A test of the vicariance hypothesis of western North American freshwater biogeography

Aim The biogeography of western North American freshwater molluscs has traditionally been attributed to vicariance associated with late Tertiary rearrangement of landscape based on distributional evidence and the putatively limited dispersal ability of these organisms. We examined the phylogeography of a widely ranging western springsnail (Pyrgulopsis wongiHershler, 1989) to test this hypothesis and evaluate the relative importance of vicariance and dispersal in structuring the distribution of this species. Location Southwestern Great Basin (California and Nevada), United States of America. Method Two mitochondrial genes (COI, NDI) were sequenced for 28 populations ofP. wongispanning its entire geographic range, which consists of 10 topographically closed drainage basins. We also sequenced eight closely related congeners, as well as the type species of the closely related eastern North American genusFloridobiaHershler & Thompson, 2002, which was used as the outgroup. Phylogenies based on the combined data set were obtained using several methods, and networks for each gene were generated as an additional means of examining relationships among haplotypes. Partitioning of haplotype variation was studied usingamova, migration between populations was estimated using a coalescent‐based method (mdiv), and divergence times were inferred using a locally calibrated molecular clock andmdiv. Results Pyrgulopsis wongiis subdivided into narrowly localized and widely distributed lineages that diverged in the Pleistocene, well after the inception of the contemporary regional landscape. While large ΦSTvalues and the localized geographic distributions of most haplotypes imply absence or negligible contemporary dispersal of this spring‐dwelling snail, the pattern of phylogeographic structuring, presence of a few widespread haplotypes, and results of themdivanalyses suggest geologically recent dispersal across drainage divides. Main conclusions Phylogeography ofP. wongiconflicts with the traditional vicariance model as it is not structured by the contemporary landscape and is instead indicative of geologically recent dispersal. In the absence of evidence that dispersal of this species occurred through surface water connections during the relevant (Quaternary) time frame, we conjecture that spread may have instead been mediated by transport on waterfowl or via upland stream capture. The non‐concordance between phylogeography and landscape reported in this and other recent studies ofPyrgulopsissuggests that members of this diverse and imperiled genus should not be managed using ana priori, ‘watershed as conservation unit’ approach.