Viability of ground beetle populations in fragmented heathlands.

Numbers of ground beetle species that are characteristic for heathlands were negatively associated with area, whereas this relationship was not found for the total number of ground beetle species or for unspecialised ground beetle species. In particular the number of heathland species with low dispersal ability was strongly related to area. Transplant experiments showed that some heathland species with low dispersal ability experienced reduced habitat quality in small habitats, whereas for others at least part of the unoccupied areas were of sufficient quality for successful reproduction. From the presence of occupied as well as unoccupied habitats and from knowledge on its possibilities for dispersal, it is inferred that Pterostichus lepidus lives in metapopulations with continuously and discontinuously occupied patches. Using allozymes, high levels of genetic variation were found in P. lepidus and Agonumericeti from large areas. In two small areas, however, populations of P.lepidus had relatively low levels of genetic variation and were genetically different from the other populations. As exchange of individuals of P.lepidus between the separate populations is possible, the divergent genetic composition of the two populations of P.lepidus probably results from founder events. Populations of A.ericeti from small patches were not genetically divergent from large areas. As exchange of individuals between populations in this species is not possible, it is concluded that dense populations in small areas have remained present since isolation. Genetic erosion could not be identified as an important process causing extinction of ground beetle populations. Environmental stochasticity is the most important threat for the survival of small and isolated ground beetle populations. The presence-absence data of ground beetles in heathlands of various sizes showed that specialized species need a habitat of more than 50 ha to provide a substantial chance to survive an isolation period between 25 and 100 year. For part of the endangered species two options are available to reduce extinction rates: enlargement of habitats and the formation of corridors between habitats