Marine Chemical Ecology - Chapter 5 docx

Benthic marine invertebrates possess a relatively discrete repertoire of reproductive modes.4–7 Typically, they fall into two categories: those that broadcast large numbers of small eggs

The Chemical Ecology

of Invertebrate Meroplankton and Holoplankton

James B McClintock,* Bill J Baker, and Deborah K Steinberg

CONTENTS

I Introduction 196

A The Problem 196

B Role in Regulation of Material and Energy Flux 197

C The Paradox of the Plankton 197

II Laboratory Studies of Chemical Defenses 198

A Meroplankton 198

B Holoplankton 206

III Field Studies of Chemical Defenses 210

A Meroplankton 210

B Holoplankton 210

IV Chemistry of Meroplankton and Holoplankton 210

V Other Modes of Predator Avoidance 212

A Size 213

B Transparency and Other Forms of Crypsis 213

C Vertical Migration 214

D Exploitation of Sea Surface or Surfaces of Other Organisms and Particles 214

E Structural Defense 215

F Aposematism 215

G Other Considerations 216

1 Speed/Swimming Behaviors 216

2 Nutritional Content 216

3 Time in the Plankton 216

VI Symbioses 216

VII Potential Antifoulants 218

VIII Summary and Future Directions 218

Acknowledgments 219

References 219

* Corresponding author.

5 9064_ch05/fm Page 195 Tuesday, April 24, 2001 5:18 AM

196 Marine Chemical Ecology

I INTRODUCTION

A T HE P ROBLEM

Meroplankton is comprised of organisms that spend some component of their life history in theplankton, usually the eggs and larvae of benthic or nektonic adults There are a few examples ofadult benthic marine invertebrates that spend brief periods of time in the plankton These includethe adult reproductive phase (epitoke) of some marine polychaetes, whose benthic life history isinterrupted at reproductive maturity by a dramatic ontogeny of swimming appendages followed byswimming behaviors that result in the swarming and bursting of a pelagic reproductive phase Somedeep sea holothuroids (sea cucumbers) spend some period of their adult life swimming in the

meroplankton in the world’s oceans are comprised of the propagules of algae or the eggs and larvae

of benthic invertebrates and fish Little is known of the chemical defenses of the propagules ofalgae or the eggs and larvae of fish Therefore, the present review and discussion of the chemicalecology of meroplankton will focus primarily on feeding-deterrent properties of marine invertebrateeggs and larvae While all marine invertebrate groups have the potential of producing defensivechemistry in their larval offspring, studies to date have focused specifically on the eggs and larvae

of sponges, cnidarians, molluscs, echinoderms, and ascidians, groups of organisms that are wellknown to possess chemical defenses in their adult stages

Benthic marine invertebrates possess a relatively discrete repertoire of reproductive modes.4–7

Typically, they fall into two categories: those that broadcast large numbers of small eggs that arefertilized and develop into small feeding planktotrophic larvae, or those that produce small numbers

of large eggs that are fertilized and develop into large, often conspicuous, nonfeeding lecithotrophiclarvae that are subsequently brooded (protected by the parent) or released into the plankton.5 Eggs

or larvae that are released into the plankton, while in some cases having limited mobility generated

by ciliary beating, are extremely sluggish and generally lack protective skeletization, and evidencewould suggest that they are exposed to considerable predatory pressure from planktivores.8,9 More-over, an extensive literature indicates that eggs and larvae face a formidable array of predators(reviewed by Rumrill10) One example of planktivory was aptly described by Emery11 for plank-tivorous pomacentrid fish when he offered that they constituted a “wall of mouths” facing theplankton Therefore, one might expect strong evolutionary pressure for selection of defensivechemicals that would decrease the likelihood of predation This is particularly the case for thosebenthic marine invertebrates that produce very small numbers of large, conspicuously colored,nutrient-rich lecithotrophic larvae that are released into the plankton, where loss of even smallnumbers of larval progeny would have strong negative effects on the probability of successfulrecruitment.6,12 In summary, information on the chemical defenses of eggs, embryos, and larvae ofmarine invertebrates is important because models of evolutionary selection of life history patternsmake assumptions about patterns of mortality of offspring.5,13–18 These models generally assumethat eggs, embryos, and larvae are vulnerable to predators, and have primarily considered marineinvertebrates with planktotrophic modes of development

Unlike meroplankton, holoplankton is comprised of organisms that spend their entire life cycles

in the plankton The holoplankton contain representatives of nearly every algal and animal group.Over 10,000 species of copepods (crustacea) alone are known, and these can reach abundances of

gelatinous zooplankton inhabit the sea, with prominent members including medusae, phores, ctenophores, pelagic molluscs, and pelagic ascidians (e.g., salps, larvaceans) The ubiquitoussalps, for example, are periodically encountered in swarms extending hundreds of kilometers,20,21

protozoa, unicellular, and colonial animals such as acantharia, foraminifera, and radiolaria, are also

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 197

invertebrates are likely subject to intense predation pressure, primarily from crustaceans and fish.Cyanobacteria (formerly known as blue-green algae), radiolarians, foraminiferans, and larvaceansmove passively with the currents, while some gelatinous zooplankton such as salps, cnidarians,ctenophores, pteropods, and heteropods are generally sluggish swimmers It is unlikely that theycan swim rapidly enough to avoid predatory crustaceans and fish One might expect that holoplank-tonic marine organisms have evolved secondary metabolites to deal with the problem of predation,particularly those species that are conspicuous Moreover, organisms symbiotically associated withchemically defended holoplanktonic organisms may derive some protection by simply associatingwith holoplankton

