Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production

Biomass and Bioenergy xxx (2017) 1e11

Contents lists available at ScienceDirect

Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe

Research paper

Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production Peter B. Woodbury a, *, Armen R. Kemanian b, Michael Jacobson b, Matthew Langholtz c a b c

Soil and Crop Sciences, 1017 Bradfield Hall, Cornell University, Ithaca, NY 14853, USA The Pennsylvania State University, University Park, PA, USA Oak Ridge National Laboratory, Oak Ridge, TN, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 March 2016 Received in revised form 4 January 2017 Accepted 17 January 2017 Available online xxx

Replacing row crops with perennial bioenergy crops may reduce nitrogen (N) loading to surface waters. We estimated the benefits, costs, and potential for replacing maize with switchgrass to meet required N loading reduction targets for the Chesapeake Bay (CB) of 26.9 Gg
1. After subtracting the potential reduction in N loading due to improved N fertilizer practices for maize, a further 22.8 Gg reduction is required. Replacing maize with fertilized switchgrass could reduce N loading to the CB by 18 kg ha
1 y
1, meeting 31% of the N reduction target. The break-even price of fertilized switchgrass to provide the same profit as maize in the CB is 111 $ Mg
1 (oven-dry basis throughout). Growers replacing maize with switchgrass could receive an ecosystem service payment of 148 $ ha
1 based on the price paid in Maryland for planting a rye cover crop. For our estimated average switchgrass yield of 9.9 Mg ha
1, and the greater N loading reduction of switchgrass compared to a cover crop, this equates to 24 $ Mg
1. The annual cost of this ecosystem service payment to induce switchgrass planting is 13.29 $ kg
1 of N. Using the POLYSYS model to account for competition among food, feed, and biomass markets, we found that with the ecosystem service payment for switchgrass of 25 $ Mg
1 added to a farm-gate price of 111 $ Mg
1, 11% of the N loading reduction target could be met while also producing 1.3 Tg of switchgrass, potentially yielding 420 dam3 y
1 of ethanol. © 2017 Published by Elsevier Ltd.

Keywords: Biofuel Switchgrass Watershed Nitrogen Maize Water quality

1. Introduction Producing perennial biomass feedstocks for bioenergy can have environmental benefits at the watershed scale. Such benefits are additional to other economic and greenhouse gas benefits of bioenergy production. If paid for, such additional environmental benefits, or ecosystem services, could increase the financial return to growers and increase investments in bioenergy industries and total bioenergy production. This is particularly important when low fossil fuel prices put downward pressure on alternative energy

Abbreviations: BMP, Best Management Practice, the best practices achievable on farms using current technologies, information and equipment; Bay, Chesapeake Bay; CBW, Chesapeake Bay Watershed. * Corresponding author. E-mail addresses: [email protected] (P.B. Woodbury), [email protected] (A.R. Kemanian), [email protected] (M. Jacobson), [email protected] (M. Langholtz).

sources, including bioenergy. Ecosystem services are the benefits that humans derive directly and indirectly from nature [1, 2]. When such services are classified and quantified, payments can be allocated to internalize what are otherwise economic and financial externalities. Such “payments for ecosystem services” (PES) can support increased social welfare and improve decision-making about investments in new industries and land management practices. One important environmental benefit of producing perennial bioenergy feedstocks compared to annual row crops is reduced nitrogen (N) leaching to surface and ground waters. For example, such reductions could occur if maize (Zea mays L.) area is replaced by switchgrass (Panicum virgatum L.) because switchgrass systems use N much more efficiently than maize due to having extensive root systems during all seasons that take up mineral N and store it in plant tissues [3]. Thus perennial crops such as switchgrass lose much less N to the environment, including leaching to streams and rivers which pollutes the coastal zone with excess N loading. This

http://dx.doi.org/10.1016/j.biombioe.2017.01.024 0961-9534/© 2017 Published by Elsevier Ltd.

