Univariate–Multivariate Integration To Assess The Consistency Of Agronomic Performance And Sustainability Potential Of Sorghum Bicolor (L.) Moench ‘Keller’ From Main To Second Ratoon Crops

Comprehensive analysis highlights sorghum ‘Keller’ capacity to sustain agronomic performance and sustainability across main and second ratoon crops. To identify sorghum ‘Keller’ as a superior variety and evaluate its sustainability potential through univariate–multivariate integration. This research comprised two stages on dry land in Sindang Jaya Village, Bengkulu, Indonesia. The first study compared four sorghum varieties (B100, Keller, Numbu, Wray). The second tested Keller and B100 up to second ratoon using univariate–multivariate analyses (ANOVA, Tukey HSD, PCA, biplot, PERMANOVA, Pearson). The results showed that in the first study, PCA score plot and biplot highlighted Keller’s superior profile, strongly associated with leaf area, stem, leaf, and root weights, clearly separating it from other varieties. In the second study, K-means clustering consistently distinguished varieties and phases, with Keller forming stable clusters and B100 showing higher variability. Keller declined in ratoon-1 vegetative traits but partially recovered in ratoon-2, while reproductive traits decreased gradually. Conversely, B100 showed significant increases in seed weight and spikelet length but sharp vegetative decline without recovery. Pearson correlation indicated strong positive associations among vegetative traits and negative correlations with seed weight, suggesting a potential trade-off, likely influenced by ratoon data. Nevertheless, Keller maintained strong capacity and higher stability in sugar content (18.0→15.0→13.0%) compared to B100’s fluctuation (8.7→13.0→8.0%). The integration of statistical approaches effectively identified ‘Keller’ as a highly adaptive variety for sustainable ratooning systems, offering valuable insights for agronomic selection strategies.