Comparison to other invasive starling populations
By comparing the three sites of invasive starling populations: Australia, South Africa and North America, we are able to begin to define the differences that have occurred in these three individual introductions. The F ST values show that Australian and South African populations are most differentiated from one another, Australian and North American populations are moderately differentiated, while South African and North American populations are most similar (Table S2). This is likely due to differences in timing, random sampling from the native range and number of introduction attempts across these three localities, with the Australian introduction occurring earlier than the others and characterized by multiple attempts.
As expected, the invasive populations had lower diversity than the native range, likely caused by genetic bottlenecks at introduction. The highest haplotype diversity and richness is found in the UK (0.954, 30.0), although only 45 individuals were sampled, we found 26 haplotypes. Of the three introduced ranges, the haplotype diversity and richness values are highest in North America (0.884, 14.6) and lowest for the Australian population (0.713, 7.7). The North American population has 17 unique haplotypes, the South African population has 14 unique haplotypes, and the Australian population has 10 unique haplotypes. The North American population had the highest number of novel haplotypes despite having a smaller founding size than Australia and the smallest sample size in this study (Table 2).
Opposite to the patterns we observe with haplotype diversity, nucleotide diversity (π) in the three invasive populations is highest in the Australian population and the lowest in the North American population. This could be caused by several factors. Firstly, the Australian population is the result of multiple introductions that began earlier than the other introductions (Jenkins, 1977). Multiple introductions could have brought haplotypes which are differentiated from one another by more nucleotide differences than haplotypes that evolved recently from a single small founding population. Secondly, there is evidence of selection on the mtDNA of the Australian population, which would have driven certain haplotypes to higher frequencies and eliminated others (Rollins et al. 2016). If the Australian population was originally propogated with more nucleotide diversity than the others, selection could explain the pattern of lower haplotype diversity evident today. Lastly, the sample size for the Australian data was 549, for South Africa 219 and for North America only 95 (Table 2). Therefore, although the nucleotide diversity indicates that Australian population has the most genetic diversity, it could be the result of the disparity in sample sizes between the three data sets here.