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

2

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

opens the possibility for switchgrass and other perennial biomass crops to be used synergistically to produce a suite of ecosystem services in watersheds including reducing nutrient pollution from agriculture. Nutrient loading to coastal zones is a global issue that causes environmental problems such as damage to populations of aquatic plants and animals and increasing harmful and toxic algal bloom [4]. The Chesapeake Bay (henceforth “Bay”) and its watershed (CBW) are emblematic of these issues (e.g. Ref. [5]. Despite decades of analyses and management, reductions in nutrient loading have not met the targets required to improve water quality [6]. States within the CBW are under increasing pressure to find feasible and cost-effective methods to reduce N loading (e.g. Ref. [7]. Replacing some maize fields with switchgrass that can be used for energy production could thus reduce N loading downstream to the coastal zone. Such reduced N loading can contribute to provision of multiple ecosystem services by healthy coastal zones including food production and recreation. We estimate the potential benefit of replacing maize area with switchgrass to reduce inorganic N loading to surface waters in the CBW. Specifically, we estimate the total area and cost required to use this approach to meet the Bay nutrient reduction targets set for 2025, and the degree to which the goals could be met given the available land base and the area suitable for improved management practices to reduce N loading. Finally, we estimate the extent to which growers might be induced to grow switchgrass in place of maize while accounting for national demand for food and feed crops based on an economic model and our estimate of the payment of ecosystem services that could be available to reduce N loading to the CBW. Our analysis is intended to support strategic decision-making about the extent to which PES could both improve water quality and produce products such as ethanol to help meet society's demand for transportation fuels. 2. Methods The analysis has six main steps to determine: 1) N loading from maize to fields in the CBW; 2) N loading reduction from replacing maize with switchgrass; 3) N loading reduction from adding winter rye (Secale cereale L.) to maize systems; 4) switchgrass price with and without payment for N loading reduction; 5) payments for N loading reduction; and 6) potential future area of switchgrass using economic modeling. To accomplish these steps, our approach was to use the best available estimates of each parameter required in the calculations based on existing literature and supplemented by model results. A description of these steps follows in Sections 2.1e2.6 and summarized in Table 1. 2.1. Nitrogen loading from maize Nitrogen loading depends in particular on N fertilizer rate, timing, and harvest removal. Therefore, we estimated N fertilizer rates and subsequent N loading to the Bay from current average maize management business as usual (MþBAU) and also from best management practice maize management (MþBMP). Most of the maize production in the CBW occurs in Pennsylvania. While average N fertilizer application rates per state are available, these data are not adequate to determine representative N application rates for a given field due to crop rotations, manure use, local yield potential and other practical considerations. To

estimate N fertilizer application rates by growers, we used data from on-farm research trials in the region, modified to reflect average maize yields in Pennsylvania. Specifically, we used data from an on-farm research study for maize grain production without manure at 50 site-year combinations from 2011 to 2013. From this study, the average N fertilizer rate was 228 kg ha
1 y
1 (Table 1, [8,9]). The average maize yield in these trials was 11.7 Mg ha
1 (15.5% moisture on green weight basis), which was higher than the average of 9.4 Mg ha
1 in Pennsylvania in 2015. We adjusted the N fertilizer rate by the ratio of the average yield in Pennsylvania to that in the 50 trials (i.e. 9.4/11.7) for a final estimate of 184 kg ha
1 y
1 (Table 1). Henceforth, this estimate of current average practice for maize will be referred to as MþBAU. The on-farm trial mentioned above indicated that there is substantial room for improvement in N management and grower profit in maize production. Thus, we estimated the average N fertilizer rate using best management practices (BMP) from an online tool called “Adapt-N”. This tool uses soil and management practice data to recommend the minimum side-dress N rate that will enable a target high yield based on the recent weather at the field location [10]. Based on the study mentioned above, the Adapt-N rate averaged 160 kg ha
1 y
1 for 50 trials (site x year) in NY [8,9]. Yet, onfarm yield in plots that followed Adapt-N recommendations were not lower than those with the higher average grower rate of 228 kg ha
1 y
1 [8, 9). Thus, we adjusted the N fertilization rate using the ratio of yield in Pennsylvania to that in the 50 trials (9.4/ 11.7) for a final estimate of 129 kg ha
1 y
1, henceforth referred to Scenario MþBMP (Table 1). This recommendation is about 25% lower than the rule of thumb of approximately 22 kg of N per Mg of expected grain yield recommended by Pennsylvania State Extension [11]. For comparison, for New York State in 2007, we estimated previously an average recommended fertilizer rate for grain maize of 129 kg ha
1 y
1 [12,13], but because yields have increased since 2007, the Adapt-N recommendation represents an improved efficiency. This reduction in N rate increased average grower profits by 91 $ ha
1 because yields were not decreased (Table 1, [9]. This reduction in N rate will also reduce N loading to surface and groundwater compared to current average management practices. To estimate changes in N loading to the CBW coastal zone we must estimate (a) the N amount leaching below the rooting zone, (b) the N amount reaching streams via interflow and surface flow, and (c) the N amount reaching the coastal zone after transport and processing in streams and rivers. The amount of N leaching below the rooting zone was modeled for the 50 site-year combinations discussed above using Adapt-N. For the average grower fertilizer rate, the average N leaching below the rooting zone was 44.8 kg ha
1 y
1. This leaching rate is close to the surplus N obtained by subtracting the N in the harvested grain from N in fertilizer and other inputs. This rate was adjusted downward as described above based on the lower average yields in Pennsylvania compared to those in the on-farm trials, resulting in an adjusted value of 36.2 kg ha
1 y
1. Losses of N due to denitrification in the subsurface soils during transport below the rooting zone to the stream were calculated using the difference between basin-wide N inputs and measured concentrations of N in rivers near the coastal zone weighted by the flow. The Susquehanna River Basin is the dominant source of water and nutrients to the CBW. For the Susquehanna Basin, we assumed that the fraction of N entering the Bay from the river was 0.21 of that entering the entire watershed as calculated by Woodbury et al. [14]. We assumed that this input/output ratio can be multiplied by the N fertilizer rate to estimate the fraction of the N that would end up as N loading to the coastal zone. Losses in the riverine system have been estimated using the SPARROW model, and the average N delivery ratio of the river system within the basin is 0.98 (Table 1,