B R OLE IN R EGULATION OF M ATERIAL AND E NERGY F LUX

Holoplankton, and to some extent meroplankton, are responsible for the regulation of material andenergy flow in oceanic food webs.24 Zooplankton grazing plays a key role in the recycling of allbiogenic elements, and the community structure of the pelagic food web determines the export ofelements from the upper water column The abundance of particular taxa as influenced by ecologicalprocesses like chemical defense provides a mechanism to affect this structure The size distribution

of pelagic producers (phytoplankton) and trophic position of consumers (zooplankton and ekton) determines the proportion of primary production that is lost and the composition andsedimentation rate of sinking particles from surface communities, and has a significant impact onnutrient cycling.24 For example, the production in most oceanic food webs tends to be dominated

micron-by microbial processes, with protozoan and small crustacean grazers in complex food webs Much

of the carbon and nutrients are recycled in the surface waters with little export, due to loss ofenergy at each of the many trophic levels Alternatively, copepods or other large grazers feedingdirectly on large diatoms in coastal upwelling areas may contribute directly to flux, and a largerfraction of the phytoplankton production is exported in this short food web There are also generalistconsumers, such as the pelagic tunicates that feed with mucous food webs with which they canfilter the smallest size particles.25,26 When abundant, pelagic tunicates can account for large exports

of material from the surface waters to the deep sea21,27–29 through the flux of their fecal pellets ordiscarded feeding webs Since they feed at the base of the microbial food web, even in an oceanicecosystem, they will short circuit the normal paradigm for material and energy flow within thesecommunities Clearly, there is a need to begin to understand how chemical deterrents may mediatethese important patterns of material and energy flow in oceanic systems

C T HE P ARADOX OF THE P LANKTON

Chemically mediated defenses among holoplankton and meroplankton may help resolve why there

is a great diversity of co-existing species, all competing for the same resources, in a seemingly

theory of competitive exclusion, one species should out-compete them all However, this uniformenvironment is characterized by small-scale spatial and temporal heterogeneity, such as development

of microhabitats in low-turbulence situations,31 and, thus, offers a variety of niches Factors such

as selective zooplankton grazing are important as well For example, if one species is consumed

by a predator, while another species is chemically defended, this will result in diversification ofboth predator and prey and make co-existence possible An excellent example of niche diversifi-cation in the pelagic environment is the association of crustaceans with gelatinous zooplankton,such as copepods associated with salps32 and the mucus feeding webs of “houses” of larvaceans.33,34

Interestingly, virtually all hyperiid amphipods are associates of gelatinous zooplankton such assalps, ctenophores, siphonophores, medusae, or radiolarian colonies for at least part of their

9064_ch05/fm Page 197 Tuesday, April 24, 2001 5:18 AM

198 Marine Chemical Ecology

evolved to live in a benthic-like habitat in midwater that is provided by the gelatinous

predation in the pelagic zone However, a benthic marine amphipod has been shown to build a

“domicile” from algal material in which it resides.39 The amphipods selectively chose for theirdomiciles algal species with secondary metabolites that deter predation by reef fishes, and arethus chemically defended against predation Whether pelagic amphipods might reduce their chance

of being consumed by associating with chemically defended gelatinous zooplankton in the pelagiczone is unknown

II LABORATORY STUDIES OF CHEMICAL DEFENSES

The vast majority of work conducted to date on the chemical ecology of meroplankton andholoplankton has employed laboratory-based approaches There are both pros and cons associatedwith laboratory studies Bioassays conducted in the laboratory can be carefully controlled, andsmall organisms or their eggs and larvae can be observed directly, whereas field observations would

be much more difficult or even impossible Nonetheless, it is fair to say that the field of marinechemical ecology has been moving ideologically towards increasingly ecologically relevant

defenses in the earlier years41–43 have given way to sympatric marine predator models, the use ofextracts rather than homogenates, employment of ecologically relevant concentrations of extracts

or pure compounds, and increasing numbers of studies that couple laboratory and field assays

numbers of small eggs that develop as planktotrophic larvae Lucas and others had noted earlier45,46

occurred in the adult body wall and eggs of A planci,47 Lucas et al.44 examined the potential role

of saponins as feeding deterrents in eggs and larvae Gelatin pellets were prepared with yeast extract

as a feeding stimulant and contained ecologically conservative concentrations of crude saponinextracts Four species of sympatric pomacentrid fish were employed as potential predators In almostall cases, fish rejected experimental gelatin pellets containing saponins while readily consumingcontrol pellets Interestingly, Lucas et al.44 noted that with decreasing hunger level, fish increasedtheir discrimination against pellets containing saponins, indicating that the nutritional condition orhunger level of fish predators can influence the ability of chemicals to effectively deter predators

pellets influenced acceptability While their study did not involve an accurate modeling of thenutritional or energetic content of eggs or larvae, they extended this observation to a comparison

of planktotrophic and lecithotrophic eggs and larvae, arguing that planktotrophic larvae may benutritionally less acceptable to fish than yolky eggs and yolky lecithotrophic larvae Lucas et al.44

concluded that saponins sequestered in eggs and larvae appear to be effective deterrents againstfish, and they offered some preliminary qualitative observations that the larvae of A planci mayalso be rejected by planktivorous invertebrate carnivores and benthic polychaetes The logicalextension of this work, posed in a question by Lucas et al.44 — “Do larvae of other sea stars containsaponins as chemical defenses?” — still remains unanswered

