Integrated cover crop innovations for biologically based no-till

Final Report for LNE06-244

Project Type: Research and Education
Funds awarded in 2006: $118,310.00
Projected End Date: 12/31/2009
Region: Northeast
State: Pennsylvania
Project Leader:
Stacy Glackin
Rodale Institute
Jeff Moyer
The Rodale Institute
Expand All

Project Information


In the three years of grant funding, over 100,000 people have been exposed to biologically-based no-till technology, and over 20 farmers in the targeted region have purchased equipment and are employing the technology on their farms. Farmers, extension agents, educators, and students were the primary participants in project activities, but policy makers and governmental officials also participated in many of the project’s workshops and field day events.

Two years of trials at three Pennsylvania locations, (PSU Rock Springs, PSU Landisville, Rodale Institute in Kutztown) produced viable crops of organic corn no-till planted into rolled/crimped hairy vetch cover crops, without any additional weed control or nitrogen fertilizer. Yields varied from 1.1 Mg ha-1 to 9.6 Mg ha-1. Due to the variety of potential reasons for corn yields fluctuation, more research is recommended to address key system issues such as winter survival and maturation of hairy vetch varieties, adjustment of no-till corn planters, and successful crop rotations to manage weeds in organic no-till corn.

Two locations (PSU Rock Springs and PSU Landisville) investigated use of rolled/crimped desiccated rye cover crops to eliminate use of post-emergence herbicide prior to no-till soybean planting for 2 years at each location. Rye was terminated at two dates in the spring, which resulted in much more biomass for weed control in the later kill date. In most cases, the rye cover crop reduced weed density and biomass compared with fallow and no post-emergence herbicide application. Delayed kill of rye only had an effect on weeds in one out of 4 site-year combinations. The rye provided control similar to post-emergence herbicide in 2 out of 4 site-year combinations. Soybean yields did not differ between treatments in 2007, the first year of the study. However, in 2008, Landisville soybean yields were the same in the “rye-no post-emergence herbicide” and the “post-emergence herbicide-no rye” treatments, while no yield differences were noted in Rock Springs. An economic analysis of the treatments indicated there was a net cost to the conventional (chemically based) system if rye was planted for weed management, even if a post-emergence herbicide application could be eliminated.

In both the corn and soybean systems, it was demonstrated that the ultimate success or failure of the systems was directly related to the density and biomass of the cover crops. Unique planter modifications and adjustments are required in order to successfully get the cash crop seed planted through the cover crop into the soil. More equipment research needs to be conducted in order to guarantee a greater chance of farmer success with either of these cover crop/cash crop systems.

On-farm research was performed at three locations in New York, in Pennsylvania, and in Maryland. (Bill Mason, MD; Kirby Reichert, PA; Peter Schuster Farm, NY). A cover crop summit was held in June 2007 at Rock Springs, central PA, attended by about 100 individuals. Two outreach publications titled ‘Roller Study News’ were published and widely distributed by The Pennsylvania State University, and many farmers were reached through conferences, workshops and outreach presentations where cover crop/roller information was presented. Two peer reviewed articles were written and entered for publication, (Agronomy Journal, Weed Technology), and three articles were posted on the New Farm on-line magazine. (Attachment items 1 thru 8). Valuable on-farm data were collected that will have an impact on future research activities.


No-till production as practiced on 17 percent of US crop land and 30 percent of crop land in the Northeast, has demonstrated an ability to conserve soil resources. However, conventional no-till practices are highly dependent on the use of herbicides. Rodale Institute has demonstrated that no-till planting, in combination with intensive cover cropping, can reduce or eliminate the need for herbicide use while capitalizing on the soil health benefits of reduced tillage and cover crops. Rodale Institute has designed a heavy duty roller/crimper to be used with cover crops to create a no-till planting system that plants crops without tillage, reduces or eliminates herbicide applications, and increases the sustainability of Northeast farms.

Rodale Institute partnered with The Pennsylvania State University and farmers in Pennsylvania and Maryland to increase adoption of cover crop roller technology among organic and conventional farmers across the Northeast. The effects of rolled cover crop management on no-till corn and soybeans were studied. Specifically, the potential of the roller/crimper to kill a hairy vetch cover crop, as well as the ability of the rolled vetch cover crop to suppress weeds and provide nitrogen for following organic corn crop, was studied at three locations for two years. Weed suppressive potential of a rolled cereal rye cover crop in soybean production was also compared to no-cover treatments, with and without post-emergence herbicide, in no-till soybeans in two Pennsylvania locations for two years. Finally, through collaboration between Rodale Institute and PSU’s Cooperative Extension Service, research findings were disseminated through dozens of farmer meetings, bulletins, and online communications.

This technology impacted many farmers across the region. Project field days and workshops attracted over 370 farmers from across the northeast region. As a result, I&J Manufacturing, the company that collaborated with Rodale Institute to put the roller/crimper into production for retail sale, sold 23 roller/crimpers in just Pennsylvania and New York since the start of the project. Some of the farmers who purchased the rollers noted that they not only used the roller themselves, but also lent the rollers to neighboring farmers who watched the implement being used and developed interest in trying the new technique in their own fields.

Performance Target:

Performance Target: We will track 160 farmers who attend training field days/events to determine their adoption of the practices. Of the 25 farmers who have been targeted as early adopters and who will use the techniques and tools demonstrated in this project, 10 will experience enough success that they will increasingly adopt the technology during the subsequent three years and will greatly reduce or eliminate the use of herbicides annually on at least 1,500 acres.

