Developing Sustainable Management Tactics for Cucumber Beetles in Cucurbits

Objective 1: Develop trapping techniques to control early season infestations of cucumber beetles.

A mixture of cucurbit blossom volatiles (TIC) is an effective attractant for cucumber beetles and rootworms when used in traps (Levine & Metcalf 1988 and others). We have improved the effectiveness of such traps through a series of studies (Hoffmann et al. 1996a). In an effort to find a better trap for cucumber beetles we investigated several designs, including a design that only requires beetles to land on the trap. Upon landing on the trap the beetles pick up a small dose of toxicant or spores of an insect pathogen. This design results in rapid control from the toxicant adhering to the beetle or slower control via infection by the pathogen. In this unique scenario it should be possible to trap enough beetles (mass trap) to reduce damage. The advantages of this tactic include: (1) an extremely small amount of toxicant is used, (2) the toxicant does not contact the crop) and, (3) the toxicant is removed from the field after a relatively short time and properly disposed elsewhere. We found that the most effective toxicant was a solution of carbaryl (0.3% ai) in mineral oil. For example, at 10 traps per acre this translates to 0.0022 lbs. active ingredient per acre. This is a 99% decrease over a traditional insecticide application. Thus this method could considerably reduce the amount of insecticides used for managing cucumber beetles. For organic growers the synthetic toxicant could be replaced with one appropriate for their needs. However, our trials showed that the traditional insecticides most often used by organic growers were only marginally effective when used in traps. Although slower in rate of control, a pathogen-containing trap may be best for organic systems.

In 1996, the effectiveness of the trapping system was evaluated in 12' X 12' field cages containing 12 young squash plants. The trap we tested consisted of a 16 oz. plastic cup baited with TIC and glued upside down on an 8 inch dia. plastic plate. The cup was covered with yellow felt, which was saturated with a mixture of mineral oil and carbaryl (0.3% ai). Three traps were installed each in 3 cages and another set of 3 cages were used as control (without traps). One hundred striped cucumber beetles were released into the control cages and 175 into cages with traps. Cages were observed 1, 3, and 5 days after release, and the number of dead and live beetles was recorded. The average percentage dead (trapped) beetles in the cages with traps was 38.7, 64.0, and 42.7 on the first, third, and fifth day, respectively. Although a high proportion of beetles were unaccounted for in cages (escaped or were not observed) these results indicate that this trapping system holds potential.

During the summer of 1997 we investigated the potential for mass trapping cucumber beetles and conducted studies to determine the optimal trap density per unit area. Studies were conducted in experimental plots located at the Cornell University, Department of Entomology Research Farm, Freeville, NY. The trap design was similar to the ones used in 1996. One, two, four, and eight traps were installed per plot (1/20 acre) (corresponding to 20, 80, 160 traps/ac) while the plants were at cotyledon stage. The attractant and toxicant were replaced or refreshed on a weekly basis. The number of beetles on plants and the injury they caused per plant were recorded.

These studies showed that plots with one or eight traps had more beetles per plant than plots with two or four traps (Table 1). When expressed as percentage of the beetles captured [(number of beetles on the traps ÷ (number of beetles on the traps + total beetles in the plot)) X 100], the percentage of beetles captured by 1, 2, 4, and 8 traps was 23, 54, 65, and 49 percent respectively. Therefore at two or four traps per plot the traps were removing over half of the beetles from the plot. One trap per plot did not remove a sufficient number of beetles (23%) whereas eight traps per plot may have attracted additional beetles from surrounding areas. Plant injury coincided with the number of beetles per plot and was less in plots with two or four traps (Table 1). For example, the mean cumulative injury per plant was less than 20% in plots with two traps. The mean number of beetles trapped/day/plot increased (~doubled) with increasing trap densities. In addition to striped cucumber beetles, hundreds of western and northern corn rootworms were also attracted to these traps. The number of western corn rootworm found dead on traps ranged from 10 - 740 per plot during the study period. We also observed that an unknown number of beetles landed on traps and left before the toxicant had its affect. Thus a potentially large number of beetles may have been killed, but not recorded as captured by the trap.

