Contents lists available at ScienceDirect

Field Crops Research

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

Long term biochar effects on corn yield, soil quality and profitability in theUS Midwest

Deborah M. Allera, Sotirios V. Archontoulisa,⁎, Wendong Zhangb, Wendiam Sawadgob,David A. Lairda, Kenneth Moorea

a Department of Agronomy, Iowa State University, Ames, IA, 50011, United StatesbDepartment of Economics, Iowa State University, Ames, IA, 50011, United States

A R T I C L E I N F O

Keywords:APSIMNitrate leachingSoil organic carbonResidue harvestingEconomic analysisSustainability

A B S T R A C T

Corn production in the US Midwest has the potential to generate a large amount of crop residue for bioenergyproduction. However, unconstrained harvesting of crop residues is associated with a long-term decline in soilquality. Biochar applications can mitigate many of the negative effects of residue removal but data and economicanalyses to support decision making are lacking. To explore sustainable and profitable practices for residueharvesting in central Iowa we used 11 years of soil, crop yield, and management data to calibrate theAgricultural Production Systems sIMulator (APSIM) biochar model. We then used the model to evaluate howdifferent biochar types and application rates impact productivity and environmental performance of conven-tional corn and corn-soybean cropping systems in Iowa under different N fertilizer application rates and residueharvesting scenarios. A cost-benefit analysis was also employed to identify the economically optimal biocharapplication rate from both producer and societal perspectives. Modeling results showed for both continuous cornand corn-soybean rotations that as biochar application rate increased (0–90Mg ha−1) nitrate leaching decreased(2.5–20%) and soil carbon levels increased (8–115%), but there was only a small impact on corn yields (–2.6 to0.6%). The cost-benefit analysis revealed that public benefits, evaluated from decreased nitrate leaching andincreased soil carbon levels, significantly outweighed the private revenue accrued from crop yield gains, and thata biochar application rate of 22Mg ha−1 was more cost-effective (per ton) compared to higher biochar rates.Overall, this study found that applying biochar once at a rate of 22Mg ha−1 allows for the sustainable annualremoval of 50% of corn residue for 32 years, is profitable for farmers even with minimal impact on grain yield,and beneficial to society through reduced nitrate leaching and increased soil organic carbon levels.

1. Introduction

The US Midwest is one of the largest and most productive agri-cultural regions globally; producing nearly one-third of the world’s corn(Zea mays L.) and soybean (Glycine max L.) crops (FAOSTAT, 2015). InIowa specifically, corn for grain is planted on nearly 5.7 million hec-tares and soybeans on more than 3.6 million hectares (USDA NASS,2016), covering roughly 80% of the landscape (Newton and Kuethe,2015). As the number of hectares devoted to growing corn and soybeanhas increased in response to rising feed, fuel, and food demand, crop-ping system diversity has largely been eliminated (Liebman et al.,2013). This has potentially negative consequences as more diversesystems that include long-term rotations are well known to have

widespread benefits for soil and environmental quality (Giller et al.,1997; Karlen et al., 2006; Russell et al., 2006; Davis et al., 2012; Lal,2015; Aller et al., 2017).

Additionally, the shift to finding alternative energy sources hasmeant corn residue is becoming a high value product. A US Departmentof Energy report in 2005 estimated biomass recovered from crop re-sidues could significantly contribute to US energy production (Perlacket al., 2005). Specifically, Iowa farms produce the highest concentrationof corn residue, independent of management scenario, for bioenergyproduction nationwide (Graham et al., 2007). However, residues left inthe field have many positive impacts on soil and ecosystem functioningincluding: maintaining soil organic matter (SOM) levels, reducing soilerosion by wind and water, and increasing soil microbial activity and

https://doi.org/10.1016/j.fcr.2018.07.012Received 8 November 2017; Received in revised form 23 July 2018; Accepted 29 July 2018

Abbreviations: APSIM, Agricultural Production Systems sIMulator; CC, continuous corn; CS, corn-soybean; C, carbon; GDD, growing degree days; ME, modelingefficiency; NO3, nitrate; RMSE, root mean square error; RRMSE, relative root mean square error; SOC, soil organic carbon⁎ Corresponding author.E-mail address: sarchont@iastate.edu (S.V. Archontoulis).

Field Crops Research 227 (2018) 30–40

0378-4290/ © 2018 Published by Elsevier B.V.

T

sequestering carbon (Wilhelm et al., 2007; Lal and Pimentel, 2007;Blanco-Canqui and Lal, 2009; Laird and Chang, 2013). Soil quality andlong-term agricultural productivity will be negatively impacted if cropresidues are continuously removed in an unsustainable way (Wilhelmet al., 2004; Laird and Chang, 2013). Some of the potential negativeeffects of residue removal, however, could be offset through the in-corporation of soil amendments. Biochar, the charcoal like co-productof biomass pyrolysis, is a soil amendment that can enhance soil qualityand agricultural productivity while simultaneously sequestering carbon(Laird et al., 2009; Lehmann and Joseph, 2015).

The wide range of options for residue management and the com-plexity, uncertainty, and long time-scales underpinning the outcome ofthese options makes it difficult to develop sound residue managementstrategies for optimizing the agronomic and environmental benefitsbased purely on experimental studies. Agro-ecosystem models that in-tegrate multiple factors influencing residue C and N dynamics over bothshort- and long-time scales are capable of predicting and explainingcomplex interactions between soil-crop processes and hence can facil-itate a systems level of understanding (Thorburn et al., 2005). TheAgricultural Production Systems sIMulator (APSIM, Keating et al.,2003; Holzworth et al., 2014) is such a model. The APSIM model hasbeen successfully used to simulate various aspects of cropping systemsin Iowa (Malone et al., 2007; Archontoulis et al., 2014a, b; Basche et al.,2016; Dietzel et al., 2016; Martinez-Feria et al., 2016; Togliatti et al.,2017).

Recently, a biochar model was developed for the APSIM platform(Archontoulis et al., 2016) to simulate the effects of biochar amend-ments and to enhance understanding of biochar’s long-term impacts onagro-ecosystem performance. Initial biochar model testing using fieldexperimental data that contained various biochar application rates(Rogovska et al., 2014) showed very good agreement between simu-lations and experimental observations (Archontoulis et al., 2016).However, additional testing of the biochar model is needed before it canbe used for scenario analysis and agricultural assessments to supportdecision-making.

The specific objectives of this study were to 1) to calibrate theAPSIM biochar model for predicting the impact of biochar amendmentson corn grain yields and soil organic carbon (SOC) levels for continuouscorn and corn-soybean cropping systems in central Iowa, and 2) to usethe calibrated biochar model to evaluate the long-term (32-years) ef-fects of different biochar types and application rates on grain andbiomass productivity and environmental performance of corn and corn-soybean cropping systems under different management scenarios thatinclude varying nitrogen application and residue removal rates. Wehypothesized that biochar applications can offset the negative effects ofresidue harvesting on SOC levels, while not impacting corn yields.

2. Materials and methods

2.1. Dataset and measurements

The experimental dataset used to calibrate the model was from acrop rotation experiment conducted at Iowa State University’s SorensonResearch Farm in Boone County, IA from 2006–2016. A continuousrecord (11 years) of end-of-season grain yields, detailed yearly man-agement records and periodic soil organic carbon data were available(for details see Aller et al., 2017).

The overall field experiment was arranged in a completely rando-mized block design with split plots that included five different croprotations: continuous corn, corn-soybean, corn-soybean-triticale/soy-bean-corn-soybean-triticale/soybean, corn-corn-corn/switchgrass-switchgrass-switchgrass-switchgrass, and continuous switchgrass.Rotations were in a six-year cycle with all phases of each rotationpresent every year in four replicate blocks and with complete removalof aboveground biomass from all plots containing corn, switchgrass,and triticale every year. Biochar was applied on one-half of the 192 split

plots over four consecutive years (2012, 2013, 2014, and 2015) fol-lowing the corn-phase of each rotation. A slow-pyrolysis hardwoodbiochar (number 10 granular charcoal, Royal Oak Enterprises, LLC.,Roswell, GA) was applied at a rate of 22.4 t ha−1 and was incorporatedto a depth of 20 cm with a single pass of a rotary tiller.

