Combining trap cropping and natural-chemical lures to attract and kill crucifer flea beetles

Final Report for SW08-102

Project Type: Research and Education
Funds awarded in 2008: $191,868.00
Projected End Date: 12/31/2011
Region: Western
State: Washington
Principal Investigator:
William Snyder
Washington State University
Expand All

Project Information


The crucifer flea beetle (CFB), Phyllotreta cruciferae, is an oligophagous pest of Brassica crops. In plots both west (Mt. Vernon, WA) and east (Moscow, ID) of the Cascade Mountains, we have been evaluating different species-compositions and optimal distance of trap crop plantings that will effectively draw flea beetles out of broccoli, yet prevent over-spilling from occurring. Flea beetle populations in trap-crops were tracked using D-vac suction sampler, while visual observations were used to monitor CFB populations and damage in broccoli. Our results suggests that multi-species trap crops protected broccoli planted at varying distance by inducing subtle changes in CFB behavior.

Project Objectives:

We examined simple and diverse trap crop plants in their ability to draw flea beetles away from the broccoli crop.


The crucifer flea beetle (CFB), Phyllotreta cruciferae (Coleoptera: Chrysomelidae), is an oligophagous pest of Brassica crops throughout North America (Palaniswamy and Lamb 1992). In the Pacific Northwest, many growers rely on Brassica crops as a major component of mixed-vegetable production. In addition, Brassica crops can be planted and harvested season-long, providing a steady source of income throughout the year. Organic Brassica crops are valued at over $60 million annually (USDA NASS 2008) and include arugula, broccoli, cabbage, kale and mustard greens. However, feeding damage by the CFB lowers marketable yields of these crops. Adult flea beetles scar foliage, resulting in produce that is unattractive to consumers, and often kill seedlings and small transplants outright. Indeed, flea beetle damage sometimes leads to total crop loss (Newton 1928, Kinoshita et. al 1979, Turnock and Turnbull 1994), and for this reason many small-scale vegetable growers in the Pacific Northwest are unable to include Brassicas in their yearly rotations.

Pesticides are commonly used to manage flea beetles in Brassica crops. Current chemical controls include broad spectrum insecticides (Howell et al., 2004) and, less successfully, several organic pesticides including pyrethrin and azidirachtin (Andersen et al., 2006). However, with the potential risks pesticides pose to the environment and non-target organisms, more research is needed on alternative control methods. Furthermore, organic producers are very limited in their options for controlling flea beetles; these being limited to the use of floating row covers, which can be costly, and organic-approved insecticides that must be applied frequently as flea beetles continuously move into the crop from surrounding vegetation. The limitations of these strategies have led the industry to look for alternatives.

Trap crops represent one such potential management alternative. Trap crops are stands of plants that protect the target crop by attracting pest insects and/or providing a more suitable host plant (Hokkanen, 1991; Shelton and Badenes-Perez, 2006). Trap crops work because nearly all pest insects display a preference for specific plant types (Hokkanen, 1991). In general, mobile insects capable of direct flight are good candidates for trap cropping (Potting et al., 2005), as opposed to insects with passive flight like aphids. This is because with direct flight, insects can orient toward host plants more easily. While trap crops diversify plantings on farms, which has shown many ecological benefits (Vandermeer, 1989; Altieri and Nicholls, 1994), the main benefit of trap cropping is reduced pesticide use. Once pests are concentrated in the trap crop they can be removed by different means, such as burning or tilling-under the trap crop (Hokkanen, 1991).

There are many successful examples of trap cropping. For example, in soybeans, Mexican bean beetles can be controlled using a trap crop of snap beans (Rust, 1977). Similarly, for over 50 years in Belorussia, early-planted potato trap crops have been used to protect later plantings of potatoes from Colorado beetle attack (Dorozhkin et al., 1975). Trap crop success depends on a number of variables, such as the physical layout of the trap crop (e.g., size, shape, location) and the pests’ patterns of movement behavior (Hannunen, 2005). Manipulating diversity within trap cropping may also provide improved pest suppression. For example, diverse trap crops are composed of plants with different physical characteristics, phenologies, structures, shapes and colors. Therefore, the crucifer flea beetle may be more attracted to diverse trap crops. We examined whether multi-species trap crop plantings were more effective than any single species at attracting the crucifer flea beetle (CFB), Phyllotreta cruciferae, away from broccoli (Brassica oleracea var. italica) plantings.


