This singular persistence of the fish to the same spot, and to the close vicinity of the great anemone, aroused in me strong suspicions of the existence of some connection between them.

These suspicions were subsequently verified...though what is the nature and object of that connection remains to be proved.

Dr. Cuthbert Collingwood: Rambles of a Naturalist on the Shores and Waters of the China Sea (John Murray, London, 1868, pages 151-152)

More than a century later, people continue to be intrigued by that mysterious "connection." Two questions come immediately to mind. We dealt in the Introduction with how small fish can survive in such a hostile environment. The other concerns the advantages and disadvantages of the relationship to the partners.


The word symbiosis literally means "living together," implying no judgment about the benefits or detriments to either partner. In some countries, however, it is used as a synonym for "mutualism", a relationship from which both partners gain advantage. "Parasitism" is a type of symbiosis in which one partner benefits to the detriment of the other. There are two problems with invoking the latter terms. First, a relationship between organisms of two species may be either positive or negative, depending on circumstance. For example, a small amount of a virus can provoke immunity without disease, but a large amount can result in illness. The illness might be fatal in a very young or very old person. Thus, the "symbiosis" between a virus and a human may be good, indifferent, or harmful. Second, a great deal of understanding is necessary to assess the pluses and minuses accurately.

Recall that clownfishes are never found in nature without an anemone; this is an obligate association for them, although in captivity they are capable of living by themselves. It seems obvious that they are protected by the anemone with which they live -- when threatened, they dive among its tentacles, from which most other fishes remain distant. We have taken clownfish far from their anemones, and have removed anemones from beneath their fish. Poor swimmers, they sooner (in the former instance) or later (in the latter) became prey of larger fish. The presence of an anemone is also essential to reproduction of the fishes: their eggs are laid beneath the oral disc overhang of the anemone, where they are tended by the male (see chapter 4).

In an aquarium, without an actinian, captive fish will bathe among air bubbles or frondose vegetation, so we infer that they obtain tactile stimulation from anemone tentacles. And the claim of some aquarists that the fish are livelier and healthier when kept with anemones suggests other benefits as well.

Indeed, aquarists have added much to knowledge of this symbiosis. Many have seen fish bring food to their anemones. This behaviour seems confined to aquaria. The normal diet of clownfishes is small plants and animals that live in the water above the anemone, or algae that grow around it (chapter 4). In nature, they do not encounter large particles of food, so they eat their food where it is found. Feeding large morsels to a fish in an aquarium produces an artifact: the fish, unable to devour the piece immediately, takes it home to work on it in the relative security of its own territory, as is typical of predators that obtain food in large amounts. But the territory in this case consumes the food!

Benefits or detriments to the anemone are not obvious. But neither do the fishes seem to harm their hosts. Therefore, many biologists have considered this a strictly one-sided relationship. The occasional anemones found in nature lacking fish support this conclusion -- they seem to survive perfectly well without fish.

Or do they? On the Great Barrier Reef, at Enewetak Atoll, and in Papua New Guinea, we removed clownfishes from hosts of the species Entacmaea quadricolor to determine how long it would take for new fish to repopulate them. But usually within 24 hours, there were no sea anemones at all! Instead, butterfly fish were poking their long snouts into crevices where the actinians had been anchored. In the absence of protection by their symbionts (which rarely happens in nature), anemones of this species fall prey to butterflyfishes. Hans Fricke made similar observations in the Red Sea, and Jack Moyer in Japan. Presumably, whatever protects butterfly fishes from the nematocysts of corals, on which they normally feed, does so as well for sea anemones. Thus, the relationship with its fishes is mutualistic for this species of anemone: the partners obviously profit by living together, protecting one another from predators.

Other, larger anemones can survive without fish, and apparently suffer no harm, but may do better when occupied by a clownfish. When threatened, a typical anemone withdraws its oral disc, closing its upper column over its tentacles. This is impossible in many host actinians because of their widely flared oral discs. Through evolutionary time, the anemone may have lost the ability to close because the fish protected it.

