Skip to content

Lots we don’t know about coral reef fishes, but some astounding things we do. More reasons to value coral reefs.


This is another in my series of stories from a coral reef – stories in which I seek to tell about the wonders of this marvelous world, and what it has been like as a scientist trying to uncover just a small part.

It was a long time ago. In 1975, E.O. Wilson coined a new term and published the book, Sociobiology. Although he got into a lot of trouble over comments he made about human behavior in the final chapters, it was a wonderful summation of what we knew up till then about the social behavior of animals. With one major failing; and I can still remember my disappointment in realizing that Ed Wilson, that amazing naturalist and leading biologist, didn’t know squat about the behavioral capabilities of fishes. The four pages he devoted to fish talked about schooling, and nothing more.

As an ecologist deeply interested in, and studying the behavior of reef fishes, I knew they were capable of far more interesting social interactions than schooling, and also that there was an abundant literature he had clearly missed. And E.O. Wilson was not the only person to underestimate them – as I remember it, this gaping hole in coverage of what are clearly the most important of all the vertebrates was not mentioned by any of the book’s numerous reviewers.

Any of you who think I am overstating when I call fish the most important vertebrates….. I’m just trying to remedy the anthropocentric view of life that pervades most discourse. An objective view of the vertebrates places all of the terrestrial forms as a single Division (the Tetrapoda), within the Subclass Crossopterygii (which also includes the Coelacanth), of the Class Osteichthyes (the bony fishes). Some authors prefer to group the Tetrapoda and other Crossopterygians with the Dipnoi (lung fishes) and others into a group called the Sarcopterygii within the Osteichthyes, but this is a minor adjustment. The major point is that we and all other terrestrial vertebrates, every mammal, bird, reptile and amphibian, are fish. Period. But, I digress…..

Reef fishes are capable critters

I was reminded of Wilson’s omission when a flurry of tweets caught my eye late this May. The tweets drew my attention to an editorial in Nature overviewing the work of the behavioral scientist, Redouan Bshary, now of the University of Neuchâtel, Switzerland, and to a recent paper in Current Biology, led by a Cambridge graduate students, Alex Vail, with his advisors, Andrea Manica and Bshary, as co-authors (this link is to a pdf of it). The flurry of tweets arose because everyone seemed amazed that groupers were capable of complex, interspecific social behavior. Indeed, the Nature editorial, by Allison Abbott, described how Bshary had awoken to the fact that fish were capable creatures while snorkeling in the Red Sea in 1998, and the tone of her article was on how amazing that fact about fishes really was. Turns out that Bshary’s early career was spent studying primates, and if there is any group of biologists that seeks to believe that cleverness is a trait shared by humans and few other forms of life, it is those who study the behavior of apes and monkeys. I read the Abbott article, downloaded the Current Biology one and read it, and sat there stunned. Because reef fish biologists have been talking about interspecific cooperation in feeding (which is what it was all about) since at least the 1970s. (This site has a whole page of examples of intriguing interspecific cooperative adventures among reef piscivores.)

None of this is meant to diminish the quality of what Vail and Bshary have done (I’ll get to that soon). I just have to tamp down the idea that they have discovered a radically new behavioral capability that nobody had ever suspected. Given that many of the tweets came from fish biologists, my effort to tamp down seems needed (and probably won’t be successful, because why should young scientists doing cutting edge research believe me?). So maybe I should get to the behavior in question.

Bshary’s 1998 epiphany occurred one day when he happened to see a grouper team up with a moray eel to go hunting for prey. To his credit, he remembered what he had seen, and adjusted his expectations. Marshall McLuhan said long ago, “If I had not believed it, I never would have seen it”, neatly summarizing one of humanity’s big problems – we tend not to see things that do not conform to our prior expectations. Bshary adjusted his expectations in 1998, and became a better behavioral biologist by doing so. The paper in Current Biology describes a novel set of aquarium experiments run about a decade later. In the interim, Bshary had ditched primates for reef fish, mastered SCUBA, and begun some productive work on the behavior of cleaner wrasses, Labroides dimidiatus; a fish I have talked about in an earlier post. Bshary’s interest in the cleaner wrasse concerns the way in which the interaction between cleaner and cleanee functions as a market transaction. (The cleaner provides a service – removing parasites and general ‘grooming’ – in return for a fee – the edible parasites and the occasional nibble at a fin.) I’ll leave discussion of that interesting story for another time.

