Final report for LNE18-366
Project Information
Cover crops convey benefits that are typically maximized by maximizing the amount of time the cover crop is in the field. For example, delayed termination of cereal rye gives the cover crop a chance to potentially produce almost double the amount of biomass and thus scavenge more nitrogen (N) that might otherwise be lost to the environment. Growing cover crops for greater biomass, especially cereal rye, comes with the concern of reduced N availability at corn planting due to N tie up in the soil and subsequent potential yield loss. Any significant reduction in crop yield would inhibit the adoption of an agricultural practice. To avoid potential yield loss, delayed rye termination may require different fertilizer application strategies to account for the effect of large rye biomass on soil N dynamics in a following corn crop (i.e. it may be beneficial to apply more fertilizer as a starter and less at corn side-dress to protect corn yield). We tested varying ratios of starter to side-dress N fertilizer in corn to fine-tune the fertilizer strategy following delayed (i.e. late) rye termination. The educational approach in this project, which was put into action in workshops and through demos, featured a curriculum involving topics such as a) benefits of cover crop use, b) benefits of maximizing time the cover crop grows in a field, c) cover crop planting and termination methods and timings, and d) corn fertility programs following late-terminated rye.
Average yields for corn across the sites ranged between 7 Mg ha-1 and 13 Mg ha-1. Hypothesis 1 was disproved: we found that the amount of N applied as starter vs at side-dress did not affect corn yield irrespective of rye biomass, so long as adequate total N (either 168 kg ha-1 or 235 kg ha-1 total) was applied. Hypothesis 2 was partially disproved: rye biomass had no significant effect on corn yield (aside from in the 0 applied N control), even when large amounts of biomass were present. Hypothesis 3 was also partially disproved: while quantity of cereal rye biomass did not affect corn yield so long as adequate N was applied, N loss was reduced because the larger the rye biomass, the more N it has accumulated.
Previous modeling research has shown an increase in rye biomass (as through delayed termination) results in increased N accumulation in tissues, which makes the N unavailable to leach. We thus estimate that N loss was reduced because, while corn yield was not impacted, cover crop biomass (and concomitant N uptake) were increased due to delayed termination. As a result of this project, 67 farmers delayed rye termination on 11,348 acres, which we estimate prevented the leaching of 1196,320 lbs N total into Maryland watersheds. Our results indicate that the standard N fertility recommendations for corn production provided by agronomists and universities are currently robust enough to compensate for the potential for large quantities of rye to tie up N in the soil.
A total of 60 farmers in DE, MA, MD, NY, and PA will terminate cereal rye 2-4 weeks later than they usually would (April vs. early March) on 2,400 acres, doubling the average amount of cereal rye biomass produced and preventing the leaching of 45,000 lbs N into local watersheds.
Cover crops convey benefits that are typically maximized by maximizing the amount of time the cover crop is in the field. For example, delaying termination of cereal rye gives the cover crop a chance to potentially produce almost double the amount of biomass and thus scavenge more nitrogen (N) that might otherwise be lost to the environment. However, it is speculated that delaying rye termination may require different fertilizer application strategies to account for the effect of large rye biomass on soil N dynamics in a following corn crop. For example, it may be beneficial to apply more fertilizer as a starter and less at corn side-dress. The research approach in this project was to test varying ratios of starter to side-dress N fertilizer in corn to fine-tune the fertilizer strategy following delayed (i.e. late) rye termination. The educational approach in this project featured a curriculum involving topics such as a) benefits of cover crop use, b) benefits of maximizing time the cover crop grows in a field, c) cover crop planting and termination methods and timings, and d) corn fertility programs following late-terminated rye. This curriculum was delivered at workshops and demonstrations.
Cooperators
Research
1) If split nitrogen (N) applications are used, the amount of starter N applied at corn planting must increase proportionally with increasing cereal rye biomass.
2) Total N requirement for optimal corn yield is not influenced by the quantity of cover crop biomass at planting.
3) Larger amounts of cereal rye biomass will have a net positive effect on corn yield and reduce N loss.
