6, 867-878. Zimmer-Faust, R. K., P. B. O'Neill, and D. W Schar (1995) Effects of predator activity on odor-mediated prey search. Bioi. Bull., 189: in press. Zimmer-Faust, R. K., and M. N. Tamburri (1994) Chemical identity and ecological implications of a waterborne, larval settlement cue. Limnol. Oceanogr., 39, 1075-1087.

THE TELEOST OCTAVOLATERALIS SYSTEM: STRUCTURE AND FUNCTION ARTHUR N. POPPER Department of Zoology, University ofMaryland, College Park, MD 20742 email: AP17 @umail. umd. edu This paper considers the detection ofvibrational signals (including sound) by the two components of the octavolateralis system, the ear and mechanosensory lateral line. Together, these systems provide fishes with a good deal of information about their surrounding environment, and enable fishes to detect both predators and prey. While the mechanisms by which fishes and zooplankton produce and detect signals may differ, it is clear that the physical principles underlying the signals themselves ate identical, no matter whether we are dealing with fish or zooplankton. Thus, an understanding of signal production and detection mechanisms by fishes can be of significant help in understanding how similar systems would function in zooplankton.

INTRODUCTION Why be concerned with the sensory biology of adult fishes when the thrust of this symposium is on the sensory ecology and physiology of zooplankton, only a small portion of which include ichthyoplankton? There are several answers to this question. The first is that there are predator-prey interactions between the two groups, and if fishes can detect signals produced by zooplankton, this improves the survival of the predator and affects the prey. On the other hand, if zooplankton evolve signals not detectable by fishes, or if they themselves can detect potential predators, this improves their chances of survival. Thus, in understanding the ecology and behavior of zooplankton, it is necessary to have some appreciation of the sensory biology of major predators. A second answer to the question is that in order to understand the sensory biology of zooplankton it is necessary to have some understanding of the physical aspects of the signals that are detected and emitted by these organisms, as well as of the physical aspects of the environment and the constraints imposed on signals by the environment. While the mechanisms by which fishes and zooplankton produce and detect signals may differ, it is clear that the physical principles underlying the signals themselves are identical, no matter whether we are dealing with fish, zooplankton, or marine mammals (or, for that matter, with terrestrial organisms). The thrust of this paper will be to consider the octavolateralis system - the ear and lateral line of fishes. In doing so, I will first consider the relationship between the ear and lateral line and basic principles of underwater acoustics that affect both fish and zooplankton. This is followed by brief descriptions of the way in which fish use sound for communication and a more extensive discussion of how fishes detect sounds using the ear and the lateral line. Finally, the paper will consider what kinds of sounds fishes detect. Though the paper will discuss the ear and the lateral line, some emphasis will be placed on the auditory system since other papers in this symposium (e.g., Bleckmann et a/., Volume 26) will provide additional insight into the function and behavior associated with the lateral line system. 51

52

ARTHUR N. POPPER

In addition, this paper will not deal with zooplankton per se since this is the subject of many other papers in the symposium. However, it should be noted that the basic physical acoustic principles that affect the evolution and function of the octavolateralis system also had to affect the evolution and function of sensory systems in zooplankton. Thus, any consideration of how zooplankton detect or produce acoustic or hydroacoustic signals will certainly benefit from studies on fishes. Since this paper is intended as an overview, no attempt has been made to be comprehensive in citing the extensive body of literature on fish bioacoustics or the octavolateralis system. The body of literature on sound production and the use of sounds by fish in behavior has been reviewed in several papers including those by Tavolga (1971), Demski et a/., (1973), Fine et al, (1977) and Myrberg (1981). More extensive reviews on fish hearing can be found in papers by Schellart and Popper (1992), Popper and Fay (1993), and Popper and Platt (1993). Again, although the lateral line is discussed in several of the papers mentioned above, more extensive discussions of the structure and function of the lateral line can be found in a volume by Coombs eta/. (1989) and in papers by Coombs eta/. (1992), Bleckmann (1993) and Montgomery eta/. (in press). MULTIPLE SENSORY SYSTEMS It is important to understand that detection of vibrational signals (which includes

sounds) by fishes really involves two sensory systems, the ear and the lateral line. (Note that references to the lateral line in this paper refer to the mechanoreceptive lateral line, as opposed to the electrosensory lateral line system that is found in many fish species,) Together, these are often referred to as the octavolateralis system. A single term is used since there are a number of features in common between the two systems (see Popper eta/., 1992 for a review of the history of this term and the phylogenetic relationship between the ear and lateral line). Both systems use similar sensory hair cells as the transducing structure for signal detection. In addition, both respond to similar types of signals, and it is possible that the input from the ear and lateral line may overlap at some higher levels in the central nervous system (CNS), although they do not overlap at least up to the level of the midbrain (ScheJlart eta/., 1984, 1992; McCormick, 1992). The two components of the octavolateralis system have considerable functional overlap. The inner ear detects sounds from well below 50 Hz to, in some species, over 2,000 Hz, and it also responds to positional information and motion of the body (vestibular senses) (e.g., Platt, 1983; Popper and Platt, 1993). The lateral line responds to the net motion between the fish's body and the surrounding water (e.g., spatial nonuniformities of the flow field- see Bleckmann, 1993 and Montgomery eta/., in press), including stimuli produced by swimming fish and other organisms, from several hundred Hz down to close to DC (e.g., Coombs eta/., 1989, 1992; Bleckmann eta/., 199lb). An important difference between the two systems is the distance from the fish over which they detect signals. The lateral line detects signals close to a fish (e.g., within one or two body lengths) while the ear detects signals to considerable distances from the fish (e.g., Kalmijn, 1988, 1989). Many signals can be thought of in terms of stimulating both the ear and the lateral line. In fact, from the perspective of modifYing the behavior of fish with sound, which sensory system is involved with the response is probably irrelevant.

THE TELEOST OCTAVOLATERALIS SYSTEM

53

UNDERWATER SOUND It is important to consider underwater acoustics briefly before any further discussion of

the octavolateralis system. Much of the literature often refers to the acoustic near field and acoustic far field (see van Bergeijk, 1967 for one of the classic discussions; also Kalmijn, 1988, 1989; Rogers and Cox, 1988). In the near field, particle displacement (which is vectorial and thus has a directional component) includes a hydrodynamic flow of the water as a consequence of the motion of the source. The far field particle displacement is only that molecular motion accompanying the far field pressure component of the signal. The measure often used for the transition between near- and far field is a function of >..!21f (>.. = wavelength) for a monopole source (pulsating sphere), which is approximately l/6th of a wavelength of the sound frequency, but its precise distance depends upon the detailed structure of the sound source (e.g., Harris and van Bergeijk, 1962; Siler, 1969). Sound sources differ in the way they radiate energy. Consequently, the attenuation rates of the pressure and particle displacement components of the signals differ, depending upon the source and the distance from the source. Within the near field, signals generated with a complex source such as a "dipole" (a vibrating sphere approximating a swim bladder producing sound), have the most rapid attenuation of signals as they leave the source, while signals from a monopole have less rapid attenuation (see Harris and van Bergeijk, 1962; van Bergeijk, 1967; Kalmijn, 1988, 1989). Moreover, for both monopole and dipole, near field velocity and displacement attenuate more rapidly than does near field pressure (l!r 3 vs l!r 2 for a dipole and 1/r2 vs 1/r for a monopole). However, in the acoustic far field particle displacements (velocities) and sound pressure from all sources attenuate as 1/r. By way of example, for a 100 Hz signal produced by a monopole source, the wavelength in water is 15 m (the speed of sound in water is approximately 1500 m/sec). The transition between the near- and far fields (.X/27r) is approximately 2.4 meters from the sound source. Even if we assume that a fish is quite sensitive to the particle displacement, it is unlikely that the particle displacement component of a 100-Hz signal would be detectable beyond about 2 m from the source unless the fish were exceptionally sensitive. The distance will be affected by sound intensity as well as by the displacement sensitivity of the fish. It must be emphasized that near- and far field components do not suddenly change at the point of >.121r and that some fishes may be sensitive to hydrodynamic motions well into the far field, while others may be insensitive even in the outer part of the near field. In fact, pressure and particle displacement are present in both the near- and far fields, but particle displacement predominates in the near field and pressure predominates in the far field. The rate of attenuation of pressure and particle displacement differ, with particle displacement attenuating much more rapidly in the near field than does pressure. This provides a very substantial particle displacement gradient with increasing distance from the source in the near field. The biological significance of these differences in the near and far field effect depends upon how readily a particular species can detect the particle displacement component of a signal with the ear vs. the lateral line. If the fish ear is very sensitive to particle displacement, then it may detect the signal for a considerable distance from the source. If the fish ear is not very sensitive to particle displacement, it will not detect the signal even as far as the transition point. The lateral line, in contrast, detects particle displacement gradients along the length of a fish (e.g., Denton and Gray, 1989, 1993). Since the gradient is greatest in the acoustic near field, the primary stimulus

54

ARTHUR N. POPPER

appropriate for this endorgan will only be found quite close to the source. Thus, ifthe fish is detecting sounds with the ears, it may be possible to detect particle displacement at a distance from the source that is a function of the initial sound level and ear sensitivity. However, initial sound level is not as relevant for detection by the lateral line since particle displacement gradients may not be sufficiently great in the far field to be detectable for a fish. While sound is important for communication behavior in a variety of species, sounds also are produced as a side-effect of other behaviors such as feeding or locomotion (see Moulton, 1963; Tavolga, 1971). (In fact, unintentional sounds may also be produced by zooplankton, thereby providing a signal to predators about their presence and location.) Another critical aspect of underwater acoustics that is relevant to fishes that live in shallower water such as rivers and streams is that the attenuation rates of sounds, and particularly oflow frequencies, is a function of water depth (Rogers and Cox, 1988). In shallow water there is far more attenuation oflow-frequency sounds (both pressure and particle displacement components) than in deep water (Rogers and Cox, 1988). The lowest frequency that can be propagated depends upon water depth - the deeper the water, the lower the frequency that can be propagated. For example, Rogers and Cox (1988) show that the lowest frequency that will be propagated in water 1 m deep is about 300 Hz, while in water 10 m deep the lowest frequency would be approximately 30 Hz. The specific cut-off frequency depends upon the nature of the bottom (e.g., rock, mud, etc.), but the cut-off frequency for other bottoms is lower than for the example used (which is for rocky bottoms). Because of this lack of propagation of low frequencies in shallow water, it is likely that fish in such an environment only detect sounds that are extremely close to them.

ACOUSTIC BEHAVIOR Acoustic communication has been well documented for many fish species (reviewed in Tavolga, 1971; Demski et al., 1973; Fine et al., 1977; Myrberg, 1981). As pointed out by Tavolga (1971), sound is a particularly useful channel for underwater communication since acoustic signals are not affected by murkiness or darkness of the environment, sound travels rapidly over long distances, is highly directional, and sound is not particularly affected by rocks or coral reefs (as long as they are small relative to the wavelength of the sound). In some species, such as toadfish (Opsanus), sounds are used over long distances by males to "call" females during the mating season (Winn, 1967). In other cases, such as the goby Bathygobius soporator, low intensity sounds are used to communicate between males and females that are very near one another (Tavolga, 1958). It has become clear from these studies that fish sounds are quite variant and depend upon the species. In addition, individual species may have more than one type of sound, with different sounds used in different behavioral contexts (Tavolga, 1971; Fine et al., 1977). Sounds are produced intentionally in a large number of species, although the emission mechanisms and the behavioral context of the sounds varies inter-specifically (Tavolga, 1971). "Stridulatory" sounds are produced by moving or grinding of body parts against one another such as using pharyngeal teeth or other hard body parts (similar mechanisms are found in snapping shrimp - Moulton, 1963). Other species produce

THE TELEOST OCTAVOLATERALIS SYSTEM

55

sounds by directly or indirectly involving the swim bladder. Some of these "swim bladder sounds" consist of short bursts of broad-band noise (especially those produced using stridulatory mechanisms), while others are tonal and contain a fundamental frequency and multiple harmonics (e.g., Tavolga, 1971; Demski et al., 1973; Fine et al., 1977). Perhaps the most elegant study of the use of sound in behavior was done by Myrberg and his colleagues on damselfish (family Pomacentridae) (e.g., Spanier, 1979; Myrberg and Spires, 1980). Several species live conspecifically on reefs and use sounds to defend territories, including nests. Each species has a sound that differs in frequency and in pulse repetition rate. Behavioral studies showed that species can discriminate between sounds ofconspecifics and heterospecifics based on differences in pulse rate, and there is even good evidence for individual recognition (Myrberg and Riggio, 1985).

STRUCTURE OF THE OCTAVOLATERALIS SYSTEM The Mechanoreceptive Lateral Line

The structure and function of the mechanoreceptive lateral line was reviewed by Coombs et al., (1989, 1992). The lateral line system consists of a series of receptors (called neuromasts) located over the body surface of a fish (Figure 1a). Each of the neuromasts contains sensory hair cells that are very similar to those found in the fish ear. The cilia on the sensory hair cells project into an overlying gelatinous cupula which acts as a "sail" in response to fluid motions of the water. Movement of the cupula produces bending of the cilia on the hair cells, and this results in stimulation of the sensory hair cells. Some of these receptors lie within canals that run along the fish's body (canal neuromasts), while other receptors, often called "free (or superficial) neuromasts," are located in pits or directly on the body surface (Figures 1e, f, g). There appear to be some differences in the functional response characteristics of the two parts of the lateral line (Miinz, 1989). Differences in canal vs. free neuromasts include the frequency range over which they respond (with canal neuromasts responding to higher frequencies than free neuromasts) (Miinz, 1989). The functional basis for these differences are not fully understood, but they may be related to variation in sizes of the sensory epithelia in the canals and free neuromasts, the number of hair cells (e.g., Denton 200m), the attenuation coefficient for 480 nm light was .0263 ± .0007 (n 3). The spectral distribution oflight was measured in waters off Cape Hatteras at several times approaching sunset. These measurements were taken at 30, 90 and 150m depth on

=

=

=

UV LIGHT IN THE DEEP-SEA 1.00E+13

189

~----------------------------,

1.00E+12

.,. ,. . . 'A ~

1.00E+11 ';"

E c:

\

/\,

1.00E+10

: 111 1.00E+09

'E (J

K(380) = 0.041

~ 1.00E+08

o.

0

0 ..r::

11. 1.00E+07

·o /

1.00E+06

·o ·o

K(380) = 0.030

0

1.00E+05 1.00E+04

+---+----+--+---+----+--+---+----+--+---+----+----! 0

50

100

150

200

250

300

350

400

450

500

550

600

Depth (m) Figure 2 Measurements of downwelling irradiance at 380 nm from 75 to 450 m water depth, in Northwest Providence Channel, Bahamas. Data have been corrected for cloud cover. Data were grouped, and lines calculated, according to an iterative procedure in which data points were added or subtracted until the best fit, according to the F -test, was obtained.

the same dive, at the times indicated on Figure 3. At 30 m depth (Figure 3A), there is a slight narrowing of the spectrum due to a reduction oflong wavelength (yellow-orange) light. This shift is in the same direction, although of a much smaller magnitude, as that reported by McFarland and Munz (1974, 1975) at similar time intervals approaching sunset in surface waters. At 90 m depth (Figure 3B), this shift in the spectrum is barely discernible, and by 150 m depth (Figure 3q, no change in the spectral distribution is apparent.

DISCUSSION The submersible based deployment of LoLAR and the PSlOOO allowed us to avoid the effects of ship shadow on optical data, a well-known problem with "over-the-side" radiometers (Jerlov, 1976; Smith and Baker, 1986; Voss, et a/., 1986; Gordon, 1985, Spinrad and Widder, 1992). In addition, measurements were made below the nonhomogeneous epipelagic zone, making the extrapolation of spectral irradiance to a depth of 600 m a valid estimate of the true photon flux at the two wavelengths measured.

TAMARA M. FRANK and EDITH A. WIDDER

190 1.0

A

0.8

_18:20 ······19:23

0.6 0.4

·. 0.2

. ..

' •

0.0

400

300

1.0

500

600

700

B

0.8

-18:33 ······19:19

0.6 0.4 0.2 0.0 400

300

500

600

700

c

1.0 0.8

-18:38 ...... 19:16

0.6 0.4 0.2

.-· ..

0.0 300

400

500

600

700

Figure 3 Spectral distribution oflight at 3 depths at various times approaching sunset, off Cape Hatteras, at 34° 14' N, 75° 44' W, on September 12, 1994. Sunset occurred at 19 : 20. Y-axis is relative spectral irradiance (in photons cm- 2 s- 1 nm- 1); X-axis is wavelength in nm. A) 30m depth. B) 90 m depth. C) 150m depth.

Previous investigators have demonstrated that diffuse attenuation coefficients change with depth (Clark and Wertheim, 1956; Clarke and Backus, 1964; Boden and Kampa, 1967). However, prior to our investigations, the minimum wavelength measured at

UV LIGHT IN THE DEEP-SEA

191

depths deeper than 200m depth was 408 nm (Kampa, 1970). Our data for UV light (380 nm) indicate that utilizing attenuation coefficients measured in shallow water will greatly underestimate the true photon flux of UV light at depth. Using our own data, it can be seen that the difference between the mean shallow attenuation coefficient (.054) and deep attenuation coefficient (.029) is substantial. These results indicate that a single attenuation coefficient cannot be used to calculate the photon flux at the shorter wavelengths. The photon flux of 380 nm light at 500-600 m (the average daytime depth range of the UV sensitive crustaceans - Table 1), extrapolated from the data in Figure 2, is approximately 2.5 x 104-3.5 x 10 5 photons cm- 2 s- 1 nm- 1. Our data were taken at a time when the surface water was comparatively "murky", due to high productivity and runoff, and at other times the water is more transparent. Therefore it is likely that the irradiance of 380 nm light is higher at times when there is less· particulate matter in the water. Nevertheless, even under these conditions, we estimate that the UV irradiance is sufficient for vision in the crustaceans with a UV spectral sensitivity peak. The LoLAR filter for 380 nm has a full width half maximum bandwidth (FWHM) of12 nm; therefore the actual values measured were 3 x 105-4 x 106 photons cm- 2 s- 1 12 nm- 1 . Since a visual pigment absorbs light efficiently over a waveband spanning 100 nm, a crustacean with a behavioral threshold to short wavelength light of 10 6-10 7 photons cm- 2 s- 1 (the mean behavioral threshold sensitivities of the crustaceans with UV spectral sensitivity peaks) should be able to detect the presence ofUV light at 600 m depth. One of the hypotheses we generated about the possible function ofUV light sensitivity to the crustaceans possessing two peaks of spectral sensitivity is that it may play a role in cueing their vertical migrations. Light is considered to be the most significant factor cueing vertical migrations in most cases, but very little is known about what characteristic of the changing light field triggers the migration (see Forward, 1988 for review). Between noon and dusk, the spectral distribution of underwater light near the surface changes significantly (McFarland and Munz, 1974, 1975; Jerlov, 1976). We suggested that the deep-sea crustaceans with two peaks of spectral sensitivity may be using this spectral shift as a cue to trigger their migrations (Frank and Widder, 1994a). Our data from Cape Hatteras indicate that this is not the case. As seen in Figure 3, a slight change in the spectral distribution of downwelling light is still visible at 30 m depth, but by 150 m depth, this shift is no longer apparent. The water in Cape Hatteras is not as transparent as the water in the Bahamas (Widder et al. in prep), but it is unlikely that spectral changes that cannot be discerned at 150m in Cape Hatteras water would be visible at 500-600 m (the daytime depth of the oplophorids in question) in the clearer Bahamian water. Boden (1961) describes a spectral shift between noon and sunset at 170 m depth for water off San Diego, but this spectral shift was towards longer wavelengths, which is in the opposite direction of both our data and those of McFarland and Munz (1974, 1975) for surface waters. In addition, Boden also describes a spectral shift towards shorter wavelengths at three times bracketing sunset, but a constant depth was not maintained during these measurements. Our measurements indicate that the spectral change that can be measured in shallower waters is not transmitted to the deep-sea in clear oceanic water, indicating that crustaceans living at 600 m depth in these waters must be responding to some other trigger to cue their upward migrations. The question of the functional significance of a dual visual pigment system in deepsea crustaceans remains. By monitoring the decrease in spectral bandwidth of downwelling light with increasing depth, such a system might function as a depth

192

TAMARA M. FRANK and EDITH A. WIDDER

gauge, as first suggested by Wald and Rayport (1977), who discovered high near-UVand blue-green light sensitivity in a deep-sea alciopid worm. There is also the possibility that UV sensitivity serves some function related to bioluminescence since all the deep-sea crustaceans with UV spectral sensitivity peaks (that we have studied to date) possess ventrally directed bioluminescent photophores. It is thought that organisms with these photophores may attempt to camouflage their silhouette from predators below them by matching the downwelling illumination with their own bioluminescence (Clarke 1963; Herring 1990). Predators with pigmented lenses, such as the yellow lenses of numerous species of deep-sea fish (Douglas and Thorpe, 1992, for review), would be able to break this camouflage when the match between the downwelling irradiance and bioluminescent irradiance is not very good (Muntz, 1976, 1983). Therefore, since sensitivity to two wavelengths would enable these crustaceans to more closely determine and match, with their own luminescence, downwelling irradiance, dual visual pigments in these species may aid them in avoiding detection by predators with pigmented lenses. Studies are currently underway to acquire an accurate picture of the spectral distribution and irradiance of light at depth, and how changes in these parameters, which can be replicated in the laboratory, affect the behavior of these crustaceans. Acknowledgements

This work was funded in part by a National Science Foundation Grant #OCE-9313972, and subcontract UNCW9410 from the University of North Carolina at Wilmington Award No. NA36RU0060-02. LoLAR was developed with support from ONR and NSF grant OCE89-17502. The authors gratefully acknowledge the captain and crew of the R. V. Edwin Link and the Johnson-Sea-Link submersible for assistance with data collection. This is Harbor Branch Contribution number 1089. References Boden, B. P. (1961) Twilight irradiance in the sea. Int. Un. Geod. Geophys. Monogr., 10, 96-101. Boden B. P. and Kampa, E. M. (1967) The influence of natural light on the vertical migrations of an animal community in the sea. Symp. zoo/. Soc. Lond., 19, 15-26. Clarke, G. L. and Backus, R. H. (1964) Interrelations between the vertical migration of deep scattering layers, bioluminescence, and changes in daylight in the sea. Bull. Inst. oceanogr. Monaco, 64, 1-35. Clarke, G. L. and Wertheim, G. K. (1956) Measurements of illumination at great depths and at night in the Atlantic Ocean by means of a new bathyphotometer. Deep-Sea Res., 3, 189-205. Clarke,W D. (1963) Function of bioluminescence in mesopelagic organisms. Nature, Lond., 198, 1244-1246. Dartnall, H. J. A. (1974) Assessing the fitness of visual pigments for their photic environments. In: Vision in Fishes- New Approaches in Research, M. Ali, ed., Plenum Press, New York, pp. 543-562. Denys, C. J. and Brown, P. K. (1982) Euphausiid visual pigments. The rhodopsins of Euphausia superba and Meganyctiphanes norvegica. J Gen. Physiol., 80, 45451-472. Douglas, R. H. and Thorpe, A. (1992) Short-wave absorbing pigments in the ocular lens of deep-sea teleosts. J mar. bioi. Ass. U K., 72, 93-112. Fisher, L. R. and Goldie, E. H. (1958) The eye pigments of a euphausiid crustacean, Meganyctiphanes norvegica (M. Sars). XV Intern. Cong. Zoo/. Lond. Proc., 533-535. Fisher, L. R. and Goldie, E. H. (1960) Pigments of compound eyes. Prog. Photobiol. Proc. 3rd Int. Congr. Photobiol., 153-154. Forward, R. B., Jr. (1988) Die! vertical migration: zooplankton photobiology and behavior. Oceanogr. Mar. Bioi. Ann. Rev., 26,361-393. Frank, T M. and Case, J. F. (1988) Visual spectral sensitivities of bioluminescent deep-sea crustaceans. Bioi. Bull., 175, 261-273.

UV LIGHT IN THE DEEP-SEA

193

Frank, T M. and Widder, E. A. (1994a) Evidence for behavioral sensitivity to near-UV light in the deep-sea crustacean Systellaspis debilis. Mar. Bioi., 118, 279-284. Frank, T M. and Widder, E. A. (1994b) Comparative study ofbehavioral-sensitivity thresholds to near-UVand blue-green light in deep-sea crustaceans. Mar. Bioi., 121, 229-235. Goldsmith, T H. (1986) Interpreting trans-retinal recordings of spectral sensitivity. J Camp. Physiol., 159, 481-487. Gordon, H. R. (1985) Ship perturbations of irradiance measurements at sea. I. Monte Carlo simulations. Appl. Opt., 24,4172-4182. Herring, P. J. (1983) The spectral characteristics of luminous marine organisms. Proc. R. Soc. Land, 8220, 183-217. Herring, P. J. (1990) Bioluminescent communication in the sea. In: Light and Life in the Sea, P. J. Herring, A. K. Campbell, M. Whitfield, and L. Maddock, eds., Cambridge University Press, Cambridge, pp. 245-264. Jerlov, N. G. (1976) Marine Optics. Elsevier, Amsterdam. Kampa, E. M. (1970) Underwater daylight and moonlight measurements in the eastern north Atlantic. J mar. bioi. Ass. UK., 50,397-420. Latz, M. 1., Frank, T M. and Case, J. F. (1988) Spectral composition of bioluminescence of epipelagic organisms from the Sargasso Sea. Mar. Bioi., 98,441-446. McFarland, W N. and Munz, F. W (1974) The visible spectrum during twilight and its implications to vision. In: Light as an Ecological Factor. II, G. C. Evans, R. Bainbridge, and 0. Rackham, eds., Blackwell Scientific Publications, Oxford, pp. 249-271. McFarland, W N. and Munz, F. W (1975) The photic environment of clear tropical seas during the day. Vision Res., 15: 1063-1070. Menzel, R. (1979) Spectral sensitivity and color vision in invertebrates. In: Handbook of Sensory Physiology, Jill. Vll/6A, H. Autrum, ed., Springer-Verlag, Berlin, pp. 503-580. Muntz,W R. A. (1976) On yellow lenses in meso pelagic animals. J mar. bioi. Ass. UK., 56, 963-976. Muntz, W R. A. (1983) Bioluminescence and vision. In: Experimental Biology at Sea, A. G. MacDonald and I. G. Priede, eds., Academic Press, London, pp. 217-238. Smith, R. C. and Baker, K. S. (1986) Analysis of ocean optical data II. SPIE Ocean Optics VIII, 637, 95-107. Spinrad, R. W and Widder, E. A. (1992) Ship shadow measurements obtained from a manned submersible. SPIE Ocean Optics XI, 1750, 372-383. Voss, K. J., Nolten, J. W and Edwards, G. D. (1986) Ship shadow effects on apparent optical properties. SPIE Ocean Optics VIII, 637, 186-190. Wald, G. and Rayport, S. (1977) Vision in annelid worms. Science, 196, 1434-1439. Widder, E. A., Latz, M. I. and Case, J. F. (1983) Marine bioluminescence spectra measured with an optical multichannel detection system. Bioi. Bull, 165, 791-810. Widder, E. A., Caimi, F. M., Taylor, L.D. and Tusting, R. F. (1992) Design and development of an autocalibrating radiometer for deep-sea bio-optical studies. Ocean. Eng. Soc. IEEE, Oceans '92. Zieman, D. A. (1975) Patterns of vertical distribution, vertical migration, and reproduction in the Hawaiian mesopelagic shrimp ofthe family Oplophoridae. Ph.D. Thesis, University of Hawaii, Honolulu, pp.l6-20.

ULTRAVIOLET VISUAL SENSITIVITY IN THE LARVAE OF TWO SPECIES OF MARINE ATHERINID FISHES ELLIS R. LOEW\ FLORENCE A. McALARY 2 and WILLIAM N. McFARLAND 2

2

1 Physiology, Cornell University Wrigley Marine Science Center, University of Southern California

Microspectrophotometry of single retinal photoreceptors oflarval California topsmelt, Atherinops affinis, and the grunion, Leuresthes tenuis, revealed the presence of a cone cell that contained an ultraviolet visual pigment (maximum absorbance at 355 nm, half bandwidth 70 nm). Topsmelt larvae 6-19 mm standard length fed on Artemia when exposed to ultraviolet (UV-A, 300-400 nm) light alone at least as well as in full sunlight. Grunion 24 hours after hatching 6-8 mm standard length also fed on Artemia in UV-A light alone. Measurement of the spectral near-surface radiance field in which both species forage indicate that UV-A photons between 300 and 400 nm constitute as much as 18-20 percent ofthe total number of photons between 300 and 750 nm. The data imply that ultraviolet vision in these obligate planktivorous larvae facilitated foraging. We suggest that the ability to detect zooplankton in the near-ultraviolet may be widespread amongst fish larvae that forage on zooplankton where ultraviolet light intensities are high.

INTRODUCTION Retinal cones containing a visual pigment maximally sensitive in the near-ultraviolet are present in several freshwater fishes (reviewed in Bowmaker, 1991), and in euryhaline and marine fishes (Harosi and Fukorotami, 1986; McFarland and Loew, 1994). Several functions of near-UV sensitivity in fishes have been suggested: orientation, navigation, and polarized light detection (Hawryshyn and McFarland, 1987; Hawryshyn, 1992), contrast enhancement and foraging on plankton (Bowmaker and Kunz, 1987; Loew eta/., 1993; Browman eta/., 1994) and the possible discrimination of discrete UVpatterns (Lythgoe and Shand, 1984; Harosi, 1985; Loew and McFarland, 1990). Because larvae of many marine and fresh water fishes are obligate diurnal particulate planktivores in euphotic environments (see Gerking, 1994) where ultraviolet radiance levels are relatively intense (Baker and Smith, 1982), ultraviolet visual sensitivity, if present, could be important in feeding success. We report the presence of retinal cone photoreceptor cells in two species of larval marine atherinid fishes containing a visual pigment having an absorbance maximum (Amax) in the near ultraviolet (UV-A) at 355 nm. We also demonstrate that each species can feed effectively on zooplankton (Artemia) when exposed to UV-A photons alone, and evaluate the radiant underwater light field of the near-surface environment in which these larvae were observed to feed.

195

196

ELLIS R. LOEW ETAL.

METHODS Collections

Larval topsmelt, Atherinops affinis, (6-19 mm standard length, SL) were collected from the surface using a dip net at Big Fisherman Cove, Catalina Island, USA, from June through early July 1994. Larvae were maintained in aquaria in running seawater for visual pigment analysis and feeding experiments. Grunion eggs, Leuresthes tenuis, were obtained from the beaches at Santa Barbara, California, transported in moist sand, and subsequently hatched at the Wrigley Marine Science Center and Cornell University. Surface temperatures through this period varied from 17 to 20 degrees Celsius. Visual Pigment Analysis

Larval fish were dark adapted, anaesthetized in cold water and killed by destroying the brain with a sharp scalpel. After enucleation the retinas were isolated and placed in a drop of physiological saline (Sigma Modified MEM, pH 7.6) or standard fish Ringer's on a cover slip and macerated with razor blades. The tissue was sandwiched with a second cover slip whose edges were sealed against the first with Dow Corning high vacuum silicone grease. The sample was then transferred to the stage of the microspectrophotometer (MSP). All procedures described above were carried out under infrared light with the aid of infrared image converter goggles or very brief exposure to dim red light. Two MSPs, one at The Wrigley Marine Science Center and the other at Cornell University were used in this study. Each is a single beam, computer controlled instrument which measures absorbance of a single cell. Both machines were fitted with quartz and fluorite optics to allow measurement in the near-UV, and identical electronic, monochromator and wavelength drive mechanisms. The condenser lens was either a Leitz Ql70 mirror condenser, N.A. 0.4 or a Zeiss UV Achromat condenser, N.A. set at 0.35. The objective was either a Leitz 63 x fluorite, N.A. 0.85 or a Zeiss 100 x Ultrafluor, N.A. 0.85. Wavelengths were scanned in both directions with even wavelengths displayed on the way down and odd wavelengths displayed on the way up. Comparison of these spectra indicated that there was no detectable bleaching of the pigments during measurements. For both machines, wavelength uncertainty was ± 1 nm as determined with a mercury line source, so all data are reported to the nearest integer. Both machines yielded identical spectra for human blood cells and a calibration filter made from a piece of Kodak CC025M-AAA color compensation film. To establish that a spectral absorbance curve was actually a visual pigment two criteria were used: first, that the outer segment of the photoreceptor that contained the visual pigment was dichroic and second, that it was photosensitive. The former relies on the fact that visual pigments are oriented in the outer segment membranes such that light polarized at right angles to the long axis has a greater probability of absorption than light polarized parallel to the outer segment axis. The latter depends on the fact that all visual pigments are inherently photosensitive. Data were analyzed in a manner similar to that described by McFarland and Loew (1994). First, each MSP curve was smoothed using a digital Fourier filter subroutine having DC restoration ("smooft", Press eta/., 1987). The smoothed data were then differentiated to provide a preliminary estimate of the wavelength of maximum

UV VISION IN LARVAL MARINE FISHES

197

absorbance (Amax). The curve was normalized to that Amax and transformed to a normalized frequency scale in the manner of Mansfield (1985) and MacNichol (1986). A linear regression was then performed on the data between 30% and 70% of the Amax on the low frequency (long wavelength) limb of the doubly normalized curve, and between 40% and 70% of the A max on the high frequency (short wavelength) limb. By comparing the calculated linear regression values at 1 nm intervals to accepted templates for visual pigment curves (Partridge and deGrip, 1991 for rhodopsins; Bridges, 1967 for porphyropsins) and transforming back to wavelength, we generated 40 estimates for Amax from the long wavelength limb and 30 estimates from the short wavelength limb for each visual pigment curve, except for the UV visual pigment curve where only the long wavelength limb could be analyzed. We then determined a mean Amax ± standard deviation (SD) using the short wavelength limb estimates, the long wavelength limb estimates and the combined estimates. For each of these three Amax values a template curve was calculated and drawn. Each template curve was overlaid on the raw and filtered original data and a decision as to which fitted best was made by visual examination. Usually, the template fit having the lowest SD also had the best visual fit. After repeating this process for each MSP curve, the Amax values for each curve of a class were averaged to yield a final estimate of Amax ± 1 SD. Feeding Experiments Outdoor trials: To determine if young topsmelt can feed under ultraviolet light, we placed groups of 5-6 freshly collected topsmelt (6-18 mm SL) unfed for 24 hours in an aquarium containing 1-liter seawater (10 em deep) into a light tight chamber (internal dimensions 12.5 x 12.5 x 20 em) painted non-reflective black. The lid of the chamber was fitted with a HOYA UV-A transmitting filter system (Figure 1). Each trial was conducted between 1000 and 1400 hours Pacific standard time when the sun was close to its zenith. The chamber was oriented normal to the solar beam to maximize exposure. Once the chamber was in place, fish were observed in the open chamber (lid oft) for three minutes before beginning a feeding trial. At the beginning ofeach trial, 1ml of seawater containing 1100-1500 one to two day old Artemia was introduced to the aquarium by pipette as the chamber lid was closed. This provided a high concentration of prey in the test aquarium ( > 1 Artemia per ml) to ensure sufficient food during each trial. The number of Artemia per ml was estimated by counting the number of Artemia in five 0.05 ml samples from each solution of Artemia used (average number Artemia per 0.05 ml = 65.6 ± 11 SD, n = 40). Water in the aquarium was agitated continuously with an aerator to distribute prey throughout the volume. At the end of ten minutes, larvae were collected by emptying the aquarium through a net and euthanized by rapid freezing. The standard length of each fish was measured, its stomach and intestine dissected free, and the number of Artemia in the gut counted using a dissecting microscope (6-25 x ). As controls, we also tested feeding responses of young topsmelt in darkness and full sunlight. The same protocol was used except the chamber lid was either covered with a piece of opaque plastic and the entire apparatus with a dark canvas eliminating all light, or left off to expose the chamber to full sunlight. The feeding responses of grunion 24 hours post hatching (6-8 mm SL) were tested outdoors in UV-A and darkness using the same procedures.

