Multi-Trait Selection Index for the Improvement of Agronomic and Yield Traits in Rice (Oryza sativa L.)

Despite the critical role of rice ( Oryza sativa L.) as a staple food in Ghana, national demand far exceeds local production. This low production is partly attributed to the limited availability of improved varieties, highlighting the urgent need for varietal enhancement. This field study evaluated 295 diverse rice genotypes from multiple Ghanaian sources for variation in 14 agro-morphological and yield-related traits using advanced multivariate analyses and the Multi-Trait Genotype-Ideotype Distance Index (MGIDI) to identify superior genotypes for breeding. Significant variation (p < 0.001) was observed across all traits. Moderate to high broad-sense heritability (76.9–86.3%) and strong positive phenotypic and genotypic correlations (r ≥ 0.85) were found among key yield traits, including grain yield, hundred-grain weight, panicle number, and tiller number, underscoring their importance in yield improvement. Conversely, days to heading and maturity were highly positively correlated (phenotypic r = 0.96; genotypic r = 0.98) but strongly negatively correlated with all yield and morphological traits (phenotypic r = − 0.84 to − 0.98; genotypic r = − 0.94 to − 0.99), highlighting the trade-off between early phenology and yield performance. Principal component and hierarchical cluster analyses revealed distinct trait groupings and genotype clusters, providing valuable insight into trait associations and germplasm diversity. Notably, accessions from KUMASI-CRI tended to associate with yield and yield-related traits, including grain yield, hundred-grain weight, panicle number, tiller number, and grain thickness. The MGIDI identified 44 promising genotypes combining desirable trait profiles (e.g., SA2-SARI, Togo Marsha, SA51-SARI, KBR 12, NERICA-L 41, JKE56-30, CRI-AMANKWATIA, AGRA-CRI-LOL-1-21, AGRA-CRI-LOL-2-29) that enable simultaneous improvement of yield, phenology, and plant architecture while overcoming limitations of traditional indices. These genotypes represent valuable breeding material for enhancing rice production and food security in Ghana. Future multi-environment testing and stress screening are recommended to confirm stability and adaptability. This study demonstrates the effectiveness of multi-trait selection indices in accelerating genetic gains in rice breeding programs toward developing climate-resilient, high-yielding cultivars.