Modeling the relationship between estimated fungicide use and disease-associated yield losses of soybean in the United States I: Foliar fungicides vs foliar diseases.

• Main Authors: Ananda Y Bandara, Dilooshi K Weerasooriya, Shawn P Conley, Carl A Bradley, Tom W Allen, Paul D Esker
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2020-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0234390

Fungicide use in the United States to manage soybean diseases has increased in recent years. The ability of fungicides to reduce disease-associated yield losses varies greatly depending on multiple factors. Nonetheless, historical data are useful to understand the broad sense and long-term trends related to fungicide use practices. In the current study, the relationship between estimated soybean yield losses due to selected foliar diseases and foliar fungicide use was investigated using annual data from 28 soybean growing states over the period of 2005 to 2015. For national and regional (southern and northern United States) scale data, mixed effects modeling was performed considering fungicide use as a fixed and state and year as random factors to generate generalized R2 values for marginal (R2GLMM(m); contains only fixed effects) and conditional (R2GLMM(c); contains fixed and random effects) models. Similar analyses were performed considering soybean production data to see how fungicide use affected production. Analyses at both national and regional scales showed that R2GLMM(m) values were significantly smaller compared to R2GLMM(c) values. The large difference between R2 values for conditional and marginal models indicated that the variation of yield loss as well as production were predominantly explained by the state and year rather than the fungicide use, revealing the general lack of fit between fungicide use and yield loss/production at national and regional scales. Therefore, regression models were fitted across states and years to examine their importance in combination with fungicide use on yield loss or yield. In the majority of cases, the relationship was nonsignificant. However, the relationship between soybean yield and fungicide use was significant and positive for majority of the years in the study. Results suggest that foliar fungicides conferred yield benefits in most of the years in the study. Furthermore, the year-dependent usefulness of foliar fungicides in mitigating soybean yield losses suggested the possible influence of temporally fluctuating abiotic factors on the effectiveness of foliar fungicides and/or target disease occurrence and associated loss magnitudes.