Summer Cover Crops for Organic No-till Broccoli

Final report for FS20-326

Project Type: Farmer/Rancher
Funds awarded in 2020: $14,820.00
Projected End Date: 03/31/2022
Grant Recipient: Wild Hope Farm
Region: Southern
State: South Carolina
Principal Investigator:
Sarah Belk
Wild Hope Farm
Expand All

Project Information


While the farm management team at Wild Hope Farm has done some small experiments with sunn hemp and millet as summer covers for no-till fall-planted crops, we would like to conduct a more controlled experiment in order to examine what mix of sunn hemp, Japanese millet, soybeans and buckwheat result in the best weed suppression and yield for our fall-planted broccoli. If we find a successful mix, this would bring our no-till crops to 75% of our total production, increasing our ecological and financial sustainability as a farm business. 

In order to investigate which summer covers work the best, we evaluated 1) weed suppression 2) total yields and 3) total labor hours for each treatment including bed preparation, transplanting and weeding.


Our experiment consists of 5 treatments: 

1) Tilled after Sunnhemp (100lbs/acre)

2) Crimped Sunn Hemp (100lbs/acre)

3) Crimped Sunn Hemp (100lbs/acre) & Millet (50lbs/acre)

4) Crimped Sunn Hemp (75lbs/acre), Millet (50lbs/acre) & Buckwheat (50lbs/acre)

5) Crimped Sunn Hemp (75lbs/acre), Soybeans (50lbs/acre) & Buckwheat (50lbs/acre)


An area of 12 tractor pulled raised beds (217ft long) will be divided into 15 sections (3 beds wide, 54 ft each) in order to provide 3 replications of our 5 treatments. A list randomizer was used to assign each area a treatment.

Before planting, the 12-bed area will be stale seed bedded 1-2 times with mechanical cultivation equipment dependent on weather. Each cover crop will be seeded 3 beds wide with a bag seeder at the densities specified above, and then incorporated with another pass of the bed shaper.

Project Objectives:


  • Weed Suppression
    • After broccoli planting, weed counts will be conducted in each treatment area every 2 weeks until the end of harvest.
    • We will randomly pick three 5ft x 5ft squares to sample within the plot at each weed count. 
  • Labor Hours
    • Labor hours for the preparation, seeding, planting, and weeding will be tracked for each plot.
  • Total Yield
    • The total yield of broccoli in pounds will be tracked for each treatment. 


Click linked name(s) to expand/collapse or show everyone's info
  • Katherine Belk - Producer (Educator)
  • Sophia Friis - Producer (Researcher)
  • Mark Dempsey
  • Shawn Jadrnicek - Producer
  • Rachel Klein - Producer


Materials and methods:

An area of 12 tractor pulled raised beds (217ft long) was divided into 15 sections (3 beds wide, 54 ft each) in order to provide 3 replications of our 5 cover crop treatments. A list randomizer was used to assign each area a treatment. Treatments were seeded with a bag seeder at the following rates:

Treatment 1: Tilled after Sunnhemp (100lbs/acre)

Treatment 2: Crimped Sunn Hemp (100lbs/acre)

Treatment 3: Crimped Sunn Hemp (100lbs/acre) & Millet (50lbs/acre)

Treatment 4: Crimped Sunn Hemp (75lbs/acre), Millet (50lbs/acre) & Buckwheat (50lbs/acre)

Treatment 5: Crimped Sunn Hemp (75lbs/acre), Soybeans (50lbs/acre) & Buckwheat (50lbs/acre)

This Plot Map depicts the randomized location of each treatment within our test plot. In 2020, Broccoli plants were planted in each plot except for plots 4, 8 and 12 where we were short on broccoli transplants and supplemented the plots with cabbage and summer squash. In 2021, all plots were planted with the same variety of broccoli. 

Materials needed for weed tracking include a 5x5' square made from PVC pipe. Materials needed for harvest and yield tracking include harvest bins and knives for the harvest crew and color-coordinated tape that correlates to each treatment in the field. In the plot, colored flags were attached to in ground stakes that indicated where treatments began and ended.


Early May

  • Winter rye and clover plot mowed and disked
  • Beds reformed with the bed shaper. 