Click linked name(s) to expand/collapse or show everyone's info
  • Sanford Eigenbrode
  • Matthew Morra
  • Todd Murray
  • Kathleen Painter
  • Cinda Williams


Materials and methods:

Trap Crop Diversity 2009.

Because CFB is a ubiquitous pest both east and west of the Cascade Mountains, we conducted experiments at the Washington State University (WSU) Mt. Vernon Research and Extension Center in Mt. Vernon, WA (west) and at the University of Idaho Parker Plant Science Farm in Moscow, ID (east). In this experiment we varied trap crop diversity; however, trap crop density remained the same. The simple trap crop treatments consisted of five monocultures of each of the trap crop species, while the diverse trap crop treatments followed a substitutive design that included every possible unique combination of four species out of the total pool of five species (Table 1). Trap crop species included Barbarea vulgaris (Yellow Rocket), Brassica juncea (Pacific Gold Mustard), Brassica napus (Rape), Brassica oleracea var. acephala (Green Glaze Collard) and Brassica rapa subsp. pekinensis (Pac Choi). Trap crop plants were randomized for each replication in the diverse treatments. There were a total of 11 different treatments: five simple (monocultures), five diverse (polycultures), and one control (bare ground). Furthermore, each treatment was replicated four times in Mt. Vernon, WA and, due to space limitations, two times in Moscow, ID.

Trap crop plantings were both direct-seeded and transplanted, depending on the variety. Seeds of B. juncea and B. napus were direct seeded at a rate of 5g/foot in Moscow, ID on May 5, 2009 and in Mt. Vernon, WA on April 4, 2009. After the direct-seeded trap plants emerged they were grown for two weeks. Greenhouse-grown seedlings of B. vulgaris, B. oleracea var. acephala, and B. rapa var. pekinensis were hand transplanted in the field at the recommended planting densities. Therefore, both direct-seeded trap crop plants and transplanted seedlings were roughly the same size. Once the trap crop plantings were established (four weeks after planting) the broccoli (Brassica oleracea var. italica) were transplanted. The experimental design consisted of four rows of the target crop, broccoli, flanked on both sides by the different trap crop treatments (Figure 1). Each row was spaced 21’’ apart.

Trap Crop Diversity 2010.

We expanded our diversity experiment and included treatments with simple, low-diversity and high-diversity schemes of our most attractive trap crop species from 2009 (Table 2). The simple trap crop treatments consisted of three monocultures of each of the trap crop species planted alone, while the low-diversity scheme followed a substitutive design with every possible unique combination of two species out of the total pool of three species included in different plots. The high-diversity scheme consisted of one unique combination of all three species. Trap crop species included Brassica juncea (Pacific Gold Mustard), Brassica napus (Rape), and Brassica rapa subsp. pekinensis (Pac Choi). There were a total of eight different treatments: three simple (monocultures), four diverse (polycultures) and one control (bare ground). Furthermore, each treatment was replicated eight times in Mt. Vernon, WA and, due to space limitations, four times in Moscow, ID. Planting methods from 2009 were followed in 2010.

Trap Crop Optimal Distance 2011.

We simultaneously assessed the effects of trap crop proximity and CFB removal in its ability to protect adjacent broccoli. The trap crop consisted of our high-diversity trap crop mixture from 2010: Brassica juncea (Pacific Gold Mustard), Brassica napus (Rape), and Brassica rapa subsp. pekinensis (Pac Choi). Trap crop plantings were both direct-seeded and transplanted depending on the variety. Seeds of B. juncea and B. napus were direct seeded at a rate of 5g/foot in Moscow, ID on June 1, 2011 and in Mount Vernon, WA on May 25, 2011. After the direct-seeded trap plants emerged they were grown for two weeks. Greenhouse grown seedlings of B. rapa var. pekinensis were hand transplanted into the rape and mustard at the recommended planting densities. Therefore, both direct-seeded and transplanted trap plants were roughly the same size. Once trap plants were established broccoli (Brassica oleracea var. italica) plants were transplanted.

