Stability of Seed Traits in Partially Interspecific Cotton Lines Across Irrigation and Fertilization Regimes

Cotton (Gossypium spp.) seeds are a valuable source of protein, oil, and minerals; however, seed-quality traits have received less attention than fiber traits, particularly in partially interspecific germplasm. This study evaluated the performance and stability of five cottonseed quality traits (1000-seed weight, crude protein, oil, ash, and crude fiber) in four partially interspecific Pa7 cotton lines (G. hirsutum × G. barbadense) and one commercial cultivar, grown under three irrigation levels and two nitrogen fertilization regimes across two Mediterranean growing seasons in Northern Greece. A strip–split plot factorial design with three replications was used, and year × irrigation combinations were treated as six distinct environments. Trait responses were analyzed using multi-way ANOVA, stability metrics (stability index and coefficient of variation), correlation analysis, principal component analysis (PCA), and genotype × environment interaction models (AMMI and GGE biplots). Multi-way ANOVA revealed significant effects of genotype, environment, and management practices, as well as their interactions, indicating complex regulation of cottonseed composition. Genotypic effects were significant for all traits, while environmental effects were particularly strong for protein content. The greater environmental sensitivity of protein content highlights the key role of nitrogen-related processes and indicates that optimized fertilization can partially offset environmentally induced variability in seed protein accumulation. Stability analysis showed that storage-related traits (protein, oil, ash, and crude fiber) were generally more stable across environments than 1000-seed weight. Among the genotypes, M4 consistently combined high trait performance with broad stability across environments, whereas M1 exhibited the greatest stability for 1000-seed weight. Multivariate and GEI analyses complemented univariate results by revealing trait associations, physiological trade-offs, and crossover responses among genotypes. Overall, using both stability indices and multivariate analyses enabled a detailed evaluation of cottonseed quality in partially interspecific material, supporting the identification of suitable genotypes and optimization of management practices under varying Mediterranean conditions.