Study sites. Field experiments were conducted at six agriculture experiment stations across the Northeast US. These locations included the Beltsville Agricultural Research Center (BARC; 39.03 ˚N, 76.93 ˚W) near Beltsville, MD; the University of Maryland Wye Research and Education Center (Wye, 38.91 ˚N, 76.15 ˚W) near Queenstown, MD; the University of Delaware Carvel Research and Education Center (38.64˚N, 75.46 ˚W) near Georgetown, DE; the Pennsylvania Department of Agriculture Livestock Evaluation Center associated with Penn State University (40.72˚N, 77.93 ˚W) near State College, PA; Cornell University Agriculture Experiment Station (42.46˚N, 76.44 ˚W) near Ithaca, NY; and the University of Massachusetts Crop and Animal Research and Education Center (42.48˚N, 72.58 ˚W) near Amherst, MA.
We established one experimental site at each location for two years. The Maryland sites conducted the experiment in 2017 and 2018; all the other locations conducted the experiment in 2018 and 2019. The sites at MD-BARC for 2017 and 2018 were both installed on a loam soil. The MD-BARC 2017 site was on a Russett-Christiana complex with 2.96% organic matter (USDA-NRCS 2020). The MD-BARC 2018 site was on a mixture of Codorus soil and Hatboro soil with 3.33% organic matter (OM) (USDA-NRCS 2020). The sites at MD-Wye for 2017 and 2018 were both on a silt loam [a Nassawango soil with 1.22% OM (USDA-NRCS 2020)]. The sites at DE for 2018 and 2019 were both on a loamy sand [a Rosedale soil with 1.25% OM (USDA-NRCS 2020)]. The sites at PA for 2018 and 2019 were on a silt loam [a Hagerstown soil with 3% OM (USDA-NRCS 2020)]. The sites at NY for 2018 and 2019 were established on a gravelly loam for 2018 (a Chenango soil with 4% OM) and a silt loam for 2019 (a Genesee soil with 4% OM) (USDA-NRCS 2020). The sites at MA for 2018 and 2019 were both established on a silt loam [a Winooski soil with 2.9% OM (USDA-NRCS 2020)].
All field experiments were conducted in corn/soybean fields in conventional rotation. Except for Wye 2017 which had corn in 2016, all sites had soybeans as the previous crop prior to the experiment. Corn was planted at all 12 sites with a no-till planter row spacing at 76cm and a population of 11,331 seeds ha-1 with a corn hybrid maturity length suitable to each of the sites locations and regions.
Experimental design. Treatments were established in an unbalanced split plot Randomized Complete Block Design with at least four replicates per field. The whole plot factor was rye treatment [early termination; intermediate termination; late termination; no rye cover crop (no cover crop control)]. Fertility treatments were the subplot factor. The unbalanced design was the result of the Nitrogen (N) and Sulfur treatment setup, which was not full factorial. The nitrogen fertility factor had three rates of total N in the treatments; 0 kg ha-1, 168 kg ha-1, and 235 kg ha-1. For each total N rate treatment, starter N was applied in four different amounts (0 kg ha-1, 28 kg ha-1, 56 kg ha-1, and 84 kg ha-1), with the rest applied at side-dress. The no cover crop control had N treatments consisting of 0 kg ha-1 total N, 56 kg ha-1 starter with 168 kg ha-1 total N, and 56 kg ha-1 starter with 235 kg ha-1 total N. For each replicate there was a total of 30 subplots: 27 subplots for the three whole plots of early, intermediate, and late termination of rye and three subplots for the whole plot of no rye cover crop. Each subplot at every site was three meters wide and at least 15 meters long, with four rows of corn in each plot (yield was only harvested in the two middle rows to reduce the potential for edge effects). This study focused on understanding corn N needs, not tracking N loss; estimates related to N loss are based on modeling methods from previous research (e.g., Mirsky et al., 2017).
