Final Report for ANE91-003
[Note to online version: The report for this project includes tables that could not be included here. The regional SARE office will mail a hard copy of the entire report at your request. Just contact Northeast SARE at (802)-656-0471 or firstname.lastname@example.org.]
Seven commercial farmers participated in winter cover crop demonstrations for corn production systems. Some were already convinced of the value of cover crop while others were skeptical. Farms varied widely in terms of soils, climates, and type of operations. Locations ranged from the relatively flat Coastal Plain to the rolling Piedmont with higher O.M. soils, more severe winters and shorter growing seasons. Data collected from these demonstrations confirmed that research results can be directly applied to the farm. For example, research has shown that soil moisture use efficiency increases when no-till corn follows hairy vetch. The primary reason Wayne Lambertson uses hairy vetch is because it conserves soil moisture for his corn crop. Data from Wayne McFarland's farm suggests that improved soil moisture may be a partial explanation for yield increases although moisture samples were not taken. Cover crops that were plowed down appeared to mineralize more rapidly than killed mulches left on the surface, resulting in more cover crop N available to increase corn yields. However, plowing the cover down and/or allowing it to grow too long appeared to reduce availability of soil moisture for corn growth.
The PSNT soil samples effectively identified fields that needed no additional fertilizer nitrogen (FN). However, it underestimated legume N contributions because cover crop mineralization was incomplete when samples were taken.
Disk meter estimates of cover crop N appeared to be more accurate for crimson clover than for hairy vetch with its more prostrate, viney growth. Also, because the compact hairy vetch canopy excluded more light, low canopy leaf losses were high and consequently not measured by the disk meter. Based on these demonstrations and earlier research, calendar date (adjusted to current soil and climatic conditions) appears to be more appropriate. Ten years of data collection and observations suggest that cover crops in both the Piedmont and Coastal Plain should be killed between April 20 and May 10 since maximum N fixation will be achieved by this date and soil moisture depletion will be low. As cover crop growth approaches maximum and ambient temperatures increase, soil moisture can be depleted, reducing corn production. The interrelationships are different for each cover crop, soil type, and climate. Experience, planning, and skilled management are required to optimize these relationships. The most common complaint from producers was that timely fall cover crop seeding, spring cover crop killing and planting corn at the best time the following spring adds to workloads during two of their busiest periods. When only the cost of a pound of N is considered, cover crop N is not always cheaper than purchased FN. However, when wind or water soil erosion, soil water availability, and improved soil conditions are factored in, cover crops will help sustain a productive agriculture.
Data from these demonstrations have been discussed at field days and workshops. Some of the cooperating farms were included in field days and/or farm tours. As a result, many individuals were introduced, for the first time, to advantages and problems associated with the use of cover crops in corn production systems. Agronomy Mimeo 34 "Winter Annual Cover Crops for Maryland Corn Production Systems" can be used by interested farmers to develop systems best suited to their specific farm situation.
A. Prepare fact sheet on benefits, use and management of winter cover crops in corn production systems.
B. Use current knowledge accumulated from Maryland cover crop research and adapt the pasture disk meter to provide a rapid method that farmers can use to accurately estimate cover crop N production and subsequent N availability to corn production.
C. Develop strategies for acceptance and implementation of the disk meter technique in cover crop/corn production systems to help fine-tune nitrogen fertilization recommendations in concert with currently used nutrient management plans which make use of the PSNT; compare economics of cover crop use to current producer practices.
Seven producer location and two research facilities were used in these demonstrations. Soil and climatic conditions varied widely from sandy soils of lower Coastal Plain, with relatively mild winters and 200 plus summer growing season, to the heavier soils of the Piedmont with significantly higher organic matter levels, more severe winters and a shorter summer growing season.
1. Objective one:
Agronomy Mimeo 34 "Winter annual cover crops for Maryland corn production systems" was completed March, 1992. It addressed cover crop benefits and constraints, cover crop selection in which specific merits of legumes, grasses, legume/grass mixtures and brassica species adapted to Maryland were discussed in detail. Cover crop fall seeding requirements and spring cover crop management options prior to seeding the corn crop were discussed. Mimeo copies were distributed to county Agricultural Agents and Nutrient Management specialists. Copies were also available at field days, farm tours, producer meetings, etc.
