Synergies between cover crops and corn stover removal

Michelle R. Pratt a, Wallace E. Tyner a,!, David J. Muth Jr. b, Eileen J. Kladivko c

a Department of Agricultural Economics, Purdue University, 403 West State Street, West Lafayette, IN 47907-2054, USAb Praxik, Inc., 2701 Kent Ave, Suite 130, Ames, IA 50010, USAc Department of Agronomy, Purdue University, 915 West State Street, West Lafayette, IN 47907-2054, USA

a r t i c l e i n f o

Article history:Received 2 December 2013Received in revised form 26 June 2014Accepted 28 June 2014Available online 25 July 2014

Keywords:Cover cropsCorn stoverRemovable biomassBenefit-cost analysis

a b s t r a c t

The potential harvest of corn stover as a feedstock for biofuels to meet government mandates has raisedconcerns about the agronomic impacts of its removal from fields. Furthermore, in order to meet thesemandates, larger quantities of stover will be required. As a result, increased attention has been placedon sustainable agronomic practices, such as cover crops. While cover crops may offer desirable benefits,adoption comes at a cost. The objective of this study was to determine the extent to which cover cropcosts could be compensated by additional stover removal and additional agronomic benefits from theuse of cover crops.

To meet the objective, we took three distinct approaches: (1) benefit-cost analysis, (2) integrated modelanalysis, and (3) representative farm analysis using the linear programming model PC-LP. Each approachwas a different take on the same issue; however, each provided different information. First, we estimatedcover crop costs and agronomic benefits and employed benefit-cost analyses, including stochastic anal-ysis in @RISK. Second, we tested cover crops with stover removal for 24 Indiana farms using PC-LP. Covercrop costs ranged from $81.76/ha to $172.50/ha, with variability being driven by differences in the seed-ing rate and seed cost. Agronomic benefits included reduced erosion, which was calculated using a newlycreated integrated modeling system.

The mean estimated reduced soil erosion with a cover crop and no residue removal was 0.72 metrictons/ha. An analysis of cover crop agronomic benefits resulted in private benefits (on-site) ranging from$91.45/ha to $192.07/ha, and $97.63/ha to $198.27/ha from society’s perspective. These benefits werehighly influenced by added or scavenged nitrogen (N) from the cover crop. For sensitivity we eliminatedthe benefit from added N and reevaluated the results. Without the N credit, benefits ranged from$74.72/ha to $134.62/ha. Benefit-cost analyses when considering the agronomic benefits of cover cropsresulted in a range of a net loss of $11.09/ha to a net benefit of $87.32/ha for the private perspective.

The integrated modeling system results indicated that, on average, while holding soil erosion constant,an additional 4.01 metric tons/ha of stover could be removed if a cover crop were used. Accounting forcover crop costs and stover removal, a benefit-cost analysis suggested that at a farm-gate stover priceof $66.14/metric ton, net benefits ranged from a loss of $3.78/ha to a net benefit of $86.93/ha. At afarm-gate stover price of $88.18/metric ton, mean net benefit ranged from $158.81/ha to $249.52/ha.

Results from the farm model (PC-LP) indicated that cover crops, along with increased stover removal,impacted crop rotations, increased the total amount of stover harvested, and had the potential to increasefarm profits.

! 2014 Elsevier Ltd. All rights reserved.

1. Introduction

As concerns with global warming increase, alternative energysources are continually being sought. The Energy Independenceand Security Act (EISA, 2007) established that by 2022, 36 billiongallons (144 billion liters) of biofuel are to come from renewable

fuel sources. More specifically, the mandate requires the produc-tion of cellulosic biofuels to increase to 16 billion gallons (64 bil-lion liters) ethanol equivalent annually by 2022 (EISA, 2007).Cellulosic biofuels are derived from several sources including corn(Zea mays L.) residue, or corn stover. Of the 16 billion gallons man-dated by 2022, 7.8 billion gallons (31.2 billion liters) are estimatedto come from corn stover (EPA, 2009). It is also favored due to thefact that it is readily available and has a high cellulosic content(Blanco-Canqui and Lal, 2009).

http://dx.doi.org/10.1016/j.agsy.2014.06.0080308-521X/! 2014 Elsevier Ltd. All rights reserved.

! Corresponding author. Tel.: +1 (756) 494 0199; fax: +1 (756) 494 9176.E-mail address: wtyner@purdue.edu (W.E. Tyner).

Agricultural Systems 130 (2014) 67–76

Contents lists available at ScienceDirect

Agricultural Systems

journal homepage: www.elsevier .com/locate /agsy

Corn stover (also interchangeably referred to as ‘‘residue’’, ‘‘sto-ver’’, and ‘‘biomass’’ throughout this study) is a crop residue whichis identified as the ‘‘above ground material left in fields after corngrain harvest’’ (Karlen et al., 2011). The components of corn stoverwere found to be 30% husks, shanks, silks, and cobs, and the reststalks, tassels, leaf blades, and leaf sheaths (Hoskinson et al.,2007). Crop residues, such as corn stover, which typically remainon the field, are responsible in numerous ways for preserving thesoil (Huggins et al., 2011). While corn stover could be a promisingsource of biofuels, several concerns have risen about its removalfrom the fields. This has brought new attention to conservationpractices, such as planting cover crops. While it has been observedthat moderate removal of stover may actually be beneficial (Swanet al., 1994), increased removal can have many adverse effects(Blanco-Canqui and Lal, 2009; Blanco-Canqui and Lal, 2007; Mekiet al., 2011; Wilhelm et al., 2004). Acceptable removal rates of cornstover vary across studies, but there is evidence to suggest thatrates of removal may be limited to around 33% of total availablestover due to the potential negative effects on soil quality and pro-ductivity (Blanco-Canqui and Lal, 2009; Graham et al., 2007; Kimand Dale, 2004; McAloon et al., 2000; Nelson, 2002; Petrolia,2006; Thompson and Tyner, 2011; Thompson and Tyner, 2014).Over the years, much research has been conducted to show theagronomic advantages and disadvantages of using various covercrops (Frye and Blevins, 1989; Dapaah and Vyn, 1998; Stivers-Young and Tucker, 1999; Kinyangi et al., 2001; Andraski andBundy, 2005; Snapp et al., 2005). More recent research has shownthat in addition to agronomic benefits associated with cover crops,there may also be an opportunity for economic gains if cover cropresidue could reduce subsequent fertilizer application and evenmore so if it can be sold as forage (Gabriel et al., 2013). Giventhe known benefits of cover crops, it is believed that these benefitscould mitigate the potentially adverse impacts of stover removal.Furthermore, the use of cover crops may allow corn stover to beremoved at higher rates, which could potentially increase farmrevenues.

While the Midwest Corn Belt region is seen as a major supplierof corn stover, cover crops have not been widely adopted. The aimof this study is to analyze the economic and agronomic impacts ofstover removal when done in combination with cover crops in theMidwest. Specifically, to what extent would it pay for famers toestablish a cover crop if it were possible to increase stover removalrates from 33% to 50% or higher. This analysis considers data fromseveral sources in order to quantify the benefits and costs of covercrops. Additionally, we evaluate the extent to which cover cropsallow for increased stover removal without adverse agronomicconsequences. Ultimately, the combination of information on sto-ver removal and cover crops is used to determine if the additionalrevenue from stover removal will compensate farmers for the costsof establishing cover crops.

2. Materials and methods

The data used (or applied) in this study comes from severalsources including the Midwest Cover Crop Council (MCCC) CoverCrop Decision Tool, farmer interviews, and anecdotal evidence.We consider six pure cover crops and two cover crop mixes forour analysis: (1) annual ryegrass (lolium multiflorum), (2) cerealrye (secale cereal), (3) crimson clover (trifdium incarnatum), (4)hairy vetch (vicia villosa), (5) oats (avena sativa), (6) oilseed radish(raphanus sativus), (7) annual ryegrass/crimson clover mix, (8)annual ryegrass/oilseed radish mix. Because of paucity of data,and in some cases poor understanding of how management prac-tices affect soils, we employ several methods and models to ana-lyze the costs and benefits of cover crops coupled with cornstover removal. Each method allows us to approach our objective

from a different angle. In doing so, each approach brings somethingto the overall picture and helps us confirm our results.

