Dissecting genetic insights and combining abilities into growth and productivity of bread wheat under rainfed environments.

Bread wheat (Triticum aestivum L.) production faces increasing instability under rainfed and semi-arid conditions, necessitating the development of cultivars adapted to these environments. However, the genetic basis of physiological growth dynamics and their connection to yield under water-limited conditions remains poorly understood, limiting breeding progress. To address this, six genetically diverse parental lines were crossed in a full diallel mating design (Method I, Model I) and evaluated with their F₁ hybrids using a randomized complete block design with three replications across seven phenological stages (Zadok scale Z-20 to Z-92). Analysis of variance showed highly significant (P < 0.01) genetic variation among genotypes for all growth, yield, and phenological traits. Among parental lines, Pakistan-2013 displayed superior performance, with early heading (110.5 days), longer grain-filling duration (46.1 days), the highest thousand-grain weight (47.3 g), and the maximum grain yield (27.6 g plant⁻¹). In contrast, NR-514 was the weakest parent, with the lowest thousand-grain weight (32.0 g) and grain yield (12.2 g plant⁻¹). Among direct crosses, Pakistan-2013 × NR-499 showed superiority, with the maximum number of tillers (16.1 plant⁻¹), the highest grains per spike (65.1), the highest thousand-grain weight (51.2 g), and the highest grain yield (31.5 g plant⁻¹). Reciprocal effects were most notable in NR-499 × Pakistan-2013, with significant positive RCA effects for thousand-grain weight (4.933**) and grain yield (4.929**), indicating substantial maternal contributions. General combining ability (GCA) effects identified Pakistan-2013 as the best general combiner for yield-related traits (GCA = 3.785** for TGW; 3.556** for GYPP), suggesting dominant additive gene action. Specific combining ability (SCA) effects were highest in Borlaug-2016 × NR-516 for thousand-grain weight (5.626**) and grain yield (7.471**). Additionally, Pakistan-2013 × NR-499 showed significant positive SCA for grain yield (2.835*), indicating strong non-additive gene action (dominance/epistasis) and heterotic potential in these elite hybrids. Stage-dependent gene action was observed for growth traits, with additive effects prevailing during early stages and non-additive effects influencing reproductive stages. Baker's ratio (> 0.70) and GCA: SCA ratios (> 1) confirmed the predominance of additive gene action for most traits, supporting early-generation selection. These findings establish a genetic framework that links growth dynamics, phenology, and yield components, offering practical strategies for developing climate-adapted wheat ideotypes for rainfed ecosystems. Future research should focus on multi-environment trials to confirm the stability of combining ability and utilize molecular marker-assisted selection, thereby speeding up the introgression of favorable alleles into breeding programs.