Since the initial studies under field cages on the potential of these traps in 1996 and subsequent elaborate trials in 1997 under field conditions gave promising results. We further tested the effectiveness of mass-trapping during the summer of 1998. Studies were conducted in experimental plots located at the Cornell University, Department of Entomology Research Farm, Freeville, NY. Trials were conducted during mid June and mid August. The experimental set-up was similar to our 1997 studies. We did not find any significant difference in the number of beetles per plant or plant injury rating during either trial. Frequent heavy rainfall during the early season trials presumably reduced the effectiveness of the traps and late season trials were inconclusive because of low beetle densities.

We also conducted trials on two farms to demonstrate the effectiveness of mass trapping. An area of 100 X 80 ft was flagged in pumpkin fields and 6 traps (corresponds to 120 traps per acre) were installed, while the plants were at first leaf stage. The set-up was replicated twice on each farm and beetle densities and plant injury monitored. On both farms, the growers applied insecticide to control the beetle infestation across the entire field including the test areas. This experience emphasized the importance of the pest in the growers’ mind and that this tactic has to be very effective for growers to adopt.

In addition, we set-up TIC baited traps during fall in a pumpkin field at Whitney Point, NY, to determine whether traps could be installed in the fall to control the beetle population that overwinters. Data on the number of dead beetles on trap were recorded on a weekly basis. Our results show that these traps could be effective in reducing the beetle population in fall (Table 2). Over about a one month period (October 2 to November 5), total catch per trap averaged >200 beetles. Results from this study suggest that the use of traps in the fall and when cucurbit vegetation has been greatly reduced, holds considerable promise.

In conclusion, our studies indicate that with little modification these traps will be effective in mass-trapping striped cucumber beetle population, both during early and late season, using an extremely small amount of toxicant. We have demonstrated that traps are effective, but they need to be made more durable and able to remain effective under even adverse environmental conditions.

Objective 2: Develop cultural and biological control methods for control of cucumber beetles.

(a) Cultural control - Trap Crop:
Cucumber beetle management can also be improved through the development of cultural and biological control tactics. Certain types of cucurbits are highly preferred by cucumber beetles and these preferred types could aggregate beetles and their progeny for more efficient control. Our ongoing cucurbit screening and breeding program with Dr. Molly Kyle, Dept. of Plant Breeding, Cornell University, focuses on developing resistance to beetle pests and has identified highly preferred genotypes for use as trap crops (Hoffmann et al. 1996b, Reiners et al. 1999). The potential for spring trap crops has been further supported by work in Maine (Radin & Drummond 1994) and more recent work in Oklahoma (Pair 1997). Trap crops could be further enhanced by the addition of the long range attractant (TIC).

During the summer of 1996, we investigated the trap crop tactic in experimental plots. A highly preferred squash type (Seneca Zucchini) was transplanted either before (June 6, 1996), the same date (June 16, 1996), or following (June 24, 1996) the planting of the main crop of "Munchkin" pumpkins. Our objective was not only to determine if trap cropping would work, but also to determine if the time of planting of the trap crop was important. Each plot consisted of 68 plants planted in four rows (spacing 3' X 5' for squash and 2'X 5' for zucchini). Fifteen percent of each plot was planted to trap crop (first 5 hills / 25 ft of each outer row on opposing corners of plot). Immediately following the transplanting of the trap crops and main crop, beetle counts were recorded 2-3 times per week. Injury caused by beetles was recorded 1-2 times per week on all plants.

Results of the early season test showed the trap crop was not effective in reducing the number of beetles on the main crop. The reasons for this lack of control include the rapid destruction of the trap crop by beetle infestations and because of the unexpected presence of some highly preferred Munchkin plants in the main crop. Apparently, there was sufficient variability in the commercial Munchkin seed that some plants were highly attractive to beetles (>50 beetles per plant). Although this indicates that commercial seed has an unacceptable level of variability and a potential limitation for this tactic, it offered another opportunity. Instead of using a different variety for trap crop and main crop, we have now initiated a breeding program to select for, and develop, a highly preferred type of Munchkin pumpkin. This would simplify the trap crop method by permitting the use of a single variety in a field, except that some of it would be used as the trap crop.

