The combined analyses of variance across years detected significant differences among genotypes for Fusarium ear rot and fumonisin content (Table 1). The year x genotype interaction was
significant for fumonisin content and was due to changes in genotype ranking rather than magnitude changes across years. For this reason, mean comparison among genotypes for fumonisin content was kept separate for each year, but comparisons for ear rot symptoms were combined across years. Since fumonisin quantification requires expensive chemical analyses and this data indicated that these analyses should be performed in several environments due to the large environment x genotype interaction, it would be advantageous if a simpler less costly trait that was genetically correlated to fumonisin content could be used as an indirect selection criterion
for increasing resistance to fumonisin accumulation.