Disease and thermal acclimation in a more variable and unpredictable climate

Global climate change is shifting the distribution of infectious diseases of humans and wildlife with potential adverse consequences for disease control(1-4). As well as increasing mean temperatures, climate change is expected to increase climate variability(5,6), making climate less predictable. However, few empirical or theoretical studies have considered the effects of climate variability or predictability on disease, despite it being likely that hosts and parasites will have differential responses to climatic shifts(6,7). Here we present a theoretical framework for how temperature variation and its predictability influence disease risk by affecting host and parasite acclimation responses. Laboratory experiments conducted in 80 independent incubators, and field data on disease-associated frog declines in Latin America(6), support the framework and provide evidence that unpredictable temperature fluctuations, on both monthly and diurnal timescales, decrease frog resistance to the pathogenic chytrid fungus Batrachochytrium dendrobatidis. Furthermore, the pattern of temperature-dependent growth of the fungus on frogs was opposite to the pattern of growth in culture, emphasizing the importance of accounting for the host-parasite interaction when predicting climate-dependent disease dynamics. If similar acclimation responses influence other host-parasite systems, as seems likely, then present models, which generally ignore small-scale temporal variability in climate(7), might provide poor predictions for climate effects on disease.