2. Objectives two and three:
Eleven commercial producers interested in cooperative research/demonstration cover crop plantings were identified by Agricultural Agents in 5 Maryland Counties. These farmers were visited during the summer of 1991 when cover crop research findings and our desire to identify on-farm benefits and/or problems associated with winter cover crop use were discussed. Each operator then selected fields to be used, cover crops to be seeded, and corn FN rates and/or other management treatments to be applied. Winter cover crops were established on seven of the farms that fall. We worked closely with each farmer, but all field operations were done by him. Arrangements were made for weigh wagons at corn harvest.
Fall cover crop stands were assessed and observed throughout the winter. Cover crop dry matter yields (disk meter readings and hand-clipped samples) were obtained just prior to planting corn to estimate cover crop N production. Pre-sidedress Soil Nitrate Test (PSNT) soil samples were obtained when corn was 12 inches tall.
In addition to the 7 commercial farms, comparisons were made between no-tillage and conventional corn planting following legume, grass and legume/grass cover crop mixtures at three Maryland Agricultural Experiment Station locations. Data for only the Piedmont and Coastal Plain locations will be reported. Soil drainage problems and heavy deer damage at the fall-line location eliminated meaningful corn data. The results from these 9 locations follow.
a. Evaluation of Disk Meter
The pasture disk meter was calibrated during the spring of 1991 for cereal rye, hairy vetch, crimson clover, and mixtures of vetch and crimson with rye. On four dates (4/8, 4/16, 4/26 and 5/3), a series of disk readings was taken for each cover crop species, and hand samples were taken from the same area measured by the disk. After determining the N concentration of each cover crop species, cover crop N content was determined. Data were used to develop regression equations relating the disk reading to actual hand-clipped dry matter (DM) yield. Equations were developed for each sample date and for all sample dates together. The data were best described by a quadratic equation. While equations for individual dates sometimes had slightly higher R2 values than the combined data set, we felt that it was more important to use the combined data set, since the objective of using the disk meter was to be able to determine at any time during the spring how much N was contained in the cover crop topgrowth. Quadratic equations relating disk meter readings (in inches) to cover crop DM yields (in lbs./acre) are presented below:
Rye DM Yield = 622 + 1314 Disk - 166 Disk²__________.95
Vetch DM Yield = 403 + 4920 Disk - 1286 Disk²_______.38
Crimson Cl DM Yield = 1590 + 1121 Disk - 84 Disk²_____.75
Crimson/Rye DM Yield = 1691 + 741 Disk - 30 Disk²_____.61
Vetch/Rye DM Yield = 2057 + 1121 Disk - 71 Disk²_____.65
These equations demonstrate that, within the range of cover crop growth that we measured, the disk meter successfully estimated rye DM yield, and was not quite as accurate at estimating crimson, crimson/rye or vetch/rye. The poorest correlation was with hairy vetch DM estimates. This was apparent even as calibrations were made in 1991. On some measurements, the disk did not compress the plant material to the ground because the vetch vines held it up higher. For other measurements on vetch, we realized that the disk was measuring uneven areas on the ground, rather than vetch topgrowth.
In 1992, the disk meter was used on all farm demonstration fields to estimate dry matter yield and yield variability across large farm fields. At each location and for each species, we again hand-clipped several samples as part of the validation of the disk meter equations developed the previous year (data are included with discussion of each farm, below). The results were disappointing. For hairy vetch, in particular, equations developed in 1991 consistently overpredicted DM yield of fields in 1992. For pure stands of crimson clover and cereal rye, predictions were relatively good. However, DM estimates of legume/grass mixtures were inconsistent.
The disk meter readings seemed to reflect variability across a field quite well, but estimation of actual biomass was not accurate. It may give a very gross estimate of cover crop biomass, but probably less accurate than an estimate based on calendar date, stand height, stand density and producer experiences.
Data for each species at each farm are presented in the sections that follow. Included are the disk estimate of topgrowth N and the hand clipped sample estimates for each field. Also included for each location are PSNT values and corn yields.
b. No-tillage corn plantings on three commercial farms.
(1) Mr. Wayne Lambertson, 1750 Boston Rd., Pocomoke City, Md 21850, annually harvests approximately 1,000 acres of corn, 1300 acres of soybeans, and 300 acres of wheat. Each year he finishes 2,000 head of hogs and operates a 136,000 capacity broiler facility. For the past 5 years he has used cereal rye, wheat, and hairy vetch winter cover crops.
In 1991 he seeded 300 acres of hairy vetch after soybeans. The following spring he no-tilled corn into the living vetch and sprayed burndown and residual herbicides 1 to 2 days later. Four strips in one vetch field were laid out and 4 different FN rates were applied. The following grain yields, PSNT, and estimated cover crop N values were obtained [Table omitted].
