Genotype selection based on multiple traits in multi-years is frequently influenced by unpredictable rainfed conditions. The main objective of the study was to apply the new methodology of genotype by yield*trait (GYT) biplot for genotype selection and trait profiles in durum wheat genotypes based on multi-traits and multi-year data under rainfed conditions of Iran. A superiority index was applied based on GYT table for ranking of genotypes by the mean of all traits. The GYT biplot ranked the genotypes based on their levels in combining yield with other key traits. Grain yield was combined with target traits and showed the strengths and weaknesses of each genotype. Based on GYT-biplots the relationships among the studied traits were not repeatable across years, but they facilitated visual genotype comparisons and selection. The breeding lines G13, G10 and G15 ranked as the best in combination of the morph-physiological traits i.e., SPAD-reading, early heading, flag-leaf length and number of grain per spike with grain yield under rainfed conditions. The results indicate that there is a potential for simultaneous improvement of some characteristics of durum wheat under rainfed conditions. The GYT biplot was a useful tool for exploring the combination of yield with traits and trait profiles of the durum genotypes to obtain high genetic gains in the durum breeding programs.
Drought is the major cause of durum wheat yield losses in the Mediterranean and many other regions where the crop is not normally irrigated. Over three years (2010–13), 24 durum wheat genotypes representing diverse genetic materials were tested under drought and irrigated conditions. The main objectives were to assess the degree of genotypic variation for drought tolerance, characterize genotypic differences in response to drought, and identify sources of germplasm with greater drought tolerance than old and new cultivars. The percent reduction in average grain yield under drought conditions as compared to irrigated conditions was maximum (69%) during 2012–13, followed by 2010–2011 (33%) and 2011–2012 (15%). The average yields of genotypes under drought conditions differed significantly, which ranged from 1174 (correspond to old variety) to 2086 kg/ha (correspond to breeding line G2). The maximin-minimax approach, yield tolerance index (YTI) and three-dimensional (3-D) plot were used to classify genotypes for drought tolerance and yield productivity. Based on the results, two genotypes were identified as resistant and high yielding (G3 and G20), and eight genotypes (G2, G22, G8, G11, G15, G1, G9 and G5) were found to be high yielding and tolerant to drought conditions. Among the methods, the maximin–minimax approach appears to be more useful in identifying high yielding and drought tolerant genotypes as it seeks to minimize percentage yield loss while maximizing yield potential. In conclusion, considerable variability in yield and drought tolerance was observed for the durum wheat genotypes, which could be exploited at improving drought tolerance in durum wheat breeding program.