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

3

Table 1 Key parameters for estimating N loading benefits of replacing maize with switchgrass. Step

Item

Value

Unit

Description

Source

1.1.1

N fertilizer rate, maize, grower

228

kg ha
1 y
1

[8,9]

1.1.2

Maize yield, 50 sites

Grower average from study with 50 site x year combinations. Grower average from study with 50 site x year combinations. Average for Pennsylvania in 2015. Grower average from study adjusted to average PA yield. Adapt-N recommended average from study. Adapt-N recommended from study adjusted to average PA yield. Adapt-N recommended rate instead of grower rate. Grower average modeled in on-farm study with 50 site x year combinations. Adjusted to average PA yield. Adapt-N recommendedraverage modeled in on-farm study with 50 site x year combinations. Adjusted to average PA yield. Average basin-wide denitrification delivery ratio from below the rooting zone to the stream. Average basin-wide riverine denitrification delivery ratio from SPARROW model. Calculated as Steps 1.3.1*1.3.3*1.3.4. Calculated as Steps 1.3.2*1.3.3*1.3.4. BMP N rate from Cycles model for Southeast PA (Lancaster) Average of data from the literature. Average of data from the literature. Calculated as Steps 1.3.3*1.3.4*2.2.1 Calculated as Steps 1.3.3*1.3.4*2.2.2 Calculated as Steps 1.3.6-2.2.3.

11.67

Mg ha
1 y
1
1


1

[8,9]

1.1.3 1.1.4

Maize yield, PA mean for 2015 N fertilizer rate, maize, grower

9.42 184

Mg ha y kg ha
1 y
1

1.2.1 1.2.2

N fertilizer rate, maize, BMP N fertilizer rate, maize, BMP

160 129

kg ha
1 y
1 kg ha
1 y
1

1.2.3

Increase in grower profit

91.43

$ ha
1 y
1

1.3.1

N leaching, maize, grower

36.2

kg ha
1 y
1

1.3.2

N leaching, maize, BMP

23.9

kg ha
1 y
1

1.3.3

Subsurface loading ratio

0.88

ratio

1.3.4

Riverine loading ratio

0.98

ratio

1.3.5 1.3.6 2.1.1

N loading to CPB, maize, grower N loading to CPB, maize, BMP N fertilizer rate, switchgrass

31.2 20.7 75

kg ha
1 y
1 kg ha
1 y
1 kg ha
1 y
1

2.2.1 2.2.2 2.2.3 2.2.4 2.3.1

N leaching, switchgrass w N N leaching, switchgrass, w/o N N loading to CPB, switchgrass w N N loading to CPB, switchgrass w/o N Reduced N loading, BMP maize to switchgrass w/o N Reduced N loading, BMP maize to switchgrass w N Reduction in N leaching due to winter rye cover crop Reduced N loading, BMP maize to winter rye cover crop Switchgrass w N price, to equal profit of BMP maize Switchgrass w/o N price, to equal profit of BMP maize Price paid for cover crop (ha)

2.9 2.0 2.5 1.7 18.9

kg kg kg kg kg

18.2

kg ha
1 y
1

Calculated as Steps 1.3.6-2.2.4.

ratio (reduction)

See Table 2.

2.3.2 3.1.1 3.1.2 4.1.1 4.1.2 5.1.1 5.1.2 5.2.2

5.2.3

0.54

ha
1 ha
1 ha
1 ha
1 ha
1

y
1 y
1 y
1 y
1 y
1

[38] (calculated from above) [8,9] (calculated from above) [8,9] [8,9]

[8,9]

See text. [15]

[17,18] [3,17,19]

11.2

kg ha
1 y
1

Average of data from the literature (see Table 2.). Calculated as Steps 1.3.6*3.1.1.

110.83

$ Mg
1 y
1

Oven dry basis. See Table 3.

See Table 3.

Oven dry basis. See Table 3.

See Table 3.

Price paid for implementation of a winter rye cover crop in a maize field. Calculated as Steps 5.1.1*3.1.2.

