Tailoring fire blight predictions to your site: The Fire Blight Interactive Predictor

A new feature for predicting fire blight disease of apples blossoms has been added to the Michigan on-line weather system Enviroweather (www.enviroweather.msu.edu). The new tool, dubbed the Fire Blight Interactive Predictor, allows the user to modify temperature, rainfall, bloom dates, and other factors to more closely match conditions on your farm. We will explain the features of the Fire Blight Interactive Predictor on Enviroweather and how we think this can be useful to growers, consultants and others involved with fire blight.

The underlying model for blossom blight in the Fire Blight Interactive Predictor model is the same as is used in the Fire Blight Assist Chart that has been on Enviroweather for several years. The Fire Blight Interactive Predictor version provides more versatility and allows the user to gain a better understanding of conditions that favor fire blight epidemics.

The Fire Blight Interactive Predictor is based on Maryblyt, a popular computer program developed by Dr. Paul Steiner and Mr. Gary Lightner, Department of Botany at the University of Maryland. The output for the Fire Blight Interactive Predictor version closely resembles the original Maryblyt model by Steiner. The basic model predicts blossom infection by fire blight on a day when four conditions are met:

  •  Open blossoms (B).
  • Sufficient warm temperatures over the past several days to increase fire blight bacterial growth on blossoms (H).
  • A wetting event such as rain, dew, fog or spraying (W).
  • Average daily temperature of at least 60°F (T).

The fire blight risk for each day during bloom is given as low, moderate, high, or infection when one, two, three, or four of these conditions are met (Figure 1). For many users, this will be all they need to know to make spray decisions. Bactericide applications are definitely needed for predicted infections and possibly for high fire blight risk ratings. The Fire Blight Interactive Predictor keeps track of multiple infection events and calculates when symptoms are expected to show in the orchard. The Predictor is useful for summarizing fire blight conditions for past seasons, which helps in comparing years and unraveling reasons for disease outbreaks.

A feature in the module that will be prized by advanced users is the ability to temporarily change rainfall, temperature, bloom start and end dates, bactericide spray, and trauma events like hail or high winds to better match local conditions or to try “what if” scenarios.

For example, if the user suspects that the maximum temperature may or has been higher than predicted or recorded for a nearby Michigan State University MAWN weather station, the higher temperature can be substituted to see the resultant prediction. The Fire Blight Interactive Preictor Predictor can be used to work backwards to discover when the infection took place that was responsible for unexpected symptoms showing up in the field.

In some respects the Fire Blight Interactive Predictor is a step backwards closer to the original Maryblyt model by Steiner and Lightner still used by some growers and consultants in its original PC DOS format, the only version released and no longer available. The Predictor uses the essence of Maryblyt model in the MSU Enviroweather network to deliver real-time fire blight predictions to busy people.

Did you find this article useful?