Developing a Perennial Living Mulch System to Manage Insect Pests in Northeastern Cantaloupe Fields

Progress report for GNE20-236

Project Type: Graduate Student
Funds awarded in 2020: $14,955.00
Projected End Date: 04/30/2022
Grant Recipient: University of Maryland
Region: Northeast
State: Maryland
Graduate Student:
Faculty Advisor:
Dr. Cerruti R. R. Hooks
University of Maryland
Expand All

Project Information


Drawbacks of excessive pesticide usage in cantaloupe have created a demand for alternative pest control strategies that can reduce producers’ reliance on chemical inputs. The economic burden of growing cucurbits such as cantaloupe can be considerable due to the cost of frequently applied chemicals to protect cucurbits with low pest tolerance. Further, chemical sprays can have a negative impact on natural enemies inhabiting cucurbit fields and consequently cause secondary pest outbreaks. Interplanting cash crops with cover crops, also known as companion planting, has been used for centuries to improve soil quality and suppress weeds. However, interplanting a live cover crop (living mulch) into a cucurbit crop can be used to also help manage insect pests and diseases that they vector. Recent research has shown that interplanting a red clover living mulch into cucumber plantings can be an effective strategy to reduce insect pest and increase natural enemy populations without negatively impacting crop yield. This is a promising finding for other crops in the Cucurbitaceae family that share similar insect pests. This study will determine if similar benefits can be conferred to a cantaloupe system. Specifically, the objective is to determine whether interplanting cantaloupe into two perennial living mulches (i.e., alsike clover or Virginia wildrye) systems that are structurally dissimilar can be used to reduce insect pest and increase natural enemy populations in cantaloupe. Additional objectives are to compare the ability of these living mulches to enhance cantaloupe yield and serve as overwintering refuge for natural enemies.

Project Objectives:

The overall objective is to investigate and disseminate an alternative tactic for managing insect pests in cantaloupe plantings. Specific study objectives are as follows:

  • Compare the impact of a monoculture cantaloupe treatment with two additional treatments: 1) cantaloupe interplanted into a perennial bunch grass (Virginia wildrye, Elymus virginicus) and 2) interplanted into a perennial legume (alsike clover, Trifolium hybridum) on herbivore abundance,
  • Compare treatment impact on the abundance of natural enemies,
  • Determine treatment impact on yield and fruit quality, and
  • Determine the potential of different treatments to serve as overwintering refuge for natural enemies

The purpose of this project is to investigate living mulches in cantaloupe plantings as a sustainable alternative to insecticides for managing insect pests. Insect pests pose a challenge for Northeast cantaloupe growers; and if left unmanaged, can cause significant yield reductions and lost revenue. However, using an intensive chemical regimen for their mitigation can contribute to environmental and human health problems (Geiger et al., 2010; Relyea, 2016; Weisenburger, 1993). For example, a common pest mitigation practice used in cantaloupe is the application of systemic neonicotinoids to protect seedlings which is typically the most vulnerable stage (CMCC, 2003). Unfortunately, these products have received public scrutiny worldwide because of their potential contribution to pollinator decline (Van der Sluijs et al., 2013). Further, the high susceptibility of cantaloupe to pest damage and insect vectored diseases contributes to many growers administering prophylactic sprays. Cantaloupe is the second most susceptible cucurbit crop to insect damage (ATTRA, 2008), and tolerance to insect feeding damage is low in cantaloupe due to the threat of bacterial wilt. As such, some extension programs suggest foliar application of pyrethroids when only one cucumber beetle per plant is observed (Brust & Foster, 1999). In the year 2000, insecticides were used on 74% of US cantaloupe crops. Most treatments were applied to plants after crop emergence to protect them after fruit-set (CMCC, 2003). Concerns for cucumber beetles alone cause some growers to apply pesticides as many as seven times over a single season to prevent significant lost (Brust & Foster, 1999).

In recent studies, performed in our lab, crops in the Cucurbitaceae family (e.g., zucchini and cucumber) experienced a significant reduction in insect pest and increase in natural enemy abundance when interplanted with a living mulch (Hinds & Hooks, 2013; Kahl et al. 2019). Increasing the vegetational diversity within a crop field by adding multiple plant species can confer associational resistance to cash crops. Associational resistance is when neighboring plant species decrease the likelihood that vulnerable plants will be discovered or exploited by insect herbivores. This phenomenon can be explained by two non-exclusive hypotheses: 1) The resource concentration hypothesis states that increased plant diversity makes it more difficult for herbivores to detect their host plants, and 2) the natural enemy hypothesis suggest that natural enemies can more effectively colonize and reduce herbivores on crop plants in more vegetationally diverse environments due to increased structural diversity and the presence of alternate prey (Barbosa et al., 2009).  This research will advance our knowledge on living mulch usage by investigating responses of pests and natural enemies to their presence in cantaloupe plantings. Further, I will compare the effect of two perennial living mulches with structural differences on populations of pests, beneficials, over wintering natural enemies, and crop yield. Past studies have shown that leguminous cover crops (clover) can be used to reduce pest colonization and tenure time and that perennial grasses can serve as refugia for overwintering natural enemies (Bowers, Toews, Liu, & Schmidt, 2020).


