Chemical Ecology in Antarctic Seas

James B McClintock & Bill J Baker. American Scientist. Volume 86, Issue 3. May/Jun 1998.

In predation, as in love and war, anything goes. All three of these situations-and the way species respond to them-determine to some degree whether an animal species survives in a particular niche. As a result, some animals develop flashy coloration to attract mates or camouflage to avoid enemies. And for the contest between predator and prey, organisms have developed an impressive armamentarium that includes spines, teeth, claws and quills.

Yet not all the implements of love and war are quite so obvious. Many interactions between species take place on a covert level and are quite difficult to perceive. These interactions are chemical, not physical, and rely on substances such as pheromones that attract mates as well as toxins that repel and even kill predators, competitors and enemies. Chemical interactions alter, sometimes profoundly, the conventional scenarios posited by ecologists studying predators and their prey Traditionally, ecologists have made their speculations on who eats whom and who dominates whom based on the obvious cues of relative size, swiftness and armaments. But when one factors chemicals into the mix, then the slow, the sessile, the small and the toothless may be just as fearsome as creatures that are the more obviously well-armed.

In our work we are looking specifically at chemical relations between marine organisms. In particular, we are interested in secondary metabolites, chemicals that do not seem to be required for any of the primary metabolic processes, such as the energy-producing reactions of respiration or photosynthesis. There is a general consensus that these compounds evolved for specific purposes, but a few may be evolutionary baggage, chemical leftovers from functions long lost from an organism’s repertoire.

Conventional wisdom holds that competition and predation among marine species are most intense in tropical waters, and as a result the chemical ecology of marine organisms in tropical waters has received much more attention than that of organisms dwelling in cooler waters. When we started our studies, we were interested in finding out whether this was justified, so we focused our attention on the Southern Ocean. We began our studies by looking at creatures, such as sponges, soft corals and molluscs, that dwell on the ocean floor in the antarctic. Rooted in biogeographic theory was the assumption that these polar bottom dwellers were not much threatened by fish predators and would therefore not require elaborate chemical defenses.

This was simply not the case. We found that the sessile and sluggish organisms on the antarctic ocean floor were very much threatened by invertebrate predators and competitors and had developed chemical defenses to ward them off. Stimulated by these findings, we then considered what kinds of interactions might be found farther up the water column, where we expected predator-prey interactions to be at least as intense.

We were not disappointed. In fact, looking up we discovered a unique relation between two species that is perhaps without precedent in the animal kingdom. This unusual arrangement, which does not fit into any existing paradigm of interspecific relations, would be a complete mystery if it were not for the chemistry between the two groups.

At the Bottom of the Ocean It was in the middle of the 1980s when we first had the opportunity to dive under the sea ice of McMurdo Sound, Antarctica. The seafloor, so rich in marine life, is covered with a dynamic community of sponges, soft corals, molluscs, snails, tunicates and echinoderms. Many of these organisms are immobile and cannot move to a less densely populated region if the one they are in becomes overgrown with competitors. Nor can they escape predators. Nevertheless they manage to survive. This suggested to us one of two things. Either sessile organisms in the antarctic ocean have no predators, or they have some other ways of defending themselves.

It did not take long for us to determine that bottom dwellers do in fact have many predators. Potential predators indude swarms of the voracious Paramoera antarctica, a one-centimeter-long crustacean resembling a shrimp and dense populations of sea stars. So our next course of action was to look into the chemistry of the bottom dwellers to see whether we could identify any substances that could potentially deter predators.

To date we have isolated dozens of secondary metabolites from bottomdwelling antarctic marine organisms. The sponges have been the most productive. Recently we discovered that the polychaete sponge Isodictya erinacea produces a novel derivative of the amino acid tryptophan we have named erebusinone. This compound appears to mimic a natural compound that keeps crustaceans from molting. When erebusinone is put into food and given to crustaceans of the amphipod family, the amphipods no longer molt and eventually die. The red sponge Kirkpatrickia variolosa, the green sponge Latrunculia apicalis and the cactus sponge Dendrilla membranosa produce variolin pigments, discorhabdin pigments and picolinic acid, respectively. All of these compounds deter sea stars by making their sensory tube-feet retract when they touch the outer sponge tissue.

