Progress report for FNE25-125
Project Information
We wish to see for which crops the effort of using paper mulch for weed control is worthwhile on a small-scale, no-till market farm.
By effort we mean the time involved in from bed preparation to harvesting (and including all steps in between including weeding).
By worthwhile we mean the weight or dollar value of crop harvested.
We expect the main expenses of paper mulch will be in time spent on bed preparation and seeding or transplanting, and the main benefits will come from reduced weed pressure and reduced time weeding. However, our aim is to measure benefit as the amount of crop harvested regardless of weeds or other differences. Our output variable is how much marketable product we can harvest for a given effort.
Weed control is a major activity on organic farms. There are many techniques used, with the most common being tilling. Other options include solarization, cover crops to reduce weed growth, burning with a weed torch, as well as hoeing and tining between crops. Also used are various organic mulches such as straw or well-aged compost. Quite commonly, a physical barrier is put down and the crops sown into holes in the barrier. This barrier is typically plastic mulch. On larger operations the plastic is laid out by tractor, while smaller operations will do so by hand. Often they use a very thin plastic that is thrown away, or sometimes a thicker mesh row cover than can be reused. Sometimes the thin plastics are advertised as biodegradable.
Plastic mulch, biodegradable or not, is made from petrochemical products that are not long-term sustainable. Moreover, there is a growing body of evidence that plastic mulches leave behind residues in the soil that may not be healthy for the soil, or for the people eating crops grown in that soil. These chemical residues don’t degrade quickly so using plastic mulch can have long-term effects on agricultural land (Gao et al, 2019), on the environment in general (Huang et al, 2020), and potentially on the health of those who eat crops grown in fields with plastic mulch (Yuan et al, 2022).
Carrot Corner Farm has tried to avoid the use of petrochemicals, including using hand and electric tools rather than gas-powered, and not using single-use plastics (even our bags at the farmers market are paper bags). But we do have the same weed problems as other farms. Our one acre is surrounded by naturally occurring “weeds” – native and non-native prairie plants whose seeds and seed bank constantly sprout in our crop rows. Weeding is one of the time-consuming activities on the farm, and we would be able to expand with the same labor costs with better weed control. We imagine many other small scale farms such as ours face the same desires to be more sustainable and produce less microplastic toxins while at the same time facing challenges controlling weeds.
Last year, the NOFA bulk order included an option for a paper-based mulch, and we tried it on a couple of our rows. Anecdotally, it seemed to reduce the weed pressure on those rows significantly, but added quite a bit of up front effort both laying down the paper and keeping it in place, as well as time in transplanting through the paper. It’s not clear to us the balance between extra time up front and less time weeding later on. By rough measurements, some crops seemed to do well in the paper mulch, while with others there was not a noticeable difference. We also were not able to notice whether soil health changed under the mulch. We would like to investigate the reward to effort ratio of using paper mulch. We are particularly interested these questions for small scale no-till commercial organic farms in the Northeast such as ours.
Cooperators
- - Technical Advisor
Research
We did paired comparisons of paper vs. no paper plots with a total of nine crops in the 2025 season. The ground was prepped the same in each plot. Typically we used the shuffle hoe to remove weeds, loosened the dirt with a broadfork, added compost, smoothed the bed out, then laid paper on whichever part was designated and transplanted equal numbers of the same varieties of each crop into both paper and no paper plots. Of the nine crops where we did this, two crops (kale and strawberries) did not grow well and we don't consider the data very useful. For the rest, we were able to collect data on both time devoted to various tasks involved in the crop and harvest totals.
Unfortunately, give the small scale we work on, we switch tasks often, and we weren't always great at keeping track of time on each specific task (we had a nice app but would sometimes forget to switch from task to task and then have to guess). So the time on task data have a decent amount of error in them. Nevertheless, we think they give a good enough picture to see any major patterns. We also sometimes forgot to record harvest totals, but in those cases we would forget to record both from paper and no paper plots on the same day, so that should not skew the data.
To lay paper we would roll it out on top of the bed and then weight it down by digging a bit of dirt next to it and holding the sides down with dirt. This was effective for holding the paper, but meant that there was a small gutter in the walkway and weeds could grow on the edges of the paper where the dirt was piled on. Using ground staples was much faster, but the paper tore right through them in the wind. Late in the season we came up with another method where we cut 1/4" strips off 2x4 studs on a table saw, put those on top of the edges of the paper, and then held those down with ground staples. So far, that seems the most effective method and is less labor intensive than shoveling dirt.
For six of the crops we attempted to replicate the setup at least twice, but only two crops (tomatoes and zucchinis) had useful data from the replicates. We had a nice replicate on onions as well, but the harvest was not weighed properly in the second replicate. For lettuce and cucumber we did not do a good job distinguishing the second replicate from the first so the time on task and harvest data are summed across both replicates. We had replicate plots of strawberries, as mentioned above those did not grow well for us last year.
