Detecting spatial dynamics of range expansions with geo-referenced genome-wide SNP data and the geographic spectrum of shared alleles

Abstract

Uncovering the spatial dynamics of range expansions is a major goal in studies of historical demographic inference, with applications ranging from understanding the evolutionary origins of domesticated crops, epidemiology, invasive species, and understanding specieslevel responses to climate change. Following the surge in advances that make explicit use of the spatial distribution of genetic data from georeferenced SNP variants, we present a novel summary statistic vector, the geographic spectrum of shared alleles (GSSA). Using simulations of twodimensional serial expansion, we find that the information from the GSSA, summarized with Harpendings Raggedness Index (RI), can accurately detect the spatial origins of a range expansion under serial founder models, even with sparse sampling of only ten individuals. When applying to SNP data from two species of the holarctic butterfly genus Lycaeides, the suggested origins of expansion are consistent with hindcasts obtained from ecological niche models (ENMs). These results demonstrate the GSSA to be a useful exploratory tool for generating hypotheses of range expansion with genomewide SNP data. Our simulation experiments suggest high performance even with sampling found in studies of nonmodel organisms (one sampled individual per location, no outgroup information, and only 5,000 SNP loci).