Harvest Timing of Standing Corn Using Near-Infrared Reflectance Spectroscopy.

Abstract

Harvesting corn at the proper maturity is important for managing its nutritive value as livestock feed. Standing whole-plant moisture content is commonly utilized as a surrogate for corn maturity. However, sampling whole plants is time consuming and requires equipment not commonly found on farms. This study evaluated three methods of estimating standing moisture content. The most convenient and accurate approach involved predicting ear moisture using handheld near-infrared reflectance spectrometers and applying a previously established relationship to estimate whole-plant moisture from the ear moisture. The ear moisture model was developed using a partial least squares regression model in the 2021 growing season utilizing reference data from 610 corn plants. Ear moisture contents ranged from 26 to 80 %w.b., corresponding to a whole-plant moisture range of 55 to 81 %w.b. The model was evaluated with a validation dataset of 330 plants collected in a subsequent growing year. The model could predict whole-plant moisture in 2022 plants with a standard error of prediction of 2.7 and an R 2 P of 0.88. Additionally, the transfer of calibrations between three spectrometers was evaluated. This revealed significant spectrometer-to-spectrometer differences that could be mitigated by including more than one spectrometer in the calibration dataset. While this result shows promise for the method, further work should be conducted to establish calibration stability in a larger geographical region.

Plant phenotyping relevance

携帯型近赤外分光計でトウモロコシの耳水分を測定し、全植物体水分（成熟度の表現型）を推定する手法を開発・検証しており、校正の機器間移植性も評価しているため、方法論が中心です。

Code and data availability

The paper's NIRS spectra and moisture reference data (610 calibration plants from 2021, 330 validation plants from 2022) are not publicly deposited; the Data Availability Statement says they are available only upon request from the corresponding author. No author analysis code, trained model, or supplement with data is

No evidence-backed public reproduction asset is currently recorded.