Feeding responses of young topsmelt were tested at three intensities of UV-A light in the laboratory. In each trial, 5-6 top smelt (10-19 mm SL) unfed for

Laboratory tests:

ELLIS R. WEW ETAL.

198

50

-.....

40

>


"§ c:

100

.S!

-:;;

11> 0>

E

iQ

10

i2

~

Ir

I

8 +I

I

~I 0>

~I ....5 0

0

I I ---------I ------ -I--- ---

~

T

I II

....

-~

(.)

11>

w

-I

c .I Food density

10 100 ( 1-19 ml- 1 )

1000

Figure 1 Relationship between total food density and: (A) ingestion of Rhodotorulaglutinis and (B) ingestion of Euglena gracilis by the rotifer Brachionus calyciflorus. (Q Electivity for Euglena based on the relative mortalities of the two foods in each preparation. Symbols represent two experiments, in all cases means ± 95% confidence intervals, n = 3, 20-60 rotifers per sample.

SENSORY BIOLOGY OF ROTIFER FEEDING

263

alga, the rotifer consumes food independently of its availability over a wide range of densities (1.0~100 p,g dry wt mi- 1) (Starkweather and Gilbert, 1977). This difference is consistent with the food-specific coronal behaviors used by the animals when feeding on the two cell types. Euglena elicit the formation of pseudotrochal screens over the buccal funnel when at moderate-high density; R. glutinis cells do not (Gilbert and Starkweather, 1977). In the light of these different feeding behaviors, John Gilbert and I examined the responses of B. calyciflorus which were exposed to 1: 1 mixtures (by dry weight) of yeast and Euglena cells over a range offood densities between 0.1 and 250 p,g mi- 1• R. glutinis cells were ingested at rates directly proportional to their availability in suspension (Figure 1A), a result similar to that found in pure yeast preparations. The series of ingestion rates simultaneously obtained for Euglena (Figure 1B), on the other hand, differed from both the yeast curves and the characteristic, pure suspension Euglena pattern. Below 10 p,g mi- 1 total food density, the uptake of the alga mimicked that found in pure suspensions, including an interval (1~10 p,g ml- 1) of density independence. At higher densities uptake of Euglena in the mixtures departed radically from that found for the alga in pure suspensions (Starkweather and Gilbert, 1977); ingestion rate increased sharply with food density, roughly in parallel with yeast ingestion. The important point in the context of rotifer feeding behavior and food selectivity is that electivity indices (Jacob's "D", Jacobs, 1974) calculated for rotifers in the 1 : 1 food mixtures varied substantially with total food density (Figure 1C). An electivity of zero indicates no preference for either food type, while a positive or negative value implies selective ingestion of one food type or the other. These results indicate that the apparent preference for a particular food cell type can be changed dramatically depending on overall food (and perhaps total particle or seston) density. This may help explain the variable results of several in situ or laboratory tracer cell experiments (like several listed in Table 3); the differing particulate background of lake waters or experimental suspensions might have switched, modulated or accentuated the apparent electivities of some rotifer/food type combinations. FUTURE APPROACHES TO ROTIFER SENSORY BIOLOGY Physiology

Rotifers may be relatively intractable for conventional electrophysiology. They are, of course, very small. They often have resilient integuments (loricae) and frequently have indistinct cell boundaries including some syncytia; therefore, appropriate placement of electrodes may be difficult. Rotifers are also, as noted above, continuous swimmers, with feeding processes closely coupled to locomotion. Some suspension-feeding species normally swim helical paths (for Brachionus: right-handed, ventral side in) which may be important in feeding regulation. While many basic feeding functions certainly continue in restrained animals (Gilbert and Starkweather, 1977; Starkweather, 1995), more subtle regulatory responses (e.g. mechanoreception) may be subject to artifact. One promising approach may be laser killing of individual or suites of putative sensors or their opposed nerves (Sengupta et al., 1993). This may be possible with rotifers treated with appropriate anesthesia. Subsequent behavioral changes may indicate the functional classes of cells destroyed.

264

PETER L. STARKWEATHER

Also feasible may be probing rotifer tissues or ciliary structures with modality-specific labeled oligonucleotides, proteins or antibodies (see Snell et al., 1995). One recent study has shown tissue localization of olfactory function by probing for olfactomedin in ciliated olfactory tissues (Yokoe and Anhalt, 1993). This work, while admittedly vertebrate, might be extended to invertebrate taxa with straightforward molecular tools. Ecology

The complexity of sensory/behavioral processes in rotifers demands continued examination of differential feeding and its sensory bases. New work should take special care to consider details of the experimental environment, with, for instance, increased attention to overall food density effects. It also appears that experimental work on interacting food chemistries and the relationship between chemical and mechanical stimulation will be valuable in the context of better describing field behaviors. Characterization of sensory capabilities of additional members of the zooplankton community, crustaceans as well as rotifers, is essential for enhanced understanding of pelagic structure and function. The fields of sensory physiology and physiological ecology may then converge on a more complete understanding of the dynamic interface between the ecophysiological demands of the organism and its surrounding aquatic world. Acknowledgements

Sincere thanks are due to J. J. Gilbert who collaborated on and/or inspired many of the described studies on Brachionus feeding, and to R. L. Wallace, who provided valuable insight and references on the biology of Cupelopagis. Supported in part by NSF grant DEB 78-02882 (Dartmouth College) and by the UNLV Limnological Research Center Endowment. References Andrews, J. C. (1983) Deformation of the active space in the low Reynolds number feeding current of calanoid copepods. Can. J Fish. Aquat. Sci. 40, 1293-1302. Atema, J. (1988) Distribution of chemical stimuli. In: Sensory Biology of Aquatic Animals. J. Atema, R. R. Fay, A. N. Popper and W N. Tavolga (eds), Springer-Verlag, Berlin. pp. 29-56. Bevington, D. 1., White, C. and Wallace, R. L. (1995) Predatory behavior of Cupelopagis vorax (Rotifera; Collothecacea; Atrochidae) on protozoan prey. Hydrobiologia, in press. Bogdan, K. G. and Gilbert, J. J. (1987) Quantitative comparison offood niches in some freshwater zooplankton. A multi-tracer-cell approach. Oecologia 72, 331-340. Buskey, E. (1984) Swimming pattern as an indicator or the roles of copepod sensory systems in the recognition of food Mar. Bioi. 79,165-175. Carlisle, M. J. (1975) Taxes and tropisms: diversity, biological significance and evolution. In: Primitive Sensory and Communication Systems. M. J. Carlisle (ed) Academic Press, N.Y pp.l-28. Charoy, C. (1995) Modifications of the swimming behavior of Brachionus calycifWrus (Pallas) according to both food environment and individual nutritive state. Hydrobiologia, in press. Charoy, C. and Clement, P. (1993) Foraging behaviour of Brachionus calycifWrus (Pallas): variations in the swimming path according to presence or absence of algal food (Chlorella). Hydrobiologia 255/256, 95-100. Clement, P. (1987) Movements in rotifers: correlations of ultrastructure and behavior. Hydrobiologia 147, 339359. Clement, P., Amsellem, J., Cornillac, A.-M., Luciani, A. and Ricci, C. (1980) An ultrastructural approach to feeding behaviour in Philodina roseola and Brachionus calyciflorus (Rotifers). Hydrobiologia 73, 137-141. Clement, P., Wurdak, E. and Amsellem, J. (1983) Behavior and ultrastructure of sensory organs in rotifers. Hydrobiologia 104, 89-130.

SENSORY BIOLOGY OF ROTIFER FEEDING

265

Clement, P. and Wurdak, E. (1991) Rotifera (Chapter 6). In: Microscopic Anatomy of Invertebrates. F. W. Harrison and E. E. Ruppert (eds) Vol. 4. Wiley-Liss, N.Y pp. 219-297. DeMott, W. R. (1982) Feeding selectivities and relative ingestion rates in Daphnia and Bosmina. limnol. Oceanogr. 27,518-527. DeMott,W. R. (1986) The role of taste in food selection by freshwater zooplankton. Oecologia 69,334-340. Egloff, D. A. (1988) Food and growth relations of the marine microzooplankter, Synchaeta cecilia (Rotifera). Hydrobiologia 157, 129-141. Fenchel, T. (1980) Suspension feeding in ciliated protozoa: feeding rates and their ecological significance. Microb. Ecol. 6, 13-25. Gilbert, J. J. (1980) Feeding in the rotifer Asplanchna: behavior, cannibalism, selectivity, prey defenses, and impact on rotifer communities. In: Evolution and Ecology of Zooplankton Communities. W. C Kerfoot (ed) Special Symp.Vol. 3 American Soc. Limnology and Oceanography. University Press of New England. pp. 158-172. Gilbert, J. J. (1985) Escape response of the rotifer Polyarthra: a high-speed cinematographic analysis. Oecologia 66,322-331. Gilbert, J. J. (1987) The Polyarthra escape from response: defense against interference from Daphnia. Hydrobiologia 147, 235-238. Gilbert, J. J. and Bogdan, K. G. (1984) Rotifer grazing: in situ studies on selectivity and rates. In: Trophic Interactions Within Aquatic Ecosystems. D. G. Meyers and J. R. Strickler (eds) AAAS Selected Symposium 85. pp. 97-133. Gilbert, J. J. and Starkweather, P. L. (1977) Feeding in the rotifer Brachionus calyciflorus I. Regulatory mechanisms. Oecologia 28, 125-131. Gilbert, J. J. and Starkweather, P. L. (1978) Feeding in the rotifer Brachionus calyciflorus III. Direct observations on the effects offood type, food density, change in food type and starvation on the incidence of pseudotrochal screening. Ji?rh. int. Ji?rein. Limnol 20, 2382-2388. Hellebust, J. A. (1974) Extracellular products. In: Algal Physiology and Biochemistry. W. D. P. Stewart (ed.), University of California Press. pp. 838-863. Hong, S. -J., Chai, J. -Y and Lee, S. -H. (1991) Surface ultrastructure of the developmental stages of Heterophyopsis continua (Trematoda: Heterophyidae).J Parasitol. 77,613-620. Hyman, L. H. (1951) The Invertebrates: Acanthocephala, Aschelminthes, and Entoprocta. The pseudocoelomate bilateria. McGraw Hill, N.Y Jacobs, J. (1974) Quantitative measurement of food selection. A modification of the foraging ratios and Ivlev's electivity index. Oecologia 14,413-417. Johnson, B. R., Borroni, P. F. and Atema, J. (1985) Mixture effects in primary olfactory and gustatory receptor cells from the lobster. Chemical Senses 10, 367-373. Kirk, K. L. (1991) Inorganic particles alter competition in grazing plankton: the role of selective feeding. Ecology 72, 915-923. Koehl, M. A. R. (1984) Mechanisms of particle capture by copepods at low Reynolds number: possible modes of selective feeding. Trophic Interactions Within Aquatic Ecosystems. D. G. Meyers and J. R. Strickler (eds) AAAS Selected Symp. 85. pp. 135-166. Koehl, M. A. R. and Strickler, J. R. (1981) Copepod feeding currents: food capture at low Reynolds number. Limnol. Oceanogr. 26,1062-1073. Koste, W. (1'973) Ein merkwiirdiges festsitzendes Riidertier: Cupelopagis vorax. Mikrokosmos 4, 101-106. Laverack, M. S. (1974) The structure and function of chemoreceptor cells. Chemoreception in Marine Organisms. In: P. T. Grant and A. M. Mackie (eds) Academic Press. pp.l-48. Laverack, M. S. (1988) The diversity of chemoreceptors. In: Sensory Biology of Aquatic Animals. J. Atema, R. R. Fay, A. N. Popper and W. N. Tavolga. (eds.). Springer-Verlag. pp. 287-312. Makarewicz, J. C and Likens, G. E. (1979) Structure and function of the zooplankton community of Mirror Lake, New Hampshire. Ecol. Monogr. 49, 109-127. Nielsen, C (1987) Structure and function of metazoan bands and their phylogenetic significance. Acta Zoologica (Stockh.) 68, 205-262. No grady, T. and Alai, M. (1983) Cholinergic neurotransmission in rotifers. Hydrobiologia 104, 149-153. Nogrady, T., Wallace, R. L. and Snell, T. W. (1993) Rotifera. fiJI. I: Biology, Ecology and Systematics. SPB Academic Publishing, The Hague. Pourriot, R. (1977) Food and feeding habits ofRotifera. Arch. Hydrobiol. Beih. 8, 243-260. Price, H. J. (1988) Feeding mechanisms in marine and freshwater zooplankton. Bull Mar. Sci. 43, 327-343. Rassoulzadegan, F., Fenaux, L. and Strathmann, R. R. (1984) Effect of flavor and size on selection offood by suspension-feeding plutei. Limnol. Oceanogr. 29, 357-361. Remane, A. (1933) Rotatorien. In: Klassen und Ordnungen des Tier-Reichs. H. G. Bronn, ed., Vol4 (Part 2, Sections 1-4). C F. Winter, Leipzig. pp.l-576.

266

PETER L. STARKWEATHER

Sengupta, P., Colbert, H. A., Kimmel, B. E., Dwyer, N. and Bargmann, C. I. (1993) The cellular and genetic basis of olfactory responses in Caenorhabditis elegans. In: The Molecular Basis ofSmell and Taste Transduction. D. Chadwick, J. Marsh and J. Goode (eds.), Ciba Foundation Symposium 179. John Wiley & Sons, N.Y pp. 235-250. Snell, T. W, Rico-Martinez, R., Kelly, L. N. and Battle, T. E. (1995) Identification of a sex pheromone from a rotifer. Marine Biology, in press. Spittler, P. (1988) La seleccion del tamano de las particulas alimenticias por Brachionus plicatilis (Rotifera). Revista de Investigaciones marinas 9, 91-96. Starkweather, P. L. (1980a) Aspects of the feeding behavior and trophic ecology of suspension-feeding rotifers. Hydrobiologia 73, 63-72. Starkweather, P. L. (1980b) Behavioral determinants of diet quantity and diet quality in Brachionus calycifiorus. In: Evolution and Ecology of Zooplankton Communities. W C. Kerfoot (ed) University Press of New England. pp. 151-157. Starkweather, P. L. (1987) Rotifer energetics. In: Animal EnergeticsT. J. Pandian and F. J. Vernberg (eds) Vol. I. Academic Press, N.Y pp. 159-183. Starkweather, P. L. (1995) Near-control fluid flow patterns and food cell manipulation in the rotifer Brachionus calycifiorus. Hydrobiologia, in press. Starkweather, P. L. and Bogdan, K. (1980) Detrital feeding in natural zooplankton communities: discrimination between live and dead algal foods. Hydrobiologia 73, 83-85. Starkweather, P. L. and Gilbert, J. J. (1977). Feeding in the rotifer Brachionus calycifiorus II. Effect of food density on feeding rates using Euglena gracilis and Rhodotorula glutinis. Oecologia 28, 133-139. Strickler, J. R. (1984) Sticky water: a selective force in copepod evolution. In: Trophic Interactions Within Aquatic Ecosystems. D. M. Meyers and J. R. Strickler (eds.) AAAS Selected Symp. 85. pp.l87-239. Vasisht, H. S. and Dawar, B. L. (1969) Anatomy and histology of the rotifer Cupelopagis vorax Leidy. Research Bull. (NSJ Punjab University 20,207-221. Verity, P. G. (1988) Chemosensory behavior in marine planktonic ciliates. Bull. Mar. Sci. 43, 772-782. Wang, M. M. and Reed, R. R. (1993) Molecular mechanisms of olfactory neuronal gene regulation. In: The Molecular Basis of Smell and Taste Transduction. D. Chadwick, J. Marsh and J. Goode (eds.) Ciba Foundation Symposium 179. John Wiley & Sons, N.Y pp. 68-72. Yokoe, H. and Anholt, R. R. H. (1993) Molecular cloning of olfactomedin, and extracellular matrix protein specific to olfactory neuroepithelium. Proc. Nat/. Acad. Sci. USA 90, 4655-4659.

CHARACTERISTICS OF THE MATE-RECOGNITION PHEROMONE IN BRACH/ONUS PLICATILIS (ROTIFERA) TERRY W SNELL and ROBERTO RICO-MARTINEZ School of Biology, Georgia Institute of Technology, Atlanta, GA 30332-0230 e-mail: terry. [email protected]. edu The chemical recognition systems used by aquatic invertebrates to discriminate mates are poorly understood. We describe the results of analysis of a surface glycoprotein that serves as a mate-recognition pheromone (MRP) for the marine rotifer Brachionus plicatilis. Binding of a polyclonal antibody to the MRP signal on females is described. The binding of anti-MRP to purified MRP in nitrocellulose membranes is non-linear. AntiMRP also can be used to detect MRP in crude homogenates of rotifer proteins where MRP comprises 0.06 to 0.14% of total protein. As little as 1-2 ng ofMRP can be detected in crude homogenates. We estimated that a single B. plicatilis female has about 0.11 ng ofMRP concentrated primarily in her corona. Males can discriminate conspecific females from those of other species, females from different geographic populations, and males from females in their own populations. Monosaccharide analysis of the MRP demonstrated the presence of mannose, glucose, N-acetylglucosamine, fucose, and N-acetylgalactosamine in molecular ratios of 64 : 23 : 2.2 : 1.4 : 1, respectively. Comparison to known biantennary oligosaccharide structures reveals a large excess of mannose and glucose, suggesting further structural analysis is required. The rotifer mate-recognition system is likely to be useful for investigating the molecular basis of speciation resulting from the differentiation of recognition molecules.

INTRODUCTION Substantial effort has gone into investigating sexual recognition systems of terrestrial insects so that now the molecular structure and mechanism of action of many sex pheromones is well understood (Carde and Baker, 1984). As is often the case, Drosophila has been a useful model for examining the role of sex pheromones in recognition and reproductive isolation. Sexual recognition is mediated by contact pheromones that are long chain hydrocarbons on the cuticle surface (Coyne, 1992; Coyne et al., 1994). Closely related but sexually isolated species differ in the composition and quantity of these compounds which induce male courtship through stimulation of chemoreceptors on male forelegs and proboscis. The genetic basis of this recognition system is currently the subject of investigation. In contrast to insects, little is known about the type of molecules employed by aquatic animals for sexual recognition and the basis for their detection and discrimination. However, there is renewed interest (Atema, 1985; Carr et al., 1987; Dunham, 1988; Gleeson, 1991; Knowlton, 1993; Larsson and Dodson, 1993; Palumbi, 1994). There also is a growing appreciation for the constraints on small organisms attempting to communicate chemically in aquatic environments (Dusenbery and Snell, 1995). Progress has been made in elucidating the molecular mechanisms of recognition systems in a few aquatic organisms. Surface glycoproteins responsible for mating type recognition in some algae and ciliates have been isolated and characterized (Weise, 1965; Kubota et al., 1973; Miyake and Beyer, 1974; Luporini and Miceli, 1986; 267

268

TERRY W. SNELL and ROBERTO RICO-MARTINEZ

Miceli eta!., 1992). The molecular basis of gamete recognition has been described in sea urchins and abalone. The sperm protein bindin in sea urchins attaches to the vitelline coat of conspecific eggs, promoting sperm-egg fusion (Foltz and Lennarz, 1993; Minor et al., 1993). Bindin has been cloned and sequenced, revealing highly conserved amino acid sequences (Gao et al., 1986; Glabe and Clark, 1991). The egg receptor also has been isolated, cloned and sequenced (Foltz et al., 1993). In abalone, the sperm protein lysin allows the sperm to penetrate the outer chorion of the egg (Vacquier et al., 1990). Lysins act efficiently only on chorions of conspecifics, limiting fertilizations between species (Lee and Vacquier, 1992). Lysin has been purified, sequenced, and its secondary structure determined (Shaw eta!., 1993). Mate recognition in the marine rotifer Brachionus plicatilis has been shown to be determined by a surface glycoprotein that functions as a contact sex pheromone (Snell et al., 1995). This 29 kD glycoprotein has been purified and found to be glycosylated with oligosaccharides containing mannose, glucose, N-acetylglucosamine, and fucose. This glycoprotein enables males to discriminate conspecific females from those of other species. Males also are capable of discriminating conspecific females from different geographic populations and males from females in their own population (Snell and Hawkinson, 1983; Rico-Martinez and Snell, 1995). Male rotifers cannot, however, discriminate conspecific amictic females from mictic females. Our objective in this paper is to extend the understanding of how the rotifer materecognition pheromone (MRP) functions in discrimination. We examine the quantity of MRP per female, the binding of a polyclonal antibody to purified MRP, and the carbohydrate composition of the oligosaccharide moiety of the MRP. We expect that rotifer species may differ in the oligosaccharide components of their MRPs and that a compositional analysis can reveal these differences. These data will provide insight into how changes in recognition molecules lead to the formation of reproductive barriers between species. METHODS A polyclonal antibody against the MRP was prepared in a rabbit (Snell eta/., 1995) and used in all experiments. A suspension of total proteins from B. plicatilis was prepared by homogenizing about 10 g of wet-weight biomass with sample buffer (0.5 M Tris-HCl, pH 6.8, with 2% sodium dodecyl sulfate, 10% glycerol, and 5% ,8-mercaptoethanol). Electrophoresis in 12% polyacrylamide-SDS gels was performed to separate the proteins according to molecular weight (Snell et a/., 1995). Proteins then were transferred to a nitrocellulose membrane (Sigma) with > 90% efficiency as demonstrated using ovalbumin as a standard. The MRP protein was detected by Enhanced Chemoluminescence (ECL, Amersham). Briefly, this protocol consists of overnight blocking of proteins in the membrane with a 5% dry-milk solution diluted in PBS-Tween 20 (10 mM Na2 HP0 4 , pH 7.2, with 137 mM NaCl and 0.1% Tween-20). The membrane then was washed with Tris-HCl saline once for 15 minutes, then twice for 5 minutes. After the washes, the membrane was exposed for one hour to a solution of 1 : 500 anti-MRP in rabbit serum diluted with blocking solution. The membrane was washed again as described above, and then exposed for one hour to a solution of 1 : 1500 horseraddish peroxidase labelled anti-rabbit IgG in donkey serum diluted with blocking solution. Washes were performed as described, except that there were four

269

ROTIFER MATE-RECOGNITION PHEROMONE

5 minute washes instead of two. Horseraddish peroxidase activity was detected by exposing the membrane for one minute to an equal volume of detection reagents l and 2 (Renaissance kit, DuPont) in a final volume of 0.125 ml per cm2 of membrane. Excess detection reagents were drained and the membrane wrapped in plastic wrap. The membrane then was placed in a film cassette with a sheet of autoradiography film (DuPont). After 30 seconds, the film was removed and developed. An image of the film was digitized and image analysis was performed to estimate band density with NIH Image software. Protein concentration was determined by comparing the absorbance of the MRP sample at 220 nm to an ovalbumin standard curve. Oligosaccharide analysis was perfomed with a DIONEX HPLC fitted with a CarboPac column specifically designed to separate monosaccharides. A 100 11g sample of affinity purified MRP was separated into two aliquots, one was hydrolyzed in 2 M trifluoroacetic acid at l00°C for 4 h and the second in 6 M HCl. The trifluoroacetic acid hydrolysis is for analysis of neutral sugars like mannose and the HCl hydrolysis is for analysis of hexosamines. After cooling, the acid was removed by evaporation and the residue reconstituted in distilled water. The samples then were applied to the column and the monosaccharides eluted isocratically with 14 mM NaOH. Detection of carbohydrates was accomplished with a pulsed amperometric detector with a

100 90

125 ng MRP in single band

80 70

Band Density

60 50 40 30 20 10 0 500

1000

1250

1500

2000

Anti-MAP Dilution Figure 1 Detection of MRP in nitrocellulose membranes using anti-MRR In each lane, 125 ng of purified MRP was loaded, electrophoresed, and transblotted onto a nitrocellulose membrane. The anti-MRP dilutions on the x-axis are ratios of 500 : 1 to 2000 : 1. Band density on they-axis refers to anti-MRP detected as optical density of bands on autoradiography film quantified using image analysis. Vertical bars indicate standard errors.

TERRY W. SNELL and ROBERTO RICO-MARTINEZ

270

monosaccharide detection limit of 10-20 ng. Monosaccharides in our samples were identified by comparison to standards obtained from Sigma Chemical Company. RESULTS To determine the amount of antibody required to detect the MRP, we investigated five dilutions: 1 : 500, 1000, 1250, 1500, and 2000. A sample of 125 ng of purified MRP was western blotted onto a nitrocellulose membrane and detected with ECL. Antibody dilutions of 1 : 1250 or greater were undetectable (Figure 1). A dilution of 1 : 1000 produced a good signal, but a 1 : 500 dilution produced about a 20% stronger signal. Dilutions less than 1 : 500 resulted in non-specific binding. As a result of these experiments, a dilution ofl : 500 was used in all subsequent experiments. The smallest sample of purified MRP detected in western blots with ECL was 25 ng; 12 ng of MRP was not detected. The relationship between the ng of MRP in a band and

120 100

Band Density

80

Y = 28.5 In X- 43.0

60

R

2

= 0.92

40 20 0 0

50

100

150

200

250

ng MRP Figure 2 Binding of anti-MRP to purified MRP in nitrocellulose membranes. The x-axis indicates the ng of purified MRP loaded into each lane of a polyacrylamide gel, electrophoresed, and transblotted onto nitrocellulose membrane. The band density on they-axis is the optical density ofeach band quantified in arbitrary units using image analysis. A digitized image ofeach band can be seen in the lower portion of the figure. 200 ng point not plotted owing to break in band.Vertical bars indicate standard errors.

271

ROTIFER MATE-RECOGNITION PHEROMONE

light intensity from ECL is described by Y =28.5 lnX- 43 with R 2 =0.92 (Figure 2). One explanation for the non-linearity of this relationship is that as binding sites get occupied, steric hindrance limits further binding. Anti-MRP antibody also can be used to detect MRP in crude rotifer homogenates (Figure 3). This relationship is described by Y = 10.4 lnX + 0.19 with R 2 = 0.99. A minimum of 2 flg of rotifer protein is required for detection. Because MRP is 0.06 to 0.14% of total protein (unpublished observations) 2 flg of total protein corresponds to 1-3 ng of MRP. This is about 1/10 the amount of MRP needed for detection in purified samples. Carbohydrate analysis of purified MRP clearly demonstrated the presence of five monosaccharides. The most abundant of these was mannose with 20.6 nM, followed by glucose with 7.4 nM (Figure 4). The coefficients of variation in two replicate samples were 5% and 31%, respectively. N-acetylglucosamine, N-acetylgalactosamine, and fucose also were detected in approximately equal nannomolar quantities of 0.69, 0.32, and 0.44, respectively. These nannomolar estimates were exactly repeated in two replicate samples yielding coefficients of variation of 0. In addition, there was a small peak at 15 minutes suggesting the presence of trace amounts of galactose. The 25.-------------------------+-~

20

Band Density

15

Y

= 10.41n X- 0.19

2 R = 0.99

10

5

0

2

4

6

8

10

ug crude protein Figure 3 Binding of anti-MRP to MRP in crude homogenate of rotifer proteins. The x-axis indicates the Jlg of protein loaded into each lane of a polyacrylamide gel, electrophoresed, and transblotted onto nitrocellulose membrane. The band density on they-axis is the optical density ofeach band quantified in arbitrary units using image analysis.Vertical bars indicate standard errors.

TERRY W. SNELL and ROBERTO RICO-MARTINEZ

272

J

gaiN 11.3

1.0 run

0

uc

fucose 5.5

0

gluN gal glu man 13.4 14.9 16.0 17.4

standards

I

Minutes 0.2-

I

I

5

20

15 20.6 man 17.3

MRP sample 0.44 run fucose 5.5 min

uC 0.1-

0.32 ? 9.6

7.4 g1u 15.9

0.69

gaiN 11.3

0 0

5

10 Minutes

15

20

Figure 4 The monosaccharide composition of the MRP. The x-axis is the minutes of retention time on the column. Commercial standards (upper panel) were injected at 1 nM concentrations and their retention times in minutes are indicated below each monosaccharide label. The y-axis is the current in microcoulombs registered by the amperometric detector. The purified MRP sample is shown in the lower panel with nM estimates above each monsaccharide label and retention times below.

molecular ratios of the these five monosaccharides were 64:23: 2.2:1.4: 1 for mannose, glucose, N-acetylglucosamine, fucose, and N-acetylgalactosamine, respectively.

DISCUSSION The amount ofMRP on a single B. plicatilis female can be estimated. The dry weight of a single, non-ovigerous female is about 0.2 J-Lg, 57% of which is protein (Watanabe et al., 1983). Our calculations indicate that MRP is 0.06 to 0.14% of total protein. Using a middle value of 0.1 %, we estimate that a single B. plicatilis female has about 0.11 ng of MRP, most concentrated in the corona (Snell et al., 1995). This raises the interesting question of what density of MRP is required for male detection. Different populations of a species may differ in the MRP density on females which could lead to male discrimination. Male discrimination of females from other populations has been reported (Snell and Hawkinson, 1983; Rico-Martinez and Snell, 1995). The basis of

ROTIFER MATE-RECOGNITION PHEROMONE

273

male discrimination of females of other rotifer species also could be related to differences in MRP density and location as well as qualitative changes in oligosaccharide and/or protein structure. Patterns of deglycosylation, differential lectin binding, and monosaccharide composition analysis provide insight into the structural features of carbohydrate moeities of glycoproteins (Osawa and Tsuji, 1987). The structure of carbohydrates bound to cell surface glycoproteins are classified into two groups: serine or threonine linked and asparagine linked. The latter group is further divided into high mannose, hybrid, and complex types. The complex types are classified as triantennary or biantennary. Deglycosylation of the MRP by N-glycanase markedly reduced females' ability to elicit mating responses from males (Snell eta/., 1995). This demonstrated that anN-linked oligosaccharide bound to asparagine residues of the MRP is essential for male recognition. The oligosaccharide portion of the MRP can be classified further as a complex type with biantennary structure based on its lectin binding characteristics. Lectins are proteins of non-immune orgin that bind glycoproteins with high specificity (Leiner eta/., 1986; Sharon and Lis, 1989). Of the 10 lectins tested on the MRP, significant binding was detected with only four, all having affinity for terminal mannose and glucose residues (Snell and Nacionales, 1990; Snell et a/., 1993). Later, a lectin with fucose affinity also was found to bind the MRP (Snell eta!., 1995). These characteristics suggest that the oligosaccharide bound to the MRP is an N-linked, biantennnary structure with molecular ratios of 4 : 3 : 2 : 1 of N-acetylglucosamine, mannose, galactose, and fucose, respectively (Osawa and Tsuji, 1987). Our monosaccharide compositional analysis of the MRP demonstrated the presence of these four monosaccharides plus glucose. However, the molecular ratios of the these were 64 : 23 : 2.2 : 1.4 : I for mannose, glucose, N-acetylglucosamine, fucose, and N-acetylgalactosamine, respectively. Compared to the oligosaccharide structure described by Osawa and Tsuji (1987), the MRP oligosaccharide had a great excess of mannose and glucose. Furthermore, on the MRP, galactose is present as N-acetylgalactosamine. This leads to the conclusion that the structure of the MRP oligosaccharide is still poorly understood. Perhaps there are two kinds of oligosaccharides, one a high mannose type and the other biantennary. Resolution of these structural details awaits further analysis. Analysis of rotifer mate recognition pheromones could contribute understanding of the molecular basis of speciation, one of the great unsolved mysteries in evolutionary biology (Coyne, 1992). In this context, it is important to know to how variation in recognition molecules of natural rotifer populations contributes to reproductive isolation. What is the relative importance of the MRP protein and its oligosaccharides in mate recognition? How many genes are involved and can simple genetic changes produce reproductive isolation? Do speciation events leave a distinctive signature on mate recognition systems? The rotifer mate recognition system based on surface glycoproteins seems poised to provide answers to such questions. Acknowledgements

The authors gratefully acknowledge the support of the National Science Foundation (OCE 8600305, OCE 9115860). We thank Dr. Anna Plaas for performing the monosaccharide analysis.

