Social learning by mate‐choice copying increases dispersal and reduces local adaptation

Abstract

In heterogeneous environments, dispersal may be hampered not only by direct costs, but also because immigrants may be locally maladapted. While maladaptation affects both sexes, this cost may be modulated in females if they express mate preferences that are either adaptive or maladaptive in the new local population. Dispersal costs under local adaptation may be mitigated if it is possible to switch to expressing traits of locally adapted residents. In a sexual selection context, immigrant females may learn to mate with locally favoured males. Mate‐choice copying is a type of social learning, where individuals, usually females, update their mating preferences after observing others mate. If it allows immigrant females to switch from maladapted to locally adapted preferences, their dispersal costs are mitigated as mate choice helps them create locally adapted offspring. To study if copying can promote the evolution of dispersal, we created an individual‐based model to simulate the coevolution of four traits: copying, dispersal, a trait relevant for local adaptation, and female preference. We contrast two scenarios with copying—either unconditional or conditional such that only dispersers copy—with a control scenario that lacks any copying. We show copying to lead to higher dispersal, especially if copying is conditionally expressed. This leads to an increase in gene flow between patches and, consequently, a decrease in local adaptation and trait‐preference correlations. While our study is phrased with female preference as the learned trait, one may generally expect social learning to mitigate dispersal costs, with consequent feedback effects on the spatial dynamics of adaptation.