Of all the antarctic sponges, we have found that the most brightly colored are also the most amply endowed in chemical deterrents. It is tempting to conclude that this is related to warning coloration. However, the antarctic benthos, unlike the floor of more temperate waters, has very few visually oriented predators. One would expect, therefore, very little evolutionary pressure for sponges to maintain their bright coloration in the antarctic. We feel that to explain this pattern, we must look to the sponge’s origins in the more temperate waters of the lower latitudes, where visually orienting predators are more apt to be deterred by brightly colored warnings. Most of the sponges in McMurdo Sound lost their coloration long ago and are now colorless or dully colored. We believe that those that have retained their color may have done so because the pigments also serve as defensive toxins.

We have begun to evaluate this theory by examining the biologically active properties of antarctic sponge pigments. In support of our predictions, we have found that pigments from four colored sponges we have studied, including the red sponge K. variolosa and the green sponge L. apicalis, the cactus sponge D. membranosa and the polychaete sponge Isodictya erinacea have either antifoulant or antipredator properties.


Having found many examples of chemical deterrence in marine organisms on the ocean floor, where little was predicted, we decided to look further up the water column, where we expected competition to be as fierce. We set up a laboratory tank and filled it with a mix of species from the middle of the water column.

Among the species included in the tank was the sea butterfly Clione antarctica, a creature whose bright colors may serve to warn off predators. This bright orange snail, which has lost its shell and evolved wing-like extensions of the outer mantle used in swimming, can be found in abundance throughout the water column. In the tank in our McMurdo Station laboratory, John Janssen of Loyola University in Chicago observed these sea butterflies being carried on the backs of Hyperiella dilatata, a crustacean that resembles a small shrimp. Other members of the amphipod family Hyperiidae are well known to associate with gelatinous zooplankton and are often seen riding on the swimming bells of jellyfish, perhaps using them as a platform to more efficiently dart into the water column to capture prey. Some may even feed on the tissues of their host jellyfish. But what we were observing was substantially different. There seemed no question as to who was carrying whom, as amphipods swam about with sea butterflies in tow. Indeed, this appeared to be an outright abduction.

Concerned that this unique behavior might somehow be an artifact of mixing the two species together in a laboratory tank, we set out to determine whether we could document it in the field. To do that, we set up coastal sampling stations. We drilled holes in the sea ice at each site and placed small fish huts over the holes to protect us from the elements while we conducted our sampling. At a near-coastal site over shallow water no deeper than 30 meters, we placed a plankton net 10 meters deep. At two additional sites further offshore over deeper water, 500 meters and greater, we placed nets at depths of 10 and 50 meters. We positioned our nets horizontally and allowed them to fish passively for 24 hours.

The results of our sampling were quite remarkable. We sampled more than 1,700 amphipods collectively at the three sites. A total of 6 percent of the amphipods at the offshore deep site, 33 percent of the amphipods at the offshore shallow site, and 74 percent of the amphipods at the coastal site were carrying sea butterflies. By peering down into the clear seawater through holes drilled in the sea ice, we made independent observations in situ of sea butterflies being carried by amphipods. Clearly, these amphipods were actively searching out and capturing sea butterflies in the wild, and in some environments doing so frequently.

We found that the amphipod-sea butterfly pairs remained coupled, even after being held in the laboratory for a week. Uncoupled amphipods presented with solitary swimming sea butterflies swam rapidly toward the sea butterfly and grasped its outer tissue, or mantle, with their sixth and seventh sets of swimming appendages, called pereiopods. The amphipods used one each of the sixth and seventh sets of pereiopods to attach to the mantle tissue by gently pinching it. Subsequently, the periopods were used to rotate the sea butterfly on the amphipod’s back. Then, the opposing set of amphipod swimming appendages attached to the sea butterfly, securing it snugly against the amphipod’s upper abdominal segments. The amphipods grasped the sea butterflies with such tenacity that they remained attached during and after handling, as well as after exposure to a 10-percent solution of magnesium chloride, a substance often used as a relaxant.