Over the three years of the grant, over 370 people participated in Rodale Institute Field Days that featured no-till organic practices. However, attempts to track workshop and field day attendees proved to be challenging as farmer contact information was limited and information collected from the few who were contacted was extremely difficult to categorize. As such, the decision was made to look at roller/crimper purchases as a better metric of whether famers adopted the technology. Because this methodology may have missed capturing some producers who may have chosen to build their own roller or borrowed one from another producer, this metric errs on the side of under-estimating rather than over estimating the success of the project work.

Over the life span of the grant-funded work, I&J Manufacturing sold twenty one roller/crimpers to farmers and two to researchers at The Pennsylvania State University (I&J Manufacturing is the main builder and seller of this tool in the region). This far surpasses the milestone of having ten producers adopt the technology onto their farms and is a testament to the adaptability of the system and the value of the project. Other roller/crimper builders may exist, and some farmers may have built their own roller/crimpers, due to the fact that plans for building the device are available for free on Rodale Institute’s New Farm web site. Several of the farmers who purchased and are using roller/crimpers reported loaning their equipment to neighbors to try the system on their farms. Purchased roller/crimpers range in size from 5 feet to 20 feet in length, indicating that they are being use on a diversity of crops and on farms of varied sizes, from large-acreage grain to smaller vegetable operations.

The names and contact information for each of farmers who purchased the roller/crimpers are known but not supplied in the public version of this report to protect their privacy.


Materials and methods:

Field trials were established in 2006 to study rolling/crimping of hairy vetch cover crops, rye cover crops, and other cover crop mixtures, to assess their value in suppressing weeds and supporting viable corn and soybean crop yields in the Mid-Atlantic region. Trials were established at Rodale Institute’s research farm, Kutztown, PA, “Rodale”, 40°33/ N, 75°43/ W) and at The Pennsylvania State University’s Russell E. Larson Agricultural Research Center in Centre County, PA (“Rock Springs”, 40°44/ N, 77°57/ W), and at collaborating farms in New York, Pennsylvania, and Maryland. Another location at The Pennsylvania State University’s Southeastern Research and Education Center in Landisville, Lancaster County, PA (“Landisville”, 40°05/ N, 76°24/ W) was added in 2007. The inclusion of these sites allowed us to study the potential of this management system under a range of climatic conditions ranging from relatively mild winters (Landisville) to colder winter temperatures at Rock Springs (Rodale Institute was intermediate) because overwinter survival of hairy vetch is sometimes poorer and phenological development in the spring can be slower in colder winter conditions.

At Rock Springs and Landisville, the experimental design was a modified split plot with corn planting date as the main plot, cover crop vs. no cover crop as a sub-plot, and herbicide vs. no herbicide as a sub-subplot treatment. “Planting date” refers to corn planting date, which was the same day as cover crop rolling/crimping date. Soils at these locations are Hagerstown silt loams (well-drained, fine, mixed, semi-active, mesic Typic Hapludalf). At Rodale, the study was managed using certified organic practices and only corn planting date was examined in a randomized complete block design in which all plots were planted to cover crops in the fall. The soil type at Rodale is a Berks-Weikert silt loam, (well-drained, loamy-skeletal, active, mesic Typic Dystrudept).

At Rodale, cover crops were drilled into tilled soil in late August of 2006 and 2007: Hairy vetch was planted at 20.8 kg ha–1 with 54 kg ha–1 oat (Avena sativa L.) as a nurse crop, and rye was planted at 3 bu/ac. At Rock Springs and Landisville, hairy vetch was no-till seeded into small grain stubble across the entire study area in late August or early September at 22.5 kg ha–1 with 54 kg ha–1 oat as nurse crop, and rye was planted at 3 bu/ac. To establish no-cover crop plots at Rock Springs and Landisville, 0.56 kg ha–1 2,4-D low volatile ester ((2,4- dichlorophenoxy)acetic acid) was applied after hairy vetch emergence in mid- to late-November to control the cover crop. At Rodale, the hairy vetch seed used in 2006 was produced in Nebraska (variety not stated), while in 2007, the seed was of the variety ‘Auburn Early Cover’ (Teasdale et al., 2004). At the other two locations only ‘Auburn Early Cover’ was used. The ‘Auburn Early Cover’ seed had been reproduced for many years by a Pennsylvania farmer, which probably caused a gradual adaptation of the vetch to Pennsylvania conditions. Aroostook rye was the variety used for the soybean trials unless otherwise noted.

The cover crops were rolled and crimped at four different dates with the roller/crimper in late spring and early summer of 2007 and 2008. The roller/crimper used in this study (I & J Manufacturing, Gap, PA) was 3.04 m wide and similar in design to the roller tested by Kornecki et al. (2006). The roller/crimper was constructed using a 41 cm diameter steel cylinder with blunt metal blades welded to the outside in a chevron pattern (Ashford and Reeves, 2003). The roller/crimper was filled with water at each rolling-crimping and weighed about 900 kg. The implement was front-mounted to the tractor and driven at 7.2 km /h. At Rock Springs in 2008, a slightly heavier (1520 kg) roller/crimper of similar design was used (Mirsky, 2008).

In the vetch trials, because the roller/crimper was followed in the same pass by corn planting, the combination of the roller/crimper and no-till corn planter on hairy vetch control was assessed. The depth gauge wheels on the corn planter provided additional control of the vetch by crushing it. In the same pass, a 95 d corn hybrid ‘Blue River 40M21’ was no-till planted in a 76 cm row spacing into the hairy vetch residue. At both Rock Springs and Rodale a Monosem vacuum planter (Monosem Inc., Edwards, KS) was used, and at Landisville a White no-till planter (Landoll Corporation, Maryville, KS) was used. The corn seeding rate was 79,000 seeds/ha at Rock Springs and Landisville and 84,000 seeds/ha at Rodale.

