Estimación de curvas de progreso de la incidencia de podredumbre blanca (Sclerotium cepivorum Berk.) en cultivos de ajo mediante un modelo no lineal mixto

this paper models the progress curve of white rot in garlic crops using nonlinear mixed models taking into account the effect of concomitant factors in the development of epidemics. between 2001 and 2003 in cruz del eje and jesus maria, argentina, the initial density of sclerotia (die) and the incidence of the disease were evaluated every two weeks until harvest. with die high (>15 sclerotia/100 g of soil) the final incidence was high (64-100%), while with die low (≤15 g soil sclerotia/100) the final incidence ranged from 0-100%. the mixed logistic model that was selected had random “intercepts” and rates which were different for each combination of “environment” (location and year) and die category (high and low). the representation of the epidemic curves was made through three site-specific curves, the typical, with zero random effect, which expresses the general form of the model, and the percentiles q1 (0.25) and q3 (0.75), expressing the variability. fifty percent of the curves had rates of disease progress between r±0.67 v. the variability of the “intercepts” and slopes depended only on the die, and was lower in the “intercept” of curves with die “high” that those with die “low”. the opposite effect was observed on the slopes.