Groupers Invite Morays to Hunt: The Science of Fish Gestures and Interspecies Cooperative Hunting

The interspecies hunt taking place on coral reefs

Animals that hunt in groups are nothing unusual. Wolves, orcas, lions, ants — in every case they cooperate with members of their own species. On coral reefs, however, two fish from utterly different lineages corner their prey by covering each other's weaknesses. The relationship between groupers and morays is the classic example.

The daylight hunter and the crevice hunter

Groupers (family Epinephelidae, order Perciformes) are diurnal ambush predators that specialise in lying in wait and short bursts of speed. They engulf small fish whole with their large mouths, but their thick, rigid bodies cannot enter narrow gaps in coral or rock. The moment their prey slips into a crevice, the hunt is over.

Morays (family Muraenidae, order Anguilliformes), by contrast, thread their long, flexible bodies through cracks in the coral. Most are nocturnal, spending the day tucked into rock holes. Inside a crevice they are close to unbeatable, but in open water they swim slowly and cannot run down a fleeing fish.

How complementary hunting skills multiply success

Crucially, the two do not share the catch. Whichever one swallows the fish takes all of it. The partnership still works because the expected payoff for each individual is higher together than alone. As long as the other's presence raises your own probability of success, cooperation remains stable without any division of the spoils.

406 hours in the Red Sea: evidence that groupers "call" morays

The study that put this phenomenon on a scientific footing was published in PLOS Biology in 2006 by Redouan Bshary and colleagues. It was a large-scale survey built on underwater observations made in Ras Mohammed National Park, Egypt, between September 2002 and December 2004.

The scale of the study and its data

• Subjects: 14 groupers, Plectropomus pessuliferus (total length 55–100 cm), and 7 giant morays, Gymnothorax javanicus (total length 130–200 cm)
• Groupers were followed for a total of 406.4 hours
• Within 286 hours of those records, groupers spent about 31.5 hours (roughly 11%) in the company of morays
• 207 interactions in which the two species associated closely were recorded

If these were chance encounters, the time spent together should be very short. The team calculated the expected duration of an accidental association from the swimming speeds and home ranges of both species. The observed associations exceeded that expectation in 56% of the 207 interactions. This is active company-keeping, not random overlap.

A shake of the head means "let's go hunting"

The grouper positions itself right in front of the moray's den and shakes its head violently from side to side while keeping its body axis steady. All 14 observed individuals performed this "head shake," a movement never seen during normal swimming.

The results were clear. In 70 of the 120 head-shake events (58%), the moray left its den and a joint search or hunt followed. Without a signal, morays moved in only 11 of 38 cases (29%) (chi-squared test, χ²=8.4, p<0.01). In other words, a grouper's signal roughly doubles the chance that a moray sets out.

Roughly five times the catch rate of hunting alone

The payoff has been quantified too. A grouper in the company of a moray caught 0.19 prey per hour, against 0.04 per hour when hunting alone — a difference of about five times. The morays also benefited, taking 5 prey in 829 minutes (0.36 per hour) while with a grouper. A statistical test of the 16 observed successful hunts confirmed a bias that chance cannot explain (binomial test, p=0.007).

Pointing out "it's in there": referential gestures in fish

If the head shake is an invitation to go hunting, another gesture — the one performed after prey escapes — carries an even deeper meaning. It is the headstand signal reported in Nature Communications in 2013 by Alexander Vail, Andrea Manica and Redouan Bshary.

A headstand directly above the hiding place

When the small fish it was chasing escapes into a gap in the coral, the grouper swims to a point directly above that hiding place, turns head-down into a vertical posture and shakes its head. In this posture the grouper was positioned above the spot where the prey it had failed to catch was hiding. Nearby morays or Napoleon wrasses then head to that spot and probe the crevice.

The five criteria for a referential gesture

In animal behaviour research, five criteria have been proposed for recognising a movement as a referential gesture — a gesture that points something out. The team argues that the grouper's headstand satisfies all of them.

