Use of High-Residue, Winter-Killed Cover Crops in No-Till Organic Tomatoes

Project Overview

Project Type: Partnership
Funds awarded in 2016: $29,998.00
Projected End Date: 05/03/2018
Grant Recipient: FairShare CSA Coalition
Region: North Central
State: Wisconsin
Project Coordinator:
Claire Strader
FairShare CSA Coalition

Annual Reports

Information Products


  • Agronomic: sorghum sudangrass
  • Vegetables: tomatoes


  • Crop Production: cover crops, no-till
  • Production Systems: organic agriculture


    Using cover crops to create an in situ mulch is one way that organic farmers can explore no-till techniques.  While there has been some success with organic no-till row crops, organic no-till vegetables remain a conundrum.  Inadequate weed control, narrow cover crop termination windows, and planting delays related to termination are all challenges.  This project explored the use of season-long managed fallow concluding with high-residue, winter-killed cover crops to create a weed free mulch that does not need exact timing or special equipment for termination.  The primary cover crop was sorghum sundangrass, which is known for producing large amounts of biomass.  Farmer cooperators chose tomatoes as the vegetable crop because they generally benefit from mulch and have high value.

    In 2016 we used two rounds of cover crops to reduce weeds in the trial areas on three cooperating farms in Dane County, Wisconsin.  In July we established no-till treatment plots with (1) sorghum sudangrass alone, (2) sorghum sudangrass and sunn hemp, and (3) sorghum sudangrass and cow peas.  The control was sorghum sudangrass to be managed with conventional tillage and plastic mulch.  In the fall, just before frost, we rolled the cover crops with a disengaged rotovator to align the residue into an even mulch mat.

    By May of 2017, it was clear the cover crop residue would not provide adequate weed control for the cropping year.  The project then pivoted to look at three supplemental mulch materials used to exclude weeds in the no-till tomatoes:  (1) plastic, (2) landscape fabric, (3) marsh hay.  We tracked labor time, soil temperature, plant survivorship, and yield and analyzed all the data for differences.

    Key Findings

    Sorghum Sudangrass Residue for No-Till:

    • Even with 10,000 or more pounds of dry sorghum sudangrass biomass in the fall, the remaining spring residue covered only 20% - 60% of the ground and was not adequate for use as a no-till mulch.
    • Rolling the sorghum sudangrass with a disengaged tiller just prior to frost was quick, easy, and did a good job of aligning the residue.
    • Sorghum sudangrass smothered not only weeds but also the sunn hemp and cow peas planted as leguminous companion cover crops.
    • The sunn hemp attracted Japanese beetles, which makes it a less desirable cover crop on farms with vegetables that are susceptible to Japanese beetle damage.


    Mulches for No-Till Tomatoes:

    • Tomato yields were the same for all the no-till mulches and the conventionally tilled and plastic mulched control.
    • Given that there was no yield penalty, it is worth experimenting with no-till tomatoes, especially when wet soils prevent or delay tillage.
    • Soil temperature under the marsh hay was significantly lower than the other two treatments and the control.
    • Tomato harvest from the marsh hay plots was equal to, but peaked later than, the other treatments, likely due to the lower soil temperature.
    • Labor in each no-till system was different, with the marsh hay requiring the least time and the plastic requiring the most. The control required less time than any of the no-till systems.
    • Weed control under all the no-till mulches was excellent, even through they were applied directly on top of sprouted weeds.


    Project objectives:

    Objective 1:  Assess three cover crop combinations of high-residue, winter-killed mulch for production of no-till organic tomatoes as compared to a conventionally tilled plastic mulch.

    We collected population density and biomass for weeds and cover crops in the fall and again in the spring.  We also did a visual assessment of the cover crop residue in the spring.  When it became clear the residue would not be enough to exclude weeds in the cropping year, we adjusted the project and added objective 2 below.

    Objective 2:  Assess three supplemental mulches for use in no-till tomatoes as compared to a conventionally tilled, plastic mulched control.

    To measure the effect of the mulches and associated management costs, we tracked labor time, soil temperature, plant survivorship, and tomato yield.   An analysis of variance (ANOVA) was conducted using the R package lme4 (Bates et al 2015).  Means and 90% confidence intervals for each treatment were calculated with the R package lsmeans (Lenth 2016).  A confidence level of 90% was used, meaning that for each comparison that is statistically significant, we are 90% confident that the difference is due to the treatments and not to chance variation.

    Objective 3:  Share information with growers through a field day in 2017.

    We held a field day at Crossroads Community Farm 8/31/17.  The 32 attendees toured the trial plots, received a handout with the results to date, and discussed pros and cons of organic tomato no-till.

    Objective 4:  Create an illustrated info sheet and short video with project results and recommendations to be posted on the FairShare website and disseminated through the FairShare listserv.

    Two bulletins and two videos were completed and distributed in May 2018.  See information products.

    Objective 5:  Contribute information to the development of longer, multi-year reduced tillage vegetable rotations.

    Results of the project were used to draft farmer recommendations for organic no-till tomatoes and distributed through the bulletins and videos.


    Bates, Douglas, Martin Maechler, Ben Bolker, Steve Walker (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1-48. doi:10.18637/jss.v067.i01.

    Lenth, Russell V. (2016). Least-Squares Means: The R Package lsmeans. Journal of Statistical Software, 69(1), 1-33.  doi:10.18637/jss.v069.i01

    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.