Genetic analysis of combining ability, gene action, and heterosis for morpho-agronomic and grain quality traits in hybrid rice ( Oryza sativa L.)

Abstract Background : Hybrid rice ( Oryza sativa L.) technology utilizing heterosis is essential for improving food security. This study assessed the combining ability, gene action, and heterosis for grain yield (GY), agronomic traits, and grain quality in newly developed rice hybrids. Methods : A line × tester (L × T) mating design was used involving three cytoplasmic male sterile (CMS) lines and nine restorers to generate 27 hybrids. Field trials were conducted at Rice Research Department Farm, Egypt, during 2023 and 2024 seasons. All lines, testers, and hybrids were transplanted in 5 m 2 plots using a randomized complete block design (RCBD) with three replications. Standard crop husbandry practices were followed. Morpho-agronomic and quality traits data were collected using standardized protocols. The analysis of variance (ANOVA), correlation, principal component analysis (PCA), and clustering analysis were performed using RStudio. Combining ability and gene action were estimated using GENES software, while heterosis was calculated based on better-parent performance. Results : Specific combining ability (SCA) variances exceeded general combining ability (GCA) for most traits, indicating dominance gene action, except for hulling percentage (HP) and amylose content (AC) showing additive effects. Lines significantly influenced fertility percentage (FP), thousand grain weight (TGW), and AC, while testers affected days to heading (DTH), panicle weight (PW), and HP. L × T interactions contributed most to trait variation. Hybrid C12 showed the highest GY, while C27 exhibited significant heterosis for flag leaf traits and GY. Tester T1 had positive GCA effects for plant height (PH), chlorophyll content (CHC), total spikelet number per panicle (TSN), and elongation percentage (EP). CHC strongly correlated with TSN and GY. The PCA revealed over 54% of total phenotypic variation for first three principal components, with PC1 linked to yield-related traits (CHC, PW, TSN, TGW, GY, and harvest index [HI]), PC2 for developmental and architectural traits (DTH, flag leaf greenness [FLG], panicle length [PL], and gelatinization temperature [GT]), and PC3 for milling quality traits (PL, milling percentage [MP], head rice percentage [HRP], and AC). Cluster analysis clearly separated CMS lines (Cluster I) from high-performing testers and hybrids (Cluster II). CHC positively correlated with TSN, GY, and HI highlighting it as a reliable indirect selection criterion for yield in hybrid rice. Among hybrids, based on considerable phenotypic diversity for morpho-agronomic and milling quality traits and high SCA, C12, C4, and C11 are promising candidates for hybrid rice breeding in the Mediterranean environments. Conclusions : Traits TSN, HI, and GY are key targets for hybrid rice improvement with spikelet number as a critical selection trait due to its association with higher yield and biomass efficiency. With superior mean performance, significant positive SCA effects, and desirable better-parent heterosis (BPH) exceeding 40% for GY, hybrids C12, C4, and C11 are ‘promising candidates’ for use as parent material in hybrid breeding. Additive gene action in several traits supports the selection of favourable parental lines and provides a strategic foundation for breeding programmes to optimize hybrid combinations and target key traits for enhanced yield and grain quality.