1. It is directed towards a specific object (the hidden prey)
2. It is mechanically ineffective (the grouper is not extracting the prey itself)
3. It is directed towards a potential recipient (a moray, for instance)
4. It receives a voluntary response from that recipient
5. It shows hallmarks of intentionality (persisting until the partner responds, looking at the partner, and so on)

Before this paper, referential gestures had been confirmed only in humans, great apes and ravens — all animals with large brains relative to their bodies. The report in fish shook a core assumption of comparative cognition by showing that "pointing something out" does not require a large brain.

Morays are not the only partners

The study covered the Red Sea grouper Plectropomus pessuliferus marisrubri and the coral trout Plectropomus leopardus of the Great Barrier Reef. Besides giant morays, the recipients of the signal included Napoleon wrasses (Cheilinus undulatus) and day octopuses (Octopus cyanea). The fact that the same signal works across different partner species means the gesture is not locked into a single one-to-one relationship.

When to collaborate and with whom: evidence that fish choose

Being able to issue an invitation is not, on its own, enough to call something cooperation. It is a different matter if an animal calls a partner only when one is needed and can choose a partner who is actually useful. Vail and colleagues answered that question head-on in Current Biology in 2014.

Translating a chimpanzee experiment for fish

The model was the famous "rope-pull task" used with chimpanzees. Using an apparatus in which a food tray only moves if two individuals pull ropes simultaneously, researchers test whether an animal (1) calls a partner only when it cannot succeed alone, and (2) can choose the more capable of several candidates. Apart from humans, only chimpanzees were thought to manage both.

The team translated this into terms relevant to coral trout ecology. The setting was the reef around Lizard Island Research Station, run by the Australian Museum. They prepared a "good" moray model that came to help when signalled and a "useless" model that swam off in the opposite direction, and observed which one the coral trout signalled to.

The result: calling when needed, learning who is effective

The coral trout raised the frequency of calling a moray precisely in situations where they could not reach the prey themselves. And over successive trials they came to choose the model that would help. The paper concludes that, in combining these two abilities, the coral trout performs on a par with chimpanzees.

These collaborative abilities are not specific to apes and may be more closely linked to ecological need than brain size or relatedness to humans.

― Summarised from the conclusion of Vail, Manica & Bshary (2014) Current Biology 24(17): R791–R793

What the moray gets out of it, and why the partnership holds

So far the story has been told from the grouper's side. What about the moray being called out? For a nocturnal animal, being roused in daylight and swimming around costs energy.

There is a payoff for the moray too

In the Red Sea observations, morays accompanying groupers caught 0.36 prey per hour. Because a moray has very few opportunities to hunt alone during the day, the grouper's guidance delivers a meal it would otherwise never have had. And when the grouper marks a hiding place with a headstand, the moray is spared the effort of searching crevices at random.

Nothing is shared, so nothing is fought over

Cooperative relationships tend to collapse over how the rewards are divided. With groupers and morays, however, the catch is swallowed whole by whoever gets it — the step of dividing it does not exist. Which of them wins any given fish is close to luck, and over the long run both do better than they would alone. The structural absence of an opportunity to cheat is what keeps this relationship stable.

Are there familiar faces?

The Red Sea observations did record a tendency for particular pairs of individuals to appear together repeatedly. Whether groupers and morays remember each other as individuals, however, is still unsettled. The finding that coral trout learned which model was effective in the 2014 experiment shows that they can at least update their assessment of a collaborator from experience.

When does this partnership break down?

Cooperative hunting depends on a surprisingly narrow set of conditions. First, without complex terrain full of crevices, the moray's speciality simply ceases to exist. If bleaching or crown-of-thorns starfish outbreaks collapse the coral skeleton and the reef flattens into rubble, both "prey that escapes into a gap" and "any reason to call a crevice specialist" disappear at the same time.

