Effects of metapopulation processes on measures of genetic diversity

Abstract

Many species persist as a metapopulation under a balance between the local extinction of subpopulations or demes and their recolonization through dispersal from occupied patches. Here we review the growing body of literature dealing with the genetic consequences of such population turnover. We focus our attention principally on theoretical studies of a classical metapopulation with a ‘finite–island’ model of population structure, rather than on ‘continent–island’ models or ‘source–sink’ models. In particular, we concern ourselves with the subset of geographically subdivided population models in which it is assumed that all demes are liable to extinction from time to time and that all demes receive immigrants. Early studies of the genetic effects of population turnover focused on population differentiation, such as measured byFST. A key advantage ofFSTover absolute measures of diversity is its relative independence of the mutation process, so that different genes in the same species may be compared. Another advantage is thatFSTwill usually equilibrate more quickly following perturbations than will absolute levels of diversity. However, becauseFSTis a ratio of between–population differentiation to total diversity, the genetic effects of metapopulation processes may be difficult to interpret in terms ofFSTon its own, so that the analysis of absolute measures of diversity in addition is likely to be informative. While population turnover may either increase or decreaseFST, depending on the mode of colonization, recurrent extinction and recolonization is expected always to reduce levels of both within–population and specieswide diversity (πSand πT, respectively). One corollary of this is that πScannot be used as an unbiased estimate of the scaled mutation rate, θ, as it can, with some assumptions about the migration process, in species whose demes do not fluctuate in size. The reduction of π in response to population turnover reflects shortened mean coalescent times, although the distribution of coalescence times under extinction–colonization equilibrium is not yet known. Finally, we review current understanding of the effect of metapopulation dynamics on the effective population size.