Bobcat (Lynx rufus) home range size and habitat use in northwest Wisconsin.

INTRODUCTION

Public concern for the status and management of bobcat (Lynx rufus) populations in Wisconsin has generated considerable political and biological debate regarding the efficacy of current management practices. This debate, coupled with a lack of regional information, demonstrate the need to describe demographic and behavioral characteristics of bobcat populations in the Upper Midwest. Extensive geographic variability in bobcat morphology, demographics and behavior (Hall, 1981; McCord and Cordoza, 1982; Gittleman, 1989) preclude extrapolation of population characteristics from other regions to bobcats in Wisconsin.

Seasonal variations in home range size, habitat use and social structure are important because they reflect density of conspecifics, mating strategies, inter- and intraspecific resource partitioning, and density and distribution of prey resources. Past research addressing temporal and spatial aspects of bobcat behavior in the Upper Midwest has been limited to one investigation in North-central Minnesota (Berg, 1979; Fuller et al., 1985), which has served as a basis for statewide population estimates and management strategies. A better understanding of factors affecting bobcat density and distribution is needed to direct sound conservation planning strategies.

We initiated a study in 1991 to estimate and compare annual and seasonal home-range size and habitat use by female and male bobcats in northwestern Wisconsin. We tested hypotheses that spatial patterns of bobcat activity and distribution were independent of sex, season and habitat availability. Surveys of small mammal populations, and existing reports on habitat use by snowshoe hare (Lepus americanus) and white-tailed deer (Odocoileus virginianus) were used to compare potential prey resources within cover-types bobcats selected and to understand seasonal variations in bobcat habitat use.

METHODS

Study area. - The research was conducted in the St. Croix Study Area (SCSA), a 288-[km.sup.2] area in southern Douglas Co., Wisconsin, along the St. Croix National Scenic Riverway. Topography was flat to moderately rolling. Most of the region was underlain with poorly drained ground moraine consisting of stones, silt loams and red clays and was characterized by scattered lakes, swamps and marshes. The southeastern portion of the study area was a pitted outwash plain with light-textured, sandy soils. The study area was dominated by co-niferous-hardwood and pine savanna forests (Curtis 1959) and consisted of three lowland and five upland cover-types: (1) lowland conifer stands dominated by white cedar (Thuja occidentalis), black spruce (Picea mariana), balsam fir (Abies balsamea) and occasionally tamarack (Larix larcinea); (2) lowland deciduous stands of black ash (Fraxinus niger) and alder (Alnus spp.); (3) unforested areas including sedge (Carex spp.) meadows, and occasionally fields; (4) upland conifer stands including plantations of red pine (Pinus resinosa), mature stands of jack pine (Pinus banksiana), and occasionally white pine (Pinus strobus); (5) upland deciduous stands dominated by mature aspen (Populus tremuloides), red maple (Acer rubrum), white birch (Betula papyrifera), and occasionally basswood (Tilia americana) and red oak (Ouercus borealis); (6) mixed upland stands containing a 50:50 mix of deciduous (aspen, maple, birch) and coniferous (jack pine, white pine, balsam fir) forest; (7) young dense stands of regenerating aspen usually [less than] 10 yr old; (8) mixed oak savanna often containing small plantations of red pine, isolated stands of regenerating jack pine and an area burned during a wildfire in 1977.

Field methods. - Bobcats were captured in #1.75 offset and #3 Soft-catch[R] (Woodstream Co., Lititz, Pa.) foot-hold traps during the springs and summers of 1991-1993, and were immobilized with an intramuscular injection of ketamine hydrochloride (17 mg/kg body weight) and acepromazine (1.7 mg/kg body weight) (Anderson, 1987b). Individuals were ear-tagged, weighed, measured and classified as adults or juveniles, based on body weight and tooth eruption (Crowe, 1975; Jackson et al., 1988). Adult bobcats ([greater than]1 yr) were radio-collared with 150-g transmitters (Advanced Telemetry Systems, Isanti, Minn.) and were located 1-12 times/24-h period using a single, vehicle-mounted, 4-element yagi antenna and an electronic compass (Lovallo et al., 1994). Only locations separated by [greater than]4 h and only two locations per 24 h period were used in the analyses to avoid autocorrelation of bobcat location estimates. Coordinates for telemetry stations (n = 204) were determined using 1: 24,000 USGS topographic maps and a global positioning system (Magellan Nav Pro 5000, Magellan Systems Corp., San Dimas, Calif.). Bobcat locations were estimated from 2-4 bearings using the Maximum Likelihood Estimator (Lenth, 1981). Locations were only estimated when all bearings were collected within 30 min to reduce errors due to animal movement. A standard deviation of angular error (2.4 [degrees]) was determined by …