Hyperspectral soybean reflectance to predict soybean aphid (Hemiptera: Aphididae) feeding under greenhouse conditions

Remote sensing of plant spectral reflectance holds promise to increase the efficiency of pest management for the most economically important soybean insect pest in the North Central United States: the soybean aphid, Aphis glycines Matsumura (Hemiptera: Aphididae). Here, we aimed to determine the most promising narrowband spectral wavelengths to predict aphid-induced stress under greenhouse conditions. Four independent caged-trials were conducted using a completely randomized design with six replications of two treatments (i.e., aphid-infested and aphid-free plants). Aphid densities and leaf spectral reflectance were recorded by trifoliate twice a week. The reflectance of 2,151 narrowband wavelengths were recorded between 350 nm and 2,500 nm (i.e., ultraviolet, visible, near-infrared and far-infrared spectral ranges). Analysis of variance was used to determine the cumulative effect of aphid injury, the within-plant position of the aphid feeding and its interaction on the reflectance of each narrowband wavelength. Ongoing analysis indicated that, under greenhouse conditions, there is no significant interaction between the cumulative aphid injury and the position of the aphid feeding across wavelengths of the ultraviolet, visible and near-infrared spectral ranges. Nonetheless, cumulative aphid injury affected soybean reflectance of most wavelengths of these spectral ranges. Comparison of models with the significant effect of cumulative aphid injury will be presented. This study may support the development of sensors that use few wavelengths that can best describe the spectral changes caused by injury of the soybean aphid.