Improving Nitrogen Utilization with Rotation and Cover Crops

The impact of rotation of corn, soybean and wheat, including cover crops, was compared with continuous corn planting in south-central Michigan. Three studies were conducted as part of this research:

* Eighteen farms along a transect from southwestern across south-central Michigan, covering eight counties were paired for high and low crop diversity.

* A detailed verification study of clover frost-seeded into wheat was done over two years on a single farm from this transect.

* An intensive replicated rotation and cover crop study, the Living Field Laboratory, located along the defined transect, was monitored in its second and third year for comparative analysis.

Objective 1. High levels of crop diversity significantly increase microbial activity.
R. Harwood, E. Franco, T. Willson

Soil quality measurements of 23 parameters were conducted across the 18 paired farms of the transect. Some individual pairs of farms where rotation and cover crops could be compared with continuous corn showed increased nitrogen mineralization, increased microbial biomass, more rapid water infiltration and lower penetration resistance. Overall statistical analysis of the entire data set showed low correlation between crop diversity and most soil quality parameters, probably due to the overriding influence of soil type combined with differences in tillage and in manuring across varying levels of crop diversity. High crop and manure residue diversity tended strongly to result in increased accumulation of soil carbon, reducing the carbon to phosphorus ratio. This led to lowered bulk density, increased water infiltration rate and higher microbial biomass.

In controlled studies results are available only for short-term rotation effects. Patterns of response are nevertheless beginning to emerge. In experiment station studies during 1994, the second year of the rotation, soil microbial biomass in soybeans showed slightly higher levels with previously-applied compost and cover crops (Figure 1). In first year corn following wheat, where the cover had a longer growth period, microbial biomass was increasing substantially by the time of corn maturity (Figure 2). It is felt that with continuing rotation effect microbial biomass can be shown to increase when comparisons are made within a single soil type and field. Soil texture, and particularly its clay content, has major influence on both carbon accumulation and microbial biomass irrespective of rotation and cover crops.

Objective 2. Crop rotation and cover crops increase crop-available nitrogen.
R. Harwood, M. Jones, N. Dehne, T. Willson

Nitrogen mineralization potential as determined by laboratory incubation was higher in soybeans following composting (Figure 3), but the greater impact of clover in wheat was evident in the mineralization potential early in the season in succeeding corn (Figure 4). The incubation curves (Figure 5) illustrate the enhanced rates of mineralization with both clover cover and compost being plowed down before corn. These results are identical to those from on-farm studies at the Hall farm in southeastern Ingham County. The nitrogen mineralization in the field from the clover cover in this on-farm study gave significantly higher soil nitrate levels by time of sidedress, allowing a 70 lbs/A reduction in nitrogen application (Figure 6). Actual nitrogen equivalent of the clover was shown to be 120 lbs/A in a succeeding corn crop.

The most dramatic changes in soil nitrogen-supplying potential were seen in 1995 in Year 3 of the rotation in the experiment station trial.

In first year corn following wheat, the combination of compost and cover crop gave an increase in nitrogen mineralization to 10 ppm of soil nitrate by May 25, whereas corn without cover and fertilizer only reached 10 ppm of nitrate by June 20, nearly 3 weeks later (Figure 7). Corn yields were 169 bu/A in the compost/cover plots, and 144 bu/A in the fertilizer-no cover plots (Table 1). The fertilized plots without cover received 20 lbs/A of starter N and 140 lbs/A of topdressed N. Fertilized plots with cover yielded 158 bu/A, with a total of 110 lbs/A of applied N. This was a 10% yield increase with 50 lbs/A less nitrogen applied. Continuous corn without cover reached 10 ppm in soil nitrate in early July, and yielded 150 bu/A with 160 lbs/A of added nitrogen (Figure 8).

Nitrogen leaching loss is lower using compost and with cover crops. During the winter of 1993-94 in rotations using fertilizer, residue of first-year corn leached 36 lbs of nitrogen per acre. Second year corn leached 29 lbs with cover and 40 lbs without cover. Continuous corn leached 35 lbs with cover and 55 lbs without cover.