At Rock Springs and Landisville, the plots without hairy vetch received N fertilizer applied at 112 kg N/ha before corn planting. This intermediate rate of fertilizer (90 to 100 kg /ha N to a corn crop) was selected to approximate the N contribution expected from the hairy vetch cover crop (Ebelhar et al., 1984). No N was applied to corn grown in plots with a hairy vetch cover crop. At Rock Springs, tefluthrin (2,3,5,6-tetrafluoron 4-methylphenyl) methyl-(1?, 3 ?)-(Z))-(±)- 3-(2-chloro-3,3,3-trifluoro-1-propenyl)-trifluoro-1-propenyl)- 2–2-dimethylcyclopropone carboxylate) was applied at 3.7 kg a.i./ha with the seed to control soil insect pests. Cyfluthrin (0-[2–1,1-Dimethyl)-5-pyrimidinyl]-0-ethyl 0-(1-methylethyl) phoshorothioate or cyano(4-fl uoro-3-phenoxyphenyl)-methyl 3-(2,2-dichloroethenyl0–2,2-dimethylcyclopropanecarboxylate) at 3.4 kg a.i./ha was used at Landisville.

At Rock Springs and Landisville, weed-free sub-subplots with and without hairy vetch received a foliar plus residual herbicide application (2.26 kg a.i./ha s-metolachlor (2-chloro-N-(2-ethyl-6- methylphenyl)-N-[(1S)-2-methoxy-1-methylethyl]acetamide), 0.84 kg a.i./ha atrazine (2-cholor-4-ethylamine-6-isopropylamino-1,3,5-triazine), 0.23 kg a.i./ha mesotrione (2-[4-(methylsulfonyl)- 2-nitrobenzoyl]-1,3-cyclohexanedione), and 0.86 kg a.e. ha–1 glyphosate (N-(phosphonomethyl)glycine)) shortly after corn planting to provide complete hairy vetch and weed control. The other sub-subplot treatments did not receive any herbicide. The intent of the herbicide treatment was to provide complete control of both the hairy vetch and weeds to compare the effects of the rolled/crimped cover crop and weeds on corn yield. The size of the main plots at Rock Springs in 2007 was 12.2 by 24.3 m, and the subplots were 3 by 12.2 m. Individual sub-subplot size that
received herbicide was 3 by 7.6 m, but the weed-free sub-subplot was 3 by 4.6 m. In 2007, main plots at Landisville and Rock Springs were 12.2 by 30 m, subplots 3 by 15.2 m, and sub-subplots 3 by 7.6 m. At Rodale, plots were 7.6 by 15 m in both years.

Hairy vetch growth stage development was determined immediately before rolling/crimping and planting by two methods: (i) visual assessment method based on 0 to 100% flowering where 0 = no flowers present and 100 = complete flowering on the entire stem length, and (ii) count method, where ten stems per plot were removed and the number of nodes, buds, flowers (any purple color on the raceme) and pods were determined. Percent flowering was calculated by taking the ratio of flowers and pods to the total number of nodes. The visual assessment method tended to overestimate vetch flowering stage and did not capture pod formation. The count method used in 2007 more accurately captured flower stage but still did not capture pod set because it considered pods to be flowers. In addition, this method was cumbersome to use.

In 2008, the second method was therefore modified to include pod-set stages and to facilitate faster data collection. The first five nodes below the apical meristem were counted on two stems per plot. Nodes had to be at least 5 cm apart or the next sequential node was counted. Each node was recorded as a bud, flower, or pod to classify the growing stage as follows:
• Vegetative stage—all five nodes were buds
• 20% flowering—one node flowering out of five
• 40% flowering—two nodes flowering out of five
• 60% flowering—three nodes flowering out of five
• 80% flowering—four nodes flowering out of five
• 100% flowering—five nodes flowering out of five
• Early pod set—one or two nodes out of five contained pods with immature seeds (flattened and not hard)
• Mid pod set—three nodes out of five contained pods
• Late pod set—four or five nodes out of five contained pods and/or seeds were maturing in pods

Data Collection
For all the trials, cover crop biomass was collected in two 0.5 m2 quadrats per plot the same day the cover crop was rolled and crimped. Weeds were separated from the cover crops, and were not included in biomass measurements. However, few weeds were present at corn planting time. Samples were dried at 55°C for 72 h and weighed.

Four weeks after planting (WAP), cover crop control in the entire plot was assessed visually for each plot on a 0 to 100% scale, where 0 = no control and 100 = complete control. The visual assessment was done by averaging across the entire plot area. In addition, to estimate regrowth after rolling, living cover crop was collected 4 WAP from two 0.5 m2 quadrats in each cover crop plot that did not receive the herbicide treatment; these samples were dried and weighed. Some cover crop flowering and re-growth data were not collected at Rodale due to inclement weather and labor constraints.

Corn stands were determined 4 WAP by counting the number of emerged plants in a center row of each plot for a 5.28 m length. Weed density was determined by counting all emerged weeds in three 0.5 m2 quadrats per plot at 4, 8, and 12 WAP. Weed density was not measured at Rodale due to labor shortage. Weed biomass was harvested at all three locations 12 WAP from three 0.5 m2 quadrats, dried at 55°C for 72 h and weighed. Weed species were grouped as annual broadleaves, annual grasses, and perennials. Grain yield was harvested either mechanically (small plot combine) or by hand from the center two rows of each plot. In 2007, because of the smaller size of the weed-free sub-subplots at Rock Springs, grain yield was determined by hand-harvesting all four rows. All crops were harvested using a small plot combine equipped with a scale and moisture recorder. All yields were standardized to 155 g/kg moisture.

