In resistance tests to Fusarium head blight (FHB), the mixing of inocula before inoculation is normal, but no information about the background of mixing was given. Therefore, four experiments (2013-2015) were made with four independent isolates, their all-possible (11) mixtures and a control. Four cultivars with differing FHB resistance were used. Disease index (DI), Fusarium damaged kernels (FDK) and deoxynivalenol (DON) were evaluated. The isolates used were not stable in aggressiveness. Their mixtures did not also give a stable aggressiveness; it depended on the composition of mix. The three traits diverged in their responses. After the mixing, the aggressiveness was always less than that of the most pathogenic component was. However, in most cases it was significantly higher than the arithmetical mean of the participating isolates. A mixture was not better than a single isolate was. The prediction of the aggressiveness level is problematic even if the aggressiveness of the components was tested. Resistance expression is different in the mixing variants and in the three traits tested. Of them, DON is the most sensitive. More reliable resistance and toxin data can be received when instead of one more independent isolates are used. This is important when highly correct data are needed (genetic research or cultivar registration).
Keywords: deoxynivalenol (DON); disease index (DI); fusarium damaged kernels (FDK); host-pathogen relations; phenotyping FHB.