The experimental design consisted of five rows of trap crop adjacent to two rows of each broccoli plot. Both the trap crop and broccoli rows were 5 m in length. Five rows of trap crops were planted which corresponded to the broccoli-to-trap-crop ratio used the previous seasons. Broccoli plants were planted at three different distances away from the trap crop: 59 cm, 3.5 m and 7 m (Figure 2). With this design, we were able to observe if over spilling (meaning, large numbers of CFB accumulating in the trap crops then “spilled over” into adjacent broccoli) was occurring. Broccoli plants were spaced 18’’ apart within rows, and both broccoli and trap crop rows were spaced 23’’ apart between rows. Randomly selected trap crop plots were treated with a pyrethroid insecticide (Mustang Max) to kill CFB in the trap crop. Control plots consisted of no trap crop (bare ground adjacent to the broccoli). All plots were spaced 3.5 m apart and separated by bare ground.

CFB Sampling.

For all experiments, samples within the trap crop were taken with a D-vac suction sampler (Rincon Vitova, Ventura, California) at two week intervals. In addition, flea beetles were counted visually on the broccoli by carefully examining all broccoli leaves and florets. At the end of the season, approximately 73 days after planting, subsets of mature broccoli in each treatment were harvested to determine final whole plant biomass.

Statistical Analysis.

All data were analyzed using JMP, version 9 (SAS Institute Inc., Cary, North Carolina). For the our field experiments repeated measures, ANOVA was used to analyze the interactions between flea beetle numbers in trap crops and broccoli and broccoli yield. In both experiments, flea beetle counts and whole plant dry weights were log10 transformed to homogenize variances.

Farm Trials 2009.

We conducted farm trials across the states of WA and ID in 2009. We examined whether mustard meal, a byproduct of making biodiesel, enhanced the trap crop’s effectiveness. In the process of making biodiesel, oil is removed from mustard seed, leaving behind a flaky-meal byproduct. When water is added to the mustard meal the biologically active products of glucosinolates (isothiocyanates) are released. It is possible that amending the trap crop with mustard meal may enhance the trap crop’s attractiveness to CFB, which are attracted to these mustard oils. In addition, mustard meal is currently being used as a biofumigant, similar to mustard green manure, to control soil pests and diseases and also as a fertilizer because it releases nitrogen. To examine the effectiveness of amending trap crops with mustard meal, four field trials collaborating with local organic growers were conducted. These collaborators included Greentree Naturals Certified Organic Farm (Sandpoint, ID), GT’s Farm Foods (Moscow, ID), Garden Treasures Nursery and Organic Farm (Arlington, WA) and Rents Due Ranch (Stanwood, WA). Using a split plot and a monoculture trap crop design, B. juncea was planted in one row with half of the trap amended with B. juncea mustard meal, along with broccoli planted adjacent to the trap crop. The same procedure for sampling and documentation as utilized in the larger field experiments were followed for these smaller farm trials.

Farm Trials 2010.

We conducted paired-farm trials to investigate the effectiveness of a mustard trap crop monoculture. The distance at which a trap crop monoculture of mustard attracts CFB is not known. Therefore, we used paired farms located near each other to help guide us in this question. Our collaborators were located east and west of the Cascade Mountains. East collaborators included Greentree Naturals Certified Organic Farm (Sandpoint, ID), GT’s Farm Foods (Moscow, ID), University of Idaho Soil Stewards (Moscow, ID), Washington State University Organic Farm (Pullman, WA) and Washington State University Tukey Orchard (Pullman, WA). Collaborators in our west location included Zestful Gardens (Tacoma, WA) and Terry’s Berries (Tacoma, WA). Paired farms were randomly assigned either a mustard trap crop treatment adjacent to a broccoli monoculture (trap crop treatment) or a broccoli monoculture (control). We compared and recorded CFB densities in broccoli adjacent to the trap crop treatment and broccoli without the trap crop using the same procedure for sampling and documentation as utilized in the larger field plots.

