Phylogeographic signal variation in mitochondrial DNA among geographically isolated grassland butterflies

Aim We surveyed mitochondrial DNA (mtDNA) sequence variation among regionally isolated populations of 10 grassland-associated butterfly species to determine: (1) the utility of phylogeographic comparisons among multiple species for assessing recent evolutionary patterns, and (2) the respective roles of isolation attributable to range disjunction versus isolation attributable solely to geographic distance in establishing divergence patterns. Location The Peace River grasslands of northern Alberta and British Columbia, Canada, which are isolated by 300+ km from similar communities to the south. Methods We sequenced mtDNA (1420 bp of cytochrome c oxidase subunit I) from five grassland-restricted butterfly species that have geographically disjunct populations and from five ecologically broader species that have more continuous distributions across the same regions. Using analysis of molecular variance (AMOVA), Mantel and partial Mantel tests, and haplotype networks, we compared population structure within and between species in order to assess the validity of single-species phylogeographic characterizations. We then contrasted variance components between disjunct and continuously distributed species to assess whether divergences were correlated more with disjunction or with geographic distance. Results Single-species analyses varied substantially within both the disjunct and the continuous groups. One species in each of these groups had mtDNA with unusually deep intraspecific mitochondrial lineage divergences. On the whole, however, the five species with disjunct ranges exhibited greater divergence between geographically distant populations than did the five species with continuous distributions. Comparison of variance components between disjunct and continuous species indicated that isolation attributable only to geographic distance was responsible for up to half of the total sequence variation between disjunct populations of grassland butterflies. Main conclusions Our findings show that single-species phylogeographic analyses of post-Pleistocene butterfly distributions are inadequate for characterizing regional biogeographic divergence histories. However, comparison of mtDNA sequence divergences between groups of disjunct and continuously distributed species can allow isolation attributable to range interruption to be quantitatively distinguished from isolation attributable solely to gene flow attenuation over the same geographic area.