Models for predicting Fusarium head blight (FHB) and deoxynivalenol (DON) content in wheat provide farmers with a tool for preventing yield loss and mycotoxin contamination. Due to the complex nature of FHB, these models typically take only a subset of explanatory variables as inputs, such as weather conditions during anthesis and, to a lesser extent, field-specific variables and crop characteristics. We present a thorough statistical analysis of a broad spectrum of variables that affect FHB and DON content, using an unique database of field observations covering nine years. This study unveils new variables that play an important role in FHB incidence and DON content. Evidence is brought forward demonstrating the effect of species interactions on DON content and the effect of weather conditions during the vegetative growth stage. Indeed, temperature from November till May was positively correlated with FHB and DON content, whereas relative humidity was negatively correlated with DON content in November and December. In a broader view of the divergent results obtained regarding FHB and DON content in wheat crops, this study underscores that the prediction of FHB and DON content should be seen as two distinct goals.
