**1. Introduction**

Mixing of isolates is a general methodical procedure used to produce inoculum for artificial inoculation. In most cases, no reason is given as to why it is used. It is known that the isolates of the *Fusarium* spp. have a strong variability in aggressiveness [1–3]. As mixing in seedling tests strongly influences aggressiveness [1], it is important to know what the influence of mixing on the disease-causing capacity is. It is clear now that *Fusarium graminearum* and *Fusarium culmorum* do not have vertical races and the resistance is race-non-specific [4–7]. Another important feature is that the resistance is also species-non-specific [8,9], meaning that the same quantitative traits locus (QTL) gives protection against all the *Fusarium* species tested. Highly significant differences were detected in aggressiveness within the *F. graminearum* and *F. culmorum* populations [4,10–12]. In addition, the aggressiveness does not seem to be stable [13], as proven by the many significant isolate/year interactions [14,15].

In this paper, and our previous publications, we used the aggressiveness term for the disease-causing capacity of the given inocula, as virulence is taken for the race-specific pathogens like rusts. The term pathogenicity is referred to the disease-causing capacity of the genus itself [4].

Table 1 shows a cross section of the literature working with mixtures. Research task, plant, media for increasing inoculum, conidium concentration, number of participating isolates in the inoculum and the data of visual symptoms, Fusarium damaged kernels (FDK) and deoxynivalenol (DON) were followed.





= culmorum, Fg + Fc: mixture of the two species. No. of isolates: *v*/*v* volume/volume, when 4, each have 25% in the pooled inoculum, nFg, nFc, = mixture of the two species without giving the number of the isolates, therefor "n" before, Conidium concentration: 1 × 104, 1 × 105: = two different concentrations were used in the same paper, FHB (Fusarium head blight) Visual: \* = 20 days after inoculations, I = Incidence, S = severity, VSS 1–9 = visual scale 1–9, DI = disease index, DAI = Days after inoculation.: AUDPC: area under disease progress curve, FDK = Fusarium damaged kernel, DON: deoxynivalenol.

Numerous authors used spray inoculation (*n* = 22) [18,20,22,24,28,31,39,46,62,63], and point inoculation was used in 20 cases [16,23,40,49,53]. The Chinese authors work mostly with point inoculation [21,26,27,29,37,41,45]. Many American sources also use this [5,25,26], partly with Chinese scientists working in the US, or from US–China collaboration. However, in increasing numbers, spray inoculations. In some cases, papers are found where both inoculation methods are used parallel [16,24,25,37,40]. Mixtures are made mostly from di fferent isolates of the same *Fusarium* species; in several cases, the di fferent chemotypes are mixed. However, without mixing, no tests were made, so nothing can be said about the e ffect of the mixing. We have an example that the inoculation was made separately with *F. graminearum* and *F. culmorum*, and then data were pooled for ANOVA [24]. The number of isolates in the mix varied from 2 to 39. In the eight cases, the participating inocula were adjusted before mixing to the given concentration, and then in three inocula, one-third of the amount was pooled to secure the same rate of the given inocula in the pooled version. For the others, we do not have such information, and in several cases no isolate number was given; this case is marked with "more" in the column no. of isolates in Table 1. Aggressiveness before the test was made only in one case [40]; for others, no test was performed. In several cases, the selection of the isolates was made based on experience of earlier years. The conidium concentration is very variable from 10,000 to one million. There is no explanation for this. This means that besides the mixing, the adjusting conidium concentration can also cause problems. There are two conclusions. There is no control of aggressiveness from side of the mixing and diluting. Therefore, only after finishing the test will be clear, whether the necessary aggressiveness could have been secured to achieve the necessary reliability of the experiment. The fifty papers were listed, but in four papers, two lines were used as the authors have applied di fferent inoculation methods or di fferent *Fusarium* spp. Thus, the total number of the cases is 54. The aggressiveness level was evaluated by the presented visual data in this paper. Nineteen cases were found in the high to —very high aggressiveness group, 18 were classified medium or medium/high, eight had low or medium/low level, and in ten cases, no data were printed (not tested or not given). From the 54 cases, only 17 proved good and acceptable, the others were of lower level with moderate di fferentiation power or even less. This shows, clearly, that securing the necessary aggressiveness could be secured at 36% of the cases. In many cases, disease index was found; in other cases, severity was mentioned, but looking at the data average, severity was indicated, so these were also considered as disease index. In older literature, this was normal. FDK severity was tested only in eleven cases; five cases were high, three cases low, and two medium severity. In 37 cases, we have no data. DON was measured in 17 cases, five cases had high numbers, two were medium, and 10 were low or low/medium and medium/low qualifications. It is important that, in several cases, high aggressiveness in visual symptoms resulted in low DON yield in grains [27,34,42,50]. However, in one case, one poor visual rate showed high DON contamination [39]. The data show that the response to visual symptoms, FDK and DON is not the same. The most important task is the reduction of the DON contamination. The problem is that the least research is done in this field, and only in five cases were the data suitable to analyze DON response; this is less than 10% of the cases.

In the cited literature, the number of isolates in the mixtures varied between 2 and 39. The conidium concentration was set to between 5500 and 5 × 106. This leads to the following question: is mixing and adjusting isolates not significant, or does it have a significant influence on inoculation results? From the papers, we did not ge<sup>t</sup> any information. The fact that everybody worked with the best thought conidium concentration and mixing—the published results do not support this probable conviction. However, we thought that the questions should be answered. Therefore, one should know what really happens when di fferent isolates are mixed. After the test, we will know more, how the mixing is working and whether the aggressiveness of the composite inoculum could be.

An important thing should also be considered. Suppose that the aggressiveness problem can be solved for the one inoculum used normally (single isolate or mixture); the question remains whether the single inoculum can provide the reliability of the testing needed for scientific purposes in genetic analyses, variety registration trials, etc. Snijders [64,65] applied four *F. culmorum* isolates from Research Institute for Plant Protection IPO-DLO, Wageningen, NL (IPO 39–01, IPO 329–01, IPO 348–01 and IPO 436–01). Ranking of isolates and the height of the infection were di fferent and variety responses showed high variability. Further results also showed significant isolate-year interactions [66–71], e.g., changing ranks in di fferent years. Besides the changing isolate ranking, the variety ranking di ffered, that also can be a problem in resistance classification. It seems [1] that the more aggressive isolates keep their aggressiveness much better (following dilution) than the less aggressive ones. It is supposed that the mixtures may have a similar picture.

Therefore, this study focused on three main objectives. First, making inoculations with four *Fusarium* isolates in every possible combination to observe the range of plant reactions as widely as possible. Second, to gain more reliable information about the response of cultivars with di ffering resistance levels, and the structure of resistance expression in order to understand the behavior of the isolates and their mixtures, depending on their aggressiveness level, and to study FDK and DON responses. Here, the changing variety ranks are especially important. Third, as the di fferent traits (FHB, FDK and DON) often do not respond the same way, obtaining more information that would promote regulation of these traits at di fferent aggressiveness levels.
