Farmer-led cover crop trials and demonstrations for vegetable and corn silage fields

Vegetable Farms
Vegetable Farm 1:  High Biomass cover for roll-down
The early roll was largely unsuccessful as more than 50% of the rye stood back up after being rolled at seed head emergence, and most plants did not terminate without chemical (glyphosate) application.  The late roll (during pollination) was more successful with approximately 75% of the rye rolled down and 25% of the rye stood back up.  Both roll-downs (at seed head emergence and at pollination) were considered unacceptable for the farms organic vegetable transplant conditions, without chemical use and without threat of the rye going to seed.  The whole field was later terminated with chemical burndown. No crop was transplanted to this field. Cover crop dry matter to estimate biomass was collected at the time of roll-down.  The earlier roll-down (seed head fully exposed) had an average of 2.7 tons dry matter per acre (ton DM/ac).  The later roll-down (pollination) had an average of 3.5 tons DM/ac.  The other two fields were accidentally plowed under, however there was opportunity to collect cover crop for dry matter estimates of biomass from a few missed spots still standing in the field.  Their average dry matters at plow-down were 1.0 ton DM/ac in the wet field, 2.5 ton DM/acre in the adjacent, slightly drier plot, and 2.8 ton DM/ac in the third trial field.  The third trial field had a split field with one half rye and a second half hairy vetch, which had been discussed for possible roll-down, as it reached maturity.  Purely out of curiosity, a biomass sample was taken in the vetch, at flowering stage, for a 1.6 ton DM/acre average.  Across the whole experiment, the farm was targeting 3 to 4 tons DM/ac based on recommendations from the 2016 Baltimore SARE Cover Crop conference, and specifically Dr. Ron Morse, Virg. Tech, where his talk included tips on successes and failures of organic cover crop roll-down for weed control (Fitting Cover Crops in Vegetable Rotations;  https://www.youtube.com/watch?v=77wQXZlYwSs).  Combined with the recent arrival of a no-till transplanter, the farm was very interested in organic methods to reduce weed pressure (labor saving) and reduce plasticulture use, though only on appropriate crops.  The farm advised that some crops did so much better (productivity & sales) on plastic, they would continue to use plastic cover in some instances (tomatoes/peppers), but the farm was interested in eliminating plastic use where feasible, such as some of the vine crops like cucumber and pumpkins.  Keeping the vegetables up off the bare soil and away from the immediate bacterial host or soil splash was very attractive to the farm.  But the learning curve to get there was not a one-n-done process.  The cover crop trials from 2016 and 2017 were used to refine future attempts at high biomass cover crop and no-till transplanting.  

Vegetable Farm 2:  High Biomass cover for roots and for roll-down
The original experimental design of a randomized complete block with four replications and three mixes was laid out and seeded entirely by the farm in perfect form.  However, due to the very dry conditions in the summer and fall of 2016, the seed mixes did not take.  The seeds were broadcast on the soil surface, the growing conditions were dry, and the seed mixes did not germinate nearly as well as the weed seeds germinated.  The farm made a valiant effort to allow the late rains to help the seed mixes grow/catch-up, but the weeds absolutely dominated the site and a quick decision was made to mow the plot before the weed seeds developed further and caused future problems.  After some quick consultation, the farm established two additional plots to evaluate and learn the specifics of managing large biomass cover crop.  Field 2 had large biomass with a 150 lbs/ac rye seeding.  Biomass sampling revealed 2.9 ton DM/ac on average for the 0.2 acre plot, planned for curcubits in the second planting, around early to mid-June.  The plan was for the rye to be rolled with a chevron roller-crimper, but at the time of need, the roller was unavailable and the transportation was not easily accessible.  The farm is not a fan of rye because they've had so much trouble controlling it as a volunteer crop in past years.  Because of their experience, they wanted to be pro-active and not allow the rye to mature to viable seed.  After word came back the roller was not easily available, the field was flail mowed, then solarized, and lastly plowed, to completely kill the rye cover. The third field, planted to peas/vetch/oats (PVO) for high biomass, developed nicely and grew in as intended.  The opportunity to test rolling & crimping the mature plants for use as a rolled mulch weed barrier, in place of plastic rows or mulch, was looking good.  However, the same roller/crimper was unavailable at flowering.  Due to the grave concern for volunteer plants in the vegetable fields, this field was also mowed.  No biomass samples for dry matter were obtained from the PVO mix.

