Fig. 3: The relationship between temperature and the fitted model mosquito infection parameter values.

a μ (the mean oocyst load among infected mosquitoes), b \({{{{{\delta }}}}}_{{{{{O}}}}}\) (the human-to-mosquito transmission probability, as determined by the presence of oocysts), c \({{{{{\delta }}}}}_{{{{{S}}}}}\) (the probability a mosquito infected with any oocysts will develop any sporozoites at the mosquito scale: conversion), and d \({{{{{\delta }}}}}_{{{{{O}}}}}{{{{{\delta }}}}}_{{{{{S}}}}}\) (vector competence; the human-to-mosquito transmission probability, as determined by the presence of sporozoites). For (a) and (b) black points show the sampled mean oocyst load and oocyst prevalence disaggregated by day of dissection with the 95% confidence intervals respectively, which were estimated from n = 520 dissected mosquitoes. All mosquitoes were dissected at sufficient days post infection that all oocysts were expected to have developed. For all plots blue points show the independent model estimates and yellow lines show the pooled model median posterior estimates with the 95% credible intervals.