A digression to talk about Lizard Island

Lizard Island
Lizard Island, Great Barrier Reef. The research station is in foreground at right. The more luxurious Lizard Island Resort is in the center (they get the better beach). The outer barrier can be seen on the horizon. Photo from

Alex Vail had the luck to grow up on Lizard Island where his parents were, and still are, the Co-Directors of the Lizard Island Research Station, perhaps the best-run field research facility on the Great Barrier Reef over its 40 year history. Lizard Island, offshore from Cooktown is a ‘high’ island, meaning a continental island rather than a reef-built one. In July 1770, while waiting for his ship, Endeavour, to be repaired on the beach at the mouth of the Endeavour River, Capt. James Cook sailed to Lizard Island and climbed to its highest peak. From there, he was able to discern a route our through the maze of reefs, beyond the outer barrier and into the Coral Sea, so, in one sense, Lizard Island made it possible for Cook to complete his first circumnavigation.

Cook's ship at Endeavour River 10113

Engraving of Cook’s ship under repair at the mouth of the Endeavour River, Queensland.
Image © British Library.

I climbed that hill once, and walked about on top, but the day was cloudy and I could not even see the outer barrier, yet alone a path through it. Still, my pacing about on top of the hill ensured that I had “walked in James Cook’s footsteps”, which is almost as satisfying as wearing his shoes, or riding on his shoulders. But I am digressing again…. Alex Vail undoubtedly climbed that same hill; more relevant for our story, he also got the opportunity to know Redouan Bshary over Bshary’s several visits to Lizard Island, and ended up (presumably he did an undergraduate degree along the way) becoming a student at Cambridge, doing field research back home on Lizard. And thus we come to the point of this bedraggled tale….

Using models in experiments on cooperative piscivory in the coral trout

Labroides dimidiatus & Plectropomus image_full J Fatherree
A coral trout, Plectropomus leopardus, being attended to by a cleaner wrasse, Labroides dimidiatus. Photo © J. Fatherree

The coral trout, Plectropomus leopardus, is a common grouper on Australian reefs. Alex set out to determine experimentally whether the coral trout makes ‘rational’ decisions about whether or not to solicit the help of a moray eel when foraging. To do this he used a technique as old as the study of animal behavior, the use of model animals.

The first use of models to test the behavior of a fish, to my knowledge, occurred in the late 1930s when Niko Tinbergen undertook his classic studies of social behavior in the three-spine stickleback, a tiny fish that is common in small streams and ponds in his native Holland. Building simple models fastened to wires that allowed them to be manipulated in the aquarium occupied by a male stickleback, he showed that a model could elicit courtship behavior if it had a swollen belly mimicking a female with eggs, or aggressive behavior if it had a red belly like another male. Beyond that, the model could be quite crude, scarcely resembling a stickleback at all. Since that time, simple models have proved useful in many studies of fish behavior.

Each of the crude models with a red belly elicits aggressive behavior from territorial male sticklebacks, Gasterosteus aculeatus; the realistic model without does not.

While Tinbergen used sticklebacks about 5 cm in length, housed in small aquaria, Alex Vail worked with coral trout 46 to 59 cm in length. All test animals had been collected from the nearby reefs, and housed for a few days prior to testing. For the main experiment, they were placed singly into a tank about 2 m in diameter and 50 cm deep that contained two structures – a moray hiding place to one side and a prey shelter at the other side of the tank. The prey shelter was surrounded by a clear plastic cylinder so the trout could not actually catch the prey. The prey, a dead frozen pilchard, on a wire could wiggle about in a crevice beneath its shelter, outside the crevice and at the base of the shelter, or in the water column on top of the shelter. The moray eel was a life-size, two-sided, laminated photograph cut to shape and weighted to sit upright on the bottom of the tank. This was rigged on fishing line so it could be moved backwards into, or forwards out of its shelter, and towards the prey. By moving the moray towards the prey while simultaneously moving the prey up to the top of its shelter, Alex could simulate the flushing out of the prey by the eel.

