Aflatoxin occurrence in corn is highly erratic, due in part to the small levels at which it is measured. The erratic nature of aflatoxin contamination poses difficulty in separating more-resistant hybrids from less-resistant ones in variety field trials using traditional techniques. An alternative approach is to use regression analysis to evaluate the aflatoxin content of individual hybrids compared with the average aflatoxin content of all hybrids across multiple environments. This procedure provides a more powerful statistical test. Statistical models are used to predict expected aflatoxin contamination and provide a better picture of how individual hybrids perform across a range of environments differing in aflatoxin contamination.

Twenty-three commercial corn hybrids were grown at Alexandria, Bossier City, Jeanerette and St. Joseph in 2001 and 2002. Ten ears from each plot were inoculated with Aspergillus flavus spores after silking. The inoculated ears were harvested at maturity along with 10 non-inoculated ears. After rating for Aspergillus flavus fungal growth, harvested ears were shelled and kernels ground to a meal. Samples were then analyzed for aflatoxin. Data from two locations in 2002 were not used in hybrid analyses because of water contamination to grain stored in a warehouse caused by Hurricane Lily. Data for individual hybrids that exceeded the upper confidence interval after initial regression analyses were also discarded.

Average aflatoxin concentration for inoculated and non-inoculated ears ranged from 923 ppb to 3,728 ppb among the four locations over two years. Inoculation with Aspergillus flavus increased the overall average of aflatoxin concentration in corn grain from 443 ppb to 3,742 ppb. Bossier City had the highest aflatoxin contamination. This location normally has the highest temperatures and least rainfall. Alexandria followed Bossier City in degree of contamination. Both of these locations appear to be good and stable sites for aflatoxin biosynthesis and would be expected to serve well as test locations for screening hybrids and breeding lines.

Aflatoxin contamination of all hybrids was highly responsive to the environment. A straight-line increase of aflatoxin in each hybrid was directly proportional to the average aflatoxin content of all hybrids at each location. Aflatoxin in some hybrids followed environmental parameters more closely than others.

Graphing the interaction between hybrids and the environment provided readily visible comparisons. Using this approach, aflatoxin content of individual hybrids is depicted on the vertical axis and mean aflatoxin of all hybrids is depicted on the horizontal axis. Some hybrids had above-average aflatoxin in all environments. Others had below-average aflatoxin content in some environments but above-average aflatoxin in others.

A small group of hybrids was shown to have predictable below-average aflatoxin across all environments and was deemed to be generally superior in resistance to aflatoxin. Dekalb DK697, perhaps the best example of a resistant hybrid, is compared to the average aflatoxin content of all hybrids in Figure 1. The black line in Figure 1 is the mean aflatoxin content of the 23 hybrids. Dekalb DK697 had lower-than-average aflatoxin in all environments, making it a good choice when selecting a hybrid for resistance to aflatoxin. Other hybrids that had below-average aflatoxin predicted across all environments were Croplan Genetics 733BT, Dyna-Gro 5516 RR, Croplan Genetics 827 and Garst 8288.