First we estimate the cost of cover crops. Next we quantify thebenefits of cover crops. Quantifying cover crop benefits involvestwo separate cases, both of which involve the use of an integratedmodel: one for agronomic benefits, and another for additional sto-ver removal. Once costs and benefits are quantified and estimated,a benefit-cost analysis with risk distributions is conducted. Finally,cover crop costs are used in a linear programming model to simu-late the impacts of cover crops and corn stover removal at the farmlevel based on real data for 24 Midwest farmers.

2.1. Cover crop cost estimates

We develop a method of cover crop cost estimation whichbreaks down costs into three components: (1) establishment, (2)termination, and (3) unexpected costs. Establishment costsassumed in this analysis are those costs that are required to aeri-ally inter-seed the cover crop in the fall into the standing cash crop.The components of the establishment cost therefore include therecommended cover crop seeding rate, seed cost, and the cost ofaerial application.

Recommended aerial seeding rates for each cover crop comefrom the MCCC Cover Crop Selector (MCCC, 2012) (available atwww.mccc.msu.edu) as a range and are measured as pounds ofpure live seed (PLS) per acre. This measure is adjusted to accountfor the percent purity and the percent germination of a cover crop.The percent purity for seeds is usually about 98–99% and percentgermination ranges from 85% to 90% (E. Kladivko, personal com-munication, September 4, 2012). Since an actual plot is not beingtested, we increase the recommended PLS rates for each cover cropprovided by the MCCC by 10%. Given the recommended aerialseeding rate for each cover crop and taking into account the costof seed, it is possible to estimate the seed costs of cover crops.Using quotes from seed suppliers listed in Clark (2007) and pricesstated by several farmers, a range of seed costs for each cover cropor cover crop mix is generated. The final component of the estab-lishment cost is the cost of aerial application, which is often doneat custom rates. Data for this component comes from anecdotalevidence (Vollmer, 2011) and farmer interviews. The mean esti-mated aerial application cost is $30.39/ha, the minimum is$24.71/ha and the maximum is $37.06/ha.

Termination costs assumed in this analysis are those costs asso-ciated with chemically killing the cover crop in the spring beforeplanting a cash crop, which very often is also done at a custom rate.Using estimated custom rate costs from anecdotal evidence (USDA,2011a) and farmer interviews, the estimated mean terminationcost is $15.76/ha. The minimum is $11.12/ha and the maximumtermination cost is $22.24/ha. It should be noted that in somecases, this chemical application would occur regardless of the pres-ence of cover crops or not. Therefore, since we cannot differentiatethe proportion of this cost that could be attributed to the use ofcover crops or standard field operations, the full cost is used inour analysis.

Due to the inherent risk that planting cover crops carries to afarmer, a cost item has been included to account for an unexpectednegative event, such as needing more than one pass of cover croptermination (chemical or mechanical) if it does not kill initially,untimely termination, the cover crop becoming a weed issue inthe following cash crop and/or the need to disc an area twice inthe spring. Snapp et al. (2005) reports on similar events as ‘‘indirecton-farm costs’’ where the establishment of a cover crop may inter-fere with the following cash crop or where the cover crop hasexcessive growth or becomes a weed. For this analysis, the totalunexpected cost is the probability that the unexpected cost willbe incurred multiplied by the associated cost per acre. Based on

68 M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76

two probability estimates of 10% (Ostendorf, 2010) and 15%(A. & C. Ault, personal communication) at an estimated cost of$32.12/ha (A. & C. Ault, personal communication, 2012), the meanunexpected cost is $4.03/ha, the minimum is $3.21/ha, and themaximum is $4.82/ha.

The total cost of a cover crop is then the summation ofestablishment, termination, and unexpected costs.

2.2. Cover crop benefits

In this analysis we consider the agronomic benefits that covercrops offer, as well as the additional benefits that may be associ-ated with the ability to sustainably harvest higher levels of residue.Cover crop agronomic benefits are estimated for four benefit cate-gories: (1) added nutrient content, (2) increase soil organic matter(SOM), (3) reduced compaction, and (4) reduced soil erosion.

Added nutrient content accounts for the ability of legume covercrops to add N to the soil, as well as the ability of non-legume covercrops to scavenge N and make it available for the subsequent cashcrop. Data on added N comes from the MCCC Cover Crop DecisionTool as a range of values. Adjustments were then made to this data.For annual ryegrass/oilseed radish mix and the oilseed radish covercrop, the values for added N are quite high and would most likelyonly be possible if the soil has manure applied to it as well.Therefore, the annual ryegrass/oilseed radish mix is adjusted to11.21–44.82 kg/ha, and oilseed radish cover crop is adjusted to22.41–56.03 kg/ha (E. Kladivko, personal communication, Septem-ber 4, 2012). Furthermore, data is adjusted to account for the factthat some of the N contributes to building SOM, while some willbe available for the next crop. This avoids double counting whenthe increase in SOM is considered. The assumption is that 50% ofthe N could be available for the next crop (E. Kladivko, personalcommunication, September 4, 2012). Added N is then valued. Thevalue of N comes from United States Department of Agriculture(USDA) historical US average farm prices of N fertilizers (USDA,2012). Average prices for three N fertilizers, anhydrous ammonia,nitrogen solutions (30%), and urea 44–46% are considered for2008–2012. By accounting for the percentage of N in each of thefertilizers, the price of the fertilizer in dollars per ton is convertedto the price of N in dollars per kilogram. Combining prices for all Nfertilizers, the mean cost is $1.15/kg, the minimum is $0.66/kg, andthe maximum is $1.48/kg.

Increased SOM is the percentage increase in SOM, which is aproxy for soil health, soil carbon and nutrient content, and therebylinked to soil productivity and crop yields. SOM is converted fromthe dry matter produced by each cover crop (Hoorman, 2012), andthe assumption here is that 25% of the dry matter produced fromthe cover crop becomes decomposed organic matter. The percent-age increase in SOM is that quantity (in tons) divided by the baseSOM. That % increase in SOM can be converted to a value per acreusing Eq. (1):

Increased SOM !$=acre" # SOM Increase !%"$ Value of SOM !$=!1%"" !1"

Reduced compaction accounts for the benefit of not having todeep rip fields, as well as enhanced root growth of the followingcash crop. This is the cost of deep tilling, assuming that the useof cover crops reduces the need for deep tillage of a field due toroot growth, which alleviates and/or prevents compaction of thesoil. The value of this is estimated to be between $74.13 and$86.48/ha (Hoorman, 2010). However, it is unlikely that a farmerwould deep rip more than once in every five years if there is a com-paction problem (E. Kladivko, personal communication, September4, 2012). Therefore, the estimates provided by Hoorman are

adjusted to reflect the probability of deep ripping once infive years, resulting in a reduced compaction value range of$14.38–$17.30 per hectare per year.

The reduced erosion is the difference between soil erosion(wind and water) with and without a cover crop. This estimatecomes from an integrated model, which will be discussed in moredetail below. The value of reduced soil erosion comes from theUSDA-NRCS (2011b) and is the cost to replace soil function andremediate off-site damage. There are on-site and off-site valuesof soil erosion. The on-site value represents the cost to the farmerof soil erosion while the off-site value represents the cost of soilerosion to society. The on-site value of soil erosion is $11.21/metricton, which accounts for reduced yields and water and nutrient loss.The off-site value of soil erosion is $19.83/metric ton, whichincludes impacts on air quality (health and property) and waterquality (USDA, 2011b). Although there may be other benefits, theyare not considered to avoid overlap and due to lack of data. Oncethe benefits of each category listed above have been quantified, arange of values, based on anecdotal evidence, is assigned to eachbenefit.

