Estimating the abundance of lygus in canola fields

Key Result

An efficient sweep sample method was developed that allows producers to efficiently assess the risk of lygus damage and minimize yield loss and control costs, resulting in yield savings of up to about 20% in some fields.

Project Summary


As of 1994, lygus bugs are a newly recognized pest of canola in western Canada. They feed on the growing points of canola plants and can cause damage to canola particularly when they attack pods and feed directly on developing seeds. This damage is usually invisible to the producer, but may result in a lower than expected yield. Although there is no doubt that lygus bugs can be damaging to canola, it has been difficult to estimate how many lygus bugs it takes to cause economic losses.


A lygus bug in a sweep net
  1. To determine where lygus bugs occur on canola plants at various stages of growth;
  2. To determine whether lygus bugs are distributed evenly from the edge to centre of canola fields;
  3. To compare the efficiency and accuracy of four sampling methods (i.e. vacuum, sweep net, beating tray, absolute counts);
  4. To devise an efficient method for producers or pesticide applicators to estimate lygus abundance in canola and determine whether control is necessary.
    • The goal was to devise a sequential sampling scheme that provides sufficient accuracy, can be completed with minimal equipment, and requires about 1 hour to complete per field;
  5. To devise a precise method for researchers to estimate abundance and sample lygus in experimental plots.


One aspect of this study was to estimate an economic threshold for lygus in canola. Producers also need a method to efficiently assess their crop to determine whether the threshold is exceeded and control is required. Based on work supported by the canola producers of western Canada, a sampling method for lygus in canola has been developed.

To determine whether or not lygus densities have exceeded the economic threshold, producers need to use a sweep net to sample the edges of their fields when the crop has almost finished blooming. A sequential sampling scheme has been developed to minimize the number of samples required. We estimate that up to 12 twenty-sweep samples (usually fewer) are required per field. The sampling method should take no more than 30-45 minutes to complete, and could result in yield savings of up to about 20% in some fields.

Resulting publications:

Wise, I., & Lamb, R. (1998). SAMPLING PLANT BUGS, LYGUS SPP. (HETEROPTERA: MIRIDAE), IN CANOLA TO MAKE CONTROL DECISIONS. The Canadian Entomologist, 130(6), 837-851. doi:10.4039/Ent130837-6

Further experimental details

Research Conducted – 1 May 1994 to 1 March 1995:

  1. A single large plot of canola was seeded at our experimental farm at Glenlea, Manitoba. This plot was sampled weekly by sweeping to determine when lygus enter the crop, when the various age classes appear, and what type of sampling devices would be effective. Sweep sampling proved to be much more effective for assessing all stages of lygus bugs in the crop than searching individual plants. The younger stages of lygus bugs occurred at lower levels in the crop and were more difficult to sample. This should not interfere with accurate sampling of fields by produceō€€¸s because at the time sampling and control are appropriate, most lygus have reached a stage that can be sampled.
  1. Two 100 m transects were designated in each of 12 commercial fields of canola in early August. These were oriented at right angles to the edge of the field. Sweep sampling was conducted from these transects by four samplers working at distances of 0, 10, 20, 50 and 100 m into the fields. These data allowed us to compare the efficiencies of different samplers, and to determine how much variation we might expect among samplers. Most importantly, these samples allowed us to determine that lygus bug populations are sufficiently evenly distributed throughout fields that they can be estimated by sampling the edge of the field. This result assures that a cost-effective sampling scheme for lygus bugs in canola can be developed for canola producers.
  2. Beating trays and bag samples of individual plants quickly proved to be inadequate for lygus bugs and were abandoned. Vacuum sampling is effective but impractical for producers. We used a combination of insecticide applications, vacuum sampling and hand searching in large experimental plots to compare sweep sample estimates with absolute abundance of lygus. These data suggest that sweep sampling collects about one-half of the lygus present in the sampled area, and that the age distribution of the lygus bugs are similar in sweep and absolute samples.
  3. Replicated sampling was conducted along the edges of 15 commercial fields using sweep nets. The beating tray method was abandoned as ineffective. The time taken to collect and count samples was recorded for 40 of the samples. Analyses show that an adequate estimate of the lygus bug population could be conducted in 30-45 minutes in one commercial field by one sampler.
    Field data collected this summer were entered on the computer and most analyses of the data have been completed. In addition, sampling data collected in previous years in Manitoba from about 100 commercial fields and :many experimental plots are being incorporated into the data set. Also, data from northern and central Alberta have been made available to us by Dr. R.A. Butts, Lethbridge Research Center 1 and are being included in the analyses. This large data base is allowing us to develop an effective and well-verified sampling scheme that producers can use efficiently to assess the threat to their canola crop from lygus bugs. Analyses of these data sets began in December and we expect them to be completed in May of l995. Work to date shows that an accurate estimate of the density of lygus can be made with 12 or fewer groups of sweep samples. A single group of 20 sweeps can be completed and counted in about 2 minutes, assuring that sweep sampling will be practical for canola producers. The sequential scheme that is currently being developed will assure that only the minimum number of sweeps need be taken. The number of required sweeps is lower when lygus densities are very low or very high, so that a control decision can usually be reached quickly with a minimum of effort.
  1. It proved to be more difficult than we expected to devise a highly precise method for researchers to use when estimating the abundance of lygus in experimental plots. Fortunately, The more important objective of developing an efficient method for producers was achieved. The absolute method described in item 3, above, appears to be the only practical method for researchers, although it is quite time consuming. We want to confirm the calibration of the relationship between sweep counts and absolute counts in the summer of 1995, so that researchers can also use sweep sampling, but with a more precise understanding of the proportion of lygus collected in a sweep sample.