The following is an abstract for the selected article. A PDF download of the full text of this article is available here. Members may download full texts at no charge. Non-members may be charged a small fee for certain articles.

Novel Near-Infrared Sampling Apparatus for Single Kernel Analysis of Oil Content in Maize

Volume 62, Number 4 (April 2008) Page 423-426

Janni, James; Weinstock, B. André; Hagen, Lisa; Wright, Steve

A method of rapid, nondestructive chemical and physical analysis of individual maize (Zea mays L.) kernels is needed for the development of high value food, feed, and fuel traits. Near-infrared (NIR) spectroscopy offers a robust nondestructive method of trait determination. However, traditional NIR bulk sampling techniques cannot be applied successfully to individual kernels. Obtaining optimized single kernel NIR spectra for applied chemometric predictive analysis requires a novel sampling technique that can account for the heterogeneous forms, morphologies, and opacities exhibited in individual maize kernels. In this study such a novel technique is described and compared to less effective means of single kernel NIR analysis. Results of the application of a partial least squares (PLS) derived model for predictive determination of percent oil content per individual kernel are shown.