The effects of human-caused fragmentation of boreal forest on the abundance of red fox Vulpes vulpes L. and pine marten Martes martes L. were studied by combining the Finnish wildlife-triangle snow-track data (1990-94) with land-use and forest resources data employing the GIS. Two study areas (each 45 000 km(2) ) located in northern and southern Finland were selected for the investigation. The extent of landscape that best explained predator abundance (tracks per 10 km 24 h(-1) ) was the same (about 100 km(2) ) in both species and study areas. The decreasing proportion of older forest and the increasing proportions of young forest and agricultural land in the landscape positively affected track density of red fox. The relationship between agricultural land and fox abundance, however, was characterized by a convex curve peaking at 20-30% of agricultural land. With the habitat classification used, landscape composition explained 26% and 11% of the spatial variation in fox abundance in the northern and southern study area, respectively. The relationship between landscape composition and pine marten abundance was not as clear as in that of red fox. Landscape composition explained 10% and 6% of spatial variation in pine marten abundance in the northern and southern study area, respectively. In both areas a positive impact occurred with the increasing proposition of young forest in the landscape, but in the northern area the negative effect of increasing proportion of agricultural land was dominant. The abundances of red fox and pine marten were not negatively correlated, indicating that competition or intraguild predation by red fox do not determine abundance of pine marten on a landscape scale. A general increase in predation pressure by generalist predators in fragmented forest landscapes has been an intensively discussed conservation problem during recent years. We conclude that the red fox is a species potentially able to cause elevated predation pressure in boreal landscapes fragmented by human activities, but that the evidence against the pine marten is weaker.
