Comparing Alternative Weed Management Practices to Black Plastic in CBD Hemp Production

Progress report for ONE20-366

Project Type: Partnership
Funds awarded in 2020: $29,993.00
Projected End Date: 11/30/2022
Grant Recipient: University of Maine Cooperative Extension
Region: Northeast
State: Maine
Project Leader:
Dr. John Jemison, Jr.
University of Maine Cooperative Extension
Expand All

Project Information


After an eight-decade hiatus, hemp production is back.  Farm diversity and profits can be increased if production costs can be controlled.  Approximately 2625 acres have been grown in Maine since 2017.  Hemp grown in the northeastern United States for cannabidiol (CBD)-rich flowers must be produced on wide (5x5) spacing to reduce disease due to high humidity.  No herbicides are labeled for hemp, and with the wide spacing, there is little crop shading to slow weed growth without using plastic mulch.  More than 19,000 ft2 of plastic is landfilled each year per acre of hemp produced. This waste and the fuel required to till and prepare hemp beds is unsustainable.  Planting hemp into no-tilled mowed or rolled/crimped cereal rye cover crops could reduce weed pressure and reduce cost.  Winter rye protects the soil from erosion, transpires water to dry soils in the spring, feeds microbes through root exudates, and inhibits weeds when rolled or mowed into a mulch.  In this 2-year, 2-location project, Sarah Hewitt of Victory Hemp Farm and I will evaluate CBD hemp performance in a no-till rolled/crimped or mowed rye mulch and assess possible allelopathy or nitrogen immobilization issues.  We will also evaluate novel stacked cultivation tools to see if intensive cultivation can reduce weed pressure. These systems will be compared to a tilled, standard black plastic mulch production system.  We will assess plant growth/development, flower and total plant yield, weed biomass, and cannabinoid content.  Results will be shared at grower meetings and on a hemp production website.  

Project Objectives:

Through replicated research trials, we seek to explore novel means to manage weeds in CBD hemp production.  We will compare a standard black plastic mulch (control treatment) to rolled/crimped winter rye, sickle-bar mowed winter rye, and stacked cultivation tools.  This will be done in two experiments: one conducted with Sarah Hewitt of Victory Hemp (VHF) in Union Maine, and the other at the University of Maine Whitter Center Rogers farm (UMRF). 


The project seeks to assess the effect of: 1) tillage on CBD hemp growth and development; 2) allelochemicals and nitrogen immobilization slowing plant growth and development; 3) rye cover crop destruction method (roller crimper or mowing) on continuity and thickness of cereal rye mat; 4) weed management treatments on weed density by species; and 5) weed density and plant stress on THC and CBD levels in the plant.  Study results will benefit farmers by providing them with research-based information on the risks and benefits of managing weeds with rye cover crops, reducing tillage, and the effects of plant stress on the total cannabinoid content in CBD hemp. 


The Farm Bills of 2014 and 2018 legalized hemp production across the US, leading to renewed farmer interest in this crop.  Maine initiated its hemp licensing program in 2016, and producers have grown approximately 2500 acres of hemp, primarily for high-CBD flower, in the state over the past two years.  Most farmers transplant hemp into black plastic mulch on a 5x5 or 6x6 spacing.  There are no legal herbicides for use in hemp, and plastic mulch serves to both control weeds and warm the soil.  However, the input costs and waste generated to produce CBD hemp in the Eastern US is not sustainable.  Black plastic can only be used for one year; so, following harvest, almost 19,000 ft2 of plastic is landfilled for each acre of production.  Alternative weed management systems using cover crops are needed that can produce adequate biomass to protect the soil from erosion over winter, improve soil health, slow weed germination, and make eastern hemp production competitive with our western competition.


