Interpreting the effects of genotype x environment interaction on cane and sugar yields in sugarcane based on the AMMI model

Field experiments were conducted in six environments (2 water regimes x 3 crops) during 2005–07 to study the effect of genotype x environment interaction (GEI) on cane and sugar yields in sugarcane (Saccharum spp.). Analysis of variance of 11 genotypes revealed that genotype, environment, and GEI were highly significant. Highly significant GEI effects indicated the necessity for testing newly developed genotypes under different environments. The additive main effects and multiplicative interaction (AMMI) model was used to interpret the GEI. The first two IPCAs of the AMMI analysis of GEI accounted for 90.39% of this variation for cane yield and 92.54% for sugar yield. AMMI-1 biplot showed that genotype LG 03001 had very good cane yield with specific adaptation to the 1st year and 2nd year plant crop under optimum condition. For sugar yield, LG 03001 and LG 03002 were better adapted to the 2nd year plant crop under normal and moisture-deficit conditions, respectively, Considering only the IPCA-1 scores, BO 91 (average yielder) and LG 03007 (poor yielder) were the most stable for cane yield, whereas BO 91 was more stable in terms of sugar yield. AMMI-2 biplot indicated that LG 03003 for cane yield and LG 03007 and BO 91 for sugar yield had small interaction with environments, reflecting their stability in yield performance across environments.