Offset between GPS collar-recorded temperature in moose and ambient weather station data

Abstract

GPS collar-recorded temperature is often considered as a proxy for the ambient temperature in wildlife ecology studies, yet few studies actually test its reliability as well as the correlation with ambient temperature. Here, we address this question and demonstrate a strong correlation between collar temperature and weather station data, indicating that GPS collar sensor data can be regarded as a reliable index of ambient air temperature. Using data obtained from 384 free-ranging moose equipped with GPS collars between latitude 57° N to 68° N in Sweden, we analyzed 1,467,361 paired observations of collar temperature and air temperature of the nearest official Swedish Meteorological and Hydrological Institute station. We found a systematic offset that varied across months, being larger during the warm summer months than during the winter period. We found an average correlation of .91 (r s; range .75 to .93, median .91) between collar and ambient temperature of the nearest weather station. Thus, temperature sensors in, e.g., a GPS collar, may be used to study animal behavior, movement and habitat choice in relation to ambient air temperature. This aligns with the calls for using animals as not only subjects but also as the samplers of the environment. It also opens up possibilities for large-scale projects on animal ecology and physiology in the absence of ground measuring stations on higher spatial scales like home range and landscape. As an application of collar temperature, we show that changes in the movement patterns seem to be highly influenced by temperature-induced heat stress that moose experience during summer.