In addition to the N fertility treatments, a sulfur treatment of 84 kg ha-1 in the form of gypsum was applied at MD-BARC 2018, MD-Wye 2018, DE 2018 and 2019, and PA 2018 and 2019. This was done as a split-split plot in the nitrogen treatment plots consisting of 56 kg ha-1 starter with a total N rate of 235 kg ha-1. This sulfur treatment was only applied to the higher total N rate treatment to ensure N was not a limiting factor in the treatment effects for S.
Field operations. Corn was planted at each site at least two weeks after the late termination of rye. Corn was planted at 76cm row spacing and a population of 11,331 seeds ha-1 using a no-till drill set to a depth of 5 cm. The starter N was applied at planting in the form of urea-ammonium nitrate. The starter N was applied to the soil surface 7.6 cm from the corn rows. The side-dress application of N was applied at corn growth stage V6. The side-dress application of N was also in the form of urea-ammonium nitrate and was applied to the soil surface 38 cm from the corn rows. Corn was planted and starter applied in April-May of each year, while side-dress occurred in June-July and corn was harvested September-October (Table 1). Corn planting, fertilizer application, and corn harvest were conducted by hand or with small-plot research equipment (Table 2). In the sulfur treatment, the gypsum was applied before corn planting by hand.
This study was conducted on non-irrigated land at all sites but emergency irrigation was permitted if there was a risk of the experiment failing due to water stress. At the DE site for 2019, emergency irrigation was needed; 5.1 cm of water was applied on July 11th, 2019 to the whole site using a forward driving irrigation system. Emergency irrigation was not needed elsewhere.
Table 1. Field operation dates.
|
Location |
Planting/Starter Date |
Side-dress Date |
Harvest Date |
|
MD-Wye 17 |
April 16, 2017 |
June 28, 2017 |
September 29, 2017 |
|
MD-Wye 18 |
April 30, 2018 |
July 2, 2018 |
October 1, 2018 |
|
MD-BARC 17 |
April 16, 2017 |
June 22, 2017 |
September 28, 2017 |
|
MD-BARC 18 |
April 11, 2018 |
June 21, 2018 |
October 22, 2018 |
|
DE 18 |
June 1, 2018 |
June 25, 2018 |
October 4, 2018 |
|
DE 19 |
May 20, 2019 |
June 11, 2019 |
September 23, 2019 |
|
PA 18 |
May 25, 2018 |
July 9, 2018 |
October 19, 2018 |
|
PA 19 |
May 22, 2019 |
June 1, 2019 |
October 8, 2019 |
|
NY 18 |
May 29, 2018 |
June 29, 2018 |
December 12, 2018 |
|
NY 19 |
June 3, 2019 |
July 12, 2019 |
November 26, 2019 |
|
MA 18 |
May 5, 2018 |
July 3, 2018 |
October 31, 2018 |
|
MA 19 |
May 5, 2019 |
June 3, 2019 |
November 15, 2019 |
Table 2. Field equipment used in this study.
|
Location |
Planter |
Side-dress Application |
Harvest Equipment |
|
MD-Wye 17 |
Kinze 2600 6 row planter |
Demco 12.2 meter boom sprayer |
Massey Ferguson 8XP 2 row combine |
|
MD-Wye 18 |
Kinze 2600 6 row planter |
Demco 12.2 meter boom sprayer |
Massey Ferguson 8XP 2 row combine |
|
MD-BARC 17 |
John Deere 7200 4 row planter |
Demco 12.2 meter boom sprayer |
Massey Ferguson 8XP 2 row combine |
|
MD-BARC 18 |
John Deere 7200 4 row planter |
Demco 12.2 meter boom sprayer |
Massey Ferguson 8XP 2 row combine |
|
DE 18 |
Kinze 2600 4 row planter |
Fabricated tractor mounted rig |
Massey Ferguson 8XP 2 row combine |
|
DE 19 |
Kinze 2600 4 row planter |
Fabricated tractor mounted rig |
Massey Ferguson 8XP 2 row combine |
|
PA 18 |
John Deere 1755 6 row planter |
Crop Care AGX300 |
Almaco SPC-40 2 row combine |
|
PA 19 |
John Deere 1755 6 row planter |
Crop Care AGX300 |
Almaco SPC-40 2 row combine |
|
NY 18 |
John Deere 7000 |
Hand application |
Almaco SPC-20 2 row combine |
|
NY 19 |
John Deere 7000 |
Hand application |
Allis Chalmers K2 Gleaner 3 row combine |
|
MA 18 |
Monosem NG+4 planter |
Hand application |
Almaco PMC20 2 row combine |
|
MA 19 |
Monosem NG+4 planter |
Hand application |
Almaco PMC20 2 row combine |