Two of the five cropping systems are considered in this study, thecontinuous corn (CC) and corn-soybean (CS) rotation. We chose thesetwo systems because together they comprise 93% of the cropping sys-tems in Iowa (USDA NASS, 2016) and the corn and soybean models inAPSIM have been well tested and validated in this environment(Archontoulis et al., 2014a, b; Basche et al., 2016; Dietzel et al., 2016;Puntel et al., 2016; Martinez-Feria et al., 2016; Togliatti et al., 2017).Biochar applications made in 2012, 2013, and 2014 for the CC systemand in 2012 and 2013 for the CS rotation system were included in themodel calibration effort.

Daily weather information to drive APSIM simulations were re-trieved from the Iowa Environmental Mesonet (2017), which reportsdata from a weather station approximately 2 km north of the field site.The baseline soil profile used in the model was the Clarion soil series(Fine-loamy, mixed, superactive, mesic Typic Hapludolls) as this is thedominant soil series at the site. Soil parameters obtained from the soilsurvey database (USDA NRCS, 2017) were adjusted based on analysis ofsamples collected from the field site (Aller et al., 2017). Parametersadjusted included the drained upper limit (DUL), lower limit (LL), sa-turation point (SAT), bulk density (BD), and soil organic carbon (SOC)content. We used the average of SOC measurements for all 0–30 cmsamples collected from no-biochar plots between 2006–2016. The SOCvalues for the sub-soil (30–220 cm) and DUL, LL, and BD values for theentire soil profile were based on analysis of deep cores collected in 2016(unpublished data). The soil profile parameter values can be found inthe supplementary material (Tables S1 and S2).

2.2. The APSIM model - initialization and calibration

The APSIM model version 7.9 was used in this study along with thefollowing component models: corn and soybean crop models (Wanget al., 2002; Keating et al., 2003), Soil N (soil N and C cycling modelwith the default soil temperature model; Probert et al., 1998), SoilWat(a tipping bucket soil water model; Probert et al., 1998); SURFACEOM(residue model; Probert et al., 1998; Thorburn et al., 2001, 2005); andthe biochar model (Archontoulis et al., 2016).

Management practices specified for each crop within APSIM in-cluded rules for crop rotations, tillage, biochar tillage (to incorporatethe biochar immediately after application in the plots and years ap-plicable), sowing, fertilizing, and harvesting. Nitrogen fertilizer wasapplied at a rate of 190 kg ha−1 annually for continuous corn and135 kg ha−1 in corn years only for the corn-soybean rotation. A 110-day relative maturity corn hybrid and a maturity group 2 soybeanvariety were used.

To initiate the SOC, soil N pools across the profile and other modelparameters (e.g. residue amount, carbon-to-nitrogen ratio), we ran si-mulations for 6-years prior to the start of the analyzed simulations,similar to Dietzel et al. (2016). The initial 6-years of simulated datawere excluded from model analysis and corn yield and SOC values wereoutput. Soybean yields were simulated but not analyzed in this study.

During model calibration we replaced annual seeding rates (targets)with plant stand counts measured in the field. This additional man-agement information substantially improved agreement between mea-sured and predicted corn yields because the model assumes that thenumber of plants sown equals the number harvested. It does not ac-count for plant losses between sowing and harvest due to managementpractices (e.g., poor germination) or environmental factors (e.g., frost).The next step was to adjust the maize cultivar specific parameters as weinitially used a 110-day relative maturity hybrid from the APSIM da-tabase. We made the following changes: 1) time from emergence to theend of vegetative stage was increased from 214 to 250 growing degree

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

31

days (GDD) to achieve a final leaf number of about 18 leaves that iscommon in this environment, 2) time from flowering to maturity wasdecreased from 885 to 820 to achieve earlier and more realistic ma-turity dates GDD, 3) time from flowering to the start of grain fill wasincreased from 150 to 170 GDD similar to Puntel et al. (2016), and 4)time from maturity to ripening was increased from 1 to 180 GDD toallow for the kernel dry-down period since we ran the model sequen-tially to better represent the system (see also Table S4). These changeswere guided by previous modeling exercises in this region(Archontoulis et al., 2014a, 2016; Puntel et al., 2016; Togliatti et al.,2017). The same corn cultivar was used in every year of the simulation.In the field, different corn hybrids were used but all of them had thesame relative maturity. Also, no changes were made to the parametervalues in the soybean, soil water, and nitrogen models.

The simulation process was continuous, without annual reset. Atotal of 96 simulations were performed for the CC and CS systems,which included 48 plots with biochar and 48 plots without biochar. Weran the model on a plot by plot basis because biochar was applied todifferent subplots over a four year period, as described previously (fordetails see Aller et al., 2017).

The biochar model was included in simulations for plots wherebiochar was applied. Within the biochar model we adjusted the fol-lowing management information: timing of application, amount ofbiochar applied, depth of biochar incorporation; and the followingmeasured soil and biochar parameter information: sand and clay con-tent, biochar carbon fraction, biochar labile fraction, and biochar C:Nratio (Aller et al., 2017). All other biochar parameter values were de-fault values derived from Archontoulis et al. (2016) during biocharmodel development, parameterization, and testing. All biochar para-meter values are presented in the supplementary material (Table S3). Abrief description of the biochar model follows.

The biochar model adds a new biochar C pool to the APSIM soilmodel that contains both labile and recalcitrant components. The dailyrate of biochar decomposition is simulated by a double exponentialdecay function that accounts for water, temperature and N limitations.Parameters in this function are biochar properties such as biochar ap-plication amount, labile/recalcitrant biochar fraction, carbon fractionin the biochar, and mean residence time of each pool. The decomposedbiochar C is partitioned into C emissions to the atmosphere and Ctransfers to the microbial biomass (BIOM) and humic (HUM) pools. Inaddition to the changes in the size of the SOC pools (BIOM and HUM)imposed by biochar decomposition, additional functionality has beenadded to the biochar model to account for the positive and negativepriming effects that were not utilized in this study. The biochar modelsimulates soil N mineralization or immobilization of inorganic N basedon the biochar C:N ratio, which is an input to the model. Furthermore,the biochar model simulates changes in soil cation exchange capacityand soil pH dynamics that affect soil N dynamics. Adsorption anddesorption of NH4 due to biochar addition are included in the modeland default values were used in this study. In terms of biochar effects onsoil hydrology and aeration, the biochar model dynamically simulateschanges in soil bulk density (and porosity) due to biochar C addition orSOC loss as well as changes in field capacity. The emergent changes insoil water and nitrogen affect crop water and nitrogen uptake by thecrop model and therefore crop yield simulations. For more details onthe biochar model refer to Archontoulis et al. (2016).

2.3. Model performance evaluation

Agreement between simulated and measured values was assessedusing the root mean square error (RMSE), relative root mean squareerror (RRMSE), and modeling efficiency (ME) statistics.

=∑ −

−RMSES O

N( )i

N1

2

(1)

=RRMSE RMSEO

* 100avg (2)

= −∑ −

∑ −ME

S OO O

1( )

( )avg

2

2 (3)

where N is the total number of observations, S is the simulated value, Ois the measured value, and Oavg is the average of the measured values.Lower values of RMSE and RRMSE indicate better model fit, as theyprovide the absolute and relative error between the simulated andmeasured values, respectively. Modeling efficiency (scale< 0–1),which describes the average model performance across all observations(Archontoulis and Miguez, 2015), was determined as an indicator ofoverall goodness of fit. Modeling efficiency is useful when comparingmodel performance between different datasets because it normalizesthe data (Wallach, 2006).