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 199

While no bioassays were conducted on discrete gametes, embryos, or larvae, De Vore andBrodie48 examined the palatability of gravid ovaries of the temperate sea cucumber Thyone briareus

pieces of ovaries or control food over an 11 day period, with the first day consisting of a feedingtrial on control food (mussel tissue) and the subsequent 10 days consisting of a randomizedpresentation of either experimental gravid ovary or control food Fish demonstrated a strong andsignificant rejection of the gravid ovaries as compared to the controls, and investigators suggestthat this may reflect the concentration of a toxin within the ova to protect the progeny As thisspecies possesses a vitellarian larva that develops in the water column, it is possible that theselarvae may possess a chemical defense However, further work is needed to verify that the deterrentproperties of the gravid ovary are indeed related to feeding deterrents sequestered in the ova andnot simply in the supporting ovarian tissues

coloration) in their study of the large, brightly colored larvae of the subtropical colonial ascidian

acceptance or rejection noted Importantly, these investigators examined whether rejected larvaewere still capable of normal swimming behaviors, an indication that fish mouthing did not causesubsequent mortality Swimming larvae were rejected by pinfish The researchers further demon-strated that rejection was chemically based by observing rejection of agar pellets containing larvalhomogenates; defensive chemicals were effective at a concentration approximately 170 times lowerthan they occur in the larva The identity of the defensive chemicals was not determined but shown

to have a molecular weight of less than 14,000 Da, and to unlikely be proteins since boiling did

examined the question of larval aposematism in E turbinata Utilizing the palatable larvae of theascidian Clavulina oblonga, they dyed these larvae a color similar to that of the unpalatable larvae

larvae of C oblonga Indeed, these fish appeared to learn how to avoid colored larvae rapidly,whereas fish that had not been conditioned on the larvae of E turbinata consumed the larvae ofdyed or undyed larvae of C oblonga in equal frequency Their observations provide a unique test

of whether warning coloration operates at the group or kin level, since larval prey die if sampledduring the learning process.50 Their results argue against group selection since the majority of larvaesurvive the learning process, and fish retain their recognition of unpalatable larvae for only a veryshort period of time (Young and Bingham, unpublished) Instead, they argue that if larval survivaldecreases with successive strikes, then individual selection51 should be invoked as an explanationfor the evolution of aposematic coloration in the larvae of E turbinata, even given the short memory

of the pinfish This study raises intriguing questions about how widespread aposematism might beamong marine invertebrates or even fish eggs, embryos, or larvae Certainly, visual predators such

as fish can comprise a significant component of the planktonic predator population, and otherinvestigators have indicated that colored and thus conspicuous larvae may be more likely to possesschemical defenses.52,53 A larger data base from carefully controlled studies of aposematism in marineinvertebrate larvae is needed before a general pattern can be adequately evaluated

Although the chemical deterrent properties of the eggs and planktotrophic larvae of the tropicalnudibranch Hexabranchus sanguineus were not investigated, Pawlik et al.54 demonstrated that theegg ribbons of this nudibranch are defended from fish predation by macrolides (see kabiramide,Figure 5.2) Sequestering these compounds from its sponge diet, H sanguineus incorporates thecompounds into the mantle tissues and egg cases after chemical modification The presence ofdefensive macrolides in the egg cases raises intriguing questions about whether these defensivecompounds might also be provisioned in the eggs and subsequently serve a defensive role inmeroplanktonic larvae

McClintock and Vernon52 furthered the study of chemical defenses of echinoderm offspring byexamining feeding-deterrent properties of the eggs and embryos of Antarctic sea stars, sea urchins,9064_ch05/fm Page 199 Tuesday, April 24, 2001 5:18 AM

200 Marine Chemical Ecology

and a sea cucumber They chose to use an allopatric fish model as a predator (the marine killifish

no exposure to Antarctic echinoderm eggs or embryos over evolutionary time and would not beexpected to have co-evolved adaptations of resistance to toxins Other chemical ecologists haveargued that it is more important to employ sympatric predators in order to effectively evaluatewhether chemical deterrents are truly effective against ecologically relevant predators.40 Both sides

of this argument likely have merit, and perhaps a dualistic approach is best if possible, with bothsympatric and allopatric models tested Because many Antarctic echinoderms have been shown torelease their eggs into the water column, in contrast to Thorson’s Rule,7 and have extremely slow

predicted that chemical defenses may be relatively common in Antarctic echinoderm eggs andembryos They found that lyophilized egg and embryo tissues of four species of sea stars embedded

at ecologically relevant concentrations in agar pellets deterred feeding in killifish Three of thesespecies produced large yolky eggs or embryos (lecithotrophs), and one of these, Perknaster fuscus,did not brood but rather released large yolky eggs into the water column where they developed aspelagic lecithotrophic larvae A fourth species, Porania antarctica, produced intermediate-sizedeggs that developed as relatively large planktotrophic larvae The nature of the chemical com-pound(s) was not determined, but they are likely to be saponins44 such as steroid oligoglycosides

numbers of lecithotrophic or large numbers of planktotrophic eggs and embryos can employchemicals to defend their offspring, but that chemical defenses may be somewhat more common

in lecithotrophic species Interestingly, both lecithotrophic species that brood and broadcast theireggs and embryos were found to be chemically defended One obvious question that arises fromthese observations is why brooding species should invest defensive chemistry in their offspringwhen they are presumably protected from predation by the adult The answer may be associatedwith the especially vulnerable period of time when the juvenile leaves the protection of the adult,presumably provisioned with defensive chemicals until some refuge in size is attained This may

be particularly important in polar marine environments where slow growth rates in sea stars andother marine invertebrates may result in periods of years spent in the vulnerable juvenile phase.56

