Caves, crevices and cooling capacity: Roost microclimate predicts heat tolerance in bats

1. The microsites that animals occupy during the rest phase of their circadian activity cycle influence their physiology and behaviour, but relatively few studies have examined correlations between interspecific variation in thermal physiology and roost microclimate. Among bats, there is some evidence that species exposed to high roost temperatures (T-roost) possess greater heat tolerance and evaporative cooling capacity, but the small number of species for which both thermal physiology and roost microclimate data exist mean that the generality of this pattern remains unclear. 2. Here, we test the hypothesis that bat heat tolerance and evaporative cooling capacity have co-evolved with roost preferences. We predicted that species occupying roosts poorly buffered from high outside environmental temperature exhibit higher heat tolerance and evaporative cooling capacity compared to species inhabiting buffered roosts in which T-roost remains well below outside conditions. 3. We used flow-through respirometry to investigate thermoregulation at air temperatures (T-a) approaching and exceeding normothermic body temperature (T-b) among six species with broadly similar body mass but differing in roost microclimate (hot vs. cool roosts). We combined these data with empirical measurements of T-roost for each study population. 4. Hot-roosting species tolerated T-a similar to 4 degrees C higher than cool-roosting bats before the onset of loss of coordinated locomotion and non-regulated hyperthermia. The evaporative scope (i.e. ratio of maximum evaporative water loss [EWL] to minimum thermoneutral EWL) of hot-roosting species (16.1 +/- 2.4) was substantially higher than that of cool-roosting species (5.9 +/- 2.4). Maximum evaporative cooling capacities (i.e. evaporative heat loss/metabolic heat production) of hot-roosting species were >2, while the corresponding values for cool-roosting species were <= 1. 5. The greater heat tolerance and higher evaporative cooling capacity of hot-roosting species compared with those occupying cooler roosts reveal variation in bat evaporative cooling capacity correlated with roost microclimate, supporting the hypothesis that thermal physiology has co-evolved with roost preference.