2.4. Model application

The calibrated model was used to investigate the long-term (32-years) impacts of biochar application on corn yields, SOC levels, andNO3 leaching under various management conditions for CC and CSsystems. Nitrate leaching was defined as the amount of nitrate leachedbelow 1.5m, the typical root zone for corn and soybean crops in Iowa(Ordóñez et al., 2018). Specifically, the factorial simulation experimentevaluated three different nitrogen application rates (75, 150, and225 kg N ha−1), three residue removal rates (0, 50, and 90% removal),four biochar application rates (0, 22, 45, 90Mg ha−1), two biochartypes (hardwood, slow pyrolysis with a high C:N ratio and corn stover,fast-pyrolysis with a low C:N ratio), and two crop sequences (con-tinuous corn and corn-soybean rotation). This resulted in3×3×4×2×2=144 simulations. Simulations were run sequen-tially over a 37-year period (1980–2016), with the first five years(1980–1984) prior to biochar application excluded from the analysis.The sequential simulation resulted in roughly the same starting SOCvalue as in 2006 (calibration). Historical weather data were used tomake inferences regarding future changes in SOC and other soil vari-ables. Biochar was applied once in year 1985. Overall, 4608 data (yearx management combinations) on corn yield (kg ha−1), SOC (%), andNO3-N leaching were used for the scenario study. From these data wecalculated the average percent difference between biochar and no-biochar treatments on corn yield and NO3 leaching values over the 32-year period. For SOC we calculated the percent difference between SOClevels in 1985 (pre-biochar application) and 2016 (32-years after bio-char application).

2.5. Economic analysis

A cost-benefit analysis using a partial budget approach was em-ployed to evaluate the net present value of expected future private andpublic net economic benefits of the various biochar application andresidue removal rate scenarios over the 32-year simulation periodstarting from 2017. We assumed that the effects of biochar on cropyields, SOC levels, and NO3 leaching for 1985–2016 (percentagechange) would apply for the next 32 years, and evaluated the futurebenefits from 2017 to 2048 following biochar application in 2017. Thisallowed for greater insight into whether or not a producer should applybiochar in the present day and at what rate it should be applied. Weused the agronomic and environmental outcomes of the 144 simula-tions and assigned literature-driven values to determine the economiccosts and benefits relative to the respective baseline scenario for eachcropping system (Calkins and Dipietre, 1983). The net present value ofbenefits were the summation of discounted annual benefits over the 32-year simulation using a commonly employed 5% discount rate(Nordhaus, 2017). We focused the analysis on the prevailing productionbaseline scenarios, which were represented by the 2.75 ha experimental

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

32

field site that had no biochar, no residue removal, and an N applicationrate of 225 kg N ha−1 every year in the CC system and 150 kg N ha−1 incorn years of the CS rotation system. This entailed varying the C:N ratioof the biochar (high or low), the biochar application rate, applicationrate of N fertilizer, and rate of residue removal in both the CC and CSsystems. From the analysis we not only derived the net present value ofthe private and public benefits and the corresponding equivalent an-nualized cost, but also determined the private and public breakevencost, which represents the maximum per-ton price of biochar the pro-ducer could afford to pay.

2.5.1. Net private benefitsThe net private benefits were determined from increased cash crop

yields relative to the baseline plus net-revenues due to the sale of re-sidue. As previously discussed, we assumed the same percentage changein crop yields above the USDA long-term projected trend line yieldsfrom 2017 to 2048 (USDA OCE, 2017) as revealed from the APSIMsimulations for 1985–2016. Annual long-term projected nominal cornand soybean prices and a time-constant stover price were used (USDAOCE, 2017; Edwards, 2014). The $13 per ton net-revenue from cornstover removal was determined from the price of stover as feed minusthe farmers harvest and transportation costs. We assumed that theprivate costs were labor and machinery expenses from a one-timebiochar application, plus the additional cost of nitrogen beyond the rateapplied in the respective baseline scenarios for each system. We alsoassumed that all other corn and soybean production costs were iden-tical across the 2.75 ha field in all scenarios. The net present value ofthe net private benefits acquired annually and summed over the 32-yearperiod were calculated using a 5% discount rate by the followingequation:

Net private benefits= value from yield increase above future trendline yield+ value from sale of stover – biochar application cost

To calculate the private breakeven cost for a farmer applying bio-char, the net private benefits were divided by the quantity of biocharapplied in each scenario. The final value represents the maximum pricea farmer would pay to not lose any revenue by applying biochar, re-lative to the baseline scenario.

2.5.2. Net public benefitsThe net public benefits were determined from the reduction in NO3-

N leaching and the increase in SOC, which was given a value based onthe corresponding reduction in CO2 emissions from each croppingsystem. We quantified the value of decreased NO3-N leached relative tothe baseline scenario annually using an equilibrium price implied fromwater quality trading, $3.13 for each pound nitrate saved (Ribaudoet al., 2014), and summed that over the 32-year simulation period foreach scenario using a 5% discount rate. The reduction in CO2 emittedfor each scenario was calculated annually relative to the baseline sce-nario, based on average SOC measures across all 0–30 cm samples, as-suming that for every ton increase in SOC there is one less ton of CO2-Cemitted in the future. We only consider changes in SOC (in-field) and donot attempt to quantify system level CO2 changes that may be asso-ciated with biochar transportation and application. The economic valueof future CO2 emission reductions was converted into a dollar amountusing a $36 per ton social cost of carbon and a 3% discount rate(Nordhaus, 2017). Considering the global nature of greenhouse gases, alocalized carbon sequestration measure was also calculated by scalingthe global value of reduced carbon emissions by Iowa’s share of theglobal population (Hoque and Kling, 2016) to estimate local level CO2

emissions. Annual net public benefits were calculated using the fol-lowing equation:

Net public benefits= value of reduced NO3-N leaching+ reduced CO2

emissions through improvements in SOC (in-field)

2.5.3. Total net benefitsThe net present value of the total net benefits were calculated as the

sum of the net present value of the net private benefit and net publicbenefit. We also calculated an equivalent annualized cost assuming a5% discount rate and a 32-year project life. Equivalent annualized costis equal to the net present value divided by the present value annuityfactor. The public breakeven cost for each scenario was calculated asthe total net benefits divided by the quantity of biochar applied in eachscenario. This value represents the maximum price at which it wouldbenefit all of society, including the farmer, if the farmer applied bio-char. Lastly, the potential benefits associated with changes in landvalue were not included in the analysis.

3. Results

3.1. Model performance for corn yields

Across treatments and years, the calibrated model performed well insimulating annual corn yields in both the CC and CS systems (Table 1;Fig. S1). In the CC system, RMSE decreased from 3168 to 1482 kg ha−1,the RRMSE from 34.6 to 16.2%, and the ME increased from -1.03 to0.56 before and after model calibration, respectively. Similarly, in theCS rotation, the RMSE decreased from 2415 to 1526 kg ha−1, theRRMSE from 23.5 to 14.7%, and the ME increased from −0.3 to 0.49from the uncalibrated to calibrated model, respectively.

In the no-biochar plots, year-to-year variability in corn yields wassimulated well by the model (data not shown). In only two out of the 11years (viz. 2006 and 2015) the model underestimated measured yieldsfor both cropping systems. The model performed slightly better in si-mulated corn yields in the CS rotation (RRMSE=15.0%, ME=0.60)than in the CC system (RRMSE=15.9%, ME=0.57).

In the biochar plots, model results showed satisfactory agreementbetween measured and simulated yields following different years ofbiochar application (Fig. 1). For the CC system, both measured andsimulated grain yields increased following biochar application for thethree different years of biochar applications (2012, 2013, and 2014, seeFig. 1). A similar increase was observed for the control plots (Fig. 1),however, the yield increase in the biochar plots, as determined from themeasured data, was 837, 943, and 760 kg ha−1, in 2012, 2013, and2014, compared to the no-biochar plots, respectively (data not shown).But this observed yield increase in the biochar plots is confounded bytillage and thus may not be a true biochar effect (the experimental sitewas under strict no-till management except for a single roto tillageoperation to incorporate the biochar into their respective subplots). Inthe 2012 and 2013 biochar application years the model showed lowerME compared to their respective controls. Model performance wasequal when the 2014 biochar and no-biochar plots were evaluated(Fig. 1).