Other, minor benefits could be the fish eating parasites of the anemones, and fanning their hosts to create increased circulation of water over them. Clownfish do not, however, lure small fishes so that they can be eaten by the anemone, as once was believed. In fact, clownfish tend to keep other fish away from the anemone.


The relationship between fishes and anemones varies with the species of partner. One cannot speak of the anemone/fish symbiosis: there are nearly as many variants of it as there are combinations of species. To begin with, of the 28 fishes, only Amphiprion clarkii naturally occurs with all 10 host anemones; about a third of the fish are known from a single species of host, and the rest occur with several hosts. Conversely, E. quadricolor, H. crispa, and possibly S. mertensii all host 12-14 fishes, whereas two play host to fish of only one species.

What governs which species occur together? Clearly, the fishes, being the shorter-lived and mobile partner, are responsible for this pattern. But how do they choose? Only species that occur in the same geographical area and have similar ecological preferences -- sand or reef, deep or shallow -- can potentially live together (although not all of them do). The sand-dwelling anemone S. haddoni provides an example of how ecology affects species specificity. Amphiprion polymnus, one of the largest and most aggressive of anemonefish, lives with it in relatively clean sand and deep water. Where does such a fish lay its eggs? Egg clusters are attached to a solid object tucked beside the anemone's column. We have seen sand dollar tests, chunks of wood, and even a soft drink can or bottle used in this manner. In several instances, a groove in the clean sand extending some distance from the object on which the eggs had been laid provided evidence that the fish had dragged the object to the anemone. Because a hard substratum is necessary for reproduction, a fish large enough to drag an appropriate object to its anemone is required for success in that environment.

Beyond geographical and ecological coincidence, we believe three factors affect specificity between partners. 1) Fish have an innate or learned preference for only some of the anemones potentially available for colonization. They may even choose based on habitat of the host: A. melanopus inhabits clones of E. quadricolor on tops of reefs, whereas P. biaculeatus lives in solitary specimens of E. quadricolor on reef slopes. 2) From among that subset of acceptable hosts, there is competition. The most host specific fish are those that are, generally, competitively superior for preferred anemones (for whatever reasons some anemones are preferred over others). 3) Chance.

If a fish's own mucus protects it from being stung, presumably it can occur only with actinians to which it has evolved a protective mucus. A fish larva settling into any anemone other than one of the "right" species would be killed. Indeed, some anemonefish are killed when placed in host actinians with which that fish species does not normally occur. There is evidence, however, that specificity is due to more than simply the deaths of all larvae that do not happen to settle into an appropriate host. Miyagawa found, in aquaria, that newly metamorphosed fry of some species locate an anemone by chemicals that are constantly being released by the anemone, much as a salmon senses its home stream, and that vision plays no role. These chemicals differ among species, so larval fish are attracted to anemones of species with which fish of that species naturally occur, but not to anemones of other species. However, fry of other fish are not attracted to anemones with which they naturally occur. So clownfishes may differ in how they select and locate hosts, as well as how they are protected from them.

If fish are protected from being stung by a coating of actinian mucus, it follows that they should be able to adapt to anemones with which they have had no previous contact (as an individual or a species). Indeed, a popular host anemone for home aquaria in the US is the Caribbean species Condylactis gigantea, and some clownfishes have also been adapted to European and American actinians.

In either case -- whether protection is from fish or anemone mucus -- chemical attraction of larval fish to host actinians is theoretically possible, and is consistent with some fishes being general in their host preferences and others being specific. The recognition of chemicals may be innate. Alternatively, it may be learned early in life. Recall, anemonefish eggs are incubated beside an actinian. During incubation, chemicals from the anemone may penetrate the egg case and imprint the embryonic fish. This is also analogous to how a salmon learns the "smell" of its home stream.