Materials and methods:

Experimental layout and design

To fulfil my objectives, field trials were conducted at the University of Maryland’s Central Maryland Research and Education Center in Upper Marlboro, MD. The experiment consisted of three treatments: 1) cantaloupe interplanted into alsike clover living mulch, 2) cantaloupe interplanted into Virginia wildrye living mulch, and 3) monoculture cantaloupe (grown in bare-ground). The experiment was replicated four times and arranged in randomized complete block design. Each plot measured 12.1 m by 12.1 m and was separated by 7.6 m alleys. Within each plot there were eight rows of cantaloupe with plants spaced 91 cm apart within rows and a between row spacing of 1 m.  Each row consisted of 13 cantaloupe plants each (Figure 1).

Cover crop plot diagram
Figure 1: Diagram of cover crop treatment plot

Alsike clover and Virginia wildrye were seeded in their respective treatment plots with a no-till drill in late September 2019 for the 2020 field season, as well as September 2020 for the 2021 field season. Monoculture plots were left fallow over the winter. Sugar cube was the variety of cantaloupe used, partially due to its high level of disease resistance. Cantaloupe seedlings were grown out in a greenhouse for two weeks before they were transplanted into the field on June 30th. One week before the cantaloupe was transplanted, monoculture plots were rotary tilled, and living mulch treatment plots were be mowed with a rotary mower to reduce potential early season competition with the cantaloupe and strip tilled approximately 38 cm wide where cantaloupe rows were established. The remaining interrow areas consisted of the alsike clover of Virginia wildrye.


Objective 1: Compare treatment impact on insect pest (herbivore) abundance

Several monitoring techniques were used to quantify herbivore abundance. The primary monitoring techniques included visual counts on the cantaloupe foliage and the deployment of yellow sticky cards. For visual counts, a square meter quadrat was randomly placed in each interior cantaloupe row with a minimum of one meter from the border. All arthropods encountered on the foliage or visible on the ground within the quadrat were field-identified to the lowest taxonomic level possible and recorded. Insect herbivores that were expected to colonize the cantaloupe included striped and spotted cucumber beetles, squash bugs, and melon and green peach aphids. Visual counts were conducted every week initiating two weeks after planting up until the beginning of harvest. Visual counts were always conducted in the mornings close to the same time and the order of reps was randomized to maintain observational integrity.

Yellow sticky cards were used mainly for detecting aerial arthropods at or above the cantaloupe canopy. They were deployed in the center row of each plot in the inter-row areas. Cards were held in place on fiberglass poles with a pair of clothes pins at canopy height. Sticky cards were deployed two weeks after planting and remained in place for one week prior to collection. Cards were deployed on 3 occasions spaced 3 weeks apart to provide an image of insect abundances in the early, middle, and late stages of the cantaloupe development. Sticky card specimens were counted and identified to the lowest taxa possible. An advantage of using sticky cards is that they can remain in the field for an extended period of time and continue to capture insects throughout the day and evening. Once collected, cards were placed in zip-tied bags and stored in the lab freezer until their contents could be identified using a stereomicroscope.

Objective 2: Compare treatment impact on the abundance of natural enemies

Natural enemies were counted along with herbivores during sampling tasks described in objective 1. However, in addition to foliar and sticky card counts, pitfall traps were be used to quantify the activity density of ground dwelling predators, including wolf spiders, rove beetles and ground beetles. Two pitfall traps were installed in each plot, one in the within (intra) and between (interrow) cantaloupe row areas. Pitfall traps consisted of two nested plastic cups placed flush with the surface of the ground. The top cup was removable and partially filled with propylene glycol as the trapping/killing agent. A 30x30 cm square black plastic cover supported by three large screws was placed over each trap with two to three centimeters of clearance to prevent traps from being flooded by rain or disturbed by wildlife. Pitfall traps were deployed and retrieved on the same dates as sticky cards. Once trap contents were retrieved, samples were brought back to the lab and vacuum filtered to separate them from the propylene glycol. Specimens were then transferred to a 70% ethanol solution to be preserved until they could be identified under a stereomicroscope.

