The milker-killer dilemma in spatially structured predator-prey interactions

Consumers can often exploit their resources' potential for growth or regeneration. For example, predators can reduce predation pressure to allow prey density to increase. However, in a well-mixed environment such 'prudent predators' are vulnerable to selfish mutants. In a spatially structured environment, selfish mutants do not have a global effect, and more prudent local exploitation strategies may evolve. Gilpin investigated how group selection may affect predation rate. In this article, we intend to investigate how within-group and between-group selection will affect dispersal strategies of the predators. If a single predator clone exploits a prey patch, the clone would increase the total number of dispersing progeny by partial dispersal before extinction of the prey (the 'milker strategy'), because then the prey population reaches higher numbers and persists for a prolonged period. However, if the prey patch is coexploited by many clones, a mutant cannot increase the interaction time, and dispersal before extinction of the prey is not favoured (the 'killer strategy'). Thus, optimal exploitation strategies depend on the number of foundresses. This number will depend on the exploitation strategies that are prevalent in the predator population at large, and therefore there is 'ecogenetical feedback'. Assuming a simple ecological setting, we investigate how ecology and evolution interact to determine ESS dispersal.