Khosravi, Saeed , Fazeli, Arash , Mohammadi, Reza
2026-08-01 JOURNAL OF CEREAL SCIENCE 2026 130(卷), null(期), (null页)
Durum wheat (Triticum turgidum L. subsp. durum) is essential for pasta and semolina production in the Mediterranean basin, where grain yield and quality are highly sensitive to genotype & times; environment (G & times;E) interactions, particularly under drought-prone environments. This study evaluated 20 durum wheat genotypes (16 elite breeding lines from ICARDA and CIMMYT, plus 4 national checks) across six environments (three cropping seasons and two water-regimes: rainfed vs. supplemental irrigation) to dissect G & times;E effects, trait relationships, and identify stable, high-performing lines for multi-trait improvement. Thirteen agronomic and quality traits were assessed. Variance partitioning revealed that environmental effects, driven largely by inter-annual climate variability, accounted for 73.9% of grain yield variation, whereas key quality traits were predominantly genotype-driven (e.g., semolina color: 53.2%; gluten index: 41.5%). Significant G & times;E interactions were observed for most traits, with strong context dependency for functional quality parameters. Correlation analyses confirmed a negative yield-protein trade-off under irrigation (r = -0.29 to -0.51), which weakened under severe drought (r = -0.02 to -0.41), indicating opportunities to select moderate-yielding, high-protein ideotypes adapted to water-limited environments. Genotype-by-trait (GT) biplots and heatmap clustering analyses identified consistently high-quality, stable breeding lines (e.g., ICARDA-derived G6, G7, G8) and high-yielding, drought-adapted lines (e.g., CIMMYT-derived G17, G18). Environment-specific trait syndromes emerged: irrigated conditions favored co-expression of yield, semolina extraction, and bright color, while rainfed adaptation was associated with higher protein, improved sedimentation values, and larger kernel size. These findings support environmentstratified breeding and multi-trait selection frameworks, integrated via GT biplots and pattern-based clustering, to concurrently advance yield resilience and end-use quality in durum wheat for Mediterranean and other semiarid agroecosystems.