[24]


1

248.11

$ Mg

148.26

$ ha
1

Price paid for cover crop N loading reduction Price required for switchgrass w/o N, harvested, N loading reduction

13.29

$ kg
1

29.75

$ kg
1

Price required for switchgrass w/o N, not harvested, N loading reduction

39.50

$ kg
1

[15]. The input/output ratio and the riverine system delivery ratio allow solving for the losses between the field and stream entry point, or subsurface N delivery ratio (although it also includes N transported via runoff). This subsurface N delivery ratio was calculated to be 0.88 (Table 1). To estimate N delivered to the Bay, the amount of N leached below the rooting zone was multiplied by the subsurface N delivery ratio and the riverine delivery ratio. For MþBAU and MþBMP, 31.2 and 20.7 kg ha
1 y
1 of N were delivered to the Bay, respectively (Table 1). 2.2. Nitrogen loading reduction from switchgrass For switchgrass, we calculated results for two management scenarios. The first scenario uses BMPs for N fertilizer rate and timing (Scenario SþN). The second scenario uses zero added fertilizer, and it is designed to reduce N loading as much as possible,


1

y

Calculated as Steps 5.1.2*4.1.2/4.1.1 (so that price for N accounts for the lower yield without fertilizer). Calculated as total value of production (from Table 3) divided by reduction in N loading (Step 2.3.1.)

but will result in lower switchgrass yields over time (Scenario SNþH). There are few data on appropriate switchgrass N fertilizer rates, particularly average values applicable to the CBW. Switchgrass is slower to establish than cool-season grasses, and experts recommend adding no fertilizer during the first year or two. Our goal was to estimate long-term average rates. We used the Cycles cropping system simulation model, an outgrowth of CropSyst [16] for a representative location in Lancaster County, PA and found a recommended rate of 75 kg ha
1 y
1 (Table 1). This is close to the average fertilizer rate of 76 kg ha
1 y
1 developed in a previous study for New York State [13]. To estimate N leaching from switchgrass we calculated the average rates based on literature from long-term experiments in the USA, and then adjusted for losses from fields to the Bay as described in Section 2.1 for maize. Due to the paucity of data for

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

4

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

switchgrass alone, we supplemented with data for mixed prairie grasses. With N fertilizer, N losses were found in two reports: 5 kg ha
1 y
1 [17] and 0.8 kg ha
1 y
1 [18] for an average of 2.9 kg ha
1 y
1 (Table 1). Without N fertilizer, the N losses reported were 1.4, 2.0, and 2.5 kg ha
1 y
1 [3,17,19], respectively, for an average of 2.0 kg ha
1 y
1 (Table 1). These values were multiplied by the N delivery ratio for subsurface soils (0.88) and for rivers (0.98). N loading estimates for Scenario SþN management were 2.5 kg ha
1 y
1 and for Scenario S-NþH were 1.7 kg ha
1 y
1 (Table 1). We estimated this benefit for two comparisons: (1) by assuming BMPs for maize and using fertilized switchgrass, subtracting from the maize the N loading rate to the CBW, i.e. Scenario MþBMP - Scenario SþN ¼ 20.7e2.5 ¼ 18.2 kg ha
1 y
1 (Table 1), and (2) by assuming no N fertilizer for switchgrass, i.e. Scenario MþBMP - Scenario S-NþH ¼ 20.7e1.7 ¼ 18.9 kg ha
1 y
1 (note that results may differ slightly due to rounding, Table 1). Although 18.2 and 18.9 kg ha1
y
1 are indistinguishable for any practical purpose, the fertilization of switchgrass increases yield and production costs, which are considered in the economic analysis. 2.3. Cover crop We estimate the benefit of adding a winter rye cover crop to maize. This is a critical component of our analysis, because there are currently payments for ecosystem services being made in the watershed for adding a winter cover crop to maize that could be a competing practice for replacing maize with switchgrass to improve water quality. We reviewed the literature for multi-year field studies of adding a winter rye cover crop to maize in the USA and calculated an average N leaching reduction of 0.54 as shown in Table 2. Applying this factor, the N loading reduction from adding a winter rye cover crop to maize is as follows: Maize BMP N loading * cover crop reduction ¼ 20.7*0.54 ¼ 11.2 kg ha
1 y
1 (Table 1). We call this scenario MþCC. 2.4. Switchgrass price We obtained maize enterprise budgets from the USDA Economic Research Service (ERS) for our region (the “Northern Crescent”) for the last 10 years [20]. This region contains New England, New York, New Jersey, and most of Pennsylvania. These budgets do not include any government payments. We averaged these values for the 10 years from 2005 to 2014

