Temperature is a strong driver of the physiology of ectotherms, including the mosquitoes that transmit infectious disease. Mathematical models are often used to predict how temperature affects the transmission of mosquito-borne diseases. These models are mechanistic in that the thermal response of transmission emerges from the responses of organism-level traits. However, the thermal responses of traits are typically phenomenological. In contrast, the metabolic theory of ecology (MTE) is based on physical, chemical, and biological principles that predict how metabolism and related physiological traits scale with body temperature and body size. We use a combination of modeling and empirical analyses to synthesize these two approaches and bodies of research. In particular, we show how MTE can be used to make informed estimates of how physiology changes with temperature even in cases where direct data are sparse and phenomenological curves cannot be fitted. To do this, we built a mathematical model for transmission of mosquito-borne disease (indexed as R0) based on principles of MTE.
Results/Conclusions
The general result is similar to prior results from trait-based models parameterized with phenomenologically-fit thermal responses, but provides different mechanistic interpretations for the lower and upper thermal limits on disease transmission. Additionally, our model helps explain existing variation in the skew of R0 among results from these previous models. Finally, we tested several predictions from MTE for how specific traits should respond to temperature, using a large dataset of temperature-dependent trait data that was assembled for parameterizing the previous trait-based models. While most traits respond to temperature predictability in accordance with MTE, the thermal response of vector competence varies widely across mosquito and pathogen species in a way that is currently unexplained by MTE.