VALIDATION OF FORECASTING SYSTEMS OF FUSARIUM HEAD BLIGHT IMPACT IN WHEAT CULTIVARS
Ricardo C. Moschini, Mario Acuña, Enrique Alberione, Julio Castellarín, Facundo Ferraguti, Homero F. Lozza, Malvina I. Martínez
Instituto de Clima y Agua. CIRN INTA Castelar, Buenos Aires, Argentina
CONAE, CABA. Argentina
EEA INTA M. Juárez, Córdoba, Argentina
EEA INTA Oliveros, Santa Fe, Argentina
Manuscript received on 23rd July 2015, in final form on 1st January 2016
ABSTRACT
In Argentina, wheat Fusarium head blight (FHB) is a disease predominantly caused by the fungus Fusarium graminearum (Schwabe) (teleomorfo Gibberella zeae (Schwein)). FHB is considered a floral disease because the infection process begins at the anthers. Severe FHB epidemics are associated to the occurrence of long wet periods from anthesis to milk-dough ripe wheat stages. The sporadic weather-induced nature of FHB in the Pampa region led to the development and validation of empirical and fundamental-empirical weather-based incidence and Fusarium index (IndF%) forecasting systems. Both systems recognize infective events by head wetting combining the simultaneous occurrence of precipitation (Pr) and high relative humidity. The current study contrasted the Fusarium index values predicted by the weather-based fundamental-empirical system, in its original version and adjusted version (IndFPOri: original and IndFPAju: adjusted), with those observed in many wheat cultivars in M. Juárez (Córdoba, 2011/12 and 2012/13), Gdor. Mansilla (Entre Ríos) in 2012/13 and in Oliveros (Santa Fe, 2012/13 y 2013/14). The greatest observed and predicted IndF values were registered in 2012/13, especially in wheat cultivars that headed before 10 October. Overestimated IndF values were obtained using the original forecasting system in Gdor. Mansilla and Oliveros, in 2012/13. The overestimation was sharply reduced calculating adjusted IndF values, defining a day with precipitation that with occurrence of Pr>0,5mm and thermal amplitude <11oC. Differences between predicted IndFPOri and IndFPAju values were not clearly observed in M. Juarez (2011/12 and 2012/13) and Oliveros (2013/14). Besides, IndF predictions without using anthesis date as input data were calculated. Greater accuracy of simulated anthesis date with the model DSSAT-CERES wheat led to a more precise calculation of predicted IndF values. Genetic coefficients of current wheat cultivars have to be updated for improving the simulation of wheat growing stages.