Genotype × Environment Interactions effects on Grain Yield in Winter Wheat grown under rainfed conditions

Wheat (Triticum aestivum L.) is widely grown across Central and West Asia. Evaluation of wheat genotypes in different environments is essentials to estimate Genotypes x Environment interactions (Fan et al., 2007). Breeding new cultivars require evaluation of yield stability and the adaptability of high yielding genotypes across different environments. Statistical approaches such as GGE Biplot analysis reveals genotypic main effects, as well as G × E interactions. GGE biplots analysis provides a graphical depiction of relationships among environments, genotypes and their interactions in an effective manner (Yan et al., 2000). This method has been used to examine the uniformity of different environments and to identify superior genotypes in multi-environment trials. This study aimed to determine the effects of genotype by environment interactions on the performance of 22 promising wheat lines along with two check cultivars and to identify the ideal genotypes for the examined environments.

Wheat promising lines including 22 elite lines along with two commercial cultivars Azar2 and Ohadi were evaluated in a randomized complete block design (RCBD) experiment with 4 replications under rainfed dryland conditions. Uniform experiments were carried out for 3 years (2013-16) at the agricultural research stations of Maragheh, Sararod, Ghamlo, Zanjan, Ardabil, Shirvan and Arak representing cold and moderate cold dryland wheat growing regions of Iran. Entries were planted in plots with 6 x 1.2 dimension (6-m long rows, 17.5 cm apart and seeding rate of 380 kernel/m2). Morphological characteristics, phenological stages and yield components were scored during growth period. Grain yield and thousands kernel weight were determined for each plot after harvest. The collected yield data were subjected to combined ANOVA considering environment as random and genotypes as fixed effects using SAS (9.1) statistical software. Yield stability parameters were estimated and GGE biplot analysis was carried out using open-source software R packages.

The ANOVA revealed significant differences among genotypes for yield performance. Significant genotype by environment interactions were also identified. The main effect for environments and genotypes by environment interactions accounted for 82.3 and 6.5 percent of total variances in the experiment. The grand total mean grain yield for the evaluated wheat genotypes was 2165 kg/ha. Genotypes G21 and G23 produced the highest (2378 kg/ha) and the lowest (1747 kg/ha) yields, respectively. Genotype G20 was found as the most stable according to the estimated stability parameters (Table.3).Based on polygonal GGE biplot, two environmental groups and five different groups for the examined genotypes were detected (Fig. 1). The first environmental group included Maragheh (M92, M93), Ghamlo (G92, G93, G94), Ardabil (A93), Zanjan (Z92, Z93), Arak (Ak92, Ak93, Ak94) and Shirvan (Sh93, Sh94), with the G21 being as the best performer in the group. The second environmental group included 8 remaining environments with the G1 (Azar2) as the highest yielding genotype in the group (Fig. 1). Simultaneous evaluation of genotypes for yield and yield stability identified genotypes G20, G13 G2 and G6 as high yielding with high yield stability (Fig. 2).An environment is considered ideal tester for evaluating a set of genotypes if it is able to effectively differentiate genotypes based on traits of interest and if it properly represents other environments (Yan & Tinker, 2006). Our findings revealed environments Z92, Sh92 and Ak92 as the most ideal environments and environments A92, S93, Ak92, M92, M94 as the highly discriminative of the examined genotypes (Fig.3). The genotype G20 with higher average yield and the least contribution into GxE interactions showed the characteristics of an ideal genotype (Fig. 4).

Analysis of yield data revealed that, the variations due to G × E interactions were largely explained by the environment effects. In this study, cold environments were grouped closely, showing a similar behavior in discriminating the examined genotypes and were easily distinguished from the moderate cold environments. Acute angles among environments M92, G92, Sh93, Ak93, Sh94, G94, Ak94, Z93, A93, Ak92 and also among environments S94, Sh92, M94, Z94, S92 indicated high correlations in discriminating the studied genotypes. We found the genotype G20 as the ideal genotypes possessing high yield performance and high yield stability.