McClintock et al.57 extended the analysis of potential chemical defenses to another phylum ofAntarctic marine invertebrates in their analysis of the biochemical and energetic composition and

studies were limited to an examination of the palatability of the gonad to a sympatric planktonicfish (Pagothenia borchgrevinki) Unable to trigger a spawning response in order to collect eggs orembryos or raise larvae, they were only able to test small pieces of intact ovitestes using similarlysized pieces of cod muscle as controls They found significant rejection of ovitestes by Antarcticfish While homogenates or extracts were not tested in feeding pellets, it is unlikely that deterrencewas related to structural defenses (no skeletal material) or low nutritional content (ovitestes werefound to be very high in energy57) Their findings, nonetheless, must be interpreted with cautionsince feeding deterrence could be attributable to chemicals in the sperm (although unlikely) or thenongametic gonadal tissues The data do suggest that the bright orange planktotrophic eggs,

chemical deterrent was not investigated, but the ovitestes was determined to be mildly acidic (pH

= 5.86), a factor that may have contributed to their rejection by fish Feeding deterrence in theouter tunic of some ascidians has been attributed to sulfuric acid sequestered in small bladders.58

Using the eggs and larvae of marine ascidians as a model system, Lindquist et al.53 examinedthe question, “Why are embryos so tasty?” posed by Orians and Janzen,59 who had pointed out thatbirds, reptiles, amphibians, fish, and insects all seem to lose large proportions of their eggs andlarvae to predators and that evolution should strongly favor those organisms that produce distastefuleggs Orians and Janzen59 speculated that (1) actively developing tissues such as those in eggs andembryos are incompatible with toxic chemicals (autotoxicity), (2) there are energetic constraints

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 201

that limit the ability of eggs and embryos to produce toxins, and (3) there may be tradeoffs between

the production of deterrents and potential development rates Lindquist et al.53 noted that the focus

of much of this speculation revolved around vertebrates, and that there was a need to extend the

evaluation to marine invertebrates before generalizing about any apparent lack of chemical defenses

in eggs and embryos

Lindquist et al.53 selected ascidians as a model group of organisms for such a study because

they have large conspicuous eggs and embryos that are amenable to chemical evaluation, and their

larval ecology is generally well known.49,60–62 Shipboard laboratory-based bioassays included an

investigation of chemical defenses of both living larvae and larval crude extracts of the Caribbean

maintained on a diet sufficient to prevent starvation Using a rather innovative approach, Lindquist

et al.53 employed the eyes of freeze-dried krill as a larval mimic since they were similar in size

and color to larvae and were readily consumed by these fish Replicate groups of fish were presented

a krill eye, then a T bifasciatum larva, and then yet another krill eye to ensure that a rejection

response to the larva was not due to satiation Crude extracts of single larvae were impregnated

into single krill eyes and were presented to fish along with controls and fish ingestion experiments

conducted in a similar fashion The researchers found the tadpole larvae and the krill eyes treated

with crude extract of T bifasciatum to be highly unpalatable to the groups of bluehead wrasse

Coupled with field observations of ascidian larval chemical defenses (see below) and a review of

the literature on the unpalatability of ascidian larvae conducted to date (Table 5.1), Lindquist et

al.53 argue that brooding ascidians that produce large conspicuous larvae, and often release their

larvae over short durations and distances during daylight hours to ensure that larvae can employ

photic cues to enhance settlement, are under strong selection pressure to evolve chemical defenses

Lindquist et al also point out that many predatory reef fish have limited home ranges,63 and that

clumping of unpalatable larval prey may increase the likelihood that fish will learn to avoid ingesting

49) They also propose thatchemical defenses among larvae that require several weeks or more to develop in the plankton may

be less common because pelagic eggs and larvae are likely to be transported offshore where

predation levels are lower.63,64 However, it would seem that while predation may indeed be lower

in these pelagic offshore habitats, the longer duration of exposure may offset any benefit attributable

to habitat-specific differences in predation level

Importantly, Lindquist et al.53 also document that ascidians can exhibit chemical differences

between defensive secondary metabolites among adults and larvae For example, larvae from

tambjamine F as compared to adults.65 Moreover, larvae of Trididemnum solidum contain only four

of the six didemnins found in adults.53 This could be the result of different selective pressures

during planktonic vs benthic life history phases In contrast, Lucas et al.44 found no differences in

the saponin chemical defenses of the embryos, larvae, and adults of the sea star Acanthaster planci.

Clearly, additional studies are needed to expand the evaluation of ontogenetic shifts in defensive

chemistry in marine organisms

Based on their findings as well as those of others for ascidians, Lindquist et al.53 question the

adequacy of the autotoxicity, energetic, or developmental constraints suggested by Orians and

Coupled with other reports of chemical defenses in the eggs and embryos of amphibians,66 insects,67

and additional marine invertebrates,44,48,52,54,68–70 there appears to be ample evidence to question

the validity of these presumed constraints However, Slattery et al.70 recently suggested that the

lack of chemical defenses in the larvae of the soft coral Sinularia polydactyla may be attributable

to autotoxicity constraints

In yet another study focusing on the larvae of a colonial ascidian, Lindquist and Hay71 evaluated

not only whether secondary metabolites in the large brooded larvae of Trididemnum solidum cause

9064_ch05/fm Page 201 Tuesday, April 24, 2001 5:18 AM

chemically defended eggs and larvae (also see Young and Bingham

Reproductive Mode

Predator(s)