For the CS rotation the model showed similar results as for the CCsystem (Fig. S2). Average corn yields were largely underpredicted inonly 2006 and 2016 for the 2012 biochar applied plots and in 2013 and2015 for the 2013 biochar applied plots. The relative model error andthe ME were both lower in the 2012 biochar plots compared to the no-biochar controls. While the 2013 biochar plots showed greater relative

Table 1Model performance before and after calibration in simulating corn yields incontinuous corn (CC) and corn-soybean (CS) systems (N=48 plots x 11years= 528 per cropping system).

Before calibration After calibration

Cropping system RMSE RRMSE ME RMSE RRMSE MEkg ha−1 % – kg ha−1 % –

Continuous corn 3168 34.6 −1.03 1482 16.2 0.56Corn-soybean 2415 23.5 −0.30 1516 14.7 0.49

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

33

model error and lower ME than the control plots. Indicating that overallmodel performance is slightly better at estimating corn yields in thecontrol plots then the biochar plots for the CS rotation system (Fig. S2).

3.2. Model performance for soil organic carbon

The model estimated well SOC levels in both biochar and no-biocharplots for both the CC and CS systems across the 11 years analyzed(Figs. 2 and S3). Across both biochar and no-biochar plots the model ingeneral overestimated SOC levels in 2012 and 2013. While in the bio-char plots, regardless of year of biochar application, the model under-estimated SOC levels in 2016 only. In both cropping systems modelpredictions of SOC varied little, while measured data were more vari-able (Figs. 2 and S3). This is attributed to sampling variability fromfield measurements.

Relative model error was higher but overall model performance wasbetter in the biochar plots compared to the control plots for both the CCand the CS systems (Figs. 2 and S3). Further, for both cropping systems,model agreement between the predicted and measured SOC data wasbest for the 2014 biochar application year. Overall, both simulated andmeasured data showed similar trends over time for SOC levels.

3.3. Model scenario analysis

We used the calibrated model and historical weather data to simu-late the impact of two biochar types, four biochar application rates,three residue removal rates, and three nitrogen fertilizer rates on cornyields, nitrate leaching and SOC levels over a 32-year period. Overallthe changes in average corn yields attributable to biochar were smallover the 32-year simulation period and across all scenarios. The largestsimulated effect was a yield decline of 2.6% for the CC system withapplication of 90Mg ha−1 high C biochar, 0% residue removal, and afertilization rate of 75 kg N ha−1 (Fig. 3), which would put the corncrop under considerable N stress. The 2.6% yield loss equates to a total

of 6112 kg ha−1 (97 bu ac−1) or an average of 191 kg ha−1 yr−1 (3 buac−1 yr−1) over the 32-year simulation period.

Although the overall effect of biochar on corn yields was small, themodel indicated a complex cropping system by N fertilization rate bybiochar application rate by residue removal rate interaction. Simulatedcorn yields for the CC and CS cropping systems showed different trendsfor residue removal rates for the 75 and 150 kg N ha−1 fertilization ratescenarios (Fig. 3). For the CC system, the 75 kg N ha-1 fertilization ratewithout residue removal resulted in the largest yield decline followedby the 90% and then 50% residue removal rates, regardless of biocharapplication rate. At the same fertilization rate but for the CS rotationsystem, a residue removal rate of 50% had the largest yield declinefollowed by the 90% and then 0% residue removal rates, regardless ofbiochar application rate. At the 150 kg N ha−1 fertilization rate cornyields decreased for all biochar application and residue removal ratescompared to the no-biochar control for the CC system. Whereas in theCS rotation system corn yields increased at the 0% and 50% residueremoval rates and decreased at the 90% residue removal rate regardlessof biochar application rate (Fig. 3). These differences are attributedprimarily to biochar and crop residue effects on N immobilization/mineralization with a lesser effect due to soil water. Specifically, themodel predicted that biochar applications would decrease N availabilityto the crop while the labile C in biochar is being decomposed due to Nimmobilization. The effect of biochar treatments on corn yields waspositive at the 225 kg N ha−1 fertilization rate regardless of biocharapplication and residue removal rates (Fig. 3). At the 225 kg N ha−1

fertilization rate, N availability is no longer limiting to crop growth andhence the positive aspects of biochar on soil quality boost yields.

The high and low C biochar treatments had similar effects on si-mulated corn yields with only a decreased magnitude of change relativeto the control observed for the low C biochar compared to the high Cbiochar treatments (Fig. 3). When the low C biochar type was appliedthe greatest yield decline was less than 2% in the CC system scenario of75 kg N ha−1, 0% residue removal, and 90Mg ha−1 biochar (Fig. 3).

Fig. 1. Impact of biochar applied in 2012, 2013, or 2014 on cornyields for a CC cropping system across 11 years. Measured values(black diamonds) are the average of all plots with biochar (leftside) and no-biochar plots (right side) with error bars indicatingstandard deviations. Predicted yields (blue line) are the average ofall plots with biochar (left side) and no-biochar plots (right side)with shaded standard deviation bands. (For interpretation of thereferences to color in this figure legend, the reader is referred tothe web version of this article).

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

34

The difference is attributed to the lower C:N ratio of the low C biocharrelative to the high C:N ratio of high C biochar, which means less Nimmobilization during decomposition of the low C biochar relative tothe high C biochar.

The simulations showed similar patterns for NO3 leaching as wereobserved for corn yields (Fig. 4). Simulation results indicate that ap-plication of both the high and low C biochars decreased NO3 leachingthrough the root zone (NO3 leached below 1.5 m) for all N fertilizationand residue removal rates and for both cropping systems (Fig. 4). Thepercent reduction in NO3 leaching attributed to the biochar treatmentsrelative to the no-biochar controls increased with the N fertilization rateand the amount of biochar applied. The maximum impact on NO3

leaching for both cropping systems was a reduction of about 10% whenthe low C biochar was applied and a nearly 20% reduction when thehigh C biochar was applied (Fig. 4). This difference is attributed to theC:N ratio of the biochar and the availability of biochar C for N im-mobilization.

Residue removal decreased NO3 leaching at the 75 and 150 kg Nha−1 N fertilization rates, but increased NO3 leaching at the 225 kg Nha−1 fertilization rate (Fig. 4). These differences are attributed to Nbeing limiting to the crop at the low and medium fertilization rates, soas more residue is removed the availability of N increases, enhancingroot growth, and reducing NO3 leaching. Whereas, at the high fertili-zation rate, N is not limiting plant growth, thus as residue removal

Fig. 2. Percent SOC across 11 years as impacted by biochar ap-plied in 2012, 2013, or 2014 for the CC system. Measured values(black diamonds) are the average of all plots with biochar (leftside) and no-biochar plots (right side) with error bars showingstandard deviations. Predicted yields (blue line) are the average ofall plots with biochar (left side) and no-biochar plots (right side)with shaded standard deviation bands. (For interpretation of thereferences to color in this figure legend, the reader is referred tothe web version of this article).

Fig. 3. Differences in average corn yields (kg ha−1) betweenbiochar and no-biochar control treatments for the 32 year simu-lation (1985–2016) in CC and CS rotation systems under differentN fertilization and residue management scenarios and when twotypes of biochar (a low C content and low C:N ratio biochar and ahigh C content and C:N ratio biochar) were applied in 1985.

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

35

increases, N immobilization decreases, and NO3 leaching increases.The 32-year simulation indicates that biochar applications, biochar

type, and residue removal rate have large impacts on SOC levels whilecropping system, and N fertilization rate have relatively small effects onSOC levels (Fig. 5). There was a direct relationship between biocharapplication rate and the increase in SOC content at the end of the 32-year simulation. For the no-biochar control scenarios, SOC levels in-creased over time for the 0% residue removal treatment but decreasedover time for the 50% and 90% residue removal treatments. For thescenarios with 22, 45, 90Mg ha−1 biochar application rates, as percentresidue removal increased, the percent increase in SOC levels decreasedbut were higher than the initial SOC levels because of the biochar C.When the low C biochar was applied, the simulations indicate a smallerincrease in SOC levels over 32 years relative to the simulations usingthe high C biochar. This observation is attributed to the lower C contentand C:N ratio of the low C biochar.