Analysis (Objective 1 and 2)

Important or frequently appearing arthropod taxa observed across the sampling period are being analyzed. Visual counts were converted into number per square meter densities. All quadrats within a plot were averaged to represent the estimated density of arthropods for that entire plot for the given sampling period. Densities of notable arthropod species were compared over time between treatments using a linear mixed-effect model, and where treatment effects were detected pairwise comparisons of each treatment on each date were conducted to determine where the treatment effects were arising from. Results of visual counts and yellow sticky cards were also compared with action thresholds of an average of one cucumber beetle per plant or 15 beetles per sticky card. If there is no effect of pitfall trap location (intra vs inter-row) content of both traps will be pooled otherwise comparison will be made by trap location.

Objective 3: Determine treatment impact on yield and fruit quality

During each harvesting event, the total number and weight of harvestable and marketable fruit more than 4ft from the plot border was measured. Harvested fruit was evaluated by quantifiable measures of fruit quality by sampling 4 fruit of equal ripeness from each experimental plot during a single harvesting event. These measured characteristics included weight, polar and equatorial diameter, soluble solid (sugar) content, titratable acidity, skin and flesh hue, chroma and lightness, as well as texture characteristics including firmness, adhesiveness, springiness, cohesiveness, chewiness, and resilience. Treatment effects were checked for using a one-way ANOVA, and when treatment effects were detected a Tukey-Kramer test was completed to determine which treatments were statistically unique from each other.

Objective 4: Determine the potential of different treatments to serve as overwintering refugia for natural enemies 

After harvest, the living mulch remained undisturbed in the plots through the fall and winter into 2021. In March 2021, two 60cm x 60cm soil emergence traps (cages) were left in place in two randomly selected interrow areas in each plot. Trapping will lasted 10 weeks up through May. Collection jars in the top of each emergence cage were partially filled with propylene glycol. Trap contents were collected and replaced weekly. After collection, contents were filtered and transferred to a 70% ethanol solution for storage until they could be identified to the lowest possible taxa. Arthropod counts will be converted into number per 0.36 m2 (area covered by the trap). Traps within the same plots will have their weekly values averaged. Weekly insect trap content for each treatment will be compared using a linear mixed effect model and total seasonal emergence will be compared between treatments.

Research results and discussion:

Research results and discussion:

The study is currently ongoing, and data and trap captures are still being analyzed. However, insects captured during the 2020 and 2021 field seasons have largely been identified and are in the process of being analyzed. All sticky cards and pitfall traps samples were sent to an entomologist with greater taxonomic experience to verify ID’s and ensure consistency in identifications. Pitfall samples are still undergoing processing, but sticky cards have been returned. It is hoped with this higher resolution data that analyses regarding the diversity and compositions of the insect communities may be conducted.

Objective 1: Compare treatment impact on insect pest (herbivore) abundance

Overall, both the 2020 and 2021 field seasons saw relatively low abundances of major pest insects across the cantaloupe growth period. Striped cucumber beetles (Acalymma vittatum), the most damaging cucurbit pest, never surpassed their action threshold of 1 insect per square meter in any of the experimental treatments for either year, including the control. Striped cucumber beetles did however appear to respond somewhat to the living mulch treatments.  A linear-mixed effects model revealed a treatment response for striped cucumber beetles using both foliar counts and sticky cards, and pairwise comparisons of treatments revealed which treatments bore statistically different insect abundances. In 2020 foliar counts the treatment effect was significant (p=0.03994) with clover having fewer cucumber beetles than the conventional treatment (p=0.0369) (Figure 2). Foliar counts in 2021 were just shy of significant (p=0.05884), though daily averages of beetles in the conventional treatment consistently higher than in the other two treatments (figure 3).


Striped cucumber beetle foliar detections, 2020
Figure 2: Striped cucumber beetle detections via foliar counts in the 2020 field season.    
Striped cucumber beetle foliar counts 2021
Figure 3: Striped cucumber beetle detections via foliar counts in the 2021 field season.


For the sticky card captures in 2020 there was no treatment effect in the inter row space (the living mulch strip between the rows of cantaloupe) (p=0.4333) though there was a significant treatment effect in intra row space (the cantaloupe rows themselves)(p=0.0201). (Figure 4) Beetles reduced most from levels in the conventional treatment in wildrye treatment (p=0.0279). (Figure 5)


Striped cucumber beetles, 2020 inter-row
Figure 4: 2020 inter row captures of striped cucumber beetles on sticky cards.
row striped CB detections 2020
Figure 5: 2020 intra row captures of striped cucumber beetles on sticky cards.


