PROJECT RESULTS

A Producer-Friendly Economic Threshold for Control of Wheat Midge in Manitoba Wheat

Applicant: 

Dr. Robert J. Lamb

Cereal Research Centre
Agriculture and Agri-Food Canada
Winnipeg, Manitoba  R3T 2M9  Canada

Table of Contents:

• Background and Objectives
• Procedure and Project Activities
• Results and Discussion
• Conclusions

ARDI Project:

#98-060

Total Approved:

Date Approved:

Project Status:

Completed January, 2000

Background and Objectives:

Yield loss and crop downgrading resulting from wheat midge damage in Manitoba wheat can be prevented by a timely application of insecticide. The decision on whether or not to apply insecticide is currently made by producers, but without the necessary information and tools to make a reliable, cost effective decision. Producers need to know the economic threshold for wheat midge in their wheat, and they need a method for assessing whether the threshold is exceeded. The objective of this project was to develop that threshold, and an inexpensive, easy-to-use action threshold that farmers could adopt to make control decisions for wheat midge in wheat.

Procedure and Project Activities:

The project consisted of two steps: 1) defining an economic threshold for wheat midge feeding damage for the various grades of wheat; 2) developing a simple trapping method that allows farmers to anticipate in a timely way whether or not the economic threshold is exceeded and make a control decision.

The first step involved measuring the losses in weight of individual wheat seeds attacked by various numbers of wheat midge larvae. We tested plants infested artificially in the laboratory and naturally in plots and commercial fields. We also determined the relationship between visual damage to the seeds and the weight loss of seeds, to relate grading factors with yield loss and financial losses. We assessed the role of harvesting in grade protection by measuring how effectively a combine eliminates wheat midge damaged seeds. We compared the grade of rail car shipments with the actual damage that occurred to the seeds to determine how damage influences grade. Finally, we assessed the effects of wheat midge damage on germination and seedling growth to determine the potential losses for seed growers. All this information was integrated to produce an economic threshold based on the percentage of infested or damaged seeds in a field.

The second step involved designing and testing a trapping system that fits with current agronomic practices, is inexpensive to use, and allows farmers to decide in a timely way whether or not the economic threshold is exceeded in a field of wheat, i.e. an action threshold. Traps consisted of 75 mm x 125 mm yellow sticky cards, affixed to stakes. In cooperation with Manitoba Agriculture and Food and research collaborators, traps were placed in wheat fields, with the long axis of one sticky surface exposed vertically at the height of wheat heads. Ten traps were spaced 10 m apart in a row 10-15 m from the edge of the field at the time wheat heads began to emerge and collected after they were exposed for three successive nights. The number and spacing of traps were chosen arbitrarily, for the convenience of farmers. The three-night exposure at heading allowed farmers time to control the pest at the correct crop stage, if required. Two to three weeks later 20 wheat heads were collected from each field, two from each trap location. We recorded the total number of midges on 10 traps, and the number of damaged seeds and larvae in each wheat head in 1998, 1999 and 2001. We related the percentage of damaged seeds to the number of wheat midge adults caught on the traps, and determined the accuracy with which the trap catches could predict whether or not the economic threshold was exceeded.

Results and Discussion:

In the first step an economic threshold was estimated. The economic threshold is the pest level (7% infested seeds) causing a financial loss that balances the cost of controlling the pest. Another way of saying the same thing is – insecticidal control for the wheat midge should only be applied if the cost of control is less than the value of the wheat saved by the control. For wheat, the calculation of this economic threshold is complicated by the grading system. The value of the product depends not only on the total weight of harvested seed, but also on the grade. Wheat midges reduce the weight that is harvested and also the grade. Therefore, the effect of wheat midge damage on both weight loss and grade must be determined to calculate the appropriate economic threshold for each grade. This was accomplished by measuring the weights of damaged and undamaged seeds at various levels in the production chain, from heads in farmers’ fields, to harvested seeds fresh from a combine, to rail cars samples.

The weight of seeds from different parts of a wheat head varied, but seeds in all parts of a spike were infested, and so the smaller seeds that were infested tended to be more affected than the larger seeds. Most infested seeds had one to three larvae, but at least 11 larvae could mature on a single seed. Wheat plants did not compensate for wheat midge damage, and no indirect damage was detected for undamaged seeds. The distribution of weight for infested seeds was bimodal (two weight classes) with over 40% less than 8 mg when hand harvested, whereas infested seeds harvested mechanically had a unimodal distribution (one weight class) with nearly all of the most severely damaged seeds removed during harvest. A visual rating system of six damage categories was related to the weight of the seeds. The germination and early growth rate of infested seeds was reduced in comparison to uninfested seeds. Based on the weight relationships for the insect-plant interaction and the visual rating of damage, high protein #1 grade common and durum wheat and #1 grade durum wheat can tolerate 4-7% of the seeds being infested by larvae, before downgrading is likely. For other grades, the economic threshold is 10% of the seeds infested, based on yield loss. Seed growers can adopt the threshold for #1 wheat (4% infestation) to prevent downgrading, which would also reduce the effects of infestation on seed germination to an acceptable level. The economic thresholds are summarized in Table 1.

