Modeling Yield Losses and Fungicide Profitability for Managing Fusarium Head Blight in Brazilian Spring Wheat

Abstract

Fusarium head blight (FHB) and wheat yield data were gathered from fungicide trials to explore their relationship. Thirty-seven studies over 9 years and 11 locations met the criteria for inclusion in the analysis: FHB index in the untreated check ≥ 5% and the range of index in a trial ≥ 4 percentage points. These studies were grouped into two baseline yields, low (Yl ≤ 3,631 kg ha−1) or high (Yh > 3,631 kg ha−1), defined based on the median of maximum yields across trials. Attainable (disease-free) yields and FHB index were predicted using a wheat crop and a disease model, respectively, in 280 simulated trials (10 planting dates in a 28-year period, 1980 to 2007) for the Passo Fundo location. The damage coefficient was then used to calculate FHB-induced yield loss (penalizing attainable yield) for each experiment. Losses were compared between periods defined as before and after FHB resurge during the early 1990s. Disease reduction from the use of one or two sprays of a triazole fungicide (tebuconazole) was also simulated, based on previous meta-analytic estimates, and the response in yield was used in a profitability analysis. Population-average intercepts but not the slopes differed significantly between Yl (2,883.6 kg ha−1) and Yh (4,419.5 kg ha−1) baseline yields and the damage coefficients were 1.60%−1 and 1.05%−1, respectively. The magnitudes and trends of simulated yield losses were in general agreement with literature reports. The risk of not offsetting the costs of one or two fungicide sprays was generally higher (>0.75) prior to FHB resurgence but fungicide profitability tended to increase in recent years, depending on the year. Our simulations allowed us to reproduce trends in historical losses, and may be further adjusted to test the effect and profitability of different control measures (host resistance, other fungicides, etc.) on quality parameters such as test weight and mycotoxin contamination, should the information become available.