The nature of this relationship is puzzling. It is difficult to discern how carrying another organism around can be advantageous. We speculated that in fact, amphipods must be considerably slowed down. By carefully measuring the swimming speeds of amphipods with and without sea butterflies, we found that amphipods carrying sea butterflies moved only half as quickly as similarly sized solitary amphipods. The situation simply made no sense. In reducing their mobility, the amphipods became more vulnerable to predators and less adept at capturing prey. Why, then, would amphipods go out of their way to abduct and carry sea butterflies?

Our suspicion was that chemistry played a role in this complex behavioral association. We predicted that the sea butterflies were producing some chemical that deters a predator of the amphipod. To determine whether this was indeed the case, we set about locating a suitable, ecologically relevant predator. We found one in the antarctic planktivorous fish Pagothenia borchgrevinki. This fish is common in midwater- and seaice-associated communities and is one of the most well known predators of Hyperiella dilatata, the key amphipod player in our study. This amphipod is also found among the stomach contents of at least five additional species of antarctic fish. The evidence would suggest that the amphipod is under significant predation pressure. These fish predators are several orders of magnitude greater in size than the amphipods. We devised a series of experiments to determine whether the fish would accept or reject the sea butterflies.

In our experiments, we presented live sea butterflies to the fish. In each trial, a fish approached a sea butterfly took it into its mouth and then, after only a few seconds, spit it out violently, with head shaking. We also presented the fish with the sea butterflies paired with amphipods. These were similarly violently rejected. In general, the amphipod-sea butterfly pairs remained intact throughout these experiments, and both creatures seemed to remain healthy.

In a striking contrast to these experiments, the fish readily consumed pieces of antarctic cod muscle in favor of live sea butterflies. Amphipods were consumed unless carrying a sea butterfly. Our experiments demonstrated that whether alone or on the back of an amphipod, the sea butterflies present a significant deterrent to the fish. We noticed that as we repeated feeding trials, the fish seemed to learn to avoid the sea butterflies. They would approach sea butterflies, stop a centimeter or two away to examine their potential prey, only to turn and swim in another direction.

Safer Living through Chemistry Our experiments clearly demonstrated that something about the sea butterflies was repelling the fish, and we suspected that this deterrence was chemical. To find out, we conducted a second set of feeding experiments. We homogenized the sea butterflies and mixed the homogenate with fish-meal powder to make food pellets. As a control, we also made food pellets containing just the fish-meal powder. We offered both the experimental and the control pellets to fish, which always ate the control pellets and always rejected the pellets containing the homogenate. This provided compelling evidence that chemical compounds might be responsible for the feeding deterrence.

We needed to determine whether the chemical was actually an organic substance made by the sea butterflies or some kind of trace metal or salt that deterred the fish. Ultimately, we hoped to isolate the specific secondary metabolite or metabolites responsible for feeding deterrence. Patrick Bryan, a postdoctoral fellow at the University of Alabama at Birmingham, and Wesley Yoshida, a research technician at the Florida Institute of Technology, both members of our antarctic research team, took the lead in isolating, purifying and elucidating the sea butterfly’s chemical deterrent.

First, they captured more than 4,000 sea butterflies in plankton nets. These were frozen and freeze dried.They then processed the frozen sea butterflies in such a way as to extract the lipophilic substances, those soluble in oil, and separate them from the hydrophilic substances, which are soluble in water. Each of these extracts was then used to make food pellets as described earlier. Bryan and Yoshida then fed the pellets to one of two ecologically relevant antarctic fish. The first is the planktivorous P. borchgrevinki, which we had used in the earlier feeding experiments. The second species was Trematomus bernacchii, which feeds primarily on the ocean floor, including, occasionally, on planktonic organisms. P. borchgrevinki is known to include sea butterflies in its diet, but not the one we were interested in, Clione antarctica. T. bernacchii is not known to feed on sea butterflies, but it does feed on plankton, and we had observed it sampling and rejecting live C. antarctica sea butterflies.

To the best of our ability, we made sure that each group of fish was equally as hungry when we offered them our prepared food pellets. Both species of fish readily ingested all the control and experimental pellets with the exception of one. This was a pellet that contained one of the lipophilic extracts from the sea-butterfly homogenate. These pellets were very occasionally mouthed, but never ingested by the fish.