274

TERRY W. SNELL and ROBERTO RICO-MARTINEZ

Reforences Atema, J. (1985) Chemoreception in the sea: Adaptations of chemoreceptors and behavior to aquatic stimulus conditions. Soc. Exp. Bioi. Symp., 39,387-423. Carde R. T. and Baker, T. C. (1984) Sexual communication with pheromones. In: Chemical Ecology of Insects, W. J. Bell and R. T. Carde, eds., Sinauer Associates, Sunderland, MA. CarrW. E. S., Ache B.W. and Gleeson, R. A. (1987) Chemoreceptors of crustaceans: Similarities to receptors for neuroactive substances in internal tissues. Environ. Health Perspect., 71,31-46. Coyne, J. A. (1992) Genetics and speciation. Nature, 355, 511-515. Coyne, J. A., Crittenden, A. P. and Mah, K. (1994) Genetics of pheromonal difference contributing to reproductive isolation in Drosophila. Science, 265, 1461-1464. Dunham, P. J. (1988) Pheromones and behavior in Crustacea. In: Endocrinology of Selected Invertebrate 'Ij!pes., Alan R. Liss, New York. Dusenbery, D. B. and Snell, T. W. (1995) A critical body size for use of pheromones in mate location. J Chern. Ecol., 21:427-438. Foltz, K. R. and Lennarz, W. J. (1993) The molecular basis of sea urchin gamete interactions at the egg plasma membrane. Dev. Bioi., 158, 46-61. Foltz, K. R., Partin, J. S. and Lennarz, W. J. (1993) Sea urchin egg receptor for sperm: sequence similarity of binding domain and hsp 70. Science, 259, 1421-1425. Gao, B., Klein, L. E., Britten, R. J. and Davidson, E. H. (1986) Sequence ofmRNAcodingforbindin, a speciesspecific sea urchin sperm protein required for fertilization. Proc. Nat!. Acad. Sci., USA, 83, 8634-8638. Glabe, C. G. and Clark, D. (1991) The sequence ofthe Arbacia punctulata bin din eDNA and implications for the structural basis of species-specific sperm adhesion and fertilization. Dev. Bioi., 143,282-288. Gleeson, R. A. (1991) Intrinsic factors mediating pheromone communication in the blue crab, Callinectes sapidus. In: Crustacean Sexual Biology, R. T. Bauer and J. W. Martin, eds., Columbia University Press, New York. Knowlton, N. (1993) Sibling species in the sea. Ann. Rev. Ecol. Syst., 24, 189-216. Kubota T., Tokoroyama T., Tsukuda Y, Koiama H., and Miyake, A. (1973) Isolation and structure determination ofblepharismin, a conjugation initiating gamone in the ciliate Blepharisma. Science, 179, 400-402. Larsson P. and Dodson, S. (1993) Chemical communication in planktonic animals. Arch. Hydrobiol., 129, 129-155. Lee, Y H. and Vacquier, V D. (1992) The divergence of species specific abalone sperm lysins is promoted by positive Darwinian selection. Bioi. Bull., 182, 97-104. Liener I. E., Sharon N., and Goldstein, I. J. (1986) The Lectins., Academic Press, New York. Luporini P. and Miceli, C. (1986) Mating Pheromones. In: The Molecular Biology ofCiliated Protozoa, J. G. Gall, ed., Academic Press, New York. Miceli, C., LaTerza, A., Bradshaw, R. A. and Luporini, P. (1992) Identification and structural characterization of eDNA clone encoding a membrane bound form of the polypeptide pheromone Er-1 in the ciliate protozoan Euplotes raikovi. Proc. Nat/. Acad. Sci., USA, 89, 1988-1992. Minor, J. E., Britten, R. J. and Davidson, E. H. (1993) Species-specific inhibition of fertilization by a peptide derived from the sperm protein bindin. Mol. Bioi. Cell, 4, 375-387. Miyake, A. and Beyer, J. (1974) Blepharmone: A conjugation-inducing glycoprotein in the ciliate Blepharisma. Science, 185, 621-623. Osawa, T. and Tsuji, T. (1987) Fractionation and structural assessment of oligosaccharides and glycopeptides by use of immobilized lectins. Ann. Rev. Biochem., 56, 21-42. Palumbi, S. R. (1994) Genetic divergence, reproductive isolation, and marine speciation. Ann. Rev. Ecol. Syst., 25,547-572. Rico-Martinez, R. and Snell. T. W. (1995) Male discrimination of female Brachionus plicatilis and Brachionus rotundiformisTschugunoff(Rotifera). J Exp. Mar. Bioi. Ecol., in press. Sharon, N., and Lis, H. (1989) Lectins as cell recognition molecules. Science, 246,227-234. Shaw, A., McRee, D. E.,Vacquier,V D. and Stout, C. D.1993. The crystal structure oflysin, a fertilization protein. Science, 262,1864-1867. Snell, T.W. and Hawkinson, C. A. (1983) Behavioral reproductive isolation among populations of the rotifer Brachionus plicatilis. Evolution, 37, 1294-1305. Snell, T. W., Morris, P. D. and Cecchine, G. A. (1993) Localization of the mate recognition pheromone in Brachionus plicatilis (0. F. Muller) (Rotifera) by fluorescent labeling with lectins. J Exp. Mar. Bioi. Ecol., 165, 225-235. Snell, T. W. and Nacionales, M. A. (1990) Sex pheromone communication in Brachionus plicatilis (Rotifera). Comp. Biochem. Physiol., 97A, 211-216.

ROTIFER MATE-RECOGNITION PHEROMONE

275

Snell, T. W, Rico-Martinez, R., Kelly, L. S. and Battle, T. E. (1995) Identification of a sex pheromone from a rotifer. Marine Biology, in press. Vacquier,V. D., Carner, K. R. and Stout, C. D. (1990) Species specific sequences of abalone lysin, the sperm protein that creates a hole in the egg envelope. Proc. Nat/. Acad. Sci., USA, 'ifl, 5792~5796. Watanabe, T., Kitajima, C. and Fujita, S. (1983) Nutritional values oflive organisms used in Japan for mass propagation of fish. A review. Aquaculture, 34, 115~ 143. Wiese, L. (1965) On sexual agglutination and mating type substances (gamones) in isogamous heterophallic Chlamydamondes. I. Evidence of the identity of the gamone with the surface components responsible for sexual flagellar content. J Phycol., 1, 46-54.

LECTIN BINDING TO SURFACE GLYCOPROTEINS ON COULLANA SPP. (COPEPODA: HARPACTICOIDA) CAN INHIBIT MATE GUARDING D. J. LONSDALE, T. W SNELL1 and M. A. FREY Marine Sciences Research Center, State University of New York at Stony Brook, Stony Brook, NY 11794-5000 1 School of Biology,

Georgia In~titute of Technology, Atlanta, GA 30332-0230

We tested the hypothesis that surface glycoproteins found on Coullana spp. are important signals in contact mate-recognition. Female copepodites (mostly molt-stage V) of Coullana canadensis (Maryland) and Coullana sp. (Florida) were treated with 0.1 mg ml- 1 of four lectins that represent a variety of carbohydrate affinities. The females were then washed and exposed to males. Binding of some lectins significantly reduced the ability of males to recognize potential mates and initiate precopulatory mate guarding. Other lectin treatments had no significant effect on this behavior. These data show that surface glycoproteins on female Coullana spp. are important mating signals for males in the recognition of conspecifics. Our results also suggest that differences in chemical signals among these sibling species may have evolved.

INTRODUCTION Mate recognition and reproductive barriers in zooplankton include behavioral, morphological, and biochemical factors (e.g., Katona, 1973; Blades and Youngbluth, 1979; Uchima and Murano, 1988; Snell and Nacionales, 1990; Snell and Morris, 1993; Snell and Carmona, 1994). For example, body size may be an important determinant in reproductive success between male and female copepods. Within a locale, sizeassortative mating was found to be related to successful mating and clutch formation in the calanoid copepod Diaptomus birgei (Grad and Maly, 1988). Latitudinally separated populations or sibling species of copepods frequently exhibit differences in body size, such as found in the harpacticoid copepods of Coullana spp. (Lonsdale and Levinton, 1985). In some cases, it was found that the percentage of mate-guarding behavior by males decreased with increasing distance of Coullana male and female source locales from Florida to Maine (Lonsdale et al., 1988). Smaller males from low latitudes had a lower percentage compared to larger males from high latitude when paired with larger females from high latitudes. However, measures of size-related differences in mateguarding frequency within a Coullana population were not significant. Thus, it seems that other biological cues such as behavioral or chemical cues are important in mate recognition in Coullana spp. Specificity in lectin-carbohydrate binding affinities may be an important mechanism in cell-cell recognition, such as among pathogens and hosts or in cell aggregation and differentiation during development (Sharon and Lis, 1989). Snell and Carmona (1994) screened Coullana sp. originally collected in Florida for 12 lectins representing four major classes of oligosaccharide binding affinities (see Table 1 in Snell and Carmona). In general, no lectin bound to the prosome or fork of the caudal furca of females. Strong, localized binding was found on several urosome sites, the genital pore area, and the margin of the caudal furca (Table 3 in Snell and Carmona). As examples, the 277

278

D. J. LONSDALE, T. W. SNELL and M. A. FREY

lectins Pisum sativum and Triticum vulgaris showed strong binding at all these regions, while no binding was found on the genital pore using Tetragonolobus purpureas. Binding of the lectin Erythrina corallodendron was found on both the urosome and caudal furca, but not measured for the genital pore. Snell and Carmona (1994) also measured lectinbinding affinities in adult males of Coullana sp., and found that binding was similar to females but also intense on the antennae. These results suggest that surface glycoproteins may play a critical role in mate recognition in Coullana, such as found in rotifers (Snell and Nacionales, 1990; Snell eta/., 1993; Snell eta/., 1995). We tested the hypothesis that surface chemicals, i.e. glycoproteins, serve as important contact mate recognition cues in Coullana sp. (Florida) and Coullana canadensis (Maryland). Although these two sibling species have been shown to exhibit some degree of mating compatibility in the laboratory, they are reproductively isolated from one another (Lonsdale eta/., 1988). We also investigated whether there are interspecific differences in the relative importance of various glycoproteins for mate recognition, Table 1 Percent survival ofcontrol (Q and treated (T) copepods of Coullana canadensis and Coul/ana sp. after 48 h for four lectin-binding experiments (PS = Pisum sativum, TV = Triticum vulgaris, TP = Tetragonolobus purpureas, EC = Erythrina corallodendron). Lectin Treatment

c Coullana sp. 'jl C. canadensis 'jl C. canadensis 6

100 100

97

PS T

c

100 100 100

100 100 100

TV T

c

71 100 100

100 100

83

TP T

c

100 100

100 100

95

97

EC

T

100 100 100

which could thereby contribute to prezygotic reproductive isolation among these sibling species. METHODS Field Collection and Laboratory Culture of Copepods Coullana spp. are meroplanktonic, brackish water copepods (Coull, 1972). Planktonic nauplii of Coullana sp. were collected in spring from the Sebastian River on the east coast of Florida, USA, in 1984 (see Lonsdale eta/., 1988) and maintained in culture when these experiments were begun in 1993. Coullana canadensis nauplii were collected in late spring 1994 from an embayment of the Patuxent River, located in Lusby, MD. The laboratory culture of the copepods was the same as previously reported (e.g., Lonsdale and Levinton, 1985). Briefly, copepod cultures were maintained at a water temperature of20-21 oc and salinity of15 %o. Two phytoplankton species, lsochrysis galbana (ISO) and Thalassiosira pseudonana ( 3H), were grown in f/2 culture medium (Guillard, 1975) at 15 %o, and fed to copepods in culture. Approximately once a month, 50% of the water in copepod cultures was replaced with autoclaved seawater taken from Stony Brook Harbor, NY that previously had been filtered through a 0.2 pm cartridge filter and diluted to 15%o with distilled water (AFSW). Cultures of copepods from a particular collection site were mixed monthly to maintain genetic homogeneity among cultures.

MATE-GUARDING IN COULLANA SPP.

279

Lectin-binding Experiments

Copepod pairs exhibiting mate-guarding behavior were isolated from batch culture for lectin-binding experiments. To separate paired males and immature female copepodites (mostly molt stage V), pairs were transferred to wells of a multi-depression dish and prodded with a micropipette. Copepods inadvertantly damaged in this procedure were discarded. For each lectin-binding experiment, 30-40 females and males from these pairs were separated into two 50-ml Stendor dishes filled with AFSW and algae. Additional single males from batch cultures were added to the male dish to total about 60 males. Female copepodites were randomly sorted into a control and treatment group, twice transferring them to dishes containing only AFSW. Based upon findings of Snell and Carmona (1994), we selected four lectins, each having a different carbohydrate-binding specificity, to test our hypothesis that glycoproteins are important for contact mate recognition in Coullana spp; lectins Pisum sativum (Glucose/Mannose group), Triticum vulgaris (N-Acetylglucosamine group), Erythrina cora/lodendron (N-Acetylgalactosamine/Galactose group), and Tetragonolobus purpureas (L-Fucose group). Suspensions of lectin in AFSW were prepared at a concentration of 0.1 mg lectin ml- 1 (after Snell and Carmona, 1994). The lectin was weighed on a CAHN electrobalance. The one exception to this concentration was with the lectin T purpureas in treating female Coullana sp.. We treated these copepodites with a 0.025 mg ml- 1 lectin concentration because the higher concentration resulted in significant mortality of treated females ( < 30% survival after 48 h). Wells of a multi-depression dish were filled with about I ml of the lectin suspension and 3 female copepodites were placed in each well for 7-15 minutes to allow for binding of lectins to surface glycoproteins. Control copepods were similiarly added to wells containing only AFSW. Control and treatment females were then twice transferred to a Stendor dish containing AFSW, the latter to· wash off unbound lectin, and then to a dish containing an algal suspension. The algal suspension was a 1:1 mixture by cell number of I galbana and T pseudonana added to AFSW to reach a concentration of 2.5 x 105 cells ml- 1. Following the wash, control and treatment copepodites, up to 18 per group, were placed individually in wells of a multi-depression dish that contained 1 ml of algal suspension. Two males were immediately placed in each well. The dishes were contained within an airtight, white translucent plastic box. Distilled water in the bottom of the box reduced evaporation from the wells. The boxes were placed in an environmental chamber, in indirect light under a 14 : 10 h light:dark cycle, and at 20-21 oc for Coullana sp. and 15°C for C. canadensis. Lower water temperature was necessary for the latter species because of faster development times, and high molting rates of copepodites within the first 24 h observation period. Molting would result in the elimination of the bound lectins on the female's carapace. Observations on mating postures of males were made at 1, 2, 3, 4, 8, 21, 24, 28, 44, and 48 hours following placement with females. Only those pairs that showed mate-guarding behavior were counted in this analysis. Sometimes males would be observed hanging to the caudal furca or dorsal side of females, but were not considered to be exhibiting mate-guarding behavior as shown in Figure I. During mate guarding, males are positioned on the ventral side of females, near the genital pore. If a female copepodite molted during the 48 h period in either the control or treatment, it was excluded from any further counting. Males were replaced if they died. Copepod survivorship was monitored throughout the experiment to determine if lectin

280

D. J. LONSDALE, T W. SNELL and M. A. FREY

Figure 1 Mate-guarding in Coullana canadensis. The smaller copepod is the male, and is approximately 0.6 mm in length.

treatments had lethal effects. The frequency of pre-guarding contact between the sexes (e.g., male grasping the female caudal furca) was also noted, and served as a measure of sublethal effects oflectin treatment on behavior. Three series of experiments using all four lectins were conducted; lectin treatments of copepodite females of Coullana sp. and Coullana canadensis, and of males of C canadensis. For the latter experiments, males were exposed to lectins as described above, and female copepodites were not exposed. Statistical Tests

The effect of lectin binding to females on mate-guarding behavior of males was tested with a one-tailed paired t-test (Sokal and Rohlf, 1981) of angular transformed proportions during the first 28 h time interval. Our hypothesis was that the frequency of male mate guarding would be higher with control females compared to lectintreated females. Although observations continued for 48 h, we restricted the statistical test to 28 h because it was probable that copepods could produce additional glycoproteins, and that bound lectins were degraded by bacteria. The same statistical approach was also used to test the hypothesis that lectin binding to males negatively impacted mate-guarding behavior. To evaluate if lectin treatment had negative, sublethal effects on behavior, we compared the frequency of pre-guarding contact between males and females in the same manner. Observations on Mate-guarding Duration

To evaluate the likely duration of a mate-guarding pair between the above described time intervals, 18 pairs were removed from Cou/lana canadensis batch cultures and placed in

281

MATE-GUARDING IN COULLANA SPP. -Treatment c::::J Control

60.----------------------------

C. canadensis P. sativum, n.s.

50

100 .---------------------------~

C. canadensis T. vulgaris**

80

40

60

30 40

20

~ 10 0.05).

wells with an algal suspension as in the lectin-binding experiments. Observations were made every 15 mins. for four hours, noting the number of pairs.

RESULTS E.ffocts of Lectin-binding on Female Copepodites

Binding of some lectins to the surface of female Coullana sp. and Coullana canadensis significantly reduced the frequency of mate guarding by males (Figure 2). Treatment of Coullana sp. females with the lectin Pisum sativum resulted in a significant reduction in mate guarding by males compared to control females (p < 0.05). This effect persisted for 48 hours when 42.9% of control females were guarded, as compared to only 7.1% of lectin-treated females. In contrast, treatment of C canadensis females with the lectin P. sativum had no significant effect on male mate guarding (p > 0.05). When females of both species were treated with the lectin Triticum vulgaris, male mate guarding was significantly reduced (p < 0.001 for C canadensis and p < 0.05 for Coullana spJ. The effects of T vulgaris treatment became less apparent for C canadensis at 28 hours, and at 44 hours for Coullana sp.

D. J. LONSDALE, T. W SNELL and M.A. FREY

282 60

Treatment

c:::=J Control

C. canadensis T. purpureas, n.s.

50

(])

40

80

30

60

20

40

10

t>ll

ro ....., p

0

I~

1n In

I

C. canadensis E. corallodendron, n.s.

100

20 0

I~

(])

~ 30,---------------------------~ Coullana sp. T. purpureas, n.s.

30,-----------------------------

Coullana sp.

(])

0..

E. corallodendron, n.s.

20

20

10

10

2

3

4

8

21

24

28

44

48

2

3

4

8

21

24

28

44

48

Hours Figure 3 Effects oflectin binding to females of Coullana sp. and Coullana canadensis on mate-guarding behavior of males using the lectins Tetragonolobus purpureas and Erythrina corallodendron. No significant effects oflectin treatments were found.

Lectin treatments of females using Tetragonolobus purpureas and Erythrina corallodendron had no significant effect on mate guarding by males for either Coullana sp. or C. canadensis (p > 0.05; Figure 3). In both these lectin experiments using Coullana sp., however, the percentage of mate guarding in controls was lower than controls in P sativum and T. vulgaris binding experiments. The highest percentage

of mate guarding in controls ranged between 6.7-14.3% compared to 42.9%, respectively. A difference in time interval for mate guarding to be intiated was noted between C. canadensis and Coullana sp. males. In most experiments, mate-guarding behavior of the latter was not observed in the first four hours of observation. In all experiments, survivorship of treated and control female copepodites was high, ranging from 71-100% over the 48 h duration (Table 1). The frequency of pre-guarding contact was not significantly lower (p > 0.05) with treated female copepodites of either species for the four lectins used (Table 2). Efficts ofLectin-binding on Males

Only treatment with the lectin T. purpureas of C. canadensis males significantly reduced the frequency of mate guarding by males (p < 0.05; Figure 4). The frequency of

283

MATE-GUARDING IN COULLANA SPP.

100

-Treatment c::J Control

100

P. sativum, n.s.

. 30 (.) s::::: 20 Q)

::s

C"

~

LL

10 0 N

'It

(Q

OCI

Cl .....

N .....

'It

.....

(Q

.....

...0 Cl)

::iii:

-

-

40

s::::: Q)

20

C"

10

0~

Tumbling

>. 30

(.)

::s

~

LL

0 N

(Q

OCI

Cl .....

N .....

'It .....

(Q

.....

...0

Cl)

::iii:

--

40

s::::: Q)

20

C"

10

~ 0

'It

Sinking

> . 30 (.)

::s

e LL

0 N

-

(Q

OCI

Cl .....

N .....

'It

.....

(Q

.....

...0 Cl)

::iii:

-

40

s::::: Q)

20

C"

10

0~

'It

Individual Path

>. 30

(.)

::s

~

LL

0 N

'It

(Q

OCI

Cl .....

N .....

'It .....

Swimming Speed (mm/sec)

(Q

.....

~

0

::iii:

Figure 5 Histograms of percent frequency of swimming speeds for A) mating pair displaying "hop sink" pattern, B) mating pair displaying "tumbling" pattern, C) mating pair displaying "sinking" pattern, and D) an individual copepod swimming in a swarm. The trajectories of the mating pairs are shown in Figure 3, and the trajectory of the individual copepod is shown in Figure 4.

COPEPOD MATING IN SWARMS

297

Dioithona oculata may use contact pheromones (see Stadler, 1984 for review) to distinguish virgin females from C5 copepodids and nonvirgin females. This selectivity would imply that the contact pheromone is present in greatest concentration on virgin females. Snell and Carmona (1994) found that Mesocyclops edax, which probably has similar mating behavior to Mesocyclops leuckarti and Dioithona oculata, displayed six types of glycoproteins on the tip of the caudal furcae. Snell et al (1988) have identified a group of glycoproteins that occur on the surface of female rotifers and cause males to initiate mating behavior. The presence of similar glycoproteins on female copepods is circumstantial evidence that glycoproteins may also function as contact pheromones for copepods. For cyclopoid copepods such as Mesocyclops spp. and Dioithona oculata, contact pheromones may exist on the female's fourth legs which are usually the first mating contact. Swarm formation did not depend on the presence of adult males or females. Swarms formed with groups of only adult males, adult females, or copepodid stage (C5). Although copepodid stages 3 and 4 were attracted to the light, their densities were an order of magnitude lower than that of the other groups that swarmed. In field collections of swarms, adults and copepodid stage five are the dominant stages (Ambler eta/., 1991). No dramatic differences were found between swimming behaviors of adult males and females, which is in contrast to other cyclopoid copepod species (Uchima and Hirano, 1988; Williamson, 1991). Although males of other cyclopoid species usually swim faster than females (Williamson, 1991), Dioithona oculata females had faster swimming speeds than the males (Figure 2). The swimming behavior of Dioithona oculata mated pairs was very distinctive: bursts of rapid swimming followed by sinking, presumably while the male was extruding spermatophores to place on the female. Both male and female may be contributing to the high velocities of a mating pair. These high bursts of speed in mating pairs, up to 78 mm sec-1 were comparable to D. oculatds escape velocities from fish. Buskey (unpublished data) found that D. oculata had a mean escape velocity of 101 mm sec- 1 (± 23 S. D., n = 20), and a maximum escape velocity of 154 mm sec- 1 • Although the primary purpose of the mating pair's high swimming speed was presumably to maintain their position in the swarm and thus safety, they may also be able to escape from fish predators. Mating pairs swimming in swarms would be especially vulnerable to fish predation, since the mating pairs are a larger target which moves in a distinctively different swimming pattern. Two other studies on swarming zooplankton suggest that mating may be the primary selective advantage for swarming. Ratzlaff (1974) observed swarms of the cladoceran Moina affinis near the shore of a backwater area of the Susquehanna River. He documented that the swarms contained mostly males and sexually reproducing females in contrast to collections of nearby nonswarming cladocerans which had higher percentages of immatures. Hebert et a/. (1980) observed the freshwater predatory copepod Heterocope septentrionalis swarming above pale substrates in arctic ponds. Since this copepod was the dominant predator in these ponds, swarming was not an adaptation to avoid fish or other predators. Swarming did not help H septentrionalis locate its zooplankton prey, since these smaller animals were not present in the swarms. Hebert eta/. (1980) concluded that mating must be the main advantage for swarming in these ponds. Swarm formation of D. oculata in the mangroves may enhance mating encounters as well as protect animals from predation. The emerging model for swarm formation of Dioithona oculata is a fairly simple case (Buskey et a/., this symposium). Copepodid stage five and adults are attracted

J. W. AMBLER ET AL.

298

independently to a swarm marker, which is a light shaft caused by light filtering through the overhanging mangrove leaves. The light intensity in the light shaft must be above a certain threshold intensity for the animals to swarm. From the present study, we have learned that typical swarm densities only occur when stage five copepodids or adult copepods are present (Table 1). After the swarm has formed, D. oculata males appear to search for mates by briefcouplings ("swings") with other copepods, and presumably use their geniculate antennae to grasp other copepods. The brief couplings suggest that contact pheromones may be necessary to identifY virgin females which were preferentially selected for mating. Acknowledgements

This research was funded by a grants (OCE 9217422 and OCE 9349834) from the Ocean Sciences Division of the National Science Foundation. This study was part of the Caribbean Coral Reef Ecosystem (CCRE) Program of the National Museum of Natural History. We thank Dr. Klaus Ruetzler, CCRE Director, and Dr. Frank Ferrari, Department of Invertebrate Zoology at the Smithsonian Institution, for permission to use the Carrie Bow Cay Laboratory and for assisting us in arrangements. *We thank Dr. James Parks (Millersville University) for the loan of his JVC videorecorder for this work. This paper is Caribbean Coral Reef Ecosystem Program contribution 453, and University of Texas Marine Science Institute contribution 954. *We thank Dr. Guy Steucek (Millersville University) for the loan of his LICOR quantum meter and introduction to the Sigma Scan image analysis software. References Ambler, J. W., Ferrari, F. D. and Fornshell, J. A. (1991) Population structure and swarm formation of the cyclopoid copepod Dioithona oculata near mangrove cays. J Plank. Res., 13, 1257-1272. Broadwater, S. A. (1995) Mating behavior of Tropocyclops prasinus (Copepoda: Cyclopoida). Departmental Honors Thesis, Millersville University, Millersville, PA. Buskey, E. J. (1984) Swimming pattern as an indicator of the roles ofcopepod sensory systems in the recognition of food. Mar. Bioi., 79,165-175. Buskey, E. J., Peterson, J. 0. and Ambler, J. W. The role of photoreception in the swarming behavior of the copepod Dioithona oculata. (this symposium) Carde, R. T. and Baker, T. C. (1984) Sexual communication with pheromones. In: Chemical Ecology of Insects, W. J. Bell and R. T. Carde, eds., Chapman and Hall Ltd. Dunham, P. J. (1978) Sex pheromones in crustacea. Bioi. Rev., 53, 555-583. Emery, A. (1968) Preliminary observations on coral reef plankton. Limnol. Oceanogr., 13, 293-303. Ferrari, F. D. (1977) A redescription of Oithona dissimilis Lindberg 1940 with a comparison to Oithona hebes Giesbrecht 1891 (Crustacea: Copepoda: Cyclopoida). Proc. Bioi. Soc. Wash., 90,400--411. Gophen, M. (1979) Mating process in Mesocyclops leuckarti (Crustacea: Copepoda). Israel J Zoo/., 28, 163-166. Griffiths, A. M. and Frost, B. W. (1976) Chemical communication in the marine planktonic copepods Calanus pacificus and Pseudocalanus sp. Crustaceana, 30, 1-8. Hamner, W. M. and Carleton, J. (1979) Copepod swarms: attributes and role in coral reef ecosystems. Limnol. Oceanogr., 24,1-14. Hebert, P. D. N., Good, A. G. and Mort, M. A. (1980) Induced swarming in the predatory copepod Heterocope septentrionalis. Limnol. Oceanogr., 25, 747-750. Hill, L. L. and Coker, R. E. (1930) Observations on mating habits of Cyclops. J Elisha Mitchell sci. Soc., 45,

206--220.

Jacoby, C. A. and Youngbluth, M. J. (1983) Mating behavior in three species of Pseudodiaptomus (Copepoda: Calanoida). Mar. Bioi., 76,77-86. Katona, S. K. (1973) Evidence for sex pheromones in planktonic copepods. Limnol. Oceanogr., 18, 574--583.

COPEPOD MATING IN SWARMS

299

Lazzaretto, I., Salvato, B. and Libertini, A. (1990) Evidence of chemical signaling in Tigriopus fulvus (Copepoda, Harpacticoida). Crustaceana, 59,171-179. Nishida, S. (1986) Structure and function of the cephalosome-flap organ in the family Oithonidae (Copepoda, Cyclopoida). In: Proc. Second International Cmiference on Copepoda, Ottawa, Canada, 13-17 August 1984, G. Schriever, H. K. Schminke and C. T. Shih, eds. Syllogeus No. 58, National Museum of Natural Sciences, Ottawa, Canada. Ratzlaff,W (1974) Swarming in Moina affinis. Limnol. Oceanogr., 19,993-994. Snell, T. W, Childress, M. J. and Winkler, B. C. (1988) Characteristics of the mate recognition factor in the rotifer Brachionus plicatilis. Comp. Biochem. Physiol., 89 A, 481-485. Snell, T. W and Carmona, M. J. (1994) Surface glycoproteins in copepods: potential signals for mate recognition. Hydrobiologia, 292/293, 255-264. Stadler, E. (1984) Contact chemoreception. In: Chemical Ecology of Insects. W J. Bell and R. T. Carde, eds., Chapman and Hall Ltd. Strickler, J. R. and A. K. Bal. (1973) Setae of the first antennae of the copepod Cyclops scutifer (Sars): their structure and importance. Proc. Nat. Acad. Sci. USA, 70, 2656-2659. Ueda, H., Kuwahara, A., Tanaka, M. and Azeta, M. (1983) Underwater observations on copepod swarms in temperate and subtropical waters. Mar. Ecol. Prog. Ser., 11, 165-171. Uchima, M. (1985) Copulation in the marine copepod Oithona davisae Ferrari & Orsi. I. Mate discrimination. Bull Plank. Soc. Japan, 32,23-30. Uchima, M. and Hirano, R. (1988) Swimming behavior of the marine copepod Oithona davisae: internal control and search for environment. Mar. Bioi., 99, 47-56. Uchima, M. and Murano, M. (1988) Mating behavior of the marine copepod Oithona davisae. Marine Biology, 99,39-45. Watras, C. J. (1983) Mate location by diaptomid copepods. J Plank. Res., 5, 417-423. Williamson, C. E. (1991) Copepoda. In: Ecology and Classification of North American Freshwater Invertebrates, Academic Press, Inc.

THE ROLE OF PHOTORECEPTION IN THE SWARMING BEHAVIOR OF THE COPEPOD DIOITHONA OCULATA E. J. BUSKEY, J. 0. PETERSON and J. W. AMBLER The University of Texas at Austin, Marine Science Institute, Port Aransas, Texas, USA (EJB, JOP) and Department of Biology, Millersville University, Millersville, Pennsylvania, USA (JWA) The copepod Dioithona oculata forms dense swarms near mangrove prop roots that are centered around shafts oflight penetrating the mangrove canopy. Swarms can be created in the laboratory within light shafts created with a fiber optic light pipe. Laboratory observations of swarming behavior were recorded using video cameras, and the swimming behavior of the copepods and density of the swarms were quantified using video-computer motion and image analysis techniques. Swarm formation results from a combination of phototactic and klino-kinetic behavior. Dark adapted copepods initially exhibit a photophobic response to a light shaft, but become positively phototactic within 3-5 min after exposure to the light. Copepod aggregation rates under the light fit a saturation model, suggesting that copepods are attracted independently to the swarm marker. Copepods reverse their swimming direction when they encounter light intensity gradients near the edge of a light shaft, which aids in maintaining the swarm. Swarm formation can occur in the laboratory at light intensities as slow as 0.1 J1M photons m - 2 s- 1 , which is similar to light intensities at dawn when they are first observed to form in nature. Swarm formation appears to have an endogenous rhythm, as copepods will not form swarms at night under a light shaft.

INTRODUCTION Copepods have often been reported to form dense aggregations, called swarms, in a wide range of habitats, including temperate and sub-tropical bays (Ueda eta/., 1983), coral reef ecosystems (Emery, 1968; Hamner and Carlton, 1979) and near mangrove cays (Ambler eta/., 1991). Densities in these swarms range from 100-23,000 copepods per liter, and are at least two orders of magnitude higher than densities in water surrounding the swarms (Hamner and Carlton, 1979; Ueda eta/., 1983; Alldredge eta/., 1984; Wishner eta/., 1988; Ambler eta/., 1991). The formation and maintenance of these swarms must include a behavioral component since they are species specific and turbulent diffusion would lead to dispersal (Okubo, 1980). -The cyclopoid copepod Dioithona oculata (Farran) is a common copepod around coral reefs and mangrove cays, and has been reported to form dense swarms in both habitats (Hamner and Carlton, 1979; Ambler eta/., 1991). These swarms can take the form of extensive wedge-shaped bands (0.5 m 2 in cross sectional area) along the edges ofembayments extending for up to 100m in length (Hamner and Carlton, 1979) or smaller spherical or cylindrical swarms tens of em in diameter that form behind coral heads (Hamner and Carlton, 1979) or between mangrove prop roots (Ambler eta/., 1991). How do copepods maintain cohesive swarms when the movement patterns of individuals appears to be random and chaotic? Although the behavioral mechanisms and sensory modalities used in copepod swarm formation have not been studied directly, previous studies provided anecdotal evidence of the visual responses of swarming 301

302

E. J. BUSKEY ETAL.

Dioithona oculata to divers or other objects in the water (Hamner and Carlton, 1979). Vi-

sion in the sense of image formation and perception is probably not possible with the simple photoreceptors of most copepods (Eloffson, 1966; Land 1984). Image formation may not be required to explain the observed swarming behaviors; they could be the results of other photobehaviors such as phototactic or klino-kinetic behaviors. In mangrove cays, D. ocu/ata forms swarms between the prop roots of the red mangrove. These swarms are often associated with shafts oflight that penetrate through the mangrove canopy with swarms appearing at dawn and dispersing at dusk (Ambler eta/., 1991). Swarms of D. oculata also have been reported to form over white-colored substrates and on the downstream side of coral boulders (Hamner and Carlton, 1979). In this study we examine the role of photoreception in the formation and maintenance of swarms of Dioithona oculata collected near mangrove cays of Belize. Copepods were induced to form swarms in the laboratory in response to an artificial light source. Their behavioral responses were videotaped and their movement patterns quantified using a video-computer system for motion analysis. MATERIALS AND METHODS Field work was carried out at the National Museum ofNatural History's Caribbean Coral Reef Ecosystems Program field station on Carrie Bow Cay, Belize. Swarms of Dioithona oculata were collected in the Twin Bays area of Twin Cays and transported back to the field station for study. Swarms were collected in light shafts between the mangrove prop roots while snorkeling along the shore of Twin Bays by enclosing the swarm within a clear plastic bag. Copepods from several swarms were collected and transferred into a large insulated cooler for the short boat trip back to Carrie Bow Cay. Fresh copepods were collected for each day's experimentation. Examination of the swarms showed that all copepods collected were D. oculata and that the swarms were composed primarily of adults with smaller numbers oflate copepodite stages. Experimental studies were carried out in a small room of the field station, that could be darkened during the day, allowing for control oflight intensity. To further control lighting within the experimental area, experiments were performed within a black fabric enclosure that excluded any stray ambient light. Approximately 500 copepods for each experiment were placed within a clear acrylic aquarium (10 x 10 x 15 em) filled with seawater filtered through a 20 J..Lm mesh to remove any other zooplankton. Swarming behavior was induced in the laboratory using a vertical artificial light shaft produced by a fiber optic illuminator (Cole Parmer, 150 watt quartz-halogen lamp) with a 5 mm diameter light pipe. The output of the light pipe was further collumnated as it passed through a black, light-tight box with a 3 x 4 em square opening. This produced a vertical light shaft with a 3 x 4 em cross-sectional area at the water's surface, and a 4.5 x 6 em cross-sectional area at the bottom of the aquarium. Light intensity was regulated by combinations of neutral density filters (Oriel Corporation) placed within the black box. Light intensity was measured with LI-COR model158A photometer equipped with a quantum probe. Copepods were videotaped from the side of the aquarium, viewing copepods in the vertical plane. A Cohu 3315 black and white CCD video camera equipped with a macro lens (Micro-Nikkor 55 mm f 2.8) was focused on an area directly under the fiber optic light pipe over the center of the aquarium, where the center of the swarm would form. The field ofview of the camera (1.8 x 2.2 em) was completely within the area illuminated by the vertical light shaft (except as noted below). Image contrast ofthe copepods under

COPEPOD SWARMING BEHAVIOR

303

low light conditions was enhanced with illumination produced by a ring of infrared light emitting diodes (peak wavelength 890 nm) on the side of the aquarium opposite to the camera, that were arranged to produce dark field illumination. These wavelengths of light should not be perceived by the copepods (Stearns and Forward, 1984; Buskey eta/., 1989).Video records of copepod's swimming behavior were recorded on a Sony FX-710 camcorder, integrated into a field portable video system (Furhman Diversified FieldCam WCMS). Swimming behavior of swarming copepods were initially recorded at the center and edges of the swarm to compare swimming behavior in different regions of the swarm. Since photobehavior appeared to be responsible for swarm formation, experiments also were designed to examine the phototactic and photokinetic behavior of Dioithona oculata. The role of phototaxis in swarm formation was examined by studying the aggregation rate of dark adapted (for 30 min) copepods to a swarm. Swarm formation was recorded on videotape over a period of 10 minutes, and the density of copepods at the center of the swarm was determined at 15 second intervals. The aggregation response was fit to models for logistic and saturation responses (Okubo and Anderson, 1984) using the curve fitting program in the SigmaPlot computer software package. To simulate the photic environment experienced by copepods entering and leaving a light shaft, the response of copepods to changes in light intensity from 50 11M photons m- 2 s- 1 to 5 11M photons m- 2 s- 1 was examined by sliding a 1.0 neutral density filter in front of the light pipe to change light intensity. Copepods were allowed to adapt to the initial light intensity for 10 minutes and then were recorded in the center of the swarm at each light intensity for one minute. Behavioral parameters were averaged over 15 second intervals. Experiments to determine the light intensity needed for swarm formation were performed between 13:00 and 17:00 h. A group of approximately 500 copepods was placed in the plexiglas aquarium, enclosed within the black fabric light shield and held in the dark for 30 minutes. The swimming behavior of the copepods was videotaped in darkness for 5 minutes using clarkfield infrared illumination. The fiber optic illuminator was turned on, and the copepods were videotaped for 10 minutes within the light shaft formed by the fiber optic illuminator. The light was then turned off and the copepods were filmed for 2 minutes in darkness. Light intensities oflO 11M photons m- 2 s- 1 were produced by adjusting the rheostat on the fiber optic illuminator. Subsequent intensities ofl, 0.1, 0.01 and 0.001 pM photons m - 2 s- 1 were produced with combinations of neutral density filters. Each experiment was performed with a fresh group of copepods, and eight experiments were performed at each light intensity. Experiments designed to examine the diel nature of the photobehavior of Dioithona oculata were carried out following a similar procedure to that described above, except that the light intensity used was always 10 11M photons m- 2 s- 1, and the experiment was carried out at several times during the day with the same group of copepods. Photobehavior was recorded in the afternoon (16:00 h), early evening (20:00 h), midnight (24:00 h), near dawn (06:00 h) and morning (10:00 h). Copepods were videotaped for five minutes in darkness, ten minutes in the light shaft and two minutes again in darkness at each time of day. The diel experiments were repeated six times. Swimming behavior of Dioithona oculata was quantified from videotapes upon returning to the University of Texas, using an Expertvision Cell-Trak video-computer motion analysis system. Videotaped experiments were digitized using the Motion Analysis VPllO video-to-digital processor, and digital outlines of the copepods were sent to a host computer at a rate of 15 frames per second. These digitized images were processed to produce records of the swimming speeds (mm s- 1), turning rates (deg s- 1) and net-

304

E. J. BUSKEY ETAL.

I~

Figure 1 Swimming paths of Dioithona ocu/ata at the right edge of a swarm. Shortly after passing from the higher light intensities within the lift shaft to the decreasing light intensities outside the shaft, the copepods make a sharp turn which often directs them back into the swarm. The region of rapidly decreasing light intensity extended from the left margin to the right margin of the figure.

to-gross displacement ratios (NGDR) of the copepod's paths of travel (Buskey, 1984). NGDR is a ratio of the linear distance between starting and ending points of a path (net displacement) and the total distance covered by the path (gross displacement). The density of copepods within the field of view of the video camera was measured using Bioscan Optimas image analysis software. A single frame was digitized at 15 or 30 second intervals, and the number of copepods in the field of view was measured. The depth of field for the macro lens at f 2.8, viewing an area of ca. 4 cm2 , was calculated to be ca. 1 em, and this value was used to estimate the density of copepods in the field of view (copepods m1- 1).

COPEPOD SWARMING BEHAVIOR

305

10

5

8

9

10

Figure 2 Time course for swarm formation by dark adapted (30 min) Dioithona oculata during a ten minute exposure to a light shaft oflO J-!M photons m- 2 s- 1. Copepod densities were measured within the light shaft at 15 second intervals. Each point represents the mean ± 1 SD for 6 replicate trials.

RESULTS During the day, Dioithona oculata were induced to form a swarm within the beam oflight created by a fiber optic light pipe. When this was attempted in a laboratory with normal room lights, additional swarms also formed in one or more of the corners of the aquarium, the copepods probably being attracted to other sources oflight. Best results were achieved in a darkened room where the only source oflight was from the fiber optic light pipe. Under these conditions, a swarm quickly formed which appeared to have its center of density a few em below the water surface and centered within the beam oflight. Most of the copepods remained within the light shaft; copepods swimming out of the light shaft often reversed direction and returned into the swarm (Figure 1). The swarm appears to be maintained in the light shaft by a combination of phototactic behavior and a klino-kinetic response (increase in turning frequency) to changes in light intensity. Dioithona oculata shows a clear positive phototactic response to a vertical beam ofwhite light oflO J.LM photons m- 2 s- 1. Dark adapted copepods exposed to a light beam initially respond with a photophobic response (burst of swimming speed) similar to those observed in other marine copepods (Buskey et al., 1987). During the first two minutes, copepods exhibit negative phototaxis and move to the bottom of the container. After approximately 3 min exposure to the light, copepods switch to a positive phototactic response and begin aggregating within the light shaft. Copepod density within the light shaft increased from< 1 copepod m1- 1 in the dark to approximately 15 copepods m1 - 1 in a period of about 5 minutes (Figure 2). The curve describing the rate of aggregation of D. oculata was found to fit a saturation response model better than a logistic

E. J. BUSKEY ETAL.

306 ~

4

~

I 10

I

i\

en

E 3 E '-"'

0

"'C

CD CD

a.

"-o-o...-o

0\

1

E

.i (/)

50

J'M photons m-2 s -1

y·

0 450

~ ~

I

en

400

g' 350

"'C '-"'

0

()

.I

---·-----

300

0::

250 200

50 p.M photons m-2 s -1

0

I

o

0...-

I I I I 5 J'M photons I 50 p.M photons m-2 s -1 m-2 s -1 I

(/)

·-Ec:

:

o--o-o-..L...o/ ".

I

2

I I

• ~· ./"

.

I .I '.--·-I I I I I I 5 p.M photons I 50 p.M photons I m-2 s -1 m-2 s -1 I I

Time (15 second intervals)

Figure 3 Time course of change in copepod swimming behavior when exposed to a rapid decrease in light intensity from 50 J.LM photons m- 2 s- 1 to 5 J.LM photons m- 2 s- 1 followed by a rapid increase in light back to the original intensity.

response model. The saturation response model suggests that copepods are attracted independently to the light shaft; a fit to the logistic response model would have suggested that aggregation was enhanced by the presence of other swarming individuals (Okubo and Anderson, 1984). Dioithona oculata also shows a klino-kinetic response to rapid decreases in light intensity. As D. oculata passes through the edge of the light shaft, it reverses direction, return-

COPEPOD SWARMING BEHAVIOR

307

CJ Light On

. . Dark

400~----------------------------------~

350 _.. 300 '(/) 250 Ol Q)

"'0 ....._

0

200 150

(.)

a:: 100 50

_.. ~

I

E

(/)

"'0

0

a. Q) a. 0 0 ....._ :?;·a;

0

*

12

*

9 6

*

3

c

Q)

0

0

0.01

0.1

1

10

Light Intensity (J.£ M photons m-2 s-1) Figure 4 Comparison of Dioithona oculata turning rates (top) and density of copepods beneath a fiber optic light source (bottom) for dark adapted copepods exposed to vertical light shafts of various intensities. Black bars represent values for copepods swimming in the dark before the light was turned on; white bars represent values for copepods swimming within the light shaft beginning 5 minutes after the light was turned on. These are the mean values for eight experiments at each intensity; error bars represent one standard deviation. An asterisk indicates a significant difference based on a paired comparison Student's t-test (a = 0.05); ns indicates no significant difference.

308

E. J. BUSKEY ETAL.

-

CJ Light On

Dark

400~-----r-------------------.---------,

350

DAY

NIGHT

DAY

DAY

NIGHT

DAY

*

........... 300 250 200 150 100 50 0 I

E en

12

*

*

-o 0

a. Q)

a.

0

0 ..........

.?;-

·en

c:

Cl)

0

16:00

20:00

24:00

Time of day

06:00

10:00

Figure 5 Comparison of Dioithona oculata turning rates (top) and density within the light shaft (bottom) for dark adapted copepods exposed to light shafts of!O {LM photons m- 2 s- 1 at different times of the day. Black bars represent values for copepods swimming in the dark before the light was turned on; white bars represent values for copepods swimming within the light shaft. These are the mean values for six die! experiments. An asterisk indicates a significant difference based on a paired comparison Student's t-test (a = 0.05); ns indicates no significant difference.

COPEPOD SWARMING BEHAVIOR

309

ing to the light shaft (Figure 1). A similar klino-kinetic response is produced when the entire swarm is exposed to a rapid decrease in light intensity. Decreasing light intensity withinalightshaftfrom 50 J.LM photons m- 2 s- 1 to 5 J.LM photons m- 2 s- 1 causes a transient increase in rate of change of direction in a copepod swarm; the opposite increase in light intensity (increase from 5 to 50 J.LM photons m- 2 s- 1) results in little change in turning rate (Figure 3). When dark adapted copepods are exposed to a light shaft during the daytime hours (13:00-17:00 h), the copepods formed swarms at light intensities of 0.1, 1.0 and 10 J.LM photons m- 2 s- 1, but not at the lowest light intensity of 0.01 pM photons m- 2 s- 1 (Figure 4, Table 1). In comparison to the dark period, the copepods forming the swarm in the light shafts show a significant increase in rate of change of direction, net to gross displacement ratio and in the density of copepods in the area under the light shaft (Student's t-test with paired sample design, a = 0.05; Table 1). Table 1 Swarming behavior of Dioithona oculata exposed to light beams of differing intensity. Behavior of co-

pepods in the dark is reported first, followed by their behavior 5 min after light beams ofvarying intensity were turned on. Each value represents the grand mean from mean parameters from eight replicate experiments. RCD is mean rate of change of direction in degrees s- 1. NGDR is mean net to gross displacement ratio. Density is number of copepods counted in field of view of the video camera, directed below the light pipe. An asterisk (*)indicates significant difference (Student's t-test with paired design, a = 0.05). D/L

Dark Light Dark Light Dark Light Dark Light

Light (JIM m- 2 s- 1)

10

0.1 0.01

Speed (mms- 1)

RCD (o s-1)

NGDR

Density (copml- 1)

2.70 3.18 2.57 2.67 2.54 2.32 2.31 2.19

156 281 * 190 289* 188 228* 208 202

0.86 0.68* 0.80 0.72* 0.83 0.73* 0.80 0.77

0.28 9.81* 0.64 7.96* 0.53 1.93* 0.30 0.51

Dark adapted Dioithona oculata form swarms in response to a 10 J.LM photons m - 2 s- 1 light beam only during daylight hours (Figure 5, Table 2). At 20:00 h, when copepods had been held in the dark for 2 hours, they failed to swarm in the light. A similar lack of response was seen at midnight (24:00 h). However, at 06:00 h, after having been held in the dark for a period of 12 hours, copepods began forming swarms in response to the light beam. The density of the swarm at 06:00 h was ca. one half that formed during daylight hours, but those copepods responding showed the same behavioral changes characteristic of copepods responding to light beams during daylight hours (Table 2). DISCUSSION

The copepod Dioithona oculata forms swarms in the laboratory within the light shafts created by a fiber optic light pipe. The behavioral mechanisms responsible for the copepods aggregating and remaining in this small area appear to be a combination of phototactic behavior and klino-kinetic behavior. The phototactic behavior may serve to both attract the copepods initially to the light shaft and to guide them back to the swarm when they have strayed away. A klino-kinetic behavior of increased turning frequency in areas

E. J. BUSKEY ETAL.

310

of decreasing light intensity (Figure 3) may further help them remain within the light shaft. These behavioral mechanisms help explain the frequently observed swarms of D. ocu/ata that form within shafts oflight in the mangrove cays of Belize (Ambler eta/., 1991), but they do not completely explain the swarms of this same species that have been observed to form in open waters around coral reefs (Hamner and Carlton, 1979). These swarms have been reported to form over white areas of the bottom; the reflective nature of the bottom substrate might also create a photic environment suitable for swarm formation. We have observed swarm formation over white plastic circles placed on the bottom of an aquarium with a black bottom and black sides. Hebert et al (1980) report that the freshwater, predatory calanoid copepod Heterocope septentrionalis forms aggregations over areas of pale substrate, and forms swarms over white or yellow plastic panels placed on the bottom of ponds. Swarms have also been reported to form in the lee of objects rising above the bottom, where chances of being dispersed by currents are lower (Kimoto et a/., 1988). It is possible that other senses besides photoreception may also contribute to swarm formation and maintenance, but these have not yet been investigated. Table 2 Swarming behavior of Dioithonaoculata exposed to light beams of!O J.LM photons m- 2 s- 1 at different times of the day. Behavior of copepods in the dark is reported first, followed by their behavior 5 min after light beams were turned on. Each value represents the grand mean from mean parameters from six replicate experiments. RCD is mean rate of change ofdirection in degrees s- 1. NGDR is mean net to gross displacement ratio. Density is number of copepods counted in field of view of the video camera, directly below the light pipe. An asterisk(*) indicate significant difference (Student's t-test with paired design, a = 0.05). Time

D/L

Speed (mms- 1)

RCD (o s-1)

NGDR

Density (copm1- 1)

16:00h

Dark Light Dark Light

3.70 3.46 3.09 3.50 3.10 3.31 2.72 3.02 2.82 3.29

145 291* 161 173 210 196 187 260* 197 308*

0.89 0.70* 0.86 0.82 0.84 0.81 0.85 0.73* 0.81 0.68*

0.11 11.22* 0.36 1.03 0.48 0.64 0.25 3.67* 0.54 9.97*

20:00h 24:00h 06:00h lO:OOh

Dark

Light Dark Light Dark Light

Swarm formation was demonstrated to occur in response to light intensities as low as 0.1 jiM photons m - 2 s- 1. This is lower than the minimum observed ambient light intensity (1.1 jiM photons m- 2 s- 1) at which swarms are reported to form at dawn in the field (Ambler eta/., 1991). A larger proportion of copepods responded to light intensities of I jiM photons m- 2 s- 1 than at 0.1 (Table 1), so swarms that are dense enough to observe in the field may not form until this intensity is reached. It is unclear from this study why swarms fail to form at lower light intensities. The minimum photosensitivity threshold of Dioithonaoculatato white light has been calculated to be 1.4 x w- 4 p,M photons m- 2 s- 1 (Ambler and Buskey, unpublished), which is comparable to photosensitivity thresholds calculated for other species of copepods; minimum photosensitivity is usually in the range of 1-100 x w- 6 p,M photons m- 2 s- 1 (Swift and Forward, 1983; Stearns and Forward, 1984; Buskey eta!., 1989). Photosensitivity of D. oculata is several orders of mag-

COPEPOD SWARMING BEHAVIOR

311

nitude greater than the threshold we observed for swarm formation. It is not unreasonable that light intensities required for swarm formation may be higher than those for minimum visual perception; Cronin (1986) hypothesized that light dependent behaviors such as setting circadian rhythms may also have thresholds higher than for visual perception. Swarms of Dioithona oculata have been observed to form at dawn and disperse at dusk (Ambler eta/.. , 1991) providing additional evidence of the role oflight in swarm formation. In this study we observed a diel pattern in photobehavior; copepods that had been held in the dark for two hours or more would not form swarms during the night at 20:00 or 24:00 h. However, copepods that had been held in the dark for 12 hours would form swarms at 06:00 h, approximately the time of sunrise. The normal diellight cycle may act as a releasor stimulus that initiates the behavior pattern so that it occurs at an appropriate time and place. Based on these results, it would be predicted that copepods in the field would not respond to light stimuli at night. On three occasions we visited the areas adjacent to the mangrove prop roots at night, and tried to capture small fish by attracting them to the beam of a flashlight held in place for 15-20 minutes. On two nights the sky was cloud covered and dark, and no D. oculatawere attracted to the light. However, when the same procedure was tried during a full moon on a clear night, dense aggregations of D. oculata swarmed under the flashlight beam within minutes. The dim light of the full moon may have acted as a premature releasor stimulus, but was of insufficient intensity to cause swarming behavior. We have determined that light intensity, gradients oflight intensity and diel rhythms serve as proximal cues for swarm formation. The adaptive significance of swarm formation includes increased encounter with potential mates, protection from predators and reduced dispersion by currents (Hamner and Carlton, 1979). Swarming Dioithona oculata may experience an increased frequency of mating encounters, as has been suggested for swarm-forming freshwater copepods (Hebert et al., 1980). Many insect taxa, notably mayflies, midges and oceanic water striders form mating swarms (Peckarsky eta/., 1990). Ratzlaff (1974) hypothesized a mating function for cladoceran swarms of Moina a./finis, since these swarms included almost exclusively adult males and females. During the diel studies, mating pairs of D. oculata were only observed during swarm formation at 06:00 hours, corresponding to the time of natural dawn. The initial formation of swarms at dawn may provide the first encounters of the day between adult copepods that were dispersed throughout the adjacent bays during the night. On average, 64% of swarming D. oculata are adults, and 62% of these are female (Ambler eta/., 1991). Swarming in zooplankton may also protect mating pairs from higher predation, since mating pairs would present a larger target for predators than would individuals. Jakobsen and Johnsen (1988) showed experimentally that sticklebacks selected larger Bosmina longispina when individuals were dispersed than when the cladocerans formed swarms. By swarming, Dioithona oculata may reduce its chances of being encountered by planktivores than if they were evenly dispersed. This seems to be particularly true in the mangrove prop root habitat where large schools of small planktivorous fish (4nchoa, Harengula, Jenkisia) are often found at the edge of the mangrove habitat, but not between the prop roots where D. oculata swarms form. D. oculata can maintain its position in the mangrove prop roots in tidal currents of up to 2 em s- 1 (Buskey eta/., in prep); swarm formation in light shafts may be a means of using a visual marker to maintain their position within the prop roots and prevent them from being washed out into the dense schools of planktivores. Once a visual predator encounters a swarm, it may be difficult for the predator to select an individual to attack (Milinski, 1979; Jakobsen and

E. J. BUSKEY ETAL.

312

Johnsen, 1988). As the benefits of reduced predation risk increase with increasing swarm size, the costs arising from increased competition for food also increases, however (Bertram, 1978). The diurnal pattern of swarm formation during the day, when risk of visual predation is high, and dispersal at night may represent a trade-off between reducing risk of predation and maximizing feeding rate. Acknowledgements

This study was funded by a grant from the National Science Foundation (OCE 9218516). This study was part of the Caribbean Coral Reef Ecosystem Program of the National Museum of Natural History. We thank Dr. Klaus Reutzler, CCRE program director and Dr. Frank Ferrari, both of the Smithsonian Institution, for allowing us to use the facilities at Carrie Bow Cay and for assisting with arrangements. This is University of Texas Marine Science Institute contribution number 938 and Caribbean Coral Reef Ecosystem Program contribution number 437. References Alldredge, A., Robison, B., Fleminger, A., Torres, J., King, J. and Hamner,W (1984) Direct sampling and in situ observation of a persistent copepod aggregation in the mesopelagic zone ofthe Santa Barbara Basin. Mar. Bioi., 80,75-81. Ambler, J. W, Ferrari, F. D. and Fornshell, J. A. (1991) Population structure and swarm formation of the cyclopoid copepod Dioithona oculata near mangrove cays. l Plank. Res., 13, 1257-1272. Bertram, B. C. R. (1978) Living in groups: predators and prey. In Behavioral Ecology, J. R. Krebs and N. B. Davies, eds., Blackwell Scientific Publications, Oxford, pp. 64--96. Buskey, E. J. (1984) Swimming pattern as an indicator of the roles ofcopepod sensory systems in the recognition of food. Mar. Bioi., 79,165-175. Buskey, E. J., Baker, K. S., Smith, R. C. and Swift, E. (1989) Photosensitivity of the oceanic copepods Pleuromamma gracilis and Pleuromamma xiphias and its relationship to light penetration and daytime depth distribution. Mar. Ecol. Prog. Ser., 55, 207-216. Buskey, E. J., Mann, C. and Swift, E. (1987) Photophobic responses of calanoid copepods: possible adaptive value.l Plank. Res., 9,857-870. Cronin, TW (1986) Photoreception in marine invertebrates. Amer. Zoo!., 26,403--415. Eloffson, R. (1966) The nauplius eye and frontal organs of the non-malacostraca (Crustacea). Sarsia 25, 1-128. Emery, A. (1968) Preliminary observations on coral reef plankton. Limnol. Oceanogr., 13, 293-303. Hamner, W and Carlton, J. (1979) Copepod swarms: attributes and role in coral reef ecosystems. Limnol. Oceanogr., 24,1-14. Hebert, P. D. N., Good, A. G. and Mort, M. A. (1980) Induced swarming in the predatory copepod Heterocope septentrionalis. Limnol. Oceanogr., 25,747-750. Jakobsen, P. J. and Johnsen, G. H. (1988) Size specific protection against predation by fish in swarming waterfleas, Bosmina longispina. Anim. Behav., 36, 986-990. Kimoto, K., Nakashima, J. and Morioka, Y (1988) Direct observations of a copepod swarm in a small inlet of Kyushu, Japan. Bull Selkai Reg. Fish. Res. Lab., 66,41-58. Land, M. F. (1984) Crustacea. In Photoreception and vision in invertebrates, M. A. Ali, ed., Plenum Press, New York, pp. 401--438. Milinski, M. (1979) Can an experienced predator overcome the confusion of swarming prey more easily? Anim. Behav., 27,1122-1126. Okubo, A. (1980) Diffusion and ecological problems: mathematical models. Springer-Verlag, Berlin. Okubo, A. and Anderson, J. J. (1984) Mathematical models for zooplankton swarms: their formation and maintenance. EOS 65, 731-732. Peckarsky, B. L., Fraissinet, P.R., Penton, M.A. and Conklin, D. J. Jr. (1990) Freshwater macroinvertebrates of northeastern North America. Comstock Pub. Assoc. of Cornell University Press, Ithaca. 442 pp. Ratzlaff,W (1974) Swarming in Moina affinis. Limnol. Oceanogr., 19,993-995. Stearns, D. E. and Forward, R. B., Jr. (1984) Photosensitivity of the calanoid copepod Acartia tonsa. Mar. Bioi.,

82,85-89.

COPEPOD SWARMING BEHAVIOR

313

Swift, M. C and Forward, R R, Jr. (1983) Photoresponses ofthe copepod Mesocyclopsedax. J. Plank Res., 5, 407-415. Ueda, H., Kuwahara, A., Tanaka, M. and Azeta, M. (1983) Underwater observations on copepod swarms in temperate and subtropical waters. Mar. Ecol. Prog. Ser., 11, 165-171. Wishner, K., Durbin, E., Durbin, A., Macauley, M.,Winn, H., and Kenney, R. (1988) Copepod patches and right whales in the Great South Channel off New England. Bull. Mar. Sci., 43, 825-844.

EFFECTS OF SEDIMENT LOADING ON FOOD PERCEPTION AND INGESTION BY FRESHWATER COPEPODS NANCY M. BUTLER Flathead Lake Biological Station, The University of Montana, [email protected] Seasonal influx of suspended sediments to lakes is common in many aquatic systems, particularly during periods of high flow. Suspended particles not only affect visibility and net primary productivity, as light penetration is reduced, but may also affect the ability of herbivorous zooplankton to locate and ingest food. In Swan Lake, an oligotrophic lake in northwestern Montana, the close association between zooplankton population development and seasonal maxima in turbidity levels suggests that the community present during peak turbidity is minimally impacted by suspended sediments. This report presents the results oflaboratory investigations into the effect of suspended sediments on feeding by the copepod Diaptomus ashlandii, the dominant zooplankter in Swan Lake. Ingestion rate for copepods feeding on 32 P labeled algae was significantly reduced in the presence of suspended particles at all turbidity levels tested (5 to 200 NTU's) compared to ingestion rate in the absence of suspended particles (0 NTU's).

INTRODUCTION The oligotrophic, lowland lakes of northwest Montana are typically clear water lakes with seasonal influxes of sediment which result in a visible change in water clarity. In Swan Lake, water clarity is high for a major part of the year, with turbidity levels of approximately one NTU (nephelometric turbidity units). However, during spring runoff, turbidity rapidly increases to more than 20 NTU in the upper water column, with a concurrent decrease in Secchi depth to less than 3 meters (Figure 1). In large part, the elevated suspended sediment loads are a consequence of erosion of naturally unstable banks that line many of the stream channels. However, there is also evidence that sediment loads vary annually in response to land use activity. Lake bottom core samples collected from Swan Lake indicate marked increases in sedimentation rates in the lake basin associated with timber harvest and road building activities and an overall trend of steadily increasing annual deposition rates (Spencer, 1991b). As the area undergoes rapid socio-economic development and upstream land use activities increase, there is growing concern about the potential impact of these increasing sediment loads on the lake ecosystem and water quality. The seasonal influx of suspended sediments (e.g. sands, silts, and clays) can have a substantial effect on the biological, physical, and chemical components of a lake. These effects may be manifest through a variety of direct and indirect influences on community structure and trophic interactions. Reduced phytoplankton density is a common response to increased suspended sediment and can be attributed to a variety of causes. The increase in turbidity causes reduced light penetration through the water column, resulting in a decrease in the euphotic zone (Kirk, 1985; Grobbelaar, 1985, 1992). This reduction in available light results in a subsequent decline in primary production 315

NANCY M. BUTLER

316

A

10

9 8 7

g

6

..c a.

5

Q

4

Cl)

3 2

0

B

June

1-

~

:s...

Oct

Dec

Feb

Apr

June

Aug

Oct

Feb

Apr

June

Aug

Oct

25 ---i:x--- 5 meters

------ 30 meters

20

5

Aug

---a--

Integrated (0-30 m)

15

>

=s

::I

1-

10

5

0

June

Aug

Oct

Dec

Figure 1 Secchi depth (A) and turbidity (B) data during the period of June 1992 to November 1993. Secchi depth measurements were taken in the south basin of Swan Lake and are expressed in meters (m). Turbidity measurements were made on whole water sa"1ples collected in the south basin of Swan Lake at 5 m and 30 m depth and in an integrated 0-30 m sample.

SEDIMENT EFFECTS ON COPEPOD FEEDING

317

(Canfield and Bachman, 1981; Cuker, 1987). Phytoplankton densities can also be reduced due to precipitation of algal cells in clay-algal floes (Avnimelech eta/., 1982). In addition, suspended clays can bond with phosphorous, reducing its availability to the phytoplankton (Ellis and Stanford, 1988; Cuker eta/., 1990). Phytoplankton-herbivore and herbivore-herbivore interactions may change under conditions of increased turbidity (Kirk and Gilbert, 1990; Kirk, 1992; Jack and Gilbert, 1993) and precipitate changes in plankton community structure (Jack et al., 1993). Because copepod ingestion and clearance rates are a function of cell density (Marshall, 1973; Williamson and Butler, 1986), the decrease in primary production and cell density that result from increased turbidity may lead to decreased feeding efficiency. In addition, many zooplankton utilize a combination of chemical and mechanical sensory cues to locate food items and distinguish between potential food items (DeMott, 1986; Butler eta/., 1989; Starkweather, this symposium; Lenz eta/., this symposium). These food detection, location, and capture mechanisms may be hindered by the presence of high densities ofsuspended inorganic particles, further increasing the potential for decreased feeding efficiency under turbid conditions. In addition to affecting the feeding efficiency of grazing zooplankton, the efficiency of visual predators (e.g. fish) decreases with the decrease in light penetration in the presence of suspended sediments. Consequently, zooplankton prey may be released from predation pressure under conditions of increased turbidity. This benefit, however, may be offset if predation pressure from non-visual predators (e.g. Chaoborus) increases as these populations also are released from predation by visual predators, as demonstrated by Cuker (1993). While the above suggests an overall negative effect on zooplankton communities, there is evidence that those effects can be modified when the zooplankton are conditioned to the presence of suspended sediments. When maintained under conditions of high turbidity, cladoceran and copepod species from naturally turbid waters have enhanced development times compared to species from sediment-free waters and may even be unable to develop in the complete absence of suspended sediments (Koenings eta/., 1990; Hart, 1991, 1992). These observations suggest that zooplankton from naturally turbid systems may develop adaptive responses appropriate to living under turbid conditions. The zooplankton community in Swan Lake initiates annual population increase coincident with the spring influx of sediment, reaching peak numbers by late summer (Spencer, 1991a). The close link between initiation of zooplankton community development and the spring sediment pulse suggests the zooplankton are not negatively impacted by the turbid conditions. Because copepod reproduction is strongly influenced by food availability (Williamson et a/., 1985; Williamson and Butler, 1987), the initiation of reproduction and offspring development suggests feeding efficiency is not severely affected during turbid conditions. It is possible that the zooplankton community is adapted to the physical presence of suspended particles, as suggested by researchers studying zooplankton from naturally turbid lakes. However, unlike systems studied by other researchers, Swan Lake is only seasonally turbid, with the water relatively clear for the majority of the year, and the turbidity levels, even during the seasonal peak, are far below those typically studied. Here I present results of research investigating the feeding behavior of the calanoid copepod, Diaptomus ashlandii, the dominant zooplankter in Swan Lake, in response to natural turbidity levels.

