Impacts of Silage Tarps on Soil Arthropods, Soil Properties and Crop Yields

Progress report for GNE19-205

Project Type: Graduate Student
Funds awarded in 2019: $15,000.00
Projected End Date: 11/30/2022
Grant Recipient: University of Vermont
Region: Northeast
State: Vermont
Graduate Student:
Faculty Advisor:
Gillian Galford
University of Vermont
Expand All

Project Information

Project Objectives:

The main objective of this project is to enhance sustainable agriculture pursuits in New England by ensuring that management practices will not harm biological communities and the ecosystem services they provide. The specific research objectives of this project are to 1. Measure the effects of silage tarps on soil arthropod diversity and soil properties; 2. Measure the effects of silage tarps on crop outcomes of relevance and interest to the farmers; and 3. Determine the relationships between soil arthropod diversity, soil properties, and crop properties. 



The purpose of this project is to determine the ecological and agronomic impacts of silage tarping. Silage tarps are impermeable plastic films that are placed over crop beds with the primary purpose of reducing weed prevalence. They are widely used in the Northeast United States and can reduce tilling needs. However, while silage tarps have had promising initial results, there is very little research regarding their impact on soil ecology.

In particular, the impacts of silage tarps on soil arthropods (organisms such as beetles, mites, springtails, ants, and millipedes) is largely unknown. This is concerning because soil arthropods make up an important part of the soil community and play important roles in ecosystem functioning. Through actions such as carbon mineralization and fragmentation, aggregation, litter translocation, soil structure creation, and interactions with microbes, soil arthropods can contribute to agricultural sustainability and soil restoration.  

Given the contributions of soil arthropods to soil health, it is important to be wary of the impacts of silage tarps on their populations. It is unideal if silage tarping inadvertently causes soil degradation through thwarting soil arthropod communities. Monitoring the effects of silage tarps on soil arthropod diversity in a variety of soil types and over time are important steps in ensuring that this management decision can provide long-term benefits to farmers. This research will help guide farmers to making decisions that harmonize farm sustainability and environmental stewardship.


Click linked name(s) to expand/collapse or show everyone's info
  • Dr. V. Ernesto Mendez (Researcher)
  • Dr. Victor Izzo (Researcher)


Materials and methods:

Location and design:

Field work for this project took place in the summer of 2021 on three farms in Chittenden County, Vermont (Intervale Community Farm, Diggers’ Mirth Farm, and Catamount Farm). We tested two types of plastic tarps, silage tarps and clear plastic tarps, along with a hoed control. We laid out six randomly arranged replicates of these three treatments on each farms (18 plots on each farm; 54 plots total across the three farms). Treatment plots were 4.5m by 1.5m (1.5m was the approximate width of our crop beds), with a buffer space of 0.5m between plots. 

Before the tarp treatments were applied, all fields were tilled and prepared with fertilizer. We additionally watered the fields 1-2 days before applying the tarps to stimulate weed germination and increase effectiveness of tarps for killing weeds. We applied the tarps on May 28 and May 29th, and sealed the tarps edges by burying them. The tarps were left on the fields for 25 days. During this time, we hoed the control plots weekly to create stale seed beds. 

After tarps were removed in late June, we planted two rows of lettuce seeds in each plot using a Jang seeder. We did not remove weeds in any plots (treatment or control) for the remainder of the experiment. This reflects the reality of many small organic farms that do not have the labor or time for extensive weeding. 

Arthropod sampling:

We sampled arthropods 5 times throughout the field season: 1 week before tarping (to get a baseline of arthropod diversity), three weeks into the tarp treatment, and then 1, 3, and 5 weeks after tarps were removed (5 sampling periods, 54 treatment plots = 270 samples / method). 

We sampled arthropods using two methods: pitfall traps (for surface and litter organisms) and the Berlese Funnel method (for belowground organisms). For our pitfall trap method, we buried collection cups (95mm diameter lid, 120mm deep) in the center of each plot with their lids level to the surface. We used a killing/preserving agent of 50% propylene glycol (non-toxic anti-freeze) and 50% water, and collected organisms for 3 consecutive days. 

For the Berlese funnel extraction method, we took three randomly positioned soil cores (5 cm diameter, 10 cm deep) from each plot and aggregated them. We then extracted arthropods by placing collected soil in a funnel apparatus and exposing it to a 60 Watt light bulb for 72 hours, during which time organisms crawled downward through the funnel and fell into a collection vial with 95% ethanol. 

