Abiotic constraints on the activity of tropical lizards

Abstract

Many tropical ectotherms are considered vulnerable to anthropogenic climate change because they have evolved to become thermal specialists. Indeed, several recent studies have suggested that even small increases in mean operative temperature may lead to a reduction in activity and the subsequent extinction of populations. Within the tropics, lizards are considered particularly vulnerable due to the potential for climate change to directly impact physiology and alter community interactions. However, models usually focus on the effects of mean operative temperature at the expense of other climate variables that may also affect lizard physiology. We used daily variation in operative temperature, humidity and wind speed to examine how changes in climate influence activity in two species of lizards from the island of Cayo Menor, Honduras. Anolis lemurinus is a forest species, whereas A. allisoni is an open-habitat species. We conducted daily surveys for active lizards in habitat typical to each species, while simultaneously measuring operative thermal environments with physical models. The effects of the thermal environment were considered in the context of the thermal sensitivity of locomotor performance for each species and compared with the effects of the hydric (humidity) and convective (wind) environments. When all surveys were combined into a single analysis, the activity of the forest species Anolis lemurinus was positively correlated with wind speed, the spatial heterogeneity of operative temperature, and the mismatch between mean operative temperature and the optimal temperature for sprint performance. Mean operative temperature did significantly affect Anolis lemurinus activity, but only when it was above their thermal optimum. Activity of the open-habitat species A. allisoni was negatively correlated with wind speed, but was not related to any other climate variable. Whereas the mismatch between mean operative temperature and the thermal optimum for performance predicted the activity level of the forest species in ways partially consistent with its use in models for the response of lizards to climate change, the effects of the abiotic environment were habitat dependent. Our results suggest that successfully predicting the biological impacts of climate change will require holistic models that account for more than changes in mean temperature alone.