How do fish use the movement of other fish to make decisions?: From individual movement to collective decision making

Recent experiments by Ward et al. have shown that fish a moving fish group detects hidden predators faster and more accurately than isolated individuals. The increase in speed, in particular, seems to be a consequence of the movement-mediated nature of the interactions used by fish to share information. The present work aims at investigating the link between movement and information transfer underlying collective decisions in fish. We define an individual-based self-propelled particle (SPP) model of the decision-making process analyzed by Ward et al. We fit it to data in order to deduce the smallest set of interaction rules consistent with the experimentally observed behaviour. We infer the relative weight of different social forces on fish movement during the decision-making process. We find that, in order to reproduce the observed experimental trends, both the social forces of alignment and attraction have to be introduced in the model, alignment playing a more important role than attraction. We finally apply this model to make theoretical predictions about fish ability to detect and avoid a moving predator in a natural environment such as open water.