Having narrowed the search to the lipophilic extract, we refined our extraction further, separating out of the lipophilic extract five pure compounds. We made five new batches of feeding pellets, with each batch containing one of these pure compounds, and again offered them to the fish.

Only one of these compounds was a potent deterrent to fish feeding. We then used a combination of nuclear magnetic resonance and mass-spectrometric analyses to determine the identity of this compound. We found it to be a previously undescribed Beta -hydroxyketone, whose formula is C^sub 14^ H^sub 24^ O^sub 2^. We named the compound pteroenone to reflect the taxonomic description of the compound’s origins as well as its chemical class. Hence, we used “ptero,” meaning “winged,” which we felt referred to the idea of the butterfly. In addition, the pteroenone contains an olefin group, which is expressed in chemical nomenclature as “en,” as well as a ketone group, which is expressed as “one.”

Pteroenone is the first bioactive secondary metabolite ever reported from a planktonic snail. Compounds chemically similar to it have been isolated from bottom-dwelling false limpets and ascoglossan molluscs. Since these molluscs live in shallow water where levels of ultraviolet radiation can be high, it has been suggested that these compounds provide protection from the sun as well as deter predatory reef fish.

We were curious to learn whether pteroenone is actually synthesized by C. antarctica, or whether it might have been derived from another organism eaten by the sea butterfly. It is not unusual for molluscs to derive their chemical deterrents from their prey. Interestingly, sea butterflies, including C. antarctica, frequently prey on other species of sea butterfly. C. antarctica feeds primarily on Limacina helicina, which we were able to collect in abundance in our plankton nets. We pooled 2,000 individuals and extracted their lipophilic compounds. We then analyzed the extractions to determine whether pteroenone was present and found none. This suggests to us that the compound is not derived from diet, but rather that C. antarctica actually makes pteroenone.

The Arrangement

Our experiments demonstrated to us that the sea butterfly C. antarctica synthesizes a deterrent substance that the amphipod H. dilatata exploits for its own protection. This unique association-the abduction of one species by another-is unprecedented in the annals of behavioral and chemical ecology. Some decorator crabs are known to cover their upper carapace with a variety of objects, including the occasional sponge that might harbor defensive chemistry. But this appears to be a nonselective behavior. Crabs haphazardly decorate themselves with whatever is at hand.

The association between the sea butterfly and the amphipod falls within the definition of symbiosis, where two dissimilar species live together in an intimate association. However, none of the relationships defined within the broad context of symbiosis-parasitism, commensalism or mutualismappear to suitably describe the sea butterfly-amphipod relationship.

Parasitism implies that one species associates with another to the detriment of one of them. Often, the parasite feeds off the tissues or body fluids of its host. Mutualism describes a relationship where both species benefit from the association. Commensalism describes organisms that live in benign and neutral association with one another. Neither of the latter two interactions appropriately describes what we have observed, and parasitism provides only a very weak analogy to it. We feel that a new term might be needed to describe the “antagonistic symbiosis” that seems to describe most accurately this unique interaction.

In this association the antagonist benefits greatly from the relationship. Remember that the amphipod must sacrifice its mobility and speed. Clearly the defense it acquires offsets these drawbacks. The sea butterfly, on the other hand, is at the mercy of the amphipod. While it is being carried around, it cannot feed to sustain its energy requirements. Nevertheless, we never found amphipods carrying dead sea butterflies, which clearly indicates that they must periodically release the sea butterflies in order to capture fresh ones.

We were curious to learn how significant this association might be in terms of the general energy flow of the antarctic ocean. To start to get at this question, we tried to assess what percentage of the total biomass is contributed by sea butterflies. We pulled plankton net trawls along the water column under the sea ice at a depth of 10 meters. We determined that at this depth, C. antarctica swarms could be found at densities as high as 300 individuals per cubic meter of seawater. Although sea-butterfly populations can be patchy, large “blooms” represent a significant amount of assimilable carbon and nutrients potentially available to organisms included in oceanic food webs. Similarly, hyperiid amphipods can occur in high densities and likely represent a considerable source of energy and nutrients. That both employ chemical defenses, one directly and one indirectly, has important consequences on population dynamics. In addition, this association adds to the number of sinking particles; that is, it helps to create flocculent detritus and dead organisms that sink to lower levels of the water column.

