Increasing biological control of brassica pests through overwintering

Final Report for ONE06-064

Project Type: Partnership
Funds awarded in 2006: $9,903.00
Projected End Date: 12/31/2008
Region: Northeast
State: Connecticut
Project Leader:
Kimberly Stoner
Connecticut Agricultural Experiment Station
Expand All

Project Information


This project was intended to address the question of whether the practice of overwintering Brassica crops (such as tatsoi, mizuna, mustards, collards, and kale) would increase biological control of caterpillars on Brassica crops on organic farms in Eastern Connecticut in the following year. We were not able to answer that question directly, but we did collect valuable information on the complex of caterpillars on Brassica crops, the parasitic wasps attacking the caterpillars, and also on the beneficial insects feeding on flowering plants on organic farms over the season. Through this research we have documented that cross-striped cabbageworm is now a major pest of Brassica crops in eastern Connecticut. We have also shown that a parasitoid, probably Cotesia orobenae, has followed its host to eastern Connecticut and is parasitizing this host at a substantial rate and through the season from mid-May until early September. We have also found, pending expert identification of our specimens, that Cotesia rubecula is established in eastern Connecticut and is now the major parasitoid attacking imported cabbageworm on organic vegetable farms. In addition, we presented several workshops with “how-to” information on an inexpensive way of producing greens for market through the winter using low tunnels. We know of six local growers who have tried overwintering Brassicas and other greens under low tunnels.


Bryan O’Hara, a diversified organic farmer, thought that his pest problems in Brassicas had been much reduced since he began overwintering a substantial area of Brassicas under low tunnels and allowing them to flower in the spring. One possible explanation for such a reduction in pest problems is that the overwintered Brassicas increase the numbers and diversity of parasitoids attacking his pests – either by providing habitat for the parasitoids to overwinter, by providing early season Brassica flowers as sources of nectar, or both.

In 2007, we compared Bryan’s farm to other diversified organic vegetable farms as described below. At each of these farms we took vacuum samples of any blooming Brassicas and other abundant flowering plants from early May to the end of June and again in September of 2007. We also collected weekly samples of caterpillars from Brassica crops (primarily collards and cabbage) and from wild mustard, and brought them back to the laboratory, reared them out to pupation, and recorded how many were parasitized.

Farm sites in 2007 (all organic):
1. Tobacco Road Farm (Bryan O’Hara, farmer). On this farm, low tunnels are used to carry Brassicas (and other leafy greens) through the winter on a substantial area (about ¼ acre) and then several of these leafy Brassicas, including tatsoi and mizuna, are allowed to flower and set seed the following spring, providing an abundant source of spring floral resources for parasitoids, and also producing seed for future crops. In addition, the farmers here deliberately collect parasitoid pupal masses (probably Cotesia orobenae) when harvesting fall cabbage and put them in other Brassica crops that will be left standing through the winter. They also grow plants intended to flower through the summer to provide resources to beneficial insects. They do not use any pesticdes at all (including Bacillus thuringiensis against caterpillars), relying on biological control and cultural methods for pest management. This farm was also used for the collection and rearing of caterpillars and the parasitism experiment with cross-striped cabbageworm in 2008.
2. Wayne’s Organic Garden (Wayne Hansen, farmer). This farm does not normally overwinter Brassicas, but did over the winter of 2006-2007 because some fall-planted kale survived the relatively mild winter without any protective covering. Because there was an infestation of imported cabbageworm threatening the marketability of cabbage and other Brassicas during our sampling period in 2007, a single treatment of Bacillus thuringiensis was used on cabbage on June 13, 2007, and we switched to sampling broccoli, collard, and kale after that point.
3. Full Moon Farm (Rob Miller, farmer). We used two sites for sampling – Utley Road, which had only recently been converted to a vegetable farm and had no previous Brassica crops before 2007 (and no Brassica weeds that we found), and Station Road about ½ mile away, which did not have any Brassica crops, but did have a large stand of wild mustard overwintering each year as rosettes. The Brassica crops at Utley Road (broccoli and cauliflower) were planted mainly for this experiment and abandoned after harvesting.

