A Comparison of Row Cover Materials for Use in Excluding Major Insect Pests from Cucurbit Crops

Final report for OS18-120

Project Type: On-Farm Research
Funds awarded in 2018: $14,933.00
Projected End Date: 03/14/2021
Grant Recipient: Oklahoma State University
Region: Southern
State: Oklahoma
Principal Investigator:
Dr. Eric Rebek
Oklahoma State University
Co-Investigators:
Dr. Jim Shrefler
Oklahoma State University
Expand All

Project Information

Abstract:

Vegetable production is an important component of various types of farms in Oklahoma and surrounding states, and farm sustainability depends on achieving consistent production of a marketable crop to have continuity in the market. The long, warm growing season in the region makes cucurbits a well-adapted enterprise. However, cucurbit crops are not free of production risks, and a general hazard is that posed by insect pests. For some, effective control measure are available and typically involve the use of synthetic insecticide applications. For others, control measures can be either difficult to employ, of limited efficacy, or non-existent.

Recent research (Driever et al., 2016) showed that row covers can be used to exclude pests, and at the same time, provide crop access to insect pollinators, thereby enabling production without using insecticides. Although effective, row cover materials have several drawbacks, including high temperatures under cover that could interfere with fruit set, insect pest entry through open covers, and a high incidence of disease, such as powdery mildew. Thus, while the development of this pest management technique for cucurbit crops would add to the sustainability of farms in the region by improving the predictability of producing marketable cucurbits, we need to find a suitable row cover material that will not only exclude insect pests, but one that will be convenient to use, not be conducive to diseases, and be cost effective.

We propose to compare several products for use as row covers for the explicit purpose of excluding insect pests from cucurbit crops, specifically squash.

Project Objectives:

We will determine the effectiveness for excluding insect pests, effects of cover materials on the incidence of diseases, and the influence of the row cover materials on crop microenvironment and light transmission.

We will conduct field trials over two years at two locations using the treatments: no row cover as an untreated control; a frost blanket type material; a woven mesh netting that is used for insect exclusion in the tree fruit industry; and a formed mesh material used for insect exclusion purposes.

Squash plants will be examined weekly for insect pest pressure, fruit set, disease incidence and other possible treatment effects.

Cooperators

Click linked name(s) to expand
  • Dr. Tracey Payton (Educator and Researcher)
  • Dr. Lynn Brandenberger (Educator and Researcher)
  • Dr. Bizhen Hu (Educator and Researcher)

Research

Materials and methods:

We established the row cover trial at two locations in Oklahoma, Langston and Stillwater, in 2019. In 2020, we conducted the trial at these locations and added a third site at an experiment station in Lane, Oklahoma. The Langston site was a field planted previously in sweet potato and a cover crop of radish and clover. The Stillwater site was a vegetable garden that had been planted with tomatoes and other vegetable crops in previous years. The Lane site was previously planted with a mix of vegetable and leafy green crops, but had not been planted to cucurbits for many years. Due to space restrictions in both years, we had to modify Objective 3 and only plant squash once at both sites. Squash ('Lioness' summer squash) was planted in the center of each row, placing 12 seeds on 1-foot centers. Each plot measured 15 feet long by 4 feet wide. Squash plants were later culled so that six healthy plants remained in each plot. Due to saturated field conditions in 2019, squash planting was delayed until late May at the Langston and Stillwater sites. In 2020, planting was done in late April at all three sites. Treatments consisted of the following: 1) heavy fabric - DeWitt row cover deluxe plus 1 oz material; 2) woven mesh fabric - and 3) light fabric - DeWitt row cover 0.5 oz material. Treatments were compared against control plots (no row covers). At each site, all treatments and the control were replicated three times for a total of 12 plots, arranged in a completely randomized design. Each plot was covered with one of the three row cover fabrics corresponding to treatment, except for control plots which were left uncovered. Assail 30SG (acetamiprid) was applied at a rate of 5.3 oz/acre to plants in control plots for early season control of squash bugs and squash vine borer. These were protected with insecticide as a precaution against losing young plants under heavy insect pressure. A pre-emergent herbicide was applied at each site to reduce pressure from annual grasses and annual broad-leaf weeds.

Row covers were constructed from the treatment fabrics listed above and aluminum tubes measuring 1-inch in diameter. Each tube was bent to form a U-shaped frame. Three frames were erected in each treatment plot, one in the center and two at each end of the plot, and secured to the ground by inserting each end into rebar, which was driven into the ground at a depth of 12 inches. Row cover fabrics were cut to sufficient length for full coverage of treatment plots and stretched over the metal frames, allowing enough slack to prevent tears. Fabric was secured to the metal frames using 2-inch binder clips and covered on all sides with enough soil to prevent insects from crawling under the fabric.

