GGE biplot analysis of genotype × environment interaction and forage yield stability in grass pea (Lathyrus sativus) genotypes

Abstract

In crop breeding programs, biplot analysis is a well-known statistical method. This study aimed to survey the genotype ? environment interaction (GEI) on grass pea genotypes in Iran. The experiment was conducted in twelve environments (four separate sites: Gachsaran, Kuhdasht, Mehran, and Shirvanchardavol) over three sequential years (2017, 2018, and 2019) with sixteen grass pea genotypes. The purpose of this research was to utilize the GGE biplot as a tool to identify the superior genotypes of grass peas. The results for the combined analysis of variance, genotypes, and the GEI revealed a significant impact (p < 0.001) on forage yield. Moreover, genotype ? environment interaction responded differently under various climatic conditions. The interaction components evaluated by the biplots revealed the genotypes' predominant effect and the significant genotype ? environment interactions (GEI). The first two principal components (PCs) interpreted up to 93.11% of the total variation in the GGE model (PC1 = 53.30%, PC2 = 37.80%). GGE biplot analysis categorized the studied environments into two mega-groups for forage yield. Genotype G11 (Russia) was superior in terms of mean forage yield (5.362 t/ha). The genotypes that performed best in each environment, were genotypes G11 (Russia) and G8 (Bangladesh-I). Among these genotypes, G11 (Russia) was the highest-yielding genotype in the field. The Kohdasht site was the most discerning and representative test environment for crop yield. The selected genotypes are recommended for breeding programs aimed to improve forage yield in the tested sites or similar agroecological areas.