After collection, the specimens were classified to morphospecies. All organisms were classified at least to the order level, with certain organisms, including ants, beetles, and true bugs, classified at least to family and at times to genus or species. We then calculated richness separately for organisms caught in pitfall traps (aboveground arthropods) and Berlese funnel methods (belowground arthropods), and for each time period. We consider the pre-tarp sampling a “baseline” for arthropod richness, and subtract richness values from other samplings from this baseline to ultimately determine how tarps change richness from their baseline. 

Soil sampling:

At the beginning of the experiment, we collected soil to assess baseline soil properties on each farm, including pH, soil organic matter, phosphorus, and soil texture. During the tarp treatment, we measured soil temperature using iButtons, which log temperature every 30 minutes. In each plot, we buried two iButtons, one at the surface and one 10cm below the surface. We additionally manually measured soil moisture using a TDR-350 soil probe, which we used every time we collected soil arthropods. Finally, in each plot we measured soil nitrate before and after the tarp treatment to understand how tarps impact soil nutrient levels. These variables (soil temperature, moisture, and nitrate) were chosen because they represent effects of tarps, are of interest to farmers, and potentially correlate to arthropod diversity or crop outcomes.

Crop and weed sampling:

After tarp removal, we measured weed and crop growth weekly for five weeks. To sample weeds, we identified every present weed species in each plot and approximated its cover in the plot. From this, we calculated weed species richness as well as total weed coverage per plot. Total coverage could be greater than 100% if there were multiple layers of weeds present. 

To sample crops, we measured number of leaves, height, and widest leaf width of 10 randomly chosen lettuce plants in each plot. At the end of the 5 weeks, we harvested lettuce from each plot, counting individual lettuce heads and weighing total biomass per plot. 


All analyses were done in R version 4.0.4. We tested differences of soil temperature, soil moisture, soil nitrate, and crop yields between the treatments using an ANOVA with farm as a random effect. We tested differences of aboveground and belowground arthropods and weed coverage using a repeated measures ANOVA with again farm as a random effect. We made multiple comparisons among treatments using a least square means test adjusted using the Tukey method. 

Finally, we tested the relationship between arthropod diversity and environmental factors in our study system. We created four sets of linear models, representing aboveground and belowground arthropods when tarps were on the fields and after they were removed. Before running models, we eliminated highly correlated variables to avoid redundancy. We were left with four variables for the model when tarps were on (treatment, farm, soil moisture, and soil temperature) and five variables when tarps were removed (treatment, farm, soil moisture, weed coverage, and sampling time). We then multiple linear regression to test the importance of each variable to soil arthropod richness, reporting variable importance (average R2 from model averaging) and P values from type 3 ANOVAs. 

Research results and discussion:

Soil properties

Tarps significantly increased temperatures at the surface and 10cm below the surface (all P < 0.0001), with clear plastic causing higher surface and soil temperatures than silage tarps (both P < 0.0001; Table 1). Extremely hot temperatures under the clear plastic tarps likely relate to the greenhouse effect, where heat is trapped under the tarps. On the other hand, heat under silage tarps likely is due to black tarps absorbing heat. Higher temperatures under both tarps raise potential concerns if they create inhospitable environments for biological life and affect soil functioning.

Soil moisture results slightly varied by farm, though in general we found lower soil moisture under clear plastic tarps (both P < 0.01). Moisture was similar between the control and silage tarps (P = 0.86). Lower moisture under clear tarps may reflect condensation on the tarp surface, which was visible during the experiment. Many factors may contribute to moisture trends, including soil texture and recent precipitation and temperature patterns. 

Nitrate results also varied by farm. At Catamount Farm and Diggers’ Mirth Farm, tarps showed more positive change in soil nitrate availability than control plots. However, at ICF we found no differences between treatments. Overall, there were no significant differences, though clear plastic showed a slightly positive effect on soil nitrate (P = 0.09). These results may relate to fertilizer type, as each farm used a slightly different fertilizator. In particular, ICF used a slower-releasing fertilizer which may make the soil nitrate effects more delayed. An experiment testing soil nitrate availability for several weeks after a tarp treatment may better elucidate trends. 


