Critical-point yield loss models based on incidence and severity of wheat head blast epidemics in the Brazilian Cerrado

Abstract

Wheat head blast (WHB), caused by the fungus Pyricularia oryzae Triticum pathotype, is one of the most damaging diseases of wheat. In Brazil, the disease has limited the expansion of wheat in the tropical regions of the Cerrado biome. In this study, wheat yield (Y) and visual estimates of WHB incidence (I) and severity (S) (assessed 25 to 30 days after flowering) were measured in a single location in the Cerrado (Minas Gerais state) where field experiments were conducted during seven sequential years (2013 to 2019) using three cultivars and 12 planting dates each year. Pearson correlation (r) analysis, used to study the association between each of the three disease variables (I, DHS = diseased-head severity and S) and Y, under a meta-analytic framework, showed the strongest association between incidence and yield (rI-Y = −0.92) and severity and yield (rS-Y = −0.90). The functional relationship between WHB intensity (I or S) and Y was studied based on the fit of a random coefficients model to I or S and the log-transformed Y data. The population average estimated intercept and slope for the I or S and log-Y relationships were 7.798 log units (2436.6 kg/ha) and − 0.020 for the I-Y relationship and 7.721 log units (2255.5. kg/ha) and − 0.030 for the S-Y relationship. The relative (percent) yield loss when I or S reaches 10% was predicted at 18.4% and 26.1%, respectively. This study provides a basis for understanding and predicting yield losses due to wheat blast based on the visual assessment of two disease variables assessed at a single time during the epidemics.