Sponges

star, amphipod

113

Soft Corals

Erythropodium

caribaeorum

Larva Lecithotrophic Coral, fish 68, 72

Hard Corals

Hydroids

Echinoderms — Sea Stars

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 203

feeding deterrence, but rather extended their evaluation to measure whether changes in consumergrowth or survivorship result from physiological effects of the ingested noxious compounds Such

an evaluation is appropriate considering that many consumers ingest small amounts of noxiouscompounds when sampling prey, often with minimal apparent detrimental effects First, they

presented the pinfish Lagodon rhomboides with alginate pellets (larval mimics) containing

ecolog-ically relevant concentrations of larval didemnin cyclic peptides53 and squid puree as a feedingstimulant, or, alternatively, control alginate pellets containing only squid They found that fishrapidly learned to avoid the larval mimics and consequently found it impossible to evaluate long-term impacts of the ingestion of noxious secondary metabolites on fitness Nonetheless, Lindquist

long-term effects on fitness Sea anemones were presented experimental and control pellets dailyfor a period of 32 days At the end of this time period, it became clear that while didemnins werecapable of causing an emetic response, sea anemones did not become conditioned to avoid ingestion

of the experimental pellets Therefore, the sea anemones consumed some minimal level ofdidemnins over the experimental period The results of this minimal consumption of noxiouscompounds were profound Growth of the adult sea anemones was reduced by 82% in the exper-imental group Moreover, the production of asexual clones was reduced by 44%, and the averagemass of a clonal offspring was reduced by 41% as compared to clones produced by control seaanemones that had not ingested didemnins These findings clearly demonstrate that ingestion ofecologically relevant concentrations of noxious secondary metabolites can cause a significantreduction in consumer fitness This was the first demonstration of an effect on fitness resulting

argue that such dramatic decreases in fitness could clearly select for consumers that recognize andreject prey containing defensive chemicals, even when they form a very small portion of a

generalist’s diet In the case of the ascidian Trididemnum solidum, larvae are released year round,

during daylight hours, and remain in the vicinity of the adult It is likely that predators would

ascidian larvae are not killed by predators, group or kin selection need not be invoked to explain

additional support for the hypothesis that there should be strong evolutionary selection of chemical

anemone

113

anemone, amphipod, fish

52, 113

anemone, amphipod

113

Ascidians

Reproductive Mode

Predator(s)

defenses in the eggs, embryos, and larvae of meroplanktonic marine invertebrates, particularlythose that have lecithotrophic reproductive modes.12,49,52,53

Lindquist68 extended the analysis of the palatability of lecithotrophic marine invertebrate larvae

in his comparative investigation of the larvae of a variety of temperate and tropical sponges (ninespecies), gorgonians (nine species), corals (three species), hydroids (two species), and bryozoans(one species) Noting that while larvae of benthic invertebrates search for an appropriate settlementsite they encounter a variety of sessile invertebrate predators, he selected three species of coralsand a species of sea anemone as model larval predators The methodological approach involvedplacing predators held on a maintenance diet individually in containers and presenting them firstwith a palatable control food, comprised of a sodium alginate pellet containing squid mantle flesh

to ensure they were feeding, and then subsequently presenting them a larva All larvae presented

to corals or sea anemones were monitored to see if they were rejected, ingested, or ingested andregurgitated Larvae that were regurgitated were followed to ensure that they developed andmetamorphosed to the juvenile stage normally Feeding assays were done across a period of severaldays to prevent predators from becoming satiated during feeding trials Lindquist found that many

of the species tested were unpalatable to these predators; indeed, only the larvae of three of ninespecies of sponges and two of nine species of gorgonians were consumed Importantly, both larvalsurvival and metamorphosis were not significantly different in regurgitated larvae and control larvaethat were never attacked Although Lindquist68 did not evaluate the basis of rejection, it is likelythat this was chemically based, since the larvae had no potential skeletal or behavioral defenses

In yet another survey that focused on fish rather than invertebrate predators, Lindquist and

gorgonians, in addition to the brooded larvae of 2 species of temperate hydroids and a bryozoan,were unpalatable to fish In contrast, brooded larvae of three species of temperate ascidians, atemperate sponge, and three species of Caribbean hard corals were consumed Larval laboratoryassays were conducted by first presenting a single brine shrimp to a species of sympatric fish thathad been held on a maintenance diet Only fish that consumed the brine shrimp were presentedlarvae, and only larvae that had been sampled by fish and either ingested or rejected were considered

showed no significant decrease in metamorphic competence Of the species with unpalatable larvae,five were further examined to determine whether noxious chemicals were responsible for deterrence

In all five cases, fish rejected alginate pellets containing a feeding stimulant and ecologically relevantconcentrations of larvae extracts, while consuming control pellets containing only feeding stimulant,indicating a chemically based deterrence While not directly tested, it is likely that deterrentchemistry is responsible for the unpalatable nature of many, if not all, of the larvae tested Inter-estingly, brooded larvae were most likely to be unpalatable, while broadcasted larvae (both leci-thotrophic and planktotrophic) were generally consumed Providing additional and more broadlybased evidence to their conjecture71 that larvae of the tropical ascidian Trididemnum solidum are

chemically defended in part because they release conspicuous larvae during daylight hours,

day (89% of total species investigated), while palatable larvae were seldom found to do so (23%

of total species investigated) Many of the unpalatable larvae were brightly colored (60% of totalspecies investigated), while all palatable larvae lacked coloration, supporting earlier predictionsthat aposematism may operate in chemically defended lecithotrophic marine invertebrate larvae.49

Extending the analysis of the palatability of marine invertebrate lecithotrophic and totrophic eggs, embryos, and larvae to the polar regions, McClintock and Baker12 examined a suite

plank-of Antarctic marine invertebrates with contrasting modes plank-of reproduction These included thespawned eggs and larvae of a sea urchin and the intraovarian eggs of a sea star, both withplanktotrophic larvae, and the lecithotrophic embryos and larvae of three sea stars with eitherbrooding or broadcasting modes of reproduction Moreover, a nudibranch and sponge with eggribbons and brooded lecithotrophic embryos, respectively, were examined Gravid ovaries, spawned

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 205

eggs, and developing embryos and larvae were tested for palatability against three common

sym-patric predators with very different feeding patterns: the antarctic sea star Odontaster validus, a

benthic omnivorous scavenger suggested by Dayton et al.73 to be sufficiently abundant to act as alarval filter for settling marine invertebrate larvae; the large, abundant, and voracious sea anemone

