The detection and differentiation of soybean aphid and brown stem rot on soybean using hyperspectral remote sensing

The use of hyperspectral remote sensing for the detection of crop stress has gained considerable attention and appears to have potential for further development as a scouting tool for integrated pest management. Previous studies have shown hyperspectral reflectance to be sensitive to nutrient-, disease-, water-, and pest-related stress. Identifying areas of stress in the field is an important step toward implementation of hyperspectral technologies as an informative tool in agriculture; however, there is a need for more research into hyperspectral models that accurately differentiate between stressors in the field. As a model system for development of hyperspectral remote sensing, we are working with soybean aphid (Aphis glycines Matsumura) and brown stem rot (Cadophora (Phialophora) gregata), which are two important pests of soybean (Glycine max) in the North Central U.S. In a caged field study we manipulated soybean aphid and brown stem rot pressure to examine the effect of these stressors on ground-based hyperspectral reflectance (350 to 2500 nm) of soybean plants. Regression analysis was used to examine trends between cumulative aphid days and reflectance values. ANCOVA was used to observe how the presence of brown stem rot modified the initial correlations. Results will be discussed in the context of understanding how both reflectance measurements and timing of sampling can be used to improve crop stress detection and differentiation under situations with potential confounding factors. The findings may shed light on similar systems involving hyperspectral reflectance and multiple stressors on a shared host.