Local Distribution and Coexistence of two Species of Peromyscus in Vermont

Abstract

In New England, the woodland deer mouse (Peromyscus maniculatus gracilis) and whitelooted mouse (P. leucopus noveboracensis) are forest-dwelling species that may be found on the same site. Choate (1975) reported that P. maniculatus preferred habitats associated with coniferous forests or northern hardwoods and that P. leucopus was found in slightly more xeric habitats dominated by white pine (Pinus strobus), hemlock (Tsuga canadensis), and oaks (Quercus spp.), with populations occasionally extending into adjacent, more mesic northern hardwood forests. Several other authors also have reported that P. maniculatus is found in moist, cool, or coniferous habitats and that P. leucopus is found in dry, warm, or deciduous sites (Klein, 1960; Smith and Speller, 1970; Wrigley, 1969), although Lovejoy (1970) found P. maniculatus generally using dry habitats in the absence of P. leucopus. Sympatric species coexisting in a stable environment may select different foods, differ in temporal or spatial activity, and/or utilize different habitats, thus presumably reducing competition for niches. Hamilton (1941) noted no differences in food habits of P. mantculatus and P. leucopus. Temporal activity has been found to differ between these species in response to weather parameters (Drickamer and Capone, 1977), but neither seasonal nor circadian differences have been reported in New England. Both species are semi-arboreal and may partition niches spatially, but results of studies of arboreal habits of Peromyscus species do not permit a conclusion for P.m. gracilis and P. 1. noveboracensis (see Tadlock and Klein, 1979; Stahl, 1980; Wolff and Hurlbutt, 1982). Peromyscus maniculatus and P. leucopus appear to occupy similar ecological niches in the Northeast (Klein, 1960; Smith and Speller, 1970). The association of microhabitat structure with the distribution of rodents has received increasing attention in recent years (e.g., Carey et al., 1980; Dueser and Shugart, 1978; Holbrook, 1978; M'Closkey, 1976; M'Closkey and Fieldwick, 1975). Discriminant function analysis (DFA) has been used commonly in small mammal studies to describe microhabitats, demonstrate niche separation, or explain coexistence of sympatric species (Cavallaro et al., 1981). Our study was designed to determine how the distribution of coexisting Peromyscus species is related to microhabitat components of a northern hardwoods forest in southern Vermont. METHODS This study was conducted in a northern hardwoods forest in the Grafton Forest Resources Project area, Grafton, Vermont (45'10'N, in the eastern foothills of the 'Green Mountains. Elevations ranged from 250 to 400 m. Sherman live traps (7.6 by 8.9 by 22.9 cm) were set from 21 May through 20 August 1980 on 88, 0.08-ha plots for a total of 6,600 trap nights. Traps were set 15 through 19 July 1981 on 20 new plots, for 500 trap nights. To sample systematically, each plot contained five live traps, one at plot center and the others equidistant from plot corners and center, along diagonals of the plot (Fig. 1) Plots were at least 100 m apart to reduce the frequency of capturing an individual on adjacent plots. By using many plots widely dispersed throughout the study area, we incorporated much of the variability of forest types and structure. If a standard trap grid (e.g., traps spaced 10-15 m) had been used, the capture of an individual at an adjacent grid location would have been less independent of previous captures. The