4.9. Population structuring and gene flow

According to Excoffier et al. (1992) the analysis of molecular variance (AMOVA) produces estimates reflecting the correlation of haplotypic diversity at different levels of hierarchial subdivision. Thus, AMOVA was conducted based on subspecies, lineages, and populations with or without the subspecies as a structural variable (VI, Table 5). Based on this, the genetic variation depended rather on these qualifiers than on the individual variation. Pairwise f_ST values between populations (VI, Table 6) showed, that most populations were significantly differentiated from each other. A surprising result was that the Irish population did not differ from geographically distant Finnish (FI1), Greek or Bulgarian populations. Pairwise φ_ST values between subspecies showed that birds from the expected areas of subspecies P. p. sphagnetorum and P. p. perdix did not differ from each other (Table 2).

Population pairwise φ_ST values were also reflected in gene flow estimates (VI, Tables 6, 7). In most cases the gene flow estimates were less than one individual per generation between populations. Estimated gene flow tended to be high for populations, which were not significantly differentiated from each other. The geographical distance and gene flow between populations were significantly and negatively correlated.