Stability of the expression of the maize productivity parameters by AMMI models and GGE-biplot analysis

The objective of this study was to estimate genotype by locality, by year, by treatments (GxLxYxT) interaction using AMMI model, to identify maize genotypes with stable number of rows of grains performance in different growing seasons. The trials conducted with seven maize lines/genotypes, four treatments, two years and at the two locations. The results showed that the influence of genotype (G), year (Y), locality (L), and GxL, GxT, GxLxT, GxYxT, GxYxLxT interaction on maize number of rows of grains were significant (p lt 0.01). The genotype share in the total phenotypic variance for the grains number rows of was 53.50%, and the interaction was 21.15%. The results also show that the sums of the squares of the first and second major components (PC1 and PC2) constitute 100% of the sum of the squares of the interaction GxL. The first PC1 axis belongs to all 100%, which points to the significance of the genotype in the total variation and significance of the genotype for overall interaction with other observed sources of variability. The highest stability in terms of expression of the grains number of rows had the genotype L-6, followed by the genotypes L-4, L-5 and L-3. The lowest stability was demonstrated by the genotypes L-2 and L-1, which confirmed that these genotypes are not important for further selection in terms of this trait