Final Report for FNC06-626
Project Information
PROJECT BACKGROUND
We operate a 170 acre integrated farm in southeast Wisconsin. Our products include: field crops; corn, soybean, winter wheat, oats and hay; fresh market vegetables sold roadside and at farmers markets, and sheep; breeding stock, feeder lambs, direct-market lambs and wool products. Our crops are produced using a “chem-lite” approach to production, maximizing use of management and on-farm resources to limit use of purchased inputs and to protect the farm’s environment. Our forage-based sheep enterprise uses rotational grazing and supplemental hay. Manure nutrients are recycled in forage production to maintain soil fertility, so cover crops are vital to meeting the nitrogen demand of field crops.
Cover crops are critical for supplying N in grain rotations for those that don’t have access to manure, yet want to reduce or eliminate purchase of external inputs. At the same time, they provide additional benefits of soil conservation and building, but these gains are easily negated by tillage to manage the legume. My primary reason for cover cropping is to reduce fossil energy use, improving the farms net energy balance, while capturing the associated benefits. We’ve used cover crops as the primary source of nitrogen for corn production off and on for the last 15 years and consistently for the past 5.
Red clover, frost-seeded into winter wheat before corn has been a reliable green manure crop for us and others when it can be incorporated with tillage. It is not however, well suited to no-till corn establishment because its rank growth prevents soil drying in spring and it survives winter, requiring herbicide to kill it. We converted to 100% no-till in 2003, so this has been a challenge. Recently, published research from Illinois also suggests that perennial legumes may attract the eastern variant of the western corn rootworm during their egg laying period, creating additional risk for the following corn crop and possibility for insecticide use to the detriment of the bottom line, farmer health and the environment. Finally and ideally from farm integration and risk management standpoints, a cover crop would be harvestable as hay should factors reduce yield in established hay fields. Red clover is very difficult to harvest in storable form and has inherent feeding problems, denying this benefit to producers who can only harvest dry hay. For these reasons, we are seeking an alternative.
PROJECT DESCRIPTION AND RESULTS
Our objective was to evaluate alternative legume cover crops for adoptability in wheat-corn crop sequence. Desirable cover crop attributes include the ability to be seeded after harvest of short season crops, acceptable productivity, no over-winter survival potential, and harvestability as dry hay of acceptable quality. These cover crops were compared to red clover, our standard cover crop and the “control” for this study. The evaluation was made on agronomic and economic comparisons.
PROCESS: A two-year field study was conducted near East Troy WI (N42o48.902’, W088o29.123’) in 2007 and 2008 to compare the productivity of annual legume cover crops seeded after winter wheat to that of medium red clover interseeded with wheat the previous spring. Site characteristics and growing season conditions can be found in table 1. [Editor’s Note: For copies of the tables mentioned in this report, please contact NCR-SARE at: ncrsare@umn.edu or 1-800-529-1342.]
Wheat (c.v. Kaskaskia) was no-till planted into soybean stubble at a rate of 1.6 million seeds/a. the previous fall for both years. Medium red clover (vns) was “frost seeded” the following spring at a rate of 10 lb/a. using a “spinner type” broadcast seeder. The summer seeded annual legumes: berseem clover (vns), annual sweetclover (c.v. Hubam) and chickling vetch (c.v. AC Greenfix) were no-till drilled into wheat stubble following grain and straw harvest at rates of 12, 10 and 50 lb/a. respectively. All seed was treated with the appropriate inoculant strain at seeding, even if the seed was preinoculated. Legumes were allowed to grow undisturbed until early November.
Pherocon AM yellow sticky traps were used to monitor western corn rootworm beetle activity in each plot during August. A description of this problem and scouting methods can be found at: http://www.entomology.wisc.edu/cullenlab/site_pages/extension/PDFs/Sticky_trap_protocol.pdf Traps were mounted on stakes at canopy level and were replaced every 7 to 10 days. Beetle population data was used to estimate “attractiveness” of the various species.
One-half of each plot (except the red clover treatment) was harvested as forage in early November. The other half of the plot was left undisturbed. Forage was sampled to determine yield and quality using near infrared spectroscopy (NIR). Plots were also sampled for aboveground biomass production by clipping random areas in the plots (total area 14 ft2). Samples were analyzed for dry matter (DM) and nitrogen (N) content to estimate aboveground nitrogen production. In 2008, red clover roots were also sampled and both tops and roots analyzed for total carbon content. The red clover treatments were killed with an application of glyphosate and 2,4-D ester after sampling.
