Abstract: In order to explore the genetic diversity of sweet sorghum, 31 varieties of sweet sorghum were selected in this experiment. Through variance analysis, correlation and principal component analysis, cluster analysis of biological traits and SSR molecular marker test, the following results were obtained. Thirteen biological traits had significant differences among varieties, indicating there were differences in yield and quality of varieties. There was certain correlation between biological traits. Plant height was positively correlated with stem diameter, fresh weight and leaf number, and stem diameter was positively correlated with fresh weight, crude protein and ether extract at 0.01 level. There were significant negative correlations between tiller number and stem diameter, and between stem diameter and coarse fiber at 0.01 level. Principal component analysis could take the first four principal components as principal components factor of quality, main factor of yield, sugar content and secondary factor of yield, auxiliary factor of yield. Cluster analysis based on biological characteristics could divide all varieties into six groups. Group Ⅰ belonged to high yield type. Group Ⅱ belonged to high yield but low quality type. Group Ⅲ belonged to high sugar content type. Group Ⅳ and Ⅴ had little utilization value, and Group Ⅵ belonged to high tillering type. According to the results of SSR molecular marker analysis, the polymorphism information content (PIC) of sweet sorghum cultivar resources was 0-0.757. All varieties were divided into six groups based on SSR molecular marker analysis, and this was similar to the clustering results of biological traits, but it was not exactly the same.

Key words: sweet sorghum (Sorghum bicolor L. Moench), biological characters, SSR marker, genetic diversity, cluster analysis