Agro-Morphological Characterization and Diversity Analysis of Selected Sorghum Genotypes in Kenyan Agro-Ecologies

Abstract

Sorghum (Sorghum bicolor) is a very important cereal crop cultivated globally, primarily in arid and semi-arid regions. It ranks fifth among the most important cereal crops globally, after wheat, rice, maize, and barley. Despite its crucial role in food security and climate resilience in arid and semiarid regions, sorghum production remains suboptimal, with yields consistently falling below the crop’s genetic potential. Understanding its morphological diversity is essential for effective breeding programs and genetic resource conservation. This study aimed to determine valuable morphological variation among selected sorghum genotypes against released varieties using a diverse set of traits and their correlations. The experiment was laid out using a Randomised Complete Block Design with three replications using 13 genotypes sourced from the University of Eldoret, and 03 checks from the Kenya Seed Company in Kenya. The genotypes were grown in Endebess and Sigor for one season and evaluated based on morphological traits. Using GenStat statistical software 14th Edition, data on qualitative and quantitative traits were analysed at 5% level of significance. The significant differences among the sorghum genotypes were tested using analysis of variance (ANOVA). Correlation Matrices, first, second and third principal component (PCA) were performed. Principal component analysis revealed the three most important PCs that contributed 81.78%, 15.33% and 1.5% of the total variation, respectively. At the Endebess site, grain yield exhibited the highest genotypic variation among the evaluated sorghum genotypes. E1291 recorded the longest leaves (67.87 cm), whereas Kalatur exhibited the shortest (36.13 cm). Moreover, mean comparisons between the two environments showed that Sigor recorded a higher mean grain yield (2.01 t ha⁻¹) compared to Elgon Downs (1.73 t ha⁻¹). Plant height (0.889) was the trait that contributed most to the variation in the first PC. Number of days to harst (0.814) contributed most to the variation in the second PC, whereas leaf length (0.842) was the largest contributor to the variation observed in the third PC. Correlation analysis showed significant positive relationships between 50% days to emergence to 50% days to flowering and days to maturity (r=0.7 and r=0.9), respectively, suggesting that these traits can be used as selection criteria in breeding programs. The frequency distribution analysis indicated a high occurrence of pigmented leaves (93.75%) and brown grain colour (68.75%), reflecting the natural variability within the studied population. The phenotypic evaluation of sixteen sorghum genotypes revealed significant agro-morphological diversity, confirming the genetic variability. These findings support informed selection and genetic improvement to boost yield and stress resilience in sorghum breeding for Kenya and similar regions.