(see Table 3). Then we reduced selected operating costs for switchgrass, specifically seed, fertilizer, and chemicals because switchgrass does not require these inputs every year. Therefore, costs were amortized over the projected stand life of 10 years (Table 3). All costs were adjusted for inflation using a price deflator based on Federal Reserve Economic Data [21]. Assuming a BMP for N fertilizer minus total costs for maize in the Northern Crescent region from 2005 to 2014, we calculated that the average value of production is 141 $ ha
1 planted, including all costs, or 767 $ ha
1 planted including only operating costs. For switchgrass yield, we used an estimate from the Cycles model for Lancaster PA that is representative for maize land in the CBW. The estimated yield was 9.9 Mg ha
1 (yield of 12.0 Mg ha
1 multiplied by 0.825 for establishment years, and with harvest efficiency of 0.9 already included in the 12.0 Mg ha
1 yield estimate). The value of 0.825 is based on data from field trials in New York State in which there was no harvest in the first year, average 50% of maximal yield in the second year, 75% of maximal yield in the third year, and maximum yields for the remaining seven years in the 10-year cycle. The final estimated yield is very close to the average yield of 9.8 Mg ha
1 found in field trials for multiple cultivars and multiple locations throughout NY, PA, NJ, MD, and VA [22]. Finally, we calculated that over the 10-year period (2005e2014) the break-even cost of switchgrass to provide the same average profit (value of production minus total costs) as maize equals 111 $ Mg
1 (Table 3). To estimate the ethanol that can be produced from the switchgrass biomass we used a conversion factor of 318 L Mg
1 [23]. 2.5. Payment for N loading reduction The State of Maryland offers a cover crop incentive program to reduce N leaching in the CBW. Growers are paid 99e247 $ ha
1 depending on prior crop, manure use, and other factors [24]. Maryland growers planted 193,440 ha of cover crops for 2014e2015 through the state's Cover Crop Program [24]. Including the base payment for maize land, using winter rye only, the payment amount for 2014e2015 was 148 $ ha
1 as shown in Table 1, Step 5.1.1 [24]. By dividing 148 $ ha
1 by our previously calculated reduction in N leaching with this practice of 11.2 kg ha
1 y
1 (Table 1, Step 3.1.2) we derive a cost of 13.29 $ kg
1 of N (Table 1, Step 5.1.2). For switchgrass with N fertilizer (Scenario SþN), the PES price paid for N loading reduction is estimated to be 13.29 $ kg
1 of N (Table 1, Step 5.1.2), the same as currently paid for a rye cover crop in maize. For switchgrass without N fertilizer (Scenario S-NþH), the price paid per kg of N loading reduction is estimated to be 29.75 $ kg
1 (Table 1, Step 5.2.2), because there is a lower yield of switchgrass, and the reduced payment for switchgrass production

Table 2 Reduction in nitrate leaching from US maize fields using a rye cover crop. State

Oregon Iowa Iowa Iowa Kentucky Delaware Maryland Maryland Minnesota MEAN

Crop/rotation

Sweet corn maize/soy maize/soy maize/soy maize maize maize maize maize/soy

Length of study

NO3 reduction due to cover crop

y

ratio

2 4 4 5 3 2 2 6 3

0.33 0.69 0.61 0.47 0.94 0.19 0.77 0.73 0.13 0.54

Citation

[39] [18] [40] [41] [42] [43] [44] [45] [46]

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

5

Table 3 Enterprise budget for maize and switchgrass for Northeast USA year 2015.a Maize production costs and returns

Maize (USDA-ERS)d

Switchgrass (Our estimate)e

Item

Avg 2005-2014

Avg 2005-2014

Switchgrass without N fertilizer (Our estimate)f

Switchgrass without N, no harvest (Our estimate)f

$ ha
1

$ ha
1

$ ha
1

1,436.78 12.15 1,448.94

1,097.19 0.00 1,097.19

900.63 0.00 900.63

0.00 0.00 0.00

183.78 251.75 64.77 45.84 78.70 52.54 4.77 682.15

19.77 120.75 8.03 45.84 78.70 52.54 4.77 330.41

19.77 0.00 8.03 33.14 56.90 37.99 4.77 160.60

19.77 0.00 8.03 7.56 12.98 8.67 4.77 61.78

9.30 87.31 198.58

9.30 87.31 198.58

6.73 63.13 198.58

1.53 14.40 198.58

246.30

246.30

246.30

246.30

26.48 58.19 626.17

26.48 58.19 626.17

26.48 58.19 599.41

26.48 58.19 545.50

1,308.33

956.58

760.02

607.28

140.61

140.61

140.61

-607.28

766.78

766.78

740.02

-61.78

9.90

3.63

3.63

110.83

248.11

$ ha Gross value of production Primary product: Corn grain Secondary product: Corn silage Total, gross value of production Operating costs: Seed Fertilizerb Chemicals Custom operationsc Fuel, lube, and electricity Repairs Interest on operating capital Total, operating costs Allocated overhead: Hired labor Opportunity cost of unpaid labor Capital recovery of machinery & equipment Opportunity cost of land (rental rate) Taxes and insurance General farm overhead Total, allocated overhead Total, costs listed Value of production less total costs listed Value of production less operating costs Supporting information: Yield 8.50