3.4. Economic analysis

The cost-benefit analysis indicated that sufficient N is needed toensure a positive net benefit from biochar applications. If the crop isalready short on N, a biochar application will exacerbate the N shortageresulting in a significant yield drag; lower yield relative to prevailingpractices (Tables S5, S6, and S7). The assessment of private benefitsindicated that as residue removal rates increase both the corn croprevenue and stover sale value increase (Tables S5, S6, and S7). Whenthe N rate is low the corn crop revenue is negative relative to thebaseline but as stover removal rate increases the value becomes lessnegative, while at higher rates of N, the corn crop revenue is positiveand increases with increasing biochar application rate (Tables S5, S6,and S7). Also, the net private benefits are boosted significantly as re-sidue removal rate increases due to a higher value from the sale of cornstover, which at the same time diminishes the value of crop yield en-hancements. While some scenarios without biochar reveal a decline inrevenue coming from yield as a result of residue removal, this is

Fig. 4. Average difference in NO3 leaching for the 32 year simulation (1985–2016) between biochar and no-biochar control treatments for the CC and CS rotationsystems under different N fertilization and residue management scenarios and when low/high C content and C:N ratio biochars were applied in 1985.

Fig. 5. Average change in total soil organic carbon levels for the 32 year simulation (1985–2016) in CC and CS rotation systems under different N fertilization andresidue management scenarios and when low/high C content and C:N ratio biochars were applied in 1985.

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

36

outweighed by the private revenue gained from stover sales. However,we do not consider potential future crop yield losses, especially in thecontinuous corn production system, due to the depletion of SOC levelsthat result from the long-term continuous removal of residue withoutreplenishment from biochar.

Our results reveal that the public benefits of biochar application,when coupled with the ability to harvest more crop residue, sig-nificantly outweigh the private benefits enjoyed by the farmer (Figs. 6and S4). Importantly, the global public benefits are overwhelminglydriven by the benefits from reduced CO2 emissions, especially for sce-narios with higher biochar application rates and higher stover removalrates. In fact, at 50 or 90% residue removal rates, the public benefitsfrom the reduction in NO3 leaching are substantially lower than thebenefits that result from avoided CO2 emissions (Tables S5, S6, and S7).Moreover, the public benefits of reduced nutrient runoff and CO2

emissions are often external to a farmer’s decision to apply biochar.This is unless farmers are recognized and monetized in the form of cost-share funding or market price premiums by preserving long-term soilhealth via biochar applications. In this study, the private benefits alonewere not sufficient to guarantee a positive net private return from ap-plying biochar. Furthermore, the marginal public benefit of reducingCO2 emissions due to additional biochar application is stagnant orlower; in other words, a higher biochar application rate actually causesa lower maximum price a farmer could afford to pay for biochar. Incontrast, a higher biochar application rate leads to a greater reductionin NO3 leaching, while residue removal reduces NO3 leaching when N isinsufficient for crop needs but increases NO3 leaching when N is suf-ficient to meet crop needs.

The breakeven price, both public and private, for farmers applyingbiochar increases as residue removal rates increase and decreases asbiochar application rates increase for both cropping systems and highand low C and C:N ratio biochars. Lastly, results show that a high C andC:N ratio biochar results in greater benefits then a low C and C:N ratiobiochar across all scenarios (Figs. 6 and S4; Tables S5, S6, and S7).

4. Discussion

This is the first study that evaluates the long-term (32-years) effectof a set of management practices (biochar application rates x biochartypes x N fertilization rates x residue removal practices) on agro-eco-system and economic performance for decision-making purposes. In theera of corn stover residue harvesting for bioenergy production it iscritical to determine the environmental and economic trade-offs ofvarious management strategies to ensure long-term system

sustainability. Our results suggest that a single application of 22Mgha−1 biochar can support the sustainability of harvesting> 50% cornresidue annually for 32 years. Here sustainability means that SOC levelsincreased and NO3 leaching decreased when biochar was applied re-lative to no-biochar controls.

4.1. Model performance in simulating corn yields and SOC

This study calibrated the APSIM biochar model for a new site incentral Iowa and used the calibrated model to evaluate various man-agement scenarios. Overall, the calibrated model with and withoutbiochar additions performed well in predicting year-to-year variabilityin corn yields and SOC levels (Figs. 1 and 2). The RRMSE of 15% forcorn yield is comparable with other simulation studies in this region(Malone et al., 2007; Puntel et al., 2016). We achieved this good per-formance with only two changes: cultivar specific parameters andprecise management information from year to year (plant densitycounts). The cultivar parameters were manually selected to reflect finalleaf number, silking dates and maturity dates that are common for thestudy environment based on past experience (Archontoulis et al.,2014a; Puntel et al., 2016). No attempts were made to further improvethe RRSME of 15% because according to He et al. (2017) targetingmodel calibration with even lower RRMSE values is pointless becausethe simulated error cannot be lower than the inherent error in themeasured data used for calibration.

The model showed generally good agreement with measured SOClevels for both plots with and without biochar (Fig. 2). The modelsresponse to biochar application was an immediate increase in SOC le-vels (Fig. 2). The biochar used during model performance evaluationwas a high carbon hardwood biochar (76% C) that also had a high C:Nratio of 232:1. Thus, the immediate increase in the predicted SOC va-lues was not surprising. Similarly, the more gradual increase observedin the measured data was expected because of environmental condi-tions and sampling variability. The rate of increase and the duration ofthe higher SOC levels simulated by the model will change if a lowcarbon biochar is used. This is why it was important to consider theimpacts of applying two different types of biochar in the scenarioanalysis. In this study, the priming coefficient in the biochar model wasset to zero, which is the baseline assumption that biochar neither sti-mulates degradation of biogenic SOM nor does it enhance the formationof new biogenic SOM. We used a value of zero for the priming coeffi-cient because the literature on the effects of biochar on SOM formation/mineralization is so far inconsistent. The fact that simulated and mea-sured SOC values followed the same trends suggests that the priming

Fig. 6. Net private and net global public benefits for the CC andthe CS rotation systems for different biochar application rates, 0%residue removal, and when a high C biochar (blue bars) and low Cbiochar (red bars) is applied. Values are relative to the baselinescenarios of no biochar, 0% residue removal, and 225 kg N ha−1

and 150 kg N ha−1 for the CC and the CS rotation systems, re-spectively. (For interpretation of the references to color in thisfigure legend, the reader is referred to the web version of thisarticle).

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

37

effect of biochar was negligible in our study, but more research onbiochars effect on priming is needed.

4.2. Model application

The long-term APSIM simulations revealed a range of options fordecision makers to consider. The negligible effect of biochar applica-tions on corn yields and the observed increasing corn yields as N fer-tilization rate increased were expected for the high-quality soils incentral Iowa. By contrast, variable effects of biochar applications oncorn yields have been reported in other field studies (Major et al., 2010;Uzoma et al., 2011; Cornelissen et al., 2013; Rogovska et al., 2016;Laird et al., 2017). Whether the effect of biochar is positive or negativeis environment, climate, and soil type dependent and other scenariosare likely to indicate different results.