The trap crop tactic was repeated late in the 1996 season, but modified to include TIC-baited traps (0.3% ai carbaryl + mineral oil design as described above) in the trap crop. Also, 15-20 zucchini (trap crop) plants were planted in a solid row instead of 5-6 plants in the previous experiments to minimize the total loss of the trap crop due to beetle feeding damage. The addition of TIC to the trap crop provided a long range attractant, while the traps provided a means to control beetles attracted to the trap crop. Results showed that the average number of live beetles on munchkins was reduced substantially (>75%) when the trap crop and TIC-baited traps were present (Fig. 1).

The trap crop tactic was tested again during summer 1997. The experimental set-up was similar to the late season trials in 1996. Immediately following the transplanting of the trap crops and main crop, beetle counts were recorded twice per week. Injury caused by beetles was recorded twice during the experimental period. No reduction in the average number of live beetles was observed. However, the data on the cumulative injury on munchkins two weeks after planting showed some significant differences between treatments and control. Average cumulative damage on munchkins was ~30% on treatment plots compared to 50% on control plots. The addition of traps with TIC reduced the number of beetles on the trap crop thereby minimizing the total loss of trap crop due to beetle feeding damage as observed during the first trial. Also, there were 3-4 times more beetles (including beetles dead on the trap) on the trap crop segment of the plots with TIC traps than beetles on the trap crop segment on plots with trap crops alone (Table 3).

Since the results of our 1996 and 1997 experiments showed that trap cropping with highly preferred cucurbits in combination with TIC traps is effective in reducing striped cucumber beetle numbers and injury, we conducted trials on three farms in 1998 to demonstrate the trap cropping technique. Unfortunately, all the growers applied insecticides upon the first arrival of beetles and therefore the effect of trap cropping on the protected crop could not be observed on any of these demonstration plots.

The results of our experiments shows that trap cropping is effective in reducing the injury caused by striped cucumber beetle adults to pumpkin seedlings at their early stages of development. Also, the addition of traps with TIC lures attracts and kills many adult beetles thereby reducing the population of striped cucumber beetle adults in the field.

(b) Cultural Control - Fall Cultivation
In addition to trap cropping, cultural practices such as fall cultivation may affect survival of overwintering cucumber beetles. The beetles present in the field late in the summer and fall, overwinter to infest fields the next spring and their control could minimize infestations the next season. Our objective was to monitor the fall population of beetles and also to study the effect of fall cultivation and fall clean-up on the beetle mortality.

In 1997, one field located on the Strawberry Valley Farm, Whitney Point, NY, and two fields located at Freeville, NY were monitored with yellow sticky cards and weekly visual sampling, starting early fall to study the beetle activity during fall. Ten sticky cards were installed at random inside each of the fields. Sticky cards were also installed around the fields to monitor the dispersal of beetles (if any) away from the field to overwintering sites. The beetle population at Freeville was very low when compared to Whitney Point. During late September when the plants were still green the beetles were found spread throughout the field, as the fall season progresses leaves started drying leaving only few green patches for the beetles to aggregate. After the first frost there were no more green patches and the beetles moved to decayed/damaged and some undamaged fruits left out in the field. The beetle count in the field started decreasing from the third week of Oct. Both visual sampling and sticky card counts showed almost no beetle activity during the first week of Nov.

In 1998, beetle activity during the fall and into winter was monitored in four fields located near Newark Valley, Ithaca, Whitney Point, and Freeville, NY. A combination of yellow sticky cards and weekly sampling of plants was used to monitor beetle populations. The beetle population at Newark Valley and Freeville was low and did not provide sufficient information.

Beetle populations during fall showed a similar trend to what was observed in 1997 (Fig. 2). During late September when the plants were still green the beetles were found throughout the field, as the fall season progressed leaves started drying leaving only a few green patches for the beetles to aggregate. After the first frost there was limited green vegetation and the beetles moved to decayed/damaged and some undamaged fruit in the field. The beetle count in the field started decreasing from the fourth week of October. Both visual sampling and sticky card counts showed almost no beetle activity during the second week of November. This indicates that if one wants to cause some mortality to the beetle population, the fall cultivation should be done while the beetles are still active during early fall. Waiting too long, may result in beetles leaving the field or their survival by burying themselves deeply in the soil. We installed sticky cards outside the field to monitor the dispersal of beetles, if any, away from the field. But the data obtained does not show any increase in the number of beetles on sticky cards installed outside the field, as the season progresses (Fig. 3), and therefore does not indicate that beetles move out of the field during fall. Research elsewhere, has shown that at least a portion of the beetles do overwinter in the field.

