Quantifying Alfalfa Yield Losses Caused by Foliar Diseases in Iowa, Ohio, Wisconsin, and Vermont

Abstract

Although foliar diseases of alfalfa occur throughout the United States wherever alfalfa is grown, little work has been done to quantify yield losses caused by foliar pathogens since the late 1980s. To quantify the yield losses caused by foliar diseases of alfalfa, field experiments were performed in Iowa, Ohio, Vermont, and Wisconsin from 1995 to 1998. Different fungicides and fungicide application frequencies were used to obtain different levels of foliar disease in alfalfa. Visual disease and remote sensing assessments were performed weekly to determine the relationships between disease assessments and alfalfa yield. Visual disease assessments of percentage of defoliation, disease incidence, and disease severity were performed weekly, approximately five to six times during each alfalfa growth cycle. Remote sensing assessments also were obtained weekly by measuring the percentage of sunlight reflected from alfalfa canopies using handheld, multispectral radiometers. Yield loss estimates were calculated as the yield difference between the fungicide treatment with the highest yield and the nonfungicide control, divided by the yield obtained from the highest yielding fungicide treatment × 100. Over the 4-year period, significant alfalfa yield losses (P ≤ 0.05) occurred on 22 of the 48 harvest dates for the four states. The average significant yield loss for the 22 harvests was 19.3%. Both visual and percentage of reflectance assessments were used as independent variables in linear regression models to quantify the relationships between assessments and alfalfa yield. From 1995 to 1998, visual disease assessments were performed for a total of 209 dates and remote sensing assessments were performed on 198 dates from the four states. Yield models were developed for each of these assessment dates. There were 26/209, 26/209, and 17/209 significant yield models based on percentage of defoliation, disease incidence, and disease severity, respectively. Most of the significant models were for disease assessments performed on or within 1 or 2 weeks of the date of alfalfa harvest. When the significant models were averaged, percentage of defoliation, disease incidence, and disease severity explained 51, 55, and 52% of the variation in alfalfa yield, respectively. There were a total of 68/198 significant alfalfa yield models based on remote sensing assessments, and the significant models (averaged) explained 62% of the variation in alfalfa yield. Alfalfa foliar diseases continue to have a significant negative impact on alfalfa yields in the United States and remote sensing appears to offer a better means to quantify the impact of foliar diseases on alfalfa yield compared with visual assessment methods.