Abstract Durum wheat ( Triticum durum Desf.) is a critical staple and cash crop in semi‐arid regions, yet its productivity remains highly vulnerable to climate‐induced abiotic stresses, particularly drought and rising temperatures. This study leverages data from the 30th Elite Regional Durum Wheat Yield Trials conducted across 13 rainfed and irrigated environments in Iran (2022–2025) to dissect genotype × environment (G × E) interactions and identify high‐performing, stable genotypes for deployment under increasing climatic uncertainty. Using an integrative analytical framework—combining Additive Main Effects and Multiplicative Interaction (AMMI), Genotype plus genotype‐by‐environment (GGE) biplot, and Partial Least Squares (PLS) regression with 19 climatic covariates—we quantified the relative contributions of environment, genotype, and their interaction, delineated environmental groups, and identified key climatic drivers of yield variation. Results revealed that environment accounted for 81% of total yield variance, with terminal drought (low June rainfall) imposing a universal constraint. G × E interaction, though modest in magnitude (7.65%), was over six times larger than genotype main effects, underscoring its operational relevance for varietal recommendation. Breeding lines G12 (CIMMYT‐derived) and G24 (Iranian) emerged as top candidates: G12 featured among the top‐four performers in 10 of 13 environments and ranked first in 7, while G24 displayed exceptional stability and broad adaptability. PLS modelling identified February and April rainfall—coinciding with tillering to heading—as the strongest climatic predictors of yield, surpassing total seasonal precipitation in explanatory power; late‐season (May–June) temperature also significantly modulated G × E, particularly under terminal heat stress. Environment evaluation highlighted KD3 (Khorramabad) and KH3 (Kermanshah) as the most discriminative and representative test sites, whereas MN4 (Moghan) and IM5 (Ilam) served as effective stress filters. Collectively, our findings support an empirical framework that integrates pattern recognition (AMMI/GGE) with environmental modelling (PLS) to guide environment‐targeted selection—providing a foundation for developing durum wheat varieties with improved adaptation to Iran's heterogeneous rainfed agroecologies.
Mohammadi et al. (2026) studied this question.