Intraspecific variation in dynamics: small rodents between food and predation in changing landscapes

Population densities, food and predators were examined for bank voles, Clethrionomys glareolus, and field voles, Microtus agrestis, during the period 1971-1998 in one northern (Stromsund) and one south-central (Uppsala) region in Sweden to see if trophic relationships persisted in the presence of large changes in landscape composition, and rodent dynamics, in the northern region. The main work was done on early reforestations. Whereas the limited forest areas at Uppsala were consistently situated on poor soil, forests were extensively clearcut at Stromsund in the 1960-1980s, continuously leaving poorer and fragmented old forests and extensive young thicket forests that were of little use to small rodents. Population densities and amplitudes were larger at Stromsund in the 1970-1980s than in the 1990s and densities at Stromsund reached higher peaks than at Uppsala. The proportion of M. agrestis in the rodent communities was consistently larger at Stromsund. Densities of C. glareolus were higher in forests than on reforestations but there was a continuous exchange of individuals between the two habitats. M. agrestis on abandoned fields, a main habitat for this species, did not show any temporal or regional variation and were more common than on reforestations, except at Stromsund in the 1970-1980s. Consumption of artificial seed supplies did not differ annually at Uppsala but did so at Stromsund in 1994-96; almost all supplies were there consumed at a vole peak year, particularly inside forests. Consumption of an artificial bark supply on reforestations was much larger in the 1980s than in the 1990s at Stromsund; it was consistently low at Uppsala. There was not any difference between regions in such bark consumption on abandoned fields. Vole damage to forestry seedlings was important in the 1960-1970s, of less importance in the 1980s, and did not occur at all in the surveyed years of the 1990s. Mustelid numbers (particularly Mustela erminea) were high at Stromsund after vole peaks in the 1980s but were generally low in the 1990s; they were consistently low at Uppsala. Fox tracks did not vary in any consistent way. Following interpretation emerged: Sheltering snow cover generated cycles. Peak C. glareolus numbers were generally limited by winter food, particularly by pendulous lichens that decreased during the study period. Abundant seed or fruit supply in increase autumns can irregularly cause particularly high C. glareolus peaks. Massive invasion of northern reforestations by C. glareolus from forests appears to have buffered predation on M. agrestis, permitting extensive cycles of the latter species on reforestations in the 1960-1980s. Limited numbers of C, gini glareolus on reforestations permitted only small M. agrestis populations there and generalist predators as foxes or raptors may then have terminated minor increases already in summer. Trophic interactions are intensified and spatially extended at pronounced peak numbers. More generally, regional changes in fluctuation patterns of small rodents can be related to varying use of managed landscapes, where habitat quality changes both in time and space.