At two weeks after flowering began (i.e., 50% of squash plants with female flowers), row covers were opened on the long sides of each plot for a period of 2 hours every weekday morning (approximately 0700 to 0900 h) for a total of 3 weeks, allowing pollinators access to the flowers. Insect sampling occurred twice weekly after plants were just beyond seedling stage. Counts of squash bug eggs, nymphs, and adults were made for each plant within a plot, then summed across plants for each life stage. Other insects were recorded as they were encountered, including aphids, cucumber beetles, and squash vine borers. These insects were eliminated by hand removal to minimize interference with squash bug activity within each plot. Pollinators (e.g., squash bees) were also recorded as they were encountered. Insect counts were recorded in all plots throughout the summer, terminating in mid-August as squash plants began to senesce and not yield fruit. During fruit production, fruits that had reached marketable size were harvested from each plot approximately every 3 to 5 days. Marketable fruits were separated from culls (misshapen, damaged, rotten, etc.), and the total number of marketable fruits, total weight of marketable fruit, and mean weight per plot were recorded for each plot. Harvested squash were donated to a local food bank.

For each plot, data for squash bug abundance were pooled across all sample dates, done separately for each life stage (egg, nymph, adult). Because some plants died over time, insect counts were standardized across plots by dividing the total number of each life stage by the number of plants present in the plot for each sample date. Similarly, harvest data (total fruit, total weight, and mean weight per plot) were pooled separately for each plot across all sample dates. Prior to analysis, pooled data were square-root transformed to meet assumptions of normality for parametric statistical testing. Data were analyzed using one-way analysis of variance (ANOVA) with treatment as the main effect (PROC ANOVA, SAS 9.4). Means separation testing was performed using Tukey's Honestly Significant Difference tests. All analyses were performed at α=0.05.

Research results and discussion:

Squash Bug Data

Squash bug densities varied across sites and years, presenting the opportunity to investigate the utility of each fabric type exposed to varying insect pressure. At the OSU-Stillwater Botanic Garden, we found significant differences among treatments and the control in 2019 (Table 1), a year punctuated by a very low density of squash bugs. A higher density of squash bugs was found in uncovered control plots compared to plots covered with the heavy, woven, and light fabrics (Tukey's HSD, P≤0.05). One year later, squash bugs were much more abundant at the Stillwater site, and high variability among plots resulted in no significant differences among fabric treatments and the control for any life stage (Table 2).

Table 1. Mean squash bug abundance per plant (± S.E.) for each life stage at the OSU Botanic Garden in Stillwater in 2019. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Mean no. squash bugs per plant per plot (± S.E.)
Adults Nymphs Eggs
Control 1.7 (0.2) a 30.7 (13.4) a 239.8 (68.0) a
Heavy 0.6 (0.4) ab 15.2 (3.5) ab 29.9 (7.3) b
Woven 0.2 (0.2) b 1.9 (0.6) bc 45.9 (26.0) b
Light 0.1 (0.1) b 0.1 (0.1) c 11.1 (8.4) b

Table 2. Mean squash bug abundance per plant (± S.E.) for each life stage at the OSU Botanic Garden in Stillwater in 2020. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Mean no. squash bugs per plant per plot (± S.E.)
Adults Nymphs Eggs
Control 92.2 (14.4) a 341.3 (85.5) a 424.4 (54.1) a
Heavy 38.1 (11.9) a 138.1 (50.2) a 302.8 (39.8) a
Woven 89.7 (38.8) a 143.9 (22.4) a 387.4 (56.3) a
Light 43.1 (15.0) a 216.2 (12.1) a 480.5 (39.5) a

At Langston in both years, we did not find any significant differences among treatments and the control for season totals of squash bug eggs, nymphs, and adults (Tables 3 and 4). However, there was a trend toward greater abundance of adults and nymphs in control plots compared to those under any row cover in 2019. This site was characterized by heavy squash bug pressure compared to the Stillwater and Lane sites. The Langston site was in an open area exposed to high winds, which damaged the fabric covers in some plots, allowing insects access to the squash plants until repairs could be made during the next sampling date. In 2020, squash plants did not perform well, requiring frequent reseeding and ultimately resulting in crop failure by early August.