Experimental trials all began with the moray moving 20 cm out of its shelter. Then the prey was presented at the base of its shelter and wiggled until the coral trout saw it. Then, either the prey was pulled immediately to the water column above its shelter or it was pulled to the center of its shelter, within the crevice. In the first case, the trout should ignore the moray and proceed to attempt to catch the clearly visible prey. In the second, it should solicit the help of the moray to flush out the prey.

Alex asked the question, “Will the coral trout solicit the help of the moray only if the prey is inaccessible in its crevice?” He answered that not only would it do that, but it was just as proficient as a chimpanzee in making that decision! (An earlier study by others had examined the ability of a chimpanzee to recruit a second chimp to help solve a difficult task and obtain food.)

Vail 2014 Curr Biol coral trout foraging
The main experiment (A & B) showed that over six days of testing, the coral trout was far more likely to solicit the help of the moray when the prey was in the crevice (green bars, ‘collaborative’) than when the prey moved to the waters above its shelter (yellow bars, ‘solo’). The first day’s performance suggests some learning could be involved (although these wild-caught fish had probably learned to solicit help from morays out on the reef). The results suggest the behavior was even more consistently “correct” than in the analogous experiment with chimpanzees. Results of a second experiment, which tested whether the coral trout could learn to discriminate between an effective (pink) and an ineffective (blue) collaborator moray, are in C and D. Figure © Current Biology.

Photo of the test coral trout and the moray model in the experimental tank. Image © Alex Vail

Now, how did the trout invite the moray to help? It used sign language of a sort. Coral trout ‘interested’ in securing the help of a moray have been seen to shimmy and or flick their dorsal and anal fins at the eel, creating a conspicuous rapid movement, and likely also some vibratory stimulation to the lateral line. And these were the behaviors that they performed towards the model eel in the experiments.

I’ve gone into lots of detail about this study partly because the idea of one fish species assessing a problem and seeking help from a second species is probably very new to most people. But I have also gone into detail to emphasize how simple experiments can reveal so much about what animals do and how they do it. Let us also remember that in the real situation, there is a second species involved, the moray, which responds to the solicitation and joins the hunt.

Other examples of the complexity of reef fish behavior

While behavioral experiments on fish are frequently done in aquaria, they are even more powerful evidence if they can be done out in the field. While the scientists busily studying behavior of chimpanzees may not have been aware, there have been numerous experimental studies of reef fish behavior done out in the field. Ross Robertson’s early studies of cleaner wrasses included experiments in which he removed the male fish and then watched the behavior of the females. In this case the changes in behavior as the dominant female gradually became a male commenced within 30 min and were the first evidence that sex change was taking place. Male courtship behavior was occurring within 24 hr, but viable sperm were not produced for 10 or more days.

Territorial damselfishes, the drab brown but pugnacious ones rather than the more colorful mid-water planktivores that justify the family name, have been important in my own early work. People who first learn about territoriality in animals tend to learn about birds and assume that territorial behavior is usually restricted to other members of one’s own species. But that is not the case, and in the territorial damselfishes, it is usual to defend the territory from a broad range of species on the reef. The defense is not haphazard. Species that are most likely to browse on, or disrupt the animal’s algal turf garden (yes, they cultivate gardens as well), or, for breeding males, species that are likely to feed upon eggs in the nest, are the ones that are attacked. Other species are ignored.

Three-spot damsel by Florent Charpin
The three-spot damselfish, Stegastes planifrons, claimed by Art Myrberg to be “inch for inch, the most dangerous fish in the ocean”. Photo © F. Charpin

We know this fact because of work out of the lab of the late fish behaviorist, Arthur Myrberg, of University of Miami, and particularly the Ph.D. work in the early 1970s by Ron Thresher, now retired and angling for trout in Tasmania, so far as I know. Ron Thresher used model fish experiments in the field to measure the likelihood of a territorial Three-spot Damsel, Stegastes planifrons, to defend its territory. Results appeared in Animal Behaviour in 1976 (unfortunately still behind a paywall). Ron’s models were living fish of various species, enclosed in clear plastic containers that could be positioned at specified distances from the center of the test individual’s territory. He reasoned that the distance the territory holder would go to attack was a measure of the intensity of defense against that species. He showed a reliable pattern of differences among test species that made perfect sense if you examined their feeding or other characteristics. Subsequent work by others has shown this to be a general characteristic among territorial damselfishes – territorial defense is directed to a large number of species that might disrupt the algal food, or steal eggs from a nest, but not to numerous other species. This tells us that these damselfishes are quite capable of telling one species from another and learning which need to be attacked.