Cover crops after corn gave slight reductions in soil nitrate (Figure 9) which, together with increased evapotranspiration, particularly in the early spring, reduced leaching loss.

The leaching comparisons for continuous corn, with and without cover, are shown in Figure 10. Differences were particularly striking during the March and April periods, when much of the year's leaching occurred. The cover crop treatments lost 24 lbs of nitrate N/A, whereas the covered plots lost 55 lbs, or more than double that of the cover crop treatments.

Objective 3. Comparisons of "chemical" and organic management.
R. Harwood, M. Jones, N. Dehne

The Living Field Laboratory demonstrated the crucial nature of rotation and cover crops in the early stages of conversion to organic. Corn yields decreased significantly in second or third year corn with compost as a fertilizer source. The combination of clover cover and compost produced the highest yields in the third year of conversion. Nitrogen mineralization was highest in the clover-compost plots following plowdown, giving a 1-week to 10-day earlier nitrate buildup than in comparable fertilizer-cover crop rotation plots (Figure 7). This could have been due, in part, to better clover growth.

Yields of corn in organically-grown rotation were excellent for first year corn following wheat and clover, but dramatically reduced in second year and continuous (third year) corn (Table 1). It remains to be seen whether organic production will increase in second year corn as the conversion progresses and soil nitrogen mineralization potential increases.

Leaching losses were significantly reduced in compost-treated plots as compared to fertilized plots (Figure 11), but this was due, in part, to the lower nitrate levels and lower corn yield in second and third year corn.

Objectives 4 and 5. To evaluate potential multi-year benefits from nutrient management by using enterprise budgets and tracking environmental quality parameters.
S. Swinton, W. Roberts

Key findings:
Budgeting analysis reveals that:

1. Multiple crops in rotation appear to increase component crop yields and reduce costs, thereby raising financial gross margins.
2. Manure magnifies this effect, but has much less impact by itself.

The whole-farm analysis, based on a simulated, representative farm further shows that:

3. Because some manure is high in nitrogen, the profitable use of manure decreases when restrictions are placed on nitrate leaching and phosphorus runoff.
4. Interseeding clover into a rotation becomes attractive as restrictions are placed on leaching and runoff.
5. Farm returns were lowered very little to comply with the environmental protection constraints, indicating that while interseeding clover is not profitable alone, the financial sacrifice is quite modest in order to achieve lower environmental risks.

Detailed description of findings:

1. Forty-eight recent studies comparing alternative crop production systems were reviewed to identify an appropriate methodology for a joint economic and environmental comparison of alternative cropping systems (see Roberts and Swinton, 1996, under part 3.B "Dissemination of Findings"). Empirical methods were evaluated with respect to profitability, financial stability, and environmental impact criteria. Most studies failed to incorporate environmental criteria. Most of the studies offering balanced economic and environmental analyses integrated biophysical simulation with economic optimization. The three-fold conclusion drawn for this research was (1) to gather enterprise budget data on farm fields representing the different cropping systems of interest, (2) to supplement this information with simulation of processes that cannot be observed directly under typical farm conditions, such as nitrate leaching and phosphate runoff, and (3) to combine both kinds of information in a whole-farm model that can identify the costs associated with reducing nitrate and phosphate losses by specified amounts.

2. Personal interviews and telephone calls were used to collect field enterprise budget data from the 16 farmers whose fields were sampled in 1993. Personal interviews were followed by two or three phone calls to each farmer plus a possible follow up visit to obtain information of the management practices for specific field(s) of interest. Appendix B provides a brief description of the participating farm fields. Appendix C illustrates the locations of the cooperating farms on a map of Michigan counties.

