Predictors of extinction risk in fragmented populations

Abstract

We used demographic and life-history data from natural populations of 43 species in order to predict the effect of habitat fragmentation on metapopulation dynamics. Metapopulation dynamics was quantified as the probability of extinction, stochastic growth rate and genetic diversity (allelic and gene richness). The main goals of this work were twofold: i) examine the relative impact of parameters defining habitat fragmentation on metapopulation dynamics, and ii) provide rules of thumb for the management of fragmented populations. We considered several models of fragmentation, fragmentation being characterized by different parameters (number of patches, dispersal rates, survival rate of dispersers and environmental variation). Our major finding was that the interaction between dispersal rate and survival rate of dispersing individuals explained the larger part of deviance in both extinction probability and growth rate. Moreover, we found that both an intermediate dispersal rate (of around 1% per generation per population) and an intermediate number of populations minimizes extinction rate and increases the long-term growth rate and genetic diversity for all the fragmentation scenarios examined. We, therefore, advocate that the metapopulation network and the matrix quality should be the major focus in the conservation of populations exposed to fragmentation.