Vegetable Farm 3:  Living Alley Cover and Pollinator Habitat
Pollinator Habitat:  The phacelia plots were very successful as a pollinator draw to bring pollinators to the plot and near the surrounding vegetables on the farm. At peak flower bloom near June 26, 2017, the number of bees in the plot were estimated to be (very roughly) 50 bees per cubic foot.  It was an amazing site to see - so many pollinators working in the plot.  While it was exciting getting the measurements and plant samples, the bees were very focused on their job collecting pollen and, with slow but steady movements, the data was collected and there was no concern for stings in or around the immediate area.  Biomass weights for the phacelia were 2.5 – 3 tons DM/ac with plant heights averaging 39 inches.  Plant height was measured as canopy height, as observed.  The plants were robust, lots of vegetative growth and flowers, and very intertwined.  The plant lengths, fully-extended, were very likely well beyond 39 inches, but the plants were not measured for length, only average canopy height.  The size, scope, and diversity of pollinators absolutely met and exceeded the farms goal of bringing in higher numbers and more diverse pollinators to the farm to help pollinate surrounding vegetable plots.  The farm tries to keep some cover crop in flower throughout the growing season, which is a challenging task; a task often reserved for the winter months planning sessions with the seed catalogs close at hand.  

Living Alley Cover:  The living alley cover experiment did not go quite as planned.  One of the many ideas was to go with a three treatment comparison and evaluation of:  Perennial Rye, Rye + Clover, Crimson Clover.  The farm was very interested in living alley covers to improve soil health, reduce the amount of exposed soil, and dramatically reduce the hours of weeding with the use of a managed alley planting.  The alley cover varieties were chosen based on the plants ability to germinate and grow within a vegetable season and its ability to handle the frequent travel of foot and/or wheeled traffic through the vegetable plot.  However, as the growing season evolved, and challenges arose, the weed pressure in the allies continued to build - straw was purchased to apply to the allies before the weed pressure reached the farms threshold.  Upon observation of the straw and the farms use/placement, there was a noticeable difference between what the farm used in straw and what was available after shredding (flail mowing) a tall cover crop in-place.  The cover crop was cereal rye + spring oats + hairy vetch seeded mid-September (9/18) and terminated early June (6/8).  The shredded cover crop left a nice, even cover across the plots with approximately 2 inches depth.  This did not provide adequate cover to prevent weed germination and growth, led to problems when trying to flame the weeds in the row (fire), and led to noticeable nitrogen sinks through the growing season with detrimental effects on the cash crop growth and yield.  By comparison, the straw bales were applied at 4-5 bales per alley (19-25 t DM/ac, assuming 40 lb bales at 85% DM) to meet the farms needs for weed suppression.  This resulted in mulch depths of approximately 5-7 inches (as-placed / un-compacted) and greatly reduced the time and effort spent weeding the alleys.  Additionally the vegetable rows did not have detrimental effects of N sinks with the high carbon cover concentrated in the alley, allowing for a more crop-friendly row cover crop to help ease the carbon:nitrogen (C:N) balance.  There was a very quick and strong learning curve with the mulched alley's.  In part, as a result of the farms continuous experimenting and efforts to improve their soil, reduce labor costs/time weeding, the farm no longer uses straw mulch and has converted to no-till vegetable beds.  

University of Rhode Island

No-Till transplanting of Vegetables into Killed Cover Crop Mulch
This experiment was intended to utilize the no-till transplanter purchased by a local, cooperating farm (Farm 1). Vegetable growers, particularly organic growers, are very interested in using killed cover crop mulches as an alternative to plastic mulch. Plastic mulch layers require tilled soil, and do not handle large clods, rootballs, or stones well so fields are generally rototilled or intensively disked prior to bed shaping. Petroleum-based plastic mulches result in non-degradable waste, while bio-based mulches are not permitted in Organic production. Most vegetable planters/transplanters require tilled soil, so strip tillage is a common approach. However, previous experiments have generated mixed results, largely due to problems controlling weeds in the tilled strips.

A 1.5 acre field was seeded to winter rye and hairy vetch on Sept. 13, 2016 using a rate of 60 lb rye and 23 lb vetch per acre. The field had been in summer cover crops in 2016 and was not tilled; seeding was done using a Truax no-till drill. The cover crop was terminated May 16, 2017 when rye was headed. Glyphosate was used to kill the cover crop, which was then rolled with a roller-crimper.

