Temperature is a key driver of mosquito-borne disease because it affects the physiology and life history traits of mosquitoes and the pathogens they transmit, raising concerns about the impacts of anthropogenic climate change on mosquito-borne disease. Thermal biology predicts that transmission of mosquito-borne pathogens should peak at intermediate temperatures and decline at high and low temperatures. However, thermal optima and limits remain unknown for many mosquito-borne pathogens, and empirical evidence for suppressed transmission at high temperature is limited. We built mechanistic, trait-based models for the thermal responses of a suite of seven mosquito-borne viruses transmitted by Aedes and Culex vector species, including West Nile virus (WNV), a globally important pathogen, and Ross River virus (RRV), the most common mosquito-borne pathogen in Australia. We validated the model predictions with human case data of WNV and RRV.
Results/Conclusions
Transmission peaked at moderate temperatures (23.9–26.4°C). Variation in model predictions was driven by both virus and vector identity. Across vector–virus pairs, the lower thermal limits (8.7–19.1°C) for transmission varied more than the peak or upper thermal limits (31.8–37.8°C). The model for RRV accurately predicts that transmission is year-round in the tropics but seasonal in temperate areas. Nationwide, the seasonal peak in human cases lags behind predicted transmission by 2 months. Mean incidence of WNV across US counties peaked at 23.9°C, closely matching model predictions.
Because transmission of mosquito-borne viruses is optimized at intermediate temperatures, warming from climate change will likely lead to shifts in transmission, as temperatures become more suitable in some locations and seasons but less suitable in others. For RRV, climate warming will likely increase transmission in temperate areas (where most Australians live) but decrease transmission in tropical areas where mean temperatures are already near the thermal optimum. For WNV in the US, most counties have mean summer temperatures that are currently below the optimal temperature for transmission, indicating transmission is likely to net increase with warming. These results illustrate the importance of considering nonlinear thermal responses when predicting responses to climate change.