The power of different types of markers to detect
population structure
When comparing the relative performance of microsatellites and RADseq
there was the expected discrepancy of higher heterozygosity values in
the microsatellite compared to the RADseq data, which reflect the high
mutational rate of microsatellites. Overall, both markers were able to
uncover genetic structuring. Based on the small but widely
geographically spread, number of sampling sites in the UK, both
microsatellite and SNP data agree that there is a lack of deep
phylogeographic structure within the wild British samples. Considering
the larger distance between sampling sites compared to Greek sampling
sites, for example, more than 500 km between Fife and Essex, this was
unexpected. This result is suggestive of a large, homogenous population
within the UK, but considering the low dispersal of grey partridge, this
points towards large-scale human-mediated translocations. However,
sampling from a much larger number of sites with known histories of
management and partridge releasing are required to confirm this finding
more generally. On the contrary, both molecular markers suggested
similar trends concerning phylogeographic structure in Greek
populations, with SNP data being somewhat more conclusive, although
fewer individuals were included in the analysis. The most distinct was
the population of Drama, followed by the population of Grevena while the
other three populations (Kozani, Thessaloniki and Xanthi) revealed more
mixed patterns and greater similarity with North Macedonia.
To be able to confidently assign individuals to either a Western or
Eastern origin, a marker panel of 15 SNPs has been identified, based on
the ddRAD libraries. Only wild individuals were used for this panel
development, however not a single SNP was at fixation in both
populations, i.e. only occurred in the Eastern individuals, while
completely absent from the Western individuals (see Table S5_Appendix
for a summary of the SNP frequencies). These may reflect either the
recent divergence of the two lineages from the common ancestor or could
be the result of past introduction events that affected the allele
frequency. However, this panel of 15 SNPs could be combined with a
mitochondrial DNA test for the identification of non-native stocks
within Greece or within the UK.