Price

168.99


1

(not applicable)

Oven-dry Mg ha
1

$ Mg
1

Switchgrass Notes

Our Our Our Our Our Our

estimateg estimateh estimateg estimatei estimatei estimatei

Our estimatei Our estimatei

Results from Cycles model, adjusted for establishment years (0.825) Adjusted to equal maize value of production.

Bold rows represent totals or subtotals. a Developed from survey base year, 2010. b Maize value reduced from USDA rate to reflect the best management practice nitrogen fertilizer rate for maize (see Table 1). c Cost of custom operations, technical services, and commercial drying. d 10-year average from the USDA Economic Research Service adjusted for inflation and converted to SI units. e Switchgrass budget based on maize budget except as described under “Switchgrass Notes”. f Similar to fertilized switchgrass, but yield is much lower and fertilizer cost is zero. g For 1 year in 10 (Source: [47]). h Reduced from maize value based on the ratio of fertilizer rates for switchgrass versus maize (see Table 1). i Adjusted for only harvest from non-fertilizer harvested switchgrass and for mowing only for non-harvested switchgrass (Source: [48]).

must be paid for as a nutrient reduction cost. For switchgrass without N fertilizer and without any harvest (Scenario S-N-H), the PES price paid per kg of N loading reduction is estimated to be 39.50 $ kg
1 (Table 1, Step 5.2.3). This is because there is no harvest of switchgrass, so the nutrient reduction payment must be even higher. 2.6. Biomass production To evaluate potential conversion of cropland area to switchgrass producing area, we used the Policy Analysis System (POLYSYS), a policy simulation model of the U.S. agricultural sector [25]. The POLYSYS modeling framework, which can be conceptualized as a variant of an equilibrium displacement model, simulates changes in economic policy, agricultural management, and natural resource conditions, and estimates the impacts to the U.S. agricultural sector from these changes. POLYSYS is used to estimate how growers may

respond to market signals, such as new demand for biomass, while simultaneously considering the impact on other non-energy crops. POLYSYS was used to quantify potential biomass resources in the Billion-Ton Update [26] and has been used in other agricultural and biofuels analyses [27,28]. POLYSYS anchors its analyses to the U.S. Department of Agriculture (USDA) published baseline of yield, area, and price projections for the agriculture sector, which are endogenously extended from the USDA 10-year baseline projection period through 2040 for this analysis [29]. Traditional crops currently considered in POLYSYS include maize, grain sorghum, oats, barley, wheat, soybeans, cotton, rice, and hay, which together comprise approximately 90% of the U.S. agricultural land area. Conventional crops simulated for residues include maize, grain sorghum, oats, barley, and wheat (winter plus spring). Production costs associated with herbaceous and woody energy crops include land rent, establishment, maintenance, and harvest costs.

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

6

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

POLYSYS is a linear program that solves for the most profitable mix of land-use alternatives on agricultural lands from the growers' perspective. POLYSYS calculates net present value (NPV) of agricultural alternatives over a 20-year planning horizon, and selects the most profitable mix of production alternatives. For this study, POLYSYS was run from 2015 to 2040, solving crop areas at the county level. Contracts for dedicated energy crops, including switchgrass, were simulated to begin in 2019. The model was run introducing switchgrass farm-gate prices needed to meet or exceed potential future profitability maize production on a per-hectare basis. Specifically, the area of switchgrass was modeled for three different prices: the average break-even price for switchgrass versus maize grain, and this break-even price plus 25 $ Mg
1 or minus 25 $ Mg
1. The 25 $ Mg
1 represents a payment made either to the grower (the break-even price plus $25) or to the refiner (the break-even price minus $25). 3. Results 3.1. Area and cost for meeting Chesapeake basin N targets The total N reduction required for the Bay by 2025 is 26.9 Gg (Table 4). The N loading reduction achievable with BMP for N fertilizer with maize is 10.6 kg ha
1, calculated as the N loading from current practices minus the N loading from BMP fertilizer practice (Table 4). The current area of maize in the basin is 562,244 ha. However, there are already nutrient management practices including riparian forest buffers, riparian herbaceous buffers, and filter strips. As a result, the area available for BMP maize N fertilizer management is 387,948 ha (Table 4). This is the area that does not already have any conservation practices, and thus is a conservative value because some other conservation practices might be