Table 3. Mean squash bug abundance per plant (± S.E.) for each life stage at Langston in 2019. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Mean no. squash bugs per plant per plot (± S.E.)
Adults Nymphs Eggs
Control 6.8 (2.7) a 97.5 (46.5) a 259.7 (134.8) a
Heavy 3.6 (1.9) a 66.1 (38.8) a 127.6 (62.7) a
Woven 2.6 (0.7) a 39.8 (22.8) a 125.4 (54.7) a
Light 5.7 (2.6) a 90.4 (45.9) a 110.6 (38.1) a

Table 4. Mean squash bug abundance per plant (± S.E.) for each life stage at Langston in 2020. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Mean no. squash bugs per plant per plot (± S.E.)
Adults Nymphs Eggs
Control 19.9 (4.2) a 42.3 (13.6) a 132.7 (75.8) a
Heavy 9.4 (5.2) a 31.9 (11.3) a 175.3 (86.6) a
Woven 16.1 (8.0) a 111.8 (52.2) a 140.5 (61.8) a
Light 7.1 (2.1) a 75.3 (1.5) a 198.8 (33.7) a

At the Lane site, squash bug abundance was too low to detect any significant differences among treatments and the control (Table 5).

Table 5. Mean squash bug abundance per plant (± S.E.) for each life stage at Lane in 2020. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Mean no. squash bugs per plant per plot (± S.E.)
Adults Nymphs Eggs
Control 2.0 (0.4) a 0.2 (0.1) a 0.9 (0.7) a
Heavy 1.2 (0.6) a 0.2 (0.1) a 0.3 (0.3) a
Woven 1.9 (0.3) a 0.5 (0.3) a 0.6 (0.1) a
Light 1.4 (0.3) a 0.3 (0.2) a 0.3 (0.1) a

Harvest Data

In Stillwater in 2019, we did not find significant differences among row cover treatments and the control for total number of marketable fruits and their total weight (Table 6). However, there was a trend toward more fruits and higher overall weight in plots under any of the row cover materials. The average weight of marketable fruits (i.e., total weight divided by total number) was significantly reduced in the heavy fabric treatment (Table 6), indicating that the heavy fabric negatively affects the average size of marketable squash. In 2020, there was clear separation among treatments and the control for all harvest data in Stillwater (Table 7). The total number of fruits and total weight were greatest in the woven fabric treatment and reduced in the heavy fabric treatment. The average weight of marketable fruits was lowest in the heavy fabric treatment. Additionally, there were many culls in the heavy fabric plots due to mold and rot, likely the result of high humidity that was observed in those plots.

Table 6. Harvest data at the OSU Botanic Garden in Stillwater in 2019. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Means per plot (± S.E.)
No. of fruits Total weight (lbs) Mean weight (lbs)
Control 59.0 (4.6) a 38.6 (3.4) a 10.0 (0.6) a
Heavy 85.3 (7.1) a 48.4 (6.3) a 6.1 (0.4) b
Woven 86.7 (3.3) a 60.4 (5.3) a 9.8 (0.5) a
Light 97.0 (26.4) a 59.1 (16.7) a 7.6 (1.0) ab

Table 7. Harvest data at the OSU Botanic Garden in Stillwater in 2020. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Means per plot (± S.E.)
No. of fruits Total weight (lbs) Mean weight (lbs)
Control 32.0 (5.5) ab 21.7 (4.5) ab 6.4 (0.8) a
Heavy 6.0 (2.3) c 3.3 (1.1) c 1.5 (0.4) b
Woven 60.7 (9.8) a 39.6 (6.4) a 5.3 (0.8) a
Light 20.0 (3.8) bc 12.4 (2.4) bc 3.7 (0.7) ab

At the Langston site in 2019, we found no differences among the treatments and control for any of the harvest variables (Table 8). However, there was a trend toward more fruit produced in plots covered by the woven material. In 2020, plant performance was so poor among all plots that very few fruits were produced. Most fruits produced in 2020 were underweight and/or culled due to rot.

Table 8. Harvest data at Langston in 2019. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Means per plot (± S.E.)
No. of fruits Total weight (lbs) Mean weight (lbs)
Control 11.3 (6.8) a 5.8 (4.5) a 0.4 (0.1) a
Heavy 11.3 (6.4) a 13.5 (11.5) a 0.7 (0.5) a
Woven 35.3 (17.9) a 28.7 (22.6) a 0.6 (0.2) a
Light 11.7 (8.8) a 7.3 (6.2) a 0.4 (0.2) a

In 2020, the Lane site produced more squash than the Stillwater and Langston sites. Similar to the Stillwater site, the heavy fabric treatment had significantly fewer fruits and the least total weight compared to the control (Table 9). Among the fabric treatments, plots covered by the woven material produced the most fruit and the highest total weight, although these measures were not significantly different from the light fabric treatment. Control plots produced the most fruit and highest total weight overall, but the control was not significantly different from the woven fabric treatment. The average weight of marketable fruits did not differ significantly among the control and row cover treatments. Yield in the control plots was not impacted by insect activity due to very low squash bug densities in 2020.

Table 9. Harvest data at Lane in 2020. Data for each plot were pooled across all sample dates. Means marked with the same letter within a column are not significantly different according to Tukey's HSD (P>0.05).