Isotealia antarctica that feeds benthically by scavenging organisms that drift, swim, or crawl near

its tentacles; and the swarming amphipod Paramoera walkeri that scavenges seasonally on the

benthos and in the water column Where sufficient amounts of material were available for extraction,crude lipophilic and hydrophilic extracts were prepared and imbedded at ecologically relevantconcentrations in alginate pellets containing a feeding stimulant and tested against predators usingpellets containing only feeding stimulant and the appropriate solvent carrier as a control Alginate

pellets containing lipophilic and hydrophilic extracts of spawned eggs of the sea urchin Sterechinus

neumayeri and gravid ovaries of the sea star Odontaster validus, both with broadcasting

plank-totrophic modes of reproduction, were readily consumed by all three predators Pellets containing

lipophilic and hydrophilic extracts of the four-armed plutei of S neumayeri were also readily

consumed by the sea anemone predator, indicating a lack of chemical defense In contrast, at leastone of the three predators displayed significant feeding deterrence for eggs, embryos, larvae, ortheir lipophilic or hydrophilic extracts among the remaining five lecithotrophic benthic marineinvertebrates tested The basis of rejection was demonstrated to be chemically derived in the brooded

embryos of the sea star Diplasteria brucei, and it is very likely that defense is also chemically

based in the remaining four lecithotrophic species which possess conspicuous eggs, embryos, orlarvae that are high in energy content (and thus attractive prey), lack morphological defenses, andare immobile or sluggish swimmers This study demonstrates that feeding deterrence is species-specific among predators, and supports observations that lecithotrophic embryos or larvae may beparticularly well suited to chemical defenses.49,53,71,72 Moreover, in Antarctica, where both broad-casting and brooding modes of lecithotrophy are particularly common, rates of larval developmentare several orders of magnitude longer than comparable species at temperate and especially tropicallatitudes Polar marine invertebrate larvae may literally spend 2 to 6 months on the benthos or inthe water column during development, thereby substantially increasing predation risk prior tojuvenile recruitment

Another recent study of chemical defenses of the lecithotrophic progeny of a marine invertebrate

is that of Slattery et al.,70 who examined the developing embryos and larvae of the tropical soft

coral Sinularia polydactyla Whole blastulae, early planula larvae, and competent planula larvae,

along with their extracts imbedded in alginate pellets containing a krill feeding stimulant, were

presented to the pufferfish Canthigaster solandri Rejected larvae were followed to observe survival,

and fish were presented a food pellet after an experimental pellet to make sure they were not satiatedduring the trial All whole developmental stages and their respective extracts were found to be

detected in blastulae and larvae and are likely to be responsible for feeding deterrence,70,74,75

common tropical reef fish will consume the eggs of this species of coral It is noteworthy thatSlattery et al.70 detected increasing concentrations of defensive metabolites during the ontogeny ofthe larvae, although even blastulae contained concentrations sufficient to prevent predation by

77) This suggests that developing progeny are capable of secondarymetabolite synthesis and that defensive compounds need not be derived directly from adults.The most recent study of defensive metabolites in marine invertebrate planktotrophic andlecithotrophic larvae is that of Cowart et al.,78 who examined the presence of halogenated metab-

benedicti (known to possess both lecithotrophic and planktotrophic development as a genetic

polymorphism) and Capitella sp (lecithotrophic development) Researchers found that pre-release

pufferfish (see also Harvell et al

unpublished data) in both pre- and post-release larvae of the benthic polychaetes Streblospio

larvae of S benedicti with planktotrophic development had the lowest levels of halogenated

metabolites, while post-release larvae had intermediate concentrations, suggesting that the totrophic larvae are synthesizing these defensive metabolites during their development Even higher

plank-levels of halometabolites were found in post-release lecithotrophic larvae of S benedicti, suggesting

that lecithotrophic females may be expending more energy on chemical defenses than their totrophic counterparts by supplying their embryos with greater amounts of compounds, theirprecursors, or sufficient energy for their biosynthesis.78 Pre-release larvae of the lecithotrophic

plank-Capitella sp contained the highest concentrations of halogenated metabolites when comparing both

species Levels of halogenated metabolites in the larvae of both species were greater than thoseknown to deter predation However, it should be noted that these feeding-deterrent assays wereconducted using epibenthic predators and there is a need to evaluate their effectiveness at deterringpotential larval predators

The breadth of the surveys by Lindquist,68 Lindquist and Hay,72 and McClintock and Baker,12

across a variety of phyla and from temperate, tropical, and polar geographic regions, coupled withthe data from studies reviewed above,49,53,71 indicates that previous hypotheses, predicting thatpredation on marine larvae is relatively ubiquitous10 should be reconsidered for lecithotrophic larvae.These studies also add significant new information to a growing database on the palatability ofmarine invertebrate species with lecithotrophic larvae (Table 5.1) A similar assessment for plank-totrophic larvae must await broadening of the database for species with this mode of development

To date there have been comparatively few studies of the palatability of eggs, embryos, and larvae

of species of marine invertebrates with planktotrophic larvae,12,44 although one recent study has,for the first time, examined the palatability of a suite of temperate planktotrophic larvae.80 Theseinvestigators presented three species of predatory fish and a hard coral with a selection of liveplanktotrophic larvae of benthic marine invertebrates collected from the plankton Those that wererejected were crushed to render potential morphological defenses useless and were then re-offered

to predators in order to assess whether defense was morphologically or chemically based ers found that for at least one fish predator, a significant number of gastropod veligers, barnaclecyprids, crab zoeae, and stomatopod larvae were likely morphologically defended (34% ofmeroplankton tested; these were rejected whole and then consumed once crushed) Others, includingpolychaete larvae, barnacle nauplii, bivalve veligers, shrimp zoeae, crab megalopae, phoronidactinotrochs, and hemichordate tornaria, were readily consumed (65% of meroplankton tested),apparently lacking chemical means of defense A number of nemertean pilidia, asteroid bipinnaria,and cnidaria planulae were rejected both whole and crushed, suggesting that they were chemicallydefended These findings suggest that while select taxa of meroplankton have planktotrophic larvaethat are chemically defended, a considerable proportion of species with planktotrophic larvae mayrely on the production of copious numbers of small feeding larvae to offset a lack of a morphological

