Effect of Genotype by Environment Interaction (GEI), Correlation, and GGE Biplot analysis for high concentration of grain Iron and Zinc biofortified lentils and their agronomic traits in multi-environment domains of Nepal

Abstract— Lentil (Lens culinaris Medik. culinaris ) is a cool season food legume contains the high quality of proteins and minerals. Selecting genotypes for high mean yield and yield stability has been a challenge for lentil breeders. The complexities of genotype × environment interaction (GEI) make selection difficult to identify the best performing and most stable genotypes.  Therefore, this study was carried out to apply a GGE biplot and AMMI analysis model to evaluate the magnitude of the effect of GE interaction on grain yield of 25 lentil accessions at three environments during the year of 2016 and 2017 seasons in alpha-lattice design (5x5) with three locations and to evaluate relationships between test environments for identification of favorable genotypes for lentil production areas. Combined pooled mean analysis of variance for grain yield tested at three environments over the two subsequent years 2016 and 2017 showed that highly significant differences in genotypes, environment and G x E interaction effect indicating the possibility of selection for stable accessions. The stability of the assessed genotypes using some stability statistics derived from three types of statistical concepts (variance and regression analyses), AMMI (additive main effect and multiplicative interaction) analysis and GGE biplot (genotype main effects and genotype-by-environment interaction effects) models were applied to obtain good understanding of the interrelationship and overlapping among the used stability statistics. Research results showed that lentil accession WBL-77 (1451 kg ha-1) , RL-79(1446 kg ha-1) and PL-4(1429 kg ha-1)  were the best performer and well adopted across the environments and over the years. AMMI analysis of variance for lentil grain yield (tha^-1) of lentil accessions tested at three environments over the years showed that 80.71% of the total sum of squares was attributed to environmental effects, only 8.38 % to genotypic effects and 10.90% to genotype × environment interaction effects. The partitioning of GGE sum of squares through the GGE biplot analysis showed that PC1and PC2 accounted 74.75%, and 25.24% of GGE sum of squares respectively over the years. Accessions ILL8006, RL-6, Shital, ILL3490 and simal were more close to the center point and indicated that stable across the environments. In another words, the genotypes which have low stability value (ASV) is said to be stable and the breeder chose the stable genotypes along with grain yield above the mean grand yield. In this experiment accessions RL-6(G-2) ranked 1st stability (ASV-0.53) followed by Simal (ASV-2.05), ILL-3490 (ASV-2.42) and Shital (ASV-2.72) and suitable for all environment.