Final report for FNE23-068
Project Information
This project sought to develop an improved method for growing carrots in an organic system, specifically addressing the problems of uneven germination and in-row weed pressure that many carrot growers face. We aimed to use a system of applying strips of finished compost to the bed top and seeding carrots into the finished compost, hoping that the compost would suppress weed germination and improve stands of carrots because of its moisture retention and unlikeliness to crust over. Two methods were tried, a hand-scale method for smaller farms, and a tractor-scale method for larger plantings.
We started with the hand-scale method and tried different rates of compost application and control plots. We determined that a 3" deep application of compost significantly outperformed both the shallower applications and the control plots, reducing weeds in 2 row feet from 81 to 18. This treatment would result in much lower hand weeding labor costs, which more than offset the cost of the compost applied, even before accounting for the additional fertility benefits of the compost.
In developing the larger scale method of applying compost strips, we ran into some challenges in applying even strips of compost. However, with further refinement, we believe that this method is viable for improved carrot production on our farm and others. We presented our results at the 2026 PASA conference and to other farmers on our home farm and we are eager to continue this method in the future.
This project seeks to develop a production method for direct-seeded carrots that will improve germination and reduce in-row weeds by seeding into strips of finished compost.
In year one, objectives were:
- To determine an efficient system for preparing compost strips on a small scale.
- To compare results in in-row weed control using different depths of applied compost strips and different compost products.
- To compare results in germination, and in weed species present, using different compost products.
In year two, objectives were:
- To develop a mechanized system to apply compost strips
- To determine the optimal compost product and compost strip depth by comparing the most promising four treatments from Year 1 in a replicated field trial and analyzing the data.
- To assess the financial feasibility of using this system by comparing the input and labor costs of planting into compost strips with the costs of using our standard system of producing carrots.
Carrots have been a challenging crop for us at Hearty Roots for all of the 19 seasons we have farmed, whether on a tiny scale or on a larger scale. Although we have had some great success, there have also been many frustrations and failed plantings. We face two primary challenges:
- Weed pressure. Because carrots are slow to germinate, and must be direct seeded, weeds have an opportunity to get ahead of carrot seedlings. While weeds between rows can be managed through cultivation, weeds in the row must be hand-weeded, usually at a peak time of the season when labor hours are stretched thin. We have also used flame weeding to attempt to control weeds in the row, by burning the newly germinated weeds before the carrots emerge. However, this method has several drawbacks: it is risky, as carrot seedlings can be burned inadvertently; it is fuel intensive, as it relies on propane; and it is an unpleasant and time-consuming task.
- We farm sandy and gravelly soils, which do not hold moisture for long. To encourage carrots to germinate, we need to irrigate with sprinklers almost daily. However, this practice also encourages weeds to germinate, and can also lead to crusting of the soil which prevents carrots from sprouting up.
When we have good germination and we are able to control in-row weeds by hand weeding, a clean bed of carrots is a joy to harvest, yielding quickly picked bunches of profitable, high-quality roots.
We will trial a system for growing carrots that we believe will improve both the in-row weeds problem and the germination problem. After creating narrow trenches on the vegetable bed top for each carrot row, we will fill the trench with finished compost, and then seed directly into the strip of compost. The compost strip will serve three purposes:
- The weed-free compost will smother weed seeds that are in the soil below the strip, preventing germination of weeds in the carrot row.
- The friable texture of the finished compost and its ability to hold moisture will improve carrot seed germination by resisting crusting at the surface and by keeping the seeds moist.
- The application of compost will also provide fertility to the growing carrots, a method we already use on the farm on a broadcast basis. This system will use a similar quantity of compost, but carefully targeted in its application, to provide the additional benefits listed above.
If we are able to demonstrate a reliable system for germinating and growing carrots with minimal in-row weeds through planting into compost strips, the benefits to the farm will be many. We could reliably seed carrots through the summer, without losing plantings that get too weedy because we are too busy with peak season chores; we could improve harvest efficiency as beds with fewer weeds are much faster and easier to harvest than weedy beds; and we could improve our efficiency of land use, with higher yields per foot due to improved germination rates.
We are optimistic about the potential for this project, but also aware of possible pitfalls, which is why running a careful trial is important. Possible complications that we will be on alert for as we progress through the trial include: might some of the compost products have weed seeds that end up germinating in the row? Will the high organic matter of pure finished compost cause carrots to fork? Our hypothesis is that this system will be an improvement in all of these areas over the traditional system, and this trial will give us the data to test that hypothesis.
Hearty Roots Community Farm is a diversified farm in the Hudson Valley of New York. We produce certified organic vegetables for 1000 CSA members in the Hudson Valley and NYC, as well as for our retail farm market and other outlets. Animals are also an important part of our operation, producing eggs and meat for our customers, as well as adding fertility to our fields and helping to manage our pastures.
Cooperators
- - Technical Advisor
Research
This project was planned to be carried out in two phases, over two years. However, during the second year, the results were not good due to planting abnormalities. Therefore the project was granted an additional year to carry out the trial for Phase 2.