In 2008, we set out cross-striped cabbageworms for parasitism (details under Methods below), and we also collected and reared caterpillars from Tobacco Road Farm and Lockwood Farm (the Experiment Station research farm). Lockwood Farm is not organic overall, but did have a crop of organic collards overwinter outdoors (without any protective covering) from 2007 to spring of 2008. A crop of canola (not organic, but not treated with insecticides) was planted at Lockwood farm on June 13, 2008 and we sampled caterpillars from that canola from June 25 to Aug. 15, 2008.

Caterpillars we studied and what is currently known about their biological control:

Imported cabbageworm, Pieris rapae (L.). As the name implies, the imported cabbageworm is an exotic pest, originally from Europe, although it has been in North America since at least 1860. This is a very common caterpillar in farms and gardens across the US, and it is commonly controlled by use of Bacillus thuringiensis or insecticides. For many years, the major parasitic wasp attacking this species was the braconid wasp Cotesia glomerata (L.). In 1988 and 1993, Roy Van Driesche of the University of Massachusetts introduced a new Chinese strain of a related braconid wasp, Cotesia rubecula (Marshall). (Several previous introductions of this species had been attempted, but had not apparently become established in the eastern US.) This new introduction became established and was recovered from several sites in Massachusetts, Vermont, and central and western Connecticut (Van Driesche and Nunn 2002). Although these species are closely related and adults appear very similar, one difference between these two species is easily detected – C. glomerata is a gregarious parasite, so many larvae emerge from a dead caterpillar at once and form a clump of cocoons, whereas C. rubecula is a solitary parasite, and thus only a single larva emerges from the dead caterpillar and forms a cocoon. Another gregarious parasitoid of Pieris rapae attacks the pre-pupal stage and emerges from the pupae: Pteromalus puparum.

Cross-striped cabbageworm, Evergestis rimosalis Guenee. Until about 10 years ago, this caterpillar was only considered a pest of brassica crops in the South. Most of the scientific literature, including the papers on the major parasitoid, Cotesia orobenae (Forbes) are from studies conducted in Virginia (Gaines and Kok 1995, Mays and Kok 1997, Kok and Acosta 2001, Kok and Acosta-Martinez 2001). When we began this study, we had no information on biological control of cross-striped cabbageworm from the Northeast.

Diamondback moth, Plutella xylostella (L.). Diamondback moth is the most important pest of Brassica crops worldwide, due to its rapid multiplication and resistance to many pesticides. In temperate climates in North America, however, diamondback caterpillars have high mortality due to predation and are heavily parasitized by natural enemies, particularly where broad-spectrum pesticides are not used (Muckenfuss et al. 1990). Diadegma insulare (Cresson) has been shown to be the most important parasitoid of diamondback moth in many parts of North America.

Project Objectives:

These are the original performance targets from the proposal. They have been adapted as discussed below under “Materials and Methods.”
Summer 2006: Establish a colony of imported cabbageworms at the Connecticut Agricultural Experiment Station, and test the above protocol for parasitism tests to determine if this number of larvae and length of exposure will be sufficient to measure parasitism levels on organic farms.

Fall 2006: Plan with the farmers the location of the overwintered brassicas and other flowering plants (such as parsley or parsnips) to be sampled the following spring. Collect information from farmers about when they plant and harvest their brassica crops, the methods used in overwintering brassicas, and the cost of materials for overwintering brassicas.

Winter 2006-2007: Gear up imported cabbageworm colony so that it can produce 200 larvae per week in the spring of 2007.

April – June 2007: Collect spring vacuum samples and conduct spring parasitism tests as described above. The spring parasitism tests require rearing out 1000 imported cabbageworm larvae (assuming the tests are conducted for 5 weeks) in order to determine rates of parasitism.

August – September 2007: Conduct late summer – fall parasitism tests.

June 2007 – December 2007: Sort vacuum samples, organize and enter data from the vacuum samples and parasitism tests.

December 2007 – February 2008: Analyze data from all experiments, write up results for publication and presentation.

March 2008: Outreach through grower meetings (conference of the Connecticut chapter of the Northeast Organic Farming Association ) and publication in Gleanings (CT NOFA newsletter) and on the CT Agricultural Experiment Station website.

August 2008: Additional outreach through a workshop at the regional NOFA Summer conference.


