Survey methodology for the detection of Wood Turtles (Glyptemys insculpta)

Abstract

Wood Turtles (Glyptemys insculpta) are difficult to survey because their use of aquatic and terrestrial environments varies spatio-temporally. Existing survey methodology is highly variable and typically involves searching for Wood Turtles within water and on land 0 to >20 m from the shoreline from spring to autumn. The mobility of Wood Turtles suggests that detection is likely influenced by distance surveyed from water and the amount of vegetation, which varies by season. To determine an ideal survey methodology for the Wood Turtle, we recorded distances from a waterway of 31 radio-tagged turtles at Canadian Forces Base, Gagetown, New Brunswick, in 2003 and 2004. Ordinal logistic regression was used to determine the probability of finding male or female Wood Turtles with increasing distance from water at different times of day or season. Sex and time of day were not significant factors in detecting Wood Turtles. Season was a significant factor, with highest probability (69%) of finding Wood Turtles at a distance of 0–10 m of a waterway up to July 1 (corresponding to pre-nesting and nesting periods), compared to probabilities of <10% for any 10-m distance between 10 m and 50 m from a waterway. After July 1, the highest detection probability (50%) was at distances greater than 50 m from a waterway. We recommend that Wood Turtle surveys for environmental impact assessments and population monitoring be conducted on warm days (i.e., 10–25°C) within 10 m of waterways up to July 1.