Abstract
OBJECTIVE: Combining ability analysis forms the basis for selecting suitable parents and hybrid combinations. The performance of combining ability is influenced significantly by environmental conditions. Therefore, it is essential to comprehensively evaluate advantageous parents and hybrid combinations through multi-year experiments. METHODS: In this study, seven parental lines were crossed using a complete diallel design [p(p-1)/2], producing 21 combinations. Yield-related traits, including boll number per plant, seed cotton per plant, lint cotton per plant, boll weight, lint percentage, and seed index, were measured in the parental lines and their hybrid F1 generations over two consecutive years. Combining ability and interannual stability were subsequently analysed. RESULTS: Cotton yield traits were significantly influenced by environmental factors such as weather, temperature, and soil conditions. Interannual variation and genotype-year interactions contributed substantially to variation in yield traits. The interactions between year and both general combining ability (GCA) and specific combining ability (SCA) were significant or highly significant, with GCA stability being relatively low. Yield traits were governed by both additive and dominance genetic effects, with additive effects being predominant. The comparison coefficients for stability of GCA for lint cotton per plant and seed cotton per plant across years were relatively high (51.372% and 55.187%, respectively), whereas SCA stability coefficients for lint percentage and boll weight were comparatively lower (44.986% and 48.645%, respectively). Parent Xinluzao62 (Parent 7) exhibited both high and stable GCA and can therefore be recommended as a backbone parent for yield improvement. Additionally, eight hybrid F1 combinations showing excellent and stable SCA were identified, suitable for use as dominant combinations. CONCLUSION: The GCA and SCA for yield traits in upland cotton are notably influenced by environmental factors such as climate and soil. Yield performance varies considerably across different years and locations. Therefore, selections based on multi-year and multi-location trials provide more reliable results, offering a solid theoretical basis for developing high-yield cotton varieties.