Theoretical and empirical impacts of short-term temperature fluctuations on endemic prevalence of an environmentally-transmitted infectious disease

Background/Question/Methods

Climate change has profound effects on infectious disease dynamics, yet the impacts of increased short-term temperature fluctuations on disease spread remain poorly understood. Using a Daphnia-microparasite experimental system, we tested the theoretical prediction that temperature variability should suppress endemic infection prevalence near the pathogen's thermal optimum. This prediction arises from two distinct theoretical methods. The first uses compartmental disease transmission models parameterized with metabolic scaling theory. This method is highly predictive of within-host and among-host disease dynamics responding to changes in average temperature and slowly increasing temperature but has not been tested when there are short-term temperature fluctuations. The second approach uses Jensen's inequality, which suggests that the impacts of variability may be understood by averaging a nonlinear thermal response curve over a stochastic thermal distribution. These frameworks were empirically tested in replicated experimental epidemics in Daphnia populations held in constant or variable temperature treatments. Populations in the variable treatment each received a unique sequence of daily temperature changes according to an autocorrelated temperature model. All populations were initially susceptible, and the disease was introduced over the course of the experiment at a low rate. Prevalence and infection burden were tracked every third day over 228 days of the epidemic.

Results/Conclusions

While theory predicts that temperature variability should suppress endemic infection prevalence near the pathogen's thermal optimum, experimental evidence clearly indicates that variability increased, rather than decreased, endemic infection prevalence. The data indicate that a decrease in the among-host variance of infection burden underlies the positive effect of temperature fluctuations on endemic prevalence. This result demonstrates that while combining compartmental disease transmission models with metabolic scaling theory is highly effective for predicting temperature effects under constant or slowly warming environmental conditions, it is insufficient for capturing the effects of short-term thermal fluctuations: one or more key mechanisms are missing from this framework, which likely involve scaling from within-host to among-host dynamics. Furthermore, our results show that using Jensen's inequality to analyze the impacts of climate variability on ecological dynamics can be flawed in its application to species interactions, not just in magnitude of effect but also in the direction of the effect. Our experimental results suggest that a multi-scale modelling approach is needed to reveal the impacts of variability on ecological dynamics. Such work is of clear importance for the development of predictive frameworks that assess the impacts of future climate change conditions on ecological and epidemiological systems.