1. Introduction
In the U.S. Upper Midwest, the industrialization of agriculture and lack of true integration between crop production and livestock grazing has led to a simplification of agricultural land use. As such, continuous, annual corn (
Zea mays L.) and soybean (
Glycine max L.) cropping systems have replaced more diverse long-term rotations and perennial pasture [
1]. However, increased attention to environmental issues such as extreme weather events resulting in yield instability [
2,
3], the need to reduce nutrient and sediment loss [
4,
5], and a need for improved soil health [
6,
7] have led many Upper Midwest producers to reconsider their current farm practices [
8]. Cover crops, which are characterized by a wide variety of environmental and economic benefits [
9,
10,
11], are increasingly recommended to producers as a strategy to meet the above challenges.
Despite high levels of satisfaction with cover crops by users [
12,
13], adoption of the practice is low, with cover crops making up less than 3% of the total acreage in the Upper Midwestern states [
14]. Economic uncertainty about cover crops remains one of the greatest barriers to adoption [
15]. Although cover crops provide long-term economic benefits such as more efficient water and nutrient use [
16], reduced compaction [
17], and reduced pest and weed pressures [
18,
19], the upfront cost of changing farming practices [
13,
20] is a disadvantage that prevents producers from incorporating this practice into their operations. Increased focus by public and private researchers and non-profit organizations on ways to improve the near-term economic benefits of cover crops in the upper Midwest may hold the key to increasing its adoption.
Integrating cover crops as high-value forage sources with annual crop production systems may be an opportunity for cover crops to provide more immediate economic returns to producers [
21]. Feed represents the greatest expense in animal operations [
22,
23]; as such, high-value forage cover crops that increase on-farm economic and environmental efficiency have been increasingly explored as a potential tool to improve resilience [
24,
25,
26]. A study by Gabriel et al. [
27] found that selling cover crop biomass as animal feed resulted in significantly larger economic benefits than the fertilizer saving benefits provided by cover crops. In addition, Drewnoski et al. [
28] found that in cases of both spring and fall grazing, cover crops could offset production costs and, in many cases, result in economic returns exceeding the costs of establishment.
Despite the potential for economic gains, significant agronomic challenges must be overcome in order to successfully incorporate forage into row crop production systems. Interseeding, defined as planting a secondary crop during the vegetative growth stage of a commodity crop, is one strategy to incorporate forage cover crops into corn production systems and has been well-studied [
28,
29]. The ability of a secondary crop to acquire enough water, nutrients, and solar radiation is a significant limiting factor in the establishment and survival of interseeded forage cover crops [
30]; in addition to competing with the secondary crop for water and nutrients, the primary crop may impede the delivery of adequate radiation to the cover crop upon canopy close. Prior to 1960, most corn produced in the United States was grown in row widths greater than 76 cm [
31]. However, the current best management practices, which recommend a corn row width of 76-cm rows, provides a significant challenge for producers needing to achieve a successful forage crop [
32]. Widening corn rows to allow for the more effective utilization of sunlight [
32] for forage cover crops in an interseeding system is a novel approach, and may thus optimize the establishment and subsequent yield success.
Recent studies exploring the use of wide rows in corn systems are limited. Previous work has shown a neutral or negative corn grain yield in 152-cm rows compared to a traditional 76-cm row spacing [
33,
34,
35]. However, a number of previous research projects exploring diversified production systems have found that planting forage crops adjacent to corn may improve the cropping system yield stability and resilience [
36,
37,
38]. Clarity in both the agronomic and economic considerations of this novel system is required for researchers to endorse it for a wider audience. Therefore, this research was conducted with two main objectives: (1) quantify the total system outputs including grain yields, corn population, forage cover crop yield, and forage cover crop quality; and (2) develop enterprise budgets to evaluate the total value of the forage from the cover crop biomass and corn grain to assess the economic tradeoffs.
2. Materials and Methods
2.1. Site Description
Field experiments were established over three consecutive years at three working farms in Rice County, MN, USA and Goodhue County, MN, USA: Cherry Grove Township, MN, USA (44°13′37″ N, 92°48′49″ W), Goodhue, MN, USA (44°24′42″ N, 92°41′36″ W; 44°24′31″, −92°41′41″ W), and Faribault, MN, USA (44°14′18″ N, 93°08′46″ W; 44°14′34″ N, 93°08′57″ W). At the Goodhue and Faribault location, three field experiments were imposed each year (2019–2021), while at the Cherry Grove Township Location, field experiments were imposed over two years (2019–2020). All sites were planted in soybean prior to the establishment of the experiments. Soils at the field experiment locations included a Kason Silt Loam (fine-loamy, mixed, superactive, mesic mollic oxyaquic hapludalf) at Cherry Grove Township, a Mt. Carroll-Hersey silt loam complex (fine-silty, mixed, superactive, mesic mollic hapludalfs) at Goodhue, MN, and a Marquis silt loam (fine-silty, mixed, superactive, mesic mollic hapludalfs) and Nerwoods loam (fine-silty, mixed, superactive, mesic mollic hapludalfs) at Fairbault, MN. In 2019, the soil pH was 6.2 and the average organic matter (OM) was 3.1% at Goodhue. The soil pH was 6.7 and OM was 4.1% at Faribault, and Cherry Grove Township had a soil pH of 6.2 and 2.7% OM. In 2021, the experimental sites were moved to adjacent fields. Soil pH was 6.2 and average organic matter (OM) was 2.7% at Goodhue and soil pH was 6.1% and OM was 3.7% at Faribault.