Treatment Means per plot (± S.E.)
No. of fruits Total weight (lbs) Mean weight (lbs)
Control 125.3 (4.4) a 71.7 (1.6) a 9.5 (0.2) a
Heavy 49.7 (6.2) c 31.0 (4.8) c 7.5 (1.1) a
Woven 89.3 (8.2) ab 52.7 (5.2) ab 8.6 (0.2) a
Light 63.0 (7.6) bc 35.2 (4.7) bc 7.6 (0.7) a

Conclusions

Due to high variability in squash bug density, differences among row cover treatments and the control were usually not significant, especially under high insect pressure. In general, however, there was a trend toward higher squash bug densities in the control plots that were not under row covers. Additionally, squash bugs colonized control plots earlier than plots with row covers, and there was a higher degree of plant mortality in the control plots by the end of the season. Squash bugs were able to colonize all plots because row covers had to be opened daily during bloom to allow pollinators access to the flowers. Thus, row covers are of limited utility during this critical stage of fruit development, and their use should be combined with other management strategies such as physical destruction of eggs on leaves, hand picking of adults and nymphs, and delayed planting dates to reduce insect pressure.

The purpose of using row covers is to exclude insect pests without impeding plant growth and yield. However, the light and heavy fabric (i.e., DeWitt 0.5 oz and 1 oz. row cover materials, respectively) reduced marketable yield compared to the control. Especially in Stillwater, a higher number of non-marketable, moldy fruits were observed under the heavy and light fabrics. Thus, the light and heavy fabrics reduced the quantity and quality of marketable fruit. Additionally, plant mortality increased toward the end of the season in plots covered by the heavy material. The woven mesh cover increased marketable yield and did not lead to increased moldy fruit compared to the other row cover materials. These differences in fruit quality among fabric treatments are likely due to a higher degree of humidity observed in the plots covered by light and heavy fabric compared to the woven mesh plots. Additionally, the light row cover material tore easily under high winds and storms, requiring frequent repairs and/or replacement. The woven mesh material was sturdy and breathable and did not retain excessive moisture.

In conclusion, we recommend the use of woven mesh fabric in the design of row covers for pest exclusion. Woven mesh fabric excludes insect pests (except when opened for pollinator access), does not trap excessive moisture, does not reduce yield, and is sturdy in windy and stormy weather. Row covers constructed from woven mesh fabric should be adopted as one of several strategies used in a comprehensive integrated pest management program for squash bugs.

Participation Summary

Educational & Outreach Activities

5 Consultations

Participation Summary:

Education/outreach description:

To date, we have consulted directly with approximately 5 cucurbit growers on the use of row covers for excluding squash bugs and other insect pests. Due to COVID-19, the state Master Gardener Conference scheduled for summer 2020 was cancelled. This venue was intended to be the first presentation of our research results. However, results will be presented at the 2021 state Master Gardener Conference held virtually this June and at the Oklahoma-Arkansas Horticulture Industries Show scheduled in January 2022. We will use survey instruments to evaluate knowledge gained among gardeners and commercial growers attending these conferences. In addition, research results will be shared with Oklahoma Cooperative Extension Educators in mid May 2021 during the Horticulture Update, a monthly webinar conducted for in-service credit. Research results will also be shared at the Annual Meeting of the Entomological Society of America in November 2021 and the International Integrated Pest Management Symposium in March 2022.

We will produce two extension publications in 2021 focusing on row covers as part of an integrated pest management approach to managing squash bugs. These publications will be a Pest e-Alert, produced by the Department of Entomology and Plant Pathology at Oklahoma State University, and a more detailed fact sheet. Also, we will write a peer-reviewed publication to be submitted to a yet-to-be-determined horticulture or entomology journal. 

Project Outcomes

1 Grant received that built upon this project
2 New working collaborations
Project outcomes:

Cucurbit growers in Oklahoma and elsewhere in the southwestern U.S. will benefit from this research by gaining knowledge on the effectiveness of row covers in reducing squash bug abundance on squash plants. They will also learn about the best materials (i.e., woven mesh fabric) to use in construction of the row covers to reduce pest pressure and maintain yield. Combined with other integrated pest management tactics for squash bug management, growers will reduce the use of broad-spectrum insecticides. In turn, reduced reliance on chemical management will save growers money, protect soil and water, and minimize harm to arthropod pollinators and natural enemies.

Recommendations:

In a companion study funded separately, we assessed the optimal timing of row cover removal (or opening of the fabric) for allowing pollinators access to flowers during bloom. Our preliminary data indicate that waiting to remove or open row covers until two to three weeks following 50% bloom reduces squash bug abundance in the crop without sacrificing squash yield. Thus, cucurbit growers will benefit from using row covers constructed from woven mesh fabric combined with the proper timing of row cover removal. 

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.