PREDICCIÓN DE SIGATOKA NEGRA EN 'PLÁTANO HARTÓN' BAJO CONDICIONES DE FERTILIZACIÓN DEL SUELO EN EL MUNICIPIO VEROES DEL ESTADO YARACUY, VENEZUELA 1 PREDICTION OF BLACK SIGATOKA ON 'HARTON PLANTAIN' UNDER CONDITIONS OF SOIL FERTILIZATION IN THE MUNICIPALITY OF VEROES OF YARACUY STATE, VENEZUELA 1

SUMMARY The nutritional status of the plant and knowledge of the epidemiology of black sigatoka, Mycosphaerella fijiensis should be considered in the management of the disease in 'Harton Plantain'. During 2002-2004 was installed an experiment to determine a model for predicting the disease under conditions of inorganic fertilizer in the soil. The design was a randomized block with four repetitions and treatments consisted of applying combinations of three levels of nitrogen (N: 0, 90 and 180 g pl -1 ), three levels of potassium (K 2 O: 0, 140 and 280 g pl -1 ) and one of phosphorus (P 2 O 5 : 90 g pl -1 ). The best treatment, was subjected to an analysis of Pearson correlation with climate data taken from the first to the tenth week prior and subsequent to the date of assessment of the severity of the disease. The week in which each variable had the highest correlation coefficient (R) was analyzed by multiple linear regression step by step, against the weighted average number of infection (PPI), by Stepwise procedure of Analytical Software Statistic® Version 1,0. The best combination turned out to be 90 N - 90 P 2 O 5 - 0 K 2 O g pl -1 and the selected model, which was considered the accumulated values of cloudiness (NUBOS) and heatstroke (INSOL) six consecutive weeks prior to last week assessment of the PPI, expressed by the equation: PPI= -12,6233 + 0,05382NUBOS** + 0,01179INSOL**, with an adjusted coefficient of determination (R 2 = 0.6681). The model of prediction is an important tool for managing the black sigatoka with the treatment selected.