The plan was to plant 1 acre of butternut squash and 0.5 acre of broccoli using the no-till transplanter immediately following cover crop termination. Unfortunately the no-till transplanter turned out to weigh significantly more than expected, and none of the tractors available had sufficient horsepower to tow it.

In an attempt to salvage the experiment, a parabolic chisel plow was used to cut rows, and the squash and broccoli were transplanted by hand. Squash was planted with 5 feet between rows and 2.5 feet within the row. Broccoli was planted with 2.5 feet between rows and 1 foot between plants. The equipment problems and wet weather combined to delay planting until June 16, one month after cover crop termination.

Vetch dominated the cover crop biomass at termination. Mulch cover was measured using digital images taken on June 29. At this point the vetch had entirely decomposed, leaving the rye straw as the sole mulch. Mulch cover in the squash field ranged from 10% to 87% with a mean of 53% across 20 random locations. This was not sufficient to prevent weed invasion, particularly yellow nutsedge. The area planted with squash was sprayed with halosulfuron (Sandea®) when squash plants had 4-5 true leaves to control nutsedge. The herbicide also controls other emerged broadleaf weeds. By July 28 over 50% of the surface not covered by squash canopy was covered by weeds, primarily purslane and chickweed. By late August the field had been swallowed by smartweed, with some barnyard grass and pigweed. The mulch was successful in suppressing weeds long enough for the transplanted butternut squash to form a canopy, and butternut yields were normal. However, poor weed control increased the difficulty of applying fungicides in August, and added to the weed seed bank to increase weed problems in future crops.

The initial mulch cover in the broccoli field was slightly lower at 44%. Broccoli is not tolerant of halosulfuron herbicides, so shielded applications of glyphosate were used to control nutsedge between rows of broccoli. The combination of cover crop mulch and glyphosate was sufficient to control weeds between broccoli rows, but weeds between plants in  the row were a serious problem. These had to be hand-pulled, and were much more difficult to remove than weeds in fields where soil was tilled prior to planting. Weed pressure was sufficient in some areas to prevent broccoli from forming full-sized heads.

Based on the results of this trial, URI cannot recommend killed cover crop mulches for vegetable growers in CT-RI. The method merits further exploration, but some significant obstacles exist. No-till transplanters require large tractors, limiting their use to the largest farms. The weight of no-till transplanters also makes any sort of contract operation or equipment sharing difficult, as they are challenging to load and unload from trailers. The cover crops for this study were seeded in mid-September, which is extremely early for vegetable farmers in our region. Even so, biomass production was not sufficient. Using pure rye, rather than rye and vetch, might have improved weed suppression, but would have increased the need to apply nitrogen fertilizer. Weed suppression was not sufficient to prevent yield loss in broccoli or late-season weed establishment in squash.

Teff as a Living Mulch for Pumpkins
Another approach to using cover crops for weed suppression is to plant living mulches. This is particularly useful when crops have wide area between row spacings, either because of a need for access for repeated harvest or for vine crops such as pumpkin. Studies in 2016 showed that teff (Eragrostis tef), a high-altitude tropical C4 annual grass, is well-suited to use as a living mulch in our climate. Teff grows very densely, suppressing most summer annual weeds. It is tolerant of mowing and foot traffic, and is more resistant to foliar diseases than annual ryegrass. Teff does not flower until late August, and rarely matures seed. The seed also has little dormancy, and will not overwinter in the field. In 2017 we conducted a trial to determine whether teff used as a living mulch negatively affected pumpkin yields.

The field used for this trial had not been tilled since 2015, and was planted to no-till hairy vetch at 30 lbs/acre over the winter. The vetch was incorporated with a Celli Spader in late May, and medium-sized carving pumpkins were seeded by hand on June 15. Spacing was 10 feet between rows and 2 feet between plants in the row; fertilizer was banded in the pumpkin rows. Prior to pumpkin emergence, teff was seeded at 7 lbs per acre between rows of pumpkins using a Brillion cultipacker. Three treatment plots of 3 rows of pumpkins per plot received teff; another three plots were not seeded to teff and served as the control. Mechanical cultivation was used to control weeds in the pumpkin rows and in the no-teff treatment until vines began to run. By that point, the pumpkin canopy filled the space between the teff strips in the teff treatment. Teff strips were mowed once to remove broadleaf weeds which germinated with the teff; the mower blade was set to just clip the tips of the teff foliage. Sandea® herbicide was applied to the entire trial for nutsedge control when pumpkins had 4-5 true leaves.