Garlic had already been planted the fall before, so for a small plot of garlic we laid paper after the fact by cutting lines into the paper where each garlic plant was poking out of the soil. We did this for approximately a 30' stretch of the bed and then stopped as it was quite time-consuming.
We did not see a lot of time difference on most tasks between the paper and no-paper plots. Prepping the plots prior to laying the paper was the same, although in plots without paper there was a temptation to do extra weeding where the shuffle hoe didn't take all the weeds out, whereas in the paper plots we assumed the paper would take care of the weeds. Planting time was by and large similar in paper vs. no paper plots - there was some variation in which took more time between the crops, but not consistently, and our impression was the paper might slow us down sometimes but not by a lot.
Actually laying the paper could add significant time - 40% - 100% more time above the rest of the field prep. However, in absolute terms it generally wasn't that much time - 30 - 60 minutes for each 90' row, and that will be less with our newer technique for holding down the edges with strips of wood. It would also be less time if we did it for multiple rows at once rather than dragging the roll out separately for each row.
We saw three crops where using paper was a clear win, onions, cucumbers, and tomatoes (see Harvest figure). Although we did not do a good job recording data in a replicate set of onion plots, qualitatively the result was the same - the second plot with paper had a much better harvest than a second plot without paper. We also had a paper plot with red onions, without a matched no-paper plot. That bed got particularly weedy as we didn't have time to deal with it (the weeds grew up through the holes we cut for the onions), but the onions came out nice and large anyway - the paper must have done enough weed suppression to let the onions grow.
Tomatoes were not as dramatic, but both the cherry tomatoes and the large tomatoes (we are considering those replicates) showed higher yield in paper vs no paper, and our qualitative impression as well was that the plants in paper did better. Similarly, cucumbers showed higher yields in paper, although we do not have good data from the replicate plot we planted. Garlic also showed slightly higher yields, but the plot with paper was quite small so we have less confidence in that result.
By contrast, the yields for lettuce were less in the paper than no paper. It seemed to us that small lettuce transplants, in particular, had more trouble getting going in paper than in dirt - the small leaves sometimes stuck to the paper and died, and the ones planted in paper often seemed more sickly. Peppers also showed smaller yields in paper than out, but our pepper crop was small in any case and this could have been a fluke. Zucchini plants seemed to grow better outside the paper as well, but a second planting did better in paper, so that was inconclusive.
Weeding time was consistently less in the paper plots, as one would predict (see Weeding figure), including for all the crops mentioned above. In some cases, the paper was effective enough that we didn't need to weed at all. With the tomatoes, the rows that didn't have paper had wood chips as mulch (our previous strategy), but the paper was more effective at suppressing weeds than the wood chips. The worst weeds in the paper plots was bind weed which grew along under the paper until finding one of the holes where something was planted and then climbing the plant. The paper seemed to concentrate the bind weed around the crop. Grasses also grew up through the holes and was harder to weed than in the no-paper plots.
Harvesting time did vary some but not consistently between treatments and we think the differences might have been measurement error rather than real. Qualitatively, it did not seem like harvesting was different between paper and no-paper plots.
The other difference we noticed we did not quantify, but for a while we were having trouble with our drip irrigation. That meant we had to water in some other way, as well as provide fertilizer. This was tricky in the paper plots, as hitting the hole where the transplants were was time-consuming. The paper really only worked for crop where watering was not a big issue (our onions did not require a lot of watering this year) or once we had drip irrigation operational - drip seems pretty critical to make it work.
We were able to meaningfully compare seven crops in paper vs. with no paper in the 2025 season. Several other crops which we tried to include in the study did not generate meaningful results. We only had useful replicates for three of the crops, for the rest we only had single useful comparison plots with paper and with no-paper.
The top-level summary is that using paper seems like a win for onions, tomatoes, cucumbers and potentially garlic. Harvests were higher, and total time spent on field work was lower with the paper
Paper seems neutral or had mixed results with zucchini, and sweet peppers.
Paper seemed a negative for lettuce.
We would feel much more confidence in these results with some replicates, so we will repeat the study on those crops in the 2026 season. With our garlic, we already planted some of it for 2025 in paper rather than trying to lay the paper on after the fact. We will replicate on the rest of those crops in 2026 as well.
However, we will conduct the study once more in the 2026 season to generate replicates for more crops and see if the results hold up over more replicates and whatever different conditions this season brings.
Project Outcomes
After the first year, we learned a few things.
- We lost data due to not being thorough enough in our explanations to farm assistants of the methods of the study. In year 2, we will try to do better at making sure any assistants are aware of exactly how to do things in the study, and save some critical jobs (such as some of the harvest measurements) for ourselves.
- Tracking time is difficult on a working farm. Even with an app, it is hard to be constantly switching time between different tasks, and we made mistakes. Not sure yet how we'll improve on this for year 2.