One limitation of the APSIM model is that it does not yet simulatethe effects of allelopathy on corn yield predictions. Allelopathic com-pounds are released from decomposing crop residues and can haveharmful carryover effects. Because of allelopathy, the presence of highresidue loads from the previous year can decrease early seedling growthresulting in decreased crop yields in continuous corn systems (Rogovskaet al., 2014, 2016). Removing some or all of the surface residue canimprove yields by reducing the release of allelopathic compounds. Atthe same time, in scenarios where residues are left on the field andbiochar is applied, corn yields have been found to increase from0.8–3.8Mg ha−1 depending on biochar application rate, due to theadsorption and deactivation of allelopathic compounds onto biocharsurfaces (Rogovska et al., 2014, 2016). In the scenarios evaluated herethe potential positive impacts on corn yields, in response to residueremoval and biochar applications, were not considered. The inclusionof algorithms that account for allelopathic effects and the potentialreduction in the harmful effects of allelopathic compounds due tobiochar additions should be considered as future APSIM developmentwork.

The observation that residue removal decreases NO3 leaching forthe two lower N fertilization rates (75 and 150 kg N ha−1) but increasesNO3 leaching for the higher fertilization rate (225 kg N ha−1) (Fig. 4) isattributed to the combined effects of the biochars C:N ratio and theamount of residue remaining as well as residue impacts on soil coverand soil water evaporation. When a biochar with a high C:N ratio isapplied and there is greater residue left on the field, top soil N im-mobilization limits NO3 leaching (Fig. 4). When N fertilizer is alreadylimiting, in the case of the 75 and 150 kg N ha-1 rate scenarios, andmore residue is removed, a smaller fraction of N is immobilized. Thislower rate of N immobilization means more N is available to the plantstimulating root growth, better root growth improves the efficiency ofplant N uptake and reduces NO3 leaching. However, when N is presentin excess (225 kg N ha−1 rate scenario), root growth is not limited by N,and hence there is no improvement in N uptake efficiency, howeverhigher levels of residue increase N immobilization and thus NO3

leaching is reduced with decrease residue removal. When biochar has alow C content and a low C:N ratio, the effects of a biochar applicationon N immobilization are significantly lower (Fig. 4). Therefore, biocharquality also has a big effect on N availability and potentially cropgrowth and yield. Lastly, complete validation of our NO3 leaching si-mulation results is not possible but is supported by multiple on-goingAPSIM testing efforts in the Midwest (Martinez-Feria et al., 2016, 2018;Puntel et al., 2016, 2018; Basche et al., 2016; Dietzel et al., 2016;Togliatti et al., 2017). Observational data, from nearby sites, werefurther used to validate critical model parameters such as plant rootingdepth (Ordóñez et al., 2018), temporal soybean N-fixation, soil water,and nitrate dynamics (unpublished data).

Overall, the results of these scenario simulations highlight the im-portance of evaluating biochar quality. Different biochars impactagricultural systems and the environment differently and thus need tobe produced and applied for specific end-uses (Spokas et al., 2012;

Novak and Busscher, 2012; Ippolito et al., 2012).

4.3. Economic analysis

The cost-benefits analysis assumed that if farmers could purchasebiochar for the breakeven price in a given scenario they could afford topay for biochar, hence the breakeven price is a crude measure of thefarmer’s maximum willingness to pay to apply biochar. Consideringprivate benefits alone, under current prevailing management practicesthat include no residue removal, it is difficult to justify the procurementand application of biochar. For example, assuming a 32-year projectlife, the net present value of net private benefits for a 22Mg ha−1

biochar application would translate into an equivalent annual cost of a$5 loss per hectare for a farmer with no residue removal, but a $39 gainper hectare with 50% residue removal. Thus, in cases of residue re-moval the application of biochar potentially becomes economicallyviable, assuming a sufficiently low biochar price, because the biocharmakes the long-term harvesting of residue sustainable. Without biocharapplication, continuous residue harvesting will result in the long-termdegradation of soil quality and ultimately declining yields.Furthermore, the farmer could be compensated via cost-shares or en-vironmental trading schemes for the public benefits accrued from ap-plying biochar (decreased NO3 leaching and C sequestered/decreasedCO2 emitted). In the 36 difference scenarios evaluated here we showedthat a lower biochar application rate, 22Mg ha−1, tends to yield higherper-ton benefits and a higher per-ton breakeven price the farmer wouldbe able to pay for biochar. For example, a farmer can afford to pay $28per ton for biochar when applying biochar at a rate of 22Mg ha−1,based on the private gains resulting from higher yields and the ability tosustainably harvest 50% of the residue.

Across all scenarios the application of biochar appeared to be morebeneficial in the CC system compared to the CS rotation system, whichmay be linked to a lower N application rate in the CS rotation systemdue to soybeans ability to fix nitrogen during the growing season, andthus less NO3 leached from the system. Under prevailing nitrogen ap-plication rates in central Iowa a yield drag is unlikely with biocharapplications, however, a higher initial nitrogen application rate im-mediately after biochar application is recommended.

The private benefits were predominantly driven by yield gains whenN was sufficient, which tended to be minimal compared to the potentialpublic benefits that would result from increasing SOC levels and re-ducing NO3 leaching rates. However, the private benefits alone werenot sufficient to guarantee a positive net private return from biocharapplication, as implied by the negative per-ton price if only privateyield benefits were considered and there was no corn stover removal.More research is needed to quantify the yield benefits following biocharapplication in regions with poorer soil quality, such as in U.S. southeast,to realize the greater private benefits. In contrast, the public benefits,while large, need to be monetized in the forms of agri-environmentalpayments to farmers via programs like cost shares to provide additionaleconomic incentives for greater biochar adoption. Further, while anincrease in biochar application rate, holding other things constant,would increase the total private and public benefits, the maximum pricea farmer could pay for the biochar may not necessarily increase. Lastly,the greater the residue removal rate the more cost-effective the scenariowas because there is a high value associated with selling the residue andgreater reduction in NO3 leaching rates. Assuming a sufficiently lowbiochar price, farmers could apply biochar to mitigate the negativeyield impacts resulting from sustained residue removal, thus potentiallyachieving greater benefits for both farmers and society. However, thepotential impact of greater residue harvesting on soil erosion rates wasnot evaluated in this study, but is an important factor to consider whenassessing total system sustainability. Wilhelm et al. (2004) reportedthat maintaining SOC levels, not erosion, limits the amount of residuethat can be sustainably harvested. Hence, with biochar applicationssustaining SOC levels, sustainable rates of residue harvesting for

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

38

bioenergy production could be increased.

5. Conclusions

The findings from this study demonstrated that over a 32-yearperiod biochar applications can eliminate negative effects of residueharvesting on soil quality while at the same time reducing nitrateleaching, increasing soil organic carbon, and not impacting corn yields.Specifically, at least 50% of the residue can be removed annually for32 years if a one-time application of 22Mg ha−1 biochar is made. Thesimulations revealed that corn yields are most affected by the amount ofN applied and not by the addition of biochar or residue removed. Theopposite was observed for NO3 leaching, with increasing rates of bio-char leading to the greatest reduction in NO3 leaching. Biochar appli-cations are also effective for building soil organic carbon even underincreasing rates of residue removal. This finding in particular, couldhave important positive implications for the US bioenergy industrywhile improving the sustainability of our agricultural systems. The cost-benefit analysis revealed that the public benefits that result from ap-plying biochar coupled with the ability to harvest more residue, sig-nificantly outweighed the private benefits. Biochar applications are aneconomically viable option in Iowa when at least 50% of the residue isharvested for sale; which can be done in an environmentally sustainableway.

Acknowledgments

This project is supported in part by the Global Climate and EnergyProject, Stanford Subaward Agreement No. 60413992-112883-A, theNational Science Foundation under Grant Number EPS-1101284, theUSDA-NIFA Agriculture and Food Research Initiative CompetitiveGrant no. 2011-68005-30411, and the USDA-NIFA Hatch project No.IOW03814.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in theonline version, at https://doi.org/10.1016/j.fcr.2018.07.012.

References

Aller, D., Mazur, R., Moore, K., Hintz, R., Laird, D., Horton, R., 2017. Biochar age andcrop rotation impacts on soil quality. Soil Sci. Soc. Am. J. doi 81 (5), 1157–1167.https://doi.org/10.2136/sssaj2017.01.0010.

