Integrating Morphological and Molecular Approaches for Assessment of Genetic Diversity in Advanced Breeding Lines of Castor (Ricinus communis L.)

Genetic diversity assessment in castor ( Ricinus communis L.) plays a crucial role while selecting divergent parents for hybridization programs that target high-yielding, early-maturing varieties. The present study involving 50 advanced breeding lines alongside two check varieties were evaluated at ICAR-Indian Institute of Oilseeds Research (ICAR-IIOR), Hyderabad, examining genetic variability through quantitative morphological traits and Simple Sequence Repeat (SSR) markers. The analysis of variance (ANOVA) showed significant genotypic differences across most traits examined. Plant height, primary raceme length, capsule number per primary raceme and seed yield all displayed high heritability combined with high genetic advance, a clear indication that additive gene action predominates for these characters. Seed yield correlated positively and significantly with capsules per primary raceme, 100-seed weight, and oil content. Using Mahalanobis D² analysis, the 52 genotypes were partitioned into 14 distinct clusters. Cluster I contained the largest number of genotypes (23) whereas 11 clusters had only single genotypes reflecting considerable phenotypic diversity within the material studied. Notably, Cluster VIII genotypes flowered early (51.75 days) and produced heavy seeds (36.60 g per 100 seeds). Cluster XI meanwhile has demonstrated superior performance for seed yield (136.54 g/plot) and oil content (45.48%). For molecular characterization, 32 SSR markers were employed of which eight proved polymorphic. These detected 18 alleles, averaging 2.25 alleles per locus. The Polymorphism Information Content (PIC) averaged 0.51 suggesting moderate marker informativeness. UPGMA clustering using Jaccard's similarity coefficient organized the genotypes into 10 molecular clusters. When comparing morphological and molecular datasets through the Adjusted Rand Index (ARI = 0.041) observed low congruence which indicates these approaches capture independent sources of genetic variation. Integrating both datasets provided a comprehensive framework for parental selection in breeding programs. Based on our findings, crosses between genotypes from Cluster VIII and Cluster XI are particularly recommended for developing superior, high-yielding castor hybrids.