Peter Ashcroft, Philipp M Altrock, Tobias Galla
(Submitted on 21 Jun 2014)
The environment in which a population evolves can have a crucial impact on selection. We study evolutionary dynamics in finite populations of fixed size in a changing environment. The population dynamics are driven by birth and death events. The rates of these events may vary in time depending on the state of the environment, which follows an independent Markov process. We develop a general theory for the fixation probability of a mutant in a population of wild-types, and for unconditional and conditional mean fixation times. We apply our theory to evolutionary games for which the payoff structure varies in time. The mutant can exploit the environmental noise; a dynamic environment that switches between two states can lead to a probability of fixation that is higher than in any of the individual environmental states. We provide an intuitive interpretation of this surprising effect. We also investigate stationary distributions of the population when mutations are more frequent. In this regime, we find two approximations of the stationary measure. One works well for rapid switching, the other for slowly fluctuating environments.