عنوان مقاله [English]
نویسندگان [English]چکیده [English]
Durum wheat (Triticum turgidum L. var durum) consists of only 5% of the world’s total cultivated wheat area and contributes about 10% to the total global wheat production. In recent years, the production level of durum wheat has risen to more than 30 million tons and EU, USA and Canada together representing 60% of the production. Durum wheat in Iran is grown on 300-400 thousand hectares with an average annual production of 500-600 thousand tons. Increase in yield is one of the primary aims pursued in plant breeding programs. Similar to other crops, insufficient yield stability in durum wheat is recognized as a one of the factors responsible for the gap between actual yield and potential yield. In breeding programs, the identification of superior genotypes is difficult due to environmental variability of target locations and the interaction of these variability with the investigated genotypes. Therefore, it is important to evaluate the advanced agronomic lines across various environments and over multiple years to ensure their yield stability and production. Many statistical models have been suggested to analyze G×E interaction. The additive main effects and multiplicative Interaction (AMMI) model is a multivariate statistical method that entirely justifies genotype and environment main effects as well as multiplicative G×E interaction effects. This method provides a clear interpretation of G×E interaction effect. The objectives of this study were to analyze genotype by environment (GE) interactions on the seed yield of some durum wheat lines by AMMI model and to evaluate genotype (G), environment (E) and genotype× environment (GE) interactions using statistics parameter i.e. AMMI stability value (ASV) and ecovalence (W2i).
Materials and Methods
Sixteen promising durum wheat lines (G1-G16) along with two check cultivars (durum wheat cv. Hana and bread wheat cv. Parsi), were investigated for two cropping seasons (2015-2016 and 2016-2017) at three Agricultural Research Stations (such as Karaj, Kermanshah and Neishabour cities) The experimental design at all locations was a randomized complete block design with three replications. Some agronomic attributes such as the number of days until anthesis stages, plant height, number of days till physiological maturity, 1000-kernel weight and grain yield were determined for each genotype. However, only the grain yield data was used to analyze G×E interactions. Combined analysis of variance for grain yield was performed using ADEL-R software. The GGE Biplot methodology was employed to analyze G×E interaction. The AMMI M model was used for the following purposes; (i): Evaluation of yield stability, (ii): The simultaneous selection for yield and stability, (iii): Identification of ideal durum wheat genotypes, and (iv): Assessment of the characteristics of and relationships among the testing environments.
Results and Discussion
The combined analysis of variance showed that the main effects of year and location were significant at 1% probability level, while the main effect of genotype had not significant. Genotype× year interaction and triple genotype × year × location interaction were significant at 1% probability level and also genotype × location interaction was significant at 5% probability level, indicating genotype × environment interaction. The results of AMMI ANOVA showed that about 86.5% of total variation was related to environment effect, 1.4% to genotype effect and 12.1% to genotype× environment interaction. Overall, the average grain yield of the evaluated lines ranged from 7.6 to 8.4 t.ha-1 and the G18 and G2 lines had the lowest and highest grain yield, respectively. Main effect due to environment and genotype × environment interaction as well as two first interaction principal components (IPCA1-2) were found to be significant, indicating that the agroclimatic environmental conditions were different, and that there was a differential response of the genotypes to the environments. The first two IPCA components of the GE interaction explained about 70.2% of the GE interaction. According to IPCA1, G9, G15 and G16 had the lowest scores and were the most stable genotypes whereas G17 and G18 with the highest scores were found to be unstable. The lowest ASV was observed for G16 that was the most stable genotype whose mean yield was higher than the grand mean. However, the highest ASV scores were achieved by G17 and G18. AMMI Biplot was used to visualize mean seed yield performance and stability of durum wheat genotypes. AMMI Biplot was able to distinguish stable genotypes with broad sense and narrow sense adaptation and environments with high and low genotype discrimination ability. The genotype G16 with higher seed yield than the total mean were the most stable genotypes, while the genotypes G17 and G18 with the highest contribution to GE interaction were the most unstable genotypes. Wricke’s ecovalence stability parameter (W2i) showed that the genotypes G16, G12, G5 and G4 were the most stable genotypes.
The results indicated that AMMI model and their biplots was an appropriate method for simultaneous selection of performance and stability of cultivars and lines. Also, according to all of methods, genotype G16 was selected as a stable and high genotype across all environments. Finally, it can be considered as a favorite promising line compared to the check cultivar Hana and as a candidate in the temperate climate.