compatible with BMP N fertilizer management. The potential reduction from achieving BMP fertilizer management for maize is 4 Gg (Table 4). After subtracting the potential reduction in N loading due to BMP N fertilizer practices for maize, there is still a need to reduce N loading by 22.8 Gg by 2025 (Table 4). For Scenario MþCC (maize with winter rye cover crop) the N reduction beyond that achieved by MþBMP is 11.2 kg ha
1 y
1 (Table 4). The total area that would be required to reduce N loading by 22.8 Gg is 2 million hectares (Table 4). This area is much larger than that available for the BMP practice (z0.4 million hectares, Table 4). If all available area were used, 19% of the target could be achieved at a unit cost of N of 13.29 $ kg
1 (Table 4). For Scenario SþN (switchgrass with N fertilizer), the N reduction beyond that achieved by MþBMP is 18.2 kg ha
1 y
1 (Table 4). The area required to meet the year 2025 goal is 1.2 million hectares, again a figure much larger than the available area. If all of the available area was used, 31% of the target could be achieved d (Table 4). A much greater fraction of the target could be met for this scenario than for the MþCC scenario because on a per unit area basis switchgrass is much more effective at reducing N loading than a rye cover crop (Table 4). The unit cost of the N loading reduction is 13.29 $ kg
1 because it was assumed that the same rate would be paid as is currently being paid for the cover crop practice (Table 4). For both Scenario S-NþH (switchgrass without N fertilizer and harvest) and Scenario S-N-H (switchgrass without N fertilizer without any harvest), the N reduction beyond that achieved by BMP N fertilizer management on maize is 18.9 kg
1 ha
1 (Table 4). For these scenarios 32% of the target could be met, slightly more than for Scenario SþN because the N loading reduction with switchgrass is slightly greater without N fertilizer application (Table 4). For Scenario S-NþH the cost of the N loading reduction is more than doubled to 29.75 $ kg
1 because the switchgrass yield is lower

Table 4 Area of land and cost required to meet Chesapeake Basin nitrogen reduction goals for year 2025 using improved fertilizer management combined with 4 agricultural scenarios in maize fields. Description

Value

Unit

Source

N loading reduction required for Chesapeake Basin (2025) N loading reduction with maize BMP N management (2025) Maize area (2011, assumed no change by 2025) Area of maize without N filtering practice (2011) Total N loading reduction with maize BMP N mgmt. (2025) Remaining N loading reduction required for Chesapeake Basin after maize BMP N management (2025)

26879884 10.6 5,62,244 3,87,948 40,94,422 22785462

kg kg ha
1 y
1 ha ha kg kg

[49] Calc. from Table 1 (Step 1.3.5-1.3.6) [50] [50] Calc. from above Calc. from above

Maize with winter rye cover crop (MþCC) N loading reduction per ha, beyond maize BMP Cost of N loading reduction per kg N, beyond maize BMP Area of winter cover crop required Percentage of target technically achievable

11.2 13.29 2042348 19%

kg ha
1 y
1 $ kg
1 ha of target

Table 1 (3.1.2) Table 1 (5.1.2) Calc. from above Calc. from above

Switchgrass with N fertilizer (SþN) N loading reduction per ha, beyond maize BMP Cost of N loading reduction per kg N, beyond maize BMP Area of switchgrass replacing maize required Percentage of target technically achievable Switchgrass yield Cost of N loading reduction (per Mg switchgrass)

18.2 13.29 1254916 31% 9.9 24.37

kg ha
1 y
1 $ kg
1 ha of target Mg ha
1 $ Mg
1

Table 1 (2.3.2) Set equal to cover crop Calculated, this table Calculated, this table Table 3 Calculated from above

Switchgrass without N fertilizer, harvested (S-NþH) N loading reduction per ha, beyond maize BMP Cost of N loading reduction per kg N, beyond maize BMP Area of switchgrass replacing maize required Percentage of target technically achievable

18.9 29.75 1203426 32%

kg ha
1 y
1 $ kg
1 ha of target

Table 1 (2.3.2) Table 1 (5.2.2) Calculated, this table Calculated, this table

Switchgrass without N fertilizer, not harvested (S-N-H) N loading reduction per ha, beyond maize BMP Cost of N loading reduction per kg N, beyond maize BMP Area of switchgrass replacing maize required Percentage of target technically achievable

