A cold and harsh environment: Demography and spatial ecology of a northern population of wood turtles (Glyptemys insculpta)

Natural history studies of geographically distinct populations are important for informing and implementing management strategies for species at risk. Wood turtles (Glyptemys insculpta) are a species of Special Concern (COSEWIC) in Canada and are considered Vulnerable internationally (IUCN). Their species at risk status is attributed to their habitat specificity and thus patchy distribution across the northeastern portion of North America. The purpose of this study was to assess population health, spatial ecology, and habitat requirements of an unstudied wood turtle population in the Sudbury District of Ontario at the species' northern limit while testing for variation in life history strategies (e.g. body size) across the species' latitudinal range. A mark-recapture and radiotelemetry study was conducted from 1 May 2005 to 19 October 2006 to describe home range size, movement patterns, and habitat, temperature, and hibernacula selection. A total of 900 ha of habitat were classified and a 1.5 km of river was quantified for selection studies. I hypothesized that wood turtles at their northern limit would have larger body and home range sizes and would travel greater distances during a shorter annual activity cycle compared to southern conspecifics as a result of a less productive environment in the north. I also hypothesized that optimal hibernacula characteristics (e.g. dissolved oxygen, protection from predators and freezing temperatures) may be limiting, therefore fidelity to hibernacula or communal overwintering may be common. I found that the wood turtle population in the Sudbury District was large (56 +/- 15 turtles) and healthy, with equal sex-ratios, high juvenile capture rates (35%), and that recruitment had occurred over each of the last 11 years. Moreover, latitudinal differences in body size were apparent with larger body sizes in the northern and southern portions of the species' range but relatively small body sizes in the middle of the range and home range sizes increased with latitude. Wood turtles in the Sudbury District had the largest home ranges among populations reported in the literature (59.4 ha (100% MCP)). Wood turtles selected riverine habitats throughout the annual activity cycle; however, early in spring, selection was for ephemeral pools and upland deciduous habitat that may provide a rich food resource and good basking opportunities. Wood turtles basked from 1 May to mid-June but became thermal conformers for the rest of the annual activity cycle; however, females became thermal conformers with air and males became thermal conformers with water. During winter, wood turtles selected environments that were high in dissolved oxygen (∼ 90%), had temperatures that approached 0°C, and had short ice durations compared to other water bodies used by other turtle species at the study site. Moreover, fidelity was not shown across years; therefore, hibernacula are likely not a limiting factor for the species' distribution in the north. These data are important for the creation and implementation of management strategies for the local population of wood turtles but can also be used to assess conservation needs across the species' range. I provide management recommendations for this population in each chapter and as a conclusion to the thesis.