Much is known about how dispersal influences metapopulation persistence and diversity when driven by ecological dynamics alone, but the effect of local adaptation on dynamic metapopulations is largely unknown. Population persistence and species coexistence result from the interplay between local demographic processes (births, deaths, species interactions) and spatiotemporal variation in the environment and dispersal. It is well known that when the environment is spatially heterogeneous but temporally invariant, persistence and diversity are maximized at intermediate dispersal rates. In metapopulations this unimodal relationship arises from a well-known tradeoff; increased dispersal allows for faster recolonization after local extinctions but also synchronizes dynamics, increasing the likelihood that extinct or near-extinct patches are neighbored by others with a similar fate and diminishing the pool of potential re-colonists. A similar outcome has been observed in studies of local adaptation and gene flow. High levels of gene flow can impede local adaptation, reduce local fitness, and in some cases even cause population extinction. In order to investigate these ecological and evolutionary factors together, we use a stochastic metapopulation model that incorporates demographic stochasticity - random variations in the basic demographic processes of birth, death, and migration while concurrently allowing for local adaptation to a temporally fluctuating environment.
Results/Conclusions
Here, metapopulation persistence generally increases with dispersal rate, but under certain parameter combinations we see a monotonic increase and under others the classic unimodal shape emerges. This is because when the selection environment is asynchronous, there can be a mismatch of traits – which reduces the effectiveness of dispersal for maintaining healthy population sizes. Mean immigrant fitness decreases with dispersal rate, as the benefit of moving from a dense population to a less dense population decreases with a high dispersal rate. Persistence increases with parent-offspring trait variation when dispersing across different patch habitats, for with no variation, local adaptation is too high to disperse to a new environment successfully. Some of our results contrast the commonly held view that high dispersal rates impede local adaptation and decrease metapopulation persistence. Importantly, these results would not have been evident had we focused on a purely ecological or purely evolutionary model. It is the focus on the interplay between ecological and evolutionary dynamics that has yielded these novel insights. Given increasing evidence of rapid evolution on the time scale of ecological dynamics, incorporating selection and trait evolution into models and experiments on metapopulation dynamics is an important research priority.