Abstract: The differences in quality properties between 40 major oil and feed soybean varieties in four major production areas of China were analyzed. Ten key quality indexes of soybean for each purpose were determined. One-way analysis of variance (ANOVA) and principal component analysis (PCA) were carried out, and a model was proposed using combination weighting method in order to select the quality characteristics of soybeans for different purposes and the superior soybean production areas and varieties. Significant differences in all quality properties except color, odor and carbohydrate content were found among varieties and production areas (P < 0.01 or P < 0.001). In the PCA, crude fat content, vitamin E (VE) content, impurity content, damage ratio, moisture content, and phospholipid content were extracted by factor analysis as key quality characteristics of oil soybean. For feed soybean, seven key quality characteristics including 7S/11S protein, protein dissolution ratio, crude fat content, protein content, impurity content, damage ratio, and soybean isoflavone content were identified. Using the proposed model, top 10 oil and feed soybean varieties with the highest comprehensive score and their major production areas were selected. Using cluster analysis, the oil and feed soybeans were assigned into five and nine groups with similar quality characteristics, respectively. This study provides a theoretical basis for quality evaluation of soybeans from different major production areas in China and specialty soybean variety selection.

Key words: soybean; oil soybean; feed soybean; quality properties; principal component analysis; combination weighting method; cluster analysis