The second set of benefits we consider are those associated withstover removal. Cover crops alone appear to offer many benefits.We hypothesize that cover crops will allow for additional stoverto be removed. Provided there is an existing and viable marketfor corn stover, there will be an economic benefit associated withcorn stover removal. This benefit is the value of stover beyondthe cost of removal. The benefit from corn stover removal isdefined by the value of stover multiplied by the total stoverremoved, where the value of stover is equal to a farm-gatestover price less the on-farm harvest costs associated with stoverremoval. We test two farm-gate prices of stover: $66.14/metricton and $88.18/metric ton. These prices were selected based onresults from Thompson and Tyner (2014) which indicate thatsignificant stover harvest begins around $60/Mg, and substantialharvest at $80/Mg. On-farm harvest costs are those estimated byThompson and Tyner (2014) and Fiegel (2012).

In order to estimate the value of reduced erosion and estimatethe additional amount of corn stover that can be removed with acover crop, we utilize an integrated modeling system that com-bines the Revised Universal Soil Loss Equation, Version 2 (RUSLE2),the Wind Erosion Prediction System (WEPS), and the Soil Condi-tioning Index (SCI). RUSLE2 simulates daily changes in field condi-tions based on soil aggregation, surface wetness, field managementpractices, and residue status, and is driven by daily weatherparameters. RUSLE2 is used to guide conservation planning activi-ties and previous studies have shown the model to accurately rep-resent trends in field data (Ismail, 2008; Dabney et al., 2006; Fosteret al., 2006; Schmitt, 2009). RUSLE2 has also been applied to simu-late water erosion processes within broader analysis efforts rang-ing from watershed scale soil quality assessments (Karlen et al.,2008), assessing risks at abandoned mining sites (Vaszita et al.,2009), and socio-economic impacts of biophysical processes(Halim et al., 2007). WEPS uses a process-based daily time-stepmodel to simulate soil erosion due to wind forces considering bothdirection and magnitude (Wagner and Tatarko, 2001). WEPS mod-els a three-dimensional simulation region requiring a set of param-eters describing climate, soil aggregation, surface wetness, fieldscale, field management practices (including crop rotation andgrowth) and residue status, and is driven by daily weather projec-tions. WEPS has been evaluated for erosion predictions on croplandfields (Hagen, 2004) and has been used previously for case studiesin corn stover harvest (Wilhelm et al., 2007). RUSLE2 and WEPSeach calculate components of an NRCS-developed metric for estab-lishing management practice impacts on overall soil health, namedthe Soil Conditioning Index (SCI). The SCI provides qualitative pre-dictions of the impact of cropping and tillage practices on soil

M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76 69

organic carbon, which is an important factor in sustainable agricul-tural residue removal. The SCI has been used to support watershedscale soil quality assessments (Karlen et al., 2008), evaluate crop-ping systems in northern Colorado (Zobeck et al., 2008), and inves-tigate southern high plains agroecosystems (Zobeck et al., 2007).

The model, developed by Muth and Bryden (2013) utilizes adata and software integration framework that tightly couples theRUSLE2, WEPS, and SCI scenarios for fully automated high perfor-mance computing applications. The integration framework hasbeen named the Landscape Environmental Assessment Framework(LEAF) (LEAF, 2014; Moore and Karlen, 2013; Karlen and Muth,2013). The LEAF integrated model has been used for a broad rangeof studies investigating sustainable residue removal at the nationalscale (Muth et al., 2012a), at the regional scale (English et al., 2013;Bonner et al., 2014), at the subfield scale (Muth et al., 2012b; Muthand Bryden, 2012), and additionally for developing integrated bio-energy landscape designs (Karlen and Muth, 2013; Koch et al.,2012; Abodeely et al., 2012).

The data management and model inputs for the LEAF integratedmodel are described in detail in Muth and Bryden, 2013. The meth-odologies using publicly available data resources such as the NRCSSSURGO soils database, the NASS Cropland Data Layer, NRCS devel-oped climate databases, and Conservation Information Technologytillage databases are detailed in Muth et al., 2012a. Extensive valida-tion of the individual models is documented through the previouslymentioned studies. The integrated model has been verified to deli-ver analyses consistent with validation studies (Muth and Bryden,2013).

Using the LEAF integrated model, we have defined the userinputs as follows. The spatial area to be analyzed is the state ofIndiana, which has high corn production and large potential forstover removal. The management practices specified include: covercrops and cover crop combinations, residue removal, crop rota-tions, tillage practices, vegetative barriers, and yield drag with con-tinuous corn. Outputs from the integrated modeling system are theSCI and its three sub-factors, wind erosion, water erosion, andthe amount of residue removed. Using various combinations ofthe management practices, we will be able to extract two piecesof information to be used in the benefit-cost analyses: (1) the meanavoided wind and water erosion with a cover crop; and (2) theadditional biomass that is available for removal with a cover cropwhile holding total soil erosion constant. These values were esti-mated econometrically from the nearly two million data pointsthat were obtained from all the combinations of soil types, slope,management practices, etc.

First we sum the wind and water erosion values to obtain totalsoil erosion. Next, a cover crop dummy variable was created, wherethe value equals one if any cover crop is present and zero if nocover crop is present. The mean avoided soil erosion is thus calcu-lated as the difference between the mean erosion values with andwithout a cover crop.

In order to estimate the additional total removable biomass witha cover crop holding soil erosion constant, several steps were taken.First, we wanted to separate the impacts of no-till cultivation andcover crop. This is done by first estimating the following equationfor observations under each crop rotation with a cover crop andagain for observations under each crop rotation with no cover crop.

y # b0b1X1 % b2X2 !2"

where y is the total soil erosion, X1 is a tillage dummy variablewhich equals 1 if no-till and 0 otherwise, X2 is the annual biomassremoved.

The additional removable biomass from no-till, holding soil ero-sion constant is then:

&b1=b2 !3"

Therefore, the contribution of a cover crop to the amount ofadditional biomass removable with no-till and cover crops is thedifference between additional removable biomass by no-till witha cover crop and additional removable biomass by no-till with nocover crop.

2.3. Benefit-cost analyses

After the costs and benefits of cover crops have been estimated,a benefit-cost analysis is conducted to calculate the mean net ben-efit of a cover crop as well as the probability of a loss using MonteCarlo simulation. To account for variability in the costs and bene-fits, the Palisades risk and decision analysis software @RISK(2001), was used to perform the risk analysis. All of the uncertainvariables had a minimum, most likely (mode), and maximum. Thedistributions that are normally used in this situation are the PERTand triangular. We actually tested both distributions, but reporthere only the results from the triangular distribution, as they werequite similar.

There are two different perspectives for evaluating benefits: (1)there is no stover removal, and the benefits of the cover crop areagronomic (increased N, SOM, reduced compaction, and reducederosion), and (2) there is stover removal where the benefit is thevalue of stover removed, and there are no agronomic benefits con-sidered. Furthermore, in the case for which there is no stoverremoval and cover crop benefits are agronomic, we will assumetwo cases of reduced erosion: (1) on-site value of reduced erosionand (2) off-site value of reduced erosion. That is, we consider theprivate gain for the farmer of reduced erosion as well and thesocietal gain from reduced downstream erosion.

2.4. Analyzing costs and benefits at the farm level

After the cover crop costs and benefits are estimated, variousscenarios are analyzed using farm level data in PC-LP. PC-LP is alinear programming model that was developed within the Agricul-tural Economics department at Purdue University (West Lafayette,IN). Users specify input data including land, labor, machinery, cropyields, crop prices, and costs. Given these inputs, which are farmspecific, the PC-LP model determines the profit-maximizing cropmix (Doster et al., 2009a, 2009b). Input information for the pro-gram comes from farmers participating in the Top Farmer CropWorkshop at Purdue University. PC-LP is used to combine covercrops and corn stover removal.