Archontoulis, S.V., Miguez, F.E., 2015. Nonlinear regression models and applications inagricultural research. Agron. J. 107, 786–798.

Archontoulis, S.V., Miguez, F.E., Moore, K.J., 2014a. Evaluating APSIM maize, soil water,soil nitrogen, manure, and soil temperature modules in the Midwestern United States.Agron. J. 106, 1025–1040.

Archontoulis, S.V., Miguez, F.E., Moore, K.J., 2014b. A methodology and an optimizationtool to calibrate phenology of short-day species included in the APSIM PLANT model:application to soybean. Environ. Model. Softw. 62, 465–477.

Archontoulis, S.V., Huber, I., Miguez, F.E., Thorburn, P.J., Rogovska, N., Laird, D.A.,2016. A model for mechanistic and system assessments of biochar effects on soils andcrops and trade-offs. GCB Bioenergy 8, 1028–1045.

Basche, A.D., Archontoulis, S.V., Kaspar, T.C., Jaynes, D.B., Parkin, T.B., Miguez, F.E.,2016. Simulating long-term impacts of cover crops and climate change on cropproduction and environmental outcomes in the Midwestern United States. Agric.Ecosyst. Environ. 218, 95–106.

Blanco-Canqui, H., Lal, R., 2009. Crop residue removal impacts on soil productivity andenvironmental quality. Crit. Rev. Plant Sci. 28, 139–163. https://doi.org/10.1080/07352680902776507.

Calkins, P.H., Dipietre, D.D., 1983. Farm Business Management: Successful Decisions in aChanging Environment. MacMillan and Company, New York.

Cornelissen, G., Martinsen, V., Shitumbanuma, V., Alling, V., Breedveld, G., Rutherford,D., Sparrevik, M., Hale, S., Obia, A., Mulder, J., 2013. Biochar effect on maize yieldand soil characteristics in five conservation farming sites in Zambia. Agronomy 3,256–274. https://doi.org/10.3390/agronomy3020256.

Davis, A., Hill, J., Chase, C., Johanns, A., Liebman, M., 2012. Increasing cropping systemdiversity balances productivity, profitability and environmental health. PLoS One 7,e47149. https://doi.org/10.1371/journal.pone.0047149.

Dietzel, R., Liebman, M., Ewing, R., Helmers, M., Horton, R., Jarchow, M., Archontoulis,S., 2016. How efficiently do corn-and soybean-based cropping systems use water? A

systems modeling analysis. Glob. Change Biol. 22, 666–681. https://doi.org/10.1111/gcb.13101.

Edwards, W., 2014. Estimating a Value for Corn Stover. Iowa State University Extensionand Outreach Accessed 1 July 2017. https://www.extension.iastate.edu/agdm/crops/html/a1-70.html.

Food and Agriculture Organization of the United Nations FAOSTAT, 2015. (FAO, Rome).Available at faostat.fao.org/.

Giller, K., Beare, M., Lavelle, P., Izac, A., Swift, M., 1997. Agricultural intensification, soilbiodiversity and agroecosystem function. Appl. Soil Ecol. 6, 3–16. https://doi.org/10.1016/S0929-1393(96)00149-7.

Graham, R.L., Nelson, R., Sheehan, J., Perlack, R.D., Wright, L.L., 2007. Current andpotential U.S. corn stover supplies. Agron. J. 99, 1–11. https://doi.org/10.2134/agronj2005.0222.

He, D., Wang, E., Wang, J., Robertson, M., 2017. Data requirement for effective cali-bration of process-based crop models. Agr. Forest Meteorol. 234–235, 136–148.https://doi.org/10.1016/j.agrformet.2016.12.015.

Holzworth, D.P., Huth, N.I., deVoil, P.G., Zurcher, E.J., Herrmann, N.I., McLean, G.,Chenu, K., van Oosterom, E.J., Snow, V.O., Murphy, C., Moore, A.D., Brown, H.,Whish, J.P.M., Verrall, S., Fainges, J., Bell, L.W., Peake, A.S., Poulton, P.L., Hochman,Z., Thorburn, P.J., Gaydon, D.S., Dalgliesh, N., Rodriguez, D., Cox, H., Chapman, S.,Doherty, A., Teixeira, E., Sharp, J., Cichota, R., Vogeler, I., Li, F.Y., Wang, E.,Hammer, G.L., Robertson, M.J., Dimes, J.P., Whitbread, A.M., Hunt, J., van Rees, H.,McClelland, T., Carberry, P.S., Hargreaves, J.N.G., MacLeod, N.D., McDonald, C.K.,Harsdorf, J., Wedgwood, S., Keating, B.A., 2014. APSIM - evolution towards a newgeneration of agricultural systems simulation. Environ. Model. Softw. 62, 327–350.https://doi.org/10.1016/j.envsoft.2014.07.009.

Hoque, M., Kling, C., 2016. Economic valuation of ecosystem benefits from conservationpractices targeted in Iowa nutrient reduction strategy 2013: a non market valuationapproach. Iowa State University CARD Working Papers. pp. 586.

Iowa Environmental Mesonet, 2017. Iowa Ag Climate Network. Iowa State University,Ames, IA Available at: http://mesonet.agron.iastate.edu/agclimate/ (Accessed 15February 2017).

Ippolito, J.A., Laird, D.A., Busscher, W.J., 2012. Environmental benefits of biochar. J.Environ. Qual. 41. https://doi.org/10.2134/jeq2012.0151.

Karlen, D.L., Hurley, E.G., Andrews, S.S., Cambardella, C.A., Meek, D.W., Duffy, M.D.,Mallarino, A.P., 2006. Crop rotation effects on soil quality at three northern corn/soybean belt locations. Agron. J. 98, 484–495. https://doi.org/10.2134/agronj2005.0098.

Keating, B.A., Carberry, P.S., Hammer, G.L., Probert, M.E., Robertson, M.J., Holzworth,D., Huth, N.I., Hargreaves, J.N.G., Meinke, H., Hochman, Z., McLean, G., Verburg, K.,Snow, V., Dimes, J.P., Silburn, M., Wang, E., Brown, S., Bristow, K.L., Asseng, S.,Chapman, S., McCown, R.L., Freebairn, D.M., Smith, C.J., 2003. An overview ofAPSIM, a model designed for farming systems simulation. Eur. J. Agron. 18, 267–288.

Laird, D.A., Chang, C.-W., 2013. Long-term impacts of residue harvesting on soil quality.Soil Till. Res. 134, 33–40. https://doi.org/10.1016/j.still.2013.07.001.

Laird, D.A., Brown, R.C., Amonette, J.E., Lehmann, J., 2009. Review of the pyrolysisplatform for coproducing bio-oil and biochar. Biofuels Bioprod. Bioref. 3, 547–562.https://doi.org/10.1002/bbb.169.

Laird, D.A., Novak, J.M., Collins, H.P., Ippolito, J.A., Karlen, D.L., Lentz, R.D., Sistani,K.R., Spokas, K., Van Pelt, R.S., 2017. Multi-year and multi-location soil quality andcrop biomass yield responses to hardwood fast pyrolysis biochar. Geoderma 289,46–53.

Lal, R., 2015. A system approach to conservation agriculture. J. Soil Water Conserv. 70,82A–88A. https://doi.org/10.2489/jswc.70.4.82A.

Lal, R., Pimentel, D., 2007. Biofuels from crop residues. Soil Till. Res. 93, 237–238.Lehmann, J., Joseph, S., 2015. Biochar for Environmental Management: Science,

Technology and Implementation. Taylor and Francis, London.Liebman, M., Helmers, M., Schulte, L., Chase, C., 2013. Using biodiversity to link agri-

cultural productivity with environmental quality: results from three field experimentsin Iowa. Renew. Agric. Food Syst. 28, 115–128. https://doi.org/10.1017/S1742170512000300.

Major, J., Rondon, M., Molina, D., Riha, S., Lehmann, J., 2010. Maize yield and nutritionduring 4 years after biochar application to a Colombian savanna Oxisol. Plant Soil333, 117–128. https://doi.org/10.1007/s11104-010-0327-0.

