Avaliação de sistemas de previsão de ocorrência de phytophthora infestans em batata.

Abstract

Late blight in the most important disease of potato crop and it may cause major yield losses in a matter of a few days. The use of forecast modls or systems to predict diseases has the potential of reducing the number of fungicide applications, which reduces production costs, the ocurrence risks of severe epidemics, and environmental pollution. The objective of this work was to evaluate the perfomance of Blitecast and Prophy systems for late blight prediction Asterix and SMIJ461-1 potato clones at Santa Maria, RS, Brazil. Two experiments were carried in the experimental area of the Crop Science Department during Spring 2004and Autumn 2005. Meteorological data necessary to run the models were collected in the center of the canopy at 10 and 150cm height above the soil surface and in a standard meteorological station located about 80 m from the canopy. The experimental design was a completely randomized, with four replications. Each plot was composed by 4 rows of plants with 5 m in length. Different accumulated severity values (VS), calculated with Blitecast and Prophy forecast systems, were used to indicate the time of fungicide application for late blight control in the two potato clones. The treatments wereformed by the accumulated VSgenerated by the two systems. The Prophy system predicted a larger number of fungicide applications throughout the developmental than the Blitecast system. The fungicide application scheme based on the Blitecast system increased yield by at least 34,5% compared with control untreated treatments and reduced the number of applications in 25% during wet periods and in 70% during dry periods, having similar efficiency to the obtaines with one weekly application of fungicide in controlling late blight. These results, indicate that it is possible to reduce the number of sprayings with fungicides without affecting potato yield.