Exploring Bayesian models to evaluate control procedures for plant disease

摘要: Tigernut tubers are the main ingredient in production of orxata Valencia, a white soft sweet popular drink. In recent years, appearance black spots skin tigernuts has led to important economic losses because severely diseased must be discarded. this paper, we discuss three complementary statistical models assess disease incidence harvested from selected or treated seeds, and propose measure effectiveness for different treatments against based on probability germination disease. Statistical methods these studies approached Bayesian reasoning include mixed-effects models, Dirichlet-multinomial inferential processes logistic regression models. analyses provide relevant information carry out measures palliate spot achieve high-quality production. For instance, study shows that avoiding affected seeds increases harvesting asymptomatic tubers. It is also revealed best chemical treatment, when prioritizing germination, disinfection with hydrochloric acid while sodium hypochlorite performs better if priority have reduced incidence. The reduction syndrome by agents supports hypothesis causal agent pathogenic organism.