Assessing the complex relationship between landscape, gene flow, and range expansion of a Mediterranean carnivore

Abstract

Landscape resistance is often disregarded in studies of range expansions and population connectivity. To assess those effects, we simulated the expansion of the Egyptian mongoose (Herpestes ichneumon) in relation to landscape resistance through kernel resistance modeling, confronting it with previously published data regarding the observed pattern of expansion and genetic diversity of the population in Portugal. We modeled population expansion as a function of shrub cover and elevation through iterative simulation of a resistance model and a null model. We then performed an overlap analysis to assess the congruence between the observed pattern of expansion and both resistance and null models across 30 years. We also tested whether there is an effect of allelic surfing or the central-marginal hypothesis by correlating observed allelic richness (1) with the number of simulated years that each location with sampled genotypes had been occupied by the mongoose population and (2) with the cumulative resistant kernel density (which is a measure of population centrality). Results indicated a higher similarity between observed range expansion and the simulation using the null model and a marginally significant correlation between observed allelic richness and number of years of the simulated presence of the species in the null model. The pattern of range expansion in this population is most consistent with a neutral model of uniform resistance, and genetic diversity is most correlated with null model as well. This suggests that range expansion and genetic diversity patterns in expanding populations may not always be predicted by landscape resistance models developed through association of observed genetic differentiation with landscape features.