Multispectral canopy reflectance and digital imaging to determine maturity

Abstract

Soybean breeders and producers may benefit from approaches that can quickly and quantitatively measure soybean developmental stages. To better predict developmental stages, seasonal dynamics of soybean canopies were characterized using canopy reflectance and digital imaging tools. Field experiments were conducted in 2014 and 2015 at two Missouri locations to characterize canopy characteristics of six commercial soybean [Glycine max (L.) Merr.] cultivars – two cultivars each of maturity group (MG) 3, 4, and 5 – planted at two or three planting dates (PD). Normalized difference vegetation index (NDVI), normalized difference red edge (NDRE), dark green color index (DGCI), percent canopy cover, and developmental stage were assessed every 7-10 days. Canopy cover and NDVI reached maximum values relatively early in the season, and no measurable differences were seen from R2 through R6, although, lower values were observed at R7. Canopy cover models predicted a day of maximum value (DOM) at beginning R5, while NDVI models predicted a DOM earlier in the season, corresponding with developmental stages between R2 and R5. NDRE and DGCI models had peak maxima and DOM values that coincided with beginning R5 for most site years. These results indicate that, while reflectance sensors and digital images may serve to determine soybean developmental stages, predicting stages solely based on the indices examined in this study may be difficult.