Résumé
Data of coffee berry disease (CBD) dynamics and of microclimates, collected over two consecutive years (2012-2013) on a smallholding coffee farm in Bamendjou in West Cameroon (5°24′0″N; 10°19′0″E, alt. 1600m), were used to assess the effect of shade and full sun on CBD epidemiological processes. For this purpose we developed a mechanistic SEIR model, and we inferred, within a Bayesian framework, the epidemiological parameters against microclimatic covariates. We computed the Bayesian joint posterior distribution for inference of parameters via a Markov chain Monte Carlo (MCMC) algorithm using JAGS software (Plummer, 2017) and the package MecaStat (Rey et al., 2018).