Contrasting dispersal inference methods for the greater white-toothed shrew

A species' dispersal capability is difficult to quantify but important for a general understanding of a species' ecology and for applied conservation and management efforts. One approach is to use the information from individual genotypes to estimate recent dispersal rates. These genetic methods differ in the way they use the genotype data, their assumptions, and the information they give, but choosing one method over another is complicated by the lack of work that compares these methods on simulated or real data sets. We collected detailed, spatially resolved, individual data on the greater white-toothed shrew (Crocidura russula) in western Switzerland for which past studies have found an unusual female sex-biased dispersal. We analyzed the movement from 1 cohort of juvenile shrews with 7 published methods (i.e., mark-recapture, parentage analysis; genetic assignment; hierarchical F-statistics; and the programs BayesAss, IMIG, and STRUCTURE) and used a binomial test to make quantitative comparisons between the results of the methods. Our study indicates that the methods are broadly consistent, but parentage analysis appears the most powerful method for analyzing fine-scale dispersal patterns. In a conservation context, where the evaluation of long-term translocation success is critical for species management, the species studied and spatial scale considered will dictate which is the best suited method to estimate dispersal