Fusarium head blight risk assessment models and DON prediction model

Editor’s note: This article is from the archives of the MSU Crop Advisory Team Alerts. Check the label of any pesticide referenced to ensure your use is included.

What tools are available to estimate the risk of Fusarium head blight (scab) and levels of DON (vomitoxin) and help growers determine whether fungicide applications are advisable to protect wheat against head scab?

Two models have been developed to serve as decision aids for growers to determine whether an application of fungicide should be made to control Fusarium head blight (FHB). The Penn State Model and the Ontario DONcast model are described in this article. Through research funded by Project GREEEN in 2006-2007, both the Penn State and DONcast models were validated in Michigan using weather information and collecting field data from grower locations and research plots. Although epidemic conditions for FHB did not occur during either year of the study, the predictions made by each model showed reasonable agreement with field sample observations collected from grower locations and research plots during both years of the project. Although predictions for other categories of susceptibility seemed to be fairly accurate, there was a tendency to overpredict DON levels in wheat varieties that were classified as highly susceptible (HS) using the DONcast model. Field evaluations for both models will continue in Michigan in 2008.

How does the Penn State model work?

Scientists have developed scab risk assessment models through projects with the U.S. Wheat and Barley Scab Initiative. The Penn State models were developed using data from wheat head blight epidemics around the country, and represent seven years of development and testing. Although models have been developed for both winter and spring wheat, the winter wheat model will be most useful for the majority of Michigan growers, as only a very small percentage of acres in Michigan are spring wheat. Data for the Penn State models (in Michigan) are generated from MAWN and National Weather Service locations. The addition of the Enviro-weather stations in 2007 increased the number of weather locations in the Lower Peninsula for the model from 45 to 89.

The Penn State winter wheat model uses only data from the week before flowering to estimate the risk of a scab epidemic with 10 percent or greater field severity. Risk is calculated based on the number of hours when both relative humidity is greater than 90 percent and temperatures are between 48°F and 86°F during the week before flowering. Humidity and temperatures in this range are favorable for the reproduction of the fungus (Fusarium graminearum) that causes head scab. The spores must germinate and infect the flowers in order for the pathogen to cause disease. Moisture, in the form of high humidity, rainfall, dew or ground fog has the greatest influence on infection. Generally, warm temperatures (75-86°F) and extended periods of wetness (overnight and morning) tend to be most favorable for infection. Cool temperatures tend to limit fungal growth, but long periods of wetness can compensate for cooler temperatures.  

The models can correctly identify years with severe head blight 7 out of 10 years. The color codes on the map indicate the relative risk of an epidemic. Red = High risk; Yellow = Moderate risk; Green = Low risk. The circles and triangles on the map represent weather stations that can be selected to provide site-specific estimates of disease risk. The models and evaluation of disease risk should begin about 1 to 2 weeks prior to anthesis (flowering). Many users have found that the models will produce patterns of persistent moderate disease risk, or steadily increasing disease risk during this time period.

The final decision to apply a fungicide should be made as close to flowering as possible. You should monitor local weather conditions during the week before flowering to help you decide whether infection is likely to occur in your area. Michigan specific comments will be posted to the website again this year.

How does the DONcast model work?

The DONcast model, developed at the University of Guelph, (Ridgetown, ONT, CA) makes a prediction at heading for the expected level of deoxynivalenol (DON, vomitoxin) at harvest. There are a number of components that make up the DONcast model. Environmental effects such as temperature, relative humidity and rainfall account for 48 percent of the model. The level of susceptibility of the wheat variety-moderately resistant (MR), moderately susceptible (MS), susceptible (S) or highly susceptible, (HS) account for 27 percent of the model. Whether the previous crop was a host for the FHB fungus such as corn, wheat, barley or something else represents 14 percent of the model, and tillage accounts for less than 5 percent. Using these inputs, DON predictions are made for a specific heading date, corresponding to when 75 percent of the wheat heads have emerged. This model allows for a 2-3 day window after heading to make fungicide applications if needed, depending on predicted DON values. The threshold for recommending a fungicide application for this model is if predicted DON levels equal or exceed 1 ppm. Although there is currently no Michigan based DONcast, growers who are interested in the Ontario DONcast model can access it at the Weather Innovations website, http://www.weatherinnovations.com/DONcast.cfm

When is wheat most vulnerable to FHB infection?

Before the head emerges from the boot, it is protected from Fusarium. The head is most vulnerable to infection from the time that flowering begins to the early milk stage. After mid-milk, it becomes more difficult for the fungus to infect the kernel and damage is less severe. The most severe head blight epidemics have involved multiple infection periods and cultivars susceptible to spread of the fungus throughout the head. Depending on the fungicide used, the window for management of head blight extends from heading to 50 percent flowering, with the optimum period of fungicide application being at Feekes 10.5.1 (about 15-25 percent of heads flowering).

How do I decide whether I should apply fungicides for FHB to my wheat?

Fusarium head blight has the potential to cause significant losses in both quality and yield. Discounts tend to start at lower ppm DON levels in white wheats than in red wheats. When and if a fungicide should be applied for FHB depends on the level of risk the grower is willing to accept. Although considerable effort is ongoing to develop FHB resistant varieties of wheat, the use of fungicides to control FHB and reduce levels of deoxynivalenol (DON) is still a primary management strategy. Timing of the fungicide application is especially challenging. Growers have an application window of only a few days for effective control of FHB and DON. Unlike some of the common foliar diseases of wheat, by the time scab symptoms are visible, the window for fungicide treatment is already closed. Models are meant to serve as decision aids to help growers decide whether or not an application of fungicide to control FHB is warranted.

The April 17, 2008 Field CAT Alert contained a partial budget analysis to help you determine whether an application of fungicide was likely to be economical. A table of fungicides currently registered in Michigan for FHB and their efficacy ratings can be found at the Field Crops AoE website. If additional fungicides become registered for FHB in Michigan, we will announce them in the Field CAT Alert.

Related Events

Related Articles

Related Resources