318

NANCY M. BUTLER

METHODS The organisms used in the study were the copepod Diaptomus ashlandii, the dominant pelagic copepod species in Swan Lake, and the flagellate alga Chlamydomonas reinhardtii (University of Texas, Starr Collection Strain UTEX 90). The alga was maintained in exponential growth phase via semi-continuous batch cultures with heat sterilized, modified MBL medium (Sternberger, 1981) at 15°C and constant light. To determine cell ingestion rates, copepods were fed 32 P0 4 1abeled C reinhardtii and ingestion determined as a function of label uptake by the grazing copepods. Cell labeling was based upon the technique presented in Butler et al. (1989). Twenty-four hours before each feeding trial, the medium was labeled using 32 P0 4 filtered through a 0.2 J.Lffi Nuclepore filter to remove labeled particulate matter. The culture was split, with half receiving 32 P04 labeled medium to produce 32 P04 labeled cells with an activity of 1.76 counts-cell. Immediately after each feeding trail, algal radioactivity was determined by filtering three 5 ml aliquots of the algal culture (ofknown cell density) onto 0.2 J.Lffi Nuclepore filters; the filters were transferred to scintillation vials with 20 ml DHOH and radioactivity counted using a Beckman LS-6500 scintillation counter. Turbidity was manipulated using Na-Montmorillonite obtained from the Source Clay Minerals Repository, University of Missouri (Stock SWy-2). Turbidity levels were quantified as nephelometric turbidity units (NTU), using a Hach Model 2100A Turbidimeter. A clay slurry was prepared by combining 2.5 g clay in 500 ml DHOH and distributed in aliquots among the experimental containers to yield turbidity levels ofO, 5, 10, 20, 50, and 200 NTU These levels include the natural turbidity range in Swan Lake as well as the higher turbidity levels typically used by other researchers. Clay particle sizes ranged from 1 to 12 J.Lm, with an average particle diameter of 3.5 J.tm. Average cell size of C reinhardtii was 6.8 J.tm, with cell diameters ranging from 6.0 to 9.0 J.tm. The ratio of clay particles to algal cells ranged from approximately 1 : 1 in the 5 NTU treatment to approximately 50 : 1 in the 200 NTU treatment. Diaptomus ashlandii adults were sorted from plankton tows collected from Swan Lake, maintained in filtered lake water at 15°C, and pre-fed for 24 hours on high densities (10 5 cells·ml- 1) ofunlabeled C reinhardtii. Copepods were isolated from the conditioning treatment, sorted according to sex, and randomly allocated to 250 ml narrow-necked Wheaton bottles containing filtered (0.7 J.tm GF/F filters) lake water. Each bottle received either 10 gravid females or 10 males. Aliquots of clay slurry were added to the bottles to yield desired turbidity levels. Labeled algae were added to yield a cell concentration of 104 cells·ml- 1 in each bottle. Immediately upon addition of the labelled alga, each bottle was topped off with filtered lake water (to remove air bubbles), capped, and gently rotated three times (to ensure even mixing of the alga) and timing begun. Feeding trials were conducted at 15°C in low light conditions. Tests were terminated after 15 min. by pouring the contents of the bottles through 300 J.tm Nytex screening, rinsing the bottles and the copepods three times with filtered lake water, and transferring individual copepods to scintillation vials with 20 ml DHOH. No mortality was observed during the feeding trials. Activity of the samples was counted, as described above, and cell ingestion rates determined by dividing the amount oflabel incorporated by the copepods by the radioactivity per algal cell. To determine if changes in feeding rate were a consequence of indiscriminate feeding in the presence of suspended clay particles, a second set of experiments was designed to determine clay ingestion rates. Copepods were transferred to experimental bottles with unlabeled algae and turbidity levels as described above. The bottles were

SEDIMENT EFFECTS ON COPEPOD FEEDING

-o-

c 800 E 700

w

l-It)

600

a:::::

500

20 NTU) were reached. The decrease in feeding rate could not be attributed to gross behavioral changes. Observation of swimming copepods did not indicate any readily discernible change in swimming or search behaviors associated with changes in turbidity. The copepods did not seem more or less active in any particular treatment, nor did there seem to be any difference in the time spent in the different swimming modes. It is possible that the copepods were adjusting their behavior on a more subtle scale. If the copepods switch from selective ingestion of algal cells to indiscriminate feeding of suspended particles (clay and algae mixed), one would expect an increase in the amount of clay ingested with increased turbidity increased. However, analysis of clay content in the gut after feeding for one hour did not indicate a significant pattern of sediment ingestion associated with turbidity for either sex, although clay was ingested by both sexes while feeding at all turbidity levels. In the presence of suspended inorganic particles, Diaptomus ashlandii appears capable of discriminating between algal cells and clay particles when feeding under turbid conditions, as evidenced by the only slight change in ingestion rate as turbidity increases with no apparent increase in sediment ingestion. However, the presence of sediments has a marked effect on feeding efficiency of female copepods, as evidenced by the sharp decline in ingestion at 5 NTU compared to the 0 NTU treatment. These results suggest that feeding on phytoplankton by D. ashlandii in Swan Lake would be reduced during periods of high turbidity in the lake. However, the rapid population

SEDIMENT EFFECTS ON COPEPOD FEEDING

321

growth during that period suggests the population is not food limited. It is possible that the copepods in Swan Lake are ingesting clay in addition to algae under turbid conditions, as suggested by this study, but derive a nutritional benefit from the clay entering the lake. Ellis and Stanford (1982, 1988) demonstrated that river transported sediments can develop a microbial fauna. Thus the suspended sediments in Swan Lake during the spring turbidity plume might actually provide an adequate food source in the form of a microbe/clay aggregation and the nutritional benefit gained from ingesting these particles may exceed the "cost" of decreased ingestion of phytoplankton. Acknowledgments

This project was supported by Montana State University MONTS grant #190136. All field work was conducted with the able assistance of James A. Craft. The original manuscript was substantially improved by the comments of two anonymous reviewers. References Avnime1ech, Y, Troeger, B. W, and Reed, L. W (1982) Mutual flocculation of algae and clay: evidence and implications. Science 216, 63-65. Butler, N. M., Suttle, C. A., and Neill,W E. (1989) Discrimination by freshwater zooplankton between single cells differing in nutritional status. Oecologia 78, 368-372. Canfield, D. E. and Bachmann, R. W (1981) Prediction of total phosphorous concentrations, chlorophyll a and Secchi depths in natural and artificial lakes. Can. J. Fish. Aquat. Sci. 38,414-423. Cuker, B. E. (1987) Field experiment on the influences of suspended clay and P on the plankton of a small lake. Limnol. Oceanogr. 32, 840-847. Cuker, B. E. (1993) Suspended clays alter trophic interactions in the plankton. Ecology, 74, 944-953. Cuker, B. E., Gama, P. T., and Burkholder, J. M. (1990) Type of suspended clay influences lake productivity and phytoplankton community response to phosphorous loading. Limnol. Oceanogr. 35, 822-830. DeMott, W R. (1986) The role of taste in food selection by freshwater zooplankton. Oecologia (Bed) 69, 334-340. Ellis, B. K. and Stanford, J. A. (1982) Comparative phytoheterotrophy, chemoheterotrophy, and photolithotrophy in a eutrophic reservoir and an oligotrophic lake. Limnol. Oceanogr. 27, 440-454. Ellis, B. K. and Stanford, J. A. (1988) Phosphorous bioavailability of fluvial sediments determined by algal assays. Hydrobiologia 160, 9-18. Grobbelaar, J. U (1985) Phytoplankton productivity in turbid waters. J. Plankton Res. 7, 653-663. Grobbelaar, J. U (1992) Nutrients versus physical factors in determining the primary productivity of waters with high inorganic turbidity, Hydrobiologia 238, 177-182. Hart, R. C. (1991) Food and suspended sediment influences on the naupliar and copepodid durations of freshwater copepods: comparative studies on Tropodiaptomus and Metadiaptomus. J. Plankton Res. 13, 645-660. Hart, R. C. (1992) Experimental studies of food and suspended sediment effects on growth and reproduction of six planktonic cladocerans. J. Plankton Res. 14, 1425-1448. Jack, J. D. and Gilbert, J. J. (1993) The effect of suspended clay on ciliate population growth rates. Freshwater Bioi. 29,385-394. Jack, J. D., Wickham, S. A., Toalson, S., and Gilbert, J. J. (1993) The effect of clays on a freshwater plankton community: An enclosure experiment. Arch. Hydrobiol. 127,257-270. Kirk, J. T. 0. (1985) Effects of suspensoids (turbidity) on penetration of solar radiation in aquatic ecosystems. Hydrobiologia 125, 195-208. Kirk, K. L. (1992) Effects of suspended clay on Daphnia body growth and fitness. Freshwater Bioi. 28, 103-109. Kirk, K. L. and Gilbert, J. J. (1990) Suspended clay and the population dynamics of planktonic rotifers and cladocerans. Ecology 71, 1741-1755. Koenings, J.P., Burkett, R. D., and Edmundson, J. M. (1990) The exclusion oflimnetic Cladocera from turbid glacier-meltwater lakes. Ecology 71, 57-67. Marshall, S. M. (1973) Respiration and feeding in copepods. Adv. Mar. Bioi. 11, 57-120.

322

NANCY M. BUTLER

Peters, R. H. (1984) Methods for the study of feeding, grazing and assimilation by zooplankton. In: A Manual on Methods for the Assessment ofSecondary Productivity in Fresh »flters, J. A. Downing and F. H. Rigler, eds., Blackwell Scientific Publications, Oxford. Spencer, C. N. (199la) Comparative limnology of Swan Lake and Flathead Lake, northwestern Montana. Flathead Lake Biological Station Open File Report No. 126-91. Spencer, C. N. (199lb) Evaluation ofhistorical sediment deposition related to land use through analysis oflake sediments. Flathead Lake Biological Station Open File Report No. 123-91. Sternberger, R. S. (1981) A general approach to the culture of planktonic rotifers. Can. J Fish. Aquat Sci. 38, 721-724. Williamson, C. E. and Butler, N. M. (1986) Predation on rotifers by the suspension- feeding calanoid copepod Diaptomus pallidus. Limnol. Oceanogr. 31, 393-402. Williamson, C. E. and Butler, N. M. (1987) Temperature, food and mate limitation of copepod reproductive rates: separating the effects of multiple hypotheses. J Plankton Res. 9, 821-836. Williamson, C. E., Butler, N. M., and Forcina, L. (1985) Food limitation in naupliar and adult Diaptomus pallidus. Limnol. Oceanogr. 30, 1283-1290.

THE ESCAPE BEHAVIOR OF PLEUROMAMMA XIPHIAS IN RESPONSE TO A QUANTIFIABLE FLUID MECHANICAL DISTURBANCE DAVID M. FIELDS* and JEANNETTE YEN Marine Sciences Research Center, SUNY Stony Brook Stony Brook, NY 11794-5000 jyen@ccmail. sunysb. edu *David Fields, NELHA, P Q Box 1749, Kailua-Kona, HI 96745 email:fields@uhunix. uhcc. hawaii. edu Pelagic copepods are subject to predation throughout much oftheir planktonic life. As a result, predator detection and avoidance are crucial to the survival of individuals. By using a quantifiable fluid mechanical disturbance we determine the fluid characteristics needed to elicit an escape reaction in Pleuromamma xiphias (Calanoida: Metridinidae). Four different siphon configurations were used to spatially separate regions of maximum flow speed and acceleration from regions with maximum shear surrounding the siphons. The patterns of escapes indicate that the spatial variation in fluid velocity is the proximate cue which elicits the escape reaction in P xiphias. An average threshold shear value of 15 /s was needed to elicit the escape reaction. These results suggest that small scale fluid motion, such as those caused by a predator's feeding current, with shear values greater than the 15 /s are more likely to initiate an escape reaction in P xiphias.

INTRODUCTION Predation exerts considerable influence on the structure of lower trophic level communities (Brooks and Dodson, 1965). The structuring is commonly through selective predation on one or several species depending on the efficiency of a predator, or alternatively the vulnerability of a particular prey. Understanding the critical factors which control predation in pelagic communities allows insight into the evolutionary responses of prey and predators and offers a unique opportunity to predict the role of predation in controlling community structure in particular environments. Most studies involving predation on copepods have been approached through the abilities of the predators to detect and capture their prey (Kerfoot eta/., 1980; Landry, 1978, 1980, 1981; Lonsdale et a/., 1979; Drenner and McComas, 1980; Greene, 1988; Doall and Yen, in preparation). Attack volumes then are defined based on the percentage of prey captured in different regions. However, from the point of view of the prey, success is defined by their ability to escape an attacking predator. Therefore, of equal importance, and the focus of this study, is the ability of copepods to avoid capture. Calanoid copepods are one of the most effective zooplankton at avoiding predation (Szlauer, 1964; Strickler, 1975; Kerfoot eta/., 1980; Kettle and O'Brien, 1978). Their success is a product of their ability to remotely detect predators (Yen and Fields, 1992; Lenz and Yen, 1993) and the speed that they are able to achieve during an escape 323

324

DAVID M. FIELDS and JEANNETTE YEN

reaction (Strickler, 1975; Kettle and O'Brien, 1978; Vinyard, 1980). However, the proximal cues used to detect their predators are presently unknown. It is improbable that predator detection by copepods occurs through visual cues since most copepods lack welldeveloped eyes and would be unable to assess distance from the predator. Chemoreception is also an unlikely mechanism for remote detection of predators. Entrainment by suction or feeding currents is the most common forms of feeding by teleost fish (Liem, 1980; Vinyard, 1980; Lauder, 1980; Lauder and Clark, 1984), predatory copepods (Strickler, 1985; Tiselius and Jonsson, 1990; Paffenhofer and Lewis, 1990; Yen et al., 1991; Yen and Fields, 1992; Fields and Yen 1993), bivalves (Emlet, 1990) and tunicates (Flood, 1991). Since the flow field created by the predator is towards itself and the prey is upstream of the predator, chemical detection would require that odor diffusion away from the predator is more rapid then their generated flow field. This is dubious with typical feeding current velocities for copepod predators being on the order of mm-cm/s. There is justification, therefore, in speculating that calanoid copepods react to the fluid mechanical disturbance created by their predator. The presence of mechanoreceptors on the antennae of calanoid copepods is well documented (Strickler and Bal, 1973; Landry, 1980; Friedman, 1980; Yen et al., 1992; Lenz and Yen, 1993). These receptors can detect water displacement ofless than 10 nm (Yen et al., 1992) making them highly sensitive tools for the detection of microscale water motion. In addition, Fields and Yen (1993) and Lenz and Yen (1993) have shown that the distal tips of copepod antenna are essentially outside the rapid flow region of their own feeding current and may be strategically located for detecting larger scale flow patterns. The importance of the distal tips was supported by electrophysiological studies which showed the disappearance of large neural spikes, commonly associated fast reaction times such as those involved in an escape response, when the distal tips are ablated (Lenz and Yen, 1993). Despite the fact that we know a lot about the structural morphology of mechanoreceptors, little is known about the physical characteristics of the biologically relevant mechanical stimuli which elicit different behaviors in copepods. In this study we examine the escape behavior of a calanoid copepod, Pleuromamma xiphias, to a quantifiable fluid mechanical disturbance. P xiphias is an omnivorous (Mullin, 1966; Bennett and Hopkins, 1989), bioluminescent (Buskey et al., 1989), vertically migrating copepod found at 80 meters during the night and 700 meters by day (Ambler and Miller, 1987; Haury, 1988). It has been shown to be a preferred diet for mid-water nekton (Hopkins and Baird, 1981; 1985), euphausids (Hu, 1978), and decapods (Foxton and Roe, 1974). This study details the characterization of the suction apparatus (used to elicit the escape reaction) in terms of the shear, velocity and acceleration of the in flowing water. The escape locations of the copepods are compared to the characteristics of the fluid. METHODS Animal Collection. Pleuromamma xiphias were collected from 610 meters at the Natural Energy Laboratory of Hawaii Authority (Fields and Yen, 1993). Approximately 300 animals (for each experiment) were selected and brought to 22°C, the temperature at the peak of its vertical migration (Ambler and Miller, 1987), over a 24-48 hour period prior to filming. Video Recordings. Video observations were made through a 22 liter Plexiglass vessel. A cube of approximately 900-1000 ml of fluid in the experiment tank (Figure 1)

COPEPOD ESCAPE FROM FLUID DISTURBANCE

325

was filmed for analysis using a magnification of 3.5-5X. The video cameras were two perpendicularly mounted Pulnix TM-745 equipped with 28-70 mm macro-zoom Vivitar lenses. The cameras were fixed to a Newport motion isolated table. The tank was illuminated from above with two fiber optic Shotts lamps covered with a Kodak Wratten Gelatin Filter (#24) to shift the spectrum into the red light. Images were recorded on two Panasonic AG-1960 video recorders synchronized with a Comprehensive Video Supply Corporation time code generator. To increase temporal resolution, individual video frames (0.033 s) were split and analyzed field by field (0.0167 s). Video analysis was accomplished by digitizing video images on a Gateway 2000 ® PC equipped with a frame grabbing card and Bioscan Image Analysis ® software. Siphon Tank Configuration. The siphons were made from stock Cornix Pyrex glass tubes and mounted in the center of the 22liter tank (Figure 1). The mouth of the siphon was at a height of 73-80 mm from the bottom of the tank. Three different siphon diameters each with their own hydrostatic head (measured from water surface to the mouth of siphon) were used (Table 1). The specific head heights were chosen such that the majority of escape reactions occurred at a distance of at least 3 mm from the siphon center. This was to ensure that the siphon did not interfere with the determination of the position of the copepod in one of the video cameras. A constant head pressure was maintained by simultaneously replacing the exiting water with 0.45 t-tm filter sea water. To diminish the disturbance to the siphon flow, incoming water was introduced into a 105 mm diameter vessel which drained through a 35 t-tm mesh screen located just below the water's surface (Figure 1). Siphon Flow Analysis. Flow velocities at different points within the siphon flow were determined by tracking neutrally buoyant Artemia eggs ( ,..,_ 200 f.-tiD in diameter) entrained by the flow from both cameras. By combining data points that share a similar value ( ± 200 t-tm) in theY direction, an essentially planer view of the particle trajectories could be analyzed. Velocity (8S/dt) and accelerations (8U/dt) were calculated as the change in distance and velocity, respectively, over time. The length of time used varied from 2 to 10 fields (33 and 167 ms) for velocity and 4 to 20 fields for acceleration depended on the location of the particle in the flow field. The greater the change in distance or velocity the smaller the time increment used. At the mouth of the siphon, the flow velocities became too rapid for the temporal resolution of the video cameras and could not be determined by tracking particles. To calculate the flow velocity at the origin of the siphon, the water exiting the siphon was measured by determining the length of time required to fill, 100, 250 and 500 ml volumetric flasks at different periods throughout the experiment. Since at the entrance of the siphon the velocity is nearly uniform across the diameter (Vogel, 1981), the total flow (Q:mlls) then was used to find the average entry flow velocity (V*) as: (Eq.l) Where "a'' is the radius of the siphon (Vogel, 1981). Table 1 shows the calculated flow speeds at the mouth of the four siphon configurations used. Shear was calculated as dU/dz + dW/dx. Velocities in the X (dU) and Z (dW) directions were taken from contoured velocity data of the horizontal and vertical components, respectively, using a 0.02 mm separation as 8z and 8x. The shear data were contoured with grid size of0.5 mm. Both the measured and calculated points were used

DAVID M. FIELDS and JEANNEITE YEN

326

z 73 - 80 mm

1

X Figure 1 Design ofthe siphon apparatus. Siphon was mounted in the center of the 22 L tank

~ 73-80 mm from the bottom. The mouth (top) of the siphon was 77- 177 mm below the surface and 70 mm from any walL Flow rate (Q:ml/s) was determined from the water which exited the siphon. Isolines of flow speed, acceleration and shear were computed from particles entrained by the siphon. For comparison of different siphon characteristics, profiles of flow were taken along a transect(- - -)parallel to the top of the siphon at a distance of 3 mm. To determine the proximate cue which elicits escape behavior in P xiphias the distance of the escape location from the isoline of each characteristic was used to compare the three signals tested.

COPEPOD ESCAPE FROM FLUID DISTURBANCE

A

500.00

{ 200.00

327

8

100.00 50.00

•o.oo 30.00

500.00

20.00 10.00

200.00

5.00

2.50

0.00

10.00

20.00

30.00

40.00

0.00

0.00

10.00

20.00

30.00

40.00

100.00 50.00

c

40.00

Figure 2 Isolines of flow speed into a 3.0 mm (A), 5.0 mm (B), 7.0 mm (C), 7.0 mm (D) outer diameter (O.D) siphon with corresponding escape locations of P xiphias. Isolines are calculated as the displacement of200 11m neutrally buoyant particles over a time period of 0.033- 0.167 ms depending on their location in the flow field. Isolines values are in given in corresponding color band.Values are reported in mm/s.

in the contouring. Velocity, acceleration and shear contours were generated using SURFER ® contouring software distributed by Golden Software Inc. Fluid Mechanical Signal at the Point of Escape. Escape locations were plotted with contours of velocity, acceleration and shear to determine these three characteristics of the fluid at each escape location. The characteristics from the different siphon configurations were combined and the means and variances tabulated. The characteristic with the lowest variability was considered the proximate cue which elicits the escape. Since the total variation is sensitive to the value of the means, the coefficient of variation (CV) was used to compared the three characteristics (Sokal and Rohlf, 1981). The elegance of using the siphon flow is that it separates different fluid mechanical characteristics geographically around the siphon. Therefore, it is important in the analysis of the escape location that it remain within the context of spatial distribution of the possible stimuli. To accomplish this, we compared the actual distribution of escape locations to the particular pattern of the fluid mechanical disturbance created by each characteristic. To characterize the escape pattern, the distance of each escape was plotted as a function of the angle from the siphon center at which it occurred. A 3rd order polynomial was used to smooth the data while allowing it to retain its inherent variation with the angle from the siphon center at which the angle occurred. The pattern of the fluid mechanical disturbance created by each characteristic was described by plotting the distance of the critical isoline (the average value found for each siphon) of water speed, acceleration and shear at the point of escape reaction as a

328

DAVID M. FIELDS and JEANNETTE YEN

8

A

• •

16000.00

•

14000.00 12000.00 10000.00

0.00

5.00 10.00 15.00 20.00

0.00

10.00

20.00

30.00

40.00

8000.00 6000.00 4000.00

c

2000.00

0.00

10.00

20.00

30.00

40.00

Figure 3 Isolines of flow acceleration into a 3.0 mm (A), 5.0 mm (B), 7.0 mm (Q, 7.0 mm (D) O.D. siphon with corresponding escape locations of P xiphias. Isolines are calculated as the change in velocity of neutrally buoyant 200 J.LID particles over a time period of 0.065- 0.334 ms. Isoline values are in given in corresponding color band.Values are reported in mm/s2•

function of the angle measured from the siphon center. The relationship between the pattern of the fluid mechanical stimulus and the escape pattern was determined by comparing the distances of the escape locations to the critical value isoline for each characteristic. In effect, this method measures how well the isoline for each characteristic explained the location of the escape. Since the distances were not normally distributed, the means were tested using the Mann- Whitney statistic. As a control, escape reactions to the presence of the siphon without the flow were also noted to ascertain whether the physical presence of the siphon was sufficient to elicit an escape reaction. In these controls only 1escape during a four hour period was observed. RESULTS The results of this study are divided into 2 functional parts. First the siphon is presented as a tool for the creation of a quantifiable fluid mechanical disturbance. Second is a discussion of the fluid mechanical signal which elicits the escape reaction of P xiphias. Siphon Flow. The flow into the siphon was radially symmetrical allowing its characteristics to be adequately represented in 2 dimensions. Isolines of flow speed (Figure 2), acceleration (Figure 3) and shear (Figure 4) are shown unencumbered by the prey escape locations on the left of the siphons and with the prey escape locations standardized to a common siphon orientation on the right. The shape of the speed

COPEPOD ESCAPE FROM FLUID DISTURBANCE

329

A

0.00

10.00

20.00

30.00

40.00

c D

-0.50 -1.00 ·5.00 -10.00 -50.00 -100.00

Figure 4 Isolines of flow shear into a 3.0 mm (A), 5.0 mm (B), 7.0 mm (C), 7.0 mm (D) O.D. siphon with corresponding escape locations of P xiphias. Isolines are calculated as the change in velocity of neutrally buoyant 200 J.tffi particles over space (see text). Isoline values are in given in corresponding color band. Values are reported in 1/s.

isolines surrounding the siphon is best described as concentric rings of speed anchored at the center of the siphon's mouth. Similarly, above the mouth of the siphon, acceleration of the water into the siphon is of equal value at equal distance from the siphon center. However, slightly below the siphon's mouth, acceleration is relatively slower. Shear, in contrast to both velocity and acceleration, was minimum directly above the siphon center and along the plane defined by the top of the siphon. Maximum shear values were at angles of rv35° measured from a line vertical to the siphon center. These features are interpreted more easily from profile data taken from these contours. Profiles of water velocity, acceleration and shear for each siphon are shown in Figure 5 (A-D). The cross-section was taken parallel to the bottom of the tank and 3 mm above the mouth of the siphon (Figure 1). Herein, the profiles are used as a measure of differences between the siphons for comparison, not as a measure of absolute maximums within each siphon. The effect on the flow rate of varying only the siphon radius is evident by comparing siphon 1 and siphon 2. Head height above the siphon was equal between these two experiments while internal siphon diameter was more than doubled from 1.5 mm to

330

DAVID M. FIELDS and JEANNETTE YEN

3.4 mm. Velocity calculated at the mouth of the siphons showed siphon 1 to be ,..., 40% lower than in siphon 2 (Table 1). From the profile, the maximum shear values increased with increased siphon radius from 8.5/s to 42/s.Velocity showed a similar increase from 38-130 mm/s as did acceleration which increased from 107 to 600 cm/s 2•

Table 1 Characteristics of the four siphon configurations. Flow speeds are calculated using Eq. I in text. Siphon

1 2 3 4

Siphon Diameter Outer Inner (mm) (mm)

1.5 3.4 5.0 5.0

3.0 5.0 7.0 7.0

Siphon Length (mm)

Hydrostatic Head

Avg. Flow U!locity at Origin (mm!s)

73 90 100 100

177 177 77 127

1063 1492 1273 1432

A comparison of siphon 3 (Figure 5C) and 4 (Figure 5D) demonstrates the effects of changes in head height while maintaining a constant siphon radius. Flow at the mouth of siphon 3 was 89% of siphon 4 (Table 1). However, as the profiles indicate, at a distance of 3 mm, only slight variations in the velocity, acceleration, and shear between the two siphon configurations were detected. Maximum velocity along the profile was .-v25 em/ s for both siphons and acceleration reached .-v20 m/s 2. Maximum shear along the profile for these siphons was ,..., 72 /s. Identification of the proximate mechanical cue which elicits the escape reaction in P xiphias requires that the possible stimuli be varied independently. Surrounding a siphon, each angle and distance measured from the siphon's center has a unique combination of water velocity, acceleration and shear. By comparing the suite of characteristics at one location in one siphon with those found in another siphon, the ability to independently vary the characteristics becomes evident. For example, along the 3 mm profile for siphon 2 (Figure 2B) at 1 mm from the central axis, the fluid characteristics include a velocity of 12 cm/s, an acceleration of 6 m/s 2, and a shear value of 21 /s. At a distance of 3.5 mm from the central axis of siphon 3 (Figure 2C) velocity is also 12 cm/s however, acceleration is almost double at 11 m/s 2 and shear is 3.5 times the value found for siphon 2. The ability to independently vary the characteristics of the fluid disturbance makes the siphon a powerful tool for determining the role of fluid motion in eliciting behavior in copepods. Escape Signals. The fluid characteristics at 172 total escape locations of Pleuromamma xiphias were examined from the four different drain configurations. Average values (± SD) at the point of escape for distance, velocity, acceleration, and shear are shown in Table 2. Larger siphons gave rise to escape reactions that were initiated further from the center of the siphon regardless of the flow speeds at the mouth of the siphon. This suggests that the copepods can detect fluid mechanical differences in the disturbance created by the various siphons. For each siphon, the characteristics at the point of escape were skewed to the right (due to a small percentage of escapes occurring at very high values), deviating significantly from normality (p < 0.001 for all: Kolmogrov-Smirnov). Before

COPEPOD ESCAPE FROM FLUID DISTURBANCE

120 100

331

80

60

80 -

60

40

20

20

40

0

8

s v

0 I

~ ~

I

I

4 ~ 0 2 4 6 8

~ ~

Distance (mm) 80

60

60

40

40

20

20

0

0

4

~

0 2 4 6 8

Distance (mm)

~

0 2 4 6 8

Distance (mm)

80

~ ~

4

D

~ ~

4

~

0 2 4 6 8

Distance (mm)

Figure 5 Profile of water velocity (V), acceleration (A) and shear (S) taken 3 mm above the siphon mouth and 7 mm lateral from the siphon center in siphon 1 (A), siphon 2 (B), siphon 3 (C) and siphon 4 (D). Darkened line along the abscissa represents the diameter of the siphon. Units for velocity are shown in mm/s in siphon 1 and cm/s for siphons 2-4. Acceleration is given in units of cm/s 2 in siphon I and m/s 2 for siphon 2-4. Shear is shown in units ofl/s for all the siphons.

combining the data from the different siphons, the means were tested using a KruskalWallis test (Table 2). The means were not significantly different, allowing the combined averages and variances to be used. The characteristic with the lowest coefficient ofvariation was shear with a mean value of 15.2 /s needed to elicit an escape reaction. The CV increased from a low of 1.24 for shear to a maximum of 1.46 in acceleration. Velocity had an intermediate CV of 1.40 (Table 2). In addition to examining the degree of variability of each characteristic we tested the spatial distribution of the escape behavior in relation to the distribution of the fluid characteristics. The escape locations were plotted with the water speed (Figure 2), acceleration (Figure 3) and shear values (Figure 4) surrounding each of the siphons.

DAVID M. FIELDS and JEANNETTE YEN

332

Table 2 Average distance measured from the siphon center at which Pleuromamma xiphias initiated an escape reaction. Average values (+I- SD) for the fluid characteristics at the point of escape are listed for individual siphons. Results from the Kruskal-Wallis test for differences between the siphons and the combined averages and variances are shown at the bottom of each column. The coefficient ofvariation (CV) is shown for each characteristic. Siphon

Distance (SD) (mm)

Uflter 11!/ocity (SD) (mmls)

Uflter Acceleration (SD) (mmls 2)

Shear (SD) (lis)

1 {N=l4) 2 {N=45) 3 (N=43) 4(N=70)

4.8 (4.2) 7.0 (3.1) 9.3 (2.8) 9.5 (4.0)

50.4(49.4) 57.0 (54.9) 24.2 (11.7) 37.1 (75.1)

1960 (1934) 1582 (1894) 1131 (1198) 1817 (2876)

13.1 (8.6) 27.0 (27.0) 9.3 (8.4) 11.5 (15.0)

K-W Mean Wzriance

p=O.lO 40.5 (58.9) 3470.2 1.46

p=0.76 1592 (2239) 5,043,996 1.40

p = 0.33 15.2 (18.9) 361.7 1.24

cv

Figure 6 shows the percentage of escapes as they relate to the geographical region surrounding each siphon. This figure represents the probability that an approaching P. xiphias will escape within that geographical region. In predator prey interactions this view is akin to the capture field of a predator except that it is expressed from the view point of the behavior of the prey. Immediately evident from the pattern of the escape location is the dearth of escapes in the region directly above the siphon and the cluster of escapes lateral to the mouth of the siphon. Animals entrained from above the siphon did not display an escape reaction and generally were "captured" by the siphon. This area above the siphon coincides with the region with the highest water speed (Figure 2, 5) and acceleration (Figure 3, 5) and the lowest shear values (Figure 4, 5). If either water speed or acceleration were acting as the proximate cue that elicits the escape reaction, numerous escapes would have occurred in this region. The cluster of escapes lateral to the siphon corresponds well with the regions where the isolines of shear are dense. Animals entrained within this region experienced a rapid increase in the shear rate just prior to eliciting an escape response. The most distant escapes, representing only a small percentage of the total escapes, occurred within an angle of 30°-40° from the siphon center. This region corresponds with the most lateral extent of the siphon induced shear. In general, the pattern created by the distribution of shear qualitatively agrees with the escape pattern surrounding the siphon (Figure 4). Considering the average distance of the escape from the different isolines, shear consistently showed a better fit to the escape data then either velocity or acceleration. Figure 7 shows the escape locations as compared to the contour of the three characteristics at the point of escape for siphon 3. Immediately evident is the similarity between the shape of the prey escape pattern and the pattern of the shear isoline. The distance of the escape locations from the shear isoline was significantly less than velocity and acceleration for all the siphons (MannWhitney; p < 0.01 : table 3) except for siphon 1 and 3. The lack of significance in both siphons reflects the high variability in the velocity and acceleration data.

COPEPOD ESCAPE FROM FLUID DISTURBANCE

333

00

90°

o·

0

0

D

1% 0

90

Figure 6 Frequency distributions of the escape responses of Pleuromamma xiphiaswith respect to the geographical region surrounding each siphon. Regions are demarked by both distance and angle from the siphon center. Each arc represents an equal distance from the siphon center as measured along the vertical line above the siphon. Angles are in increments of 30' measured clockwise from the vertical line through the siphons' center.

DISCUSSION Numerous researchers implied that the shape of a predator's feeding current can have strong implications on the prey they are able to capture (Strickler and Twombly, 1975; Strickler 1985; Greene, 1988; Fields and Yen, 1993). Without knowledge as to the nature

DAVID M. FIELDS and JEANNETTE YEN

334

Table 3 The distance (Mean+ I- SD) of the escape location of Pleuromamma xiphias from the isoline of mean velocity, acceleration and shear are listed for each siphon. Distance was measured from a yd order polynomial fit to the escape data. Lowest values indicate the isoline for the fluid characteristic with the best fit to the escape locations. Means were tested using the Mann-Whitney statistic.

Mean Distance (mm)

Fluid Characteristic

Siphon]

Siphon2

Siphon3

Siphon4

Velocity(V) Acceleration(A) Shear(S) VxA VxS AxS

2.56 (1.08) 2.10 (2.08) 0.60(0.37) NS

2.04 (0.89) 1.71 (0.76) 1.20 (0.29)

1.97 (2.75) 2.31 (1.78) 1.29(0.46)

2.71 (1.76) 2.48 (1.78) 1.47 (0.71)

***

NS(p~o.o7)

*** *** ***

***

NS NS

** ** **

of the mechanical signals which elicit the escape reaction, it is unclear as to what currency to use to describe the feeding current of copepod predators. Tautz (1979) suggests that determining the stimuli which elicit particular behavior patterns require experiments that independently vary fluid characteristics relative to each other. In this study we used a variety of different siphon configurations to create regions that are dominated by the effects of different characteristics. We then compared the escape locations of Pleuromamma xiphias to the pattern of the fluid disturbance. Use ofthe siphon. Although practical, the use of the siphon has both positive and negative features. Benefits to using a continuously drawing siphon is that it is temporally stable, quantifiable and simple. Temporal stability allows the generation of one descriptive contour for each flow characteristic rather than needing to consider multiple time slices. Once the flow is characterized in 3-dimensions, it can be depicted using only 2-dimensions as a result of the flow symmetry around the mouth of the tube. This facilitates comparing spatial characteristics of the flow with the behavior of the animals in that flow. A negative aspect of using the continuous flow is the loss of the oscillatory nature characteristic of many feeding currents. Oscillatory feeding currents are known to result from the rapidly expanding mouth of a fish (Lauder, 1980) or the beating of the feeding appendages of animals such as copepods (Cannon, 1928). Many marine copepods are able to phase lock to the mechanical stimulus created by an oscillating bead (Yen eta/., 1992; Lenz and Yen, 1993). In addition, vibratory motion at the frequency of copepod feeding appendages has been shown to elicit the capture response in chaetognaths (Horridge and Boulton, 1967). These studies suggest that the nature of flow can be important in eliciting animal behavior. However the present study shows that oscillatory flow is not necessary to elicit the escape behavior of copepods. Use of the siphons worked extremely well to separate the various fluid characteristics being tested. Within a single siphon and between siphons, different locations gave rise to a particular pattern of the three fluid characteristics. By comparing the escape behavior in different locations, the effect of varying individual fluid mechanical disturbances could be analyzed. By independently varying the size of the siphon mouth and the hydrostatic head pressure, particular configurations were created which mimic the dimensions of different predator types. For example, at a distance of 4 mm from the siphon center the configuration of siphon 1 gave rise to flow speeds similar to those

COPEPOD ESCAPE FROM FLUID DISTURBANCE

335

25

.-. 20

E E

Cl) (J

15

... s::

cu

tn 10

·-c

5 0 0

20 40 60 80 100 120 140 160

Angle (deg) Figure 7 The distance of individual escape locations (•) of P. xiphias from the siphon center as a function of angle (¢ ). Escape data were fitted to a 3'd order polynomial (* - - - - - - -). Also shown is the distance of the isolines ( ) for the average water speed (V), acceleration (A) and shear (S) at the point of escape, as a function of the angle measured from the center of siphon 3.

created by a large copepod such as P. xiphias (Fields and Yen, 1993). The larger siphons (2-4) are more similar to the fluid flows of suction feeding fish (Lauder and Clark, 1984). More subtle differences such as suction strength between animals of similar size can be examined by varying only hydrostatic head height. This type of analysis would be useful to determine the effects different feeding current strength have on prey escape from two similar sized copepods. Escape Signals. The antennae of marine copepods are adorned with numerous mechano - and chemoreceptors (Strickler and Bal 1973; Friedman, 1980; Barrientos Chacon, 1980). Signal differentiation allows the copepods to distinguish between predators, prey and mates (Strickler and Twombly, 1975; Yen et al., 1991; Lenz and Yen

336

DAVID M. FIELDS and JEANNETTE YEN

1992). The signals can be a cocktail of stimuli received at numerous locations along the antennae involving both chemical and mechanical signals. However, from the results of this study, only a mechanical stimulus is necessary to elicit an escape reaction in P xiphias. The past decade has seen an accumulation ofliterature which describes the flow field of teleost fish (Lauder and Clark, 1984) and calanoid copepods (Strickler 1985; Tiselius and Jonson, 1990; Yen and Fields, 1992; Fields and Yen, 1993). The currency used in these studies to describe the flow field has been flow speed. However, concurrent research on the escape reaction of rheotactic prey suggests that the flow speed is not an adequate indicator of prey escape location _(Haury, 1980; Kirk and Gilbert, 1988; Yen and Fields, 1992; for opposing view- Costello eta/., 1990; Hwang; 1991). As a result, the commonly used isotachs of water speed gives no indication of the prey's reaction to that flow structure and offers no insight into the role of flow fields in mediating predator prey interactions. A more useful flow field description would be one which accounted for the prey's behavior. The use of flow speed is in part a reflection ofthe inconclusive information as to the stimulus which elicits the escape reaction. Without qualitative information as to the nature of the stimulus that gives rise to the escape behavior, choosing a relevant currency for the feeding currents was not possible. Using flow speed as the currency has led to misconceptions about the mechanical sensors of copepods and interpreting the shape of predators feeding currents in light of predator prey interactions. This was cogently discussed by Yen and Fields (1992) but is included here for the sake of completeness. Setal bending is believed to be the mechanism by which copepods transform the mechanical motion of a fluid mechanical disturbance into a neurophysiological signal which elicits the behavior. This has been supported by the finding that setal bending along the first antennae elicits large spike neural activity commonly associated with escape behavior (Yen eta/., 1992). For a seta to bend relative to the body it must be exposed to a flow rate that is different from the rest of the body. A critical water speed does not fulfill this requirement. This is because a seta on an animal entrained within a spatially and temporally uniform flow would not have a force acting upon it which would cause it to move relative to the rest of the animal. In essence, since the animal's entire frame of reference is moving at the same velocity as the animal itself, it would lack the ability to detect that it is moving relative to a larger field of reference. In contrast, in a spatially uniform accelerating field or in a temporally uniform shear field detecting motion through setal bending is possible. In an accelerating field, the body of the animal, which has more mass then the seta, will accelerate slower. This causes the seta to bend in the direction of the flow allowing the animals to detect their relative motion and direction. In a shear field the animal moves at a speed resultant of the integrated effect of the different velocities acting on the animal's entire body. The setae, however, are only affected by the relative motion caused by the local flow regime. In a spatially variable flow, the seta will experience a relatively different force than that acting on the body and as a result will move relative to the body. Therefore, we reason that only acceleration and shear offer a mechanism for a copepod to detect that it is entrained in a larger flow field. By separating regions of maximum shear and acceleration we were able to determine that shear acts as the proximate cue to elicit the escape response, with a critical shear value of approximately 15 /s. Furthermore, the general pattern created by the escape locations mimicked the shape of the shear isolines.Very few escapes occurred in regions of low shear, such as those found above the siphon, while in regions of high shear, animals exhibited rapid, directed escape reactions.

