With the methods provided by the theoretical framework of population genetics gene flow (migration) can be measured indirectly. This has mainly been done using Wright’s Fst (standardised measure of the genetic variation among populations) to calculate Nm (number of migrants successfully entering a population per generation). Whether there are any true relationships among population structure, indirect measure of gene flow and dispersal of a species has remained ambiguous (e.g. Templeton et al. 1995, Bohonak 1999). Whitlock and McCauley (1998) have suggested that ‘the measures of genetic structure are valuable in their own right, but that transformations of these measures to quantitative estimates of gene flow or dispersal are at best not needed and, at worst, misleading’. Recently, new methods for estimating gene flow from DNA sequence data using knowledge of the phylogenetic tree of the sequences have begun to be developed (e.g. Hudson et al. 1992).
Because indirect measures of gene flow are dependent on population structure that is a result of at least thousands of generations, gene flow estimates may contain information that is relevant to the history of the species rather than reveal current gene flow (Bohonak 1999). For example, when the neighbouring populations are large, only a small fraction of gene pairs are closely related, and only this fraction gives information about the current rates of gene flow (Barton & Wilson 1995). Gene flow measures from mtDNA sequences that were suggested to be estimated by coalescence methods (Slatkin & Maddison 1989) have been criticised. The main point underlying the criticism is that because the mtDNA lineages are generally not randomised geographically, the mtDNA distributions are primarily generated by historical processes and not by current gene flow (Neigel 1997).
Templeton et al. (1995) presented a nested clade analysis of haplotype trees to solve the problems of separating current and past events. They considered three major biological factors that may cause spatial or temporal associations of haplotype variation. The first, restricted gene flow (mainly due to isolation-by distance) may be difficult to separate from the second, past fragmentation events. The third major factor is range expansion. Recurrent gene flow is indicated, when no association of haplotypes can be detected (i. e., the haplotypes are randomly scattered spatially or temporally; Templeton 1998). The haplotype trees (minimum spanning networks) constructed from the tits can be linked with these major factors. In the network from the willow tit, the haplotypes are scattered, indicating recurrent gene flow, further supported by the observed immigration rates studied in the Oulu population (Orell et al. 1999) and recovery data from the Finnish Bird Ringing Center (Saurola 1981).
The third factor, range expansion, seems to fit the networks from the great and blue tits, except that the southern lineage of the blue tit is a result of either restricted gene flow, past fragmentation events or both. Range expansion northwards is still occuring in the blue and great tits (Haftorn 1957, Veistola et al. 1994, Väisänen et al. 1998), but expansion is more recent in the blue tit. Ecological studies have shown that the northernmost populations are dependent on a continuous immigration from the south to persist. Thus present day gene flow most probably has a great impact in homogenising the populations, but the last ice-age has also had a profound contribution to the present population structure.
However, the coalescence method and nested clade analysis need further development before they can routinely be used to estimate the amount of current gene flow. Therefore, the number of migrants is still frequently estimated from Fst, but the estimates need to be treated with caution (Whitlock & McCauley 1998).