Statistical Analysis
Effects of crop planting date (which was usually the same day as rolling/crimping), cover crop presence or absence and herbicide use or absence were tested as well as their interactions using PROC MIXED and Tukey-Kramer mean separations at a P value of 0.05 in SAS v. 9.1 (SAS Institute, 2008). Visual inspection of frequency distributions showed that normality assumptions were not being violated. A test for homogeneity of variance was performed using the residual error terms to determine if data could be pooled across locations and years. This analysis showed significant differences between locations at a P value of 0.0002 and between years at <0.0001, and therefore all data were analyzed separately for each location and year. Because weed density counts were performed in fixed subplots at 4, 8, and 12 WAP, typical of a repeated measures design, time was included in the model as a factor.

Research results and discussion:

Milestone 1 - Project staff will identify seven farms for potential collaboration. Selected farmers will be interviewed about their crop management practices, interest in research and willingness to share information and experiences with other agriculture professionals. Three farmers will be confirmed as collaborators for on-farm research, demonstrations and field days on acres in grain production in Pennsylvania, Maryland, and New York.

After an initial interview process, farmer collaborators were identified in 2006. Rodale Institute, the Penn State On-Farm Research Coordinator, graduate student, and faculty members visited Kirby Reichert’s (PA) farm in May 2007, Bill Mason (MD) late May, September and mid-October, and Peter Shuster (NY) early June and at the end of August. These three farmers have been our collaborators for this project. In the summer of 2007, we identified a fourth participant, Klaus Martens (NY), who worked on establishing field planted cover crops with the expectation of rolling and planting cash crops in the spring of 2008.

In 2008, Ruth (Mick) Mischler completed her graduate studies at PSU to satisfy her requirements for a Masters Degree.

Milestone 2 - Researchers will initiate projects at the Rodale Institute Experimental Farm, Penn State’s research farm, and the three collaborator farms. Project participants will implement detailed monitoring plans to collect relevant agronomic data.

As noted in the Materials and Methods section, trials were installed in three Pennsylvania research locations for two years to investigate no-till organic corn production with hairy vetch as cover crop. The vetch was rolled/crimped at different times of flowering (early flowering-pod set). Data on crop growth and yield, weed populations, and soil parameters were followed through corn harvest each year.

Trials were also installed in two Pennsylvania locations for two years to study the possibility of using rolled/crimped rye cover crops to provide weed control in no-till soybeans, thus forestalling the need for post-emergence herbicide application. Data on crop growth and yield, weed populations, and soil parameters were followed through soybean harvest each year.

On-Station Research Results
Hairy Vetch
In the trials in which hairy vetch was rolled/crimped for no-till organic corn production at Rock Springs, Landisville, and Rodale, the vetch was established at recommended dates in late summer (end of Aug) and was rolled/crimped at four different dates in the spring, from early flowering to pod set, to determine how well the vetch could be killed with the crimper/roller.

Although vetch-kill was sometimes possible to achieve earlier, consistent kill was obtained only after pods were detected on the vetch. This stage was reached in early-mid June. Corn was planted immediately into the rolled/crimped vetch cover, either in the same pass, or on the same day. Vetch cover helped control weeds without any additional weed control such as herbicide application or in-season tillage; though annual weeds were better controlled than perennial weeds. Average weed density was reduced at least 50% by the vetch, and greater vetch biomass accumulation at later crimping/rolling dates improved weed control. No nitrogen fertilizer was applied in the organic no-till corn treatments.

The generally recommended corn planting date in the trial region is on or before May 15, suggesting that corn yield potential of this system was likely reduced when compared with timely-planted corn. Corn yields in the organic no-till system varied from 1.1 Mg/ha to 9.6 Mg/ha. Major reasons for reduced yields included: 1) hairy vetch re-growth competed with corn; 2) weed control was not sufficient (primarily due to poor vetch stand); 3) corn plant population was reduced due to poor planter performance in heavy vetch; 4) nitrogen needs of corn may not have been met; and 5) in certain locations black cutworms decimated corn stand at individual planting dates. Because there were many reasons for variable yields, it was not possible to formulate solid recommendation for this system at this time.

One clear requirement for success was high vetch biomass accumulation prior to rolling/crimping. In some cases vetch stands were poor – for example, in Landisville in 2008, a trial was abandoned due to a poor vetch stand. A light hairy vetch stand does not provide a good mulch cover for weed control and does not provide sufficient nitrogen for the corn to produce high yields. For rolled vetch systems to succeed, the planter needs to be in excellent working condition and properly adjusted to function well in these high residue situations, and guidelines to improve no-till planter performance are urgently needed. Further development of earlier maturing winter hardy vetch varieties will also greatly benefit overall system performance. More research needs to be done to control perennial weeds that may proliferate when producing corn organically with the rolled cover crop system over moldboard plow tillage systems to incorporate the vetch, or killing it with an herbicide. Currently, tillage at some other stage in the organic rolled cover crop rotation can be effective in controlling the perennial weeds, but efforts to find other means of management, such as well-timed high residue surface cultivation or changes in soil nutrient balance, need further exploration.