Research results and discussion:

Trap Crop Diversity 2009.

We found no significant effect of the presence versus the absence of trap crops, or diversity within those crops, on broccoli whole plant dry weight (control, monoculture and diverse; F 2, 77 = 2.29, P = 0.107) (Figure 3a). However, both simple and diverse trap crops attracted higher densities of CFB than bare ground controls (F1, 62 = 36.021, P = < 0.000; Figure 4a). In addition, diverse trap crops collected significantly greater numbers of CFB than trap crop monocultures (F1, 56 = 4.339, P = 0.0418; Figure 4a). Feeding damage to the trap crop compared with broccoli was more severe (Figure 5), illustrating the importance of the trap crop in relation to the protection target, broccoli.

Trap Crop Diversity 2010.

When only the most attractive trap-crop species were included within our species pool, results dramatically differed from the first experiment in 2009. In this case, we found that broccoli whole plant dry weight was significantly greater when grown adjacent to high-diversity trap crop mixes (F3,89 = 3.312, P=0.024; Figure 3b); however, diversity effects were not seen, for two-species trap crop mixes (Figure 3b). These results were consistent across both our east and west sites (F1,28 = 0.0008, P=0.9772). As before, flea beetle numbers reached higher densities in trap crops than in bare-ground controls, although flea beetle numbers did not differ between high- and low-diversity trap crop plantings (F1,40 = 1.0281; P=0.3167; Figure 4b). Results of flea beetle densities in broccoli showed that within the broccoli itself, flea beetle numbers were significantly affected by the presence of both low-diversity and high-diversity trap crops (F3,88 = 4.473, P = 0.0057; Figure 4d).

Trap Crop Optimal Distance 2011.

Results from CFB in broccoli showed that we collected significantly fewer flea beetles out of our control plots where there was no trap crop (F2, 22 = 82.760, P = < 0.0001). We also collected significantly fewer CFB in our treated trap crop plots (sprayed) (F2, 22 = 82.760, P = < 0.0001; Figure 6). Therefore, in the control plots we had very few CFB. When a trap crop was present, we saw significantly more CFB and when we killed CFB in the trap crop we found significantly fewer. Results from broccoli whole plant dry weight showed that broccoli planted in the presence of a trap crop attained the greatest dry weight (F 2, 275 = 16.516, P = < 0.0001, Figure 7). We also found no effect of trap crop distance on broccoli yield (F6, 275 = 1.409, P = 0.211). Results of CFB on broccoli were more complex. Since broccoli yields were greater adjacent to trap crops, we would expect to find an increase in CFB on broccoli in our control plots and a decrease in CFB on broccoli in the trap crop plots; however, this was not the case. Despite causing strong and clear differences in yield, trap cropping had surprisingly little effect on where CFB were found in the broccoli crop (F6,66 = 20.558, P = < 0.0001). In the low density west site, CFB were found at the edge of the broccoli crop regardless of whether a trap crop was present and sprayed. Patterns were intriguingly different at the high beetle density eastern site. Here, CFB were evenly distributed across treatments and locations in the crop in all situations but one: when the near-planted trap crop was not sprayed, CFB densities fell in adjacent broccoli. Our data suggests three points about CFB behavior. First, when densities are low CFB concentrate at the crop edge, but move further inside the crop as densities increase. Second, CFB are not necessarily heavily feeding on the plants where they are found; broccoli was protected by trap crops without changing pest densities in adjacent broccoli, regardless of whether flea beetles were at high or low densities. Finally, at least at in the high-density site, killing CFB in the trap crop reduced CFB densities in the adjacent broccoli crop (although, again this change in beetle density did not impact broccoli yields). Altogether, these results suggest that CFB are regularly moving between the trap crop and broccoli, but when given a choice will do most of their feeding on the trap crop. Apparently, because of this movement, the trap crop protected broccoli without changing apparent densities of the pests in the crop. Thus, the high-diversity trap crop mixture attracted CFB feeding away from broccoli.