Table 1. Tolerance for seeds damaged by wheat midge (%), value of wheat ($), and yield loss (%) equal to the cost of an insecticide application, in relation to grade for common and durum (in parentheses) spring wheat. 

 ^* Protein levels of 14.5 % for common and 13 % for durum wheat.

^† Grade is reduced one level if tolerance is exceeded, Canadian Grain Commission (1998).

^‡ Five year mean (1994-1999) of wheat prices paid to producers, with values for durum wheat shown in parentheses.

^§ Based on 10 year mean of yields (1989-1998) for common (2.064 t/ha) and durum (2.050 t/ha) spring wheat in Manitoba and Saskatchewan.

In the second step, the wheat midge monitoring system was developed to provide farmers with an action threshold. The action threshold is the number of wheat midge adults on sticky traps that correspond to the economic threshold for damage. Action against the midge must be taken before the economic threshold is actually reached because by then the damage is already done. An action threshold was determined from 1998 data, by dividing the sample of fields into categories based on larval densities below or above an economic threshold of two larvae per spike or 7% infested seeds. A threshold at the top of the range of 4-7% infested seeds was adopted initially to minimize unnecessary insecticide application. These categories were associated with trap catches to define a possible action threshold. An action threshold of four or more adults on 10 sticky traps correctly categorized 34% of the fields as below the economic threshold and 33% as above the economic threshold. Thirty-three percent were incorrectly classified. The validity of this action threshold was assessed in 1999, which had relatively low, and 2001, which had relatively high, pest densities. In both years, the action threshold correctly classified fields according to the economic threshold in more than 75% of the cases.

Considering all 171 fields sampled over three years and an action threshold of four or more midges per 10 sticky traps, 75% of fields were correctly categorized as above or below an economic threshold of 7% infested seeds, 15% were categorized as not needing an insecticide application when control was warranted, and 10% were categorized as needing an insecticide application when control was not warranted. If no information on wheat midge densities was available and no farmers or all farmers had chosen to apply control, the error rates would have been 42 and 58%, respectively. If a farmer adopted a lower economic threshold, such as 4% infested seeds, and dropped the action threshold to three or more adults per 10 sticky traps, the error rate would be 27%, similar to the higher threshold. Increasing the economic threshold to 10% infested seeds had little effect on the error rate for decisions based on an action threshold of four or more adults per 10 sticky traps. Therefore, the action threshold was insensitive to the small changes in economic threshold that occur as prices of wheat or cost of control vary from year to year. These results are summarized in Table 2.

Table 2. Fields (%) categorized by an action threshold of 4 wheat midge adults on 10 sticky traps and an economic threshold of 2 larvae per spike (1998, 1999) or 7% infested seeds (2001).  The frequencies of correct control decisions are provided, based on whether both action and economic thresholds were exceeded or neither was exceeded.  An incorrect control decision occurred when either action or economic threshold was exceeded and the other was not.

Conclusions:

Farmers can use sticky traps to monitor wheat midges in their fields and make accurate pest control decisions about 75% of the time. Traps are readily available, inexpensive, and easily deployed. A sticky trap with its catch is illustrated in the Appendix. Ten traps spaced at 10 m intervals within the crop provide an effective sample size. Traps should be positioned at head height when they are beginning to emerge, and left in place for three nights. Fields with two or fewer adults per 10 sticky traps rarely require an application of insecticide. High value wheat with an economic threshold of 4% infested seeds and trap catches of three or more midges per 10 sticky traps should be treated with insecticide. Wheat with an economic threshold of 7-10% infested seeds should be treated when four or more midges are caught on all 10 sticky traps. Routine use of sticky traps to monitor the wheat midge would eliminate most unnecessary applications of insecticide and assure that benefits of insecticide application usually exceed costs.

This monitoring system for wheat midge will save Manitoba farmers millions of dollars when wheat midges are abundant by reducing yield loss, retaining grade, and reducing unnecessary insecticide applications. When wheat midges are less abundant, the monitoring system will provide the peace of mind that comes from knowing that the decision not to use an insecticide has a sound basis because the farmers can determine that wheat midges really are not abundant in their fields.

Acknowledgements:

We thank many cooperating district agriculturalists and farmers who submitted samples. Financial support from the Agri-Food Research and Development Initiative, Canadian Seed Growers Association, and the Matching Investment Initiative program of Agriculture and Agri-Food Canada is gratefully acknowledged.

Appendix:

Technical details of the two studies can be found in:

Lamb RJ, Tucker JR, Wise IL, Smith MAH. 2000. Trophic interaction between Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) and spring wheat: Implications for yield and seed quality. The Canadian Entomologist 132: 607-625.

Lamb RJ, Wise IL, JE Gavloski, LA Kaminski, OO Olfert. 2002. Making control decisions for Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) in wheat (Gramineae) using sticky traps. The Canadian Entomologist 134: (in press).

ARDI Home - Contact ARDI