Click linked name(s) to expand/collapse or show everyone's info
  • Bryan O'Hara


Materials and methods:

Our plan in the original proposal was to conduct parasitism tests with reared imported cabbageworm larvae by setting out potted collard plants at four farms with imported cabbageworm eggs. The idea was to have a population of young imported cabbageworm larvae in the field each week, expose them to parasitism by the local assemblage of parasitic wasps, then bring them back to the lab, rear them out, and measure parasitism. Mortality of the reared imported cabbageworm larvae was very high both in the field and in the laboratory, so we changed our plans.

Here is what we did instead:
In 2007, we searched Brassica plantings (or at Station Road, Brassica weeds) at the four locations described above for one hour of time weekly during the summer, collected all caterpillars and lepidopteran pupae found (and also brought back wasp pupae and adult wasps found on the Brassica plants when easily collected), and reared them in the laboratory to determine parasitism. Where possible, we searched collards, since they are relatively easy to search and support good populations of caterpillars, but we were dependent on what Brassicas were currently available each week in the farmers’ succession plantings. Laboratory rearing was done in 3 oz. clear Solo cups with collard leaves grown in the greenhouse as the food source.

This was a much more successful way of determining parasitism because the caterpillars were more likely to survive to pupation or until parasites emerged. In addition, we collected much more information. We have information about which species of caterpillars were present on the Brassica crops from early June to late September, and field parasitism rates for the three major species (imported cabbageworm, diamondback moth, and cross-striped cabbageworm) over the season. However, it is unlikely that we will be able to get a clear answer to the original question about differences in parasitism with overwintered Brassicas, since so many factors (species of caterpillars, age and size of caterpillars, host plants, etc.) are varying over our sites and over the season.

In addition to the caterpillar samples, we also collected vacuum samples of flowering plants on a weekly basis through all of May and June, a single mid-summer sample in July, and then weekly samples again in September. These were collected on a broad range of flowering plants at the four sites. Our plan to evaluate these samples is to first identify the parasites reared from caterpillars so that we have a good reference collection of the parasite complex on Brassica caterpillars, and then use that reference collection to get beyond general taxonomic categories (families and subfamilies) of parasitoids on flowering plants and see if we can link specific parasitoids of Brassica caterpillars to flowers on which they feed as adults. Vacuum samples were collected with a modified Stihl leaf-blower set to suction (rather than blowing out) with a cloth collecting net attached to the inside of the plastic air-blowing tube. Each type of flowering plant was sampled twice for a count of 10 seconds.

Another change was that the farmer decided not to plant Brassica crops at the Station Road site of Full Moon Farm (where the mustards overwintered), so the caterpillars collected in mid-summer 2007, after the wild mustards bloomed and began to set seed, were minimal. We continued to collect vacuum samples from this site, but the tables below referring to the Full Moon Farm refer to the Utley Road site.

In fall of 2007, we started a laboratory colony of cross-striped cabbageworms to be used in parasitism tests the following year.

In 2008, two plants were set out each week at the Tobacco Road Farm with egg masses of cross-striped cabbageworm from the laboratory colony, and retrieved two weeks later. The number and size of egg masses per plant were not easily controlled, but we were aiming to get 5 egg masses on the two plants with 20-30 eggs per mass. After the plants were brought back to the laboratory, each individual caterpillar was put in a separate 3 oz. Solo cup with collard leaves from the greenhouse and reared until death, emergence of parasitoids, or emergence of an adult moth.

We also continued in 2008 to collect caterpillars from Brassica crops in the field at Tobacco Road Farm and also at Lockwood Farm and to rear them out as described above.

Research results and discussion:

We have not completed the data analysis or identified the parasitoids collected in this project (30 parasitoids are currently awaiting identification by a specialist in braconid wasps, the family of Hymenoptera to which most of the parasitoids belong). However, some preliminary results are given in Tables 1-4.

