AUTHOR BLOG: Speciation with gene flow in Northern Saw-whet Owls

Jack Withrow

Linked paper: Speciation despite gene flow in two owls (Aegolius ssp.): Evidence from 2,517 ultraconserved element loci by K. Winker, T.C. Glenn, J. Withrow, S.G. Sealy, and B.C. Faircloth, The Auk: Ornithological Advances.

Dorsal and ventral views of specimens of the two Saw-whet Owl forms, showing plumage differences. Photos by Jack Withrow.

Scientists have long thought that for two related populations of birds to evolve into separate species, they needed to be completely separated. This usually means the kind of total separation produced by isolation on islands or by features such ice sheets, mountain ranges, or rivers. However, the complex distributions and migratory nature of many birds mean that long-term total separation of bird populations, long the assumption in speciation research, is actually not necessary for speciation to occur. As we’ve tried to better understand how bird populations diverge, other ways of explaining speciation have begun to receive more attention, and it’s becoming increasingly clear that speciation can happen even with low levels of gene flow. Learning how this can happen in migratory taxa has become an important focus in ornithology.

Northern Saw-whet Owls (Aegolius acadicus) occur across much of North America. A small, non-migratory population is isolated on the islands of Haida Gwaii in British Columbia (A. a. brooksi), and the nominate mainland form (A. a. acadicus) passes through these islands during migration but does not stop to breed. So, these two populations overlap during some parts of their annual cycle but not during others. Presently considered a subspecies, brooksi feeds heavily on intertidal invertebrates in winter and has strikingly dark plumage compared to the migratory form, acadicus. Although their geographic context suggests the potential for gene flow, no one has ever found any intergrades between the two populations, suggesting that ecological factors are playing a role in maintaining their separation. Reconstructing how these owls diverged is difficult, because we can’t know the exact history of the distributions and ecological contexts that preceded what we observe today in this glaciated region. However, genomic data can be used to test hypotheses about their divergence and tell us what the long-term demographic history has been.

Recent technological advances have vastly increased the amount of genetic data that can be generated to look at divergence in populations. However, it’s important to examine the same parts of a genome so that you are making apples-to-apples comparisons. Ultraconserved elements (UCEs) are sections of the geneome that are relatively easy to find and can be used to assess divergence at population to family and order levels. They provide a similar set of core sequences that increase in variability the further from the UCE you go, allowing the assessment of both deep and shallow divergences at the same place in the genome. We examined over 2,500 separate UCEs to infer the processes under which these data were most likely to have evolved, including population sizes, levels of gene flow, and time of divergence. We found that the model that best fit the data included an initial split between the two populations with a low level of ongoing but skewed gene flow: an average of about 0.7 individuals per generation coming from acadicus into brooksi, and about 4.4 individuals per generation going the opposite direction.

These results support our hypothesis that the divergence between these two forms included low levels of gene flow rather than complete isolation. The low levels of gene flow from nominate acadicus into brooksi, coupled with strong evidence of ecological selection, suggests that the Haida Gwaii owls are on an independent evolutionary trajectory. Nominate acadicus are not staying on Haida Gwaii and breeding with the islands’ owls with the frequency we might expect, and we believe that selection is operating differently on the two populations due to the fact that they focus on resources that are distributed differently in time and space. Despite the history of gene flow between the two populations, brooksi appear to be a young biological species.