COPEPOD ESCAPE FROM FLUID DISTURBANCE

337

Since calanoid copepods have such well developed mechanoreceptive and escape capabilities, it is reasonable that predators have evolved different mechanisms to enhance their abilities at capture specific food items. Based on the results of this study, predators which minimize the shear values within their feeding currents are less likely to elicit an escape reaction in their prey. This is supported by the observation that Chromis verdis uses jaw protrusion with little suction to capture copepods (Coughlin and Strickler, 1990) while Chaetodon miliaris feed primarily as inertial suction feeders on Artemia which escape at much slower speeds then copepods (Motta, 1984). Predatory copepods show a similar tendency. For example, an omnivorous copepod creates a large anterior double shear feeding current that enhances their chemoreceptive abilities when grazing on algal cells (Strickler, 1982). However, when shifting to a more carnivorous diet, they begin to cruise while simultaneously moving the shear field to their side which diminishes prey escape. Mathematical models of a stationary and a moving predatory copepod show that even if the animal did not entrain the water from the side, the deformation rate in front of a moving animal is less than that created by a stationary feeder (Tiselius and Jonsson, 1990). This indicates thatthe shape of the feeding current enhances the capture of prey by minimizing the ability of the prey to detect entrainment. This is further supported by the patterns depicted in the escape field surrounding the siphons (Figure 6) which complement the reactive fields seen in many copepod predators (for fish - Luecke and O'Brien, 1981; predatory copepods -Kerfoot eta/., 1980; Doall eta/., in preparation). The ability of copepods to decode complex mechanical signals as a mechanism to evaluate the presence of predators, prey items and potential mates is crucial to their survival. An animal generated mechanical signal may outlast its creator by seconds, allowing ample time for the copepod to orient its swimming towards potential prey and mates or assume behavioral mechanisms that help them avoid predation. This study illustrates that to understand interactions among mechanoreceptive animals requires incorporating detailed investigations of the physical structure of the fluid environment with behavioral studies addressing threshold values needed to elicit specific behavioral responses. Acknowledgements

We thank Dr. T. Daniel and all the personnel at the Natural Energy Laboratory of Hawaii for the use of the facility and their help in collecting and maintaining animals and Drs. P. Lenz, D. Hartline and J. Purcell for organizing a delightful conference. We also thank Dr. J.R. Strickler for his advice in all aspects of this work and two anonymous reviewers whose comments greatly improved the quality of this manuscript. This work was supported by a Learner Gray Grant administered by the Natural History Museum of New York awarded to D.M. Fields and an ONR contract N00014-87-K-0181 to J. Yen. This is contribution number 1003 from the Marine Sciences Research Center of the State University of New York at Stony Brook. References Ambler, J. W. and Miller, C. B. (1987) Vertical habitat-partitioning by copepodites and adults of subtropical ocean copepods. Mar. Bioi., 94, 561-577. Barrientos Chacon,Y (1980) Ultrastructure ofsensory units ofthefirst antennae ofcalanoid copepods. MS Thesis, University of Ottawa, Ontario.

338

DAVID M. FIELDS and JEANNETTE YEN

Bennett, J. L. and Hopkins, T. L. (1989) Aspects of the ecology of the calanoid copepod genus Pleuromamma in the Eastern Gulf of Mexico. Contr. Mar. Sci., 31, 119-136. Brooks, J. L. and Dodson, S. I. (1965) Predation, body size, and composition of plankton. Science, 150,28-35. Buskey, E.J., Baker, K. S., Smith, R. C. and Swift, E.(l989) Photosensitivity of the oceanic copepods Pleuromamma gracilis and Pleuromamma xiphias and its relationship to light penetration and daytime depth distribution. Mar. Ecol. Prog Ser., 55, 207-216. Cannon, H. G. (1928) On the feeding mechanisms of the copepods Calanus finmarchicus and Diaptomus gracilus.1 Exp. Bioi., 26, 131-144. Costello, J. H., Strickler, J. R., Marrase, C., Trager, G., Zeller, R. and Freise, A. J. (1990) Grazing in a turbulent environment: Behavioral responses of a calanoid copepod, Centropages hamatus. Proc. Nat/. Acad. Sci. USA., 87,1648-1652. Coughlin, D. J. and Strickler, J. R. (1990) Zooplankton capture by a coral reef fish: an adaptive response to evasive prey. Environ. Bioi. Fish., 29,35-42. Drenner, R. W. and McComas, S. R. (1980) The roles of zooplankter escape ability and fish size selectivity in the selective feeding and impact of planktivorous fish. In: Evolution and Ecology ofZooplankton Communities, W. C. Kerfoot, (ed:) University Press of New England, Hanover. Emlet, R. B. (1990) Flow fields around ciliated larvae: effects of natural and artificial tethers. Mar. Ecol. Prog. Ser., 63, 221-225. Fields, D. M. and Yen, J. (1993) Outer limits and inner structure: The 3-dimensional flow field of Pleuromamma xiphias (Calanoida: Metridinidae). Bull. Mar. Sci., 53,84-95. Flood, P. R. (1991) Architecture of, and water circulation and flow rates in, the house of the planktonic tunicate Oikopleura labradoriensis. Mar. Bioi., 111, 95-111. Foxton, P. and Roe, H. S. J. (1974) Observations on the nocturnal feeding of some mesopelagic decapod crustaceans. Mar. Bioi., 28, 37-49. Friedman, M. M. (1980) Comparative morphology and functional significance of copepod receptors and oral structures. In: Evolution and Ecology of Zooplankton Communities, W. C. Kerfoot. (ed), University Press of New England, Hanover. Greene, C. H. (1988) Foraging tactics and prey-selection patterns of omnivorous and carnivorous copepods. Hydrobiologia, 167/168,295-302. Haury, L. R. (1988) Vertical distribution of Pleuromamma (Copepoda: Metridinidae) across the eastern North Pacific Ocean. Hydrobiologia, 167/168,335-342. Haury, L. R., D. E. Kenyon and Brooks, J. R. (1980) Experimental evaluation of the avoidance reaction of Calanusfinmarchicus.l Plankt. Res., 2,187-202. Hopkins, T. L. and Baird, R. C. (1981) Trophodynamics of the fish Tillenciennellus tripunculatus. I. Vertical distribution, diet and feeding chronology. Mar. Ecol. Prog Ser., 5, 1-10. Hopkins, T. L. and Baird, R. C. (1985) Aspects of the trophic ecology of the mesopelagic fish Lampanyctusalatus (Family Myctophidae) in the eastern Gulf of Mexico. Bioi. Oceanogr. 3, 285-313. Horridge, G. A. and Boulton, P. S. (1967) Prey detection by Chaetognatha via a vibration sense. Proc. Roy. Soc. B., 168, 413-419. Hu, V. J. H. (1978) Relationships between vertical migration and the diet of four species of euphasiids. Limno/. Oceanogr., 23, 296--306. Hwang, J.-S. (1991) Behavioral responses and their role in prey/predator interactions of a calanoid copepod, Centropages hamatus, under variable hydrodynamic conditions. Ph.D. Thesis, Boston University. pp. 162 Kerfoot,W. C., Kellogg, D. L. J. and Strickler, J. R. (1980) Visual observations on live zooplankter: evasion, escape, and chemical defenses. In: Evolution and Ecology of Zooplankton Communities., W. C. Kerfoot (ed), University Press of New England, Hanover, pp. 10-27. Kettle, D. and O'Brieh,W. J. (1978) Vulnerability of Arctic zooplankton species to predation by small lake trout ~alvelinus namaycush). 1 Fish. Res. Bd. Canada., 35, 1495-1500. Kirk, K. L. and Gilbert, J. J. (1988) Escape behavior of Polyarthra in response to artificial flow stimuli. Bull. Mar. Sci., 43, 551-560. Landry, M. R. (1978) Predatory feeding behavior of a marine copepod, Labidocera trispinosa. Limnol. Oceanogr., 23, 1103-lll3. Landry, M. R. (1980) Detection of prey by Calanus pacificus: Implication of the first antennae. Limnol. Oceanogr. 25,545-549. Landry, M. R. (1981) Switching between herbivory and carnivory by the planktonic marine copepod Calanus pacificus. Mar. Bioi., 65, 77-82. Lauder, G. V. (1980) The suction feeding mechanism in sunfishes (Lepomis): An experimental analysis. 1 Exp. Bioi., 88, 49-72. Lauder, G. V. and Clark, B. D. (1984) Water flow patterns during prey capture by teleost fishes. 1 Exp. Bioi., 113, 143-150.

COPEPOD ESCAPE FROM FLUID DISTURBANCE

339

Lenz, P. H. and Yen, J. (1993) Distal setal mechanoreceptors of the first antennae of marine copepods. Bull. Mar. Sci., 53, 170-179. Liem, K. F. (1980) Adaptive significance of intra- and interspecific differences in the feeding repertoires of cichlidfishes. Amer. Zoo/., 20,295-314. Lonsdale, D. J., Heinle, D. R. and Siegfried, C. (1979) Carnivorous feeding behavior of the adult calanoid copepod Acartia tonsa Dana. J. Exp. Mar. Bioi. Ecol., 36, 235-248. Luecke, C. and O'Brien, W. J. (1981) Prey location volume of a planktivorous fish: A new measure of prey vulnerability. Can. J. Fish. Aquat. Sci., 38, 1264-1270. Motta, P. J. (1984) Mechanics and functions of jaw protrusion in teleost fishes: A review. Copeia, 1, 1-18. Mullin, M. M. (1966) Selective feeding by calanoid copepods from the Indian Ocean. Some Contemporary Studies in Marine Science. 545-554. Paffenhiifer, G.-A. and Lewis, K. D. (1990) Perceptive performance and feeding behavior of calanoid copepods. J. Plankt. Res., 12, 933-946. Sokal, R. R. and Rohlf, F. J. (1981) Biometry. 2nded.,Vol.l.W. H. Freeman and Company, New York. 859 pages. Strickler, J. R. (1975) Swimming of planktonic Cyclops species (Copepoda, Crustacea): Pattern, movements and their control. In Swimming and Flying in Nature, T. Y. T. Wu, C. J. Brokaw and C. Brennen. (eds). Plenum Press, Princeton. 613 pages. Strickler, J. R. (1982) Calanoid copepods, feeding currents, and the role of gravity. Science, 218, 158-160. Strickler, J. R. (1985) Feeding currents in calanoid copepods: Two new hypotheses. In: Physiological Adaptations ofMarine Animals. M. S. Laverack (ed).The Company of Biologist Limited, Cambridge. Strickler, J. R. and Bal, A. K. (1973) Setae of the first antennae of the copepod Cyclops scutifer (Sars.): Their structure and importance. Proc. Nat/. Acad. Sci. USA., 70, 2656-2659. Strickler, J. R. and Twombly, S. (1975) Reynolds number, diapause, and predatory copepods. Int. Rev. Gesamten Hydrobiol., 19, 2943-2950. Szlauer, L. (1964) Reaction of Daphnia pulex De Geer to the approach of different objects. Pol. Arch. Hydrobiol., 12,5-16. Tautz, J. (1979) Reception of particle oscillation in a medium - an unorthodox sensory capacity. Naturwissenschaften, 66, 452-461. Tiselius, P. and Jonsson, P. R (1990) Foraging behavior of six calanoid copepods: observations and hydrodynamic analysis. Mar. Ecol. Prog. Ser., 66, 23-33. Vinyard, G. L. (1980) Differential prey vulnerability and predator selectivity: Effects of evasive prey on bluegill (Lepomis macrochirus) and pumpkinseed (L. gibbosus) predation. Can. J. Fish. Aquat. Sci. 37, 2294-2299. Vogel, S. (Ed.) (1981) Life in Moving Fluids: the Physical Biology ofFlow. lst ed.Vol. 1.Willard Grant Press, Boston. 352 pages. Yen, J. and Fields, D. M. (1992) Escape responses of Acartia hudsonica (Copepoda) nauplii from the flow field of Temora longicornis (Copepoda). Arch. Hydro bioi. Beih. Ergebn. Limnol., 36, 123-134. Yen, J., Lenz, P. H., Gassie, D. V. and Hartline, D. K. (1992) Mechanoreception in marine copepods: Electrophysiological studies on the first antennae. J. Plankt. Res., 14, 495-512. Yen, J., Sanderson, B., Strickler, J. R., and Okubo, A. (1991) Feeding current and energy dissipation by Euchaeta rimana, a subtropical pelagic copepod. Limnol. Oceanogr., 36,362-369.

PHYSIOLOGICAL AND BEHAVIORAL STUDIES OF ESCAPE RESPONSES IN CALANOID COPEPODS DANIEL K. HARTLINE, PETRA H. LENZ and CHRISTEN M. HERREN* Bekesy Laboratory of Neurobiology, University of Hawaii, Honolulu, HI 96822 USA and *Department ofBiology, University of South Carolina, Columbia, SC 29208 USA email: danh@uhunix. uhcc. hawaii. edu Electrophysiological techniques have been applied to monitoring sensory discharges from the first antennae of calanoid copepods. Extracellular nerve impulse traffic from both mechanoreceptors and putative chemoreceptors has been recorded. The first antennae of some, but not all, calanoid groups possess "giant" mechano receptive axons generating very large (mV) extracellular signals. There are two such giant antenna! mechano receptors (GAMs) innervating setae of each distal tip. These are sensitive to small ( < 10 nm) controlled hydrodynamic disturbances, including abrupt displacements and sinusoidal vibrations with frequencies up to and exceeding 2 kHz. Behavioral studies show that escape "jumps" can be triggered in Labidocera madurae by the same types of disturbances. Sensitivities as low as 4 nm were observed at frequencies of ca. 900 Hz. Behavioral sensitivities are similar to those measured physiologically and suggest that firing of the GAMs is capable of triggering escape behavior, perhaps even with a single nerve impulse.

INTRODUCTION Predation pressure in the pelagic environment is high. Zooplankton, including pelagic copepods, are preyed upon by many fish species, pelagic invertebrate predators (cnidarians, ctenophores, chaetognaths, copepods), and benthic predators (corals, sea anemones, barnacles). Copepods have a characteristic escape jump, which can effectively lower predator capture efficienCies (e.g., Browman et al., 1989; Trager et al., 1994). Involvement of sensory systems in this behavior would be expected to include detection of a predator, discriminating it from irrelevant signals, and localizing it. Behavioral studies on copepods have shown that detection of approaching predators can occur at a distance (e.g., Strickler, 1975; Kerfoot et al., 1980; Browman et al., 1989). Predator avoidance behaviors, such as escape "jumps", can be elicited by mechanical stimulation (Haury eta/., 1980; Gill, 1985; Gill and Crisp, 1985; Costello et al., 1990). Copepods can detect and react to both predatory lunges (e.g. Drenner et al., 1978; Wright and O'Brien, 1984; Kils, 1992) and flow fields of predators (Tiselius and Jonsson, 1990; Yen and Fields, 1992). Understanding the mechanosensory substrate of this escape system would give substantial insight into the functioning of a key factor in copepod survival. In deducing what information might be available to a copepod through its sensory systems, past work (e.g. Price et al., 1988; Marine Zooplankton Coloquium I, 1989) has had to rely on assumptions derived from physiological studies on large, mostly benthic, crustaceans from very different taxa (e.g. Laverack, 1963; Mellon, 1963; Tazaki and Ohnishi, 1974; Bush and Laverack, 1982; Ache, 1982; Sandeman, 1989). The issue of whether such assumptions could withstand direct scrutiny led to electrophysiological studies in our laboratory on sensory neurons in the first antennae of calanoids. Despite the small size of the animal, neural responses to controlled stimulation can be recorded 341

342

DANIEL K. HARTLINE ET AL.

successfully (Yen et al., 1992; Lenz and Yen, 1993). This approach has opened the way to analyzing the physiological substrates of behavioral sensitivity of the animals to stimuli which are crucial to finding food and mates, and avoiding or escaping from predators. In focusing initially on the latter behavior, we find involvement of mechanosensors with properties differing markedly from those inferred from other crustaceans. We will review some of our past work in this area and present some recent results, especially relating to behavioral correlates. METHODS Physiological Experiments

In most of the physiological work to be described, past and present, electrical potential differences have been recorded between the body of a copepod drawn up into an insulating layer of oil, and the sea water bath into which one of the first antennae is left projecting. This leaves the distal half to two-thirds of the sensory setae, both mechanoreceptive and chemoreceptive, exposed to the sea water under fairly "natural" conditions. The recording point can be controlled by varying the position of the oilwater interface along the antenna. The experiments to be described have focused on mechano-sensitivity in these setae, although we have also succeeded in recording from presumptive chemosensory elements with the same configuration. Controlled mechanical stimulation was produced by somewhat different means in different experiments. The most recent of these used the methods of Gassie et al., (1993), as diagrammed in Figure 1. A computer-controlled piezoelectrically-driven sphere generates near-field water disturbances the magnitude and direction of which can be calculated using classical dipole equations (Bergeijk, 1967; Kalmijn, 1988). In the particular cases we have employed so far, the sphere is located directly in front of the animal, with its axis of movement parallel to the shaft of the antenna. This produces an opposite movement of water at the antennal tip which can be calculated from the equation:

(1) where d is the displacement of the water at the antenna, D is the displacement of the sphere, a is the radius of the sphere, and r is the distance from the center of the sphere to the antenna. In the frequency ranges we have studied (100-2500 Hz) the effect of the boundary layer on the effective radius of the sphere is small and has been ignored. In another electrophysiological approach, the antennal nerve of Gaussia princeps was exposed by cutting a small window in the ventral exoskeleton at the base of the antenna. A suction electrode was lowered through the opening and applied to the surface of the antenna} nerve in this region. "En passant" recordings were thereby obtained of sensory nerve impulses ("spikes") travelling into the CNS. Electrical stimulation of motor axons in the same trunk was possible. · Behavioral Experiments

L. madurae were collected weekly using 2 to 5 min. subsurface horizontal tows (333 J-Lm mesh plankton net) in shallow ( < 10 m depth) nearshore areas off Oahu (Kaneohe Bay

COPEPOD ESCAPE

343

Forceps Electrode

"

• trig '---

Computer

~A!ol-...;;,;;;.-
1 hr). Examination of antenna! setae with the scanning electron microscope revealed that all setae of the type that stained with methylene blue bear distal pores (Figure lB). Access of

SPECIALIZATION ALONG COPEPOD ANTENNAE

357

A

0.1 mm Figure 1 A Drawing of segments 6 to 10 (ancestral segments: VIII to XIII; nomenclature ofHuys and Boxshall 1991) showing distribution of setae and pattern of methylene blue staining of neurons (dense stippling). Azocarmine-staining aesthetasc-like setae indicated by light stippling. Right antenna of adult female Pleuromamma xiphias.

Figure 18 Scanning electron micrograph of apical pore of a tapered seta. Proximal seta located on segment I (ancestral segment: III) on the left antenna in an adult male P. xiphias.

PETRA H. LENZETAL.

358

·A

B

c

1.0

pm

Figure 2 Transmission electron micrographs of three types of setae in Pleuromamma xiphias. A. Crosssection through the proximal spiniform seta on segment 17 (ancestral segment: XX; nomenclature of Huys and Boxshall 1991) on right antenna of an adult male. Section was taken near its base. B. Cross section of mechanoreceptive seta on distal segment (seta# 3, nomenclature of Weatherby eta/. 1994; ancestral segment: XXVIII) on right antenna of an adult female. Section was taken near the base of the seta. C. Cross section through aesthetasc-like seta on segment 22 (ancestral segment: XXV) on right antenna of an adult female. Section was taken near the tip of the seta. Small circular profiles are sections through distal chemoreceptive dendrites. Arrowheads= chemoreceptive dendrites; cu =cuticle; mt = microtubular arrays of mechanoreceptive dendrites.

SPECIALIZATION ALONG COPEPOD ANTENNAE

359

dye to the nerve cells may occur via this distal pore in these setae. All spiniform setae located on the distal tip except one lack distal pores (Weatherby et al., 1994) and did not stain with methylene blue. Examination by transmission electron microscope of cross-sections through spiniform setae of the antenna! shaft shows the presence of, usually, two mechanoreceptive dendrites identified by a highly elaborated microtubular complex (Figure 2B). Innervation with a single mechanoreceptive dendrite was observed in the distal tip seta bearing the apical pore (see also Weatherby et al., 1994). In the setae with an apical pore, several smaller dendrites are present as well, located peripherally, between the enveloping cell(s) and the external cuticle (Figure 2A,B). Compared to the mechanosensory dendrites, these peripheral dendrites contain few micro tubules. Near the base of the spiniform seta these micro tubules can be seen to be organized in the characteristic 9+0 ciliary arrangement. The nerve cells giving rise to these smaller dendrites appear to be the ones staining with methylene blue since their cell bodies are located close to the base of the seta (Figure 1A). In contrast, the cell bodies of the mechanosensory neurons are located more proximally within the antenna! shaft (Weatherby and Lenz, 1993; Weatherby et al., 1994). These sensory cells are likely to be chemosensory and their structure is similar to the chemosensory dendrites found in the aesthetasc-like setae. Thus, the spiniform setae appear to have a mixed-modality chemo- /mechanosensory function. Purely mechanosensory setae were found in the distal region of the first antenna. These setae are innervated by mechanosensory dendrites, but they lack the smaller putative chemosensory dendrites. The internal and external structure of the mechanosensory setae on the distal tip (segments 23 and 24) have been described elsewhere (Weatherby and Lenz, 1993; Weatherby et al., 1994). In addition to the 7 mechanosensory setae on the last two segments, feathered posteriorly-pointing mechanosensory setae are found on segments 21 and 22 (Table 1). The aesthetasc-like setae were stained only lightly or not at all by methylene blue in live preparations, but stained within minutes of application of azocarmine (Figure 1A). Azocarmine stained only setae and not more proximal parts of sensory neurons. The dye appeared to diffuse directly through the thin cuticle (Figure 2C), and no pores could be seen using scanning electron microscopy. TEM cross sections through these setae show many small, presumably chemosensory, dendrites throughout the lumen (Figure 2C). This type of seta has been identified as chemosensory in other calanoid copepods (e.g., Gill, 1986; Kurbjeweit and Buchholz, 1991; Bundy and Paffenhofer, 1993). Two additional types of aesthetasc-like setae are found on the first antennae of males: large bulbous setae on the right, geniculated antenna, and setae of intermediate diameter on the left antenna. Both types were stained by azocarmine but not by methylene blue. Table 1 shows a segmental account of all setal types on the first antenna of females: of the 80 identified setae on the first antenna, 60% are mixed-modality, possessing a distal pore and innervated by both mechano- and putative chemosensory neurons (Table 1). Aesthetasc-like chemosensory setae account for 29% of the total. Pure mechanosensory setae, showing no evidence of chemosensory dendrites within the setae and lacking a distal pore, account for 11% of the setae. The latter are primarily found distally and they include three posteriorly-pointing feathered setae (Table 1).

360

PETRA H. LENZETAL

Table 1 Distribution of identified sensory setae on the first antenna of Pleuromamma xiphias females. Segment

Ancestral segment*

Chemo-

I 2 3 to 6 7 8 9 to 18 19 20 21 22 23 24 Antenna

I-III IV VtoVIII IX-X XI XII to XXI XXII XXXIII XXIV XXV XXVI XXVII-XXVIII

3 1 I 2 1 1

1 23

Mechano-

1 1 2 5 9

Mixed modality chemolmechano

7** 2 2 4 2 2 1 I 1 1 1 48

* Nomenclature by Huys and Boxshall (1991) ** Two of the setae are modified, having short spines

DISCUSSION The copepod sensory structures share many of the general characteristics of mechano-, chemo-, and mechano-/chemoreceptors in other arthropods. Receptor function was inferred from morphology using other studies on crustacean and insect receptors (e.g., Mciver, 1985; Hatt, 1986; Laverack, 1988; Schmidt and Gnatzy, 1989). Using morphological criteria, we found three primary types of sensory setae: mechanoreceptors, aesthetasc-like chemoreceptors and mixed-modality mechano-/ chemoreceptors. The mechanoreceptors are characterized by a thick cuticle, very rigid cellular structures, and a large proliferation ofmicrotubules (see Weatherby and Lenz, 1993; Weatherby eta/., 1994 for serial-section analysis). These purely mechanoreceptive setae are located on the most distal segments. The aesthetasc-like chemoreceptors are characterized by a thin cuticle, uptake ofazocarmine dye and the presence of multiple thin dendrites. Serial sections through this type of chemoreceptor show that these small chemosensory dendrites branch at least 2 times just prior to entering the lumen of the seta (Weatherby eta/., 1994). These setae are likely to be involved in distance perception of odorants. In herbivorous copepods, it has been suggested that these setae may detect algal exudates traveling in advance of the cell (Strickler, 1985). The bulbous chemoreceptors found in males may be detecting pheromones released by females (Katona, 1973; Griffiths and Frost, 1976). Mixed-modality mechano-/chemosensory setae are the most common type of sensilla on the first antenna on P xiphias. Identification of the spiniform setae as mixed-modality sensilla is based on three lines of morphological evidence: 1) uptake of vital dye from the surrounding medium in live animals which indicates ready accessibility of the neurons to chemicals in the external medium; 2) the presence of apical pores which provides for the physical access route; and 3) the presence of two types of dendrites in the lumen of the setae: rigid, microtubule-packed

SPECIALIZATION ALONG COPEPODANTENNAE

361

mechanosensory dendrites and smaller dendrites with features typical of chemosensory neurons (see Ache, 1982). Similar mechanosensory dendrites have been found in pure mechanosensory setae on the antenna! tip (Weatherby eta/., 1994). Chemosensory dendrites are characterized by having a smaller size and smaller number of micro tubules than those of mechanosensory cells. Although this study has focused on P xiphias, mixed-modality setae may be present on the first antennae of other copepods. Gresty et a/. (1993) describe small dendritic profiles as well as larger microtubule-packed dendrites in setal cross sections of the caligid, Lepeophtheirus salmonis. Setae of mixed-modality have been described in many crustaceans, as well as other arthropods (e.g., Altner et al., 1983; Schmidt and Gnatzy, 1984; Hatt, 1986). The link between such setae and a gustatory sensory mode is well established in insects (e.g., review by Meisami, 1991). In general, it has been proposed that in crustaceans as well, mixed-modality sensilla are the typical type of "taste" or gustatory receptor, with chemical sensing occurring upon contact concurrent with detection of the mechanical contact signal (see reviews by Ache, 1982; Laverack, 1988). These receptors are characterized by limited access of odorant molecules to the dendrites as well as low chemosensitivities (Schmidt and Gnatzy, 1989). In the copepods, odorant access appears to be limited to a single, small ( < 0.5 J..Lm diameter) pore with an apical placement. Sensory specialization appears to occur along the antenna (Table 1, Figure 3). It has been previously noted that the distal setal mechanoreceptors are located in an area relatively unaffected by the flow fields generated by the beating appendages (Fields and Yen, 1993; Lenz and Yen, 1993; Hartline eta/., this volume). This location is consistent with the hypothesis that these sensory setae play an important role in predator detection (e.g. Gill, 1985; Lenz and Yen, 1993). What was unknown previously is that on the first antenna of P xiphias the pure mechanoreceptors are limited to the distal segments (Figure 3). In contrast, the highest densities of "olfactory" and "gustatory" receptors are found in the high flow velocity area. Behavioral decisions on potential food items may occur in a two-step system based on sensory input from the first antenna. The aesthetasc-like setae may sense chemical signals ("olfaction") at a distance, and thus be involved in re-routing of prey entrained in feeding currents prior to contact (Strickler, 1984, 1985). Water movements produced by prey may contribute to this distance detection via the distal antenna! mechanoreceptors. In P. xiphias if the prey then contacts the antenna, the mixed-modality setae may taste the prey item through the two associated chemoreceptive neurons, simultaneously registering the contact through the two mechanoreceptors, triggering behavioral reactions to either reject it or pass it along to the feeding appendages. The situation may be analogous to the gustatory receptors on the dactyls of crabs (Schmidt and Gnatzy, 1989). It has been hypothesized that the crab mixed-modality receptors provide the first physical contact with food and are involved in the preliminary decision whether an object should be further explored or ignored. A go-ahead signal from these receptors would then elicit a feeding behavior which would include sensory input from gustatory receptors associated with mouthparts and other feeding appendages. Acknowledgements

We would like to thank Drs. M. Dunlap and D. Hartline for continued support and scientific input throughout this study. We thank an anonymous reviewer for his/her

PETRA H. LENZ ETAL.

362

Distance (mm) R 1 - 5 mm·s-1

-E

< 1 mm ·s-1

E

~

c

·u; c

Q)

0

R

5

10

15

20

Segment(#) Figure 3 Density of sensory setae per antenna! segment in Pleuromamma xiphias adult females. X-axis: bottom - segment number; top- distance in mm from rostrum (R) to distal tip of antenna. Symbols- solid circles: chemosensory setae; solid triangles: mixed-modality mechano-/chemosensory setae; open circles: mechanosensory setae. Shaded bars indicate areas of different flow velocities at the antenna: dark stipplingvel. > 5 mm/s, intermediate stippling - vel. I to 5 mm/s and light stippling - < I mm/s. Flow velocity data redrawn from Figure 6 of Fields and Yen (1993) and scaled to a 3.3. mm long antenna.

careful review and thoughtful comments. Dr. I. Cooke graciously allowed us to use his compound microscope and photographic set-up. J. Labenia suggested the use of azocarmine, and B. Yuen advised us on the histological uses of dyes. We thank V. Hardy who shared his unpublished results with us. M. Aeder kindly collected and shipped P. xiphias from the Big Island. The Natural Energy Laboratory of Hawaii generously provided access to their deep-water pipes at Ke'ahole Point, Hawaii. This study was supported by NSF grant OCE 89-18019, Sea Grant award NA89AA-D-SG063, the Ida Russell Cades Fund of the University of Hawaii Foundation and University of Hawaii Project Development Grant to D. K. Hartline, and NIH grant RR03061 (RCMI). W. W. and K. K. W. were supported by the NSF Research Experience for Undergraduate Program (REU). References Ache, B. W. (1982) Chemoreception and thermoreception. In: The Biology of Crustacea. UJ!. 3. Neurobiology: Structure and Function, H. L. Atwood, D. C. Sandeman, eds., Academic Press, London, pp. 369-398. Altner, 1., Hatt, H., and Altner, H. (1983) Structural properties of mixed modality chemo- and mechanosensitive setae on the pereiopod chelae of the crayfish, Austropotamobius torrentium. Cell Tissue Res., 228,357-374. Bullock, T. H. and Horridge, G. A. (1965) Structure and Function in the Nervous Systems of Invertebrates, UJ!. l w.·H. Freeman Co, San Francisco. Bundy, M. H. and Paffenhiifer, G. -A. (1993) Innervation of copepod antennules investigated using laser scanning confocal microscopy. Mar. Ecol. Prog. Ser., 102, 1-14. Fields, D. and Yen, J. (1993). Outer limits and inner structure: the 3-dimensional flow field of Pleuromamma xiphias (Calanoida: Metridinidae). Bull. Mar. Sci., 53, 84-95.