Since only a handful of studies has even considered the potential role of chemical deterrents in mediating patterns of energy flow in oceanic systems, there remains much to be done to begin to understand how important such interactions may be. Recently with the collaboration of Tony Michaels and Debbie Steinberg at the Bermuda Biological Station for Research and Dan Swenson at the University of Alabama at Birmingham, we found that the planktivorous fish Abudefduf saxatilus was deterred by eight holoplankton species, representing five different phyla. These included a cyanobacterium, two solitary and one colonial radiolarian and a foraminiferan, a ctenophore, a heteropod and a salp. All of these prey items would, on the face of it, seem defenseless against the fish. We have yet to prove that these interactions are driven by chemical defenses. Still, these findings clearly suggest that relations between species in planktonic ocean systems may be more complex than previously considered.

Chemical interactions may also help explain a long-noted puzzle, what ecologists have come to call “the paradox of the plankton.” Ecologists have often observed that plankton species are unusually diverse, in spite of the fact that they live in a seemingly homogeneous environment. The evolution of chemical defenses in some of these species could reduce the number of animals grazing on them, which would increase the number of species that can coexist. Alternatively, chemical defenses may have spurred the evolution of grazers that are highly specialized and can exploit prey in spite of their defenses. For example, some of the holoplankton we examined are already known to have specialized predators-copepods on cyanobacteria, or amphipods on salps. This may in turn facilitate niche diversification. The ecosystem would grow increasingly complex if these predators themselves acquired defensive traits by association with another species, such as the one between sea butterfly and amphipod.

Holoplankton are not alone in their vulnerability in the water column. Another group of organisms is known as “meroplankton.” These take up residence in the water column for some period of their life cycle. Many benthic marine invertebrates produce eggs, embryos or larvae that are released into the water column where they spend some period of time before settling to the ocean floor, where they metamorphoses Among antarctic marine invertebrates, for example, there are many species that produce abundant small larvae that feed on plankton. There is also a relatively large number of antarctic marine invertebrates that release large pelagic larvae that do not feed on plankton. These larger larvae are particularly vulnerable to predation, as they not only lack physical defenses, but in fact almost attract predation. They are conspicuously colored and are sluggish swimmers who spend a considerable amount of time floating in the water column. In addition, they are a rewarding meal, as they are high in nutrients and energy

We recently completed a study of the palatability and chemical defenses of eggs, embryos and larvae of shallow-water antarctic bottom-dwelling marine invertebrates, including those of sponges, molluscs and echinoderms. Many of these bottom dwellers produce large numbers of small larvae that feed in the water column. This group of larvae are palatable to ecologically relevent invertebrate predators. However, there is also a second group of larvae that are larger and rely on their nutritional stores rather than feed on plankton. These larger larvae are generally unpalatable to invertebrate predators. We think it is likely that this second group defends itself chemically. Our findings are similar to those of Neils Lindquist and Mark Hay at the University of North Carolina at Chapel Hill, who have shown that the large, yolky embryos and larvae of tropical corals and sponges are chemically defended from predation by reef fish.

There is an additional benefit to studying the chemical defenses of marine animals. Some of the compounds we find may have applications in fighting human diseases. Many medicinal compounds now being used have been derived from natural sources. And it makes sense that a compound that protects one species from harm may protect another as well. Although we focus our studies on the ecological function of bioactive secondary metabolites, we have also been able to provide crude extracts and pure compounds from marine organisms to laboratories for biomedical screens against human diseases, such as cancer and AIDS. Recently, working with Gamini Jayatilake of the Florida Institute of Technology, we found a novel bioactive metabolite from the antarctic rubber sponge Leucetta leptorhapsis. This metabolite, which we have named “rhapsamine,” is an unusual lipid that appears to be highly active against a human cancer cell line. Although much remains to be done to establish whether rhapsamine has the potential to be developed for human drug therapy, its discovery does suggest a need to sustain marine biodiversity. One never knows where the next useful and important natural products may come from. Our collaboration with cancer researchers also underscores how ecologists studying issues in the basic sciences can indirectly contribute to applied sciences. As is frequently the case, much of applied science has its roots in basic fundamental science.