When an animal has to leave a safe refuge to travel to a patch, it exposes itself to predation risk. In experimental systems, this travel behaviour may be slow and indirect, rather than a fast trip that would minimises exposure time.
I describe a stochastic dynamic program that models animals deciding to move from a refuge to a distant foraging site. The animals pay attention to their energetic reserves as they are metabolising energy, and can only increase reserves when they forage. If a predator appears, the animal turns and runs to shelter, with an enhanced risk of being killed by the predator.
Several optimal behavioural policies are discussed. Typically, an animal following these policies will spend a large amount of time waiting in the exposed region between the refuge and the foraging site, mirroring the slow ‘dawdling’ behaviour seen in many empirical systems when moving to a foraging site.
|Publication status||Published - Aug 2016|