In the trials that were laid out at Rock Springs and Landisville for two years to study the ability of rolled rye cover crops to eliminate post-emergence herbicide application in no-till soybeans, rye cover crop was established in the fall (Sept/Oct), and terminated early (late April) and late (Mid-May) in the spring with glyphosate. Soybeans were planted into the rye after it had been rolled with the crimper/roller. The crimper/roller was only used to place the crop residue flat on the soil surface in one direction (not to kill the rye). The cover crop was rolled perpendicular to the direction it was drilled in to provide maximum soil cover from the rye. The soybeans were then planted into this mat in the direction of rolling to improve seed placement. Soybeans were also planted at the same dates without the cover crop at these locations.

Rye biomass reached almost 9000 kg/ha at late termination. These high biomass quantities were measured when the rye was planted relatively early (mid-Sept) and terminated late (mid-May), and the rye cover crop reduced weed density and biomass, irrespective of whether it was killed early or late in the spring. In the first year of the trial, crop yields did not vary between treatments in either location. In the second year, rolled rye cover produced the same crop yields as post-emergence herbicide application in Rock Springs, but in Landisville, the highest yields were obtained with post emergence herbicide application. In one location, the soybean stand was poor in the late-killed rye due to the inability of the planter to place the seed through heavy residue, and/or the seedlings inability to break through the residue. Research shows that rye can perform post-emergence weed control and suggests that late-April termination can be as effective as mid-May termination, when herbicides are utilized. For better crop yields, planting success through thick crop residue cover needs to be improved, which means planters and drills need to be designed specifically to plant into these conditions.

On-Farm Research

The experiences and results obtained at the three farms selected in New York, Pennsylvania, and Maryland were mixed. The team struggled with two different grain farmers in the Finger Lakes region of NY, and data of limited utility were collected there. However, with the very committed help of farmer Bill Mason of Queen Anne County, MD, the team conducted two very successful trials during each of two years. Likewise, the interaction with farmer Kirby Reichert in Lebanon County, PA was also very productive and some valuable data were generated on his farm.

New York trials: Though the NY famers had an adequate understanding of the structure and the rigor of on-farm research (OFR), work on their farms did not yield data that increased knowledge of how to better use cover crops and the roller/crimper in those locations. However, the work did reinforce the importance and value the research team must place on developing and communicating a clear OFR protocol and adhering to that plan.

A well-conceptualized four-treatment project was proposed in late summer 2006, a field was selected, and a rye cover crop was planted in late September. The plan was to establish soybeans in this field via various field operations the following June. The host farmer also expressed a strong interest in all four treatments and offered to use his own subsoiler (vertical tillage tool) and no-till corn planter to partially or completely initiate three of those treatments at the time of cover crop rolling and soybean planting. However, in reality, this farmer wasn’t equipped to conduct those operations. The OFR team learned on the day of rolling/planting that the no-till planter was incapable of no-till planting due to the fact that it had no down-pressure springs or linkages to transfer planter frame weight onto the planter units, a necessity when cutting through the heavy mulch that is created with the roller.

An attempt was made to salvage the effort with a simple two-treatment test in which the entire field was no-till drilled to soybean to compare herbicide treated with non-herbicide plots for their effects on weed competition and soybean yield. This also ended in failure when the farmer did not follow protocol and omitted the herbicide application.

A second NY farmer, an organic grain farmer and extensive user of cover crops, was identified during 2007 and agreed to take a look at the use of the roller/crimper on several fields. When the OFR coordinator transported a roller and a no-till drill to the farm during early June 2008, it was clear that a dry spring had limited the biomass production of cereal grain (barley) and legume (Austrian winter pea (AWP)) cover crops. Some limited rolling and drilling of soybean was conducted. Although the grain growing season was not excessively dry, the rolled cover crops provided marginal weed control due to limited biomass. The inadequate weed control so early in the season, coupled with mediocre soybean growing conditions, prevented development of the dense soybean canopy necessary to provide adequate competition with emerging weeds.

Maryland trials: The cooperative effort with Bill. Mason of Mason’s Heritage Farm was outstanding. This farmer’s strong interest in reducing tillage, field operations, time spent in the field, and fuel required to produce a crop enabled the OFR team to complete two projects in each of two years. In the fall of 2006 and 2007, two sites were selected for planting fall cover crops. Each year, as soybeans were maturing, or soon after soybean harvest, legume/cereal grain mixtures were planted in preparation for a following corn crop. Likewise, after corn had been harvested, the corn stover was lightly disked to enable improved drilling of the cereal grain cover crops, which soon followed. Soybean was planted following spring. Hand sampling of above-ground biomass was conducted just prior to rolling of the cover crops, and then the plots were combine-harvested and weighed on pad scales or in weigh carts when grain had dried to harvest-appropriate moisture.

Cover crops containing legumes ahead of corn

Aerial seeding of cover crops into standing grains generated extensive interest in 2006. The first experiment that followed legume cover crops with corn included two treatments that were intended to simulate “flying on” cover crop seed into standing soybeans. Crimson clover alone and crimson clover with wheat was spun onto the soybeans with a shoulder-mounted spinner spreader on September 11, 2006. Several weeks later, on October 22, after soybean harvest, two additional treatments were drilled at the same rates into adjacent plots in a replicated complete block with four replicates. The seeding rate of crimson clover was always 25 lbs per acre and wheat was seeded at 60 lbs per acre. The following spring, the entire plot area and the field outside the plots received an application of poultry manure at about 2.5 tons per acre. The area outside the research field was moldboard plowed several weeks before the cover crop was sampled and rolled.