In 2011, Thompson added stover harvest options into the PC-LPmodel by creating two new crops: BC + Stover (soybean corn rota-tion) and CC + Stover (continuous corn). Thompson (2011, 2014)assumed the stover-to-grain ratio to be 0.95 and the removal rateto be 33%. The corn harvest index is the ratio of corn grain to thesum of corn grain and stover, and is generally between 0.50 and0.55 (Michigan State University Extension, 2013). Thus, ourassumption is consistent with the normal values for corn harvestindex. The addition of stover removal into PC-LP involved account-ing for the harvest and storage costs associated with stoverharvest.

Expanding upon the methodology developed by Thompson(2011, 2014) cover crops can be added to the PC-LP model to esti-mate the impact of corn stover removal and crop mix at a farmlevel. We consider two cases: (1) a stover removal rate of 33% withno cover crop and (2) a stover removal rate of 75% with a covercrop. The difference between the two cases is the impact of covercrops at the farm level. Three cases of cover crop costs are analyzedas a means of conducting sensitivity analysis on the cover cropcost. Additionally, calculations are done at varying levels of stoverprice, beginning at $44.09/metric ton and increasing in $22.05/metric ton increments up to $132.28/metric ton.

70 M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76

There are three ways in which the addition of a cover cropaffects the current methodology in PC-LP. First, the addition ofthe cover crop is an extra cost to the farm. To account for the costsincurred by a cover crop, the cost will be added to the harvest cost;doing so allows us to adjust the crop price to reflect the use ofcover crops without changing the existing model. Second, it isassumed that a cover crop will increase the acceptable rate of sto-ver removal due to beneficial agronomic properties. Third, weassume the increase in stover removal rate will impact the harvestand storage costs associated with stover removal. Given the exist-ing methodology in PC-LP for the incorporation of stover removal,costs for all cases are estimated by ton and by hectare. Costs usedfor the base case (no cover crop) in this analysis are those esti-mated by Fiegel (2012) for stover that is harvested at 15% moisture.Costs used for the cover crop cases are those used for the base caseadjusted to reflect the increase in stover removal.

Cost components associated with stover removal include: stor-age, net wrap, labor, equipment, nutrients, and fuel. Harvest cost isthe sum of net wrap, fuel, labor and equipment used in harvestingstover per hectare. We assume that per ton, net wrap and nutrientcosts will remain the same for all cases, while storage, labor, equip-ment, and fuel costs will exhibit some economies of scale. Per hect-are we assume that storage, net wrap, and nutrient costs willincrease to match the full increase in the stover removal rate, whilelabor, equipment, and fuel costs will increase, however by a per-centage of the increase in the stover removal rate. Given the balesper hectare assumption, 15% moisture, and tons per bale assump-tion estimated by Thompson (2011) and Fiegel (2012), and ourestimated increase in the stover removal rate, we estimate thesecosts to increase by 64%.

For BC + Stover, the harvest cost was estimated at $87.47/ha forthe base case and $159.70/ha for the cover crop cases. A cost-sav-ing of $61.77/ha (Karlen et al., 2011) is assumed by Thompson(2014) for CC + Stover from reduced tillage, making the harvestcost for CC + Stover $25.70/ha for the base case and $97.92/ha forthe cover crop cases. Since tillage practice was not indicated inthe PC-LP model data, we assume reduced or conventional tillage,so the $61.77/ha savings will apply to the CC + Stover crop for allfarms in our analysis. A summary of the cost components for allcases is presented in Table 1.

Given the estimated values of stover harvest, and the assumedfarm-gate prices for stover ($/ton), the mean net value of stoverharvest is estimated. Using a stover farm-gate price of $66.14/met-ric ton, the average value of stover harvest (mean betweenBC + Stover and CC + Stover for the cover crop cases) is $22.87/met-ric ton. At a stover farm-gate price of $88.18/metric ton, the aver-age value of stover harvest is $41.92/metric ton. These mean valuesare used in the benefit-cost analysis as the value of stover removalper ton of stover removed.

3. Results and discussion

3.1. Cover crop costs

Estimation of the cover crop costs involved the use of MonteCarlo simulation in @RISK using the triangular distributions. Thereis a wide range of variability in cover crop costs. This is derivedfrom the differences in cover crop seeding rates and seed costs(Table 2). Annual rye had the lowest cost. While oats have the sec-ond highest average seeding rate, the seed costs is the second low-est on average. Similarly, annual ryegrass has a relatively moderateseeding rate and seed cost. Hairy vetch was the most expensivecover crop and appears to be a bit of an outlier since its mean costis more than $49.42/ha higher than any other cover crop. Whilehairy vetch does not have the highest seeding rate, it does havethe second highest seed cost. Oilseed radish has the highest seedcost, but its seeding rate is two to three times less than hairy vetch.As expected, the cover crop mixes have a mean cost that liesbetween the mean costs for the two individual crops that makeup the mix. This may indicate that cover crop mixes provide anopportunity for farmers to combine cover crops for maximum ben-efits at a lower cost.

3.2. Cover crop benefits

The integrated modeling system yields two results that are con-sidered as benefits of cover crops. The first is reduced soil erosionand the second is the potential for additional stover removal. Weanalyze the mean reduced soil erosion overall with and withouta cover crop. The mean difference in soil erosion with and withouta cover crop is 0.72 metric tons/ha. This value is used for thereduced erosion cover crop benefit category.

Second, using Ordinary Least Squares (OLS) regression analysisand holding soil erosion constant, we estimate the additional sto-ver that can be removed with a cover crop. Regression results areshown in Table 3 for our two sets of observations (those with covercrops and those without cover crops). Soil erosion is the dependentvariable, and total biomass removed and a dummy variable (NT)indicating no till or conventional till are the independent variables.By dividing the negative of the coefficient of NT by the coefficientof totBioRem_1 and converting to tons/ha we estimate the addi-tional removable biomass holding erosion constant. Comparingthe two results yields the benefit of a cover crop. Based upon theresults for all rotations combined, a cover crop appears to providea 4.01 metric tons/ha gain over no-till alone, with a larger gain forcorn-soybean than continuous corn. This value for additionalremovable stover with a cover crop is used for the benefit costanalysis case with stover removal.

Table 1Costs associated with stover harvest for the base case (no cover crop and 33% stoverremoval) and scaled for the cover crop cases (cover crop and 75% stover removal.

Cost component Base case Cover crop cases

$/ha $/metric ton $/ha $/metric ton

Storage 74.87 18.15 151.52 15.80Net wrap 26.02 6.17 59.15 6.17Labor 14.31 3.40 23.40 2.44Equipment 30.44 7.21 49.81 5.19Nutrients 59.20 14.03 134.57 14.03Fuel 16.70 3.96 27.33 2.85BC + stover total 221.55 52.50 445.81 46.48Tillage savings &61.77 &14.64 &61.77 &6.44CC + stover total 159.77 37.86 384.04 40.05

Table 2Total costs of each cover crop/cover crop mix as estimated by the mean of a triangularprobability distribution based on the cost of establishment, termination, andunexpected cost.

Cover crop/mix Seed ($/ha) Total ($/ha)

60% Annual ryegrass/40% Oilseed radish 43.38 94.8760% Crimson clover/40% Annual ryegrass 48.41 99.90Annual ryegrass 36.93 88.41Cereal rye 52.07 103.56Crimson clover 55.98 107.47Hairy vetch 121.01 172.50Oats 59.18 93.92Oilseed radish 53.91 105.40

Note: While uncertainty is included in all the cost components, the mean values arethe same for aerial application (30.72), termination (16.75), and unexpected costs(4.02). There is no termination cost for oats. The mean values for seed costs do varyconsiderably, so seed costs and total costs are included here.

