Avian divergence and speciation across Beringia examined using comparative mitogenomics

Abstract

Accurate knowledge of divergence and speciation processes is critical for understanding key aspects of biodiversity. As a well-known, speciose group of vertebrates, an increased understanding of how birds diverge and speciate allows us to better manage extant avian diversity and understand how it develops over time. Additionally, birds often exhibit complex and variable patterns of divergence, resulting in complexes of taxonomic uncertainty. Filling gaps in our knowledge of divergence across time and space increases our ability to correctly identify and understand not just avian diversity but clade-level patterns in speciation processes. These higher-order findings give us tools to compare and understand biodiversity more broadly across a wide range of taxa. In this thesis, I investigated both temporal and spatial elements of avian divergence, with an emphasis on the high-latitude system of Beringia, which is of particular interest for speciation due to its position at the meeting point of the Eurasian and American continental avifaunas. Chapter 1 describes my investigation of the temporal dynamics of Beringian divergence. The cyclic opening and closing of the Bering Strait due to glacial cycles intermittently isolated and reunited Asia and North America during the Pleistocene (2.6 Mya to 10 Kya). This was hypothesized to produce an uncertain number of associated 'pulses' of avian divergence events spanning that time period. I used a pairwise sampling approach among 39 taxa and a mitogenomic dataset under Bayesian modeling and found no statistical evidence for multiple vicariance events. Instead, divergence times were spread fairly evenly across a large period of time, appearing as a single vicariance event. This is biologically unusual given the system and the cyclic nature of the most likely abiotic driver (glacial cycles) and may be the result of multiple overlaid periods of divergence and gene flow in taxa with older divergence dates. In Chapter 2, I examine the relative contributions of phenotypic and genetic divergence in pairwise comparisons of diverging bird lineages in high- versus low-latitude systems in Beringia and the Philippines. Phenotypic divergence in birds is assumed to be largely due to selection (Price 2008), with genetic divergence assumed to be more driven by time in isolation. I hypothesize that the Beringian system should have less divergence overall than the Philippines, but that a greater proportion of the divergence should be phenotypic, due in part to increased population connectivity in high-latitude systems as a result of larger long- term range fluctuations as a result of Pleistocene glacial cycles. Increased connectivity should be particularly effective in removing neutral, rather than phenotypic, divergence, where selection may be in operation, in part due to a nonlinear, inverse relationship between gene flow and neutral divergence. To test this, I used standardized measures of phenetic and genetic divergence and used linear regressions to quantify the relationship between divergence metrics and the rates of divergence in each system. Beringia showed lower levels of genetic and phenotypic divergence than the Philippines, but the relationship between data types was stronger and the rate of divergence higher than in the Philippine system. I suggest that this is a result of decreased time spent in allopatry in high-latitude systems, but recognize that an increased rate of phenotypic divergence, possibly due to increased selection pressure at high latitudes, also might play a role.