By late August, all plots without teff had weeds wherever light could penetrate the pumpkin canopy. In contrast, plots with teff only had weeds in the pumpkin row itself, and then only where the pumpkins had been defoliated by downy mildew. Teff living mulch had no effect on pumpkin yields or quality. Yields averaged 21,500 pumpkins per acre with an average fruit weight of 16 lbs.

Using teff as a living mulch between rows of pumpkins reduces the need for cultivation or the amount of pre-emergent herbicide required. It also adds biomass to the soil, helping to compensate for organic matter oxidized during tillage. Teff does not compete with pumpkins for water or nutrients, and does not reduce yields.

Silage Corn for Dairy
Corn Farm 4:  Nutrient Retention
Unfortunately no data was collected from this plot due to a series of unfortunate events including weather, time, communication, and misunderstanding.  As the 2016 mid-summer warmed up and the corn crop developed and matured quickly, harvest had to happen as time, labor, and equipment allowed.  Unfortunately for the experiment, that meant reducing the treatments from five down to two.  Then in the spring, again primarily due to time and availability (reality), the cover crop was terminated at approximately 4-8 inches of growth, after a second poor fall germination due to broadcast seeding in dry conditions.  In the end, the corn field was planted without obtaining any prior cover crop samples.  

Corn Farm 5:  High Biomass Cover Crop for Soil Health and Forage
The relative biomass for the rye was 3.0 ton/ac with an average plant height of 57 inches. The relative biomass for the triticale was 3.3 ton DM/ac with an average plant height of 41 inches.  The relative biomass for the triticale & clover was 3.4 ton/ac, though there was a wide range of values (1 – 5 ton/ac) across the relatively flat, variable moisture field.  The average plant height in the triticale & clover was 42 inches.  The field was harvested as a single crop and ensiled for cattle feed, but no feed samples were taken for analysis because the plants were so mature and known poor feed quality, except as fiber or roughage in a total mixed ration (TMR).  While the plot and the crop management were difficult for this site due to weather and wet areas throughout the field, the results of the extended growth on the cover crop was a valuable change when compared to frequently used crop crop management strategies in the area.  Typically the cover crops are chemically terminated early in the spring season (early April) if possible, to allow for 2-3 weeks rest before corn planting, into the preferred short height rye.  The the short rye provides some value for soil erosion and nutrient retention on the field, but not nearly to the value and extent of a tall rye or cover.  UMass (Hashemi, et al, J. Plant Nutr. Soil Sci. 2013, 176, 69–75) has shown a very strong relation to plant growth and nutrient retention within the cover crop, at the very least keeping the nutrient (nitrogen in this specific research) in the field (as plant matter) and much reduced environmental loss.  For farms with high nutrient pressure due to large manure volumes or high soil test values, the ability to grow and manage tall cover crop would potentially be a significant cost savings specific to nutrient cycling for crop up-take and reduced liability for nutrient loss.  In addition, the ability to harvest the cover crop rather than a (pre-plant) chemical application for termination, followed by a second application for post-plant weeds was a noticeable time and management savings for the farm.  In addition, the recent headlines of glyphosate and public perception in a fairly dense, populated area has the farm looking for options to maintain and improve public relations and perceptions.  The ability to reduce spray time and acres was attractive to the farm.  Lastly, because the field was a wet field, the farm felt the ability to stay off the wet field was much to their advantage to prevent compaction or ruts, which they've been working hard to reduce or eliminate.  By growing the cover crop tall, they were not pressing to get into the field to terminate in the early spring and then plant, and they were able to work the field when the soil (moisture) was ready.  Noticeably, the extra time and the extra biomass in the filed helped manage the soil moisture much better than if the cover crop were terminated early, and by waiting for the plant to mature, helped reduce the time pressure to get onto the field (April vs May or June).  These effects got the farm thinking about how to manage or juggle the different spring activities (cover crop termination, manure spreading, corn planting, and grass harvesting) and how they could use these beneficial effects to their (soil & crop) advantage.  

