Development of models to predict fusarium head blight disease and deoxynivalenol in wheat, and genetic causes for chemotype diversity and shifting of Fusarium graminearum in Manitoba.

摘要

Fusarium head blight (FHB) is one of the most important diseases in wheat. It causes wheat yield loss and affects the quality of wheat. Disease prediction models will help wheat producers with making right decision on fungicide application.;FHB is important not only because it causes yield loss but also its agents produce mycotoxins. Deoxynivalenol (DON) is one of the most important mycotoxins produced by fusarium strains. The 3-acetyldeoxynivalenol (3ADON) chemotype of F. graminearum produces more DON than the 15 ADON chemotype and is replacing the 15ADON chemotype in Canada. These findings cause increasing concern within the wheat industry and will potentially affect the prediction accuracy of the FHB and DON models. Therefore, it is necessary to determine distribution and shifting of F. graminearum chemotypes in Manitoba.;Chapter 4 of this study revealed that the 15ADON chemotype of F. graminearum was predominant in Manitoba. The 3ADON chemotype was predominant in the southeast part of Manitoba, which included Sanford, Morris and Horndean; and shared the same percentage with the 15ADON chemotype in Cartier and Portage la Prairie. The variation of chemotype likely resulted from genetic diversity of F. graminearum populations in Manitoba, which was associated with sexual recombination, age of populations and tillage system. Wheat seed shipment and long-distance F. graminearum spore dispersal likely contributed to the gene flow in this province.;This study revealed the important roles of cumulative rainfall during different periods from seeding to wheat anthesis and in the 14 days after anthesis, and average daily temperature from the first week before to the first week after anthesis in fusarium airborne inoculum level, fusarium head blight (FHB) disease and deoxynivalenol (DON) levels. Cropping practices, which were quantified and expressed as cropping practice index (CPI), significantly affected the number of Fusarium graminearum/Gibberella zeae spores, FHB disease index and DON level in the mature kernels. The numbers of F. graminearum spores were strongly correlated with the disease index and toxin levels. Four models (Chapter 2) were developed to predict the number of F. graminearum spores on single wheat heads using CPI and weather conditions. Two types of prediction models for FHB disease index and DON level were developed based on cropping practices, actual or predicted spore counts on single wheat heads, and weather conditions. Type I models were developed using actual spore number, and Type II models were developed based on predicted spore number using the predicted model developed in Chapter 2. The average prediction accuracy was 85% for Type I models and 58% for Type II models. Type I models can be used for the purpose of market prediction by the Canadian Wheat Board. Type II models can be used with the spore prediction model by wheat producers.