M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76 71

Cover crop benefits were estimated for two cases. The first caseassumes that no stover is harvested. Therefore, the agronomic ben-efits of cover crops are accrued by the farmer. The second caseassumes that some stover is harvested. Cover crop benefits in thecase where no stover is harvested are shown in Table 4. Resultsfor the second case, where stover is harvested, are shown in Table 5along with results from the farmer perspective.

The estimated cover crop benefits suggest that a crimson clovercover crop provides the greatest benefit, while oilseed radish pro-vides the least benefit. Crimson clover has the second highest con-tribution of N and the largest SOM percent accumulation.Therefore, we can expect a large benefit overall. Oilseed radishon the other hand has low added N and low SOM percent accumu-lation. Although hairy vetch has the highest mean cost, it does nothave the highest mean benefit. The crimson clover/annual ryegrasscover crop mix offers larger benefits than the two individual covercrops. While the annual ryegrass/oilseed radish mix offers higherbenefits than the oilseed radish cover crop, the mean benefit is

slightly lower than the annual ryegrass benefit, which is due tothe low mean benefit of the oilseed radish cover crop. Cover cropbenefits from the societal perspective are uniformly higher by$6.20/ha.

In both of these cases, the benefit from crimson clover is signif-icantly larger than other cover crops. This is largely due to the Ncredit associated with crimson clover. Hairy vetch also has a highN credit. The standard assumption behind the N credit is that if acover crop is adding N, a farmer will reduce N application. How-ever, in our farmer interviews, many farmers do not abide by thisassumption. In other words, regardless of the N credit provided bya cover crop, they assume it to be zero and continue with their nor-mal regimen of N application. While we consider added N to be acover crop benefit, as there is added value into the soil, if farmersassume this value to be zero, it can impact benefit-cost analysisresults, specifically for legume cover crops such as crimson cloverand hairy vetch. Therefore, an additional case was tested to dem-onstrate the impact of the N credit becoming zero for all covercrops and it is also shown in Table 4.

While only legumes can add N to the soil, the other crops orig-inally had a value associated with the scavenged N, which is whythere is a decrease in net benefit for all cover crops. However,removing the added N benefit provides more balanced results.While crimson clover still has the highest benefit, hairy vetch nolonger has second highest benefit; cereal rye is higher. Further-more, crimson clover now has a benefit closer to cereal rye andannual ryegrass. The most commonly used cover crops from thefarmers we interviewed were annual ryegrass and cereal rye. Theseresults seem to confirm that farmers do not at present place a valueon added or scavenged N from cover crops.

The second case for which we estimate cover crop benefits iswhen there is corn stover harvest. The benefit of a cover crop withstover removal is the profit that can be made from stover once sto-ver harvest costs have been accounted for. Assuming that a

Table 3Ordinary Least Squares (OLS) regression results used to estimate additional remov-able biomass when a cover crop is present. These regressions are the result ofobservations from the integrated modeling system developed by Muth and Bryden(2013).

Variable No cover crop Cover crop

Constant 1.80771* 1.75204*

(0.02774) (0.02774)totBioRem_1 0.00048861* 0.000173250*

(0.00000527) (0.00000527)NT &1.35454* &1.10069*

(0.02761) (0.02761)R-squared 0.237 0.1035No. Observations 35679 115692

Standard errors are reported in parentheses.$ Significant at the 99% level.

Table 4The benefit of a cover crop measured in $/ha from the private perspective for the case where no N credit is assumed from the use of the cover crop versus the case where an Ncredit is accounted for.

Cover crop/mix Increased SOM No N Credit With N Credit

60% Annual ryegrass/40% Oilseed radish 69.87 93.99 106.2960% Crimson clover/40% Annual ryegrass 84.32 108.44 143.33Annual ryegrass 84.32 108.44 108.44Cereal rye 102.39 126.51 126.51Crimson clover 108.42 132.54 192.07Hairy vetch 69.87 93.99 167.89Oats 96.37 120.49 120.49Oilseed radish 50.59 74.71 93.17

Note: The benefits for reduced soil compaction (16.06) and reduced soil erosion from the private perspective (8.06) are the same for all cover crops. The increased soil organicmatter and N credit are the major differences among cover crops.

Table 5Summary of cover crop benefit-cost analysis for (i) the private perspective, where a cover crop is present and no stover is harvested, and (ii) the case where a cover crop ispresented and there is 75% stover removal. The net benefit for each case is the mean of a triangular probability distribution.

Cover crop/mix Private perspective With stover removala

Net benefit($/ha/year)

Standarddeviation

Probability ofnet benefit < 0

Net benefit($/ha/year)

Standarddeviation

Probability ofnet benefit < 0

60% Annual ryegrass/40% Oilseed radish 11.44 21.28 0.311 73.83 10.97 060% Crimson clover/40% Annual ryegrass 43.44 17.1 0.003 68.82 12.06 0Annual ryegrass 20.04 26.59 0.239 80.28 11.02 0Cereal rye 22.96 21.52 0.148 65.16 12.8 0Crimson clover 84.61 23.25 0 61.23 14.9 0Hairy vetch &4.6 24.51 0.588 &3.78 15.12 0.582Oats 26.56 29.26 0.184 74.8 22.14 0.002Oilseed radish &12.21 15.34 0.773 63.31 13.69 0

a The assumed stover price is $66.14/mt.

72 M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76

removal rate of 33% allows for about 3.36 metric tons/ha of stoverto be sustainably removed, and that an additional 4.01 metric tons/ha of stover can be removed with a cover crop, the total amount ofstover removed is 7.38 metric tons/ha. Accounting for the on-farmharvest costs, including nutrient replacement, the benefit of stoverremoval at a stover price of 66.14/metric ton is $168.69/ha and at astover price of $88.18/metric ton the benefit is $331.26/ha.

3.3. Benefit-cost analyses

We now combine the benefits and costs to obtain net benefits.In the cases with the use of a cover crop and no stover removal, thecost of the cover crop is incurred, and the agronomic benefits of thecover crop are accrued. The results from the private perspectivebenefit cost analysis are presented in Table 5. All cover cropsexcept hairy vetch and oilseed radish yield a net benefit. Althoughhairy vetch has large benefits, the seed costs are high enough thatthe mean net benefit is negative. Crimson clover has the highestnet benefit. While crimson clover had the second highest meancost, it had the largest mean benefit. As for the cover crop mixes,while crimson clover/annual ryegrass and annual ryegrass/oilseedradish had similar mean costs, crimson clover/annual ryegrass hasmuch higher mean benefits, yielding a net benefit about four timeslarger than annual ryegrass/oilseed radish.

The probability of a loss (probability that net benefit is less thanzero) is simply the probability in the stochastic analysis that netbenefit is negative. Most cover crops incur some probability of aloss, with the exception of crimson clover. Hairy vetch and oilseedradish are the two cover crops with a probability of a loss greaterthan 50% (58.8% and 77.3%, respectively). Although oilseed radishhas a negative net benefit, and the probability of a loss is high, mix-ing it with annual ryegrass yields a positive net benefit and reducesthe probability of a loss by about 50%. Furthermore, combiningannual ryegrass with crimson clover yields the second highestmean net benefit and a probability of loss very close to zero. Fromthe social perspective, the net benefit for each cover crop is uni-formly higher by $6.20/ha while the standard deviations remainthe same. The probability of a loss for each cover crop is slightlyless for the societal case.

The second benefit-cost analysis scenario is where a cover cropis present, and instead of accumulating additional agronomic ben-efits from the cover crop, benefits from stover removal, given a via-ble stover market, are included (also Table 5). The stover removedfor sale is the additional removable biomass holding erosion con-stant with a cover crop on top of a base of 3.36 metric tons/ha. Thisstover then is assigned a value. Since the market for stover has notbeen commercially established, we use a range of values based onprior research (Thompson and Tyner, 2014). In this case we analyzetwo farm-gate stover prices. The cover crop acts to hold agronom-ics constant (in other words, there should be no adverse effectsfrom stover removal). However, stover removal means that nutri-ents are removed from the ground. These removed nutrients areaccounted for by subtracting the cost of nutrient replacement fromthe value of stover. Although we account for nutrient replacement,it should be noted that other agronomic costs, such as increasedcompaction due to harvest machinery, are not accounted for. Har-vest costs associated with collecting the stover are included.