Disk reading values suggested that hairy vetch topgrowth contained 229 lbs of N/A but hand samples showed only 178 lbs N. PSNT values suggested that additional corn FN should not be needed. Grain yields increased up to 95 lbs FN/A with no increase beyond that point. This suggests that the synergistic FN/soil moisture relationship after a hairy vetch cover (observed in research plots) may be a partial explanation.
(2) Mr. Wayne McFarland, Rt. #1, P.O. Box 111A, Church Hill, Md 21623 and his son farm 350 acres that have been in the family for over 100 years. They plan to keep it a working family farm. In 1992, they completed a new milking parlor for their 120 cow dairy herd. They are located in the German Branch water project and through the use of nutrient management programs have realized significant savings by limiting commercial fertilizer to only alfalfa and soybeans. Cover crops have been used in their operation for 25 years.
In the fall of 1991, they seeded a large field of hairy vetch after a corn crop was harvested. The field had received 6-7 tons of manure per acre in 1991. In contrast to the Lambertson operation, the vetch was killed about one week before the corn was no-tilled into the vetch surface mulch. Five rates of FN were applied and in 1992 grain yields, PSNT, and estimated cover crop N values were obtained [Table omitted].
After a summer with below average rainfall, yields following hairy vetch with 90 lbs of corn FN were about the same as 125 lbs corn FN after no winter cover. Although soil moisture samples were not obtained, soil moisture conservation appeared to account for some of this increase. After vetch, 40 lbs N was inadequate for maximum yield but there was no increase above 90 lbs FN/A rate.
(3) Mr. Joseph Hottel (Egypt farms), 4524 Burkittsville Rd., Knoxville, MD. 21758 farms 1256 acres in Frederick County. Most soils are Class one and two but erosion losses can be significant in some years. Major crops are corn, wheat, barley, soybeans, alfalfa and timothy. In addition to cash crop sales the farm supports a 100-cow dairy operation. No land is currently irrigated but a 23-acre, spring-fed lake is planned that will irrigate approximately 200 acres. In addition to the Maryland farm, an additional 2500 acres are farmed in Virginia.
Twelve field strips, totalling 95 acres, were allotted to cover crop demonstration seedings. Seven of these strips had sludge applied in the late summer of 1991. Hairy vetch, crimson clover, wheat, and rye/legume cover crop mixtures were seeded in the fall of 1991 and corn was no-tilled into the killed covers the following spring. Three corn FN rates were applied at corn planting. Corn grain yields, PSNT, and cover crop N values are summarized below [Table omitted].
There were accumulative corn grain yield increases from sludge, FN and legume cover crops while winter wheat cover consistently reduced yields. Yields following legumes cover crops with 40 lbs of FN added were about the same as sludge plots regardless of the FN rate. The corn FN was applied at corn planting and about 75% of that FN was detected in the soil PSNT samples. The PSNT values from the sludge plots suggested that no FN or legume-N would be needed for top grain yields. PSNT values were highest on the wheat plots but grain yields were not as good as on the legume plots with much lower PSNT values. The wheat strip with the highest PSNT value was next to the old dairy barn, suggesting that it may have received a lot of dairy manure in the past. This past history of organic N additions was reflected by a PSNT of 218 lbs N/A.
c. Conventional Corn Planting Demonstrations
(1) Mr. Paul Towers (J.C. Ripley Farm), Rt. #1, P.O. Box 177A, Denton, Md 21629 farms 400 acres of relatively level land, droughty sandy loam soil that can be irrigated. Crops grown average 200 acres of wheat/double-cropped with soybeans, 110 acres of sweet corn/double-cropped with lima beans, 50 acres of peas/double cropped with lima beans, and 40 acres of field corn. Cover crops have been used for the past six years. In addition to 200 acres of winter wheat seeded for grain production, cover crops of wheat, crimson clover, or Austrian winter pea are seeded annually to reduce wind and water soil erosion and/or provide N production when legumes are used.
In the fall of 1991, crimson clover was seeded and a no cover strip was left for our cover crop demonstration. In 1992, 3 corn FN rates following crimson and 2 FN rates after no cover were evaluated. Corn grain yields PSNT and cover crop N values obtained were as follows [Table omitted].