Research-or chemical defense

B H OLOPLANKTON

The first experimental documentation of a potential chemical defense in a holoplanktonic marine

melea-gris, a common inhabitant of bays along the Atlantic coast of North America The study was

stimulated by observations that when physically disturbed, this jellyfish produced a sticky mucusthat appeared to have toxic effects on fish held in the same collecting container Laboratory bioassayswere conducted by collecting mucus produced over a discrete time period from an individualjellyfish of known volume This mucus was then combined with seawater, and two experimentaltreatments were prepared: one by leaving the mucus-seawater mixture undisturbed and the other

by centrifuging the mucus-seawater mixture such that the particulates and pieces of mucus wereremoved (mucus-free) A control consisted of seawater alone Three sympatric species were testedfor toxic effects including two fish, the juvenile planefish and the Atlantic bumper, and spider crabs

A fourth test animal was the allopatric pinfish Logodon rhomboides Experimental fish placed into

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 207

either the mucus-free or mucus-seawater mixture reacted by displaying gaping behaviors or laying

on their sides, and a number died within a 24-h period Spider crabs were unaffected by thetreatments While no direct measurements of feeding deterrence were conducted, the investigatorsproposed that the toxic effects observed were evidence of a chemical defense While these immer-sion toxicity assays are not directly ecologically relevant, one can speculate that fish predators

that bite into S meleagris and become exposed to mucus might find the toxins distasteful (see also

suggested that it may be derived from nematocysts trapped in the mucus Other species of gelatinous

but it is currently unknown if this indicates widespread mucus-bound chemical defenses amonggelatinous holoplankton

McClintock and Janssen85 examined the feeding deterrent properties of the shell-less Antarctic

pteropod Clione antarctica This circumpolar sea butterfly occurs in vast swarms (up to 300

individuals per cubic meter86) and is likely, therefore, to play an important role in energy transfer

subjected to laboratory feeding bioassays using an Antarctic fish as an ecologically relevant predator

The common circumpolar fish Pagothenia borchgrevinki feeds in the water column on zooplankton but does not include C antarctica in its diet.87 First, fish were presented either a live intact seabutterfly or a control piece of cod tissue of similar size and shape Fish consistently consumed thecod while always rejecting the sea butterflies In a second experiment, fish were presented eitheragar pellets containing sea butterfly homogenates or agar pellets with fish meal alone as a control.Again, fish always consumed agar pellets with fish meal while always rejecting the pellets withsea butterfly homogenates, indicating the butterflies’ chemical defense against fish predation.Employing flash and high-pressure liquid chromatographic (HPLC) techniques, pteroenone(Figure 5.2), a linear β-hydroxyketone and the first example of a defensive secondary metabolite

from a pelagic gastropod, was isolated from tissues of C antarctica.89 When embedded in alginatefood pellets at ecologically relevant concentrations, pteroenone caused significant feeding deter-

rence in P borchgrevinki and Pseudotrematomus bernachii, two Antarctic fish known to feed on

planktonic organisms.86,87 Concentrations of pteroenone were variable among pteropods, but eventhose individuals with the lowest recorded natural concentrations were effectively protected fromfish predation Further chemical analysis indicated that the primary dietary item of this carnivorous

sea butterfly, the shelled pteropod Limacina helicina, does not contain pteroenone, evidence that points to de novo production of this fish antifeedant by C antarctica.86

The cyanobacterium Trichodesmium is one of the most abundant phytoplankters in tropical

seas, contributing a major fraction of new nitrogen and fixed carbon to surface waters.90–92 To date,

however, there are anecdotal and qualitative observations of grazing by crabs and fish on

these copepods gain protection by associating with toxic algae However, Trichodesmium is lethal

to other copepods,96 and McCarthy and Carpenter97 concluded that Trichodesmium must be subject

to little grazing pressure to account for its high standing stocks O’Neil and Roman93 indicated the

necessity of quantitative experiments to determine whether Trichodesmium is chemically defended

against potential crustacean and fish predators (see preliminary studies described below) Otherphytoplankton that produce chemical feeding deterrents include the dinoflagellates.98,99 Copepods

such as Calanus pacificus will reject dinoflagellate prey containing toxins, subsequently

regurgi-tating cells and failing to maintain a full gut, or are killed by the toxins.100–102 Huntley et al.101

suggest that production of feeding deterrents and resultant release from grazing pressure allow theslow-growing dinoflagellates to form significant blooms

McClintock et al.103 examined the feeding deterrent properties of Trichodesmium and eight

other species of common oceanic holoplankton representing five phyla from Bermudian waters.Holoplankton were collected at sites 5–20 km southeast of Bermuda using plankton nets or byconducting blue-water dives and capturing individuals in handheld jars Live holoplankton wereimmediately returned to the laboratory and subjected to feeding trials using the common planktiv-

orous fish Abudefduf saxatilus (sergeant major) as a model predator Large zooplankton (salps,

heteropods, ctenophores) were cut into small pieces, while smaller zooplankton (cyanobacteriacolonies, radiolarians, and foraminiferans) were presented intact to fish, in both cases along withequivalent numbers of squid mantle tissue controls Significant feeding deterrence was detected infish presented with eight of the nine holoplankton species including the colonial cyanobacterium

