Meeting Abstract
Predation is a primal interaction between species, yet it is unclear what evasion strategies are effective for prey survival. Existing theories suggest that the prey should escape in an optimal direction that maximizes its distance from the predator or in a random and therefore unpredictable direction. Here, we propose several evasion models of zebrafish larvae, including the distance-optimal and random strategies. We built probabilistic models that account for sensory and response noise and used statistical methods to assess these models in comparison to experimental data. This novel approach allowed us to evaluate the relative merits of multiple evasion strategies in predicting the behavior of prey. We found that two strategies are best supported by experimental observations: the distance-optimal strategy and a simpler strategy where prey fish swim orthogonally to the predator’s heading. The orthogonal strategy is a special case of the distance-optimal strategy in the limit of fast predators, yet it requires less sensory effort. We argue that the orthogonal strategy is optimal when considering the neuro-sensory circuits underlying evasion. To probe these circuits, we developed a biomechanical model of the fast response of larval zebrafish that addresses the physical constraints on the motor control of evasion. Taken together, these results suggest that fish adopt a strategy that saves both the perception complexity and the physical difficulty in motor actuation.