Phase 1 (2023):
- To determine an efficient system for preparing compost strips on a small scale:
- To make trenches, we prepared a vegetable bed as we usually do with our tractor mounted bed former, and then passed over the bed with a farm-fabricated tool in order to open a 3” wide trench of the desired depth at the location of the planting row. For this phase of the trial, we applied the compost into the trench by hand.
- We planted carrot seeds into the compost strips using a Jang hand-pushed seeder, regarded as the most precise tool to use for seeding carrots on a smaller scale. We used a single line of seeds per row, 2 rows per bed.
- The carrot variety Napoli was planted, a standard variety for fall fresh market carrots. Carrots were seeded on July 26.
- We tracked the time spent and the results of this less mechanized system of preparing compost strips on beds, in order to determine the feasibility of this system for small scale growers, and will perform a cost benefit analysis of using this system on a smaller scale.
- To compare results in in-row weed control using different depths of applied compost strips and different compost products, and to compare results in germination, weed suppression, and weed seed content of different compost products:
- 18 plots were be prepared, each plot was a 10’ long section of a 6’ wide raised bed, with 2 rows of carrots to be planted per bed, in rows 32” apart.
- Nine treatment types were trialed, based on the following variables:
- Five types of planting strips were prepared:
- 3” wide x 1” deep
- 3” wide x 2” deep
- 3” wide x 3” deep
- A control area with no trench and compost scattered on the surface
- A control area with no trench and no compost
- Each depth of planting strip was prepared using each of the two types of compost to be trialed*:
- McEnroe Organic Compost
- Hudson Soil Co. Compost
- Five types of planting strips were prepared:
- Each of the 9 treatments was used on 2 plots, in a randomized arrangement, for a total of 18 plots planted.
- Plots were irrigated with micro sprinklers so that soil surface did not dry out before germination.
- 2-3 weeks after planting, the number of weeds that had germinated was assessed. Weeds were be counted within the 3” wide planting row in a defined area of each plot. Carrot seedlings were also counted in a defined area in each plot to determine an average germination percentage for each plot.
- 3-4 Weeks after planting, half of each plot was hand weeded.
- All beds were tractor cultivated between rows, as is typical with organic carrot production, first with a basket weeder and then with sweeps.
-
- *We had intended to trial municipal compost from Ulster County Resource Recovery Agency due to its low cost, but we were informed by our organic certifier, NOFA-NY, that it is not approved for organic application. We considered replacing it with Vermont Compost, but we determined that this pricier product, coupled with long shipping distance. could not be an economical choice even if it worked perfectly.
After the completion of Phase 1, data was assessed for three criteria:
- Relationship between depth of compost trench and in-row weed control
- Weed density was assessed 3-4 weeks after planting. The plots where compost was broadcasted had the highest weed density, averaging 80.8 weeds per 2 row feet, which was worse than the control plots 63 weeds per 2 row feet. The treatments of 1" and 2" of compost performed better than the control at 46 weeds per 2 row feet and 42.8 weeds per 2 row feet. The treatment of 3" of compost performed the best, with an average of only 18 weeds per 2 row feet.
- Relationship between compost product used and in-row weed control and yield and quality of harvested carrots.
- There was not a statistically significant different between compost products in weed density in a given treatment depth.
- Carrot stand counts
- Carrot stand counts did not seem to be affected by either depth of compost treatment, or by compost product type.
Based on these results, it was determined the Phase 2 of the trial would use one source of compost and would rely on one treatment depth of 3" of compost. We decided to trial different treatments of planting seeds in order to determine the best method for carrot planting in the future.
Phase 2 (2024 - 2025):
Phase 2 began in 2024, to carry out these objectives:
To develop a mechanized system to apply compost strips:
- Working with our farm’s Mechanic, we modified our existing Kennco Bed Shaper, which has a hydraulically driven hopper with a conveyor system, to create trenches of the desired depth and fill the trenches with finished compost flowing from the hopper.
- To determine the optimal carrot seeding method to be used with this system, we planned to plant plots with 3 seeding treatments:
- Single line per row planted with Jang seeder
- 2" banded planting with scatter-shoe on Jang seeder
- Single line per row planted with MaterMacc vacuum seeder
We modified our Kennco Bed Shaper to make a 3" deep trench by custom fabrication steel shoes attached to the press pan of the shaper. Then we modified the fertilizer hopper to distribute compost by switching the orientation of the hopper and fabricating a chute out of sheet metal to bring compost to the rows. We thought we had been successful in making these adjustment and prepared plots which were seeded with carrots. However, it became clear as the trial progressed that our preparation was not adequate and that rows were inconsistent. Therefore reliable data could not be gathered for the different treatments in Phase 2.
On August 16th, 2023, Ethan Grundberg and an assistant (Cornell Cooperative Extension) measured the following in one 2-foot section of each treatment plot:
- Weed density by compost depth
- Weed density by compost product
- Carrots stand counts buy compost depth and compost product
The main takeaways from the data gathered in Year 1 are:
- Compost depth had a significant impact on weed densities. The plots where compost was broadcasted had the highest weed density, averaging 80.8 weeds per 2 row feet, which was worse than the control plots 63 weeds per 2 row feet. The treatments of 1" and 2" of compost performed better than the control at 46 weeds per 2 row feet and 42.8 weeds per 2 row feet. The treatment of 3" of compost performed the best, with an average of only 18 weeds per 2 row feet.
