Realistic heat pulses protect frogs from disease under simulated rainforest frog thermal regimes

Recent emergences of fungal diseases have caused catastrophic global losses of biodiversity. Temperature is one of the most important factors influencing host–fungus associations but the effects of temperature variability on disease development are rarely examined. The chytrid pathogen Batrachochytrium dendrobatidis (Bd) has had severe effects on populations of hundreds of rainforest-endemic amphibian species but we know little about the effects of rainforest-specific host body temperature cycles on infection patterns. To address this challenge, we used body temperature regimes experienced in nature by frogs in the Australian Wet Tropics to guide a controlled experiment investigating the effects of body temperature fluctuations on infection patterns in a model host (Litoria spenceri), with emphasis on exposing frogs to realistic “heat pulses” that only marginally exceed the thermal optimum of the fungus. We then exposed cultured Bd to an expanded array of heat pulse treatments and measured parameters of population growth to help resolve the role of host immunity in our in vivo results. Infections developed more slowly in frogs exposed to daily 4-hr heat pulses of 26°C or 29°C than in frogs in constant temperature treatments without heat pulses (control). Frogs that experienced heat pulses were also less likely to exceed infection intensities at which morbidity and mortality become likely. Ten of 11 (91%) frogs from the daily 29°C heat pulse treatment even cleared their infections after approximately 9 weeks. Cultured Bd also grew more slowly when exposed to heat pulses than in constant-temperature control treatments, suggesting that mild heat pulses have direct negative effects on Bd growth in nature, but precluding us from determining whether there was a concurrent benefit of heat pulses to host immunity. Our results suggest that even in habitats where average temperatures may be suitable for fungal growth and reproduction, infection risk or the outcome of existing infections may be heavily influenced by short but frequent exposures to temperatures that only slightly exceed the optimum for the fungus. Our findings provide support for management interventions that promote warm microenvironments for hosts, such as small-scale removal of branches overhanging critical habitat or provision of artificial heat sources