Date of Award


Level of Access

Open-Access Thesis

Degree Name

Master of Science (MS)


Wildlife Ecology and Wildlife Conservation


Daniel J. Harrison

Second Committee Member

Angela K. Fuller

Third Committee Member

Robert S. Seymour


Previous studies of Canada lynx (Lynx canadensis) within the northern boreal forest region have documented that lynx respond spatially to a decline in snowshoe hare (Lepus americanus) density, as exhibited by expansion of territories and changes in social structure. I compared home range area and spatial overlap in the southeastern portion of their geographic range during periods of relatively high and relatively low hare density. Home range areas of lynx did not change between periods of high and low hare density, except that home ranges of females during the denning season expanded during the low period. The presence of kittens constrained home range areas of reproductive females during denning because females were attending kittens. Intra- and intersexual overlap did not change as hare density declined, with the exception of a decrease in overlap between females. This decrease was likely caused by decreased reproduction during the low period, which reduced potential for territorial overlap among mothers and daughters. Hare density during the nadir of cycles in more northerly populations can reach levels nearly a magnitude lower than reported for Maine during my study. This may have prevented breakdowns in territories and changes in social structure by lynx, which may have shifted life history strategies towards territorial maintenance and reduced reproduction as hare densities declined.

I also investigated changes in use of high-quality hare habitat (HQHH) at the landscape scale, and habitat selection of HQHH within home ranges of lynx between periods of high and low hare density. Lynx did not change their extent of use of HQHH at the landscape scale, suggesting lynx had adequate amounts of HQHH within their home ranges to encounter hares during both the high and low periods of hare density. Lynx exhibited stand-scale selection for HQHH during both hare density periods, but the intensity of female selection for HQHH declined as hare density declined. This suggests that lynx continued to remain focused on foraging for hares during both periods, but that females may become more generalized in habitat and prey selection during the period of lower hare density.

Lynx monitored during this study wore GPS collars during a period of low hare density and VHF collars during a period of high hare density. This presented methodological challenges when I compared lynx responses between hare density periods. Errors associated with VHF collars were known for this study, but errors associated with GPS collars were not. Failed fix attempts and location inaccuracy caused by environmental and satellite configurations can bias habitat selection and spatial analyses. I evaluated fix success and location error of GPS collars in 7 habitat classes during 2 seasons in northern Maine. I also used an information-theoretic modeling approach to investigate covariates influencing fix success and location error. Canopy cover had the greatest influence on fix success and the configuration of available satellites had the greatest influence on location error. Results were used to compensate for habitat bias and location error caused by GPS collars worn by lynx during a period of low hare density.