The Influence of Weather, Time of Season, and Time of Day on Bullsnake (Pituophis atenifer sayi) Thermoregulation and Habitat Selection

Abstract

The grasslands of Western Canada are exposed to one of the most variable terrestrial climates on Earth. Ectothermic reptiles such as the bullsnake (Pituophis catenifer sayi) must respond to this variation by altering their behaviour over daily and seasonal scales. In Chapter 2, I addressed questions regarding the extent to which bullsnakes thermoregulate. I examined how body temperatures (Tb) varied with available operative temperatures (Te) over an active season at a site in southwestern Saskatchewan, Canada. Bare ground Te was variable (x̄ min = 9.3, x̄ max = 31.4°C) and was outside voluntary limits (15 – 35°C) 45 % of the time. However, burrows created by mammals provided stable thermal conditions that remained within voluntary ranges for 93 % of the active season. Bullsnakes should not need to move far to thermoregulate because of thermal gradients (up to 27.2°C in less than 1 m) formed between burrows and patches of sun-exposed bare ground. Time of day, day of year, and bare ground Te (R2 = 0.578) were useful predictors of Tb (x̄ min = 20.1, x̄ max = 29.1°C). Bullsnakes thermoregulated by conforming to Te near their preferred range (21 to 27°C), selecting locations that enabled heating up below this range, and avoiding warmth above this range. My results support the hypothesis that predicts reptiles will thermoregulate more when environmental temperatures deviate from preferred ranges, and when energetic costs of movement are lower due to high thermal heterogeneity. Despite the extreme surface temperatures of high-latitude grasslands, thermoregulation may be easier for snakes than in high-latitude forests due to the thermal gradients available because of burrows in areas exposed to sun. In Chapter 3, I examined the influence of time (time of day, day of year) and weather (temperature, humidity, and precipitation) on a suite of behaviours in 4 bullsnake populations in Saskatchewan. I studied variation in 4th order use and selection for burrows, concealment probability, relocation probability, and movement distances. Bullsnakes altered behaviours in response to time and weather. The greatest variation was explained for concealment (R2 = 0.244) and 4th order habitat selection (R2 = 0.144 – 0.169). However, 4th order habitat use (R2 = 0.0366), relocation (R2 = 0.0419), and movements (R2 = 0.025 – 0.071) were relatively unpredictable based on time and weather variables. Movement distances were particularly variable among individuals. Overall bullsnake activity was highest in spring and during afternoons. Poor weather and unfavourable temperatures were avoided. Humidity was more important than expected. Humidity (or precipitation) was negatively associated with activity in spring or at cool temperatures, but positively associated with activity in summer or at hot temperatures. I suggest that current management recommendations for bullsnakes are valid. However, sample biases could create the false appearance of different habitat requirements among populations if weather conditions, time of day, or season were to differ among studies. Time and weather were useful for predicting the probability of finding bullsnakes on the surface (AUC = 0.8015) and should be used to forecast optimal conditions for surveys. Snake behaviour is often considered to be unpredictable, but the results of Chapter 2 and 3 suggest that weather variables and time are important for predicting bullsnake behaviour, and particularly general surface activity. Behaviours varied widely over the active season, and the effects of time and weather variables were context dependent. My thesis highlights the importance of long study periods under a wide range of conditions to understand the ecology of reptiles in highly seasonal environments.