Due to lower relative humidity, western growers don’t have to rely on plastic as they can produce hemp at increased plant densities with less concern for plant disease.  Tightly spaced hemp reduces weed competition and increases flower yield per acre.  Maine growers who have not used black plastic mulch have been unable to control weeds, leading to increased plant stress and lower flower yield.  Maine’s hemp program director stated that poor weed control is our state’s number one hemp production issue.  Along with reduced yields, CBD plants under stress can have delta-9 THC levels spike above the legal limit of 0.3%.  With the new USDA interim final rule reducing the level to 0.3% total THC on a dry matter basis, more growers with stressed plants risk failing the federal THC standard.  At $15,000 to $18,000 per acre production costs, a crop not meeting the THC standard is an extreme economic risk to the grower. 


Possible alternatives to black plastic include plant-based cover crop mulches and timely cultivation.  Planting a cereal rye cover crop in the fall would protect the soil from erosion, feed soil microflora, and when mowed or flattened/killed with a roller crimper in the spring, the rye residue will form a mulch bed through which to transplant the CBD hemp.  Seed cost for a 2 bu/ac seeding is $16/ac.  Not tilling the soil will also save fuel cost, labor, and carbon dioxide emissions. The University of Maine has access to a roller crimper to use for this purpose.  Other farmers may not have access to this tool.  Simply mowing the rye with a cycle-bar mower could be an alternative way to form a bed.  We must also explore improved cultivation methods. 


Research is needed to understand how CBD hemp will perform transplanted into a rye cover crop mulch.  Growers need to know if hemp will be affected by either nitrogen immobilization or allelopathic chemicals that might leach out of the rye mulch.  Practice using novel stacked-cultivation tools is also needed. 


Click linked name(s) to expand/collapse or show everyone's info
  • Sarah Hewitt - Producer


Materials and methods:

This project supported two replicated field trials:  one conducted in Union Maine on Sarah Hewitt’s VHF, and the other will be conducted at the University of Maine Witter Center - Rogers Farm (UMRF) in Stillwater Maine .   These will be repeated in 2022.  Each study is a randomized complete block study: the Union location will have 4 treatments:  black plastic mulch as the control, roller/crimped rye mulch, sickle-bar mowed rye mulch, and stacked cultivation with 4 replications; the UMRF location will have 5 treatments (those previously mentioned plus a weed free treatment) also with 4 replications.  Each study area will be approximately 0.5 acres in size.


The project was initiated in fields with standing pea, oat, and vetch green manure cover crops.

Figure 1.  Pea oat vetch cover crop at the University of Maine Rogers Farm, 2021.

In late august of both 2020 and 2021, cover crops were mowed and incorporated into the field.  Treatment areas were laid out and either oats or Aroostook rye were broadcast sown at 2 bu/ac and rolled to ensure seed-soil contact following the model used in Mallory and Molloy (2017) at the VHF and drilled at those rates using a grain drill at the Experimental Station.


Prior to transplanting the hemp in 2021, winter rye cover crop biomass samples were taken from each plot.  Yields ranged from 9212 lbs dm/ac at VHF and 8040 lbs dm/ac at UMRF. 


Soil and experimental preparation

At UMRF, the winter rye plots were either sickle bar mowed or rolled/crimped.  At VHF, the rye was only sickle bar mowed.  At VHF, the remainder of the field was rototilled and to establish the black plastic ‘standard’ treatment, black plastic and drip tape was laid using a bed layer.  Drip tape was run down through all the plots to provide irrigation and nutrients.  At UMRF, plots without a vegetative mulch were chisel plowed and disked.  A bed former was also used to lay the black plastic and drip tape on 10 June 2021.  The remaining plots were established the next day. The winter rye cover crop at UMRF was either mowed using a small walk behind sickle bar mower or rolled with a tractor using a 6 foot wide roller/crimper machine.  The remaining treatments included a cultivated plot in which a walk behind Power Ox cultivator equipped with three sweeps was used to kill germinating weeds on a biweekly basis at both sites. The Power Ox was directed both down the length of the plot and across the plot to control the weeds biweekly in June, July and August.  When the plants were too large to cultivate between them, that treatment was hand hoed.  As well, a control plot with no weed management was used to assess the effect of intensive weed pressure on crop yield and cannabinoid content. 