Table 1: Impacts of tarp treatments on physical soil properties




Silage tarp

Clear plastic tarp

Surface temperature (°F) average (and highest recorded)

69.0 ± 0.2 (109.7)

78.0 ± 0.2 (127.9)

81.4 ± 0.3 (138.3)

Soil temperature (°F; 10 cm below surface): average (and highest recorded)

70.1 ± 0.1 (93.9)

75.6 ± 0.1 (100.2)

83.8 ± 0.1 (118.1)

Soil volumetric water content (%)

16.3± 1.2

16.6 ± 1.6

12.2 ± 1.1


Figure 1: Soil nitrate availability in treatments, separated by farm to highlight differences



In the aboveground samplings, we collected 83,140 total arthropods, with 37,255 from Catamount, 24,305 from Diggers’, and 21,580 from ICF. Most of the arthropods we captured were springtails (around 80% of the total), with Catamount in particular having high numbers of Entomobryomorpha (elongate-bodied) springtails. We also caught high abundances and diversity of beetles. In total, we found 125 morphospecies (75 of which were beetles) and 22 total orders. 

In the belowground samplings, we collected 915 organisms, including 40 morphospecies and 17 taxonomic orders. Again, springtails were the most abundant arthropod group (comprising 43% of the collected arthropods). Mites were also relatively common (around 35% of the total sample). 

Both silage tarps and clear plastic tarps caused negative effects on soil arthropod richness when tarps were on. For aboveground arthropod richness, we found on average 13 fewer arthropod morphospecies under tarped plots than control plots (all P < 0.001; Figure 2A). Comparatively, we found lower belowground arthropod richness under clear plastic tarps than in control plots (P = 0.01), but no significant difference between silage tarps and control plots (Figure 2B).

One week after the tarps were removed, aboveground arthropod richness remained significantly lower in the treatments that had been tarped (both P < 0.01; Figure 2A). However, by 3 and 5 weeks after tarps were removed there were no significant differences between treatments. For belowground arthropods, there were no significant differences 1 and 3 weeks after tarp removal. Interestingly, though, at 5 weeks after tarps were removed belowground arthropod richness in the control plots was significantly higher than in the two tarped treatments (both P < 0.005; Figure 2B).

These results demonstrate that tarps lead to immediate impacts on aboveground and belowground arthropod diversity. However, recovery of arthropods differed between the aboveground and belowground communities, with higher recovery for aboveground arthropods. This may be because aboveground arthropods are larger and more mobile, allowing them to migrate out of the tarped area when the tarps are down, and recolonize once the tarps have been removed. On the other hand, recovery might not be as quick for belowground arthropods, which are smaller and less mobile. Thus, they may be more directly affected by the tarps’ effects and suffer greater population casualties. The long term effect of continuous tarping on these communities and the impacts of tarps on pest dynamics remain uncertain. We encourage future research on tarps effects on soil arthropods to help answer these questions.  


Figure 2: Impacts of tarp application on aboveground (A) and belowground (B) arthropods, including recovery for 5 weeks following tarp removal.


Weeds and crops

The three farms had very different dominant weed populations. While Catamount had high coverage of purslane and crabgrass, Diggers’ had high coverage of redroot pigweed, and ICF had relatively high numbers of lambsquarters and oak-leaved goosefoot. 

When tarps were first removed, weed pressure was very low in all plots. However, by 5 weeks after tarp removal, control plots had an average of 90% weed coverage while the tarped plots only had around 30% weed coverage (all P < 0.001; Figure 3A). Looking at individual weed species, we saw that most weeds were negatively affected by tarps. One exception was purslane, which grew under the clear plastic tarps at levels similar to what we found under control plots. We additionally found that, while crop yield differed among farms, yield was significantly higher under the two tarped treatments than in the control plots (both P = 0.001; Figure 3B). 

These results confirm farmers’ experiences by showing that tarps are effective at suppressing weeds. The reduction in weed coverage also likely contributed higher crop yields, due to less competition for nutrients and light. However, the inefficiency of clear plastic tarps for purslane should encourage caution. Farms with high levels of purslane may want to avoid clear plastic tarps and opt instead to use silage tarps. 


Figure 3: Accumulation of weed coverage for 5 weeks after tarp removal in each treatment (A), and total crop yield per plot among treatment, separated by farm (B).


Relationship between soil arthropods and environmental factors

During tarp application, our set of environmental and experimental variables explained aboveground arthropod richness extremely well (total model fit = 78.6%) and belowground arthropod richness moderately well (total model fit = 38.8%). We found that treatment was the best predictor of aboveground and belowground richness (variable importance = 66% and 43%, respectively). Farm was also significantly related to aboveground arthropod richness (variable importance = 8%, P = 0.017), but did not relate to belowground richness. Soil temperature was also negatively correlated with arthropod richness (Figure 4A,B), but was not a significant factor in the model due to high correlation (collinearity) with treatment. Soil moisture was not predictive of aboveground arthropods (Figure 4C), but had a slight positive, yet insignificant, relationship with belowground arthropod richness (Figure 4D). 