Data collection. Population counts of corn were conducted after side-dress and before harvest in each plot. Population was measured by physical stand counts of 6.1 linear meters representative of each plot. This was done to ensure failed populations of corn were accounted for when considering corn yield measurements so as to be able to adjust for outliers. There was no evidence in the experiment or from literature review(s) that our treatments should affect corn population and a failed plot for population should not be considered in our results. We used as a cut off 7,290 plants per hectare or less; below this population threshold our data and literature review(s) saw a significant impact to yield which should be considered as outliers. We conducted aerial imaging with drones, as discussed in this N Effects on Corn poster. Corn harvest was conducted when corn reached physiological maturity (R6) and was senescing, after the corn reached a minimum of 22% moisture. Corn was harvested with a specialized two row small plot combine with an integrated weight scale and moisture reader. Only the middle two rows from each four row plot were harvested to avoid edge effects. The front three meters and the back three meters of each plot also were not counted in the yield measurement to account for edge effect. A minimum of 6.1 meters was harvested for yield measurements at each plot in all of sites.
Statistical approach. Corn performance was modeled as a function of the preceding rye performance (i.e. rye biomass) due to the relationship between rye biomass and rye N content. Rye quantity and quality are correlated due to the treatment structure (termination date), so the variable for which we had the most complete data set was used (quantity); correlation = -0.68, P < 0.001. For estimating corn yield a linear mixed effects model (LMM) was used. The main effects were rye biomass in the proceeding cover crop, total N as a categorical factor, starter N as a categorical factor, and all of the interactions. The random effects were site, year within site, and field replicate within year within site; we included not only a random intercept but also a random slope for rye biomass. Goodness of fit was estimated using an R squared analog. For this model, means separation for the slopes and the intercepts were conducted using estimated marginal means (emmeans) using the Tukey adjustment method. Using a linear mixed effects model, we tested the interaction between S as a categorical treatment factor and cover crop termination as a categorical treatment factor as it affected corn yield with random intercepts for each site, year within site, and field replicate within year within site. Corn yield was adjusted to 15.5% moisture.
Weather. Sites varied notably in growing degree day (GDD) accumulation (Fig. 1). Precipitation varied by year, especially at DE, MA, and MD-BARC.
Starter N quantities did not affect corn yield no matter rye biomass. Average yields for corn across the sites ranged between 7 Mg ha-1 and 13 Mg ha-1 (Fig. 2). MA had the highest average corn yield as a result of soil characteristics and a history of manure use. DE had the lowest average corn yield due to having a sandier soil with lower fertility.
Rye had no significant effect on corn yield aside from in the 0 nitrogen (N) control, even when large amounts of rye biomass were present (Fig. 3). Goodness of fit was estimated at 0.71 using an R squared analog. We detected no differences among the slopes and the intercepts. There was a general positive trend (which was not significantly different from zero), which may have been due to the rye mulch providing weed suppression and water retention. As long as there was an adequate total N input, whether it was applied mostly as starter or at side-dress, rye biomass made no difference. Rye only had a statistical significance on corn yield when no total N was applied at all. We did not test total N treatments below 168 kg ha-1 so there may be a sub-optimal N rate that could capture an effect of rye biomass on corn yield. This work highlights the significance of side-dress N application as critically important because even in the 0 kg ha-1 starter treatment, corn yield potential was able to catch up so long as N was applied at side-dress.
Corn yield as affected by S application. We detected no differences in corn yield based on presence or absence of S application (Fig. 4).
Figure 1. Precipitation and cumulative GDD at study sites.
Figure 2. Corn yield by site.
