Learning strategies for optimised fitness in a model of cyclic dominance

Abstract: A major problem in evolutionary biology is how species learn and adapt under the constraint of environmental conditions and competition of other species. Models of cyclic dominance provide simplified settings in which such questions can be addressed using methods from theoretical physics. We investigate how a privileged ("smart") species optimises its population by adopting advantageous strategies in one such model. We use a reinforcement learning algorithm, which successfully identifies optimal strategies based on a survival-of-the-weakest effect, including directional incentives to avoid predators. We also characterise the steady-state behaviour of the system in the presence of the smart species and compare with the symmetric case where all species are equivalent.