Second, both species must be present at sufficient density. Groupers are prized by fisheries and are taken selectively from the largest individuals; morays too are fished in some regions. If either declines, the chance of meeting declines with it, and the partnership no longer forms. A behaviour can be lost even while the species survives, if the environment changes — which is why conservation discussions increasingly speak of protecting relationships, not just species.

It also remains unknown whether the two species always cooperate wherever they co-occur, or whether this is a behaviour learned and passed down locally. Recording it before the behaviour itself is lost is the one point on which researchers in this field agree.

Octopus and fish hunting parties: the role division revealed in 2024

Interspecific cooperative hunting is not limited to pairs. A study by Eduardo Sampaio and colleagues, published in Nature Ecology & Evolution in 2024, used three-dimensional tracking to analyse hunting groups formed by an octopus and several fish species, revealing their internal structure.

More than 100 hours of footage analysed in three dimensions

The fieldwork took place in the waters of Israel, Egypt and Australia, gathering more than 100 hours of footage. At the centre is the day octopus (Octopus cyanea), accompanied by goatfish such as the goldsaddle goatfish, the blacktip grouper (Epinephelus fasciatus) and the lyretail grouper (Variola louti) — a cast that can also be seen in the southern seas of Japan.

The fish search, the octopus decides

The analysis showed that leadership is not held by one party but divided by role. Ranging widely to locate prey and setting the group's direction falls mainly to the goatfish. Meanwhile, the octopus controls when the group sets off. The octopus pushes an arm into a crevice to flush out prey, and the fish take whatever bolts from it.

The octopus's "punch" keeps order

These groups also contain individuals that do no work and simply skim the rewards. That is where the behaviour reported in Ecology in 2021 comes in: the octopus striking fish with an arm. The 2024 study framed this "punch" as a way of shutting out freeloading fish and protecting the octopus's own share. Behind every cooperative arrangement lies the universal problem of dealing with free riders.

The same story extends to humans hunting with animals

Groupers and morays are simply the best-studied case of interspecific cooperative hunting, not the only one. Widen the view and it becomes clear that humans, too, are a species that has hunted in partnership with wild animals.

Dolphins and fishers in Laguna, Brazil

In the southern Brazilian town of Laguna, wild bottlenose dolphins and cast-net fishers have caught mullet together for more than 140 years. The dolphins herd schools of mullet towards the shore and then make a distinctive movement at the surface, such as arching the body. The fishers read that cue and throw their nets. The dolphins take the mullet that scatter as they escape the net.

A study published in PNAS in 2023 quantified the relationship using 15 years of fieldwork. The team tracked mullet schools with sonar cameras, recorded dolphin behaviour with drones and underwater acoustics, and followed the movements of 177 fishers with GPS while weighing their catches. Of roughly 3,000 net casts, about 46% were successful, and the more closely the fishers timed their casts to the dolphins' cues, the larger the catch. On the dolphins' side, individuals that take part in this cooperation have been shown to have higher survival.

Honeyguides and honey hunters in Africa

Another famous case is the relationship between the African bird known as the honeyguide and people. When the bird finds a wild bees' nest, it flies to a person and calls while leading them to it. After the person breaks the nest open and takes the honey, the bird eats the beeswax left behind. A bird that cannot split wood and a person who is poor at finding nests: a textbook complementary pair.

A study published in Science in 2016 focused on the distinctive call — a trill followed by a grunt — used by the Yao people of the Niassa National Reserve in Mozambique. Making this sound raised the probability of being guided by a bird from about 33% to about 66%, and the probability of reaching a bees' nest from about 17% to about 54%. Ordinary noises produce no such effect. A wild bird has learned to understand a specific human signal.

What they share: a signal and complementarity

All three cases come down to just two ingredients. (1) Each party can do something the other cannot (complementarity), and (2) there is a signal that conveys intent (communication). Where those two are present, neither difference in species nor difference in brain size is an obstacle to cooperation.

