91��Ƭ��

Published research co-authored by CU Boulder marine biologist shows that coral reef fish can make dynamic adjustments in their sensitivity to information from other fish to suppress the spread of misinformation

If you want to understand how wild animals deal with socially transmitted misinformation, examining the habits coral reef fish is a pretty good place to start.

In a nutshell, that’s the assessment of a research paper recently published in The Proceedings of the National Academy of Sciences, a peer-reviewed journal of the  (NAS), an authoritative source of original research that broadly spans the biological, physical and social sciences. 

The paper, titled , was co-authored by Mike Gil, University of Colorado Boulder marine biologist and assistant professor of ecology and evolutionary biology. 

At top of page: A fish feeds on algae on a coral reef, fulfilling a vital function for maintaining a healthy underwater ecosystem. Coral reef fish are social beings that are adept communicating information about food, shelter and possible predators to nearby fish. Above: In their recent paper, researchers place several video cameras on a coral reef to continuously observe fish behavior and then using artificial intelligence to help analyze the videos. Still, CU marine biologist Mike Gil has spent considerable time in the water observing fish behavior for himself.

“It turns out that Disney’s ‘Finding Nemo’ might have been closer to reality than you probably thought,” Gil says. “Fish in coral reefs are, indeed, social beings.”

He explains that information from neighboring fish can be very useful, because it can help an individual fish find resources like food or shelter—or it can provide them with an early warning for shared dangers, such as an approaching shark.

“But these reef fish—like humans—can also behave in ways that can mislead others,” he says. “For example, when they are out and about, feeding from the reef, fish commonly mistake benign environmental cues for danger, and will spontaneously flee—as if they’re being hunted by a predator.

“Such dramatic escape behavior is energetically costly to the individual,” Gil says, noting that in those instances the fish—described by researchers as a “first responder”—executes an escape maneuver involving a deep body bend followed by large acceleration and rapid turning. 

That motion, in turn, sends sensory information to surrounding individuals, “telling them that danger could be near, and they, too, may want to stop feeding and swim for their lives,” he adds. 

Given that any fish is capable of such spontaneous flight behavior, Gil says researchers’ initial hypothesis was that the larger the group of fish, the greater the chance that any one fish would trigger a false alarm that could startle the whole group.

“And so our expectation would be that larger groups of fish will be increasingly likely to fall victim to energetically costly misinformation cascades that disrupt their feeding behavior. But, to our surprise, we do not see this,” he says.

A coral reef fish swims in front of Gil, in a wetsuit. In a recent research paper co-authored by Gil, researchers determined that coral reef fish can adjust their sensitivity to information from other fish to suppress the spread of misinformation that could otherwise cause them to overreact and effectively become dysfunctional.

While some escape events involve large response cascades, most involve only one or a few responders, according to Gil. 

“What we found was that when these fish were surrounded by more neighboring fish, they dialed down their responsiveness to information from any given neighbor,” he says. “Each individual fish’s willingness to essentially ignore information from one or a few individuals behaving in an extreme way largely prevented the entire group from being misled by misinformation. Instead, these groups of fish are able to carry on eating, largely undisturbed by frequent exposure to misinformation.”

Gil says the researchers reached their conclusions by placing several video cameras on a coral reef to continuously observe fish behavior and then using artificial intelligence to help analyze the videos. 

In their paper, the researchers showed that in natural foraging collectives escape events in the absence of true predatory threats occur frequently, at a mean rate of one event per 7.7 minutes. Gil says that, if coral reef fish acted on all of the false warnings they are given, the effect would be devastating to both the fish and coral reefs. 

“Perhaps what is most remarkable about our findings is that if reef fish behaved in more conventional ways—including in ways that resemble human behavior—large groups of fish would be distracted so frequently by misinformation cascades that they could become dysfunctional—meaning their ecological role would not be fulfilled,” he says. 

“For the fish species we’re studying, that would mean algae that these fish ‘clean’ (meaning eat) from the sea floor could accumulate and kill corals that form the basis for the ecosystem. Fortunately for the many creatures that depend on coral reefs, including humans, it appears that nature has a built-in defense against such misinformation-induced dysfunction.”