In spring 2008, a corn response trial was superimposed over the 2007 plots. Corn (Jung 2587) was no-till planted into undisturbed plots at a rate of 30,000 seeds/a. in 30-inch rows. Starter fertilizer (6-24-24) was applied with the planter at a rate of 100 lb/acre. Weed control consisted of a post emergence application of atrazine and nicosulfuron with adjuvants. Presideress soil nitrate test samples were taken to a depth of 1 foot (3 cores/plot, bulked) at corn growth stage V6. Subplots were split and sidedress N (28-0-0) was applied immediately afterward at either a 0 or 60 lb N/a rate. Grain yield was estimated at physiological maturity by hand harvesting 2- 17’5” sections of row within each plot. Grain yield is reported at 15% moisture.
The experimental design is a strip-strip trial with 4 replicates. Previous legume was the main plot, harvest management the sub-plot and supplemental nitrogen the sub, sub plot. Corn yield data used only 3 replicates because of a mistake made during sidedress N application. Final plot size was 18.75 x 75’. Data were analyzed using analysis of variance procedures to determine significant main effects and interactions and means separated using a protected least significant difference (LSD) at a 0.05 level of probability. Orthogonal contrasts were used to compare red clover with the summer seeded annuals where appropriate. Not harvesting the red clover forage created an unbalanced design for corn measurements. As a result, two separate analyses were conducted: one with only unharvested treatments (all legumes); and one with the 3 summer seeded annual legumes with harvested and unharvested treatments.
PEOPLE
Dr. Eileen Cullen, Assistant Professor/ Extension Field Crops Entomologist and Kevin Shelley, Senior Outreach Specialist, UW Nutrient and Pest Management (NPM) program assisted with project interpretation and provided computer access for statistical analysis. Dr. Peg Reedy, Walworth County UW-Extension Agricultural Agent was unable to participate.
RESULTS
Growing season conditions
Rainfall in both years was well above average but distribution was uniquely different, creating a set of very contrasting growing seasons for testing legume performance (table 1). In 2007, a prolonged period of excess rainfall in August favored growth and development of all legumes. In 2008, excess rain in June and July followed by very dry conditions in late summer limited the performance of the summer seeded annual legumes.
DM and N production
Legume performance data can be found in table 2. Results are reported by year because of the extreme difference of the growing seasons and the resultant year by cover crop interaction.
Red clover significantly outperformed the annual legumes as a group as indicated by contrasts. It’s DM and N yields were similar in both years indicating it was less impacted by dry August conditions. This is probably due to the crop having better root development and the ability to use stored soil moisture deeper in the profile.
Berseem clover and annual sweetclover performed reasonably well in 2007 and total N yield, although statistically less that red clover, was at acceptable levels. However, performance of both decreased markedly in 2008 due to dry conditions. Visual assessment of the plots indicated that not only were the plants smaller, but stands also thinner. This is the result of poor or uneven germination as well as slow growth. Project photos are attached. [Editor’s Note: To see copies of the photos, please contact NCR-SARE at: ncrsare@umn.edu or 1-800-529-1342.]
Chickling vetch performed poorly in each year and does not appear to be well suited to this system.
Plots were not harvested in 2008 because of the lack of sufficient material to mechanically harvest. In 2007 with sufficient material, all three annual legumes displayed acceptable harvest qualities although chickling vetch was difficult to rake because the vines tangled on the rake tines, creating uneven and scattered windrows. When feed to sheep as dry hay, all three annual legumes were readily accepted.
Forage quality
As a group, the annuals tended to produce biomass with higher forage quality and this trend was consistently significant across all measures in 2008. In 2007 when plant growth was optimized, relative feed quality (RFQ) was not significantly different, presumably because plants were more mature. In 2008, summer annuals were delayed in germination and early growth and sampled material much more vegetative and therefore higher in quality. Red clover quality tended to be higher under conditions of drought stress which can’t be explained. In general terms, aggregate forage quality would be acceptable for maintenance diets (RFQ =120). In 2007 when acceptable yields were produced, berseem clover produced “dairy quality” forage. None of these legumes should be dismissed on the basis of forage quality, at least as they stand in the field. Productivity remains the major concern for using annual legumes for this purpose.
