A time series analysis of brown eye spot progress in conventional and organic coffee production systems

Brown eye spot (BES), caused by Cercospora coffeicola, is an important coffee disease in Brazil. Losses related to this disease have increased over the last few years. Because the factors associated with the disease dynamics are not fully understood, it is important to gather information about this in different cropping systems. BES epidemics were compared from three production systems: organic (OS), organic under shade (OSS) and conventional under full sun (CS). This study was conducted in Ervalia-MG, Brazil, from November 2004 to October 2008. Disease progress was modelled with all 4 years of data in a time series analysis. Disease intensity (severity (SEV) and incidence (INC)), leaf setting (LS) and leaf fall (LF) were assessed on a monthly basis. The highest values for SEV, INC, LF, LS, as well as a larger area under the disease progress curve and maximum disease occurred in CS and in the upper branches, whereas the lowest values for all variables occurred in OSS and in the lower branches. The highest values occurred for SEV and INC from May to July, for LF from July to September, and for LS from October to January. The disease progress was successfully modelled via a time series analysis. The seasonal behaviour of disease progress for all years and production systems was modelled with a nonlinear sinusoidal model with autoregressive moving average (ARMA) errors. Estimated parameters were generated which could be useful for comparative epidemiology, and it was shown that shade could be used in the field for BES management.