Figure 3. Corn yield as a function of N application rate and rye biomass.
Figure 4. Effect of sulfur application and rye termination timing on corn yield at sites in PA, DE, and MD.
Any significant reduction in crop yield would inhibit the adoption of an agricultural practice. Growing cover crops for greater biomass, specifically cereal rye, comes with the concern of reduced nitrogen (N) availability at corn planting. This study demonstrated that, so long as adequate N is applied (either at 168 kg ha-1 or 235 kg ha-1 and no matter the amount applied as starter vs side-dress), cereal rye biomass is unlikely to affect corn yield. This finding is irrespective of cereal rye biomass quantity present. These results indicate that the standard N fertility recommendations for corn production provided by agronomists and universities are currently robust enough to compensate for the potential for large quantities of rye to tie up N in the soil. To investigate optimum usage of N for corn production, further study would be necessary using lower total N rates to identify the threshold at which large quantities of rye biomass begin to negatively impact corn yield potential and whether the amount of N applied at starter vs side-dress has an impact at that point. Further, this study focused on corn N needs using current recommendations as a starting point, not tracking N loss. Estimates related to N loss are based on modeling methods from previous research (e.g., Mirsky et al., 2017).
Education
The educational approach used in this project included 1-1 consultations, presentations, and newsletter articles featuring topics such as a) the value of including cereal rye in a corn/soybean rotation, b) rye planting timing and methods, c) rye termination timing and methods, d) value of delayed rye termination in the spring, e) management of late-terminated rye, and f) costs/benefits of altering corn starter/side-dress fertilizer ratios after late-terminated rye. The curriculum was presented at grower conferences and field demonstration days.
Milestones
1. We contact 2,000 farmers via e-mail with an invitation to participate in project activities (May 2018).
2000
3015
May 31, 2018
Completed
October 31, 2019
The number listed above represents the number of people who heard about the project via listservs, newsletters, and events.
2. An anticipated 600 farmers sign up to the project listserv to receive updates about research results and educational events in their state (June 2018).
600
3015
June 30, 2018
Completed
October 31, 2019
The number listed above represents the number of people reached via listservs, newsletters, and events. To avoid duplication (i.e. spamming farmers’ inboxes), we used existing listservs to send out information about the project, so the numbers for Milestones 1 and 2 are the same.
3. Approximately 400 farmers attend workshops held across the collaborating states and increase their knowledge of the benefits of delaying cereal rye termination to increase biomass, the risk of N immobilization that cereal rye poses when grown before corn, and strategies for adjusting N fertilizer applications in corn based on cereal rye quality and quantity at the time of termination (March 2019).
400
820
125
March 31, 2019
Completed
December 31, 2020
Workshop/presentation venues included the Delaware Ag Week Soil Health Workshop (1/14-17/19), the Union County (PA) Crop Conference (1/25/19), the 2019 Lower Eastern Shore (MD) Agronomy Day (1/30/19), the Indiana County (PA) Crop Conference (2/5/19), the Bradford County (PA) Crop Conference (2/12/19), a Central Susquehanna Valley Organic Crop Producer Study Circle (3/22/19), a Sunbury (PA) Soil Health Twilight Meeting (8/27/19), Mid-Atlantic Crop Management School (Ocean City, MD; 11/21/19), and the Practical Farmers of Iowa annual conference (Ames, IA; 1/17/20). Webinars included one hosted by the Pennsylvania 4R Alliance on "Managing Nitrogen with Cover Crops" (10/7/2020), and another by Penn State University Extension on "Nitrogen Management with Cover Crops" (11/19/2020). A poster was presented at the annual ASA, CSSA, and SSSA meeting (virtual; 11/9-13, 2020). The coronavirus pandemic precluded in-person meetings and demos in 2020.
4. Approximately 200 farmers attend a field day at on-farm demonstration sites and increase their knowledge of the benefits of delaying cereal rye termination to increase biomass, the risk of N immobilization that cereal rye poses when grown before corn, and strategies for adjusting N fertilizer applications in corn based on cereal rye quality and quantity at the time of termination (September 2019).
