Quantification of root lodging damage in corn using uncrewed aerial vehicle imagery

• Published in: Crop, forage & turfgrass management Vol. 9; no. 2
• Main Authors: Lindsey, A. J., Allred, B., Martinez, L. R., Rouse, Greg, Thomison, P. R.
• Format: Journal Article
• Language: English
• Published: 01-12-2023
• ISSN: 2374-3832; 2374-3832
• DOI: 10.1002/cft2.20241

Accurate quantification of damage associated with root lodging events can help producers assess damage, predict potential yield losses, and help understand potential issues with grain quality that may arise post‐harvest (i.e., kernel weight reductions, premature germination on the ear, or vivipary). The objective of this research was to utilize imagery from an uncrewed aerial vehicle (UAV) to accurately quantify crop canopy height, grain yield, and identify trends in imagery data associated with grain quality after root lodging was imposed at multiple growth stages. Simulated corn (Zea mays L.) root lodging experiments were conducted in 2018 and 2019 with lodging treatments applied at two vegetative or two reproductive growth stages (V10, V14, VT/R1, and R3). At dough stage (R4), visible‐color and multispectral images were collected from each trial. Bare fields were also flown in February to obtain baseline elevation data. Imagery data were used to develop digital surface model (DSM) images and used to calculate indices of normalized difference red edge (NDRE) and normalized difference vegetation index (NDVI). Individual datapoints within each experimental plot were extracted from the imagery files and were compared to ground‐truth measurements. The DSM height values were similar to actual measured heights for most lodging treatments (Adj. R2 = .957). Both NDRE and NDVI exhibited linear trends with height and quality parameters (Adj. R2 = .25–.54), though yield patterns were best described using a quadratic model (Adj. R2 = .42–.60). These procedures hold utility in accurately quantifying canopy height following a root lodging event and hold promise in helping consultants identify yield and grain quality reductions associated with root lodging.
Core Ideas
Digital surface model image data at dough stage correlated strongly with actual height data (Adj. R2 = .957).
Indices using red or red edge reflectance at dough stage were moderate predictors of yield (Adj. R2 = .42–.60).
Prediction of grain quality after root lodging was lower, but trends were evident (Adj. R2 = .25–.50).