Adaptation and differentiation across island populations of Berthelot’s pipit (Anthus berthelotii)

The aim of this thesis was to examine how neutral and adaptive processes shape patterns of genetic diversity across populations of Berthelot’s pipit (Anthus berthelotii), a passerine bird endemic to the Canary Islands, Selvagens and Madeira archipelagos. To achieve this, I examined variation in pathogen infection, neutral microsatellites and functional major histocompatibility complex (MHC) genes. I first reviewed previous evidence for pathogen-mediated selection at the MHC, and showed that differentiating between specific mechanisms of balancing selection may not be possible, and that many studies that have attempted to do so have not fully considered alternative explanations. In Berthelot’s pipit, I found marked differences in prevalence of avian malaria and pox across populations, and showed that these differences were stable over time, largely because they were determined by biogeographic features. This, combined with an observed effect on host body condition, suggests that populations face differential selection pressures from pathogens. Microsatellite analyses indicated that the pipit colonised northwards across its range, resulting in genetic bottlenecks in the Selvagens and Madeira archipelagos. I then used the pipit system to assess how population genetic analyses were influenced by microsatellite markers with different levels of variability; lower variability loci appear to more accurately reflect population divergence, whereas higher variability loci better reflect past changes in population size. I also found that two commonly used measures of differentiation (GST and Jost’s D) are both strongly affected by marker variability, but in different ways. Finally, I found that just 11-15 MHC variants persisted through the initial colonisation event. However, since the bottleneck, at least 26 functional MHC alleles have been generated in situ across the different populations, all but two by gene conversion. Taken together, my results provide an interesting example of how founder events, mutation, drift and selection can interact to drive differentiation across natural populations