Forty-eight wood mice Apodemus sylvaticus were radio-tracked over an estimated 1500 h yielding 9000 fixes. The mice were living in three contiguous arable fields, and showed that they respond to weedy microhabitat patches within superficially homogeneous crops. Plant species' abundance and the composition of the seed bank were assessed in quadrats taken in areas in which (a) mice foraged, (b) through which mice travelled, or (c) at random within each home range. Wood mice avoided foraging in areas with a high abundance of bare earth, and selected areas with a high abundance of Alopecurus myosuroides, Stellaria media, Avena fatua, Galium aparine or Bromus sterilis. The food plants apparently selected by wood mice differed markedly between months and between sexes. This may relate to the energetic and nutritional demands incurred by the contrasting reproductive tactics of male and female wood mice. There were no differences between seed bank samples taken from quadrats where the mice foraged, travelled through, or which were selected at random. Our study, the first large-scale analysis of wood mouse microhabitat preference using radio-tracking, reveals that whereas the mice do not react to heterogeneity on the scale of the mosaic of crop fields, they are greatly influenced by the dispersion of food plants within the superficially homogeneous crop itself.

Wood mice Apodemus sylvaticus were radio-tracked in an area covering three cereal fields, which was notionally divided into 5×5 m squares; each of these squares was classified to one of four habitat types (hedgerow, wheat, barley and oil-seed rape). From a sample of 79 radio-tracked wood mice, yielding 8500 fixes, we defined home-range boundaries and estimated for two seasons: (a) the extent to which each habitat was present in each individual's home range relative to its overall availability in the surrounding landscape; (b) the extent to which each habitat was used by the mice relative to its abundance within a home range. Hedgerow ranked highest in preference (as distinct from use) for all comparisons made in both winter and summer. In winter, home ranges contained significantly more hedgerow than barley and wheat, and significantly more rape than wheat. Animals also used the habitats within their home ranges non-randomly, with a significant preference for hedgerow. In summer, home ranges contained significantly less rape than other habitats. Habitats within home ranges were used at random. Seasonal patterns in habitat use appeared to be largely a response to seasonal disturbance and the availability of cover in the fields.