We believe that anemones having the greatest number of symbionts (10 or more in nature) are preferred by fishes for some as-yet-unknown reason. It may be that their chemical attractants are especially powerful, but that, in turn, may be a result of some other advantage they impart to their fish. Conversely, if it is advantageous to an anemone to have fish, they might have evolved a particularly potent attractant. In the equatorial tropics, E. quadricolor cannot live without fish; that anemone is among those with the greatest number of associates, which may assure it of having fish wherever it is. It is no coincidence that the two actinians associated with only a single species of fish both harbour A. clarkii, the least host-specific and most geographically widespread species of anemonefish. It owes its success, in terms of numbers and geographical extent, to its ability to occupy any host anemone.

Preference of fish for only certain anemones explains part of why there are only some species combinations in nature, but at least two other factors appear to have an influence. One is competition among fishes for anemones. Once fish of a particular species occupies an anemone, they chase out newcomers of other species, with rare exceptions. However, if an anemone is empty, and fish of different species settle in at about the same time, there can be competition. Fishes with greater host specificity are better competitors than those with less specificity. This is only reasonable: a fish that can live in only one species of anemone is entirely without refuge if it cannot prevent fish of other species from occupying that anemone. A fish with wide tolerances can be chased from one kind of anemone and still have others available.

Occasionally a fish known to be a poor competitor is found in a highly preferred anemone. We suspect that that anemone missed being colonized by a member of the superior fish species, and so was available for colonization by one of the inferior species. Once the resident fish grew a bit, it could prevent small individuals of other, perhaps competitively superior, species from occupying the anemone.


Sea anemones of the species Heteractis aurora and H. malu seldom contain sexually mature pairs of fish; we term them "nursery anemones." Perhaps the hosts are inappropriate for large social groupings for some biochemical reason, although they are sufficiently hospitable that fish survival and growth to a certain size are possible. Given that a shortage of hosts limits natural anemonefish populations, we speculate that these actinians may allow resident fish the potential to move into actinians of more appropriate species when larger, jumping the queue, as it were. Some anemonefishes wander from their hosts, as juveniles or in southern Japanese waters when reproduction ceases in the winter. Fish are highly vulnerable to predation under such circumstances, but may survive sufficiently often for settlement into nursery anemones to be adaptive. Of course, a fish settling in such a host cannot know this, but (if it happens this way) the fish would have at least some chance of reproducing, whereas it would have none if it either remained in the "nursery anemone" its entire life, or, in the absence of space in a "better" species of host, settled into none at all. The basis for this effect of anemone upon fish is totally unknown.

In another instance of anemone affecting fish, the normally orange-coloured portions of the fish darken, so that the fish is black, rather than orange, with white stripes. This type of melanism differs from that associated with size and certain isolated geographical populations (see Chapter 2). Only certain species of fish react this way, and only in certain species of actinians -- for example, A. chrysopterus in S. mertensii, and A. polymnus in H. crispa. Such changes may be relatively rapid, so that an orange fish transferred to another anemone will darken within a matter of hours. Lightening, once removed from that host, generally occurs more slowly. The adaptive value of this reaction to either partner is unknown.

The fish can also affect its anemone. In the presence of a resident fish, tentacles of E. quadricolor bulge near the end, but in the absence of a fish, the tentacles commonly lack bulbs. Specimens of this anemone are often identified as different species based solely on tentacle form. But in all other respects they are indistinguishable. The transformation of an anemone from a member of the non-bulbous "species" into a member of the other putative species can be effected by placing an anemonefish among its tentacles, which develop bulbs within minutes. The reverse occurs when a fish is removed, although more slowly. The bulb exposes a larger surface area of the tentacle to sunlight, so that the algae may be able to gather more solar energy, but why that should happen only in the presence of fish, and how it occurs, are enigmas.

Several workers in the Red Sea have reported A. bicinctus diving inside E. quadricolor, although others have failed to find it. Some of the observations have been in the aquarium, where atypical behaviour often occurs. We have seen large adults of A. polymnus enter the mouth of S. haddoni under field conditions in Papua New Guinea. We do not know precisely what causes this behaviour -- it does not always occur when the fish are pursued.

No doubt other interactions specific to particular combinations of anemones and fish will become clear as the symbiosis is investigated further, and names of both partners are used in a consistent manner.

Return to Table of Contents
Proceed to Chapter 6