Early detection of crop injury from glyphosate by foliar biochemical parameter inversion through leaf reflectance measurement

In this paper, we attempt to detect crop injury from glyphosate by foliar biochemical parameter inversion through leaf hyperspectral reflectance measurements for soybean and cotton leaves. The PROSPECT model was calibrated to retrieve Chlorophyll content (Chl), Equivalent Water Thickness (EWT), and Leaf Mass per Area (LMA) of each leaf from hyperspectral reflectance spectra. The leaf stress conditions were then evaluated by examining the temporal variation of these biochemical constituents after glyphosate treatment. The approach was validated with in situ measured datasets. Results indicated that the coefficient of determination (R²) of Chl, EWT, and LMA were greater than 0.8, 0.7, and 0.5, respectively, for both soybean and cotton. The Root-Mean-Square Error (RMSE) of Chl, EWT, and LMA were reasonably low with 1.2278 μg/cm², 0.0005 g/cm², and 0.0042 g/cm² for soybean and 0.9144 μg/cm², 0.0124 g/cm², and 0.0003 g/cm² for cotton, respectively. It was further found that the leaf injury caused by glyphosate treatments could be detected shortly after spraying by PROSPECT inversion for both soybean and cotton, with Chl of the higher dose solution treated leaves decreasing more rapidly compared with no glyphosate treated leaves, whereas the EWT and LMA showed no obvious difference between injured and healthy leaves. These findings demonstrate the feasibility of applying the PROSPECT inversion technique for the early detection of leaf injury from glyphosate and its potential for agricultural plant status monitoring.