The upper limit of thermoneutrality is not indicative of thermotolerance in bats.

To assess the vulnerability of birds and mammals to climate change recent studies have used the upper critical limit of thermoneutrality (T) as an indicator of thermal tolerance. But, the association between T and thermal tolerance is not straightforward and most studies describe T based solely on a deviation in metabolism from basal levels, without also considering the onset of evaporative cooling. It was argued recently that certain torpor-using bat species who survived prolonged exposure to high ambient temperatures (i.e. high thermal tolerance) experienced during extreme heat events did so by entering torpor and using facultative heterothermy to thermoconform and save on body water. Assuming that T is indicative of thermal tolerance, we expect T in torpor-using species to be higher than that of species which are obligate homeotherms, albeit that this distinction is based on confirmation of torpor use at low temperatures. To test this prediction, we performed a phylogenetically informed comparison of bat species known to use torpor (n = 48) and homeothermic (n = 16) bat species using published thermoregulatory datasets to compare the lower critical limit of thermoneutrality (T) and T in relation to body temperature. The influence of diet, biogeographical region, body mass and basal metabolic rate (BMR) was also considered. Body mass had a positive relationship with BMR, an inverse relationship with T and no relationship with T. Normothermic body temperature scaled positively with BMR, T and T. There was no relationship between diet or region and BMR, but both influenced thermal limits. Torpor-using bats had lower body mass and body temperatures than homeothermic bats, but there was no difference in BMR, T and T between them. Exceptional examples of physiological flexibility were observed in 34 torpor-using species and eight homeothermic species, which included 15 species of bats maintaining BMR-level metabolism at ambient temperatures as high as 40 °C (and corresponding body temperatures ∼39.2 °C). However, we argue that T based on metabolism alone is not an appropriate indicator of thermal tolerance as it disregards differences in the ability of animals to tolerate higher levels of hyperthermia, importance of hydration status and capacity for evaporative cooling. Also, the variability in T based on diet challenges the idea of evolutionary conservatism and warrants further consideration.