The first stover price tested is $66.14/metric ton with results inTable 5. Hairy vetch is the only cover crop with a negative meannet benefit (under a triangular distribution). For hairy vetch theprobability of a loss is 58%. The probability of a loss for all othercover crops is essentially zero, and the mean net benefit is between$61 and $80/ha. The standard deviation is about $11–$22/ha acrossall cover crops. Comparing this case to the cases with no stoverremoval and agronomic benefit, we can see that stover removalwith a cover crop offers significantly increased benefits.

Since actual stover prices are unknown, the sensitivity of stovervalue is tested but detailed results are not included here. The sec-ond case tests a stover value of $88.18/metric ton. Although thevalue of stover has increased by $22.05/metric ton, the net benefitfor each cover crop increases by $162.59/ha. This is because oncethe harvest cost and nutrient replacement costs have beenaccounted for, the remaining value of the stover for $88.18/metricton is much higher than $66.14/metric ton, yielding higher meannet benefits in the benefit-cost analysis. Furthermore, the probabil-ity that the net benefit is less than zero is essentially zero with astover price of $88.18/metric ton.

3.4. Costs and benefits at the farm level (PC-LP)

The 24 PC-LP farms have a total of 62,632 acres available.Results are for a base case with no cover crop and three cases ofcover crops at varying corn stover prices (Table 7). The base caseassumes no cover crop and 33% stover removal, while the twocover crop scenarios are estimated using a stover removal rate of75%. Cover crop costs used include those associated with annualryegrass and crimson clover. Furthermore, since our benefit-costanalyses suggest crimson clover as significantly outperformingannual ryegrass due to added N, we test a sub-case of the crimsonclover cover crop. In this sub-case, we assume that the farmers rec-ognize the added N from crimson clover and adjust their usual Ninputs accordingly. Therefore, this sub-case considers the per acrecost of a crimson clover cover crop less the value of added N peracre.

Farms will not begin to harvest stover until the benefit of stoverharvest exceeds the costs. Results from PC-LP indicate that at a sto-ver price lower than $44.09/metric ton, no farms will participate instover harvest, while at prices of $88.18/metric ton and greater, all24 farms will participate in some stover harvest. At $44.09/metricton, 8 farms harvest some stover for the base case and 0 farms har-vest some stover for all three cover crop cases. At $66.14/metricton, 24 farms harvest some stover for the base case, 21 for annualryegrass, 19 for crimson clover, and all 24 farms harvest some sto-ver for crimson clover adjusted for N.

PC-LP also determines the profit-maximizing crop mix forfarms. For the base case, at stover price of $0 and $22.05/metricton no acres are allocated to stover acres. However, beginning ata stover price of $41.51/metric ton, there is a shift from continuouscorn with no stover removal (CCorn), corn-soybean with no stoverremoval (BCorn), soybean acres, and other (such as wheat or milo)acres to include continuous corn with stover removal (CC + Stover)and corn-soybean with stover removal (BC + Stover) acres. Stoveris first harvested from all CCorn acres, then from BCorn acres. Asstover price increases, more acres are assigned to stover acres,and increasingly to CC + Stover acres. As a result, there is a declinein the assignment of acres to other crops. This is an indication thatas stover prices increase, there will be more incentive for farms tonot only harvest corn stover, but to also allot more acres to cornproduction. This pattern of acreage assignments also holds truefor the three cover crop cases. However, the shift to more cornacres with stover removal is rapid. For example, at a stover priceof $66.14/metric ton the percentage of acres assigned to CC + Sto-ver removal for all cases is as follows: 21% for the base case, 22%for annual ryegrass, 20% for crimson clover, and 24% for crimsonclover adjusted for N. Figs. 1 and 2 illustrate the acreage allocationby stover price for the base case and annual ryegrass, respectively.

From the PC-LP results we also analyze the total amount of sto-ver harvested at each stover price. Fig. 3 illustrates these results.Since for cases involving cover crops the stover removal rate isincreased to 75%, we expect to see greater quantities of stover har-vested in the cover crop cases. We observe that crimson cloverwith the N reduction allows for the greatest amount of stover to

M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76 73

be harvested. At stover prices of $88.18/metric ton and greater, theamount of stover harvested levels off and is almost indistinguish-able among the three cover crop cases; this is due to the limitplaced on the amount of stover that can be harvested. Since weallow for 75% removal in the PC-LP model, once this limit isreached, so long as the yield and stover-to-grain ratio remain thesame, the amount of total stover harvested will eventually flatten.

Finally, PC-LP allows us to analyze total farm profit. Given thatcover crops allow for increased stover removal, and greateramount of stover to be harvested, we expect that as stover priceincreases and more farms harvest stover, profits will also increase.Fig. 4 illustrates the results for farm profit for all cases. As shown,the farm profit with crimson clover after adjusting N, yields thelargest profit after a stover price of $44.09/metric ton. While thelowest cover crop cost yields the highest profit, all three cover cropcases are relatively similar, especially after a stover price of $66.14/metric ton. Furthermore, the cover crop cases offer significantlyhigher profit than the base case.

The results from the four PC-LP simulations provide an insightto the activities of profit-maximizing farms in Indiana. This is keybecause farms such as those whose data are in PC-LP are targetedfor stover removal in the Midwest to meet biofuel standards. Ittests how farms will react to added costs associated with covercropping if the practice will allow them to increase their stoverremoval without concern for farm agronomics. Furthermore,results confirm the findings from our benefit cost analysis; basedon cost alone, annual ryegrass offers greater farm benefits thancrimson clover, but when the cost of crimson clover is adjustedto reflect the value of added N, crimson clover provides greaterbenefits than annual ryegrass.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

44 66 88 110 132

Ass

ignm

ent o

f Are

a

Stover Price ($/Mg)

CCorn

CC+Stover

BCorn

BC+Stover

Soybeans

Other

Fig. 1. Assignment of total farm area in the base case, where no cover crop is usedand stover removal is at 33%.

0%10%20%30%40%50%60%70%80%90%100%

44 66 88 110 132

Assig

nmen

t of A

rea

Stover Price ($/Mg)

CCornCC+StoverBCornBC+StoverSoybeansOther

Fig. 2. Assignment of total farm area where an annual ryegrass cover crop isassumed and stover removal is increased to 75%.

0

20

40

60

80

100

120

140

160

180

200

40 60 80 100 120

Stov

er (M

g x

1000

)

Stover Price ($/Mg)

No Cover Crop Annual Ryegrass

Crimson Clover Crimson Clover N Adjusted

Fig. 3. Tons of total stover harvested for each case run in PC-LP, where the no covercrop case is the base case.

15

17

19

21

23

25

27

29

40 60 80 100 120

Tot

al P

rofi

t ($

106 )

Stover Price ($/Mg)

No Cover Crop Annual Ryegrass

Crimson Clover Crimson Clover N Adjusted

Fig. 4. Total farm profit for each case run in PC-LP, where the no cover crop case isthe base case.

Table 6Comparison of benefit-cost analysis results for annual ryegrass and crimson clovercover crop. Results include cost estimates, private and societal benefits and netbenefits without stover removal, and net benefit with stover removal for two prices ofstover.