Mid May

  • Sections with treatment #1 and #2 seeded with Sunn Hemp (100lbs/acre)

Mid June 

  • Sections with treatment #3 seeded with Sunn Hemp (100lbs/acre) & Japanese Millet (50lbs/acre)
  • Sections with treatment #4 seeded with Sunn Hemp (75lbs/acre), Millet (50lbs/acre) & Buckwheat (50lbs/acre)
  • Sections with treatment #5 seeded with Sunn Hemp (75lbs/acre), Soybeans (50lbs/acre), & Buckwheat (50lbs/acre)

Mid July

  • Treatment #1 sections flail mowed & disked

Late July

  • Treatment #1 sections tilled (2" depth) and bed shaped
  • All other sections terminated with roller-crimper

Last week in July

  • Same variety of broccoli was transplanted in all sections by hand. Spacing at 2 rows, 12 inches.
  • 2 lines of drip tape laid per bed.
  • Treatment plots were marked in the field with colored flags.
  • Weed counts were conducted bi-weekly by laying the 5x5 square in the middle bed of each plot and observing weed coverage and weed type.

Early October 

  • Harvest of broccoli began. All plots were picked 3 times a week and yield was tracked in pounds by treatment type.
  • Harvest bins were marked with colored tape that corresponded to the treatment of that plot. A color key was posted in our wash station where yield was recorded in pounds in a chart that was then digitized for our records. 
Research results and discussion:

Cover Crop Seeding + Quality of Stand 

Summer cover crops establish well when either a rain event follows seeding or overhead irrigation is available to use. While using a bag seeder was an effective way to distribute seed evening, incorporating the seed with another bed shaping tractor pass did not produce even stands. This was particularly prominent with the larger seed (soybeans). Bed shaping did not bury the large seed enough and as a result, germination suffered. We also experience reduced germination in the pathways and higher density on the bed shoulders due to the incorporation method. 

Reflections on the stands of each treatment are as follows:

Treatment 1: Tilled after Sunnhemp (100lbs/acre) - Heavy density on the shoulders but overall good germination. 

Treatment 2: Crimped Sunn Hemp (100lbs/acre) - Same as Treatment #1.

Treatment 3: Crimped Sunn Hemp (100lbs/acre) & Millet (50lbs/acre) - Millet thrived in the second year due to more rainfall. 

Treatment 4: Crimped Sunn Hemp (75lbs/acre), Millet (50lbs/acre) & Buckwheat (50lbs/acre) - Millet seemed to be out competed by buckwheat, most likely from its allelopathic effects. 

Treatment 5: Crimped Sunn Hemp (75lbs/acre), Soybeans (50lbs/acre) & Buckwheat (50lbs/acre) - Soybeans did not germinate well and seemed to be outcompeted by buckwheat as well. 

Each year we used a hose reel to irrigation 2-3 times to establish the cover crop. In Year One, following the initial irrigation we experienced periods of drought during the cover crop’s life and this resulted in much of our Japanese Millet being stressed and either dying early on or going to seed much earlier than expected. In Year Two, we had more consistent rainfall following seeding and that extended the days to maturity of the Japanese Millet. 

In both years of this experiment, we broadcast our cover crop seed with an Earthway Bag Seeder followed with the tractor-mounted bed shaper to incorporate seed. This setup resulted in heavy seed incorporation/germination in the bed shoulders and lighter incorporation in the bed top. For the two replications of this experiment, this inconsistency in the cover crop stand created inconsistency in the effectiveness of the crimped cover crop. We would like to experiment in the future with a no-till drill or using a tine weeder to incorporate seed. 

Crimping Event

In 2020, we did not have enough time to crimp more than once as our transplants were ready to go in the ground. Due to this, we had a lot of cover crop regrowth in our plots, mostly sunn hemp and buckwheat. It seemed that the regrowth was coming from side shoot production for both of them. We had the crimper halfway full of water which was potentially too heavy and resulted in more breakage in the covers than damage. 