Rootworm beetle “attractiveness”
Rootworm beetle populations were monitored in all plots in 2007 with no significant difference between legumes, and at populations well below the threshold (5 BTD) to consider the need for rootworm control in the following corn crop. 2007 in general had lower populations than have been measured in the past. In 2008, monitoring only took place in red clover because the other legumes had not emerged and were therefore not an attractant.
Based on this one-year observation I would recommend that beetles still be monitored regardless of legume selected. Beetle populations fluctuate year to year and sufficiently high numbers may be attracted to a field to cause concern.
Carbon sequestration
In 2008, whole plant carbon content of red clover had a CO2 equivalent of 4.9 t/a, calculated from atomic mass. Although much of the carbon will be respired off during mineralization, this is a significant amount of annual sequestration and another reason to consider use of cover crops as opposed to fertilizer N. More work is needed in this area to collect additional information for the database and to estimate “carbon credits” to determine long-term sequestration. This could lead to cover crops being “creditable” on the open market, rewarding practitioners with monetary compensation.
Effect on corn grain yield
Corn grain yield following unharvested legumes is reported in table 3. Previous legume significantly affected corn yield and appears to be correlated to productivity. Red clover produced the greatest corn yield, regardless of the level of added N. Corn following all legumes responded to the addition of 60 lb N/acre, despite acceptable N yield for 3 of the 4 legumes the previous year. This response was fairly consistent across legumes, averaging 40.9 bu/a.
Presidedress soil nitrate tests reveled extremely low levels of plant available soil nitrate (<3ppm, critical value is 21 ppm) across all legumes with no significant treatment differences (data not shown). Cool, wet early season conditions presumably limited mineralization. The difference between red clover and the annual legumes is attributed to greater amounts of residue N mineralizing later in the growing season.
Although somewhat redundant, the effect of removing annual legumes as forage on corn yield is shown in table 4. Corn responded to previous legume, related to productivity, and added N, but not to the removal of biomass as forage. Although not significant, removal created a constant and interesting effect of increasing yield, regardless of legume or added N. This trend may be explained by reduced soil cover in spring, aiding warming and subsequent early growth of the no-till corn. This phenomenon also needs further investigation because the effect is counter intuitive. Removal of N-rich herbage should limit yield.
DISCUSSION
The annual legumes evaluated in this 2-year study failed to meet the criteria necessary for adoption. Although berseem clover and annual sweetclover proved productive under good conditions, their combined lack of productivity under stress conditions makes their use risky, both agronomically and economically. Medium red clover on the other hand proved productive and reliable across good and bad seasons. Chickling vetch proved unproductive in this cropping system.
Agronomic risk takes the form of highly variable yield: harvestable one year, not the next; as well as inefficient corn production the following year. Economic risk is cover crop failure, requiring a full rate of purchased N fertilizer the next year. This adds the full cost of the input on top of a substantial investment in cover crop establishment.
The average costs of each cover crop system are presented in table 5. Total cost includes seed and inoculant, establishment (either frost seeding or no-till drilling after wheat harvest) and termination in the case of red clover which will not winterkill. Standard burn-down herbicides and rates typically used in spring in no-till systems are insufficient to kill red clover, so this is a real, added cost.
While substantial, these establishment costs are approximately one-half the cost of a full (120 lb N/a) rate fertilizer, a substantial savings in the current economic climate and a good incentive to use cover crops. Table 5 also includes the cost of fertilizer used in the trial, applied at the one-half recommended rate for our soil type, or 60 lb N/a. These costs were used to conduct the economic analysis that follows. Partial budgeting was used to compare the economic impact of legume productivity and the subsequent effect on corn yield. With this analysis, only yields and costs which varied are used. This allows a simplified analysis which eliminates assumptions.
Table 6 shows the relationship of corn yield to cover crop cost with and without added N, measured as yield return (corn bushels per dollar of cover crop costs. Corn yield is related to previous legume productivity (tables 2, 3) so as expected; red clover has the greatest ratio, berseem and annual sweetclover intermediate and chickling vetch the lowest. This relationship holds up at both levels of applied N, although the ratios are lower where N was applied, indicating greater return to the initial investment in the cover crop. This relationship is identical to that seen in classic nitrogen fertilizer response trials where the first unit of applied N produces the greatest response. Although less efficient, the application of additional N was still profitable (table 8).