Figure 1b.  Roller crimper used to roll down winter rye cover crops.  


Figure 1b.  Rolled winter rye cover crop.

Four-week old hemp seedlings (variety Pasture Suver Haze) that had been sufficiently hardened off were transplanted on 10 and 11-Jun at UMRF and VHF respectively.   At UMRF, a slot was cut into the winter rye soil to facilitate transplanting the seedlings.  At VHF the soil was sufficiently loose to allow transplanting.  Each plant was hand transplanted on a 5 foot x 6 foot spacing, 15 plants per plot.  The middle three plants in the middle row provided the harvest plant material for the study. 



Irrigation intensity and timing was determined in part on rainfall and the need to apply plant nutrients. Drip lines were placed next to each plant row, with emitters every foot.  Nutrients were supplied to the plant based on the farm practices.  At UMRF, fertilizer was mixed to apply 8 lbs N/ac per application through a mix-rite nutrient application system. At UMRF, nutrients were applied weekly from May-July to apply 96 lbs of N prior to flowering.  After flower initiation, nutrients were applied four more times at 5 lbs N/ac for a total of 116 lbs of N. At VHF, fertigation was effected bi-weekly using liquid nutrients and compost teas.  The nutrient content of the mixture was analyzed, and it was expected that the total application of N was approximately 42 lbs/ac less than that done at UMRF, but with the organic matter of the soil at 4% higher, less total N was needed.  This is the way nutrients are typically applied based on Sarah Hewitt’s experience as a hemp grower.


Data Collection

To assess weed management treatment effects, plant height, diameter and node number were collected biweekly from both locations.  Key sampling times were one-month after transplanting, at flower initiation and just prior to harvest.  Leaf chlorophyll measurements were taken on the fan leaves at the first alternate and last opposite leaves (nodes 6 and 7) in July, and then the fan leaves four nodes from the top of the plant were measured using the SPAD meter, and also those 8 nodes from the top at flower initiation.  Ratios of the first alternate to last opposite (July) and 4th and 8th nodes from the top were assessed in August at flower initiation and were correlated with fan leaf total N collected from the most fully mature fan leaf at the top of the plant at flower initiation following methods described by Cochran et al., 2019, analyzed for total N. 


Harvest Methods

Harvest required collecting the middle three plants from each plot for yield assessment.  For each plant harvested, the whole plant weight was recorded. The largest of the three plants was cut at the plant base approximately 10 cm above the ground surface with loppers, and the plant was weighed and separated into individual branches and run through a chipper shredder to determine whole plant dry matter, total N content, and cannabinoid content. The smallest plant of the three harvest was field weighed and air dried.  The middle-sized plant was harvested the following day, weighed and separated into individual branches and stripped of its fan leaves.  Flowers were separated from individual branches using an MB1 Munch Machine bucker.  Wet bud weight and unmarketable bud weight were recorded. Stems were also collected and weighed.  Flower dry matter content was assessed by collecting a flower subsample and drying the flower sample overnight in a small dehydrator.  Samples of the dehydrated flower were submitted for cannabidiol (CBD) and tetrahydrocannabinol (THC) and total cannabinoid content at Nova Analytics in Portland.  The percent moisture at harvest was used to calculate total dry matter and flower dry matter yields. The whole plant total N was analyzed at the University of Maine Soil Testing Lab. 


Data Analysis

Using JMP 16 from SAS, data were analyzed using analysis of variance and specific contrasts to determine if weed management affected growth, weed biomass, whole plant and fan leaf total N content, whole plant yield, flower yield, and cannabinoid content.  Key contrasts included the black plastic standard vs all other treatments; winter rye compared to black plastic and cultivated compared to no weed management. 



Research results and discussion:



Environmental Conditions

The 2021 growing season was one of marked extremes.  Table 3 is a presentation of average temperatures,  and the very low rainfall in May and June and the abnormally warm weather in June allowed an earlier than expected planting date. 


