Technology | Fall 2008
Wheat midge forecasts gain accuracy.
By Jennifer Barber
With help from on-farm weather data, researchers are improving their ability to predict wheat midge emergence dates
As forecasters predicted, the wheat midge outbreak on the prairies this past growing season was one of the worst in recent memory. Based on information from on-farm weather stations and Environment Canada, emergence maps and commentary developed by Canadian Wheat Board agronomists were instrumental in helping farmers deal with the pest. Researchers are continuing to work with the CWB to develop increasingly accurate wheat midge forecasts.
“We used to predict the timing of an outbreak based on calendar days,” says Dr. Bob Elliot, a research scientist with Agriculture and Agri-Food Canada. “Using calendar days, we can predict emergence of adult wheat midge with an accuracy of 5-6 days.” In the early 1990s, that model was refined to one that used heat units, or degree-days, to predict emergence. Using heat units above 5°C that accumulate after March 1, emergence can be predicted with an accuracy of 2-3 days.
In fall 2007, forecasters sounded the alarm about a possible outbreak during the 2008 growing season. Surveys in Alberta and Saskatchewan showed a higher density of wheat midge cocoons than had been seen in more than 10 years. Counts far exceeded economic thresholds in these provinces.
This year, temperature information was gathered from Environment Canada stations and the network of prairie WeatherBug stations located on farms, and at grain handling facilities and ag retailers. Forecasters used a method to predict emergence that calculated heat days using minimum/maximum temperatures above 5°C starting April 1. This method was accurate most of the time, but researchers such as Elliot saw an opportunity to further increase precision.
“We are looking to refine our measurement of heat units by using what is called a sine wave calculation,” says Elliot. “With this method, we can start counting degree-days above 5°C after March 1. Looking back at the last growing season, this would have given us more heat units, increasing our accuracy.”
Elliot is currently fine-tuning the sine wave model to predict emergence. Spraying for wheat midge was fairly intensive this year. Cocoon count results for this fall will indicate whether midge levels are likely to be as high again in 2009.
“This year we had a really high population and we don't know for sure if the perception of the peak of the outbreak had more to do with the amount of wheat midge out there, or with the method of calculation,” says Elliot. “Our revised model worked very well at our test plots and so we're hopeful it will bring even better results for the coming year.”
