Validation and demonstration of forecasting models for Fusarium head blight development in wheat under production conditions of Quebec

Project Code PRR11-010 

Project Lead

Salah Zoghlami  Federation of Commercial Field Crop Growers of Quebec


To assess and validate various existing forecasting models for disease risk in wheat and recommend best model adapted for use in Quebec

Summary of results

Fusarium head blight (FHB), caused mainly by Fusarium graminearum, is the most destructive disease affecting wheat production across Canada. The disease reduces the yield and grade of wheat grains. This disease-causing fungus also contaminates kernels with mycotoxins such as deoxynivalenol (DON), making the grains unfit for human and animal consumption. Because there were no completely FHB-resistant wheat varieties available commercially, FHB control has relied mainly on fungicide use and cultural methods, including crop rotation and the use of cultivars that are less susceptible to the disease.

However, Quebec producers do not have access to a reliable model to determine the proper timing for effective fungicide applications. The goal of this project was to identify the most accurate model for forecasting disease risk and providing spray warnings for FHB control under Quebec conditions.


Several experimental trials and demonstrations were carried out in 2011 and 2012 at four participating research stations (L’Acadie, Saint-Mathieu-de-Beloeil, Saint-Augustin-de-Desmaures and Normandin) and in commercial wheat fields (31 in 2011 and 15 in 2012) across all three agroclimatic zones in the province of Quebec. The geographical distribution of the sites made the differences between the regions visible.

The performance of the following nine forecasting models (including variants of some models) were tested with weather data from experimental stations: DONcast (Canada), Hooker et al. (Canada), De Wolf A, B and I (United States), Molineros (United States), Moschini (Argentina), and Rossi “Infection” and “Toxins” (Italy). The forecasting models were selected based on their availability in the literature and on the climate of the different regions these represented. The models were incorporated into the CIPRA (Computer Centre for Agricultural Pest Forecasting) software, developed by Agriculture and Agri-Food Canada, so that they could be used effectively in the trials.

Two spring wheat cultivars (the tolerant AC Barrie and the moderately susceptible Torka) and one winter wheat cultivar (Warthog) were tested in the trials. Only one application of Prosaro fungicide was made at most of the sites.

At the experimental sites, the interactions between a number of parameters (weather, seeding date, cultivar, phenological stage, fungicide application, et cetera) were assessed to determine their influence on FHB severity and incidence in wheat. Spraying programs established according to the forecasting models were compared with conventional spraying programs for their effectiveness in terms of disease suppression, Fusarium-damaged kernels (FDKs), grain yield and DON content in grains at harvest. In the commercial fields, trials were carried out to compare fields that received fungicide treatment with fields that did not.

This project was a partnership between many collaborators, including provincial specialists in cereal crops and cereal diseases, researchers at the Centre de recherche sur les Grains inc. (CÉROM), Laval University and Agriculture and Agri-Food Canada, wheat producers, and extension specialists. Extension activities aimed at technology transfer and educating wheat producers about these tools were carried out during the project.


Fusarium head blight pressure was higher in 2011 (up to 10.4% infected spikelets, 20.0% FDKs and 2.51 parts per million (ppm) of DON) than in 2012 (up to 6.7% infected spikelets, 1.8% FDKs and 1.34 ppm of DON). At most sites in 2011, one fungicide application reduced the percentage of infected spikelets (less than 46%), percentage of FDKs (less than 22%) and DON content in the grains (less than 49%) and also increased yields.

Among the nine assessed forecasting models for FHB development in wheat, the De Wolf B model provided the best forecasts and was selected for use under Quebec climatic conditions. That model demonstrated a superior level of effectiveness and provided the best decision threshold (greater than and equal to 90% of forecasts were adequate). In 2015, the De Wolf B model will be assessed in operational mode, using the CIPRA software, in several Quebec regions by the team of specialists at the Réseau d’avertissements phytosanitaires (RAP), the plant protection warning network. In the fall of 2015, the model will be incorporated into the Agrométéo Québec system (, to ensure that the province of Quebec is covered more comprehensively by the RAP’s specialists. The information generated by this tool will enable wheat producers to effectively determine the need and best time for plant protection responses.

In conclusion, FHB severity varies from year to year, and this study showed the importance of phenological observations in the field given that the developmental period during which wheat is susceptible to this disease is short. The use of bioclimatic models which can predict the dates when the crop reaches specific phenological stages based on seeding date are recommended to also be considered, along with field visits.

For more information on this project, contact Salah Zoghlami.

Date modified: