Abstract: In order to screen wheat（Triticum aestivum L.） genotypes with high yield, wide adaptability, and coordinated agronomic traits, 33 advanced wheat lines were used as materials, and seven traits including plant height, flag leaf area, spike number, grains per spike, 1 000-grain weight, internode length below the spike, and yield were measured. A systematic evaluation was conducted through variation analysis, correlation analysis, multiple regression, principal component analysis (PCA), and cluster analysis. The results showed that there was rich variation in traits within the population. The flag leaf area had the largest variation (with the coefficient of variation of 22.35%), while the plant height had the lowest variation (with the coefficient of variation of 6.17%). Spike number, 1 000-grain weight, and flag leaf area were positively correlated with yield, and multiple regression analysis showed that these three traits had significant (P<0.05) or extremely significant (P<0.01) positive effects on yield, indicating that yield improvement could be achieved by synergistically improving these traits. The results of principal component analysis showed that the cumulative contribution rate of the first three principal components reached 72.69%, which represented yield composition, the balance of photosynthesis and plant type, and the balance of grain traits. The lines were divided into four categories by cluster analysis. Group Ⅰ(seven lines) was the high-yield type, and the average yield was 8 265.56 kg/hm², among which the lines J32 and J15 were prominent.

Key words: wheat (Triticum aestivum L.), resources, agronomic characters, principal component analysis, cluster analysis, comprehensive evaluation of multiple characters