In contrast to the 2020 field season, in 2021 there was no significant effect of treatment in either the inter-row space (p=0.3502) (Figure 6) or intra row space (p=0.7558) (Figure 7) on striped cucumber beetle abundances. In fact, the results between treatments was virtually identical. 

2021 Striped CB intra sticky cards
Figure 6: 2021 inter row captures of striped cucumber beetles on sticky cards.
Striped CB 2021 intra row sticky cards
Figure 7: 2021 intra row captures of striped cucumber beetles on sticky cards.

Aphids, as the other major vector of cantaloupe diseases, are also worth highlighting. Unlike with the striped cucumber beetles, there were no significant treatment effects on aphid detections via foliar counts in either 2020 (p=0.2067) (Figure 8) or 2021 (p=0.2893) (Figure 9). Overall aphid detections incredibly low however. In 2020 there was less than one aphid detected on average in every quadrat. 

2020 Aphid foliar counts
Figure 8: Aphid detections via foliar counts in the 2020 field season.
2021 Aphid foliar counts
Figure 9: Aphid detections via foliar counts in the 2021 field season.

For sticky cards, in 2020 the treatment effect on aphid captures in the inter-row position was significant (p=0.01563) (Figure 10) but looking at detections across time this effect appears limited to the final date. The intra row position similarly had a significant treatment effect (p=0.04965) (Figure 11), with wildrye having the fewest aphids detections on each date.

2020 Aphids inter row cards
Figure 10: 2020 inter row captures of aphids on sticky cards.
2020 aphid intra card
Figure 11: 2020 intra row captures of aphids on sticky cards.


In 2021 there was a significant treatment effect on aphid detections in the inter-row position (p=0.002166) (Figure 12) with wildrye having significantly fewer aphids than clover (p=0.0351) and control (p=0.0069). There was also an effect in the intra-row position (p=0.00794), with wildrye being statistically different from clover (p=0.0182) (Figure 13).

2021 aphid cards inter
Figure 12: 2021 inter row captures of aphids on sticky cards.
2021 aphids intra cards
Figure 13: 2021 intra row captures of aphids on sticky cards.


Objective 2: Compare treatment impact on the abundance of natural enemies

Like herbivores, the abundance of natural enemies was in general fairly low across both 2020 and 2021 field seasons as well. Analysis was carried out in an identical manner to pest species, with a linear mixed effects model determining if treatment effects were present, and pairwise comparisons between treatments determining the nature of those effects. Spiders, one of the more important groups of natural enemies, did show a significant treatment effect in their foliar count detections in 2020 (p=0.004294)  (Figure 14) with a significant difference between clover and wildrye (p=0.006), as well as a very notable effect in 2021 (p=4.545e-11) (Figure 15) with clover being statistically distinct from both wildrye (p=<.0001) and control (p=<.0001) treatments. Spiders were not detected in large numbers on sticky cards, however there were many detections in pitfall traps. Pitfall trap sample content is being analyzed by a scientist highly knowledgeable in taxonomy in order to accurately ID significant cucumber beetles predators, especially wolf spiders and parasitic wasps and flies, so pitfall trap results will not be shared at this time.

2020 Foliar spiders
Figure 14: Spider detections via foliar counts in the 2020 field season.
Spider foliar 2021
Figure 15: Spider detections via foliar counts in the 2021 field season.


As with many insects, parasitism poses a significant top down control on most cantaloupe pests, so insects identified as parasitoids (most of which were wasps) were counted on sticky cards to determine if this highly important group of biocontrol agents were affected by the treatments at all. Their specific identities will be investigated to determine if the taxa present contribute to control of important pests such as aphids and cucumber beetles. In 2020 there was not a significant treatment effect on parasitoid detections on sticky cards in the inter-row position (p=0.1845) (Figure 16), there was however an effect in the intra row position (p=0.005877) (Figure 17), with a significant difference between the clover and conventional treatments (p=0.0096). 

Parasitoids 2020 inter cards
Figure 16: 2020 inter row captures of parasitoids on sticky cards.
2020 intra parasitoids
Figure 17: 2020 intra row captures of parasitoids on sticky cards.


Despite an impact in the 2020 intra-row position, this pattern didn’t carry over to the 2021 field season. In 2021 there wasn’t a significant treatment effect on parasitoid detections in either the inter (p=0.8889) (Figure 18) or intra row (p=0.3283) (Figure 19) positions. (Figure 10)

2021 parasitoids inter row
Figure 18: 2021 inter row captures of parasitoids on sticky cards.
2021 parasitoids intra
Figure 19: 2021 intra row captures of parasitoids on sticky cards.

