Past hybridization between two East Asian long-tailed tits (Aegithalos bonvaloti and A. fuliginosus)

Incomplete lineage sorting and hybridization are two major nonexclusive causes of haplotype sharing between species. Distinguishing between these two processes is notoriously difficult as they can generate similar genetic signatures. Previous studies revealed that the mitochondrial DNA (mtDNA) differentiation between two East Asian long-tailed tits (Aegithalos bonvaloti and A. fuliginosus) was extremely low, even lower than intraspecific differentiation in some other long-tailed tits. Using a combination of multilocus and coalescent analyses, we explored the causes of the anomalous lack of mtDNA differentiation between the two species.

The mtDNA divergence between the two species was shallow, while the nuclear DNA (nuDNA) divergence was considerably deeper. The IMa analyses based on the mtDNA dataset suggested relatively high gene flow from A. fuliginosus to A. bonvaloti, while negligible gene flow in the opposite direction. In contrast to mtDNA, the migration rates at autosomal and Z-linked nuDNA loci were negligible or much lower. The NEWHYBRIDS analysis assigned all individuals except one to pure parental species with high posterior probability. The Bayesian skyline plot showed that both species underwent population expansions during the Last Glacial Maximum (LGM), and the ecological niche modelling suggested that their ranges overlapped more during the LGM than at present.

We suggest that historical hybridization, in combination with selective sweep and/or genetic drift might be the main causes of the extremely low mtDNA differentiation between the two species. The hybridization probably occurred mainly between A. fuliginosus females and A. bonvaloti males. The LGM distribution expansion might have facilitated hybridization, while the post-LGM distribution contraction could have facilitated some mtDNA sorting. Ongoing hybridization between the two species might be very limited, but further studies with more samples from the contact zone are needed to test this conclusion.