Come to think of it, the cleaner wrasse has to be recognized as a cleaner by all those larger, piscivorous species it cleans, and the cleaner wrasse has to ‘know’ that it is OK to fiddle about around the delicate gills, or enter the mouth of a fish ‘waiting’ to be cleaned. And coral trout have to ‘know’ that morays might be able to help them catch prey in difficult to access spots, while morays have to ‘know’ that teaming up to help a soliciting grouper can pay off.

None of these instances of ‘knowledge’ would be too surprising if reef fish grew up in stable family groups and could learn from their parents. But they don’t. With almost no exceptions, reef fish either begin life as eggs spawned into mid-water on an out-going tide, or as eggs in a nest that is cared for only until they hatch (which hatching usually coincides with evening or an out-going tide). They then spend from one to a dozen or more weeks as part of the open ocean plankton, before they reach the stage at which they are ready to settle down on a coral reef. Some of their knowledge about other species may be innate; other must be learned exceedingly quickly because there is not much time for trial and error, mistake-accumulating learning when you are bite-sized and nutritious.

Of course, the piece of ‘knowledge’ that fascinated me when I began working with reef fishes, was the fact that, despite the swift departure from the reef in the earliest hours of life, and the several weeks at sea, reef fishes return not just to coral reefs, but to quite particular places on reefs. The many species that occur on a coral reef are not distributed willy-nilly; they occur in precise types of places, often in a succession of different types of place as they pass from early juvenile to adult life. Having spent many hours looking for, and recording the presence of reef fishes within hours of settlement to a reef, I know personally that you hardly ever see newly arrived fishes in the ‘wrong’ habitat. Either there is a massive kill that weeds out all the unfortunate fish that chose inappropriately, or, more likely, they do indeed somehow ‘know’ where they are supposed to live.

When I was completing my PhD research, the prevailing view was that the dispersal of larval reef fish was a largely passive process – they were carried by currents and tides. Presumably, they got to reefs at an appropriate time to settle also purely by chance. One prominent hypothesis at that time was that large eddy systems that develop downstream of islands or reefs could entrap larvae, which would circulate within the eddy, only to be thrown back towards the reef when they had completed a circuit. This imaginative idea conveniently avoided noting that reef fishes were plentiful on the windward side of reefs and islands as well, and said nothing about how, on settlement, the different species of fish happened to occur in the correct reef habitats! We’ve come a long way since then, and the gradual unraveling of how reef fishes get back to reefs is an exciting example of how nature can always find ways to surprise. Perhaps I’ll write about it some time.

For now, I’ll summarize by noting that reef fishes are very capable creatures. The underwater neighborhoods that I somehow knew existed from the very first time I put my head underwater and looked around are very real, and are occupied by creatures (because it is not only the fish) that know ‘where’ they are, ‘who’ else is around, ‘how’ they may cooperate or not, ‘what’ the activities, postures, or gestures of other inhabitants mean, ‘whether’ it would be sensible to do A and/or B, and just perhaps ‘why’ life is unfolding the way it is. They may not be able to write it all down in English and explain it to us, but they ‘know’ far more than we tend to give them credit for. People like Redouan Bshary and Alex Vail are following in the footsteps of people like Ross Robertson, Art Myrberg, and Ron Thresher by using simple, yet subtle experiments to reveal just a bit of what these creatures are capable of. We need more of this kind of curiosity. If we need another reason to value coral reefs, and therefore to fight for their continued existence on this planet, this could be it; they are home to an amazingly rich community of organisms, living rich and wonderful lives that we are even now barely aware of, and they make our lives infinitely richer just by being there.

2 thoughts on “Lots we don’t know about coral reef fishes, but some astounding things we do. More reasons to value coral reefs.”

  1. Great article. I spent a few days snorkeling at Waterlemon Cay, St John, in April. There were a few coral formations home to the Three Spot Damsel. I wondered why they spent so much time facing me, staring at me. Didn’t know of their territorial aspect until now. Example photos at my website,, are the first image of each gallery.

    1. Yes, just checking you out. If you had got a bit closer, or hung around longer, the larger ones will come out and physically attach, nipping at your mask, reg, arms, etc. Good thing they are small!

Comments are closed.