Modeling canopy senescence to calculate soybean maturity date using NDVI

Soybeans [Glycinemax (L.) Merr.] are the second largest crop in the United States, and cultivar development is important to improve productivity on existing farms. High-throughput phenotyping can assist breeders in characterizing new lines particularly for phenotypic traits of interest such as crop canopy color, leaf distribution, and seed yield. However, current methods for characterizing the maturity date of soybeans needs to be improved to further leverage high-throughput phenotyping platforms. The objective of this study was to examine if relative differences in leaf senescence patterns of soybean cultivars could be used to accurately calculate the observed maturity date. A field study was conducted in 2016-2017 at two locations in Ohio. Leaf canopy senescence was measured using the normalized difference vegetation index (NDVI), and observed maturity was measured for each cultivar. Senescence was modeled for each cultivar, and relative differences in senescence calculated from NDVI reductions were compared with the relative differences in observed maturity for each cultivar. The relative differences between cultivars in the date when NDVI was reduced by 85% during senescence (compared with NDVI at the R5 growth stage) were accurate to calculate maturity differences to within 1.5 d of the observed values for 21 of the 24 cultivar x site-year combinations. This practice has potential to contribute additional data to enhance the high-throughput phenotyping process using existing equipment (or data being collected) in many programs.