Selection Vs Gene Flow

The adaptive traits representing overall wing size and shape followed the same pattern as the neutral markers. Individuals from the isolated region were significantly differentiated from the peripheral and central while the peripheral and central populations showed no signs of differentiation. If predicting that selection pressures are equal in the peripheral and isolated region, low or lacking levels of differentiation in both neutral and selected traits in the peripheral populations in combination with a significant differentiation in both marker types in the isolated populations suggest that gene swamping is causing the observed pattern. This is supported by the fact that the peripheral and isolated regions are found on the same latitude while the central populations are found on an order of magnitude more distant, suggesting that environmental conditions should be more equal in the peripheral regions rather than the peripheral and central ones.

However, I cannot from current data exclude that there are actually different selection pressures in the two peripheral regions. For at least two reasons both species experience a more fragmented landscape in Sweden than in the Baltic states. Firstly, Southern Sweden has a history of great changes in landscape configuration due to an intensive afforestation over the last century (Nilsson 1990), while the

Fig. 2 Expected heterozygosity (┬▒ S.E.) for (a) 10 allozyme loci plotted against longitudinal position for 13 populations of C. hero and (b) 18 allozyme loci plotted against latidudinal position for 16 populations of C. arcania. Populations originate from three parts of the species' distributions; one core (Ural for C. hero and central Europe for C. arcania, population means indicated as straight line), one peripheral (Estonia for both, population means indicated as stars) and one isolated (Sweden for both, population means indicated as rombs)

Table 2 Fixation index (FRT) and their significances among three regions of the two species C. hero and C. arcania within their distribution range


C. hero

C. arcania










See text for definition of regions landscape structure has not changed considerably in, for example, Estonia during the same time period (Palang et al. 1998). Secondly, due to a significant decline in low intensive agriculture in Sweden over the past century (Dahlstr├Âm et al. 2006), the proportion of semi-natural grasslands (primary habitats of Coenonympha species (Eliasson et al. 2005) is definitely lower in Sweden compared to the Baltic states. Both these processes have lead to lower amounts of available habitat for the Swedish butterflies. As a consequence, both Coenonympha species may occur to a larger extent as locally isolated populations in Sweden, whereas the distribution approaches a continuous one in the peripheral Baltic (states) region.

Was this article helpful?

0 0

Post a comment