In 2021 we had 1.5 inches of rain on July 26 and as a result, we pushed the crimping date back to July 30th. By this time, much of the Buckwheat and Sunn Hemp was leaning or partially lodged due to the rain. We filled our rear-mounted roller-crimper halfway full of water and after one pass in the Sunn Hemp and Buckwheat plot, the Sunn Hemp snapped at the base and the Buckwheat was chopped by the chevrons of the crimper. We emptied the crimper of water for subsequent passes to reduce the weight load on the cover crop. At the time of crimping, the oldest Sunn Hemp (Treatment 1, seeded May 15th) was 7ft tall and 10% flowering, all the Buckwheat was in full flower and 5ft tall with minimal seed development, and the Japanese Millet had 40% of the crop in a general grain-filling phase.

In 2021 we were able to crimp again 5 days later, which helped damage any regrowth that had started and resulted in a better weed suppression mat with less competition for our broccoli planting. 

In our attached PowerPoint there are photos of the cover crops throughout their lifecycle, at the time of crimping and their breakdown.

Transplant Success

In 2020, the farm was dealing with several additional challenges that impacted our research - it was our first season farming 12 acres (double the acreage of the previous season), we had an excessively wet spring followed by a dry summer, we experienced heavy aphid and armyworm pressure in our greenhouses and heavy fire ant pressure/girdling of all our broccoli transplants.

In 2021, we invested in a new potting soil mix which allowed for more robust broccoli transplants and reduced our fire ant transplant loss to around 10%. In the tilled sections, we noticed an unknown deformation in 30% of the young broccoli plants (excessive wrinkling of the foliage) and these plants produced small heads. One explanation for the deformation could be that we experienced higher pest pressure in the tilled plots since they did not have the same level of population of predatory insects that lived in the crimped cover crop beds. In addition to the excessive wrinkling, we had a lot of aphids on the tilled section. This could have been a result of excess nitrogen available to the plant due to tilling in the Sunnhemp seeded at 100lbs/acre.  

Cover Crop Breakdown + Weed Analysis

A primary goal of crimped cover cropping systems is to suppress annual weeds. In the first year of this experiment, one of the ways we assessed the effectiveness of our cover cropping was through weed tracking. We tracked perennial & annual weed presence in randomized 5x5 plots in each of the treatments on a bi-weekly basis. We experienced significant cover crop regrowth & reseeding. The following Weed Count Graphs depict the percentage cover of weeds and cover crop regrowth/reseeding. Negative space in the graphs shows properly mulched and covered bed space and broccoli plant coverage.

We experienced the regrowth of cover crops in most of the plots due to ineffective timing of cover crop crimping since at the time of crimping, the millet and buckwheat had produced viable seed and reseeded itself in the plots, and the Sunn Hemp regrew from new growth points. In Treatment 1 Sunn Hemp alone did not suppress perennial weed growth at first but over time the cover crop regrowth ultimately outcompeted both perennial weeds and broccoli. In Treatment 2 we experienced low suppression of both perennial and annual weed growth. In Treatment 3 we experienced high competition from both buckwheat reseeding and soybean regrowing. In Treatment 4 we believe the millet cover crop was water-stressed prior to crimping so it did not effectively suppress perennial weeds however the presence of annual weeds was not significant. In Treatment 5 we experienced a lot of buckwheat reseeding and Sunn Hemp regrowing.

We think there was excessive elongation in the broccoli stems due to light competition with the regrowing cover crop. We tried to knock back the Sunn Hemp regrowth by running the flail mower over the top of the beds. We tracked labor hours by treatment type and this bar graph shows how many times we had to visit the plot with the tractor.  The no-till plots required fewer trips which equated to less time and fuel needed to produce the crops.  The no-till labor hours would have been even lower but we had to make several trips to mow the regrowth of the cover crop.  

In the second year of the experiment, our crimping was successful and there was insignificant regrowth so we decided that in order to have better success with harvest, we would hand weed the treatments and track our labor hours doing so instead of letting the cover crop and weeds grow. This proved to be very effective at knocking everything back and only had to be done once. The primary perennial weeds we saw in our plot were horse nettle and dock. 

Broccoli Yields

In 2020, broccoli yields were highest in the Sunn Hemp and Millet plot with a total yield of 0.249 lbs per bed foot. Overall, there was significant crop loss in the research plot due to transplant loss and plant loss due to disease most likely caused by the cover crop regrowth. In 2021, broccoli yields were almost three times higher per bed foot with the highest yields coming out of the tilled plots, however, the broccoli coming off this plot was of lower quality. The second highest yields came out of the Sunn Hemp, Millet and Buckwheat plot. This mix also created the densest crimped cover crop, with proper matting in the understory and successful crimping of the Buckwheat crop. 

