Populations of Aspergillus flavus are highly variable. We have shown this variability to be reflected both in the ability of isolates from a single toxigenicities of populations from different fields. This variability provides a resource for selecting A. flavus isolates with many potentially beneficial characteristics. We have selected isolates with both competitive ability under field conditions and reduced toxigenicity. These isolates have been utilized in our continued efforts to develop techniques to modify A. flavus population of strains that do not produce aflatoxins and thereby reduce aflatoxin contamination of crops.

In two years of field experiments, in different fields, we have successfully displaced endemic toxigenic strains of A. flavus with and applied atoxigenic strain during infection of the developing cotton crop. Displacement was achieved by applying wheat seed colonized by and atoxigenic strain of A. flavus to soil underneath the crop canopy prior to first bloom. Displacement was associated with a significant reduction in aflatoxin contamination in both years without an increase in boll infection as measured by bright-green-yellow-fluorescence. Furthermore, no increase in the number of A. flavus propagules on the mature crop was associated with the displacement. In both years the concentration of toxin in the infected cottonseed was inversely correlated with the incidence of the released atoxigenic strain. Thus, replicate plots with very low toxin and those with low incidences had high toxin.

The results of our field test to date suggest that A. flavus populations infecting cottonseed can be actively modified to reduce their toxigenicity. Population modification may be a stable long term method to reduce aflatoxin contamination of cottonseed to acceptable levels.