Low levels of beetles during fall prevented us from testing a range of cultivation equipment at Freeville, NY as planned. In 1996, we conducted simulated cultivation trials under field conditions. This trial consisted of placing 50 beetles each in a series of small cages. Cages contained crop residue or no crop residue and were either disked (simulated with hand shovel) or left undisturbed. Each treatment was replicated 5 times. Activity of beetles was observed into the late fall and the number of surviving beetles under the cages were recorded next spring. Unfortunately, no beetles survived overwinter in any treatments.

To better investigate the effect of cultivation on beetle survival we set up simulated cultivation trials under greenhouse conditions. We used 2 ft long sections of PVC pipe and buried fifteen beetles at 6, 12, and 18 inches below the soil line (to simulate burial by different cultivation tactics). Also, the effect of fall clean-up was tested by including plant debris or no plant debris on top of the soil in PVC pipes. Each treatment was replicated 5 times. The top and bottom of the PVC tubes were covered and the number of beetles surviving was recorded two weeks later. Results showed that 58 and 61 percent of beetles that were buried under 6 and 12 inches, respectively, were able to crawl up to the soil surface and survive (Fig. 4). However, none of the beetles that were buried at the 18 inch depth made it to the surface. This shows that deep cultivation in the fall can cause mortality and should help reduce overwintering populations. Shallower cultivation will also result in some mortality.

Suppression of immature cucumber beetles with entomopathogenic nematodes appears possible and could be especially effective if applied to trap crop plants to control progeny of aggregated adults. Unfortunately, the commercial nematode industry has dwindled drastically since this project started and we have not been able to obtain commercial formulations of the nematode species that we intended to test. However, complimentary research to that discussed herein has discovered two species of striped cucumber beetle parasitoids not previously recorded in this region: Celatoria setosa (Coquillett) (Diptera: Tachinidae), and Centistes (Syrrhizus) diabrotica Gahan (Hymenoptera: Braconidae). Maximum rates of C. diabroticae parasitism were more than 3 X greater than the only other published rate for this species (Fig. 5). This new biological control information demonstrated higher rates of parasitism than anticipated and indicated promise for enhancement of these parasitoids for biological control, thereby reducing the usage of insecticides to control beetles.

Objective 3: Improving the effectiveness of botanical insecticides.

In 1994, our preliminary experiments on rotenone dust plus the feeding stimulant, buffalo gourd root powder (BGRP), gave encouraging results. We therefore proposed to conduct studies to determine efficacy of rotenone and cryolite alone and in combination with a feeding stimulant that might improve control and permit reduced use of these insecticides. Pumpkin (Munchkin) seedlings were transplanted at cotyledon stage in groups of three and immediately covered with cages consisting of 2'X2'X3' PVC frame covered with netting material. Hills within cages were separated by 1' and cages were separated by ~5 m. Treatments include: rotenone full (2.5 lbs/acre), rotenone full + BGRP (0.5 lbs/ac), rotenone half (1.25 lbs/ac+ BGRP), cryolite full (10 lbs/ac), cryolite full + BGRP (0.5 lbs/ac), neem, an untreated control. Treatments were applied by hand, sprinkled onto the plants from a 7 dram polystyrene snap cap vial affixed with a cloth mesh top. After application of treatments, 25 striped cucumber beetle adults were released into each cage. Cages were examined and number of live and dead beetles and percent defoliation (rating scale: 0=0%, 1 =1-20%, 3=41-60%, 4=61-80%, 5=81-100%) on each leaf were recorded.

The result of our experiments show that all the botanicals tested caused a reduction in the damage caused by beetles. Rotenone at full (with and without the feeding stimulant), and 1/2 rate (with the feeding stimulant) had the lowest damage and highest beetle mortality (Table 4). The addition of feeding stimulant seemed to enhance the effectiveness of rotenone and permitted its use at half the full rate without significant reduction in the effectiveness. However, feeding stimulants did not alter the effectiveness of cryolite. Neem did not have any significant effect on the beetle survival or mortality but being an antifeedant it significantly reduced damage caused by beetles to plants. Overall, rotenone at 1/2 rate + feeding stimulant gave good control which permits reduced rates of rotenone.