EVALUATION OF EXPERIMENTAL MAIZE HYBRIDS TESTED IN MULTI- LOCATION TRIALS USING AMMI AND GGE BIPLOT ANALYSES

Multi-environment yield trials are essential in estimation of genotype by environment (GE) interaction and identification of superior genotypes in the final selection cycles. The objective of this study was to evaluate stability and adaptability of grain yield of experimental maize hybrids by GGE (Genotype and Genotype by Environment Interaction) biplot and AMMI (Additive Main Effects and Multiplicative Interaction) analyses. This study comprised 19 experimental NS (Novi Sad) maize hybrids tested across twelve environments in northern Serbia in a randomized complete block design. The ANOVA (Analysis Of Variance) indicated significant effects of genotypes (G), environments (E) and their interaction (GE). On the same time, the highest percentage of variation was explained by E (77.83%) while G and GE effects together explained the rest of variation (<30%). Joint effects of genotype and interaction (G+GE) was partitioned using GGE biplot analysis where the first two components were significant, explaining 62.40% (44.34% PC1 and 18.06% PC2) of the GGE sum of squares. The two principal components in AMMI analysis were significant, explaining 53.99% (30.87% PC1 and 23.12% PC2) of interaction variation. Stability of analysed genotypes was similar in both methods. It is concluded that there is no large difference between the AMMI and GGE biplot analyses in evaluation of experimental maize hybrids in different climatic conditions and that both methods can be used equally successfully.