Phylogeography of Diptera in northern North American glacial refugia

Background/Question/Methods

The Pleistocene glaciations are a significant abiotic driver of current distributions of North American species and populations; contemporary species distributions are the result of dispersal from glacial refugia beyond ice margins. While most species survived glaciations south of the ice sheets, many also survived in the Beringian refugium in Alaska and the Yukon. Numerous reconstructions show additional unglaciated areas in the Canadian Arctic Archipelago, such as Banks Island, however, phylogeographic evidence is divided on whether it was a refugium. In Canada, Diptera (flies) are the most species-rich and diverse group of animals, but studies on the impacts of glaciation on their distribution are lacking. Our objective was to establish whether molecular patterns of diversity in flies support a Banks Island refugium. We sampled flies at five arctic sites: Kugluktuk, NU, Dempster Highway, YT, Iqaluit, NU, Bank Island, NT and Cambridge Bay, NU.  We sequenced two mtDNA gene regions (cytochrome c oxidase subunit I and cytochrome b) from two Diptera species: Zaphne barbiventris (Anthomyiidae) and Norrbomia fumipennis(Sphaeroceridae). Haplotype and nucleotide diversity were calculated. Haplotype networks were constructed using statistical parsimony to view relationships among clades. Molecular dating was performed to establish divergence time of clades.

Results/Conclusions

Analyses were based on 30 sequences of each gene for N. fumipennis and 25 for Z. barbiventris. Both species had high nucleotide diversity in the Yukon, however only Z. barbiventris also had high haplotype diversity there. Neither species had high values of haplotype or nucleotide diversity in Banks Island. Haplotype networks for both species had mostly separate Yukon haplotypes and clusters, but few separate Banks Island haplotypes. However, in the haplotype network of Z. barbiventris Banks Island haplotypes were shared with Cambridge Bay, on adjacent Victoria Island. Molecular dating analysis suggested that Yukon haplotypes of both species likely diverged from others early during the Pleistocene glaciation. However, the Z.barbiventris analysis also showed that the Arctic Archipelago haplotypes could have experienced secondary contact during glaciation. Ultimately, patterns from both species indicate that they survived Pleistocene glaciations in Beringia. Although Banks Island did not have high haplotype diversity or many unique haplotypes, a pattern often seen for refugia, current populations there could have arrived from another Arctic Archipelago refugium. Although our study did not find strong evidence for a Banks Island refugium, it does provide evidence for a possible Arctic Archipelago refugium.