Corn Farm 6

Germination
The immediate difference between broadcast and drilled cover crop seeding method was evident soon after planting. The drilled plots had a faster germination time and higher plant diversity of the target seeds. The broadcast plots had very minor differences (non-discernible) between them, specific to germination. There was observation and speculation whether the manure initially helped with seed germination or plant growth, but the data did not show any significant difference in this instance.  As the growing season continued, there was tremendous weed pressure and distinct high water patterns across the field.  By the end of the growing season, there were continued differences in successful target plant growth and intended diversity with the better germinated drilled plots.  The harvest yield, as dry matter, had some surprising outcomes, which should be pointed out - are suspected adverse effects of the heavy weed pressure and less germination of the broadcast (target) seed.  The Drilled plots had the highest rate of target plants, the fewest weeds (visually), but then had the lowest overall yield of the four treatment methods (3.4 t DM/ac).  The broadcast plots had fewer target plants, greater weed pressure, and higher yields (4.0 - 4.6 t DM/ac).  Immediately following harvest, upon walking the plots, questions developed about the seeming presence of treatment effect on the tubers (radish & turnip).  This led to a stem count, a tuber count, and a categorical rating of the tuber count (low = 1 - high = 5) and tuber size (small = 1 - large = 5).  The presence of all stems (target and weeds) was greatest in the Drilled and the Cultipacked plots (est. 23 stems/sq ft).  The Broadcast (only) plots averaged just less than 20 stems/sq ft, and the Manured plots averaged just over 18 stems/sq ft.  The stems could not be identified by plant species (or even categorically as target or weed), but if allowed to extrapolate from the visual presence of weeds - the drilled plots had strong germination and plot growth of the target species where the Cultipacked had the equivalent prevalence in weed stems.  Where the weeds seemed to grow especially well in the broadcast plots - more weeds were present in the Cultipacked, however the Manured plots had the highest weed + target yield (4.6 t DM/ac).  As to the presence of and the growth of tubers in each treatment, the drilled plots had the lowest number of tubers and the smallest tubers of all four treatments.  The Manure plots had the highest number and the largest tubers of all the treatments.  A rationale is not readily available to try and explain or postulate this effect on tuber presence and growth.  However, one start may be to look at how quickly the tubers had to compete for sunlight.  The broadcast plots had more open canopy for longer, before weed pressure increased.  Going back to the original questions:  which seeding method has the highest percent and fastest germination of cover crops;  because of the success of the germination trial (seeing which methods had the most seeds came up the fastest), the farm did call in an order to purchase a seed drill before the experiment was even over, and they've been very happy they did so.  The farms original goal had been met and their question had been more than satisfactorily answered with the on-farm trial.  

Termination
The early and the late termination cover crop plots were planted to silage corn on the same date, and developed at the same rate (V1 – R6).  However, the later termination corn (planted into the green cover crop) measured as much as 2 - 3 inches taller than the early termination corn during the first stages of growth (V1-V6).  All the corn was planted at the same time, across the two treatments.  The corn emerged and developed at the same rate, but throughout the V1-V6 plant stage, the plots not chemically terminated prior to planting, were noticeably taller.  After the V6 stage of growth, both corn treatments plant growth and plant height became indiscernible, and both treatments had similar yields at the end of the year (23 t/ac).  There was very little cover crop pressure on the late termination plot as the cover crop had a tough establishment in fall 2016 due to dry conditions and broadcast seeding.  The final dry matter content of the cover crop across the field (both treatments) averaged 700 lbs DM/ac.  Despite the cover crop difficulties, the farm was happy with the success of the trial and has since planted corn into successive years of green and growing cover crop, then terminated with a single, post-planting chemical application approximately three weeks after seeding.  The farm feels this helped get a valuable three additional weeks worth of work and growth out of the cover crop, the primary interest being to help manage the moisture content in the spring soils, which are typically wet.  The farm has been a strong advocate for plants helping to manage soil moisture both in times of excess water and insufficient rainfall.  Which ever of the weather or soil conditions prevail, the farm feels having active and growing plants in the soil helps manage the moisture for the better, and makes the fields more accessible.  Plus, the benefit of getting three more weeks of growth from the cover crop before termination also means more nutrients in the cover crop and cycling through the root zone, and less subject to environmental loss.