Table 3.  Average monthly temperature and rainfall collected at UMRF*


Avg Monthly Temperature (C)

30-year average temperature (C)

Monthly Rainfall (mm)

30-yr average rainfall (mm)































*the two farms are 60 miles apart.  No weather measurements were collected at VHF.



Early Growth and Development

Weed management treatments led to differences in hemp growth.  We took growth measurements at one month after transplanting, at flower initiation, and before harvest. Perhaps not surprisingly, the black plastic standard created the best growing conditions and the plants responded accordingly.  The sandy loam, high organic matter soil at VHF was significantly more conducive for early growth and plants were almost twice as tall in Union than at the UMRF (Figure 1).  There was a statistically significant interaction in plant area at one month following transplanting.  In both locations, the black plastic and control plots were taller than the mowed rye.  Whether the plants growing in the mowed rye were stunted due to 1) the untilled soil conditions slowing root development, 2) N immobilization, or 3) allelochemicals leaching out of the rye stunting growth was not clear.  It could have been a combination of factors. 


Figure 1.  Plant growth at one month after transplanting.


Plant Analysis

We used a Minolta SPAD meter to evaluate leaf color on the last opposite and first alternate fan leaf by experimental treatment one month after transplanting.  We found a statically significant interaction in both the last opposite fan leaf SPAD value and the ratio of the last opposite to first alternate fan leaf SPAD values for the VHF and UMRF locations (Figures 2a and 2b). The lowest SPAD value on the last opposite fan leaf was found with the mowed rye treatment at both sites.  This could suggest that the lower color (lighter green color) is due to lower N due to possible immobilization, poorer root formation due to the untilled soil, due to allelopathic chemicals leaching out of the rye causing leaf yellowing, or a combination of the three factors.  The ratio of the last opposite to the first alternate SPAD values were also different at the two sites.  At the UMRF location, the ratio of the last opposite to first alternate fan leaf SPAD values were below 1, and the lowest value was in the control plot suggesting that early weed growth where no weed management was done might have tied up N.  In contrast, at VHF, except for the black plastic standard treatment, the SPAD ratios were greater than one with the highest ratio value found in the mowed rye plot. This was somewhat unexpected. If the ratio of the last opposite fan leaf to first alternate fan leaf were less than one, it might indicate that the plant is robbing N from below to the growing points.  But it might also just suggest that the newer (younger) plant leaves might just be a deeper green color because it was growing out of a allelochemical toxicity setback?  Given the lack of other work with this measurement in the literature, this was difficult to determine.  If the winter rye mulch were immobilizing N, the mowed rye treatment should have had the lowest leaf SPAD value, low fan leaf total N, and the ratio of the last opposite fan leaf to first alternate fan leaf SPAD value should also have been less than 1. In fact the mowed rye treatment had a statically lower fan leaf N than the cultivated cannabis, but was not different from the black plastic or cultivated cannabis. If the ratio is >1, it might mean that the newer growth is affected by allelochemicals slowing growth and slowing photosynthesis in the newer leaves of the plant.   Although the difference was greater at VHF than UMRF, the ratio was highest in the mowed rye plot indicating that N was not limited, and that allelopathic chemicals might have slowed growth and development.   


Unfortunately, we did no additional soil N tests and we did only leaf tissue total N at VHF to verify immobilization in the rye treatments.   But the low fan leaf N at VHF would indicate some N immobilization.

Figure 2a – Fan leaf SPAD values of the last opposite leaf as affected by weed management – July 2021.


Figure 2b – Ratio of the first alternate to last opposite fan leaf SPAD values as affected by weed management – July 2021.



Flower Initiation – growth and development

By flower initiation, the resinous hemp plant growth had increased by three times at VHF and four times at UMRF.  In all cases, the VHF site had larger and more advanced plants, but a statistically significant interaction in plant size was likely caused by the difference in size in the cultivated plots at VHF relative to UMRF.  The number of plant nodes was affected differently at VHF and UMRF by weed management (Figure 3). 