At Tobacco Road and Wayne’s Organic Garden the total numbers and percentages of the different caterpillar species were very similar, and were different from the overall numbers and percentages at Full Moon Farm. At Full Moon Farm the total number of caterpillars was lower, the percentage of diamondback larvae and pupae higher, and the percentage of imported cabbageworms lower compared to the other farms (Table 1). This may be due to the fact that this was the first year Brassicas were grown at Full Moon Farm. When the collected caterpillars were reared in the laboratory, the loss due to death was lower for Wayne’s Organic Garden than for the other two sites (Table 1). I would guess that this may be due to the fact that the field schedule had us collecting caterpillars at Wayne’s at the end of the day, and therefore the caterpillars collected there spent less time in transport back to the laboratory.

Given this difference in death rates, the best available way to compare the rates of parasitism at the three farms may be to compare the ratio of those that successfully pupated to those from which parasitoids emerged (Table 1). This ratio was relatively high on Full Moon Farm (3.6) for both the imported cabbageworm and diamondback larvae, and was also similarly high for Wayne’s Organic Farm (3.4) for imported cabbageworm. The pupated:parasitized ratios were generally lower for Tobacco Road Farm (0.8 for cross-striped cabbageworm to 1.4 for imported cabbageworm). However, it should be noted that the higher relative rate of parasitism did not translate into lower numbers of caterpillars in the field at Tobacco Road Farm.

Table 2 shows the amount of parasitism through the 2007 and 2008 season on the naturally occurring cross-striped cabbageworms collected from the field at Tobacco Road Farm and Table 3 shows the amount of parasitism through the 2008 season on the cross-striped cabbageworms reared in the laboratory and set out in the field each week. (We do not yet have a confirmed identification of the parasitoids, but nearly all of the parasitism of cross-striped cabbageworm was due to a wasp similar to Cotesia orobenae in appearance and life history characteristics.) Although we started searching for caterpillars in the field on May 14 in 2008, we were not able to find any naturally occurring caterpillars until July 9, 2008. However, the parasitoids were active much earlier, attacking the caterpillars we set out as early as May 14 (Table 3).

Another important observation is that the vast majority of parasitoids of imported cabbageworm at all of these sites were soilitary braconids (suggesting a probable identification as Cotesia rubecula) rather than the gregarious Cotesia glomerata. This suggests that C. rubecula has expanded its range in eastern Connecticut to these farms in Lebanon, East Hampton, and Oneco, towns where it has not previously been reported. It also suggests that C. rubecula has replaced C. glomerata as the major parasitoid of imported cabbageworms over a wide area of eastern Connecticut.

Table 4 shows the plants flowering through the period of vacuum sampling at Tobacco Road Farm in 2007, and the number of wasps and bees counted per vacuum sample for each date. We have similar data for the other farms (and for both sites at Full Moon Farm) and also for fall-blooming plants, but Tobacco Road is the only farm where we could compare blooming Brassica crops (and ornamental plants in the Brassicaceae, such as sweet alyssum) with other flowering plants on-site. We have not yet identified the wasps collected from these samples, so we do not know if the wasps on the Brassica flowers include parasitoids of Brassica pests. We do know that the Brassica flowering period includes the time when the parasitoids of cross-striped cabbageworm first became active, and that there were few other flowering plants during this period (just chickweed at first, and later chervil and ground ivy). In terms, of numbers of parasitoids per sample, the Brassica crops had relative low numbers (0.5-5.5 wasps per sample), while sweet alyssum had higher numbers (3.5 -18 wasps per sample), and other flowering plants also had higher numbers, including white clover (4 -22 wasps per sample) and red clover (6.5 -21 wasps per sample). In general, the plants flowering later in June had more wasps, which may simply be because more wasps are active during that time.

Research conclusions:

Through this research we have documented that cross-striped cabbageworm is now a major pest of Brassica crops in eastern Connecticut. We have also shown that a parasitoid, probably Cotesia orobenae, has followed its host to eastern Connecticut and is parasitizing this host at a substantial rate and through the season from mid-May until early September. In addition, we have found (pending confirmation of specimen identification) that C. rubecula is established on organic farms in eastern CT and is now the major parasitoid of imported cabbageworm.

Both of these findings could be important in assisting vegetable farmers in measuring and increasing biological control of Brassica caterpillars on their farms. Additional specific information on methods to increase biological control are needed, however.

