ABOVE: A pair of Labroides dimidiatus cleaner fish eats dead skin cells and minuscule parasites off of a puffer fish. 

Online videos of children doing things they are not supposed to when a parent looks away often rack up views and shares. But do other animals also behave themselves better when they know someone’s watching? Some research in primates has addressed this question; for example, chimpanzees have been found to steal food from potential competitors when their actions are concealed from the competitor’s view. 

Now, in a study published on September 30 in Communications Biology, researchers have found that fish known as cleaner wrasse (Labroides dimidiatus) also behave differently when watched. Females were more likely to cheat, eating forbidden foods for which their male partners would normally punish them, when they knew the males could not see them. These insights support the notion of complex cognition occurring in the tiny brains of these fish, the study authors say.

The work is “among really the first efforts to try and understand whether fish might have some of these really rich perspective-taking abilities that we see in humans,” says Christopher Krupenye, a primate researcher at Durham University and Johns Hopkins University who was not involved in the study.

Cleaner wrasse are tropical reef fish that typically work in pairs to nibble away at the dead skin cells and parasites on their customers, which can be sharks, rays, or other bony fish. “Cleaner fish are an amazing model for cooperation outside of humans . . . they show interspecies cooperation in that they clean these client fish that come to the reef. And they also show within-species cooperation where they work in these mixed sex pairs to clean client fish,” Boston College cooperative behavior researcher and study coauthor Katherine McAuliffe explains. 

However, sometimes the female member of the couple will snack on the mucus of the client fish, an action considered to be “cheating” because it jeopardizes the client-cleaner relationship. If a cleaner wrasse is caught cheating, the client could terminate engagement by swimming away. The female fish is typically the partner who cheats, and the male fish, which is larger than the female, will punish her with aggressive and harassing behavior like biting and chasing. 

McAuliffe worked with a team to investigate whether cleaners are influenced by their partner’s perception when deciding whether or not to cheat. The research group included experts in psychology, behavioral sciences, and cleaner fish biology. They caught wild cleaner fish in French Polynesia and habituated them to tanks set up to mirror natural reef interactions. The experiment’s “client fish” were plexiglass plates with miniscule bits of food dotted on them, with flakes of mixed fish food and prawn representing dead skin cells and mashed prawn for the forbidden mucus. If the cleaner fish cheated and ate the prawn mash off the plexiglass, the researchers quickly removed the plexiglass to simulate a client fish terminating the relationship and swimming away. 

There’s a clear relationship whereby females are more likely to cooperate—i.e., less likely to cheat—when the male is in view.

—Katherine McAuliffe, Boston College

After training the fish, the scientists placed the male partner behind a clear partition while the female fed on a plexiglass client fish, and recorded what she ate. They then released male and observed him for 30 seconds to see whether he punished the female. Next, they tried the same procedure with the male behind an opaque partition that blocked the fishes’ view of one another. Each fish pair underwent 12 trials in each condition. The female cleaner wrasse tended to cheat less and eat the “dead cell” flakes when the male partner was behind the transparent partition and had a clear view. Curiously, the male fish punished the females whenever they cheated, regardless of which partition had been between them. Some males were more aggressive than others in their punishments.    

To test whether females would hide their cheating behavior from male partners if given the chance, in a second, 12-trial phase of the experiment, the fish were put in a setup in which females chose to feed on one of two identical plates, one of which was behind a transparent partition and the other behind an opaque partition separating them from the males. The females whose partners were more apt to aggressively punish them for cheating were more likely to feed behind the transparent partition, allowing the males to see them while they ate. Notably, males were more likely to punish females if they hid behind the opaque barrier, even if the female hadn’t cheated. At the same time, as in the first phase of the study, if the females cheated, males were more likely to punish them than if they ate only flakes, regardless of whether the males had visual access or not. 

Together, the results indicate that cleaner fish are perceptive to what their partners can view them doing, the researchers write in the study. “There’s a clear relationship whereby females are more likely to cooperate—i.e., less likely to cheat—when the male is in view,” says McAuliffe.

See “The Mirror Test Peers Into the Workings of Animal Minds

“The result[s] that I find really interesting, and we still can’t explain, are . . . about how male punishment seemed to not depend on whether he had visual access to her. . . . I think it’s a really interesting possibility that females somehow advertise that they’ve cheated, or that there’s some sort of behavioral signature of cheating that males can pick up on and that’s driving their punishing behavior,” McAuliffe says. In future experiments, she says she’s interested in exploring individual variations of cheating behavior between cleaner fish and identifying any indicators of cheating by the female that instigate male punishment. 

Zegni Triki, who researches the behavioral endocrinology of cleaner fish at Stockholm University and was not a part of this study, tells The Scientist that “the findings are absolutely exciting.” Triki, who has previously collaborated with study coauthor Redouan Bshary, says the work could help put cleaner fish more on the map as research subjects, particularly in fish cognition. 

Krupenye notes broader questions that arise from the paper for the field of behavioral and cognitive sciences, including “how does cognition evolve?” Specifically, he says he wonders, “if you look across fish and birds and mammals and reptiles, would you really see that frogs can [take others’ perceptions into account] too, and flamingos, and is it really this sort of deep fundamental aspect of animal social behavior or vertebrate social behavior? Or is this something that has evolved independently in different lineages?”