I conducted a multilocation trial on coffee in three locations for two years (six environments) to see if there a GEI. The result showed GEI but when I tried to classify the environment using GGE analysis. the two years of one location grouped under one mega environment so how can justify it?