Figure 3a.  Plant size as affected by weed management at flower initiation


Figure 3b.  Plant node number as affected by weed management at flower initiation. 


The more intensive broadleaf weed population at VHF likely limited N and affected growth more than at the UMRF location.  As well, plants in the mowed rye treatment were three times smaller than the black plastic at VHF while only two times smaller at UMRF.  Black plastic appeared to be more useful at UMRF than at VHF.  Again the soil type was more conducive for cannabis production at VHF than UMRF, and the difference in development in the mowed rye plot was more significant at VHF than UMRF. 


Flower initiation – leaf SPAD

Although the fertility management was different at the two locations, leaf color was surprisingly similar across locations.  To check the fertility status (Cockson, 2018) recommends sampling the top most fully developed fan leaf from the top of the plant at flower initiation.  To evaluate a second hemp fertility assessment with the Minolta 501 SPAD meter , we measured fan leaf SPAD values at the 4th and 5th node from the top of the plant.  We also counted four nodes below that and collected SPAD measurements at the 8th and 9th node from the top.  We then developed ratios of the 8th to the 4th and 9th to the 5th values to see if N might have been limited by weed management at either location (Table 4).  Fan leaf total N was also collected at VHF but due to a misunderstanding was not done at UMRF.


Table 4.  Fan leaf SPAD values at different plant nodes and the ratio to assess apparent plant health and N status.

Weed Management

4th Node

5th Node

8th Node

9th Node

Ratio 8:4

Ratio 9:5

Black Plastic





















Mowed Rye





























In all these cases, leaf SPAD values response to the weed management were generally similar across locations.  The 4th node leaf SPAD value reflected well the plants response in size and node number (Figures 3a and 3b).  Reducing weeds by black plastic or cultivation generally led to the largest plants and deepest green plant color.  Interestingly, the fan leaf SPAD values at the 4th node correlated the most highly to the most fully mature fan leaf and the relationship was significantly significant (Figure 4).


Figure 4.  Relationship of 4th node fan leaf to plant N status.



Harvest - Weed Biomass


Weed biomass samples were collected, dried, and weighed to assess weed management effects just prior to harvest.  The black plastic weed biomass was collected from between the rows.  The remainder of the samples were collected within the plot at random.  The weed dry matter production is presented below in Table 5. 


Generally, the weed management treatments led mostly to similar weed levels by treatment and location.  Tilling the soil appeared to bring up more annual broadleaf weeds which when not controlled led to significantly higher biomass compared to the mowed rye which had not tillage.

Figure 4a.  Annual weeds flush following tillage at VHF. 

  Another important difference was the mowed rye treatment at UMRF had higher levels of perennial broadleaf weeds due to a large clover population.  Despite plowing under the previous crop residue and clover remaining in the plot from the previous year, the rye residue was not sufficient by itself to keep the clover from growing through the residue and competing.  


Figure 4b.  Hemp plant size difference where soil was and was not tilled.  


Table 5.  Weed dry matter yield at harvest 2021 across location.






Weed Management

Annual Grass

Annual Broadleaf

Perennial Broadleaf

Total Weed Biomass


*******     g/m2    *******

Black Plastic


8.0 b


77.3 a



980.2 a


1027.9 c



33 b


55.0 a

Mowed Rye


202.3 b


423.8 b






















By the end of the season, the plants at UMRF had almost caught up with the size of the VHF location (Figure 5).  Although VHF plants were still statistically greater in size, but possible reasons for the increase in UMRF plants to that at VHF during flowering were that septoria leaf spot disease advanced greatly between flower initiation and harvest greatly limiting plant development at VHF and that disease was not present at UMRF in 2021; also nitrogen (N) may have also played a factor.  More N was applied at UMRF than at VHF, which at this time might have influenced plant development.  Both location and weed management methods significantly influenced plant growth at harvest.  The resinous hemp grown in the mowed rye plots remained the smallest throughout the trial and plant yield and all other components reflected the poor growing conditions.  Interestingly, the poor growth was perhaps not an N immobilization effect but was more likely due to poor rooting or allelopathic chemical leaching out of the rye and hurting growth as the whole plant total N was highest in this treatment in part due to the small plant structure. 