Analysis Annualryegrass

Crimsonclover

Cost ($/ha) 88.41 107.46Private benefit ($/ha) 108.45 192.07Society benefit ($/ha) 114.63 198.27Private benefit ($/ha); N credit = $0 108.45 132.54Private agronomic net benefit ($/ha) 20.04 84.61Probability of net benefit < 0 0.239 0Societal agronomic net benefit ($/ha) 26.22 90.81Probability of net benefit < 0 0.177 0Net benefit at stover price = $66.14/metric ton 35.81 27.31Probability of net benefit < 0 0 0Net benefit at stover price = $88.18/metric ton 108.34 99.85Probability of net benefit < 0 0 0

Note: The benefit and cost values are means of the triangular probabilitydistribution.

74 M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76

4. Conclusions

The results from the two cases, which involved cover crops andstover removal, tell us several important things: (1) from theagronomic benefit analysis, in most cases cover crop alone offerpotential net benefits to farmers, (2) the net benefit of cover cropwith stover removal is sensitive to the value of stover (farm-gateprice), (3) cover crops with stover removal appear to have the abil-ity to substantially increase farm profit over cover crops alone, and(4) adding a cover crop to stover removal can, in many cases, paythe cost of the cover crop while allaying fears of increased erosionand SOM loss from corn stover removal. These cases are fairly gen-eralized and contain a certain amount of risk.

Overall, we can draw several key conclusions: (1) cover cropcosts and benefits vary by the selected cover crop, (2) the use ofa cover crop allows stover removal to sustainably increase byabout 4.0 metric tons/ha, and (3) the increase in stover removal,along with increases in stover price, changes farm acreage alloca-tions, increases the total amount of stover available, and increasesfarm profit. However, based on farmer interviews, and the resultsfor cover crop benefits when the value of added N is eliminated(Table 6), the benefits of cover crops perceived by farmers maybe lower than those estimated in this analysis.

Acknowledgements

Monsanto Corporation provided funding for this research. Addi-tionally, several Monsanto personnel were consulted during thestudy. Several farmers from Indiana were also consulted for infor-mation on cover crop usage and corn stover harvest.

References

Abodeely, J., Muth, D., Bryden, K.M., 2012. Integration of the DAYCENTBiogeochemical Model within a Multi-Model Framework. In: 2012International Congress on Environmental Modelling and Software. ManagingResources of a Limited Planet: Pathways and Visions under Uncertainty, SixthBiennial Meeting, Leipzig, Germany.

Andraski, T.W., Bundy, L.G., 2005. Cover crop effects on corn yield response tonitrogen on an irrigated sandy soil. Agronomy J. 97, 1239–1244.

Blanco-Canqui, H., Lal, R., 2007. Soil and crop response to harvesting corn residuesfor biofuel production. Geoderma 141, 355–362.

Blanco-Canqui, H., Lal, R., 2009. Corn stover removal for expanded uses reduces soilfertility and structural stability. Soil Sci. Soc. Am. J. 73 (2), 418–426.

Bonner, I., Muth, D., Koch, J., Karlen, D., 2014. Modeled impacts of cover cropsand vegetative barriers on corn stover availability and soil quality. BioEnergyRes. 7, 576–589.

Clark, A. (Ed.), 2007. Managing Cover Crops Profitability, 3rd ed. SustainableAgricultural Network (SAN), Beltsville, MD.

Dabney, S.M., Yoder, D.C., Foster, G.R., Nearing, M.A., 2006. Application of RUSLE2 topasturelands. In: Proceedings of the 14th International Soil ConservationOrganization Conference, Marrakesh, Morocco.

Dapaah, H.K., Vyn, T.J., 1998. Nitrogen fertilization and cover crop effects on soilstructural stability and corn performance. Commun. Soil Sci. Plant Anal. 29(17&18), 2557–2569.

Doster, D.H., Dobbins, C.L., Griffin, T.W., 2009. B-21 Input Form Guide Book.Department of Agricultural Economics, Purdue University, June 2009.

Doster, D.H., Dobbins, C.L., Patrick, G.F., Miller, W.A., Preckel, P.V., 2009. Purdue PC-LP Farm Plan. Department of Agricultural Economics, Purdue University, July2009.

Energy Independence Secutiry Act [EISA], 2007. Energy Independence and SecurityAct (EISA) of 2007. H.R. 6 (110th).

English, A., Tyner, W., Sesmero, J., Phillip, O., Muth, D., 2013. Environmentaltradeoffs of stover removal and erosion in Indiana. Biofuels, Bioprod. Biorefin. 7(1), 78–88.

Environmental Protection Agency [EPA], 2009. Draft Regulatory Impact Analysis:Changes to Renewable Fuel Standard Program. United States EnvironmentalProtection Agency, May 2009.

Fiegel, J., 2012. Development of a Viable Corn Stover Market: Impacts on Corn andSoybean Markets. MS thesis, Purdue University.

Foster, G.R., Romkens, M.J., Dabney, S.M., 2006. Soil erosion predictions from uplandareas: A discussion of selected RUSLE2 advances and needs. In: Proceedings ofthe Sino-American Workshop on Advanced Computational Modeling inHydroscience and Engineering, Beijing, China.

Frye, W.W., Blevins, R.L., 1989. Economically sustainable crop production withlegume cover crops and conservation tillage. J. Soil Water Conserv. 44 (1), 57–60.

Gabriel, J.L., Garrido, A., Quemada, M., 2013. Cover crops effect on farm benefits andnitrate leaching: linking economic and environmental analysis. Agric. Syst. 121,23–32.

Graham, R.L., Nelson, R., Sheehan, J., Perlack, R.D., Wright, L.L., 2007. Current andpotential U.S. corn stover supplies. Agronomy J. 99 (1), 1–11.

Hagen, L.J., 2004. Evaluation of the Wind Erosion Prediction System (WEPS) erosionsubmodel on cropland fields. Environ. Model. Softw. 19 (2), 171–176.

Halim, R., Clemente, R.S., Routray, J.K., Shrestha, R.P., 2007. Integration ofbiophysical and socio-economic factors to assess soil erosion hazard in theUpper Kaligarang Watershed, Indonesia. Land Degrad. Dev. 18 (4), 453–469.

Hoorman, J.J., 2010. Cover Crop Economics. Ohio Ag Manager, June. Ohio StateUniversity Extension.

Hoorman, J.J., 2012. Will that Cover Crop Check Bounce? [PowerPoint slides]. In: 8thAnnual Adams County No-till Meeting, Gettysburg, PA. <http://mercer.osu.edu/topics/agriculture-and-natural-resources>.

Hoskinson, R.L., Karlen, D.L., Birrell, S.J., Radtke, C.W., Wilhelm, W.W., 2007.Engineering, nutrient removal, and feedstock conversion evaluations of fourcorn stover harvest scenarios. Biomass Bioenergy 31, 126–136.

Huggins, D.R., Karow, R.S., Collins, H.P., Ransom, J.K., 2011. Introduction: evaluatinglong-term impacts of harvesting crop residues on soil quality. Agronomy J. 103(1), 230–233.

Ismail, J., 2008. RUSLE2 model application for soil erosion assessment using remotesensing and GIS. Water Resour. Manage. 22 (1), 83.

Karlen, D.L., Muth, D.J., 2013. Landscape management for sustainable supplies ofbioenergy feedstock and enhanced soil quality. Agrocencia 17 (2), 121–130.

Karlen, D.L., Tomer, M.D., Neppel, J., Cambardella, C.A., 2008. A preliminarywatershed scale soil quality assessment in north central Iowa, USA. Soil Till.Res. 99 (2), 291–299.

Karlen, D.L., Varvel, G.E., Johnson, J.M.F., Baker, J.M., Osoborne, S.L., Novak, J.M.,Adler, P.R., Roth, G.W., Birrell, S.J., 2011. Monitoring soil quality to assess thesustainability of harvesting corn stover. Agronomy J. 103 (1), 288–295.

Kim, S., Dale, B.E., 2004. Global potential bioethanol production from wasted cropsand crop residues. Biomass Bioenergy 26, 361–375.