Objective 3: Determine treatment impact on yield and fruit quality

In neither 2020 or 2021 was there much of an impact of treatments on most of the quality metrics the cantaloupe were subjected to. There were minor differences in some color metrics, but they would be imperceptible to any consumer. More importantly however, there were issues with yield reduction during both years. In 2020 the Virginia wildrye plots saw a statistically significant reduction in yield (p=0.0028644) (Figure 2020). The clover plots fared better but they were still on the cusp of yielding a statistically lower total average weight between them and the control plots. In 2021 the living mulch treatments had a much more dramatic impact on yield, with clover (p=0.000003) and rye (p=0.0000058) having much lower total yield than control. (Figure 21)

Yield 2020
Figure 20: Average yield by treatment in 2020
Yield 2021
Figure 21: Average yield by treatment in 2021

Objective 4: Determine the potential of different treatments to serve as overwintering refugia for natural enemies 

There currently isn't any data related to the overwintering communities of natural enemies in the plots, but overwintering traps were be deployed in March and samples were collected over a two month period. Sample identification will be carried out and the analysis of the abundance and density of notable taxa will be measured and analyzed using similar methods to the insect surveys conducted via foliar counts and sticky and pitfall trap captures over the past two summers.


Research is still ongoing, and it is still premature to come to any broad conclusions. There has so far been an inconsistent impact of the living mulch treatments on the abundance important pests and natural enemies in the plots. With some modification, it is possible that this technique may impart some benefit, and the impact of the treatments on spider abundances appears promising. Broad conclusions regarding the impact of the treatments may be difficult to determine however due to the generally low abundance of arthropods in the environment for either year.

The consistence in quality of the fruit across the treatments is encouraging. The yield reductions, however, are an element that need to be addressed if this or a similar practice is ever going to be adopted in cantaloupe. It was determined that yield reductions in 2020 were most likely a product of competition with weeds in the early growth stages of the cantaloupe. The first 30 days of the cantaloupes' growth cycle is essential for determining the ultimate yield of the plant. When the living mulch was mowed prior to transplanting the cantaloupe to the field it took much longer than expected for it to recover and reestablish its weed suppressing characteristics.

To ameliorate this, in the summer 2021 the seeding rate of the living mulch was increased by 50 percent to improve the living mulches' density and reduce the chances of weeds growing among them and spreading into the cantaloupe rows. A feature of the living mulch varieties that helped inform their selection was that they were lightly competitive and usually go into a low-growth semi-dormant state during the height of the summer. Unfortunately, an uncharacteristically wet second half of July stimulated the living mulches to start growing rapidly again, and this likely contributed to increased competition with the cash crop. While we will continue to look into the data to better understand the living mulches' impacts on the arthropod communities of the fields, there will likely need to be further changes made to the methodology if growers are to ever adopt a similar practice.



Participation Summary

Education & Outreach Activities and Participation Summary

3 Webinars / talks / presentations

Participation Summary:

Education/outreach description:

While the Covid-19 pandemic disrupted many of the meetings that took place over the last two years, I was able to take advantage of several successful virtual outreach opportunities as well as speak in person at several meetings to share the objectives and current findings of this ongoing project with members of the agricultural and scientific community.

In place of the annual University of Maryland Central Maryland Research and Education Center's (CMREC) annual twilight wagon tour event which was canceled due to the pandemic in 2020, I gave a short talk on the progress of the project for the University of Maryland Extension - Charles County Youtube channel,  which is titled Developing a perennial living mulch system to manage cantaloupe pests. I also virtually presented the project's current progress to the Northeast IPM Research Update Conference on November 17, 2020 and the Entomological Society of America 2020 Virtual Meeting. In the latter I was entered in the student 10-minute paper presentation competition, where my presentation on this project was awarded first place in the IPM-Horticulture category. My presentation was available on-demand to attendees in recorded format for the duration of the conference from November 11-25, which was widely attended by scientists from across the country. I also intend to submitted a poster presentation of my project to the Mid-Atlantic Fruit and Vegetable Convention's poster session from February 8-11, 2021 which won second place in the student competition.

By August 2021 in person events had largely returned, and I presented my research in person at the UMD CMREC Twilight tour. I was also able to present at the Entomological Society of America's national meeting in Denver CO on November 1st, 2021, and more recently I gave presentations at the Frederick MD, and Denton MD Pesticide Applicator Conferences on both February 1st and 3rd, 2022 respectively.  I currently have plans to present the results of this project at the Entomological Society of America's Eastern Branch Meeting on April 26th 2022.

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.