Malone, R., Huth, N., Carberry, P., Ma, L., Kaspar, T., Karlen, D., Meade, T., Kanwar, R.,Heilman, P., 2007. Evaluating and predicting agricultural management effects undertile drainage using modified APSIM. Geoderma 140, 310–322.

Martinez-Feria, R.A., Dietzel, R., Liebman, M., Helmers, M.J., Archontoulis, S.V., 2016.Rye cover crop effects on maize: a system-level analysis. Field Crops Res. 196,145–159.

Martinez-Feria, R.A., Castellano, M.J., Dietzel, R., Helmers, M.J., Liebman, M., Huber, I.,Archontoulis, S.V., 2018. Linking crop- and soil-based approaches to evaluate systemnitrogen-use efficiency and tradeoffs. Agric. Ecosyst. Environ. 256, 131–143.

Newton, J., Kuethe, T., 2015. Changing Landscape of Corn and Soybean Production andPotential Implications in 2015, Farm Doc Daily. Dept of Agricultural and ConsumerEconomics, University of Illinois, Urbana-Champaign, IL.

Nordhaus, W.D., 2017. Revisiting the social cost of carbon. PNAS 114 (7), 1518–1523.https://doi.org/10.1073/pnas.1609244114.

Novak, J.M., Busscher, W.J., 2012. Selection and use of designer biochars to improvecharacteristics of southeastern USA coastal Plain degraded soils. In: Lee, J.E. (Ed.),Advanced Biofuels and Bioproducts. Springer Science, New York. https://doi.org/10.1007/978-1-4614-3348-4-7. (in press).

Ordóñez, R.A., Castellano, M., Hatfield, J.L., Helmers, M.J., Licht, M.A., Liebman, M.,Dietzel, R., Martinez-Feria, R., Iqbal, J., Puntel, L.A., Córdova, S.C., Togliatti, K.,Wright, E., Archontoulis, S.V., 2018. Maize and soybean root front velocity andmaximum depth in Iowa USA. Field Crop Res. 215, 122–131. https://doi.org/10.

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

39

1016/j.fcr.2017.09.003.Perlack, R.D., Wright, L.L., Turhollow, A.F., Graham, R.L., Stokes, B.J., Erbach, D.C.,

2005. Biomass as Feedstock for a Bioenergy and Bioproducts Industry: the TechnicalFeasibility of a Billion-ton Annual Supply. DOE/USDA Tech. Rep. DOE/GO-102005–2135 ORNL/TM-2005/66. www.ornl.gov/∼webworks/cppr/y2001/rpt/123021.pdf[verified 23 Oct. 2007]. USDOE, Office of Scientific & TechnicalInformation, Oak Ridge, TN.

Probert, M.E., Dimes, J.P., Keating, B.A., Dalal, R.C., Strong, W.M., 1998. APSIM’s waterand nitrogen modules and simulation of the dynamics of water and nitrogen in fallowsystems. Agric. Syst. 56, 1–28. https://doi.org/10.1016/S0308-521X(97)00028-0.

Puntel, L.A., Sawyer, J.E., Barker, D.W., Dietzel, R., Poffenbarger, H., Castellano, M.J.,Moore, K.J., Thorburn, P., Archontoulis, S.V., 2016. Modeling long-term corn yieldresponse to nitrogen rate and crop rotation. Front. Plant Sci. 7, 1630. https://doi.org/10.3389/fpls.2016.01630.

Puntel, L.A., Sawyer, J.E., Barker, D.W., Thorburn, P., Castellano, M.J., Moore, K.J.,Vanloocke, A., Heaton, E., Archontoulis, S.V., 2018. A systems modeling approach toforecast corn economic optimum nitrogen rate. Front. Plant Sci. https://doi.org/10.3389/fpls.2018.00436.

Ribaudo, M., Savage, J., Aillery, M., 2014. An Economic Assessment of Policy Options toReduce Agricultural Pollutants in the Chesapeake Bay. ERR-166. U.S. Department ofAgriculture, Economic Research Service.

Rogovska, N., Laird, D.A., Rathke, S.J., Karlen, D.L., 2014. Biochar impact on MidwesternMollisols and maize nutrient availability. Geoderma 230, 340–347. https://doi.org/10.1016/j.geoderma.2014.04.009.

Rogovska, N., Laird, D.A., Karlen, D.L., 2016. Corn and soil response to biochar appli-cation and stover harvest. Field Crops Res. 187, 96–106. https://doi.org/10.1016/j.fcr.2015.12.013.

Russell, A.E., Laird, D.A., Mallarino, A.P., 2006. Impact of nitrogen fertilization andcropping system on soil quality in midwestern mollisols. Soil Sci. Soc. Am. J. 70,249–255.

Spokas, K.A., Cantrell, K.B., Novak, J.M., Archer, D.A., Ippolito, J.A., Collins, H.P.,Boateng, A.A., Lima, I.M., Lamb, M.C., McAloon, A.J., Lentz, R.D., Nichols, K.A.,2012. Biochar: a synthesis of its agronomic impact beyond carbon sequestration. J.

Environ. Qual. 41, 973–989. https://doi.org/10.2134/jeq2011.0069.Thorburn, P.J., Probert, M.E., Robertson, F.A., 2001. Modelling decomposition of su-

garcane surface residues with APSIM-Residue. Field Crops Res. 70, 223–232. https://doi.org/10.1016/S0378-4290(01)00141-1.

Thorburn, P.J., Meier, E.A., Probert, M.E., 2005. Modelling nitrogen dynamics in su-garcane systems: recent advances and applications. Field Crops Res. 92, 337–352.https://doi.org/10.1016/j.fcr.2005.01.016.

Togliatti, K., Archontoulis, S.V., Dietzel, R., Puntel, L., VanLoocke, A., 2017. How doesinclusion of weather forecasting impact in-season crop model predictions? FieldCrops Res. 214, 261–272. https://doi.org/10.1016/j.fcr.2017.09.008.

USDA-NASS (National Agricultural Statistical Service), 2016. Quick-Stats. (4.12.16.).https://www.nass.usda.gov/Quick_Stats/Ag_Overview/stateOverview.php?state=IOWA.

USDA-NRCS (Natural Resource Conservation Service), 2017. Web Soil Survey. Availableat. https://websoilsurvey.nrcs.usda.gov/app/.

USDA-OCE, 2017. USDA Agricultural Projections to 2026. https://www.usda.gov/oce/commodity/projections/USDA_Agricultural_Projections_to_2026.pdf.

Uzoma, K.C., Inoue, M., Andry, H., Fujimaki, H., Zahoor, A., Nishihara, E., 2011. Effect ofcow manure biochar on maize productivity under sandy soil condition. Soil UseManag. 27, 205–212. https://doi.org/10.1111/j.1475-2743.2011.00340.x.

Wallach, D., et al., 2006. Evaluating crop models. In: Wallach, D. (Ed.), Working WithDynamic Crop Models: Evaluations, Analysis, Parameterization, and Applications.Elsevier, Amsterdam, pp. 11–53.

Wang, E., Robertson, M.J., Hammer, G.L., Carberry, P.S., Holzworth, D., Meinke, H.,Chapman, S.C., Hargreaves, J.N.G., Huth, N.I., McLean, G., 2002. Development of ageneric crop model template in the cropping system model APSIM. Eur. J. Agron. 18,121–140.

Wilhelm, W.W., Johnson, J.M.F., Hatfield, J.L., Voorhees, W.B., Linden, D.R., 2004. Cropand soil productivity response to corn residue removal: a literature review. Agron. J.96 (1), 1–17.

Wilhelm, W.W., Johnson, J.M.F., Karlen, D., Lightle, D., 2007. Corn stover to sustain soilorganic carbon further constrains biomass supply. Agron. J. 99, 1665–1667. https://doi.org/10.2134/agronj2007.0150.

D.M. Aller et al. Field Crops Research 227 (2018) 30–40

40