2.2. Weather Data
Precipitation data were obtained from the NOAA reporting weather stations at Goodhue County (Zumbrota, MN, USA) and Rice County, MN, USA (Faribault, MN, USA) and departures from the 30-year average (1981–2010) were calculated (
Table 1). Due to the failure to collect the temperature data at the Zumbrota and Faribault NOAA reporting sites, the temperature data were obtained from the nearest NOAA reporting weather stations within 24 km of the Rice (Owatonna, MN, USA), and Goodhue County sites (Red Wing, MN, USA); departures from the 30-year average (1981–2010) were calculated (
Table 2).
2.3. Experimental Design
Field experiments were designed as a two-factor (i.e., corn row width and location), randomized complete block with three (2021) to four (2019, 2020) replications depending on the field-size constraints. The individual plot size was 18 m wide by 221 m in length. The row width treatments included: 76 cm with no forage cover crop (BMP), 76 cm with forage cover crop (BMP + CC), 76 cm with forage cover crop and two skip rows every 4
th row (Balanced), and 152 cm with forage cover crop (WIDE). Experiments were rain-fed. Each field experiment used conservation-tillage methods [
39] to prepare the seedbed prior to planting (
Table 3). Fertility management differed among locations (
Table 3) but followed the management practices outlined by the University of Minnesota Crops Extension program [
40]. All locations maintained the optimum fertility for corn production and were fertilized via a combination of N–P–K (as urea), and manure as recommended by soil testing prior to cash crop planting (
Table 3). Corn planting date, variety, and harvest dates varied among the study locations (
Table 3). The corn planting rate was 90,000 seeds ha
−1 in 76-cm rows and 180,000 seeds ha
−1 in 152-cm rows. Herbicide was applied pre- and post-emergence (
Table 3). Forage cover crops were established mid-June in the V3–V4 growth stages (
Table 3). Forage cover crops were broadcast or broadcast incorporated via tractor (Deer & Company, Moilene, IL) (
Table 3).
All forage cover crop mixtures were selected for high forage biomass yields and ruminate nutritional quality along with the recommendations of regional seed suppliers for the two counties (
Table 4).
2.4. Sampling & Analysis
Corn populations were calculated from late-September to mid-October (
Table 3) by counting a three meter length of row at three locations in each plot. The corn grain yield was measured by hand-harvesting a three meter length of row at three locations in each plot. Seed moisture was determined and adjusted before analyses to 150 g kg
–1. Aboveground forage cover crop biomass was harvested on the same day as the corn population counts at three random locations in each plot using 0.5 m
2 quadrats designed to capture a sample representative of the spatial arrangement of corn and forage in each treatment. Weeds were included in sampling as forage biomass. All forage cover crop biomass samples were dried at 60 °C in a forced-air oven until constant mass, after which all samples were weighed. The dry forage cover crop biomass samples were ground to pass through a 6 mm screen using a Thomas Wiley mill (Thomas Scientific, Swedesboro, NJ, USA). The coarse ground samples were mixed, subsampled (~30 g), and ground to pass through a 1 mm screen using a Cyclotec Sample Mill (FOSS North America, Eden Prairie, MN, USA). Ground samples were analyzed for forage nutrient composition by a commercial forage testing laboratory (Equi-Analytical, Ithaca, NY) using the following methods: crude protein (CP) was calculated as the percentage of nitrogen multiplied by 6.25 [
41]; NDF and acid detergent fiber (ADF) were measured using filter bag techniques [
42,
43,
44]. Relative forage value (RFV) was subsequently calculated using equations described by Jeranyama and Garcia (2004).
Statistical analyses were performed using the MIXED procedure in the statistical software SAS (SAS Institute Inc., Cary, NC, USA). Fixed effects were location, row width treatment, and their interactions. Random effects were block nested within year by location and corresponding interactions with fixed effects. Means for all response variables were separated using Fisher’s protected LSD at α = 0.05.
The economic analysis used a decision tool and partial budget format (
Table 5) to evaluate the differences in the costs and returns of wider than normal corn row widths (152 cm) with a cover crop planted between the rows when compared to the 76-cm rows. The cover crop forage was assumed to be grazed or mechanically harvested along with the corn stover. This wide-row scenario was compared with one of two base scenarios of 76-cm row widths. The base scenario included in the economic results below assume that only the corn grain is harvested, in contrast to the wide row scenario where the corn stover/forage mix is grazed. This comparison implies that the producer is only producing cash crops in the base scenario, and then somehow brings livestock (i.e., beef cows or backgrounding stock) onto the farm to graze in the wide row scenario.
The mix of forage cover crop and corn stover was valued based on the crude protein and TDN they contained, with the value per pound of crude protein and TDN if purchased as corn grain and 48% soybean meal. The expected price of corn and SBM were added by the user. As stover typically sells for less than corn and SBM containing the same quantity of these nutritional measures, it was discounted at a value determined by the user. The calculation also required the entry of the assumed corn grain yield at 76-cm rows for the normal row width, the reduction in corn grain yield in wide rows, and expected forage cover crop yield in wide rows; for our purpose, we used the experimental data averaged over all years and sites. Costs associated with the forage cover crop establishment, nutrient removal due to the stover harvest, and grazing or harvesting costs were included. Costs of growing the corn grain were not included in the analysis because they were the same in both scenarios.