Redouan Bshary is best known for studying cleaner wrasse—tiny underwater hygienists that pick parasites from much larger fish, like the roving coral grouper. In 2006, Bshary decided to follow one of the groupers to see whether it sought the services of several cleaners in a row. Instead, he saw something wholly unexpected. The groupers repeatedly swam up to giant moray eels and made a vigorous head-shaking signal. It was a call to arms—a signal that meant “Hunt with me”.
The eels respond by swimming off with the groupers. They can slink through crevices and flush out hidden prey, while the groupers are lethal in open water. When they hunt together, little fish have nowhere to flee.
Working with Bshary, Alexander Vail from the University of Cambridge found that the groupers also use a different signal—a headstand—to tell the morays where hidden fish can be found. It’s the equivalent of a human pointing a finger—a gesture that says, “The prey’s in here.” These sorts of referential gestures had only been seen in intelligent animals like humans, apes, ravens, dolphins, and dogs. Their use was often taken as a sign of intelligence. The fact that fish—a group hardly known for their smarts—can use similar signals was surprising.
Vail and Bshary made their discoveries by observing fish in the wild. Now, they’ve brought a closely related species—the coral trout*—into their lab, and tested its partnership with morays through experiments. And they’ve found that the fish’s behaviour is even more sophisticated than we thought. It doesn’t just recruit its partner willy-nilly—it can decide when and with whom to collaborate. It recruits morays when the situation demands it, and it picks the more effective of two possible partners. And it performs just as well as chimpanzees did, when confronted with a similar task in an earlier study.
In 2006, Alicia Melis from the Max Planck Institute for Evolutionary Anthropology presented chimps with an out-of-reach food platform connected to some rope. If chimps could pull the platform closer on their own, they generally did. If they needed a partner, they were more likely to recruit one. And if there was more than one partner available, they chose the most effective one. Melis published the results in a straightforwardly titled paper: “Chimpanzees Recruit the Best Collaborators”.
Vail and his colleagues tried to duplicate the gist of that experiment with coral trout. They placed eight captive trout in tanks with a fake moray eel—a life-size plastic cut-out, nestled in a rocky crevice. They also added a similar cut-out of a small prey fish, which either sat out in the open (where the grouper could snatch it) or hidden under a rock (where a moray was necessary). Right from the first day of testing, the trout tried to recruit the moray far more often when the prey was hidden than when it was exposed.
But not all morays are equal. In the wild, the team saw that some eels are consistently more willing to team up with groupers and trout, while others are reticent collaborators. They simulated this by presenting their captive trout with two morays—one that would launch forwards at the right signal, and one that refused to leave its crevice. On day one of testing, the trout had six chances to recruit a partner, and they went after both eels equally. On day two, they went for the cooperative eel five times out of six.
Vail’s experiment featured the same number of subjects and trials as Melis’s study, and his trout performed as well as her chimps. Of course, there are important differences between the two set-ups. “The trout just had to do something very natural for them, something they’ve practiced for their whole lives,” says Vail. “But the rope-pulling thing was relatively novel for the chimps. It was fairly removed from something they normally do.” However, he adds that the chimps “received quite extensive training in each aspect of their task”, before being exposed to the whole experiment. By contrast, he gave his trout no such training.
“The findings are very exciting. Their results suggest that the fish’s behavior is highly adaptive, and I am not surprised to see similarities in how [chimps and trout] cooperate or choose partners,” says Joshua Plotnik, who has studied cooperation in elephants. “However, as the authors rightly point out, similarities in behavior do not necessarily suggest similarities in intelligence. The authors note that much of the fish’s behavior could be due to learning mechanisms, which do not necessarily require the flexibility of more complex cognition.”
In other words, this doesn’t mean that trout are as intelligent as chimps. They almost certainly don’t show the same wide ranging of sophisticated behaviours. But they can behave in complex ways in the situations that benefit them. They have specific smarts driven by ecological needs, rather than all-round smarts driven by big brains. They remind us, again, that complex behaviour doesn’t necessarily imply complex minds.
Still, this experiment, together with the earlier observations in the wild, suggests that the fish are going beyond simple algorithms like “IF want prey, THEN find moray”. They seem to understand the eel’s role, they recruit it in the right circumstances, and they can direct it to the right place. “It’s always going to be hard to get inside the mind of an animal, but my hunch is that the groupers have an idea of what’s going on,” says Vail. “It shows some of the hallmarks of intentional communication that apes have.”
The study adds to the evidence that fish can be smarter than we thought. We know that several species hunt in teams. Lionfish work together to corral prey with their expansive fins, and have a “Let’s hunt” signal for recruiting their peers. Some electric fish flush out their prey with formation attacks, which they coordinate through electric pulses. And the yellow saddle goatfish hunts in packs where individuals assume specific roles—some chase, others block, not unlike a team of lions, wolves or chimps.
Fish also seems to be particularly good at hunting with other species. The roving coral grouper will also partner with the humphead wrasse, the coral trout will form hunting parties with octopuses, and goatfish sometimes team up with banded sea kraits (a type of sea snake).
And recently, a European team found that three captive cod learned to manipulate a feeding device with a tag attached to their fins, after accidentally getting the tag entangled in the device. It was “one of the very few observed examples of innovation and tool use in fish.”
In fact, Vail suggests that you could view the cooperative hunts of coral trout and morays as a kind of social tool use. “A chimp can get a stick and probe honey out of a hole,” he says. “A grouper has no hands and can’t pick up a stick. But it can use intentional communication to manipulate the behaviour of a different species with the attribute it needs.” In other words: When your prey’s in a hole, and you don’t have a pole, use a moray…
Reference: Vail, Manica & Bshary. 2014. Fish choose appropriately when and with whom to collaborate. Current Biology http://dx.doi.org/10.1016/j.cub.2014.07.033
* Common names are letting us down here. The roving coral grouper and the coral trout are both part of the genus Plectropomus. The coral trout is not closely related to the freshwater trout that you might eat.