When tarps were removed, our models were less effective at explaining arthropod richness (total model fit for aboveground and belowground arthropods: 35.5% and 20.5%, respectively). Treatment history remained an important and significant predictor of both aboveground and belowground arthropod richness, with a better fit with belowground richness (variable importance = 35%, compared to 6% for aboveground richness). Farm was the most important and significant variable in the models, with variable importance of 46% and 38% for aboveground and belowground arthropods, and both P < 0.005. Weed coverage also had a slightly positive relationship with belowground arthropod richness (Figure 4F), though this was insignificant. Finally, soil moisture and sampling time did not relate to either aboveground or belowground richness. 

These results highlight that many factors within the tarped treatment contribute to arthropod diversity. In particular, the negative relationship between soil temperature and arthropod diversity, and the high collinearity between treatment and temperature, may suggest that tarps’ heat effect causes negative arthropod effects. Farm also was a highly predictive and significant factor, which highlights that even farms that are relatively geographically close can have varying arthropod diversity patterns. We were surprised that weed coverage did not have a more positive impact on arthropods, but it is possible that the relatively short time period of the experiment and small treatment plots contributed to less detectable impact. 


Figure 4: Relationships between aboveground and belowground arthropod richness and soil temperature (A,B), soil moisture (C, D), and weed coverage (E, F).


Research conclusions:

We found that tarps have major impacts on the ecosystem. Many of these impacts are positive to farmers and may support their future use, including causing potentially higher soil nitrate, lower weed coverage, and higher crop yields. However, other impacts of tarps warrant caution. The particular, the ineffectiveness of clear plastic tarps at suppressing purslane could be problematic for some farms, and may mean that silage tarps are a better option for some farms. In addition, tarps' negative impacts on arthropod communities is concerning. While species living aboveground generally recovered within 3-5 weeks, we encourage future research into this topic to ensure that impacts don't intensify with continuous tarp usage. Promoting practices that help conserve arthropods is important because of the multi-faceted benefits they have in the farm ecosystem, including pest control and contributing to decomposition, soil structure creation, and nutrient cycling. We hope that this study motivates future research into tarps, increases awareness of arthropod communities in agriculture, and helps contribute to a more sustainable agricultural system. 


Participation Summary
3 Farmers participating in research

Education & Outreach Activities and Participation Summary

1 Curricula, factsheets or educational tools
8 Webinars / talks / presentations

Participation Summary:

150 Number of agricultural educator or service providers reached through education and outreach activities
Education/outreach description:

I have presented findings of this research on multiple occasions. In particular, because this project is part of my dissertation, I have presented it to academic peers and advisors multiple times. Conferences in which I have presented this research include CART (Sustainability Food Systems Transitions; UVM, July 2019), UVM RSENR's Annual Symposium (October 2021), and the Vermont Veg and Berry Growers Association (January 2022). I also reported my findings back in 1-on-1 presentations to 4 farmers (including the three farmers who participated in my research). I also plan to present this research at the annual conference for the Society of Conservation Biology (Reno, Nevada, July 2022), and at my PhD defense in early March 2022. 

I am currently working on a number of outreach materials, including a fact sheet for distribution to growers and practitioners, and two academic papers. I expect the fact sheet to be completed in the next month, and plan to submit both academic papers before the summer.

Project Outcomes

1 Grant applied for that built upon this project
1 Grant received that built upon this project
$19,350.00 Dollar amount of grant received that built upon this project
4 New working collaborations
Project outcomes:

Because of receiving this SARE grant, I have been connected to other researchers in the New England area who also work with tarping systems. In particular, I am working with Sonja Birthisel, Suzanne Ishaq, and Alicyn Smart at the University of Maine and Natalie Lounsbury at the University of New Hamsphire. We successfully submitted a proposal to the Northeast IPM center to fund a working group focused on tarping systems (award received February 2020). This working group has met twice and brings together around 20 researchers, farmers, and extension personnel to enhance knowledge, communications, and research on tarping systems in the Northeast. As part of this working group, we are compiling a report of tarping knowledge for wide dissemination. 

Knowledge Gained:

Through research and initial communication with farmers, researchers, and extension personnel, I have learned an incredible amount about sustainable agriculture systems in the Northeast. Most notably, I've found gained a huge appreciation for the farmers and extension personnel who mix innovation, skill, and their own research to ensure food production for their community. They do this in the midst of changing climates and increasingly aggressive pests and diseases. As I continue my research on this project, I am excited to continue to learn from these people. 

I plan to continue to work on questions of sustainable agriculture, entomology, and land management throughout my career. 

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.