Date:
Location:
Dr. Roy Harpaz
Weizmann Institute, Israel
Groups of organisms, from bacteria to humans, exhibit complex patterns of collective behavior. Understanding the individual computations and social interactions at the basis of these behaviors and characterizing the advantages they confer on the individual are of great ethological and theoretical interest. To infer individual computations underlying collective behavior, we studied freely swimming groups of zebrafish. We found that individual fish transition between at least two distinct modes of social information processing and perform movement decisions based on the spatiotemporal social information available to them. Following these findings, we devised a new mode-dependent ‘Receptive Field’ model of fish social computation. Our model successfully predicted fish swimming trajectories, outperforming previously suggested models that assume a single continuous computation. Characterizing the effects that social interactions have on individual and group performance, we further studied social foraging behavior of fish in a novel semi-ethological experiment. We found that individuals can detect shoal-mates’ unique behavior while finding food, and adjust their foraging decisions accordingly. Based on these observations we constructed a model of fish social foraging behavior and compared social and independent foraging using model simulations. Our suggested social foraging mechanism increased foraging efficiency in simulated groups, and closely matched the efficiency found in real groups. These findings suggest that social computations of the individual are more complex than previously assumed, hinting towards the information processing capabilities of these fish and towards the neuronal computations that underlie these behaviors.