Vector-borne diseases constitute a significant source of the global disease burden. Climate warming likely to alter global patterns of disease burden, as the large majority of disease vectors are ectotherms whose life history traits and contact with hosts depend directly on the environmental temperature. Hence, it is imperative that we be able to accurately predict how climate warming affects vector influence on disease transmission and prevalence. The prevailing approach to predict warming-induced disease spread is to use a formula for the basic reproductive rate (R0) that is derived under the assumptions of constant population sizes and thermal environments. Given the importance of making accurate predictions that can inform policy decisions, determining whether the formula-based predictions are robust to the non-linear interactions between vector species’ thermal reaction norms and non-stationary thermal regimes such as warming is an important research priority. Here we present a trait-based framework that incorporates mechanistic descriptions of vector trait responses to temperature into a dynamical model of vector-host interactions. This framework allows us to calculate both R0 and disease prevalence from the same dynamical model, and to determine whether the R0 formula derived under stationarity assumptions accurately predicts disease prevalence under typical seasonal variation and warming.
Results/Conclusions
We find that the R0 formula derived under constant population and constant environment assumptions under-predicts the temperature range over which a disease can spread when compared with the R0 derived from the dynamical model using the next generation matrix method. The degree of underprediction becomes more severe when we compare the temperature range for disease spread predicted by the R0 formula with the temperature range of disease prevalence calculated from the trait-based dynamical model. This mismatch raises concerns about relying solely on Ro to make predictions about disease spread in non-stationary environments that can induce complex and unpredictable fluctuations in vector populations. Climate warming is likely to facilitate vector range expansion by making regions previously beyond the vectors’ tolerance range suitable for colonization. This is of particular concern in the temperate zone where rising temperatures are already facilitating the spread of vector-borne diseases with tropical/subtropical origins. The attendant public health issues make it imperative that we make reliable predictions about disease spread. Doing this requires developing dynamical models that incorporate mechanistic descriptions of trait response to temperature and yield comparative predictions based on both R0 and disease prevalence.