Meeting Abstract
The thermal relations between ectotherms and their environments determine body temperature, which is, in turn, a fundamental driver of physiology, life history, and ecology. Measuring responses of ectotherms to spatial and temporal variation in temperature is therefore a key goal of thermal ecology, particularly in the context of ongoing global climate change. Despite the potentially striking effects of short-term (seconds to minutes) temperature variation on ectotherms, most studies make inferences based on monthly or, at best, daily temperature measurements. Is higher frequency temperature variation noise that can be ignored? Or are we missing critical temperature effects and biasing inferences by discarding this “noise”? To determine the effects of changes in sampling frequency on ectotherm thermal ecology, we deployed operative models of insects of three size classes (2, 6, and 15 mm diameter spheres) on leaf surfaces and sampled model and air temperatures every second. We then subsampled these data and estimated the change in extreme temperatures, in exposure time, and in fitness (integrated from thermal performance curves). We found little change in estimates of extreme temperatures at sampling periods up to 30 min, with sampling periods 1 h or longer resulting in underestimates of extreme temperatures and of performance. The common approach of implicitly assuming a stair-step interpolation between distantly sampled temperatures overestimates exposure time to extreme temperatures. These findings suggest that even daily temperature measurements (which are often considered high frequency) are too coarse for making accurate inferences about the thermal ecology of small ectotherms.