Cash Crop Health

In the first year of this experiment, cover crop regrowth probably also contributed to poor air circulation and may have increased diseases like Alternaria head rot and Bacterial head rot. The second year of the experiment had much greater success and proper formation of broccoli heads with almost three times the poundage per acre as the previous year. 

In our attached materials, there is a powerpoint presentation that is mostly photo-based and provides visual comparisons across treatments and replication years of the projects. 

Participation Summary
5 Farmers participating in research

Educational & Outreach Activities

6 Consultations
6 Curricula, factsheets or educational tools
1 Webinars / talks / presentations
1 Workshop field days

Participation Summary:

45 Farmers participated
Education/outreach description:

In November of 2020, we facilitated a workshop summarizing our findings from year one of the research at the Carolina Farm Stewardship Association (CFSA) 2020 Sustainable Agriculture Conference. This conference occurred virtually with a powerpoint presentation and opportunities for participants (~30) to ask questions using the Zoom chat feature. The workshop lasted an hour and was recorded for later access by conference participants.

Through our partnership with the CFSA, we consulted with Mark Dempsey over both years of our research. Mark studied the research plots with us days after crimping the cover crop and a month later after the broccoli had been planted. We discussed cover crop regrowth, possible disease apparent in the broccoli crop, and cover crop management changes for the next season. He also sat in on our CFSA workshop in November of 2020.

We created multiple visual representations of our research. This fact sheet describes our projects goals and a preliminary summary of results with images of our cover crop and graphs of weed counts to analyze the cover crop effectiveness. 

The second visual is a 7 min video shot throughout the growing season and provides a great overview of our cover crop at all stages. This video describes the scope of our project and the benefits of cover cropping in a no-till agricultural system specific to the southeast. The video also includes conversation among our farmers, Mark Dempsey’s consultations, and possibilities to improve upon our experiment next season. The video was shown at our 2020 CFSA conference workshop. The video can be found here:

As a final presentation of our research, we created this power point presentation and shared it in a live zoom meeting in February of 2022. The presentation discussed our research process, how to fit a research project into a busy farming season, and how to navigate funding on-farm research. We had 25 attendees and a good Q&A discussion following the presentation. This event was advertised on our Instagram and Facebook. 

Learning Outcomes

40 Farmers reported changes in knowledge, attitudes, skills and/or awareness as a result of their participation

Project Outcomes

Project outcomes:

In regenerative, organic vegetable systems, cover crops provide several benefits to the producer - they prevent erosion, reduce water use, increase fertility, and improve the biological and physical properties of the soil. In addition to these benefits, cover crops that are terminated with the roller-crimper to produce mulch for the following cash crop can also provide a significant labor reduction for the farmer. Most research with roller-crimper vegetable production has focused on fall-planted cover crops, leaving little information about summer cover and prompting us to investigate further. With a longer growing season in the southeast, we have more of an opportunity to utilize summer cover crops. Having summer cover or mulch in place during the hurricane season (June-Dec) can significantly reduce damage and erosion to the field. Summer cover crops have the potential to reduce labor hours and increase soil resilience and retention in the southeast.

Upon completion of the project, several more trials need to be conducted to investigate the cover crop regrowth and exactly why it is happening. For further investigation, we would reduce the number of treatments we were experimenting with and stick with a straight sunn hemp plot as well as a sunn hemp and buckwheat plot. While we were excited about Millet's ability to form a thick mat for weed suppression, we noticed a lot of plant stress in those sections and believe it is too difficult to properly estimate days to maturity for crimping. In narrowing our focus, future experiments could look at sunn hemp seeding density, method of cover crop incorporation, and crimping at various maturities for the sunn hemp. 

Additionally, more research needs to be done around incorporating fertility into the experiment as we are still falling short of our ideal yields. Hopefully, this experiment continues to push research forward and provides other producers with ideas to start incorporating summer cover crops into their rotations. 

Information Products

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.