The stands of the aerially seeded treatments were observed to be slightly less uniform than were those that were drilled. Above-ground cover crop biomass was sampled during late May, and the biomass treatment means were not significantly different, ranging from 5042 to 5330 lbs per acre. The cover crops were rolled and corn was planted about 1 week after the surrounding field had been planted into tilled soil. While seed was dropped at approximately 30,000 per acre, the resultant stands were variable. The mat of rolled biomass quickly proved to be insufficient to provide satisfactory weed control for the corn crop. Some of the stand variability was likely due to weed competition, but some corn plant loss was also suspected to be due to cutworm damage. Where weed control was poor, the incidence of small ears or barren plants was high. Grain yields were disappointing, with treatment means ranging from 34 to 51 bu/acre (not significantly different). In contrast, corn plant counts in the surrounding tilled and cultivated field approached 28,000 per acre and weed control with row cultivation was very good, resulting in grain yields that approached 150 bu per acre.

For political reasons, aerial seeding was not of great interest in the fall of 2007. The State of Maryland had implemented a cover crop cost-sharing program that included barley either alone or in mixtures with other species. As such, the 2007 planting included drilled plots of barley (40 lbs/acre) with either crimson clover (15 lbs/acre) or with AWP (80 lbs/acre). A second set of crimson clover plus barley plots, to be moldboard plowed in the spring, were located within the two rolled cover crop treatments in the second year’s experiment.

Spring biomass of the crimson plus barley treatment was 5848 lbs/ac, while the AWP plus barley produced 4669 lbs/ac of biomass (no significant difference). Similar to observations the year before, considerable weed pressure was noted in both rolled cover crop treatments (averaging 11,000 lbs/ac fresh weed weight in mid-October just prior to corn harvest), especially when compared to the crimson clover plus barley treatment that incorporated the cover crops with the moldboard plow, followed by in-row cultivation for weed control (5000 lbs/ac fresh weed weight). The grain yields for the plowed treatment (138.5 bu/ac) were far superior to the rolled crimson clover plus barley (22.5 bu/ac) or the rolled AWP plus barley treatment (61.5 bu/ac). These data clearly indicate that the amount of biomass necessary to provide adequate weed control in a corn crop planted into 30-inch wide rows must be greater than 6000 lb/ac at the time of rolling and should include some vegetation that does not degrade easily once it has been rolled. Striking the right balance of legume and grasses in a cover crop mix is challenging but vital to maximize the valuable N-fixation of the legume component with the physical recalcitrance of the grasses, generating a C:N ratio that is highly favorable to the soil microbes responsible for that organic matter decomposition.

Cereal cover crops ahead of soybean

The MD cereal cover crop trials were simpler than the legume trials. The first trial was planted during October 2006, consisting of four replications of two varieties of rye (Aroostock and Wheeler) and one triticale (Trical 815), seeded at a high rate of 3 bu/ac. Spring biomass yields were not significantly different; the Trical 815, Aroostock rye, and Wheeler rye produced 6443, 5511, and 5034 lbs/ac respectively. Weed control was good in all three treatments with a few summer annuals growing as tall as or taller than the soybeans by late summer. Soybean yields were statistically different between the Trical 815 (55.5 bu/ac) and Aroostock rye (51.1 bu/ac). while the Wheeler rye generated intermediate soybean yields of 53.1 bu/ac. These yields were very respectable and are considered acceptable by farmers in the region.

The second cereal trial produced less favorable results. In fall 2007, Aroostock and Wheeler rye were compared with Alzo triticale. Planting rates fell between 2.1 and 2.4 bu/ac for the three treatments. Site soils clearly did not carry over much nitrogen from the previous corn crop; fall tillering of the cereals was poor and plant color was paler than the typical medium to dark green associated with adequate N supplies. While May 2008 plant height of Aroostock approached five feet, the stand was somewhat sparse and the leaves were small. The crop rolled down well and, due to the presence of corn stover that remained on the field surface, weed pressure was not as great as anticipated. Limited soil N appeared to affect weed growth, reducing weed height and biomass similar to the lack of growth noted in the cereal cover crops. Soybean yields were statistically similar between treatments, ranging from 44.8 to 46.6 bu/ac.

Pennsylvania trials: OFR trials in Pennsylvania began well, followed protocol, and generated some viable grain crop data.

Three years of no-till research at Kirby Reichert’s farm produced several failures before finally achieving successful corn yields in 2008. Problems arose from equipment in one year and poor cover crop stands in another.

The 2006 trial was plagued with planter issues. Mr. Reichert had two White no-till planters. The older planter was initially used to roll vetch and plant corn for a short trial pass, and it worked well. However, one of the sprockets then broke on the older planter so the rest of the field was planted with the newer model, which was lighter in weight and sported a different closer wheel design. As a result the seed was not as effectively placed into the soil and covered; the resulting corn stands were not satisfactory for the system. Support was provided to modify the newer no-till planter for the following year, however during a field day demonstration of the planter, the tractor axle broke, requiring a welding repair preventing an adequate demonstration.

In the fall of 2006, randomized replicated plots were established at two Reichert farm site locations, and cover crop biomass data was collected in spring of 2007. In general, dry conditions in the spring of 2007 led to poor cover crop stands. For example, oats mixed with vetch from Oregon, Minnesota, or Nebraska yielded 3734, 4819, or 5073 lbs/ac dry matter respectively, while oats mixed with crimson clover yielded 6065 lbs/ac dry matter. Similarly, Aroostoook, Kreider, and bin (VNS) ryes yielded 4604, 3484, and 2894 lbs/acre dry matter, while the triticale yielded 3680 lbs/ac.

