No abstract

No abstract

Northern ectotherms must seek refuge from winter conditions for a large portion of their annual activity cycle. The objective of this study was to quantify physical properties of overwintering sites selected by wood turtles ( Glyptemys insculpta (LaConte, 1830)) at the species’ northern range limit. We mapped all structural features (e.g., root balls and log jams), water depth, and sediment types along a 1.5 km stretch of river that was available to turtles outfitted with radio transmitters (N = 8) during winter. Temperature selection was assessed by comparing thermal profiles from data loggers on turtles and temperature stations within the river and other riparian habitats (e.g., ephemeral pools and oxbows). Dissolved oxygen (DO) was measured at each temperature station and turtle location. Wood turtles overwintered in the river, which was colder (~0 °C), had more stable temperatures, and provided higher DO (12.64 ppm) compared with adjacent habitats. Some turtles selected structured refuges for overwintering. Winter movements were not related to temperature or DO, but may be related to maintaining a certain distance from shore and water depth to protect against accidental relocations during winter. We discuss hibernacula as potential factors limiting the northern distribution of wood turtles, a species at risk in Canada.

Life-history traits such as age at maturity, body size and clutch size tend to vary across a species' distribution. The purpose of our study was to describe the demography of a newly discovered population of North American wood turtles Glyptemys insculpta at the species' northern range limit, and to compare our findings to those of other studies to test hypotheses about adaptive life-history variation. Turtles were hand-captured from May to October 2005 and 2006 along a 4.5 km stretch of river located in the Sudbury District, ON, Canada (461N). Fifty-five captured individuals provided a population density estimate of 1.3 turtles/100 m of river. Juveniles comprised 35% of wood turtles captured, and growth ring counts (i.e. age estimates) indicated recruitment in each of the past 11 years. Among populations, we found a nonlinear pattern in body size variation with the largest turtles in the north, smallest turtles in the centre of the range, and intermediate-sized turtles in the south. This nonlinear pattern in body size was reflected in clutch size variation. Selective pressures to overcome years of low recruitment may have resulted in larger body sizes and hence large clutch sizes at northern latitudes while conspecifics at southern latitudes can achieve larger body sizes because they live in a more productive environment. Population density decreased with latitude, likely as a result of a gradient in habitat productivity.