Figure 5.  Plant area at harvest.




Harvest:  Yield and Cannabinoid Content


Plant yield, whole plant N, and cannabinoid concentrations and yield are presented in Table 6. 


Table 6.  Whole plant and flower dry matter yield and cannabinoid concentrations and yield as affected by weed management measures.






















Whole plant Total











Black Plastic



























Mowed rye













































As one might expect, whole plant and flower dry matter yield generally reflected plant size differences found throughout the study.  The yield differences found by location likely would have been greater without the infestation of Septoria leaf blight at flower initiation.  What was unexpected was the lack of a response of cannabinoid concentration by weed management.  It used to be thought that plant stressors would cause a spike in cannabinoid concentration.  However these data and work recently published by Toth et al., 2020 indicate that genetics and time to harvest influence cannabinoid content more than environmental stressors.  That was also supported by the fact that we harvested the VHF location two weeks earlier than UMRF and total CBD, total THC and total cannabinoid content were all significantly higher at UMRF than VHF.  Cannabinoid yield did reflect the weed management based on its effect on whole plant or flower dry matter yield.



Research conclusions:


The project team learned a tremendous amount this year with respect to resinous hemp production.  While I expected that not tilling the soil would slow down production, we didn’t expect that degree of a set back to the hemp in those treatments.  Next year, we will repeat all the sampling measures, but we will also collect fan leaf total N at a month following transplanting and at flower initiation to correlate with leaf SPAD values and ratios to confirm if N is limited in the rye mulch treatments.  We also learned that the extra weed pressure apparently stressing the plants did not influence cannabinoid content, and that length of time plants are allowed to grow and genetics appear to greatly affect cannabinoid content more than weed pressure.

Participation Summary
1 Farmers participating in research

Education & Outreach Activities and Participation Summary

8 Consultations
1 Curricula, factsheets or educational tools
1 On-farm demonstrations
2 Online trainings
1 Tours
2 Webinars / talks / presentations
1 Workshop field days

Participation Summary:

15 Farmers participated
19 Number of agricultural educator or service providers reached through education and outreach activities
Education/outreach description:

As mentioned previously, weed management and reducing cost of hemp production were the top production issues last year based on surveys from the Maine state hemp program staff.  Interest in alternatives was high.  We conducted one field day at the University of Maine Rogers Farm, and 15 farmers and the state hemp program staff attended.  I am putting this weed study information into my on-line hemp course which will be offered this fall.  I am conducting an on-line hemp program on weeds and diseases this week at the Agricultural Trades Show on-line in Augusta. 

The research report will be put on the Ag and Water Quality Research Report section of the University of Maine Cooperative Extension webpage.  

Learning Outcomes

12 Farmers reported changes in knowledge, attitudes, skills and/or awareness as a result of their participation
Key areas in which farmers reported changes in knowledge, attitude, skills and/or awareness:

At the end of the field day, all farmers participating reported that they were interested in learning more about the results of the rolled crimped rye, but they were uncertain as to whether it would be successful.  This was because they heard from Dr. Gallandt at the field day that it was his experience that the mulch slows weed growth for a period of time, but more mulch needs to be added to be ultimately effective. 

It turns out that Dr. Gallandt was correct.  I am doing more outreach at our Ag Trades Show this week.  I'll be interested to see comments from those attending if they still have the same level of interest in the no-till mulch systems.  

Project Outcomes

2 New working collaborations
Project outcomes:

This will be reported upon at the end of the project.  we are still midstream.  

Assessment of Project Approach and Areas of Further Study:

This will be reported upon at the end of the project.  we are still midstream.  

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.