Poor stands of hairy vetch resulted in incomplete soil cover, weak weed suppression, and low corn yields. The crimson clover performed better than the hairy vetch in that very dry year, but overall, only a few of the legume plots produced adequate biomass to successfully roll for corn planting. As such, production constraints demanded that the few good plots be sacrificed and the field was then plowed for clean tillage production. Identical issues with cereal grain cover crops to be rolled for soybean production yielded the same result: sacrificed research plots for clean tillage production. As such, the research design was modified in the following year to establish strip trials in lieu of replicated random plots.

The final trial conducted at Mr. Reichert’s farm consisted of four replications of two legume cover crop treatments. The legume cover crops, established into tilled soil with a conventional grain drill in late August 2008, included hairy vetch (HV) drilled at approximately 25 lbs/ac, and a mixture of AWP (40 lbs/ac) and crimson clover (10 lbs/ac). After cover crop biomass was sampled, all plots were rolled in early June and corn was planted on June 6, 2009. Biomass yields from the two treatments were not statistically different: The HV produced 4616 lbs/ac and the mixture of AWP plus crimson clover yielded 5359 lbs/ac. Although weed biomass data were not collected, weed pressure was high and appeared to negatively affect corn plant growth and production. Surprisingly, corn grain yields were better than anticipated with the HV alone and the legume mixture producing 76.4 and 75.1 bu/ac, respectively. The slightly cool and damp growing season may have contributed to the respectable yields.

Milestone 3 - Results from the on-farm research will be shared among project collaborators and prepared for dissemination to farmers and agriculture professionals.

Throughout this project, farmers were inspired by the simplicity of the rolled/crimped no-till cover crop system and its ability to manage weeds without the use of primary tillage. To this end, our outreach efforts have expanded far beyond those first envisioned, as seen in the extensive list of presentations on the subject of the project.

Over the course of this grant, over 80 presentations were made directly to farmers, agricultural professionals, students, and policy agents as noted below; reaching thousands. For example, a cover crop summit was held in June 2007 at Rock Springs, PA, attended by about100 individuals including 25 county extension educators, 15 organic and 18 conventional producers. Knowledge gain was considerable, and 34 individuals indicated they were very or moderately likely to make a change in their farm operations due to the event.

Several field day activities took place: Two on-farm field day events took place in Maryland at the farm of collaborator Bill Mason, and two field day events were hosted at Rodale Institute.

In 2008, using funds from an NRCS conservation innovation grant, Jeff Moyer wrote a book on use of roller/crimper technology to support cover crop-based no-till systems, as covered in this report. The book is expected to be published in 2010 and will contain information generated from work covered by this grant, as we seek to outreach the information to as many farmers as possible.

Information obtained from these and previous years’ trials concerning cover-cropping practices, weed management and organic no-till was outreached by Jeff Moyer, Paul Hepperly, Dave Wilson and/or Tim LaSalle at many events, included in the attached events table.

Milestone 4 - Two field days will be attended by 160 farmers who will learn the new concepts and techniques through lectures, workshops, and farm demonstrations. At least 120 farmers will respond to baseline surveys to gauge their interest in the new techniques. At least 80 farmers will respond to evaluation follow-up surveys to assess rates of adoption of the techniques.

Each of our field day events brought together farmers, researchers and policy folks in an educational environment to transfer the knowledge gained to date on the specific activities funded by this grant. Following each event, participants were surveyed to determine what impacts the event and the technology might have on changes to their farm operations. A sample of a survey and the responses is included as an appendix.

On July 18th 2008, Rodale Institute hosted an on-site field day event specifically focused on cover crops and the roller technology as a management tool to eliminate the need for tillage as a weed management tool during specific cropping years. The day was attended by 120 farmers, educators, and policy folks. In assessing the impact of the field day in advancing knowledge, it is impressive to find that a sizeable 70% (44 respondents) indicate that they learned a new sustainable practice at the field day. What is even more impressive is that 68% (43 respondents) indicate that they would definitely make a change in their practice in the next 2 years as a results of attending the field day. Finally, 100% of respondents made suggestions for future topics and/or activities for Rodale Institute educational events. Combined, these findings clearly show interest, potential for measurable impact in the adoption of the roller/crimper technology and other new techniques, and the need for more information and training in the future. The complete results from the field day survey are included in Attachment #10.

Project collaborators were also fortunate to be able to host two extremely successful field day events on the farm of Mr. Bill Mason on the Eastern Shore of Maryland. The first event was held on May 22nd , 2008, when the roller/crimper was demonstrated and soybeans were planted. Then the second event took place on October 23rd , 2008 when farmers (many who attended the May field day event) were invited back to see the end result of the growing season as the soybeans were harvested. This two-part field day format was well received by the farmers, as it allowed them to be able to experience the entire crop production process, and see the system performance results at the end of the season.

Penn State also held field day events at the Landisville research facility in Lancaster County. This event also brought in farmers from across the region, exposing them to this technology.

Milestone 5 - At least 25 farmers will enroll in a collaborative program to try one of several rollers on their farms during the project period. Project staff will provide technical assistance and monitor progress and adoption of the new techniques.

Attempts to track workshop and field day attendees proved to be challenging as farmer contact information was self-limited and information collected from the few attendees who permitted follow up contact was extremely difficult to categorize. As such, the decision was made to look at roller/crimper purchases as a better metric of whether famers adopted the technology. See the information below in Milestone 6.

Milestone 6 - Eight users will fully incorporate the techniques into their cropping plans on 1,500 acres.

