【摘要】 以108个粳稻品种为试材,测定糙率等25项品质指标,运用差异性分析和相关分析探索指标间的相互关系,利用因子分析和聚类分析确定代表性品质指标。结果表明:25项指标中变异系数超过10%的共有10项。相关性分析共得到300个相关系数,其中在0.05水平存在显著相关性的为31个,在0.01水平存在显著相关性的为69个。因子分析提取了9个公因子,解释的累积方差贡献率达到80.78%。应用系统聚类方法将所有指标聚成8类,包括粳稻的淀粉特性、成熟度、营养品质、籽粒特性、蒸煮品质、粒型特征、颜色特征和碾磨品质。综合分析最终确定8项粳稻品质代表性指标,即消减值、长宽比、垩白度、蛋白质、糙米白度、直链淀粉、千粒重和整精米率。本文旨在为粳稻品质评价提供研究基础。更多还原

【Abstract】 108 japonica cultivars were used as test materials to determine 25 quality indexes such as roughness and to explore the interrelationship among indexes by using difference analysis and correlation analysis. The representative quality indexes were determined by factor analysis and cluster analysis. The results showed that there were altogether 10 items with a coefficient of variation exceeding 10% in 25 items. Correlation analysis yielded 300 correlation coefficients,of which 31 were significantly related at α = 0.05 level and 69 were significantly correlated at α = 0.01 level. Nine common factors were extracted from factor analysis, explaining the cumulative variance contribution rate of 80.78%. The clustering method was used to cluster all the indicators into eight categories, including the starch characteristics, maturity, nutritional quality, grain characteristics, cooking quality, grain type, color characteristics and milling quality of japonica rice. The comprehensive analysis finally determined eight representative indications of the quality of japonica rice, including the subtraction value, the aspect ratio, the chalkiness degree, the protein, the brown rice whiteness, the amylose, the 1000-grain weight and the milled rice rate. The purpose of this paper is to provide a research basis for japonica quality evaluation.更多还原