SPECIALIZATION ALONG COPEPOD ANTENNAE

363

Gill, C. W (1985) The response of a restrained copepod to tactile stimulation. Mar. Ecol. Prog Ser., 21, 121-125. Gill C. W (1986) Suspected mechano- and chemosensory structures of Temora longicornis (Copepoda: Calanoida). Mar. Bioi., 93,449-457. Griffiths, A. M. and Frost, B. W (1976) Chemical communication in the marine planktonic copepods Calanus pacificus and Pseudocalanus sp. Crustaceana, 30, 1-9. Gresty, K. A., Boxshall, G. A., and Nagasawa, K. (1993) Antennulary sensors of the infective copepodid larva of the salmon louse, Lepeophtheirus salmonis (Copepoda: Caligidae). In: Pathogens of Wild and Farmed Fish: Sea Lice, G. A. Boxshall, D. Defaye, eds., Ellis Horwood Ltd, Chichester, pp. 83-98. Hartline, D. K., Lenz P. H., and Herren, C. (1995) Physiological and behavioral studies of escape responses in calanoid copepods. Mar. Fresh. Behav. Physiol., this volume Hatt, H. (1986) Responses of a mixed modality neuron (chemo-and vibration-sensitive) on the walking legs of the crayfish. J Comp. Physiol. A, 159, 611--617. Heimann, P. (1984) Fine structure and molting of aesthetasc sense organs on the antennules of the isopod, Asellusaquaticus (Crustacea). Cell Tiss. Res., 235,117-128. Huys, R. and Boxshall, G. A. (1991) Copepod Evolution. The Ray Society, Unwin Brothers, Old Woking, Surrey. Katona, S. K. (1973) Evidence for sex pheromones in planktonic copepods. Limnol. Oceanogr., 18, 574-583. Kurbjeweit, F. and Buchholz, C. (1991) Structures and suspected functions of antennular sensilla and pores of three Arctic copepods (Calanus glacialis, Metridia longa, Paraeuchaeta norvegica). Meeresforsch., 33, 168-182. Landry, M. R. (1980) Detection of prey by Calanus pacificus: implications of the first antennae. Limnol. Oceanogr., 25,545-549. Laverack, M. S. (1988) The diversity of chemoreceptors. In: Sensory Biology of Aquatic Animals, J. Atema, R. R. Fay, A. N. Popper,W N. Tavolga, eds., Springer-Verlag, New York, pp. 287-312. Lenz, P. H. and Yen, J. (1993) Distal setal mechanoreceptors of the first antennae of marine copepods. Bull. Mar. Sci., 53,170--179. Mciver, S. B. (1985) Mechanoreception. In: Comprehensive Insect Physiology, Biochemistry, and Pharmacology. J-016. Nervous System: Sensory, G. A. Kerkut and L. I. Gilbert, eds., Pergamon Press, Oxford, England, pp. 71-132. Meisami, E. (1991) Chemoreception. In: Neural and Integrative Animal Physiology, C. L. Prosser, ed.,Wiley-Liss, New York, pp. 335-434. Price, H. J., Paffenhofer, G. -A., Boyd, C. M., Cowles, T. J., Donaghay, P. L., Hamner, W M., Lampert, W, Quetin, L. B., Ross, R. M., Strickler, J. R., and Youngbluth, M. J. (1988) Future studies of zooplankton behavior: questions and technological developments. Bull. Mar. Sci., 43, 853-872. Schmidt, M. and Gnatzy, W (1984) Are the funnel-canal organs the campaniform sensilla of the shore crab, Carcinus maenas (Decapoda, Crustacea)? II. Ultrastructure. Cell Tissue Res., 237, 81-93. Schmidt, M. and Gnatzy, W (1989) Specificity and response characteristics of gustatory sensilla (funnel-canal organs) on the dactyls ofthe shore crab, Carcinus maenas (Crustacea, Decapoda). J Comp. Physiol. A, 166, 227-242. Strickler, J. R. (1984) Sticky water: a selective force in copepod evolution. In Trophic Interactions Within Aquatic Ecosystems, D. G. Meyers, J. R. Strickler, eds.,Westview Press, Boulder, Colorado, pp.l87-239. Strickler, J. R. (1985) Feeding currents in calanoid copepods: two new hypotheses. Soc. Exp. Bioi. Symp., 39, 459-485. Weatherby, T. M. and Lenz, P. H. (1993) Proliferation of microtubules from discrete electron-dense bodies in the ciliary dendritic portion of mechanosensory setae of copepod antennae. In: Proc. 51st Annual Meeting of the Microscopy Society of America, G. W Bailey, C. L. Rieder, eds., San Francisco Press, San Francisco, California, pp. 340-341. Weatherby, T. M., Wong, K. K., and Lenz, P. H. (1994) Fine structure of the distal sensory setae on the first antennae of Pleuromamma xiphias Giesbrecht (Copepoda). J Crust Bioi., 14,670-685.

OPTIMAL SWIMMING BEHAVIOR OF ZOOPLANKTON STANLEY I. DODSON Department of Zoology-Birge Hall University of Wisconsin, 430 Lincoln Drive, Madison, Wisconsin 53706 email: sidodson @facstaff.wise. edu Zooplankton live in an open, exposed environment, where their every move is an open invitation for predation. Faster swimming leads to higher predator-prey encounter rates and therefore an increase in the risk of predation. Predation is a major source of mortality for zooplankton, and therefore a potentially important agent of natural selection. I review ecological constraints on zooplankton swimming behavior and present a graphical cost-benefit model that incorporates: (I) Potential benefits of swimming rates to population growth rates in the absence of predators. (2) Costs of swimming rates due to the effect of prey swimming rates on encounter rates with two different pelagic predators: visual fish and tactile invertebrates. This conceptual and graphical model helps organize our understanding of the factors affecting foraging strategies of freshwater zooplankton. The cost-benefit model allows insight into zooplankton swimming behavior.

INTRODUCTION Zooplankton live in an open, exposed environment, where greater swimming rates increase the risk of predation (O'Brien, 1987). Encounter rate depends on the predator's reactive distance and swimming speed of the predator and prey (Confer and Blades 1975, Gerritsen and Strickler, 1977; Gerritsen, 1980; Riessen et al., 1988; Blais and Maly, 1994). For example, Daphnia swimming in populations of the predator Chaoborus experience mortality rates approximately the same as the maximum Daphnia reproductive rate (Riessen, 1992). Thus, there is a clear advantage to swimming as slowly as possible when predators are present (Threlkeld, 1987). Zooplankton are constrained to swim at least as fast as their sinking rate. However, zooplankton normally swim faster than their sinking rate over a wide range of conditions (e.g. Table 7 in Dodson and Ramcharan, 1991). The field of optimal foraging has addressed some of the tradeoffs and constraints that determine patterns of animal movement in a predator-prey context. Prey movement is influenced by risk of predation, multiple habitat patches and refuges, and search strategies (many of these studies are gathered together in Stephens and Krebs, 1986; and Gerking, 1994). However, optimal foraging theory has not yet addressed the question of optimal animal movement in an environment with no refuge. The purpose of this paper is to use what is known about zooplankton (especially Daphnia) swimming behavior and ecology to develop a cost-benefit model of optimal swimming behavior. 365

366

STANLEY I. DODSON

A. B. C. D. E. F. Ill

Infrared L.E.D. Collimating lens Light beam Observation chamber First surface mirrors Video camera Light path

Figure 1 Design of the video system. The camera is about 62 em from the observation chamber, which is about 5 em square.

MEASURING ZOOPLANKTON SWIMMING BEHAVIOR The predator-prey interaction depends on a number of factors, including the distribution and abundance of predator and prey, the sensory systems of the predator, morphological aspects of the predator and prey, ability of the prey to avoid and/or escape predators, and predator and prey activity. Prey behavior is often proposed as a crucial factor in the predator-prey equation, but is the factor about which we know the least (e.g. Buskey eta/., 1993). Measurement of the importance of prey activity has only recently become feasible, with the advent of appropriate video and motion analysis technology. Previous studies estimated 3-dimensional swimming speeds from 2-dimensional film records (e.g. Stearns, 1975; Buchanan eta!., 1982). While the extrapolation of swimming speeds from 2 to 3 dimensions is acceptable for many purposes, Hamner and Hamner (1993) remind us of the inherent loss of information resulting from the extrapolation. We measure individual 3-dimensional swimming behavior of free-swimming zooplankton using a combination of video and motion analysis techniques (Dodson and Ramcharan, 1991; Dodson eta!., 1995). Two simultaneous 2-dimensional video recordings are made using one camera and an arrangement of mirrors (Figure 1). The system of mirrors (originally designed by Ramcharan and Sprules, 1989) produces two images of the same animal on each frame of the video record. Because the two images are from orthogonal angles, each frame contains enough information to calculate the

OPTIMAL SWIMMING

367

3-dimensional position of the animal, using standard Pythagorean geometry. We have modified the original system to include collimated 915 nm infrared illumination (Dodson et al., 1995). The use of infrared light (invisible to the zooplankton, even when quite bright) allows behavior to be recorded at any desired level of visible light. The visible light comes from a "Snellius Circle" (vid. Ringelberg, 1987) of translucent white plastic illuminated with an incandescent light source suspended directly above the observation chamber. We record behavior in a smaller (183 ml) or larger (1160 ml) approximately cubical observation chamber. Video data are collected at a rate of 30 frames sec- 1, for at least 3 sec. Video records are digitized using an ExpertVision computer system. The 3-dimensional positions can then be analyzed to produce estimates of a number of swimming parameters. We currently measure: 1) Swimming speed: The average speed is measured as the total distance travelled divided by the elapsed time. The total distance travelled is scale-dependent; we measure time in units of 0.033 sec and a minimum spatial resolution of about 0.034 mm (using the smaller observation chamber and the highest magnification). At average Daphnia swimming speeds, a 3 sec track is about 1.5 to 3 em long. The video records include resolution of the hops, jumps and glides characteristic of the small-scale (mm-cm) behaviors expressed by different groups of zooplankton (Ramcharan and Sprules, 1991; Dodson and Ramcharan, 1991, Buskey eta/., 1993). Note that previous studies (e.g. Buchanan and Goldberg, 1981) collected data at a coarser scale, producing tracks that leave out the mm-scale movement, and therefore produce lower swimming speeds. Variance in swimming speed is estimated using the distance travelled between each successive frame. If swimming speed is strongly canalized (as indicated by results of studies by Fox and Mitchell 1953, Buchanan and Goldberg 1981, Porter et al. 1982, Dodson et al. 1995), then it is possible that effects of environmental stresses, such as suboptimal temperature or water chemistry, toxins, predator smells, or food limitation, will manifest as an increase in variance rather than a change in the average speed. 2) Sinking velocity: The average sinking velocity can be estimated from the short sinking bouts which are part of swimming behavior, because zooplankton reach terminal velocity within a few msec- 1 (Dodson and Ramcharan, 1991). The video measurement of sinking rate allows observation of sinking oflive and active animals swimming in their natural orientation. 3) Vertical angles: Sinking is in a downward direction and the recovery power stroke is in an upward direction. The average upward and downward angles can be measured (e.g. Dodson and Ramcharan, 1991). It has been proposed that the angle of movement is a critical component of swimming behavior, because (i) of different swimming patterns during feeding in response to different colors of light (Smith and Baylor, 1953; Stearns, 1975), (ii) of differential predation risk related to horizontal and vertical swimming patterns and their effect on visual perception by fish (Ingle, 1968), and (iii) because the angle of swimming affects the ability of zooplankton prey to escape fish once detected (Jacobs, 1967; Kerfoot, 1978; Drenner et al., 1978; Zaret, 1980; Wright and O'Brien, 1984; O'Brien, 1987). 4) Small-scale erratic behavior: Some zooplankton exhibit small-scale swimming behaviors, such as hops or jerks (e.g. Daphnia and related pelagic cladocerans: Dodson and Ramcharan, 1991; Buskey eta/., 1993), glides (e.g. pelagic copepods: Ramcharan and Sprules, 1991), escapes (Alcaraz and Strickler, 1988), stop-and-go swimming

368

STANLEY I. DODSON

(Buskey et al., 1992), or walking through water (Kerfoot eta/., 1980). Erratic behavior can be quantified as the average and variance of the frame-to-frame change in direction of swimming. A complete understanding of zooplankton ecology includes an understanding of the adaptive significance of swimming behavior. Field and lab evidence suggests that zooplankton morphology, life history, and behavior are shaped by natural selection (Kerfoot and Peterson, 1979; Parejko and Dodson, 1991; Larsson and Dodson, 1993). Recent reports of Daphnia movement demonstrates that what we know about optimal foraging, physiology, functional morphology, and zooplankton ecology is insufficient for understanding zooplankton swimming behavior as an adaptive strategy. For example: 1) Swimming speed and body size: Dodson and Ramcharan (1991) predicted that large Daphnia would swim faster than small Daphnia, proportional to length. However, results of the video observations showed that Daphnia pulex (clone SBL) of different sizes swam at nearly the same rate. This independence oflocomotion rate from body size suggested that swimming behavior was being optimized, perhaps due to constraints imposed by the effects of swimming speed on feeding and predation risk. 2) Sinking rate and swimming speed: Zooplankton tend to sink when not actively swimming upwards (Dodson and Ramcharan, 1991). Daphnia sink at a rate of about 3 mm sec- 1. To stay in the same place, they therefore need to swim on average about 3 mm sec- 1. Daphnia swim 2-3 times faster than their sinking rate. This excess of movement over sinking rate increases encounter rate with predators. Faster swimming also consumes energy which could otherwise be allocated toward increasing population growth rate. 3) Swimming speed, hopping rate and food density: Swimming speed tends to be independent of food concentration (Fox and Mitchell, 1953; Buchanan and Goldberg, 1981; Porter et al., 1982) when Daphnia are observed with visible light. Larsson and Kleiven (this symposium) also report swimming speed is independent of food concentration in the light, but that faster swimming occurs in the dark at low food concentrations. Fox and Mitchell (1953) and Porter et al. (1982) reported no effect of food concentration on Daphnia hopping rate. Since zooplankton are thought to often be food-limited, this independence of swimming behavior to food concentration is surprising. 4) Swimming rate and predator smell: Dodson eta/. (1995) predicted that Daphnia swimming velocity would decrease in the presence of Chaoborus, an invertebrate predator. Standard encounter rate models suggest a slower swimming velocity would increase survivorship. While several studies have demonstrated that Daphnia do respond to Chaoborus smell by changing their vertical position in the water column (Dodson, 1989; Ringelberg 1991), the clone we studied did not change swimming velocity when exposed to Chao horus smell (Dodson eta/., 1995). The failure to swim slower when predation risk increases is surprising, and may indicate a benefit (perhaps related to feeding) of swimming that outweighs the increase predation risk.

In summary, results of studies of Daphnia swimming behaviors identifY several situations in which the behavior was anomalous or unexpected, given what is known about Daphnia ecology. These results present an opportunity for the development of a systematic cost-benefit model that can be used to explore the trade offs related to prey activity.

OPTIMAL SWIMMING

369

DEVELOPMENT OF THE COST-BENEFIT MODEL OF ZOOPLANKTON SWIMMING BEHAVIOR: Potential Benefits of Swimming Rates to Population Growth Rates in the Absence of Predators

There are a number of documented benefits to large-scale swimming behavior: diel vertical migration, location of food patches, and escape responses (Larsson and Dodson, 1993). These behaviors are expressed on a scale of em-to-m. At a smaller scale, there is an advantage to swimming at the sinking rate. In order to stay in the plankton, zooplankton must swim upward, on the average, at least as fast as their sinking rate. However, at least some zooplankton swim 2-3 faster than their sinking rate. When an animal is in a patch of food, and has migrated vertically to avoid predators, then what is the advantage of additional swimming velocity?. The excess velocity, above sinking rate, suggests there is an additional advantage to zooplankton swimming. A possible advantage of swimming is a relationship between population growth rate and swimming speed. Daphnia feeding rate (correlated with "b", the population birth rate coefficient) has been shown to be a function of food concentration (Porter et al., 1982, reviewed in Lampert, 1987) and body size (Lehman, 1976), but there have been no measurements of Daphnia feeding or reproductive rates as a function of swimming velocity. Currently it is unknown whether Daphnia reproductive rate changes with swimming rate. Figures 2A and 2B are two possible shapes of the function of "b" over a range of zooplankton swimming speeds. Figure 2A is the curve expected if there is no benefit to "b" from swimming faster, but only various (perhaps metabolic) costs (see Porter et al., 1982). Figure 2B is a curve representing a benefit to faster swimming, a benefit that is eventually cancelled by the high metabolic cost of fast swimming (Alcaraz and Strickler, 1988). The cost-benefit model requires an empirical study of the relationship between population birth rate and the average swimming velocity. One way to keep zooplankton swimming at a variety of speeds is to culture them in a flow-through chamber such as that used in a study of diel vertical migration by Dawidowicz and Loose (1992). The average swimming rate of the plankton depends on the rate at which water flows through the chamber. A second approach is to measure the reproductive rates of individual animals that vary in their life-time average swimming speed. Life table experiments using individual Daphnia with different life-long average swimming speeds may provide an alternate way to look at the effect of swimming speed on reproductive performance. Costs of Prey Swimming Rates due to the Effect ofEncounter Rates with Two Different Pelagic Predators: Visual Fish and Tactile Invertebrates

The two kinds of predators represent two major types of predation on zooplankton (Dodson, 1974b; Threlkeld, 1987). A critical difference between the two types of predation is the role of light. Fish need light to be effective predators and invertebrate tactile predators tend to have the same feeding rates in both dark and light. To understand zooplankton swimming behavior in its ecological context of day and night and vertical light gradients, it is necessary to include both kinds of predation. Realistic encounter rate curves can be generated by (i) using existing encounter rate models, and (ii) modifying the predicted encounter rates to take into account specific prey swimming behaviors (such as jumps and hops).

STANLEY I. DODSON

370

-

A

B

C1)

ca

a:

J:

t::

Ill

Prey Swimming Speed

C1)

ca

a:

>

·cat:: 0

~

Prey Swimming Speed

Figure 2 Graphs A and B: Possible shapes of the function of the birth ("b") function over the range of cladoceran swimming speeds. Graphs C and D: Possible relationships between mortality rate ("d") and swimming speed of cladocerans.

Application ofexisting encounter rate models: Species and age-specific mortality rates can be estimated using encounter rate theory and data in the literature (e.g., Dodson, 1974a). The age-specific mortality rates can then be used in a standard life-table calculation to estimate population death rates for any specific velocity. We know the general shape of the mortality vs. swimming velocity curve from encounter rate curves developed by Gerritsen and Strickler (1977), Gerritsen (1980), Riessen eta!. (1988) and Blais and Maly (1994). Encounter rate depends on the type of predator and the swimming velocities of both the predator and prey (Gerritsen and Strickler, 1977; Evans, 1989). If the predator has a sit-and-wait strategy (i.e. velocity= zero), then the encounter rate is a linear function of only the prey velocity (Figure 2C). The linear curve in Figure 2C is characteristic of larval Chaoborus (the phantom midge, a common sit-and-wait predator in ponds and some lakes). Figure 2D has a

OPTIMAL SWIMMING

371

concave shape characteristic of a system in which both predator and prey have swimming velocities greater than zero (see Figure 2 in Gerritsen and Strickler, 1977). For low prey velocities, encounter rate is constant because it depends mainly on a constant predator velocity; at high prey velocities, encounter rate depends mainly on prey velocity and the rate therefore increases linearly with prey velocity. Figure 2D is characteristic of cruising predators such as copepods, the cladoceran Leptodora, and small fish. The average encounter a lake or pond Daphnia experiences is probably more like that in Figure 2D, because lakes typically have a mixture of mostly cruising predators. Modification of encounter rates to take into account specific prey swimming behaviors: Existing predator-prey encounter rate models treat zooplankton swimming behaviors at a scale of centimeters to tens of meters (e.g. Gerritsen and Strickler, 1977; Riessen et al., 1988; Riessen, 1992; Dodson, 1990; Blais and Maly, 1994). Ecologically important components of swimming behavior exist also on a smaller scale (approximately a body length). Distinct swimming behaviors characteristic of groups of zooplankton include: (i) smooth gliding, interrupted periodically by sudden, rapid jumps (e.g., calanoid copepods and Diaphanosoma: Ramcharan and Sprules, 1991), (ii) continuous hopping (e.g., Daphnia), (iii) "walking" through the water, punctuated by jumps (e.g., cyclopoid copepods: Strickler, pers. com.; Kerfoot et al., 1980; Williamson, 1983), and (iv) cruising, (e.g., the cladocerans Bosmina and Chydorus: Kerfoot, 1978; or smaller zooplankton such as rotifers and protozoans: Epp and Lewis, 1984). These small scale swimming behaviors may affect overall prey conspicuousness, and subsequently mortality due to invertebrate predators or fish. The lack of research performed to determine the role of small scale swimming behavior in predator-prey interactions stems particularly from an inability to control prey swimming behavior in predation trials. It is impossible to coerce a given species or individual to swim in an exact, replicable manner. This problem can be addressed using a video system in which simulated prey on a computer monitor are presented to predators in an aquarium glued to the monitor (Brewer and Coughlin, this volume). With this computerized system, fish can be observed preying upon simulated prey or "virtual" plankton. The number, appearance and motion of virtual plankton can be precisely controlled with a computer program. Studies of the reaction of predators to "virtual plankton" will allow modifications of mortality curves to include the effects of individual small-scale swimming behavior (Brewer and Coughlin, this volume). The mortality curve: Encounter rates can be used to estimate age-specific probabilities of survival ("s/'). Survival probability is a function of encounter rate, the probability of escaping once encountered ("c/'), and the time interval (one day). The probability of escaping once encountered can be estimated from values in the literature for the specific predator-prey system (e.g., Drenner et al., 1978; Havel and Dodson, 1984; Riessen eta!., 1988; Riessen, 1992; and Blais and Maly, 1994). The value of ci is typically close to zero for fish, and may be nearer 0.5 to 1.0 for invertebrate predators. The agespecific survival probilities can then be used to calculate the population death rates, using standard life history techniques (e.g., Dodson, 1974a). The Cost-Benefit Model

The model is a cost-benefit comparison that predicts swimming speed for each type of prey species at a given food level, depending on the species-specific functions for population growth rate and mortality. In the graphical model, the "b" (recruitment

372

STANLEY I. DODSON

--

b

~ ca

b

"C Cl)

ca

a:

0

3

7

0

3

7

0

3

7

Swimming Speed (mm sec-1) Figure 3 Possible configurations of the cost-benefit model of optimal zooplankton swimming behavior. The birth rate curve ("b") is the same as in Figure 2B. Overlaying the birth curves are three possible mortality ("d") curves. The swimming speeds are derived from our observations of Daphnia, in which the sinking rate is about 3 mm sec 1 and the maximum (non-escape) speed is about 7 mm sec 1.

with no mortality) and "d" (mortality) functions are plotted on the same scale (Figure 3). Optimal swimming speed is predicted by the greatest positive difference between the two curves. Three possible configurations of the model are presented in Figure 3, which matches three different mortality ("d") curves (variations of Figure 2D) against the birth rate curve from (Figure 2B). In Figure 3A, the mortality curve lies mostly to the left of the birth rate curve, and crosses that curve to the left of the sinking rate. In this case, the optimal swimming velocity of a planktonic animal would be the sinking rate (about 3 mm sec- 1 in this example), but even this speed produces an excess of mortality over birth rate. The configuration in Figure 3A can be interpreted as strong selection for a littoral life. At the other extreme, shown in Figure 3C, the mortality curve is flat over the swimming range associated with a positive birth rate. In this case, selection would be for the swimming speed that maximizes birth rate (at about 7 mm sec 1 in this example). In the intermediate situation (Figure 3B), the mortality and birth curves overlap, with the curves crossing near the peak of the birth curve. In this case, the greatest positive difference between the two curves (greatest net population growth rate) is between the sinking rate and the swimming speed that gives the highest birth rate (in the absence of predation). We suspect this intermediate model is the most realistic, but the model now needs to be tested in the lab and field. The model developed through this proposed research can predict zooplankton swimming behavior for cladocerans in a number of different habitats. More generally, the model can be applied to any group of marine or freshwater zooplankton. The model is applicable to both freshwater and marine zooplankton and it is not restricted to water; it has applications in terrestrial systems as well, wherever prey in a homogeneous environment experience both a cost and a benefit by moving, and where predators are present, such as small pelagic fish searching for food, or even the herbivores of the Serengeti Plains.

OPTIMAL SWIMMING

373

Acknowledgements

The author thanks Matthew Brewer, Piotr Dawidowicz, Bart Destasio, Pat Gorski, Takayuki Hanazato, John O'Brien, Charles Ramcharan, Sumner Richman, Rudi Strickler, Luca van Duren, and numerous anonymous reviewers for valuable discussions and suggestions concerning the development of the model for optimal zooplankton swimming. Also, thanks to Ted Garland for helping to make the motion analysis possible, to Dick Ganje for manufacturing optical hardware, to Ken Olesen for help with the electronics, to Mac Passano for help with the optical system design, and to Bill Feeny for creating the figures. References Alcaraz, M. and Strickler, J. R. (1988) Locomotion in copepods: pattern of movements and energetics of Cyclops. Hydrobiologia, 167,409-414. Blais, J. M. and Maly, E. J. (1994) Differential predation by Chaoborus americanus on males and females of two species of Diaptomus. Can.1 Fish. Aquat. Sci., 50,410-415. Buchanan, C. and Goldberg, B. (1981) The action spectrum of Daphnia magna phototaxis in a simulated natural environment. Photochem. Photobiol., 34, 711-717. Buchanan, C., Goldberg, B. and McCartney, R. (1982) A laboratory method for studying zooplankton swimming behaviors. Hydrobiologia, 95,77-89. Buskey, E. J., Coulter, C. and Strom, S. (1993) Locomotory patterns ofmicrozooplankton: potential effects on food selectivity oflarval fish. Bull. Mar. Science, 53, 29-43. Confer, J. L. and Blades, P. I. (1975). Omnivorous zooplankton and planktivorous fish. Limnol. Oceanogr., 20, 571-579. Dawidowicz, P. and Loose, C. J. (1992) Cost of swimming by Daphnia during die! vertical migration. Limnol. Oceanogr., 37, 665-667. Dodson, S. I. (1974a) Zooplankton competition and predation: An experimental test of the size-efficiency hypothesis. Ecology, 55, 605-613. Dodson, S. I. (1974b) Adaptive change in plankton morphology in response to size-selective predation: A new hypothesis of cyclomorphosis. Limnol. Oceanogr., 19, 721-729. Dodson, S. I. (1989) Predator-induced reaction norms.•Bioscience, 39,447-452. Dodson, S. I. (1990) Predicting die! vertical migration of zooplankton. Limnol. Oceanogr., 35, 1195-1200. Dodson, S. I., Hanazato, T. and Gorski, P. R. (1995) Behavioral responses of Daphnia pulex exposed to carbaryl and Chaoborus kairomone. Environ. Toxicol. Chem, 13, in press. Dodson, S. I. and Ramcharan, C. W (1991) Size-specific swimming behavior of Daphnia pulex. 1 Plankton Res., 13, 1365-1379. Drenner, R. W, Strickler, J. R. and O'Brien,W J. (1978) Capture probability: The role of zooplankter escape in the selective feeding of planktivorous fish. 1 Fish. Res. Bd. Can., 35, 1370-1373. Epp, R. W and Lewis,W M. (1984) Cost and speed oflocomotion for rotifers. Oecologia, 61, 289-292. Evans, G. T. (1989). The encounter speed of moving predator and prey. 1 Plank. Res., 11,415-417. Fox, H. M. and Mitchell, Y (1953) Relation of the rate of antenna! movement in Daphnia to the number of eggs carried in the brood pouch.1 Exp. Bioi., 30, 238-242. Gerking, S. D. (1994) Feeding Ecology ofFish. Academic Press. New York. 416 p. Gerritsen, J. (1980) Adaptive responses to encounter problems. In: Evolution and Ecology of Zooplankton Communities,W C. Kerfoot, ed., University Press of New England, Hanover, NH. pp. 52--62. Gerritsen, J. and Strickler, J. R. (1977) Encounter probabilities and community structure in zooplankton: a mathematical model. 1 Fish. Res. Bd. Can., 34, 73-82. Hamner, P. G. and Hamner, W H. (1993) In situ behavior of deep-sea animals: 3-D videography from submersibles and motion analysis. U.S. GLOBEC News No.4- August 1993. pp 7-10. Havel, J. E. and Dodson, S. I. (1984) Chaoborus predation on typical and spined morphs of Daphnia pulex: Behavioral observations. Limnol. Oceano gr., 29, 487-494. Ingle, D. (1968) Spatial dimensions of vision in fish. In: The Central Nervous System and Fish Behavior, D. Ingle, ed., U. of Chicago Press, Chicago. pp. 51-94. Jacobs, J. (1967) Untersuchungen zur Funktion und Evolution der Zyklomorphose bei Daphnia, mit besonderer Berucksichtigung der Selektion durch Fische. Arch. Hydrobiol., 62,467-541.

374

STANLEY I. DODSON

Kerfoot, W. C (1978) Combat between predatory copepods and their prey: Cyclops, Epischura, and Bosmina. Limnol. Oceanogr., 23, 1089-1102. Kerfoot,W. C, Kellogg Jr., D. L. and Strickler, J. R. (1980) Visual observations oflive zooplankters: Evasion, escape and chemical defenses. In: Evolution and Ecology of Zooplankton Communities, W. C Kerfoot ed., University Press of New England, Hanover, NH. pp. 10--27. Kerfoot,W. C and Peterson, C (1979) Ecological interactions and evolutionary arguments: Investigations with predatory copepods and Bosmina. Fortschr. Zoo/., 25, 159-196. Larsson, P. and Dodson, S. I. (1993) Chemical communication in planktonic animals. Arch. Hydrobiol., 129, 129-155. Lampert, W. (1987). Feeding and nutrition in Daphnia. In: Daphnia. R. H. Peters and R. De Bernardi, eds., Mem. !st. Ita/. Idrobiol., 45, 143-192. Lehman, J. T. (1976). The filter-feeder as an optimal forager, and the predicted shapes of feeding curves. Limnol. Oceanogr., 21,501-516. O'Brien, W. J. (1987) Planktivory by freshwater fish. In: Predation: Direct and Indirect Impacts on Aquatic Communities. W. C Kerfoot and A. Sih, eds. University Press of New England, Hanover, NH. pp. 3-16. Porter, K. G., Gerritsen, J. and Orcutt, Jr. J.D. (1982) The effect of food concentration on swimming patterns, feeding behavior, ingestion, assimilation, and respiration by Daphnia. Limnol. Oceanogr., 27, 935-949. Parejko, K. S. and Dodson, S. I. (1991) The evolutionary ecology of an antipredator reaction norm: Daphnia pulex and Chaoborus americanus. Evolution, 45, 1665-1674. Ramcharan, C W. and Sprules, W. G. (1989) Preliminary results from an inexpensive motion analyzer for freeswimming zooplankton. Limnol. Oceanogr., 34,457-462. Ramcharan, C W. and Sprules, W. G. (1991) Predator-induced behavioral defense and its ecological consequences for two calanoid copepods. Oecologia, 86, 276-286. Riessen, H. P. (1992). Cost-benefit model for the induction of an antipredator defense. The American Naturalist., 140,349-362. Riessen, H. P., Sommerville, J. W., Chiappaari, C and Gustafson, D. (1988) Chaoborus predation, prey vulnerability, and their effect in zooplankton communities. Can. 1 Fish. Aquat. Sci., 45, 1912-1920. Ringelberg, J. (1987) Light induced behavior in Daphnia. In: Daphnia. R. H. Peters and R. De Bernardi, eds., Mem. /st. Ita/. Idrobiol., 45, 285-323. Ringelberg, J. (1991) Enhancement of the phototactic reaction in Daphnia hyalina by a chemical mediated by juvenile perch (fercajluviatilis).1 Plankton Res., 13,17-25. Smith, F. E. and Baylor, E. R. (1953) Color responses in the Cladocera and their ecological significance. Am Nat.,87,49-55. Stearns, S. C (1975) Light responses of Daphnia pulex. Limnol. Ocenaogr. 20, 564-570. Stephens, D. W. and Krebs, J. R. (1986) Foraging Theory. Princeton University Press. Princeton, NJ. 247 p. Threlkeld, S. T. (1987) Daphnia population fluctuations: patterns and mechanisms. In: Daphnia R. H. Peters and R. de Bernardi, eds. Memorie dell1stituto Italiano de Idrobiol. v. 45, pp. 367-388. Williamson, C E. (1983) Behavioral interactions between a cyclopoid copepod predator and its prey. 1 Plankton Res., S, 701-711. Wright, D. I. and O'Brien, W. J. (1984) The development and field test of a tactical model of the planktivorous feeding of white crappie (Pomoxis annularis). Ecol. Monogr., 54,65-98. Zaret, T. M. (1980) The effect of prey motion on plankton choice. In: Evolution and Ecology of Zooplankton Communities,W. C Kerfoot, ed., University Press of New England, Hanover, NH. pp. 594-602.

FOOD SEARCH AND SWIMMING SPEED IN

DAPHNIA

PETTER LARSSON and OLE T. KLEIVEN Max-Planck Institute ofLimnology, D-24302 Pion, Germany. e-mail: Petter. Larsson @zoo. uib. no. Swimming behaviour and distribution of Daphnia magna were studied in relation to various concentrations of the food alga Scenedesmus acutus. The experiments were done in a ring-shaped flow through chamber in the light and in the dark. Daphnia behaviour was monitored with a video camera sensitive to infrared light. The daphnids showed rapid swimming in visible light and at low food concentrations. At higher food concentrations the swimming speed decreased. Only slow swimming was observed in the dark despite the food concentration. In a food-gradient, the animals congregated in the areas with the highest algal densities when there was light, whereas animals in the dark showed a weaker reaction. Daphnids do not seem to have a directed food search guided by vision or chemical signals, but seem to swim horizontally in the light until they meet food patches.

INTRODUCTION Zooplankton, like other animals, have to react to other organisms for survival, growth and reproduction. For many years, however, zooplankton were assumed to have very limited abilities to cope with the challenges presented by neighbouring organisms. Planktonic organisms were by definition, viewed as organisms that had very reduced abilities to swim and to react to neighbouring organisms in the pelagic environment. Studies during the last decade, however, have changed this view and have shown that freshwater zooplankton can recognize and react to the presence of other organisms. Many responses observed in zooplankton are elicited by chemical cues from cooccuring organisms (reviewed in Larsson and Dodson, 1993). Although a few studies have looked upon the effect of conspecifics (Hobrek and Larsson, 1990; Kleiven et a/., 1992; Folt et a/., 1993), most of the recent literature has been concentrated on how planktonic animals respond to their predators. The first such studies were on rotifers (Beauchamp, 1952; Pourriout, 1964; Gilbert, 1966), but the most popular group during the past few years has been members of the cladoceran genus Daphnia. Daphnia responds to the presence of predators with changes in morphology, life history and behaviour. In contrast to the large number of papers devoted to the study of how zooplankton cope with their predators, surprisingly few studies have dealt with how zooplankton find their food. Since planktonic animals can discriminate between signals from different predators, one should expect that they also are equally able to find their food.