The crimson was plowed down April 21 when it was 10-15 inches tall and still vegetative. During the summer, corn received approximately one inch of irrigation water per week. Grain yields following crimson, with 21 lbs of starter N, were equal to 183 lbs N/A following no cover. PSNT values suggested that FN would be needed for top yields following crimson but yields following crimson were the same regardless of FN rates.
(2) Mr. Richard A. Edwards, Rt. #1, P.O. Box 179, Ridgely, Md 21660 farms 800 acres (300 acres of corn, 300 acres of small grain/double crop soybeans, and 50 acres of alfalfa). The remaining acreage is devoted to forages associated with a 180 dairy milking herd. Two center-pivot systems are used to annually irrigate approximately 360 acres. Winter legume cover crops have not been included in past cropping systems.
Mr. Edwards was the only cooperator who selected Austrian winter pea. Good fall stands were obtained followed by satisfactory fall and early winter growth. Peas are highly susceptible to Sclerotinia crown and root rot and pea stands were essentially eliminated by late winter. Thus, there was no opportunity to measure spring growth. Spring rainfall was low and soil was dry when plowed for conventional corn planting. Low rainfall continued into the summer and irrigation was not available for this field. Three FN rates were applied at corn planting. Grain yields, and soil PSNT values are summarized below [Table omitted].
Fertilizer N was applied prior to sampling the soil and the PSNT test detected the extra N. We had no measure of legume N. It was simply too dry for corn yield responses to any N source. Cover crops can deplete soil moisture and when they are plowed down additional soil moisture is lost and there is no benefit from a moisture conserving mulch.
(3) Mr. Lemuel Kinnamon, Rt. #1, P.O. Box 96, Centerville, Md 21617 currently operates a 246 acre grain and beef cattle operation. His soils are medium textured silt loams. Landscape is relatively flat with some imperfectly drained areas. His farm is in the German Branch water project.
The field selected for the demonstration plots had municipal sludge applied in 1991. He chose to seed a crimson clover cover crop. Stands were initially quite variable, due in part to imperfect drainage, so test strips were confined to one end of the field with the more uniform stands. A no-cover crop strip was left. Corn grain yields, PSNT, and cover crop N values are summarized below [Table omitted].
Grain yields were the same regardless of cover crop or FN rates. The PSNT values were relatively low and a response to FN and cover crop N was expected. Late summer N release from the cover crop and/or the 1991 sludge application may be a partial explanation.
(4) Mr. Jeffrey Wivell, 9807 Dry Bridge Rd., Emittsburg, Md 21727 operates a 400 acre dairy farm, with an average of 75 lactating cows. He produces all of the forage and part of the grain fed. This includes 35 acres of timothy, 30 acres each of wheat, barley and oats. The remaining land is in pasture and woodlands. The soils are Penn shaly silt and Cortin silt loams that are imperfectly drained and dry out slowly in the spring.
His demonstration strips consisted of pure barley and barley/hairy vetch winter covers. He did not use a no cover check. The covers were seeded in late September and excellent stands were obtained. The field had very little slope and remained wet throughout the winter. Vetch and, to a lesser degree, barley stands were reduced on the more poorly drained areas and strips were adjusted so that no data were collected from the areas with reduced cover crop stands. In the spring of 1992, barley and barley/vetch silage was harvested 5 May. A strip of the barley/vetch mixture was left as a surface mulch. The field was to be plowed for conventional corn planting immediately after the silage harvests, but rain delayed plowing until mid June. By that time, aftermath vetch growth equalled the original growth that was harvested as silage. Additional rain delayed corn planting until July 9. The late planted corn was frosted before it matured and was harvested as silage. Corn silage yields, PSNT soil samples and cover crop N values are summarized below [Table omitted].
Total forage production was highest where both crops were harvested as silage, and removing the cover as silage did not reduce corn silage yields. This should be expected because of the heavy aftermath vetch growth before corn planting. Results of this demonstration suggest that small grain/vetch covers can be harvested early (cereal in early head to soft dough) and the vetch can then regrow 2-4 weeks and fix 30-90 extra lbs of N for the subsequent corn crop. Removing the corn as silage also makes it easier to get the fall cover crop established on time.
d. Comparisons of conventional and no-till corn production systems on Experiment Station Research Farms.
(1) Coastal Plain-Poplar Hill Res. Facility, Lower Eastern Shore R & E Center.
The soil was a Matapeake silt loam with low soil organic matter and low water holding capacity. The winters are relatively mild with a summer growing season of 200+ days.
