Climate has a key impact on animal physiology, which in turn can have a profound influence on geographic distributions. Yet, the mechanisms linking climate, physiology and distribution are not fully resolved.
Using an integrative framework, we tested the predictions of the climatic variability hypothesis (CVH), which states that species with broader distributions have broader physiological tolerance than range‐restricted species, in a group of Lampropholis skinks (8 species, 196 individuals) along a latitudinal gradient in eastern Australia. We investigated several physiological aspects including metabolism, water balance, thermal physiology, thermoregulatory behaviour and ecological performance.
Additionally, to test whether organismal information (e.g. behaviour and physiology) can enhance distribution models, hence providing evidence that physiology and climate interact to shape range sizes, we tested whether species distribution models incorporating physiology better predict the range sizes than models using solely climatic layers.
In agreement with the CVH, our results confirm that widespread species can tolerate and perform better at broader temperature ranges than range‐restricted species. We also found differences in field body temperatures, but not thermal preference, between widespread and range‐restricted species. However, metabolism and water balance did not correlate with range size.
Biophysical modelling revealed that the incorporation of physiological and behavioural data improves predictions of Lampropholis distributions compared with models based solely on macroclimatic inputs, but mainly for range‐restricted species.
By integrating several aspects of the physiology and niche modelling of a group of ectothermic animals, our study provides evidence that physiology correlates with species distributions. Physiological responses to climate are central in establishing geographic ranges of skinks, and the incorporation of processes occurring at local scales (e.g. behaviour) can improve species distribution models.