Thirty-four enterprise budgets were developed (Appendix D) to estimate returns over variable costs per acre from 15 south-central Michigan farms during the 1994 growing season. Data were collected on labor and machinery by task and variable inputs used. For fields cropped in rotation, budgets were developed for each crop in the rotation by tracking fields with the rotational crops. For instance, if the farm operator followed a corn-soybean-wheat rotation with soybeans on the sampled field in 1994, corn and wheat fields similar to the sampled field were also monitored. The budgets are based on crop prices and input costs from mid-Michigan during the winter of 1994-95, along with custom work costs developed from a 1992 Michigan survey. Since the custom rates account for labor and equipment use, the returns over variable costs cover the returns to land, buildings, and management.

Results of the analyses are listed in Table 1 (Appendix A). Analysis of variance (ANOVA) was conducted to investigate where there were significant differences between mean yields, costs, and gross margins in the surveyed fields. Fields were contrasted in four ways. The first one paired farms growing corn continuously with farms growing more than one crop in rotation. The second paired farms using manure with those not using manure. The third paired those growing more than one crop and using manure with all other farms, and the fourth group compared those rotations with cover crops to those without cover crops. The statistical F-tests revealed mean differences at the 25% and 10% significance levels.

While use of manure shows no effect on yield, both cover crops and multiple crop rotations appear to increase yield significantly at the 25% level. On an individual basis, manure appears to have the greatest cost-reducing effect with significant differences at the 10% level. Multi-crop rotations reduce costs, but cover crops increase them (both at the 25% significant level). Differences in gross margins are evident in manure and multi-crop rotations both jointly and separately. The higher yields and reduced variable costs for the multi-crop rotations combine for a greater effect on gross margins than manure alone. The use of cover crops appears to increase both variable costs and crop yields significantly. Since these affect gross margin in opposite directions, there is little to no effect on gross margins.

Historical high and low price ratios of soybean:corn and wheat:corn, derived from the past 15 years of Chicago cash prices, were used in the budgets to determine sensitivity of the gross margin results to changes in price ratios between the crops. Table 2 (Appendix A) summarizes the changes that occur with shifts in the price ratios. Table 3 lists the farm-gate prices and price ratios used. Both high and low price ratios generally resulted in smaller differences in gross margin across the groups than mean price ratios. The exception was rotation vs. no rotation under high price ratios. When the prices of soybean and wheat are low relative to corn, no significant differences existed in gross margins. The increased relative price of corn compensated for the lower yields of continuous corn, erasing the crop rotation difference. Under the high price ratio case, the higher relative prices of soybean and wheat make crop rotation advantageous. Farmers who chose to grow other crops in rotation to corn would be better off in two of the three scenarios and equally well off under high corn prices. Therefore, growing corn in rotation in this analysis would be the dominant strategy for a price risk-averse or risk-neutral farmer.

3. Having constructed and compared enterprise budgets, the next step in the analysis was to develop a representative farm that reflects as much as possible typical field-level practices and inputs used on the fields sampled. The primary objective was to determine the optimal mix of enterprises for the representative farm under various assumptions about tolerable levels of nitrate leaching and phosphorus runoff. A secondary objective was to identify conditions under which manure and interseeded crops enter the optimal activity mix. These objectives provide insight into the tradeoffs that exist between profitability and environmental impact.

A linear programming model was used to determine the optimal mix of enterprises for the representative farm. The model provides a mechanism to answer questions such as how the enterprise mix and management practices might change if restrictions were placed on tolerable levels of erosion or potential nitrate leaching. The LP model used in this study is PCLP, the Purdue Crop/Livestock Linear Programming Model, version 3.2 (Dobbins et al. 1994), a whole-farm LP software that explicitly accounts for limited field time and penalizes planting and harvest delays.

The key environmental coefficients used in PCLP came from PLANETOR, version 2.0. PLANETOR combines site-specific environmental models with farm enterprise budgeting to evaluate the impact of crop rotations or changes in levels of applied nutrients and manure. Unlike PCLP, PLANETOR is not an optimization model; rather, it is designed to evaluate individual farm enterprises. It was used by us to evaluate the environmental impact of typical field-level practices observed in the sample. While PLANETOR is able to estimate enterprise returns and environmental impacts, PCLP identifies the enterprise mix that maximizes whole-farm returns to resources while meeting environmental and economic resource constraints.