Trichodesmium, three species of radiolarians, a foraminiferan, ctenophore, and heteropod, and one

of two species of salps These findings indicate that feeding deterrence occurs across a wide diversity

of holoplankton While the basis of the feeding deterrent responses observed were not determined

in this study, these organisms are sluggish and generally depauperate in apparent structural defenses,and, therefore, it seems likely that defense is chemically based Additional studies were recentlyconducted by McClintock, Baker, and Steinberg in Bermuda to address this question (see below).Chemical defenses, as documented here for conspicuous members of the holoplanktonic community,may allow these organisms to play a prominent role in structuring marine food webs For example,short food webs of large organisms allow a large proportion of the primary production to be

production with little subsequent export.24 The size of the grazers in each community determinesthe production rate of large sinking detrital particles; large animals generally produce large, rapidlysinking waste products In addition, some of the pelagic ascidians, such as salps and larvaceans,feed with fine-mesh mucus webs that allow them to feed directly on the smallest phytoplankton,

defense may reduce predation sufficiently to allow these large bloom populations to occur Largeplanktonic protozoa (radiolaria, acantharia, and foraminifera), with associated symbiotic algae, arealso important in elemental cycles because they create a direct link between primary production(via their associated symbionts) and export (they sink during reproduction24,104) Feeding deterrentproperties, as documented here for many of these species, may allow them to play these prominentroles in elemental cycles

McClintock, Baker, and Steinberg (unpublished) conducted further bioassays to examine therole of chemical defenses in oceanic holoplankton from Bermuda Bioassays consisted of presenting

groups of the planktivorous fish Harengula humeralis (red-ear sardine) with pieces of or whole

holoplankton Also tested were alginate/agar pellets containing whole organism homogenates orlipophilic and hydrophilic extracts of holoplankton plus a feeding stimulant, along with appropriatesolvent/feeding stimulant controls McClintock and colleagues found that both whole individuals

and alginate pellets containing homogenates of the large black pelagic copepod Candacia ethiopica

radiolarians were also significantly deterrent toward fish predators, while both intact colonies and

extracts of the cyanobacterium Trichodesmium embedded in agar were also highly deterrent (Figure 5.1) Moreover, extracts of the ctenophore Mnemiopsis macrydi embedded in agar pellets

were essentially deterrent (P < 0.06) to fish predators (Figure 5.1) These findings underscore theneed to continue rigorous studies to evaluate the incidence and importance of chemical defensesamong marine holoplankton

The epiphytic hydroid Tridentata turbinata commonly occurs in association with the pelagic

Sargassum community As such, this hydroid, although not exclusively pelagic (it can occur

benth-ically as well), can conditionally be considered a component of the holoplankton Stachowicz andLindquist105 examined the palatability of this hydroid and its extracts to the sympatric pelagic filefish

Monocanthus hispidus Individual fish were presented bite size portions of hydroid followed by a

The Chemical Ecology of Invertebrate Meroplankton and Holoplankton 209

brine shrimp in order to ensure that fish were not satiated if they rejected a hydroid piece Controlgroup of fish were individually presented only brine shrimp Consistent rejection responses for

T marginata were observed and bioassay-guided fractionation pursued to determine the nature of

the deterrent An assay food consisting of pureed squid mantle tissue (control) or mantle tissue withhydroid extract, both treated with a calcium chloride solution as a hardener, was spread into a thinpaste and cut into small pellets (2–3 microliter volume) Pellets were presented to file fish Thelipophilic extract proved deterrent, and HPLC revealed three novel compounds (tridentatol A–C).Subsequent feeding bioassays revealed that only tridentatol A was deterrent to fish Interestingly,

benthic populations of T marginata contained another novel metabolite, tridentatol D (nondeterrent),

but contained neither tridentatol B nor C Benthic populations did contain the active feeding deterrenttridentatol A, suggesting that there may be intraspecific differences in chemical defenses in response

to selective pressures associated with holoplanktonic and benthic environments

FIGURE 5.1 Histograms showing the feeding-deterrent properties of holoplankton collected near Bermuda

and presented to the common planktivorous fish Harengula humeralis (red-ear sardine) A Multiple trials used whole colonies of the colonial cyanobacterium Trichodesmium sp B Three species of colonial radiolaria.

C Other invertebrates including two species of salps and the copepod Candacia ethiopica were presented

randomly to five groups of three fish in separate aquaria, along with equal numbers of controls (equivalent

sized pieces of squid mantle tissue) Shown is percent acceptance for holoplankton presented intact

(Trichodes-mium sp., radiolarian species 1 and 2, and C ethiopica) or cut into small pieces (salp species 1 and 2 and Spherozoum punctatum) Note: in B and C, stars indicate experiments that were conducted with alginate pellets

containing whole-organism homogenates embedded in alginate pellets containing dried fish powder, as a feeding stimulant vs control alginate pellets with feeding stimulant only D Chemical extracts of the colonial

cyanobacterium Trichodesmium sp., and the ctenophore Mnemiopsis macrydi, embedded in agar pellets

containing dried fish powder were presented to a group of 60 fish in a single, large aquarium Control agar pellets containing feeding stimulant only were presented in equal numbers For all experiments, holoplankton

or pellets that were taken into the mouth and then spit out within a 1-min period were considered rejected Light bars are experimental, dark bars are controls Asterisks indicate experimental treatments that differed significantly from controls (Fisher’s Exact Test; P < 0.05) Unless otherwise noted, n = 10.

Colonial Radiolarians

* 0

10 20 30 40 50 60 70 80 90 100

Hydrophilic Trichodesmium sp

Lipophilic Mnemiopsis macrydi

Hydrophilic macrydi

Cyanobacterium Trichodesmium sp.