200
375
October 31, 2019
Completed
December 31, 2020
Field days at which this work was presented included the 2019 Crop Dairy Livestock Twilight Meeting (Mendon, MA; 7/31/19), University of Massachusetts Research Farm Tour (8/19/19), New Castle County (DE) Twilight Tour (6/4/20), University of Delaware Weed Day (6/19/20), University of Delaware Research and Education Center Field Day (8/14/20), and Sussex (DE) Conservation Field Day (8/20/20).
5. Approximately 100 farmers contact project personnel to receive assistance with managing delayed termination of cereal rye cover crops (April 2020).
100
150
April 30, 2020
Completed
October 30, 2020
Nate Richards was in contact with these 150 farmers (a combination of in person and via phone calls) as part of his responsibilities related to University of Maryland Extension's nutrient management program and this project.
6. Approximately 60 farmers agree to keep records of field activities and receive personalized assistance from project personnel (via email and phone consultations) to achieve the project performance target (May 2020).
60
73
May 31, 2020
Completed
October 30, 2020
The 73 farmers noted here were a subset of the 150 farmers mentioned in Milestone #5 who were participating in Maryland's nutrient management program. Project personnel provided assistance and reviewed farmer records to achieve the performance target.
Milestone Activities and Participation Summary
Educational activities:
One ASA-CSSA-SSSA 2020 Poster attended by 20 professionals.
Participation Summary:
Learning Outcomes
The 443 farmers reported in this section represent all attendees across collaborators’ outreach efforts. A sub-sample participated in a survey at the end of Nate Richard’s presentation at the 2020 Maryland Agronomy Day event, at which we verified changes in knowledge, skills, and attitudes. Data was collected with Turning Point software using clickers for participants after the teaching sessions were conducted. We asked participants to indicate expected impacts resulting from attendance at the event (select all that apply). Results: 47.92% reported improved crop production efficiencies, 26.84% reported increased use of conservation best management practices, 42.23% reported improved nutrient management practices, and 89.64% reported the information provided will benefit their operation.
Performance Target Outcomes
Target #1
60
Participating farmers will terminate cereal rye 2-4 weeks later than they usually would (April vs. early March).
2,400 acres
Prevention of the leaching of 45,000 lbs N into local watersheds.
67
Delayed rye termination.
11,348 acres
This project resulted in rye accumulating roughly 1196,320 lbs N total across 11,348 acres that it otherwise would not have taken up, preventing it from being available to leach.
Our calculation of performance target fulfillment is based on the following. Rye that would have been terminated at tillering (i.e. early) would be expected to accumulate 2.3% N x 900 lbs biomass/acre across 11,438 acres = an uptake of 234,904 lbs N. Rye that was instead terminated at stem elongation or boot due to this project would be expected to accumulate 1.9% N x 2,000 lbs biomass/acre across 11,438 acres = 431,224 lbs N. Net increase of N accumulated due to delayed termination: 431,224 - 234,904 = 196,320 lbs N across the full acreage involved. Nate Richards verified performance target fulfillment by reviewing farmer records as part of his duties related to Maryland extension nutrient management plans. We based our calculations on findings and models from Mirsky et al. 2017 ("Characterizing cereal rye biomass and allometric relationships across a range of fall available nitrogen rates in the eastern United States"), which comprehensively evaluated rye growth and N accumulation across latitudinal gradients.
Additional Project Outcomes
Project team members noted in 2019 that this project raised some good research ideas. One team member has a graduate student determined to investigate one of these ideas in her upcoming SARE Graduate Student grant.
Overall, team members believe that corn growers have two options to deal with their winter rye cover crops. The first option is harvesting the cover crop as emergency forage. In this case, time of planting in fall would be crucial.
The second option is the topic of the current project where the interactive effect of time of termination and N application is crucial.
As a supplement to this project, team members feel the following ideas should be pursued:
- Use of roller crimper instead of disking residues into soil and if the best fertility scenario will change with type of termination.
- Focusing on root residues and their decomposition trend.
- Manure application in fall and its impact on the best fertility scenario.