Predicting single kernel and bulk milled rice alkali spreading value and gelatinization temperature class using nir spectroscopy

Authors: Armstrong, Paul; Maghirang, Elizabeth; Chen, Ming Hsuan; McClung, Anna; YAPTENCO, KEVIN - University Of The Philippines; Brabec, Daniel - Dan; WU, TINGTING - Northwest A&f University; WEI, YONG - Tianjin Agricultural University

Submitted to: Cereal Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/14/2022
Publication Date: 7/18/2022
Citation: Armstrong, P.R., Maghirang, E.B., Chen, M., McClung, A.M., Yaptenco, K.F., Brabec, D.L., Wu, T., Wei, Y. 2022. Predicting single kernel and bulk milled rice alkali spreading value and gelatinization temperature class using nir spectroscopy. Cereal Chemistry. 99(6):1234-1245. https://doi.org/10.1002/cche.10587.
DOI: https://doi.org/10.1002/cche.10587

Interpretive Summary: When rice is cooked, short grain rice can become sticky while long grain rice tends to remain intact due to its higher starch gelatinization temperature (GT). Knowing and controlling GT is thus important for breeders and processors in order to deliver a consistent product. The alkali spreading value (ASV) of rice is a widely measured quality parameter used in breeding programs and an industry accepted indicator of the GT of rice starch. The alkali test, developed in 1958, requires the soaking of rice kernels in an alkali solution for nearly 24 hours followed by visual inspection to determine the amount of dissolving that occurs in the kernel. Although labor-intensive and time-consuming, the alkali method remains the most commonly used procedure for GT estimation. A rapid technique to measure ASV and GT for both single kernel and bulk rice would provide breeders and industry an important tool for rice variety development and for quality control. A single-kernel near-infrared (SKNIR) instrument developed by researchers at USDA-ARS-CGAHR and a commercially available bulk NIR instrument were evaluated for determining ASV and identifying rice with low or intermediate GT. The study demonstrated that both instruments could rapidly and accurately classify samples. GT could be classified into intermediate and low values with 82.4% to 85.0% accuracy on the SKNIR. Bulk samples were classified at 93.6% and 84.4% accuracy for intermediate and low GT values on the commercial NIR. In conclusion, both instruments have good potential for rough screening large numbers of samples and identifying rice varieties with intermediate and low GTvalues.

Technical Abstract: The alkali spreading value (ASV) of rice is a widely measured quality parameter used in breeding programs and is a widely accepted indicator of the gelatinization temperature (GT). The alkali test, developed in 1958, requires the soaking of rice kernels in a dilute alkali solution for approximately 23 hours with visual inspection to score the dispersed kernels. It remains the most commonly used procedure for ASV measurement. An objective and rapid technique to measure ASV and GT, both for single kernel and bulk rice, will provide breeders and industry with an important tool for developing rice varieties with specific desired characteristics. In this work, a single-kernel near-infrared (SKNIR) instrument developed at the Center for Grain and Animal Health Research, U.S. Department of Agriculture – Agricultural Research Service and a commercially available bulk NIR instrument were evaluated for the quantitative determination of ASV and qualitative classification of intermediate and low GT for both single kernel and bulk milled rice applications. Quantitative prediction of ASV scores (2 to 7) demonstrated the potential of NIR spectroscopy for rough screening with the standard error of prediction (SEP) for independent validation samples ranging from 0.91 to 1.39 for the SKNIR instrument and from 0.97 to 1.19 for the commercial bulk NIR instrument. GT categorization into intermediate and low values, based on ASV scores, showed potential with 82.4% and 85.0% correct classification of GT using 1-kernel and 30-single kernel average calibration models, respectively. For bulk samples, GT was correctly classified at 93.6% and 84.4% using a commercial bulk NIR instrument equipped with a micro mirror cup (100 kernels) and a small dish (50 g), respectively. In conclusion, NIR spectroscopy (single kernel and bulk) has the potential for use for rough screening of ASV and for two-category classification of intermediate and low GT.