Wild Animals Stop the Spread of Socially Transmitted Misinformation
For wild animals, false alarms are the most widespread form of misinformation. (Photo credit: Kaylee Rose Fahimipour)
Despite the benefits of learning about the world through social ties, social connections also provide a conduit for misinformation that impedes effective decision-making.
For wild animals, false alarms are the most widespread form of misinformation. For example, when an individual animal in a group makes the decision to produce an alarm signal or initiate an escape maneuver in the absence of a real threat. This initial action produces sensory stimuli that can be perceived by others in the group as an indication of danger, resulting in a cascade of erroneous escape responses that can spread contagiously.
Behavioral and neurophysiological studies suggest that relatively simple behavioral strategies control decision-making in many of these settings. Yet, it is unknown whether these strategies somehow account for the possibility of exposure to misinformation.
To address this question, researchers from Florida Atlantic University, Cornell University and the University of Colorado Boulder deployed camera observatories in a coral reef in French Polynesia to continuously record behavioral decision-making of wild, mixed-species groups of foraging fish. For the study, they analyzed the video recordings to explore how individual decision-making affects the spread of misinformation. To understand why, the researchers reconstructed the visual sensory information available to each fish before and during escape events, and modeled the fish’s decision-making processes to respond or not respond.
Results, published in the Proceedings of the National Academy of Sciences, suggest that even in the absence of predators, escape events occur frequently in natural groups of foraging fish but rarely spread to more than a few individuals. These animals produce and perceive visual motion cues produced by others, thereby forming dynamic information exchange networks.
“These networks are surprisingly robust to false alarms that occur when one individual flees in the absence of a true shared threat,” said Ashkaan Fahimipour, Ph.D., corresponding author, and an assistant professor in the Department of Biological Sciences in FAU’s Charles E. Schmidt College of Science. “By reconstructing visual sensory inputs to each animal, we show that this robustness to misinformation about threats inherits from a specific property of their decision-making strategy: dynamic adjustments in sensitivity to socially acquired information. This property can be achieved through a simple and biologically widespread decision-making circuit.”
In fish, escape responses are controlled by specialized neural circuits that process incoming sensory stimuli, including visual motion stimuli, and route signals to premotor neurons in the hindbrain. Therefore, the researchers hypothesized that natural escape events are initiated and spread through visual stimuli produced when individuals in the group move.
Findings from this study were inconsistent with prior hypotheses that organisms average or pool the behavioral cues from neighbors to make decisions. The researchers suggest that fish dynamically adjusted their responsiveness to visual cues based on the recent history of sensory inputs from neighbors.
According to Fahimipour; Andrew M. Hein, Ph.D., corresponding author and an assistant professor of computational biology, Cornell University; Michael Gil, Ph.D., an assistant professor at the University of Colorado at Boulder, and the other co-authors, their dynamic decision-making model is more robust to misinformation and could be achieved by a simple neural circuit found in many other animals.
“It will be interesting to investigate whether the mechanisms revealed here also are important in driving individual decision-making and misinformation spread in other biological and social systems,” Fahimipour.
Fahimipour was supported by the Research Associateship Program from the National Research Council of the National Academies of Sciences, Engineering, and Mathematics. This work was supported by the National Science Foundation grants (IOS-1855956 and EF-2222478).