Search

To test two models of how wood mouse Apodemus sylvaticus populations in different patches might interact, we estimated parameters from capture-mark-recapture data in four habitats (set-aside, crop, boundary and woodlot) at two sites in arable farmland. In the source-sink model, populations in 'source' patches have fitness > 1, while in 'sink' patches fitness is < 1; dispersal is constrained so there is a net flow of individuals from sources to sinks. In the balanced dispersal model, patches may vary in quality and carrying capacity, but there are no sinks and no constraints on dispersal; fitness is equal across patches. Our results broadly support balanced dispersal. Patches close together were more likely to exchange individuals than those far apart, there was no evidence of density-dependent dispersal and little evidence for directionality of movement. Overall population growth rates were similar in all patches (range: 0.99-1.05). Times when no animals were known to be alive occurred in some patches and could be attributed to stochastic events. On one crop patch, however, extinction followed high abundance, and coincided with poor recruitment, suggesting sink dynamics. Recruitment rates were highest in boundary and crop, but surprisingly, the proportion of individuals in reproductive condition, timing of breeding and pregnancy rates did not vary between habitats. Apparent survival rates were lowest in crop, although recapture rates were high. While female body size in December did not vary between habitats, males on set-aside were smaller than elsewhere. In general, the balanced dispersal model best described arable wood mouse populations, but dynamics on one patch suggested an 'attractive' sink. Our results demonstrate that more than one model for dispersal dynamics between populations may apply within the same landscape, for the same species.

Wood mice Apodemus sylvaticus are potentially useful indicators of change in arable ecosystems. Here we focus on changes resulting from removal of land from arable production under the set-aside scheme. Wood mice were radio-tracked to compare: (a) their use of set-aside, crop and hedgerow before and after harvest; (b) set-aside configured as margins and as a 3 ha block; (c) cut and uncut 20-m wide set-aside margins. Males had larger home ranges, and were more mobile than females. Ranges were larger, and animals more mobile, before harvest than afterwards. There were no differences in range sizes of breeding and non-breeding animals after harvest, suggesting that changes in habitat use were not a function of cessation of breeding. Before harvest, wood mice used habitats within their ranges at random, and their ranges contained a high proportion of cropped area. After harvest they preferred hedgerow and avoided margin set-aside within their ranges, but did not similarly avoid the set-aside block. The proportion of cropped area within their ranges decreased after harvest, and the proportion of margin increased. Our evidence suggests wood mice avoided using the area adjacent to the hedgerow, perhaps to avoid predators. Uncut set-aside patches were favoured and cut patches avoided, possibly in response to differences in food availability and levels of protection from predators. These results confirm that wood mice are useful indicators of change in arable landscapes.

Set-aside is a potentially important habitat for field voles, ( Microtus agrestis ), populations of which may have declined in lowland Britain. We hypothesized that field vole abundance would increase with the length of time set-aside was left in situ. To test this we monitored development of field vole populations in the first 2 years after set-aside establishment, and on older set-aside we related abundance to set-aside age and vegetative characteristics. Voles were not recorded on set-aside until 9 months after establishment, and populations only began to increase after nearly 2 years. In set-aside fields 2–9 years old, vole abundances were not related to age, but increased with the proportion of grasses and litter in the sward. Management practices such as sowing with a grass seed mix, mowing at least once annually and leaving set-aside in place for >2 years are likely to benefit field voles.

The set-aside scheme, whereby land is removed from arable production, is primarily aimed at manipulating agricultural surplus within the European Community, but has potentially great consequences for biodiversity and wildlife management (Feber et al., 1995; Macdonald et al., 1998; Sotherton 1998). Managed inappropriately, set-aside could prove environmentally damaging by reducing biodiversity and directly harming certain species, such as ground-nesting birds (Thompson, 1993). Conversely, appropriately managed set-aside could foster desirable ecological communities, with enhanced populations and biodiversity.

Despite its conversation potential, little is known of whether land taken out of arable production as set-aside benefits wildlife, nor what features enhance its value as a habitat. Authors studied the presence of small mammals on one-year set-aside through a summer harvest period (278 captures in 3000 trap nights). Trapping success was low on all set-aside (0.6%), but was higher in adjoining hedgerow (30%) and cereal crops (13%). Wood mice Apodemus sylvaticus (Linnaeus, 1758) were the most commonly caught species, and the only species to occur on set-aside. There were no significant differences in trap success between set-aside configured in blocks and strips, nor between sown and naturally regenerated set-aside. Captures on set-aside increased after harvest, when the crop no longer provided cover. Data demonstrate that set-aside is not necessarily a suitable habitat for small mammals.