Multivariate Technique to Evaluate Genetic Variability and Relationship Between Physiological Quality Traits of Soybean Seeds

<p>This study evaluated the genetic variability, explored trait interrelationships, and identified efficient physiological quality traits of soybean seeds using multivariate techniques. Fourteen soybean varieties were evaluated based on twelve physiological quality traits related to germination and vigor tests. The data for each trait was analyzed using genetic variability estimation, analysis of variance, Tukey post-hoc test, and multivariate analysis. The findings indicated that most traits exhibited high genetic variability and high heritability, particularly in TSW (CV<sub>g</sub>=22.72%, h_x^2 =99.79) and SDM (CV<sub>g</sub>=26.59%, h_x^2 =94.98). There were no significant differences in SEF and SDMF among the fourteen varieties. Correlation analysis revealed significant positive relationships between TSW and SDM (r=0.95, p<0.05), as well as between G and TTZ (r=0.75, p<0.05). In contrast, TSW and SDM showed negative correlations with GSI and SEF. PCA analysis revealed two PCs explained 68.54% of the total variance. The PCA biplot indicated that G, FGC, GSI, and SL were the most efficient traits for assessing the physiological quality of soybean seeds. Heatmap clustering further revealed that six varieties (Gepak Kuning, Demas 1, Devon 1, Deja 1, Dering 1, and Dering 3) grouped in Cluster II were the most genetically divergent among those evaluated.</p>