|Delwiche, Stephen - Steve
|NELSON, LENIS - UNIV NEBRASKA, AGRON DEPT
Submitted to: Cereal Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/5/2002
Publication Date: 11/1/2002
Citation: DELWICHE, S.R., GRAYBOSCH, R.A., NELSON, L.A., HRUSCHKA, W.R. ENVIRONMENTAL EFFECTS ON DEVELOPING WHEAT AS SENSED BY NEAR-INFRARED REFLECTANCE OF MATURE GRAINS. CEREAL CHEMISTRY. 2002.
Interpretive Summary: For 30 years, near-infrared (NIR) spectroscopy has routinely been applied to the cereal grains for the purpose of rapidly measuring concentrations of constituents such as protein and moisture. A study was performed to examine the ability of NIR spectroscopy on harvested wheat to determine weather-related, quality-determining properties that occur during plant development. Twenty commercial varieties or advanced breeders' lines of hard red winter and hard white wheat were grown in ten geographical locations under prevailing natural conditions of the U.S. Great Plains. Using two separate multivariate regression modeling approaches, we demonstrated that some of the weather conditions are measurable from the NIR spectra of the ground wheat. For example, the amount of time during the grain-fill (post-flowering) period when temperatures exceeded 30 degrees Celsius was estimated by NIR modeling at a level that was more than nthree times better than by chance alone. By partial correlation analysis, we were able to show that the success of these weather property models extends beyond the relationship that these properties have to protein content, a property that is easily measured by NIR. With refinement, this technique stands to benefit the plant breeder by providing a tool for rapid quality assessment during field trials, and to wheat traders and processors who desire a rapid method for measuring wheat protein quality.
Technical Abstract: Diffuse reflectance spectra (1100-2498 nm) of ground wheat from approximately 200 samples were modeled by partial least squares one (PLS1) and multiple linear regression algorithms for the following properties: SDS sedimentation volume, amount of time during grain fill in which the temperature exceeded (30, 32, or 35 C) or was less than (24 C) a prescribed dlevel, and the amount of time during grain fill in which the humidity exceeded (80% RH) or was less than (40% RH) a prescribed level. Rainfall values associated with four pre- and post-planting stages were also examined heuristically by PLS2 analysis. Partial correlation analysis was used to statistically remove the contribution of protein content from the quantitative NIR models. PLS1 models of 9 to 11 factors on scatter- corrected and (second order) derivatized spectra produced models whose dimensionless error (RPD = the ratio of standard deviation of the property in a validation set to the model standard error for that property) ranged from 2.0 to 3.3. Multiple linear regression models, involving the sum of four second derivative terms with coefficients, produced models of slightly higher error compared to PLS models. For both modeling approaches, partial correlation analysis demonstrated that model success extends beyond an intercorrelation between property and protein content, a constituent that is well-modeled by NIR spectroscopy.