Date of Award


Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Master of Science (MS)


Wildlife Ecology and Wildlife Conservation


Daniel J. Harrison

Second Committee Member

William B. Krohn

Third Committee Member

Walter J. Jakubas


Understanding the ecological factors affecting habitat use by the Canada lynx (Lynx canadensis) and its primary prey, the snowshoe hare (Lepus americanus), could help formulate conservation strategies for this carnivore, which is federally listed as threatened and occurs in only four regions of the U.S.A. I measured vegetation characteristics and snowshoe hare densities in 15 regenerating conifer clearcuts and 21 partially harvested stands in northern Maine during the leaf-off seasons, 2005 and 2006; and the leaf-on season, 2005. Regenerating clearcut stands had been harvested between 1974 and 1985 and were subsequently treated with an aerial application of herbicide between 1982 and 1997. Partially harvested stands were last harvested between 1985 and 2004 and included selection harvests, shelterwood harvests, and overstory removal harvests. Vegetation characteristics varied widely across partially harvested stands. This variance can be described by two principal components associated with the conifer composition and understory density within these stands. Snowshoe hare densities also varied widely in partially harvested stands: 0.26-1.65 hares/ha for the combined 2005-2006 leaf-off seasons. All 21 partially harvested stands had lower hare densities than the mean hare density for regenerating conifer clearcuts (2.10 hares/ha, SE=0.22) during these two years. I modeled the relationship of individual vegetation variables to hare densities across the 36 stands surveyed using an information theoretic approach. Hare density during the leaf-off season was positively associated with conifer stem densigy and basal area removed was negatively related to the density of logs in the stand. These three variables explained 67% of the variance in observed hare densities; however, conifer stem density was the single variable that was most strongly related to hare densities. I used GIS modeling to evaluate the relationships between lynx occurrence/non-detection and hare density, bobcat occurrence, fisher harvest density, maximum snow depth, and elevation at the geographic range- and the home range-scales in Maine. At the geographic-scale, lynx occurrence was associated with: 1) areas of higher hare density, and 2) absence of bobcats. Within the geographic range of lynx, simulated home ranges centered on lynx occurrences were associated with: 1) higher hare densities, 2) absence of bobcats, and 3) an interaction between hare density and bobcat occurrence, compared to surveyed areas without lynx detections. Only two surveys detected both bobcats and lynx, but these data suggest geographic- and home range-scale allopatry between these two species. At the geographic scale, the area of land in regenerating clearcuts was positively associated with lynx occurrence, likely as a result of the high hare densities supported by regenerating clearcuts. Annual clearcutting in Maine has been decreasing since the early 1990's and this trend may result in less regenerating forest on the landscape in the future, which might have long-term negative consequences if the objective is to maintain or increase current population levels of Canada lynx in Maine.

Files over 10MB may be slow to open. For best results, right-click and select "save as..."