Fig. 1: Mechanistic model of sporogony fits sporozoite prevalence and mean oocyst intensity.

The temporal dynamics of sporogony is first fit to data from mosquitoes exposed to each constant temperature independently (independent model; blue lines). The model is also fit to all the standard membrane feeding assay mosquito dissection data simultaneously by fitting functional forms between temperature and certain model parameters (the independent model parameter estimates were used to guide the choice of functional forms) (pooled model; yellow lines). The actual (points) and predicted (lines) sporozoite prevalence and mean oocyst intensity are shown in (a) and (b) respectively. Uncertainty is shown by the 95% credible intervals of the posterior predictive means for the model estimates and 95% confidence intervals (error bars) for the experimental data, which was conducted with n = 4195 mosquitoes dissected for sporozoites (a) and n = 520 mosquitoes dissected for oocysts (b). For the binomial proportion data, confidence intervals were calculated using the Clopper-Pearson method. The facets indicate the constant temperature at which mosquitoes were maintained.