Although it is certainly true that Tobacco Road Farm is able to protect a major fraction of its Brassica greens from pest damage by growing Brassica crops during the winter, it does not appear that caterpillar abundance during the summer is any lower at Tobacco Road Farm than on other organic farms in the area.

Participation Summary

Education & Outreach Activities and Participation Summary

Participation Summary:

Education/outreach description:

Outreach events for this project:

1/27/07: Presentation by Kim Stoner on “Organic Management of Insects of Crucifers” with a section on changing caterpillars and parasitic wasps in CT. Also participated in a forum on organic management of vegetable pests. NOFA NY Winter conference, Syracuse, NY. 50 people.

3/9/07: Presentation by Kim Stoner on “Biological Control of Insects in Annual Crops” at the Connecticut conference on Natural Resources, Storrs, CT. 25 people.

12/11/07: Presentation by Kim Stoner on “Strategies to Control Insects in Successive Plantings” at the New England Vegetable and Berry Growers Conference in Manchester, NH – 150 people

3/8/08: Two presentations: Kim Stoner on “Changing Pest Complex in Vegetables” (30 people), and Bryan O’Hara on “Winter Growing” (35 people) at the Cultivating an Organic Connecticut Conference (CT NOFA) in Windsor, CT

8/10/08: Joint presentation by Bryan O’Hara and Kim Stoner on “Winter Growing” at the NOFA Summer Conference. 65 people. Presentation was also recorded as a videotape available from NOFA Mass.

10/26/08: On-farm workshop at Tobacco Road Farm on “Winter Growing” by Bryan O’Hara, with notes on pest management and our research by Kim Stoner. Sponsored by CT NOFA. 36 people attended.

On- Farm workshop featured on a blog by students from the Connecticut College Sustainable Farming Initiative. URL:

In press: News item in Gleanings (CT NOFA newsletter) about on-farm workshop with photos.

Immediate plans: Bryan O’Hara plans to write articles later this winter about his low tunnel methods for The Natural Farmer and Growing For Market

Project Outcomes

Project outcomes:

We do not have a detailed economic analysis that includes the cost of the extra labor involved in low tunnels, but we have calculated the cost of materials and expected revenue from each bed covered by a low tunnel for the winter. The initial investment in materials is $29 per 120 sq. ft. bed or $0.24 per sq. ft. This includes 3/16” steel hoops ($12), 2 layers of 2 mil plastic ($14), and 12 sandbags ($3) per bed . Our estimate is that the hoops should last at least 30 years, and the plastic and sandbags at least 3 years, so the annual cost spread over the life of the materials would be $0.05 per sq. ft. Bryan O’Hara estimates his revenue in the low tunnel beds as $2 per sq. ft. There are, of course many additional costs not accounted for here - all the normal costs of growing a crop, plus the additional labor required for putting the covers on and off, but low tunnels are a very inexpensive way of extending the season and making more intensive use of limited growing area.

Farmer Adoption

We do not have a system for tracking farmer adoption, but we do know 6 specific farms that have tried Bryan O’Hara’s methods of growing Brassicas and other greens through the winter under low tunnels, including Michael Melillo of North Haven, CT; Rob Kelly and Dina Brewster of The Hickories Farm in Ridgefield, CT; Bruce Kittredge of Hampton, CT; David Zemelsky of Durham, CT; Julia Smargorinsky of Woodbridge Farm, Salem, CT; and the University of Connecticut Garden Club of Storrs, CT. Several students from the Connecticut College Sustainable Farming Initiative also participated in the October 2008 “Winter Growing” workshop at Bryan’s farm. The low-tunnel method is particularly useful for school and university sustainable gardener/farmers, since it is an inexpensive way for them to grow crops during the school year.

Assessment of Project Approach and Areas of Further Study:

Areas needing additional study

Organic vegetable farmers are very interested in finding ways to enhance naturally occurring biological control through their cultural practices. This study was only a small-scale preliminary step toward determining if overwintering Brassicas provides overwintering habitat for natural enemies of vegetable pests and discovering what flowering plants are attractive to beneficial insects on organic farms. Both of these questions would need either rigorous labor-intensive studies under controlled conditions on larger farms or else observational studies with dependable, consistent behavior by the farmers on a large number of farms in order to get scientifically valid results.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.