Farm Trials 2009.

Results from the 2009 mustard meal farm trials showed no significant differences between trap crops amended with mustard meal and trap crops without mustard meal (F = 0.100; P = 0.753; Figure 8). When mustard meal is used as a biofumigant it has to be covered with a tarp as soon as water is applied to the meal (making it biologically active) or else the isothiocyanates dissipate quickly into the environment. This most likely represents what happened in our study. Furthermore, amending a trap crop with mustard meal would not be practical and hinder the trap crop’s ability to attract pests.

Farm Trials 2010.

Results from the 2010 paired-farm trials showed no significant differences in CFB density recorded in broccoli adjacent to the trap crop treatment and broccoli without a trap crop (F = 0.334; P = 0.570; Figure 9).

Research conclusions:

Trap crop success is dependent on a variety of factors, each specific to the pest species. Such factors include the pest’s ability to locate host plants, movement patterns and the relative attractiveness of host plants to the target pest. Our diverse trap crop, whether sprayed or not, was successful at protecting broccoli planted at various distances from CFB attack. However, we suspect that the diverse trap crop was able to provide protection without changing CFB numbers on broccoli because of the behavior CFB displays. For these reasons, making general trap crop recommendations can result in unsuccessful trap cropping systems. Therefore, understanding interactions between trap crops-trap crop diversity and pest control is essential and will allow growers to utilize more sustainable successful management options to suppress flea beetles and other pests in organic farms.

Participation Summary

Research Outcomes

No research outcomes

Education and Outreach

Participation Summary:

Education and outreach methods and analyses:

Our outreach program focused on three field days where we concentrated on natural pest management strategies that emphasized the use of trap crops. Evaluations from a previous field day revealed that basic entomology knowledge, such as beneficial insects and pest identification, is lacking and highly requested. We incorporated this information, as well as discussing the use of entomopathogenic nematodes (year 1), insectary plants (year 2) and the benefits of predator diversity and evenness (year 3) into our Natural Pest Management Field Day. A visual display was set up which showed pictures of the trap crop farm trial throughout the field season. This allowed the attendees to observe flea beetle management through time. A colored booklet describing local beneficial and pest insects and organic management strategies were also included. In addition, the field day attendees participated in a “farm walk”, during which the host grower provided a guided tour of that year’s experimental trap crop plots, discussing what worked well, what didn’t, and what they have learned so far (Figure 8). In addition, we demonstrated the D-vac suction sampler (Figure 9) and assisted attendees on identifying collected insects which included common beneficials and pests found on the farm.

When the “work” portion of the field day concluded, all of the participants gathered for a meal featuring local food prepared by a caterer. This approach of including locally-produced foods in the communal meal has been successfully used by our participating grower (Greentree Naturals Certified Organic Farm) at a series of field days at her farm and by one previous field day we have collaborated on. In total we had over 70 growers and interested community members attend. Currently we are preparing a manuscript for submission to the Journal of Applied Ecology regarding the results of our trap crop diversity experiment.

Two manuscripts will soon be submitted to peer-reviewed scientific journals, reporting the results of this Western SARE-funded project.

Education and Outreach Outcomes

Recommendations for education and outreach:

Areas needing additional study

Biodiversity’s ecological benefits have been applied in agriculture to alleviate pest problems. However, few studies have used plant diversity within trap crops to enhance pest control. Our study demonstrated how trap crop diversity can improve crucifer flea beetle control in broccoli and emphasized the importance of manipulating agroecosystems to enhance ecosystem services. More empirical data manipulating diversity at small scales can provide single farmers with practical alternative pest management options.

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.