The simulated, representative farm is a cash grain operation with 1250 tillable acres located in south central Michigan on Kalamazoo sandy-loam soil. Crops are produced using either conventional or minimum tillage. A conventional set of machinery is assumed, reflective of the equipment used by the farm operators surveyed. PCLP makes adjustments for yield and moisture levels based on the timing of planting and harvesting. This is important due to reductions in yield due to delays in planting and harvesting delays. Available field days estimates for a typical Kalamazoo producer represent the number of good working days in a ten day period at an 80 percent probability. Input costs are the same as those used in the enterprise budgets. Farm-gate commodity prices reflect historic price ratios observed over the last fifteen years for corn, soybean, and wheat harvest prices (see Table 3). No per-unit cost is attached to manure since it is assumed to be acquired at no cost from a neighboring farm. This type of arrangement existed among two of the five surveyed farms using manure. The only cost associated with manure is the cost of spreading. Restrictions are placed on pounds of nitrate leaching allowed per year and phosphorus runoff. Yields are assumed equal across practices at 135 bu/ac for corn, 43 bu/ac for soybeans, and 61 bu/ac for wheat.

Mean nitrate leaching and phosphorus runoff levels for each cropping activity were predicted from two submodels in PLANETOR 2.0. These submodels are the Phosphorus Runoff Index, developed by the Natural Resources Conservation Service Phosphorus Index Core Team, and the Nitrogen Leaching and Economic Analysis Package (NLEAP), developed by Agricultural Research Service at Fort Collins, Colorado. The PLANETOR model runs through ten years of every rotation to account for carryover effects of the soil and crop diversity over time. Levels of phosphorus runoff and nitrate leached were simulated for each rotation on each field over a ten- year period. Results generated by PLANETOR represent the annual averages in the eleventh year of each rotation.

The nitrate leaching and phosphorus runoff estimates from PLANETOR provide the numerical values used as parameters in PCLP's activity matrix. Results are generated for an unconstrained, profit-maximizing scenario, followed by restrictions on each environmental factor separately and both together. Limits are placed at the whole-farm level at an average of 40 lbs. of nitrate leached per acre and 8 lbs. of phosphorus runoff allowed per acre. These levels represent the upper limits of low environmental risk as defined within PLANETOR.

Table 4 (Appendix A) shows the returns to resources and the optimal crop mix for the initial unconstrained solution, as well as when restrictions are applied to the model. These represent the profit-maximizing solutions given the production alternatives, available resources, and current cost and price structure. The combination of enterprises that provides the highest returns is derived predominantly from a corn-soybean-wheat rotation using manure. Clover is not used in the optimal unconstrained solution. The "return to resources" of $220,016 represents the return that remains after all direct costs of production have been deducted from gross revenue. This covers returns to machinery and buildings, operator and family labor, management, and land.

The alternatives to the base model involve whole-farm restrictions on the total amount of nitrate leaching and phosphorous runoff. While all three scenarios decrease the return to resources, these reductions are very small, at the most $0.02 per acre. However, important changes do occur in the crop mix: Interseeded clover enters the optimal mix when phosphorus runoff restrictions are added to the model. The overall distribution of crops remains the same. The use of clover increases with nitrogen restrictions. While the corn-soybean-wheat rotation with manure dominates all scenarios, the use of second-year corn decreases as restrictions are placed on the model and clover is relied on more to reduce leaching and runoff.

These results are sensitive to changes in assumptions about price ratios (Table 5, Appendix A). When the prices of soybeans and wheat are high relative to corn, the optimal mix shifts away from corn to greater soybean production. At low price ratios, where the value of corn increases relative to other crops, the optimal mix shifts towards continuous corn production.