Table 7 shows the inverse relationship of table 6, the “nitrogen cost” of each legume, with and without supplemental N per bushel of corn produced. Red clover has the lowest cost at each N level, despite having a higher total system cost. This is due to better productivity of corn following red clover, offsetting the higher cost. This illustrates the point that more expensive cover crops can be cost effective if they are productive.
Table 8 contains the overall partial budget analysis, showing the net returns over cover crop and added N costs using a harvest time corn price of $3.50/bu. Drying costs are not included because they are negligible and not different between systems. The red clover system return was $86.54 greater than the berseem system (the closest alternative) with no applied N, and $126.54 more with 60 lb N/a. In a pure cover crop system, red clover was the most profitable and given it’s production stability demonstrated in this trial, clearly the cover crop of choice.
One of the selection criteria for the alternative legumes was harvestability as dry forage. Although berseem (and the others) is less appealing because of the yield stability problem, let's look at the economics of a forage harvest system using 2007 to 2008 production cycle. We never intended to harvest red clover, so its net return over nitrogen cost remains at $508.40. Harvested berseem clover produced 1.58 t/a of high quality forage (table 2) worth an estimated $80/ ton, or $126.40/acre. Recall that harvest of forage increased subsequent corn yield (table 4) and the berseem clover system yielded 167.9 bu/a for a gross income of $587.65/acre. Less cover establishment and added N costs, the net is $500.96. Add to that the value of the forage ($126.40) less a estimated harvest cost of $45.70/a (2007 Wisconsin Custom Rate Guide) and the net return over nitrogen costs for the system is $581.66/a or $73.26 more than red clover. Unfortunately we don’t yet have 2009 corn yields to calculate a 2-year, two cycle analysis to show the impact of 2008 berseem failure on the system. I have applied for further funding from SARE to complete this analysis.
The upshot
Productivity and reliability are the reasons we've use red clover in the past and will do so in the future. This trial help confirm that this is probably the best option for our location and cropping system, although we are seeking funding to complete the second cycle (corn response) and add a third cycle. This message (experience) will be shared with extension educators, conservation agency personnel and other farmers.
Meanwhile, we are back to the original problem of not having the flexibility of harvesting the clover as dry hay if supplemental forage is needed. A possible solution to this problem is harvesting it as wet hay using a custom operator, and wrapping each bale with plastic film to form “baleage” This is objectionable from a plastic use and disposal (thus sustainability) standpoint, but the system could still have a positive energy balance based on reduced fertilizer energy reliance. This system and comparison is also probably worthy of future study.
OUTREACH
We used several methods to tell people about our project and at the same time, share results.
Our planned field day for 2008 fell through when Dr. Peg Reedy, Walworth County Extension took extended medical leave. We also planned a field day for extension educators in September 2008, but that got combined with the SARE workshop described below for efficiency reasons. This is planned for Summer 2009.
We did however host several tours of the trails:
8/24/07, Michael Fields Agricultural Institute cover crop class (at the farm), 10 participants
8/22/08, Michael Fields Agricultural Institute cover crop class (at the farm), 7 participants
9/20/08, North American Clun Forest Association Annual Meeting, 28 participants
Trial information, photos and data were used at a SARE sponsored professional development workshop for educators and conservation agency professionals at the UW Hancock Research Station, 10/1/08, 45 participants, 50 in the audience. Five individuals have subsequently requested more information.
Trial data has been incorporated into UW-Extensions master cover crop data set maintained by Kevin Shelly (UW NPM Program) and myself. This data has been provided to Drs. Carrie Laboski and Matt Ruark, UW-Extension Soil Fertility Specialists.
Trial data has also been in incorporated into a soon to be published UW NPM program publication entitled: “Frost seeding red clover in winter wheat”. This publication should be available on-line at: http://ipcm.wisc.edu/Publications/tabid/54/Default.aspx in early March, 2009. Figure 2 was generated directly from trial data. Project data is incorporated into table 1 and figure 2.
PROGRAM EVALUATION
This is my only SARE funded project. I’m appreciative of the program and responsiveness of the staff. I was also impressed by how easily my progress report was found on-line. Ready access to information from grant-funded programs indicates to me an effectively administered program. I have no suggestions for improvements.