Nondestructive single-kernel oil determination of maize by near-infrared transmission spectroscopy

The utility of near-infrared transmission spectroscopy (NITS) for the nondestructive prediction of oil content in single maize kernels was explored. Calibration models were developed from spectral information gathered between 850 and 1050 nm. Nuclear magnetic resonance (NMR) spectroscopy was employed as a reference method to determine the actual oil content of samples used for calibration development and testing. Various positionings of the kernels in the light path and calibration math treatments were explored. The best NITS calibration yielded a 1.2% standard error of cross validation, which was over four times the standard error of NMR reproducibility. Although not as accurate as NMR, NITS does have utility in selecting kernels with the highest oil content from a segregating population.