• Gene tree;
  • genetic isolation;
  • Great Basin Desert;
  • Lycaenidae;
  • phylogeography;
  • Pleistocene refugia;
  • reticulation;
  • species tree;
  • USA ;
  • vicariance



The aim of this study was to investigate patterns of genetic isolation associated with populations in discrete habitat patches in a North American cold desert through analysis of the pallid dotted-blue butterfly, Euphilotes pallescens. This small butterfly is largely restricted to low-elevation habitats across the Great Basin. The apparent geographical isolation and reported morphological variation among E. pallescens populations makes this species an ideal candidate with which to investigate patterns of genetic isolation and diversification in this arid region.


Great Basin, western North America.


We used sequence data from nuclear and mitochondrial genes to investigate genetic diversity among E. pallescens populations, and among E. pallescens and a number of closely related Euphilotes species, using Bayesian phylogenetic analyses, as well as population genetic analyses. In conjunction with genetic variation, morphological variation was examined in the context of geographical and subspecific differentiation.


Our genetic and morphological analyses suggest a moderate amount of isolation among populations, consistent with the hypothesis of restricted gene flow among isolated dune habitats, and possibly associated with isolation in distinct Pleistocene refugia. The patterns of diversification within E. pallescens and among closely related species are complicated by discordance among phylogenetic reconstructions based on nuclear and mitochondrial genes. Discordance among gene genealogies suggests a complex evolutionary history, perhaps involving alternating periods of reticulation and divergence in isolation.

Main conclusions

Although E. pallescens may be a vagile species, we find that persistence on isolated dunes in the Great Basin is associated with appreciable genetic and morphological differentiation among populations. However, genetic, morphological and taxonomic axes of variation are only partially in agreement. More generally, the discordance we find among genetic regions is consistent with the ascendant paradigm in phylogenetic reconstruction: gene genealogies often do not perfectly match species trees. Thus we present Euphilotes as a model for future biogeographical and phylogenetic reconstructions employing larger data sets of independent sequence markers.