Date of Award

5-2006

Level of Access Assigned by Author

Campus-Only Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Wildlife Ecology and Wildlife Conservation

Advisor

Daniel J. Harrison

Second Committee Member

William E. Glanz

Third Committee Member

William A. Halteman

Abstract

I examined responses of Canada lynx (Lynx canadensis) and American martens (Martes americana) to habitat features at multiple spatial scales. At the stand-scale, lynx (n = 6) selected tall regenerating clearcuts (4.4-7.3 m, 11-26 years post-harvest) and established partially harvested stands (1 1-21 years post-harvest) and they selected against short regenerating clearcuts (3.4-4.3 m, 11-26 years), recent partially harvested stands (I- 10 years), and mature stands (>40 years). The highest fractal dimension of foraging paths was in tall regenerating clearcuts and established partially harvested stands, which were stands that provided intermediate to high snowshoe hare (Lepus nmericanus) density, intermediate cover for hares, and intermediate levels of canopy closure and live-tree basal area. Movement paths of lynx had increasing fractal dimension at smaller scales and were straighter than they were at broader spatial scales, suggesting that lynx were trying to avoid switching from highly preferred to lesser preferred stands. Lynx maximized time in areas with high-intermediate hare densities to facilitate increased capture success of snowshoe hares, supporting the hypothesis that lynx select for prey access over prey density. I developed models incorporating landscape composition and configuration to predict occurrence of home ranges (n = 63) for American martens in Newfoundland and to evaluate the relative influences of habitat loss versus fragmentation on this endangered subspecies. Habitat loss was the primary determinant of occupancy of landscapes by martens. The probability of occupancy declined precipitously as the percent of suitable habitat fell below 60% of home-range sized landscapes; therefore, efforts to recover marten populations should focus on maintaining suitable habitat above 60%. I also compared threshold responses in occupancy of martens to the amount of suitable habitat in the landscape between two geographically isolated subspecies (Martes americana americuna in Maine and Martes americana atrata in Newsoundland) that differed greatly in the amount of landscape-scale fragmentation and amount of suitable habitat. Drastic declines in occupancy occurred much sooner in Maine (70-80% suitable habitat) than in Newfoundland (30-40% suitable habitat), indicating that martens in Maine are more sensitive to landscape change, and the Newfoundland subspecies has evolved to be less responsive in its more inherently fragmented environment.

Share