Pure stands of cereal rye, hairy vetch, crimson clover, and rye/legume mixtures were established in the fall of 1991. The following spring half of each cover crop strip was plowed under and the other half was killed with herbicides and left as a surface mulch. Corn was planted no-till or conventional two weeks later. One third of each cover crop strip received zero, 80, or 160 lbs of FN.
Grain yields following rye with no corn FN added were lowest and there was no difference between tillage systems since most available soil N had been immobilized by the rye. The response to added N was similar for both tillage systems. Highest yields were obtained following vetch. With no FN added, yields were best when legumes were plowed down, which suggests more rapid mineralization. With 160 lbs FN, corn yields were essentially the same. There was, however, a trend for higher yields when corn was no-tilled into pure legume stands. This synergistic response has been consistently observed in replicated research studies. Yields were lower when rye was seeded with either legume. PSNT values were consistently higher when cover crops were plowed down and highest values were always after hairy vetch. With no corn FN following vetch, grain yields were slightly higher when vetch was plowed down. However, highest yields were usually associated with 160 lbs FN when the vetch was left on the surface to maximize water use efficiency.
(2) Piedmont-Forage Farm Research Facility, Central Md R & E Center
The soil was a Delanco silt loam with higher organic matter than the Coastal Plain. Winters are more severe and summer growing season is shorter than at the Coastal Plain location. Cover crops and tillage practices were the same but only 2 FN rates were used. The field had a history of cover crops followed by no-tillage corn for the past 3 years. Corn grain yields, cover crop N and PSNT values were as follows [Table omitted].
Soil type and recent cropping history resulted in high soil N availability as illustrated by 134 bu/A of grain following cereal rye plowed down with no FN added. PSNT values were relatively low and FN responses were expected. Highest yields were again following vetch plowed down. The advantage of conventional over no-till corn was larger than on the Coastal Plain. The 1992 spring was unusually cool and wet and the heavy soils under the killed cover crop mulches simply warmed up too slowly for good corn germination and early growth. In addition to slow corn germination and growth, slug damage was extensive throughout the no-tilled strips, especially under the hairy vetch mulch.
Agronomy Mimeo #34 "Winter annual cover crops for Maryland Corn Production Systems" was completed March 1992. In addition to being used by County Agricultural Agents and State Nutrient Management Specialists, the Mimeo has been distributed and discussed at cover crop workshops, field days, producer meetings, and with individual producers.
Impacts of Results/Outcomes
A. If these state-wide demonstrations result in wide adoption of cover crop research findings, cash crops can be produced more economically while ensuring a cleaner environment. This will occur because less FN will be required, less erosion and nutrient leaching will occur, and cover crop mulches will increase soil water use efficiency; this can be important during years of limited rainfall.
B. Legumes can supply fixed N to a cropping system but are less effective than grasses in reducing nutrient leaching and erosion. Wider use of legume/grass cover crop mixtures can improve the overall effectiveness of cover crops that can result in a more sustainable agriculture.
Changes in Practice
While the concept of winter cover crops is not new, their use has largely been ignored since World War II. In the last few years, however, there has been a marked increase in cover crop seedings. Precise acreage increases are not available but SCS indicated a 5-fold increase during the past year in cover crop cost-share applications. To be most effective, cover crops must be seeded early, often before summer crops are harvested. Interest in aerial seedings before summer crops are harvested is increasing. The SCS office has published names, addresses, and estimated costs of 17 vendors in the mid-Atlantic equipped to make aerial cover crop seedings.
Producer should not add winter cover crops to his cropping system until he is convinced that production costs can be lowered, soil productively improved, soil losses reduced, and/or leaching losses reduced. He must then determine if an N-producing legume, an N-conserving grass or a legume/grass mixture best fits his operation. Cover crop use requires more planning for early fall cover crop establishment and proper spring management (cover crop kill date, tillage, corn establishment methods, corn FN, etc.) that best fits his operation.
Comments ranged from "very useful" to "not worth the added cost and planning." The most frequent advantage listed were: (1) N addition (2) N conservation (3) reduced soil wind and water erosion (4) water conservation (5) potential for 2nd silage crop. The most frequent disadvantages listed were: (1) increases work in spring & fall at already busy period, (2) soil moisture depletion, (3) slow soil warm-up, (4) requires more planning and attention to details.
Producer Involvement-number in attendance at:
Field Days: 410
Tour by Research Scientists: 60
Areas needing additional study
Studies are needed to establish time required to reach soil-N equilibrium with legume, grass or legume/grass cover crop mixture with different soils, environments, FN inputs, and management strategies.