Tracking production-based equipment purchases, rather than farmer technology adoption interview methodology, may have missed capturing some producers who may have chosen to build their own roller or borrowed one from another producer. However, it is likely that this metric errs on the side of under-estimating rather than over estimating the success of the project work.

Over the life span of this grant funded work, twenty one roller/crimpers were sold to farmers and two were sold to researchers at Pennsylvania State University by I&J Manufacturing who is the main builder and seller of this tool in the Northeast region. This outcome has far exceeded the milestone of encouraging eight producers to adopt the technology onto their farms, and it is a testament to the adaptability of the system and the usefulness of the allocated funds. Other roller/crimper builders may exist that are unknown to project collaborators, and farmers may also have built their own since the plans for building the roller/crimper are available for free on the New Farm web site. Several of the farmers who purchased roller/crimpers and are using the technology also loaned their equipment to neighbors to try the system on their farms. Of those sold to farmers, the size range is from 5 feet to 20 feet in length indicating that the roller/crimper is being applied to a diversity of crops and at scales of use from large acreage of grain to vegetable size operations.

The names and contact information for each of producers who purchased the equipment are maintained at the Rodale Institute and are not supplied with the public version of this report to protect their privacy.

Milestone 7 - Research results with case studies will be published in The Rodale Institute’s online magazine New Farm and on Penn State websites. Over 100,000 visitors (world wide) will learn about the project and the new techniques. Over 150 farmers and 50 Extension specialists from around the country will request more information from project staff.

Tracking performed by Webtrends and Google Analytics indicate that the goal of 100,000 world wide visitors to the New Farm website has been far surpassed. In fact, these tracking software systems have counted 308,590 visitors to the New Farm, web site where four new stories appeared on the no-till projects over the life of the grant. Through the grant-support period, over 67,000 of those visitors specifically went into the No-Till Forum section of the website for more detailed information on the subject. A newly-launched frequently asked questions (FAQ) section also allowed farmers direct access to other information needed to move their own field projects forward. Activity also continues on the downloadable drawings of the roller/crimper, made available on Rodale Institute’s website, by more farmers wishing to build their own rollers.

Participation Summary


Educational approach:

Mischler, R., S.W. Duiker, W.S. Curran, and D. Wilson. 2010. Hairy vetch management for no-till organic corn production. Agronomy Journal 102:355-362.

Mischler, R., W.S. Curran, S.W. Duiker and J.A. Hyde. (in print). Use of rolled rye cover crop for weed suppression in no-till soybeans. Available on-line in Weed Technology.

Mischler, R. 2009. Roller/crimper technology to manage cover crops for weed suppression. A thesis in Agronomy (M.S.). The Pennsylvania State University, University Park, PA.

Kinsley, C. 2008. Mason unveils organic, no-till test plots. Article published in The Delmarva Farmer on field day at Mason’s farm in Maryland. October 28, 2008.

Mick, R. Roller Study News 1, January 2007. Penn State Cooperative Extension.

Mick, R. Roller Study News 2, January 2008. Penn State Cooperative Extension.

No-Till Revolution Webpage, published at

Sullivan, D. 2008. Organic no-till leads to updating of Farming Systems Trial. Web article published at May 29, 2008.

Wilson, D, C.Ziegler-Ulsh. 2007. Down and Dirty details of our 2006 organic no-till success. Web article published at March 15, 2007.

NE-SARE Roller Research Newsletter Debuts, published at

Cover Crop Management Summit. Held June 5, 2007 at Russell E. Larson Agricultural Research Center, Rock Springs, PA. Attended by ca 100 persons. Flier and evaluation is attached.

Winter meetings and field days focused around the OFR projects

The interest and enthusiasm for organic no-till production is great in the Northeast and on the eastern shore of Maryland. Cooperative extension has been intensively organizing and delivering winter meeting programs and field days that focus on organic grain production for over four years. The timing and location of the study at Mason’s Heritage Farm couldn’t have been better timed. With Maryland Cooperative Extension, Penn State Extension and Rodale Institute participated in two field days and three winter meetings during the conduct of this project. The field days, conducted during May and October 2008 were attended by more than 200, most of whom were farmers from at least four nearby states.

No milestones

Additional Project Outcomes

Project outcomes:

Impacts of Results/Outcomes

Primary outcomes included collection of data from all sites (research station and farmer collaborator) on many experiments, hosting of a major field day events, posting of three primary Newfarm articles to expand outreach of project information, development of an on-line organic no-till forum, and an overwhelming request for speaking engagements on cover crop management and the roller/crimper technology.

Economic Analysis

Economic analysis was not conducted as part of these organic no-till trials. However, input costs in these systems are very low, because no tillage (either pre- or post planting) or fertilizer (nitrogen) was required. So if reasonable crop yields can be achieved, this system stands to be very profitable. However, an economic analysis of the reduced use of herbicide was conducted (elimination of post-emergence herbicide application). This analysis revealed that, if rye was used as a cover crop for soybean production, the net added cost of rye cover crop was $123/ha if post-emergence herbicide was applied, and $68.50 if post-emergence herbicide was not used in the rye system.

Farmer Adoption

See Milestone 6, above.

Assessment of Project Approach and Areas of Further Study:

Areas needing additional study

Research from this project revealed the need to develop earlier maturing cover crop varieties – for example; if vetch could be bred to reach pod-set stage by mid-May, timely corn planting could be achieved. Research also indicates the need to develop planter and drill modifications or attachments that can more consistently and successfully place seeds at the proper depth in high residue conditions and close the seed slot, even if the soil is moist.

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.