- There was not a statistically significant different between compost products in weed density in a given treatment depth.
- Carrot stand counts did not seem to be affected by either depth of compost treatment, or by compost product type.
The reduction in weeds was very significant, with the 3" deep bands of compost averaging 18 weeds per 2 row feet, while the control area averaged 80.8 weeds per 2 row feet. (see chart)
While carrots at the lowest weed density would still require hand weeding, the reduction in weeds from the control would result in significant labor hours saved over the course of the season, approximately 8.5 fewer hours per 500' bed of hand weeding, assuming that 1 weed is pulled per second. We expect that the hand weeding is also done more quickly at this lower density because carrots are easier to see and there is less likely to be damage to the carrots from pulling out adjacent weeds.
The cost savings in dollars for reduction in hand weeding is approximately the same as the cost of the compost applied for a 3" deep application per row; reduced labor saves about $200 / bed, compared to the cost of compost of $195 / bed. But the compost provides additional benefits of increasing fertility and soil organic matter, and also frees up workers to focus on other things at what is usually a busy time of year.
In Phase 2 of the project, carried out in 2024 and 2025, we designed a mechanized tool for creating the compost strips, built onto our Kennco bed former and fertilizer conveyor, with custom welded shoes to create the 3" deep trench.
We custom welded shoes that we bolted to the shaper pan of the bed former. The shoes did a good job making the trench of the correct width and depth, although the trench was sometimes inconsistent depending on field conditions when the beds were made. We have since started making beds with a Forigo stone burying bed former, which we hope will improve conditions and make trenches more consistent.
We found that the fertilizer conveyor was able to dispense the compost at sufficient rate to fill the trenches, although it was helpful to have a second person on the implement agitating the compost to ensure that it did not bridge and flowed consistently.
The mechanism to convey the compost from the hopper into the trenches was changed several times over the course of the project. First we tried using sheet metal and gutter material to gravity feed the compost from the outlet of the hopper to the trenches. The material was too flimsy and would move too much in operation resulting in inconsistant placement of the compost. Later we tried bolting a plywood chute with sheet metal guides to feed the compost, which was an improvement but was still not perfect due to compost not flowing smoothly enough down the chutes.
With the combination of finer seed bed preparation from the Forigo stone burying bed former and the improved delivery system, in the future we are hopeful that this system is viable for planting on a larger scale.
We hoped to trial planting with a tractor-mounted vacuum seeder and compare results planted with a hand-pushed Jang seeder (using both single and scatter-shoe plantings), however the compost strips were not even enough to compare results from these plantings.
We sought to develop and refine a system for carrot growing that would improve stands and reduce the need for hand-weeding, at both a smaller hand-scale and a larger mechanized scale.
We succeeded in developing a system for smaller scale production that worked very well on a hand scale, reducing weed pressure in-row from 81 to 18 weeds per row foot, by applying 3" deep strip of finished compost to the row before seeding. The cost savings of reduced hand weeding ($200 / 500' bed) more than covers the cost of the compost purchased for this method ($190 / bed), even before taking into account the additional benefits of fertility from the compost.
Although we were not able to perfect a mechanized system for applying the compost strips in Phase 2 of our trail, we did make progress toward doing so, and the results from Phase 1 show us that continuing this work will pay off due to cost savings and by freeing up labor at a very busy time of the season.
Education & outreach activities and participation summary
Participation summary:
Results of the project were presented at the PASA Conference in Lancaster, PA in February of 2026. Benjamin Shute was invited to participate in a workshop at which three SARE Farmer project grantees spoke about their projects and the process of applying for SARE Farmer Grants. The panel was attended by a few dozen agricultural service providers and farmers. PowerPoint slides from the presentation are included here.
In addition, the ongoing trials and results were presented to several farmers who visited Hearty Roots Farm over the seasons that the trials took place, and to members of the general public on farm tours.
Learning Outcomes
Attendees at the PASA Conference presentation learned about this project, as well as dozens of farmers who visited Hearty Roots Farm and had informal conversations with the Project Leader and other participants about the project and its results.
Project Outcomes
This project will result in our farm changing our carrot production practices. It remains to be seen if other growers will adapt their systems based on this project, since it has just been completed. It also remains to be seen if additional grants or projects will be informed by this project, as it has just been completed.
The first stage of this project was very well designed and had very successful results. It went according to plan, was straightforward to analyze, and had clear results.
The second stage of the project was not as smooth. The development of a mechanized system for applying the compost strips took much trial and error, which led to uneven results and difficulty measuring results reliably.
We answered the questions we had upon starting the project for hand scale application of compost strips for carrots. The viability of mechanized plantings is still uncertain based on inconclusive results in phase 2.
Any growers of organic carrots who struggle with uneven germination or high hand weeding costs would benefit from these results as they try to improve their own systems.