Factor analysis of physicochemical properties of 63 rice varieties

Publisher Website
Google Scholar
Abstract
A set of 63 Chinese‐adapted non‐waxy rice varieties of the indica and japonica subspecies was measured to extract simple indicators for rice quality evaluation. Approximate methods including amylose content (AC), gel consistency (GC), gelatinisation temperature (as measured by alkali spreading value, ASV), protein content, etc and instrumental methods including differential scanning calorimetry (DSC), viscoamylography (with a Rapid Visco‐Analyser, RVA) and texture profile analysis were adopted. Four sets of data, ie ordinary testing, flour pasting property, texture of the cooled flour gel, and visual appearance of whole milled grain and flour, were obtained and analysed for mutual relationships. Wide variations were observed in most of the measurements, eg AC (ranging from 6.3 to 28.2%), gelatinisation temperature (Tp, 65.8–83.0 °C), protein content (7.2–13.5%), pasting (peak viscosity, 92–319 RVU) and texture (hardness, 2.5–24.4 g) properties and grain appearance (grain length, 4.8–8.4 mm), to mention a few. Significant correlations were also detected among the four sets of data of the rice varieties, such as AC versus SB (setback viscosity) with a correlation coefficient of −0.73, AC versus HRD (gel hardness) of 0.73, HRD versus FRC (gel fracturability) of 0.99, etc. Generally, the instrumental measurements including DSC, pasting and texture properties were highly correlated with the three ordinarily used indicators of rice quality, ie AC, GC and ASV. However, there was a lack of correlation between pasting properties and textural parameters, implying effects on gel texture of retrogradation of starch after cooling of the paste. Six factors were computed by principal component factor analysis that explained 80% of the variation in the traits, among which amylose content‐related (30.8%), gelatinisation temperature‐related (14.9%) and grain appearance‐related (11.1%) factors were most important. Gel pasting properties evaluated by RVA, and flour colour played less important roles in explaining total variation (accounting for a combined 23% of the eigenvalue loading). This report can be used to identify predictive parameters for rapid screening of rice genotypes for eating quality.© 2002 Society of Chemical Industry