375

376

PETIER LARSSON and OLE T. KLEIVEN

The few studies conducted are inconclusive, however, and partly contradictory. Nie et al. (1980), Malone and McQueen (1983) and Tessier (1983) compared the distribution of daphnids and other Cladocera with horizontal distribution of chlorophyll a. They found either a negative relationship or no relationship between the densities of animals and the chlorophyll concentration. Porter et al (1982) also studied swimming patterns of Daphnia in aquarium experiments and found no responses to algal patches. George (1983) reported contrasting results in a study using large enclosures. He found that Daphnia hyalina often aggregated at depths with maximum phytoplankton abundance. The animals were also more dispersed when the population density was high and the food concentrations low. An observation that also indicates a food search behaviour in Daphnia, was done by Ringelberg (unpublished, cited in Stearns, 1975). He found that Daphnia magna swam horizontally when hungry but vertically after feeding. Three studies done in flow-through chambers have shown that daphnids are able to find areas with higher food concentrations (Jakobsen and Johnsen, 1987; Neary et al., 1994; Cuddington and McCauley, 1994). When Daphnia pulex were given a choice in a horizontal food gradient, they concentrated at the highest density when this was under an incipient limiting level (McMahon and Rigler, 1965). At higher concentrations they were indifferent, or searched lower densities. There seem also to be differences between species, and Cuddington and McCauley (1994) found that D. pulex found high foodpatches more easily than Ceriodaphnia dubia. All three cited experiments were, however, carried out in the light, and it is unclear what senses the animals used when they became attracted to the food. Phototactic behaviour has been studied in daphnids since the middle of the last century (Bert 1869) and has been reviewed in Ringelberg (1987). The direction, the colour, the scattering and the change in intensity oflight, evidently affect the behaviour of daphnids. Light is also an important proximate factor in diel vertical migration. To what extent the daphnids use light and vision for direct food searching are unclear. The Daphnia eye is not able to form any image (Young and Dowing, 1976) but it reacts to the direction and the quality of the light. Daphnids have two forms of swimming, and Smith and Baylor (1953) introduced the terms "red dance" and "blue dance" to describe them. The red dance is the typical "hop and sink" behaviour. It is used when the animals are maintaining approximately the same position in the water column. The blue dance in contrast is a faster, mostly horizontally directed, swimming behaviour. The two terms were given because the "hop and sink" was most common in yellow-red light and the directed swimming behaviour most pronounced in blue-green light. Smith and Baylor (1953) suggested that the directed blue dance was the behaviour the daphnids used when searching for areas with high food concentrations. When there were high algal concentrations in the water one should expect that yellow-red light would penetrate it. The daphnids could stay in the high-food patch using the red dance. Blue-green light would indicate low densities of food, and thus the animals should swim horizontally in search of higher food concentrations. Stearns (1975) tested Smith and Baylor's hypothesis but found no evidence that Daphnia actively used the colour vision to search for food. The two dances and their relation to wave lengths were, however, confirmed. The horizontal movement was found in violet (440 11m) and in white light, while the vertical component was found between 480-735 11m and at night. Young et a/. (1984) suggested another visual food-finding mechanism. They found that the beating rate of Daphnia filter limbs was responsive to changes in light field distribution, but not to the total light intensity. Thus, Daphnia might use the light field distribution to estimate

FOODSEARCHINDAPHNIA

377

the ambient concentration of suspended food particles. The ratio of top light to sideways scattered light does provide a good measure of the density of suspended particles. Since daphnids sense their predators and conspecifics chemically (Larsson and Dodson, 1993), it might be possible that they also use chemical cues in locating food algae. Paffenh6fer and Lewis (1990) found indications of distant chemical sensing of algae in calanoid copepods, and thus chemical sensing may also be a mechanism used by daphnids in their food search. To get a better picture of how the daphnids find their food, and demonstrate to what extent light was necessary in an eventual food search, we conducted our experiments in total darkness in a controlled-illumination, ring-shaped, flow-through chamber. The animals could move freely in the ring, and various food concentrations could be kept and controlled in the different sections of the chamber. We looked at the swimming behaviour and the distributions, and from these findings could test for a food search strategy. We also wanted to consider which sense organs are used during the food search. MEfHODS The experimental animals belonged to a Daphnia magna clone isolated from a pond near Pion, Germany. Only adult females with eggs were used in the experiments. Scenedesmus acutus was used as the food alga. The ring shaped flow-through chamber (Figure 1) was made from Plexiglass. It had a diameter of 49.1 em (for the outer wall) and 45.9 em (for the inner), giving a chamber width of 1.6 em. The chamber was filled to a depth of 14.0 em, giving a total volume of 3.3 1, and it was divided into 11 sections. A similar design was independently used by Cuddington and McCauley (1994). In our chamber the sections were separated with vertical ribs extending about 4 mm into the chamber from the inner wall to reduce mixing currents. In each section, glass tubes extending from the surface and the bottom served as inlets and outlets respectively. The inflow and outflow were regulated by a peristaltic pump giving a flow-rate of about 4 ml min. -I in each section. The concentration of food algae in the outlet water was measured by a CASY 1 Cell analyser system (Scharfe System GmbH, 72760 Reutlingen, Germany). The daphnids were monitored with a black and white video-camera (1: 1.4, 6 mm lens), which was sensitive to infrared light. A covering cowl of stainless steel was placed above the chamber, and at the top, a rotating arm held the video-camera. The video-camera could be moved around the periphery of the chamber by an electric motor. The camera was held at a constant distance of 17 em from the chamber, sufficient to give a picture of slightly more than one of its sections. In the dark-experiments, the light source for the camera was two rings of infrared diodes, each consisting of 154 diodes (1.3 v). They produced infrared light at a wavelength of about 910 11m. One ring was placed above the chamber and one below which provided a dark field background for the image. An infrared light should not influence the daphnid behaviour (Smith and Macagno 1990), but to avoid unknown effects, the infrared light was only used when the animals were recorded or manually counted. When the infrared light was turned on, however, the animals showed no immediate phototrophic reactions. In the light-experiments, the upper ring of diodes was removed. A halogen lamp (20 w, 12 v run at 9 V) placed in the top and centre of the

378

PEITER LARSSON and OLE T. KLEIVEN

Figure 1 A schematic diagram of the ring-shaped flow-through chamber, the video-system and the control units.

covering cowl illuminated the chamber. That provided uniform illumination in all sections of the chamber. The light intensity at the surface was moderate (160-170 JLE s- 1 m - 2). It was equivalent to about 10% of the midday light on a partially cloudy winter day in Northern Germany. To increase the light conditions for the video-camera, the infrared light from the ring below the chamber was also used in the light experiments. The ring-chamber with the camera system was placed in a separate room kept in total darkness except the light used in the experiments. The video-monitor, the videorecorder, the peristaltic pump, and the flasks for the inflow and outflow water, were placed outside the experimental room. The movement of the video camera and the regulation of the light conditions and the flow-through, were also controlled from the outside. At the beginning of the experiment, 110 experimental animals were introduced through a separate PVC tube into the chamber (section no 11) from the outside. The animals were taken randomly from a well-fed mass culture, and kept overnight in !litre glass jars without food. Just before the introduction to the chamber, they were transferred to the experimental medium. Each trial lasted six hours, and each section of the chamber was video taped for 15 seconds. Since not all the animals were always visible on the video tapes, additional manual counting was also done once every hour on the video screen. For the calculations of density in the various sections, the mean of

FOODSEARCHINDAPHNU

379

the numbers found on the video and the number manually counted, was used. The video film was digitized on a personal computer. Using a computer programme we measured the swimming speed by plotting the position of the animals in sequential, one second still pictures. Both the horizontal and vertical components of the swimming behaviour were registered. The swimming speed was calculated as the hypotenuse of a right triangle formed by the two vectors. Two types ofexperiments were carried out in the dark and in the light. In the first type, food conditions were homogeneous throughout the chamber. The selected food concentrations were 0.008, 0.03, 0.125, 0.5, 1.0 and 2.0 mgC 1- 1. The first three concentrations were below, and the three next above the incipient limiting level (Kersting and Leeuw 1976; Lampert 1987). One trial at each food concentration was conducted in the light and one in the dark. After six hours, the swimming speed and the animals' distribution were registered. In the second type of experiment the animals were exposed to a food gradient. The animals were introduced in the chamber with the visible light on. All the sections received the same food concentration (0.008 mgC 1- 1). Under these conditions, the animals distributed themselves approximately randomly in the different sections. After one hour, water with a higher food concentration, 1.0 mgC 1- 1, was introduced into one randomly chosen section. In four ofthe trials the light was then turned off and in four others, the light was kept on. Daphnid distributions and swimming speeds were registered after the first, second and fifth hours. All the experiments were run at 20°C. The animals were adapted to a natural photoperiod (October-December). All light experiments were done while it was light outside (09-15 hours) and all dark experiments when it became dark (16-22 hours). RESULTS The swimming speed of the animals was very different under light and dark conditions. In the dark, the animals did the "red dance" with only a "hop and sink" motion at all food densities. The speed ranged from 0-4 mm s- 1. In the light, however, the animals displayed marked, mostly horizontally directed, swimming behaviour ("blue dance") at the three lowest food concentrations (Figure 2). The swimming speed averaged about 16 mm s- 1, but some individuals exceeded 30 mm s- 1. Some animals swam more than one "lap" in the ring chamber, while others, after encountering an obstacle or for other reasons, stopped, turned, and swam back at similar speed. At concentrations above 0.125 mgC 1- I, the swimming speed decreased, and at 2 mgC 1-I the speed was the same as in the dark. A Tukey (HS) pairwise comparison of the means from all the treatments, showed that there were three groups that differed in mean speed (p < 0.05). The slow group contained all the food concentrations in the dark and the highest food concentration (2 mgC 1-I) in the light. The group with medium speed included the runs at 0.5 and 1.0 mgC 1- 1 in the light, and the fast group the three lowest food concentrations in the light. We also found differences between dark and light conditions in the food gradient experiments. A food gradient was established after two hours, and remained fairly constant during the following three hours of the experiment (Figure 3). In the light, animals aggregated in areas with higher food within two hours (Figure 4). A similar distribution was found after an additional three hours. The decline in animal density from high to low food concentrations was highly significant, both after two and five

PETIER LARSSON and OLE T. KLEIVEN

380

MEAN SWIMMING SPEED

32,---------------------------------------------, 0 light

*dark 16

'7cn

E E

a 4

24---~--------------~~--------~---r--------~

0.01

1

0.1

Food concentration mgC

5

r1

Figure 2 Mean swimming speed of Daphnia magna in the light, open circles, and in the dark, solid stars, at various food concentrations. The error bars represent 95% confidence limits. The curve for the observations in the light is fitted as a logistic curve withoutthe upper tail > 2mgC 1-1. The fit for the dark is a straight line. For each point is n between 16 and 44.

hours. There was, however, no significant decline after two hours in the dark, and the correlation after five hours was less pronounced than in the light (Figure 5). Most of the aggregated animals showed "hop and sink" swimming in the section with highest food concentrations. In the four sections most distant to the food source they swam faster (Figure 6). Animals we followed from the low food area to the high food area did not necessarily stop when they reached the high food. They often swam right through the food patch. The mean swimming speed was, however, significantly different between high and low food areas. In the four sections with lowest food concentrations it was similar to the swimming speed at the three lowest food levels in the experiments with homogeneous food. The swimming speed in the high food areas was similar to the swimming speed found at 0.5 and 1.0 mgC 1- 1• According to the CASY counts, the sections with the lowest food input (0.008 mgC 1- 1) had about 0.05 mgC 1- 1 in the outlet water (Figure 3). That means that the average food concentration there was so low that fast swimming should be expected. In the section with high food input the concentration was 1.0 mgC 1- 1, while the outlet had about ~of that or about 0.17 mgC 1- 1 (Figure 3). Thus, the average food concentration was in the range where medium swimming speed should be found.

FOODSEARCHINDAPHNIA

381

DISTRIBUTION OF SCENEDESMUS IN OUTLET WATER

7

1H WITH EQUAL INPUT

6

5

3

2

0 7

2

3

4

5

2H WITH HIGH FOOD INPUT IN ONE SECTION

6 5

-

4

r-

3

-

2

0

2

3

4

5

6H HIGH FOOD INPUT IN ONE SECTION

-

¥ .....

6

.,

4

'I'

~

,.E E :L

r--

-

Gl

.,

~

~ 'S 1-

r--

r--

-

0

0

4

Sections from food input

Figure 3 Distribution of the food alga, Scenedesmus acutus, in the outlet water. The sections are ordered according to their distance to the section that got 1.0 mgC 1- 1 (approx. 40 11m3 ml- 1 * 10- 5). The other sections got 0.008 mgC 1- 1 (approx. 1.0 11m3 ml- 1 * 10-5 ). The upper figure shows the situation after 1 h with only 0.008 mgC 1- 1 in all sections. The middle figure shows the situation after 2 h with high food input in section 0, and the lower figure the situation after 5 h.

PETIER LARSSON and OLE T. KLEIVEN

382

WITHUGHT 30

1H WITH EQUAL LOW FOOD ... ..o.0473 P"0.76

25

.

r::

20

;:

•.. •... u

15

~

..

.5

0 0

0

0

10

0

2H WITH HIGH FOOD IN ONE SECTION 30

25

.•...

r:: ;:

•...

0

r-..0.6802 p< 0.0001

20

15

~

.,;

.5

10

0

30

5H WITH HIGH FOOD IN ONE SECTION 0 r-..0.8102 p< 0.0001

25

r::

;•

•. ...= .,; .5

20

15

10

Number ol sections !rom food Input

Figure 4 Distribution of Daphnia magna in the light. The figure shows the number of animals per section at various distances from the input of 1.0 mgC I- 1• The upper graph shows the situation after 1 h with equal 0.008 mgC 1- 1 food in all sections. The graph in the middle shows the situation after 2 h with high food in one section and the lower graph the situation after 5 h. Data for four replicates are shown. Each trial gave two points per distance to food input since the ring chamber had two decreasing gradients. For the input section it is only one value in each trial.

383

FOODSEARCHINDAPHNIA IN DARK 30

1H UGHT AND EQUAL LOW FOOD

25

.•...•"

•...

ra-0.0138

p=0.82

20 16

~

.,;

!:

10

0

2H DARK AND HIGH FOOD IN ONE SECTION

30

r= .0.0575

25

. ..••..

p=0.71 []

20

;:=

•...

15

~

,; !:

10

[]

[]

[]

Iii!

~

5

[]

8

[]

[]

4

0

30

5H DARK AND HIGH FOOD IN ONE SECTION r= .0.4245 p 2 mgC 1- 1 not included), and it is not unlike curves made for filtering rates. Under approximately 0.3 mgC 1- 1 one should expect D. magna to filter with maximum speed while at increasing food concentration the filtering activity should decrease. Although the filtering limbs and the swimming antenna are independent, they might be affected by the same stimuli. Light seems to be necessary for the horizontal swimming of daphnids, and even the low-light intensity in our experiments was sufficient. In the ring chamber, the light came directly from above, and, with the narrowness of the chamber, horizontal swimming might have been particularly stimulated. Nevertheless the swimming speed decreased when food concentration increased. The filtering rate also becomes reduced when food density increases to high levels (McMahon and Rigler 1965; Burns and Rigler 1967). It might be that the swimming and filtering activity are responding to the same sensory input. In the dark, however, the swimming was reduced to hop and sink. Our interpretation is that the daphnids need light for orientation when swimming horizontally. This is also in accordance with previous experiments done on the swimming behaviour and light directions (Ringelberg 1987). Because the animals could swim faster in the light, they also recognized the food faster. In the gradient experiment we did not expect that some animals would simply swim through the area with higher food concentrations. Why did the animals not all concentrate in the high food area? The food conditions in the other parts of the chamber were far from optimal, and the flow-through speed sufficient to make the high food area more favourable, even with all the animals there. We think it relates to how daphnids recognize food. They would probably all have stopped in the high food area if they could recognize the food visually. If they could recognize the food as particles, as a special colour, or by a degree of scattered light, they would presumably not swim through. Had they been led by an odour gradient from the algae, they would have been able to detect where it came from, and stopped. The results could, however, be explained, if they are sensing the food by how much algae the filtering limbs are collecting, or how much food they have accumulated in their gut. This is also an explanation suggested by Cuddington and McCauley (1994). When the food available to zooplankton is abruptly changed, there are distinct time lags before filtering rates and other aspects of feeding behaviour catch up (Lampert eta/., 1988; DeMott 1993). When the food concentration in one section of the chamber became higher, the animals in this section, got an increased amount of food in their feeding system. They would recognize the food first, and then slow their swimming speed. Animals in the other end of the chamber would not recognize the increased food concentration, and swim back and forth in the low food area. Some would also swim through the food rich areas, but that would take them just a few seconds, and they might not register the increased food concentration. It seems as if this situation could last for several hours. Porter et al. (1982) found that daphnids did not change swimming speed or direction when they were exposed to small food patches in tank experiments, which supports our hypothesis.

386

PETTER LARSSON and OLE T. KLEIVEN

In our gradient experiment the highest food concentration was between

0.17-1.0 mgC 1- 1. This is below the concentration that we found would completely stop

the horizontal swimming behaviour (2 mgC 1- 1), but in the range giving medium swimming speed. It could be that fewer would have passed the high food area had the maximum. food concentration been higher or if the area had been larger. In future experiments it will be important to test iffood concentration and food patch size affect the proportion of animals passing through. In the dark the D. magna did not find the food patch so easily. They congregated in the high-food areas after some time however. This could indicate that they are able to sense the food from some distance. We cannot exclude the possibility that chemical cues could attract the animals to the food patch. Other results from our experiments show that this factor, if present, is probably very weak. The mechanisms aggregating the animals in the dark are more likely to be the same as in light. Since the swimming speed was very much reduced, the daphnids needed more time to find the food. Cuddington and McCauley (1994) found that Daphnia pulex could locate and forage in regions of high food concentrations. The other species they studied, Ceriodaphnia dubia, could not, and they attributed this to the higher swimming speed of D. pulex. This is analogous to our situation where the difference in swimming speed led to more pronounced aggregation in the light. When compared with predator avoidance, food search seems much less specific and efficient. In a dilute environment, such as the pelagic zone of a lake, one would expect that the mechanisms for finding food should be very well developed. In contrast, the opposite seems to be the case. Daphnia search for food by horizontal swimming, not apparently directed towards any food patch. Hungry animals hitting a food patch even swam through without recognizing it. The food patches offered to the animals in our experiments were probably smaller than most of the algal swarms they normally would encounter in the nature. This might be why the daphnids, in spite of their slow food recognition, are actually so successful in freshwater habitats. However, it is difficult to understand why they are not utilizing every opportunity for feeding when food is scarce. The result seems to be in conflict with general theory of procurement of essential resources (Pulliam 1989), and only further studies can give a complete answer. Acknowledgement

We would like to thank Willi Scholer and Horst Hansen for building the apparatus for this study, Knut Kjeilen for developing the computer programme measuring the swimming speed, Maren Volquardsen and Heinke Clausen for their assistance in the experiments, Piet Spaak for statistical help and comments to the manuscript, Larry Weider and Dan Hartline for improvement of the English, and Winfried Lampert for fruitful ideas and comments to the manuscript. We will also thank W. R. DeMott and an anonymous reviewer for valuable ideas and comments to an earlier version of the manuscript. The study has been done with support from the Max-Planck-Gesellschaft and the Norwegian research council. References Bert, P. (1869) Sur la question de savoir si tous les animaux voient les memes rayons lumineux que nous. Arch. dephysiol. normaleetpathol., 2, 547-554. Beauchamp, P. De. (1952) Un facteur de la variabilite chez Roti!eres du genre Brachionus. C R. Acad. Sci., Paris 234, 573-575.

FOODSEARCHINDAPHNU

387

Burns, C. W. and Rigler, F. H. (1967) Comparison of filtering rates of Daphnia rosea in lake water and in suspension ofyeast. Limnol. Oceanogr., 12,492-502. Cuddington, K. M. and McCauley, E. (1994) Food-dependent aggregation and mobility of water fleas Ceriodaphnia dubia and Daphnia pulex. Can. J. Zoo/., 72, 1217-1226. DeMott,W. R. (1993) Hunger dependent diet selection in suspension-feeding zooplankton. In: Diet Selection: An interdisciplinary approach to foraging behavior, R. N. Hughes ed., Blackwell Scientific, Oxford, pp. 102-1232. Folt, C., Schulze, P. C. and Baumgartner, K. (1993) Characterizing a zooplankton neighbourhood: small-scale patterns of association and abundance. Freshwater Biology 30, 289-300. George, D. G. (1983) Irrelations between the vertical distribution of Daphnia and chlorophyll a in two large limnetic enclosures. J. Plankton Res., 5, 457-475. Gilbert, J. J. (1966) Rotifer ecology and embryological induction. Science 151, 1234-1237. Hobrek, A. and Larsson, P. (1990) Sex determination in Daphnia magna. Ecology 71, 2255-2268. Jakobsen, P. J. and Johnsen, G. H. (1987) Behavioural response of the waterfiea Daphnia pulex to a gradient in food concentration. Anim. Behav., 35,48-52. Kersting, K. and Leeuw, W. van d. (1976) The use of the Coulter counter for measuring the feeding rates of Daphnia magna. Hydrobiologia 49, 233-237. Kleiven, 0. T., Larsson, P., and Hobrek, A. (1992) Sexual reproduction in Daphnia magna requires three stimuli. Oikos 65, 197-206. Lampert, W. (1987) Feeding and nutrition in Daphnia. In: Daphnia, R. H. Peters and R. de Bernardi eds., Mem. !st. Idrobiol, 45,285-323. Lampert, W., Schmitt, R. D. and Muck, P. (1988) Vertical migration of freshwater zooplankton: test of some hypotheses predicting a metabolic advantage. Bull. ofMar. Sci., 43, 620-640. Larsson, P. and Dodson, S. (1993) Chemical communication in planktonic animals. Arch. Hydrobiol., 129,

129-155.

Malone, B. J. and McQueen, D. J. (1983) Horizontal patchiness in zooplankton populations in two Ontario kettle lakes. Hydrobiologia 99, 101-124. McMahon, J. W. and Rigler, F. H. (1965) Feeding rate of Daphnia magna Straus in different foods labeled with radioactive phosphorous. Limnol. Oceanogr., 10, 105-113. Neary, J., Cash, K. and McCauley, E. (1994) Behavioural aggregation of Daphnia pulex in response to food gradients. Functional ecology 8, 377-383. Nie, H. W. de Bromley, H. J. and Vivjeberg, J. (1980) Distribution patterns of zooplankton in Tjeukemeer, The Netherlands. J Plankton Res., 2, 317-334. Paffenhi:ifer, G. A. and Lewis, K. D. (1990) Preceptive performance and feeding behavior of calanoid copepods. J Plankton. Res., 12, 933-946. Porter, K. G., Gerritsen, J. and Orcutt, J. D. Jr. (1982) The effect of food concentration on swimming patterns, feeding behaviour, ingestion, assimilation and respiration by Daphnia. Limnol. Oceanogr., 27, 935-949. Pourriot, R. (1964) Etude experimentale de variations morphologiques chez certaines especes de roti!eres. Bull. Soc. Zoo/. Fr., 89, 555 - 561. Pulliam, H. R. (1989) Individual behavior and the procurement of essential resources. In: Perspectives in ecological theory. J. Roughgarden, R. M. May and S. A. Levin eds., Princeton University Press, Princeton, New Jersey, pp. 25-38. Ringelberg, J. (1987) Light induced behaviour in Daphnia. In: Daphnia, R. H. Peters and R. de Bernardi eds., Mem. !st. Idrobiol., 45, 285-323. Smith, F. E. and Baylor, E. R. (1953) Color responses in Cladocera and their ecological significance. Am. Nat.,

87,49-55.

Smith, K. C. and Macagno, E. R. (1990) UV photoreceptors in the compound eye of Daphnia magna (Crustacea, Branchiopoda). A fourth spectral class in single ommatidia. J. Comp. Physiol. A, 166,597-606. Stearns, S. C. (1975) Light responses of Daphnia pulex. Limnol. Oceanogr., 20, 564-570. Tessier, A. J. (1983) Coherence and horizontal movements of patches of Holopedium gibberum. Oecologia 60,

71-75.

s:

Young, and Dowing, A. C. (1976) The receptive fields of Daphnia ommatidia. J. exp. Bioi., 64, 185-202. Young, S.,Taylor,V. A. and Watts, E. (1984) Visual factors in Daphnia feeding. Limnol. Oceanogr., 29,1300-1308.

MIGRATING DAPHNIA HAVE A MEMORY FOR FISH KAIROMONES J. RINGELBERG and E. VAN GOOL Department ofAquatic Ecology, University ofAmsterdam Kairomones from juvenile perch (Perea fluviatilis) enhance phototactic descent of Daphnia to the hypolimnion oflakes following upon light intensity increases at dawn. Kairomones are not present in the hypolimnion and ascent in the evening might be insufficient for the animals to reach the epilimnion again. The hypothesis that sensitization is maintained for some time in the absence of fish exudates was tested. Experiments showed that sensitization was indeed maintained for 5-6 days. Thus a kind of "memory" for fish kairomones exists, resulting in an enhanced ascent from a fishless hypolimnion at sunset.

INTRODUCTION Diel vertical migration (DVM) is generally considered an adaptation to diminish predation by visual predators. Several studies have shown that DVM is most strongly expressed when planktivorous fish are abundant (Bollens and Frost, 1989; Stich, 1989; Ringelberg eta/., 1991a). If the chance of escape upon encounter with a predator is small, as it is in Daphnia meeting a fish, encounter must be prevented. Indeed, downward migration starts very early in the morning, before light intensity is high enough for fish to start feeding (Ringelberg eta/., 1991a). Thus, migration can be looked upon as predation preventive behaviour rather than a predator evasion reactiqn. If predation intensity changes over time and costs of preventive behaviour are high, a genetically fixed strategy might not pay. The situation is common and of yearly occurrence in a lake. Large shoals of o+ fish suddenly appear in the open water but numbers may crash within weeks and this leads to a highly variable predation intensity. Costs of migration are high because temperature is low and food condition is poor at the daydepth. Fewer eggs and a prolonged egg development time, thus, a lowered birth rate, is the result (Ringleberg eta/., 1991b; Loose and Dawidowicz, 1994). Because of this temporally high predation and of these high costs, genotypes which migrate in the absence of predators, may be outcompeted by non-migrating ones. Instead of a genetically fixed behaviour, phenotypic induction of migrating behaviour can be adaptive. Some cue must betray the presence of predators, and fish exudates (kairomones) have been shown to do so (Tjossem, 1990; Ringelberg, 1991; Loose, 1993). In Lake Maarsseveen, DVM of the hybrid Daphnia ga/eata x hyalina starts within a week after large shoals of o+ perch (J'erca fluviatilis) have appeared in the epilimnion (Ringe1berg eta/., 1991a). Induction must be phenotypic and the origin of the change in behaviour is an enhanced negatively, as well as positively, phototactic swimming reaction due to a kairomone, mediated by the juvenile perch (Ringelberg, 1991). The threshold for relative changes in light intensity is lowered and swimming velocity is increased. Epilimnion water could induce such enhanced swimming behaviour, but hypolimnion water failed to do so. This phenotypic induction is not without problems. If sensitization fades away rapidly when the kairomone is no longer perceived, animals that have migrated into the 389

390

J. RINGELBERG and E. VAN GOOL

hypolimnion would not react sufficiently to decreases in light intensity in the evening. Thus, with a higher response threshold and a lower swimming velocity, these animals would not be able to ascend enough to reach the epilimnion. This consideration led to the hypothesis of a "memory" for this kairomone. If sensitivity is maintained for at least the hours of daylight, to bridge the gap between morning descent and evening ascent, a solution for the observed problem is realised. The results of experiments, to test the hypothesis that sensitization of the phototactic response mechanism in Daphnia is maintained for some time, are presented. METHODS With slightly different methods, three experiments were performed to test this hypothesis. Essentially, the experiments consisted of sensitizing daphnids in water containing fish exudates. Subsequently, they were placed in clean water and, over about a week, sensitization was tested daily by measuring the photoresponse evoked by a continuous change in light intensity. All experiments were done in a dark room; the temperature was fluctuating between 18-21 oc; the light-dark regime was LD 16: 8; the light was on at 07.00 h local time. Two glass cylinders (height 40 em, diameter 7 em), enclosed in square glass jackets (40 x 9 em) containing water, were used. Overhead light from incandescent bulbs was filtered through several layers of filtering paper. Incident light intensity at the water surface of the cylinders was 2 J.LE · m- 2 · s- 1. The cylinders were filled with filtered (0.45 J.Lm) water from the hypolimnion of Lake Maarseveen. This clean water does not contain fish exudates but to be sure, it had been circulated over a sand filter for several days, since the active component of the fish exudate is broken down rapidly by bacteria. Each day at 09 h, the water was enriched with Scenedesmus acuminatus to a concentration of 2 x 10 6 J.Lm 3 · mi- 1 (= 0.5 mg C ·I- 1) and carefully aerated. On test days, the first increase in light intensity was given at 10 h. In between trials, the daphnids could adapt to the original light intensity for 45-60 min. About 7-8 trials were run on a day. No mixing to re-distribute the animals between trials was performed because the disturbance would have influenced reactivity. Moreover, a gradual redistribution started at the end of the light increases. The first experimental series was performed from 8 June until15 June 1990. Only one cylinder was used, containing 20 adult females of a clone of a hybrid D. galeata x hyalina. This clone was started with a female caught at a depth of about 10m in daytime during a migration period. The clone had been cultured in the laboratory. Relative light intensity increases of S 0.30 relative light units per minute were applied. An increase lasted for 15 min. and each minute, the number of individuals in the upper and the lower half of the cylinder was estimated. Counting rapidly reacting individuals of the small D. galeata x hyalina cannot be precise and often a few animals were missed or counted twice. Therefore, the number in the lower part was expressed as a percentage of the total count. As a criterion of reactivity, the highest percentage of daphnids moving down out of the upper half of the cylinder, was used. Before fish kairomone was added to the cylinder, relative light intensity increases were given to establish reactivity in clean water. Experiments started with animals that had been in clean water for 1 day. Then 25% of this water was replaced by 0.45 J.Lm filtered water from an aquarium containing a small shoal ofo+ perch (Percajiuviatilis). The next day, the daphnids were filtered from the cylinder, washed in clean water and returned to the

=

=

PREDATOR KAIROMONE MEMORY IN DAPHNIA

391

rinsed cylinder, now filled with clean water. On subsequent days, reactivity was tested. On the last day, water from the aquarium with fish exudate was added again to test whether the daphnids were as lively as on the first day and could be sensitized again. The second experiment was performed from 8 July until14 July 1994. Two identical cylinders were used and into each one 30 adult females of D. galeata x hyalina, clone M3 were put. This clone was also derived from a female that had been caught in the hypolimnion during a period of migration and is one of the stock cultures in the department. After one day of adaptation to the experimental situation, 12.5% of the water in the cylinders was replaced, the control with fresh hypolimnion water, the experimental one with water from an aquarium with a small shoal of juvenile perch. On the third day, the water in both cylinders was replaced three times in succession by filtered hypolimnion water. Since each time some water, containing all daphnids, was left in the cylinders, some fish exudate must have been left also. An experiment consisted of a continuous relative increase in light intensity of S = 0.13 min -I. At the end, the initial light intensity was restored and after about one hour of adaptation the intensity change was repeated. The vertical distribution was determined by counting the number of daphnids in segments of 5 em into which the cylinders were divided vertically. The time needed for counting was 30--45 seconds. In order to determine the average initial vertical distribution, five counts at three minute intervals were done before the light stimulus was given. A light change lasted for 10 minutes, and after 3 and 6 minutes from the start, counts were done. For each count, the mean depth was calculated as md =I: (ni* di I I::ni. (ni =number of daphnids in compartment di)· The response (r) to a light stimulus was expressed as the difference between the average initial md minus the md as calculated from the second count during the light change. The parameter to test the sensitivity was calculated as the difference in displacement between expt;.rimental cylinder and control: d = r (exp) - r (control). Thus the displacement of daphnids that were or had been sensitized was corrected for displacement by daphnids in the control cylinder without the kairomone. From 7-19 October 1994 a third experiment was done. Methods were essentially the same as in the second experiment and "migrating" clone M3 was used. After treatment with fish kairomone, the daphnids were sieved from the cylinders, flushed carefully with about 21 hypolimnion water and returned again to the cleaned cylinders. Treatment and control cylinder were switched. The value of the stimulus was S = 0.15 min- 1. The vertical distributions were determined 6 min after initiation of the stimulus. RESULTS The results of the first experiment are presented in Figure 1. From timet= 0, when the animals had just been placed in clean water again, the percentage of animals that descended from the upper half of the cylinder upon an increase in light intensity, decreased according to an exponential function. With a sensitivity decay constant of c = 0.014 h- 1, a half-life time ofT= 49.5 h = 2.1 days can be calculated. In the absence of kairomones a feeble negatively phototactic reaction also occurred with an average descending percentage ofx = 11.5% (s = 7.46, n = 14). This value was used to calculate a time of 6.4 days after which the sensitizing effect of the kairomone may be supposed to have disappeared completely.

J. RINGELBERG and E. VAN GOOL

392

100.-------------------------------------------.

80

3

en c

0

60

0

C'CI Cl)

a-

Cl) C)

40

C'CI

r:::

Cl)

20

0

a...

CI)

Q.

0

2

0

20

40

60

80

100

120

1 40

160

180

200

Time (h) Figure 1 A decreasing reactivity upon removal of the sensitizing fish exudate is shown as a function of time. The "memory"decay function is y 95.5 * exp (-0.014 * t); R 2 1.000. Triangle l represents the percentage reactions l h, after fish exudate was added, the value of triangle 2 can not be used in the calculation of the curve because of it is negative and triangle 3 represents the percentage reaction one day after fish exudate was added again.

=

=

In the second experiment, Daphnia responded quickly with an increased downward swimming upon an increased in light intensity after the addition of fish exudate. Next day, the first observation at 10 h, revealed a similar strong response. After this first observation, the contents of both cylinders was refreshed three times. During the next days, it became clear that the phototactic response to the first stimulus in the morning was the strongest ofthe day. Therefore, the first phototactic response in the morning was used as the test parameter. Through this test parameter, a quadratic function fitted nicely (d -6.12-0.0083 * t + 0.00039 * t2 ; R 2 0.993; d =corrected downward swimming distance (em), t = days after kairomone was removed), illustrating the decrease in difference between reactivity of the animals that had been in fish exudate water and the control animals. According to this function, the sensitizing effect of the fish exudate had disappeared (d = 0) after t = 136.5 h or 5.7 days. To study the time of day effect on responsiveness, the uncorrected displacements of control and treated animals were made a function of the observation time of the day.

=

=

PREDATOR KAIROMONE MEMORY IN DAPHNIA

393

2.0

c

Q)

0 0

1.5

0 liJ•

E Q)

(,)

1.0

«