Genotype × Environment Interaction and Stability Analysis of Grain Yield of Emmer Wheat (Triticum dicoccum) Landraces Using AMMI and GGE Biplot Analyses

Abstract

Emmer wheat (Triticum dicoccum Schrank) is an important traditional cereal crop in Ethiopia valued for its genetic diversity and adaptation to marginal environments. However, its productivity is constrained by strong genotype × environment interaction (GEI), limiting the identification of stable and high-yielding genotypes adapted to diverse agroecological conditions. This study aimed to evaluate GEI and identify both broadly and specifically adapted emmer wheat landraces using AMMI and GGE biplot analyses. A total of 270 emmer wheat landraces were evaluated, comprising 260 landraces obtained from the Ethiopian Biodiversity Institute and 10 landraces collected from emmer wheat-growing areas of the Tigray Region, during the 2023/24 and 2024/25 main rainy seasons using an alpha-lattice design with two replications. The trials were conducted across three locations in the first year, with two of these locations repeated in the second year, resulting in five test environments (location–year combinations). Combined analysis of variance revealed highly significant differences among genotypes and significant GEI (p < 0.001). Environmental effects contributed the largest proportion of the total variation (43.9%), followed by GEI (31.1%) and genotype (16.1%), demonstrating the strong influence of environmental conditions on grain yield. AMMI and GGE biplot analyses effectively characterized the GEI pattern. The first two AMMI interaction principal components explained 89.7% of the GEI variation, while the GGE biplot explained 79.32% of the total GGE variation and identified two distinct mega-environments. Genotype G207 exhibited specific adaptation to mega-environment II (E4-E5), whereas G158 was specifically adapted to mega-environment I (E1–E3). The discriminativeness versus representativeness biplot identified E2 as the most suitable test environment for selecting broadly adapted genotypes. Among the evaluated landraces, G59 showed the highest yield stability and potential for broad adaptation across environments, whereas G266, G196, and G56 exhibited relatively low stability. GGE analysis identified genotypes associated with specific mega-environments, indicating potential specific adaptation. Overall, the study revealed substantial genetic variability and strong GEI affecting grain yield in Ethiopian emmer wheat. Accordingly, G207 and G158 are recommended for further evaluation in their respective mega-environments, while G59 is recommended for wider evaluation as a promising, broadly adapted genotype for breeding programs aimed at improving yield stability and adaptation. Further multilocation and multiyear evaluation is recommended before release.