Undoing the assumption that fish are stupid

Many people have heard the folk claim that a fish's memory lasts three seconds. There is no scientific basis for such statements. Recent research on fish cognition has been piling up evidence in the opposite direction.

A fish that recognises itself in a mirror

Japanese research has drawn worldwide attention in this field. In 2019, the group of Masanori Kohda at Osaka City University (now Osaka Metropolitan University) reported in PLOS Biology that the cleaner wrasse (Labroides dimidiatus) passed the mark test. When a mark was placed where the fish could not see it without a mirror, such as under the throat, the fish looked in the mirror and then rubbed that spot against the bottom.

The mark test has been passed by only a limited set of animals, including chimpanzees, elephants and magpies. That a fish about 10 centimetres long cleared it forced a rethink of the assumption that self-recognition requires a large brain. The research team themselves argue that the result is not proof that fish have human-like self-awareness, but a reason to re-examine what the test itself means.

Ecological need matters more than brain size

Both the grouper's partner choice and the cleaner wrasse's self-recognition point to the same thing: what matters is whether an animal lives a life that requires the ability. Groupers live where prey escapes into crevices, so calling a crevice specialist is worth doing. Cleaner fish interact with many client fish every day, so distinguishing others from themselves is necessary.

A coral reef is a cognitively demanding environment, crowded with countless crevices and a great diversity of species. How much biodiversity that structural complexity generates is covered in Why are coral reefs called "rainforests of the sea"?, but the same complexity has also been a pressure fostering intelligence in fish.

Groupers and morays you can meet in Japanese waters

This may sound like a story about the Red Sea and the Great Barrier Reef, but relatives of the main characters are common in Japanese waters too. The coral reefs of the Nansei Islands in particular are a very similar stage.

Coral trout (akajin mibai) and the giant moray

The coral trout (Plectropomus leopardus), the star of the 2013 and 2014 studies, is exactly the prized food fish known in Okinawa as akajin mibai. It ranges from the western Pacific to the Indian Ocean and is landed mainly around the Nansei Islands in Japan. The giant moray (Gymnothorax javanicus) also inhabits the reefs of the Ryukyu Islands and, at up to three metres in total length, is among the largest of all morays.

Groupers as a fishery resource, and stock enhancement

Groupers grow slowly, and many species only begin breeding once they are large, which makes them vulnerable to overfishing. In Japan, hatchery production and release of juveniles is being carried out in several prefectures, centred on the Hong Kong grouper (akou). Some local governments, such as Yamaguchi Prefecture, also restrict the size at which the fish may be taken. For the coral trout, technical development of juvenile production has continued at facilities in Yaeyama, Okinawa.

Rules to follow when watching them

Cooperative hunting only happens when the fish are calm enough to behave naturally. If you come across it while diving or snorkelling, do not chase; keep your distance and watch. Morays in particular may bite defensively if a hand comes near their den. Never put your hand into a hole, and never feed the fish — these two rules must be kept.

Cooperation is not the exclusive property of big brains

Twenty years of research into grouper–moray cooperative hunting has quietly redrawn the outline of what we call intelligence. Sending a signal, waiting for a response, pointing out a hiding place, remembering and choosing a useful partner — behaviours once considered the privilege of primates have been found in a fish whose brain weighs a few grams.

Ability comes from how you live, not from your lineage

Crucially, these abilities were not inherited from humans. The fish and primate lineages diverged more than 400 million years ago. It is more natural to conclude that similar abilities arose independently, many times over, from similar needs. Biologists call this convergent evolution. The same thing that happened with wings in birds, bats and insects has happened with the capacity to cooperate.

Change how you look, and the sea changes

This perspective feeds back into how we relate to the sea. Reef fish are not merely "resources that happen to be there" but individuals that form relationships, make judgements and learn as they live. If the structure of a coral reef is lost, the crevices go with it — and so do the partnerships built around those crevices. What needs protecting is not only a list of species but the relationships being lived out among them.