Kinyangi, J.M., Smucker, A.J.M., Mutch, D.R., Harwood, R.R., 2001. Managing CoverCrops to Recycle Nitrogen and Protect Groundwater. Extension Bulletin E-2763.East Lansing, MI, Michigan State University Extension.

Koch, J., Muth, D., Bryden, K.M., 2012. An Integrated Modeling and DataManagement Strategy for Cellulosic Biomass Production Decisions. In: 2012International Congress on Environmental Modelling and Software. ManagingResources of a Limited Planet: Pathways and Visions under Uncertainty, SixthBiennial Meeting, Leipzig, Germany.

LEAF, 2014. Landscape Environmental Assessment Framework project coderepository. <https://code.google.com/p/leaf-tools/> (accessed 13.02.14).

McAloon, A., Taylor, F., Yee, W., 2000. Determining the Cost of Producing Ethanolfrom Corn Starch and Lignocellulosic Feedstocks. Technical Report. NREL/TP-580-28893. National Renewable Energy Lab.

Meki, M.N., Marcos, J.P., Atwood, J.D., Norfleet, L.M., Steglich, E.M., Williams, J.R.,Gerick, T.J., 2011. Effects of site-specific factors on corn stover removal

Table 7Comparison of PC-LP results at a stover price of $66.14/metric ton of the base case, where no cover crop is assumed and stover removal is 33%, to three different cases of covercrops where stover removal is at 75%. The three cover crop cases are annual ryegrass, crimson clover, and crimson clover when the N credit is accounted for.

Stover price = $66.14/metric ton Base case (no cover crop) Annual ryegrass Crimson clover Crimson clover (Adjusted)

Removal rate 33% 75% 75% 75%Farms participating 24 21 19 24% BC + stover acres 34.67% 32.46% 25.67% 33.66%% CC + stover acres 21.04% 21.60% 20.29% 23.99%Tons of stover harvested 50,352 111,152 95,887 117,736Harvest rate (metric tons/ha) 1.93 4.31 3.70 4.55Total farm profit ($) 17,677,745 17,663,419 17,370,976 18,179,671Farm profit ($/ha) 679.57 685.37 674.03 705.09

M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76 75

thresholds and subsequent environmental impacts in the Upper MississippiRiver Basin. J. Soil Water Conserv. 66 (6), 386–399.

Michigan State University Extension, 2013. Harvest index: A predictor of cornstover yield. <http://msue.anr.msu.edu/news/harvest_index_a_predictor_of_corn_stover_yield>.

Midwest Cover Crop Council [MCCC] Cover Crop Decision Tool, 2012. <http://mcccdev.anr.msu.edu/VertIndex.php>.

Moore, K.J., Karlen, D.L., 2013. Double cropping opportunities for biomass crops inthe north central USA. Biofuels 4 (6), 605–615.

Muth, D., Bryden, K.M., 2012. An investigation of sustainable variable rate residueremoval. J. Environ. Qual. 41 (6), 1796–1805.

Muth, D., Bryden, K.M., 2013. An integrated model for assessment of sustainableagricultural residue removal limits for bioenergy systems. Environ. Model.Softw. 39, 50–59.

Muth, D., Nelson, R., Bryden, K.M., 2012a. Sustainable agricultural residue removalfor bioenergy: a spatially comprehensive national assessment. Appl. Energy102, 403–417.

Muth, D., McCorkle, D., Koch, J., Bryden, K.M., 2012b. Modeling the impact ofvariability at the sub-field scale on sustainable agricultural residue removal.Agronomy J. 104 (4), 970–981.

Nelson, R.G., 2002. Resource assessment and removal analysis for corn stover andwheat straw in the Eastern and Midwestern United States—rainfall and wind-induced soil erosion methodology. Biomass Bioenergy 22 (5), 349–363.

Ostendorf, M., 2010. Annual Ryegrass. No-Till Farmer, pp. 3.Palisade Corporation, 2001. @Risk – Advanced Risk Analsysis for Spreadsheets,

Newfield, NY.Petrolia, D.R., 2006. Ethanol from Biomass: Economic and Environmental Potential

of Converting Corn Stover and Hardwood Forest Residue in Minnesota. Paperpresented at American Agricultural Economics Association Annual Meeting.Long Beach, CA, July 23–26, 2006.

Schmitt, L.K., 2009. Developing and applying a soil erosion model in a data-poorcontext to an island in the rural Philippines. Environ. Dev. Sust. 11 (1), 19–42.

Snapp, S.S., Swinton, S.M., Labarta, R., Mutch, D., Black, J.R., Nyiraneza, J., O’Neil, K.,2005. Evaluating cover crops for benefits, costs and performance withincropping system niches. Agronomy J. 97, 322–332.

Stivers-Young, L.J., Tucker, F.A., 1999. Cover-cropping practices of vegetableproducers in Western New York. HortTechnology 9 (3).

Swan, J.B., Higgs, R.L., Bailey, T.B., Wollenhaupt, N.C., Paulson, W.H., Peterson, A.E.,1994. Surface residue and in-row treatment effects on long-term no tillagecontinuous corn. Agronomy J. 86 (4), 711–718.

Thompson, J.L., Tyner, W.E., 2011 Corn Stover for Bioenergy Production: CostEstimates and Farmer Supply Response. Purdue Extension RE-3-W. WestLafayette, IN, Purdue University.

Thompson, J.L., Tyner, W.E., 2014. Corn stover for bioenergy production: costestimates and farmer supply response. Biomass Bioenergy 62, 166–173.

U.S. Department of Agriculture, 2011a. Cover Crop Costs. Conservation Systems FactSheet No. 04o. Agricultural Research Services, Auburn AL.

U.S. Department of Agriculture, 2011b. Soil Organic Matter: What is topsoil worth?Natural Resources Conservation Services. <http://soils.usda.gov/sqi/concepts/soil_organic_matter/som_value.html>.

U.S. Department of Agriculture, 2012. Average U.S. farm prices of selected fertilizers,1960–2012. [Dataset]. <http://www.ers.usda.gov/data-products/fertilizer-use-and-price.aspx#26727>.

Vaszita, E., Siki, Z., Gruiz, K., 2009. GIS-based quantitative hazard and riskassessment of an abandoned mining site. Land Contam. Reclamation 17 (3–4),513–529.

Vollmer, L., 2011. Cropping Decisions Survey & Focus Group. Partners, January 2011.<http://www.ctic.org/Partners%20Magazine/2010/December/28/>.

Wagner, L.E., Tatarko, J., 2001. Demonstration of the WEP 1.0 wind erosion model.In: Ascough II, J.C., Flanagan, D.C. (Eds.). Proceedings of the Soil ErosionResearch for the 21st Century. Honolulu, HI, 3–5 January 2001. ASAE Proc.International Symposium Report No. 701P0007, St. Joseph, MI, pp. 372–375.

Wilhelm, W.W., Johnson, J.M.F., Hatfield, J.L., Voorhees, W.B., Linden, D.R., 2004.Crop and soil productivity response to corn residue removal: a literaturereview. Agronomy J. 96 (1), 1–17.

Wilhelm, W.W., Johnson, J.M.F., Karlen, D.L., Lightle, D.T., 2007. Corn stover tosustain soil organic carbon further constrains biomass supply. Agronomy J. 99,1665–1667.

Zobeck, T.M., Crownover, J., Dollar, M., Van Pelt, R.S., Acosta-Martinez, V., Bronson,K.F., et al., 2007. Investigation of soil conditioning index values for southernhigh plains agroecosystems. J. Soil Water Conserv. 62 (6), 433–442.

Zobeck, T.M., Halvorson, A.D., Wienhold, B., Acosta-Martinez, V., Karlen, D.L., 2008.Comparison of two soil quality indexes to evaluate cropping systems innorthern Colorado. J. Soil Water Conserv. 63 (5), 329–338.

76 M.R. Pratt et al. / Agricultural Systems 130 (2014) 67–76