Influence of spatial structure on genetic isolation in Plebejus argus populations (Lepidoptera: Lycaenidae)

摘要

In industrialized countries natural habitats both decrease in abundance and increase in their patchiness. The isolation of habitat patches from each other is a major cause of long term decline of many species. For non-migratory butterflies, the recolonization of previously occupied habitats depends on the proximity to occupied patches (Thomas and Jones 1993). Populations of widely distributed species with limited dispersal abilities seem to be particularly at risk from an increased fragmentation, as these have suffered more from an increased fragmentation than more vagile species (Maes and VanDyck 2001). The severe decline of many butterfly species in northwest Europe is partly due to an increased fragmentation of their habitats (vanSwaay and Warren 1999).In order to assess the impact of habitat fragmentation of populations, one should compare situations where the distribution of a species is naturally patchy with an area where the species has a more continuous habitat. The silver-studded blue Plebejus argus is a Lycaenid butterfly widely distributed in Europe. It may be locally abundant within its distribution range and is usually patchily distributed in Britain, where it has been intensively studied (Thomas 1985, Thomas 1996; Brookes et al. 1997; Lewis et al. 1997; Thomas et al. 1998). British populations show very limited dispersal abilities, resulting in mostly closed population structure. In other parts of Europe, the situation is quite different: in SW Spain, the species habitat is fairly continuous throughout the Do?ana National park, where Halimium halimifolium, its local food plant, is abundant. The density of P. argus is also dependent upon the density of Lasius niger, its obligatory host ant species (Jordano et al. 1992; Seymour and Jordano pers. comm.). In southern Finland, the species is present in numerous islets in the Baltic Sea, which may naturally have very small habitat patches. We hypothesize that the dispersal ability of this species may vary within its European range. On one hand, individuals from the most continuous habitats (i.e. in the Do?ana N.P.) may disperse more than in naturally fragmented habitats, as they would be under less selection pressure against dispersal. On the other hand, in Finland, individuals on islands may be selected against dispersal, resulting in isolated populations. These islands must have been colonised by individuals crossing the sea, hence there is a possibility for an evolution of high dispersal abilities in a few individuals, resulting in a dual strategy of sedentarity and dispersal. This may depend on genotype and ecological and behavioural conditions (Ims and Hjermann 2001). The aim of this study is to investigate dispersal from the point of view of its consequences on the genetic population structure of populations. Basically there can be three main possible spatial structures of population differentiation. (1) If populations show a marked isolation-by-distance effect, dispersal occurs locally in a stepping stone manner, most dispersing individuals moving only between neighbouring habitat patches. (2) If populations are not genetically different from each other, then there is a high rate of dispersal, even between population far away from each other, resulting in a regional panmixia. (3) If populations are genetically different from each other but do not show isolation by distance effect, this may be due to occasional long distance colonisation where genetic differences occur due to founder effect. The aim of the present study is to identify the population structure of Plebejus argus in two contrasted landscape structures: the rather continuous habitat of Do?ana National Park with the naturally patchily distributed populations of forest clearings and islands in south Finland.