18.9 39.50 1203426 32%

kg ha
1 y
1 $ kg
1 ha of target

Table 1 (2.3.2) Table 1 (5.2.2) Calculated, this table Calculated, this table

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

without N fertilizer (Table 4). For Scenario S-N-H the cost of the N loading reduction is considerably higher at 39.50 $ kg
1 because there is no switchgrass harvest, thus all of the costs must be paid for in the PES (Table 4). 3.2. Synergy between bioenergy production and water quality The different land management scenarios render different mixtures of production of maize, switchgrass, and of reduced N loading (Table 4). When compared to the maize BAU and on a per unit area basis, three scenarios produce greater gross value per hectare: MþBMP, MþCC, and SþN (Fig. 1). From the N loading reduction perspective, however, the two switchgrass scenarios produce greater gross value than alternatives without switchgrass. Of the two switchgrass scenarios, Scenario SþN produces the greatest gross value of production (Fig. 1). This is of course because it has only slightly lower N loading reduction than S-N, but much higher switchgrass yields due to use of adequate N fertilizer. If bioenergy production is desired, Scenario SþN is the most suitable scenario. If maximum N loading reduction is the goal, all of the three switchgrass scenarios are similar, and higher than any of the maize scenarios, but the SþN scenario has the lowest-cost N loading reduction of any switchgrass scenario (Table 4), despite adding N fertilizer to the system compared to S-N scenarios (still less N than the scenarios with maize). Because these are gross values, they do not indicate which scenarios would be most profitable for the grower, which depends on the prices paid for both crop production and for N loading reduction, as well as the costs of production (Table 3). When costs are accounted for, Scenario SþN has the same profit as MþBAU (Table 3) because of our method of setting the PES price. Scenario MþBMP will have increased grower profit relative to MþBAU because of reduced N fertilizer cost while maintaining maize yield. Scenarios S-NþH and S-N-H would have lower profit than Scenario SþN because the N loading is only slightly lower, while switchgrass yields are much lower (Scenario S-NþH) or zero (Scenario S-N-H). The payment for N loading reduction (i.e. PES) for switchgrass produced with N fertilizer is 24.37 $ Mg
1 (Table 4). If this payment was paid to growers in addition to the break-even price for switchgrass, it could induce greater planting of switchgrass. Conversely, a similar reduction in switchgrass price should reduce the area of switchgrass. The potential for such changes in

2,000

Maize production Switchgrass production N reduction

Gross value ($ ha-1)

1,800 1,600 1,400 1,200

7

switchgrass production while meeting national demand for food and feed was estimated using the POLYSYS model. In addition, the POLYSYS output allows for a market-based estimate of the potential ethanol production from switchgrass. As the price of switchgrass biomass decreases from the breakeven price, the area of switchgrass planted on former maize land decreases by 1110 ha for each dollar of decrease in price per Mg, while for each dollar of increase above the break-even price causes an expansion of 2079 ha. Thus, a 25 $ Mg
1 change in price causes an asymmetric reduction versus expansion of switchgrass area. While the equilibrium area calculated with POLYSYS at the break-even price is 81,000 ha, it expands to 133,000 ha when the price increases by 25 $ Mg
1 (Table 5). The scenario of maximum biomass production, i.e. fertilized switchgrass at the highest price, gives a production of ethanol of 420 dam3 y
1, which is equivalent to the production capacity of a grain-based bio-refinery that uses feedstock from approximately 100,000 ha of maize. Using best management practices for N fertilizer on maize fields that do not already have an N reduction practice would achieve 15% of the target with increased profit to the grower (Table 4, Fig. 2). A similar reduction of 16% could be achieved by adding a winter rye cover crop to these lands (Fig. 2). Planting switchgrass on these lands would achieve 41% or 43% of the target, with and without N fertilizer respectively (Fig. 2). However, even with a PES of 25 $ Mg
1 the POLYSYS model results show that when accounting for competition between food, feed, and biomass markets, only 11% of the target could be met (Fig. 2). These results do indicate that PES could provide 420 dam3 y
1 of ethanol production while at the same time achieving a substantial fraction (18%) of water quality goals and meeting demand for food and feed. 4. Discussion Society demands a great deal from agricultural lands including the production of food, feed, fiber, and increasingly bioenergy. At the same time, society demands that these lands provide other environmental benefits such as clean water, clean air, pollination and other services, yet there are no clear market mechanisms to balance the production of these services. Producing perennial biomass crops for bioenergy could also produce environmental benefits that are additional to the economic benefits from the bioenergy and bio-products. Among the environmental benefits of perennial biomass crops when compared to annual row crops is a reduction in N loading. But how can we calculate the price for such service? To determine the biophysical and economic feasibility of replacing some fraction of maize fields in the CBW with switchgrass, a number of critical costs and benefits must be carefully estimated. Furthermore, the option of replacing maize with

1,000 800

Table 5 Predicted area of switchgrass in 2025 on former maize land as a function of three switchgrass biomass prices, and the estimated biomass and ethanol production from fertilized and unfertilized switchgrass.

600 400 200

Pricea

Area

0 ha

Fertilized switchgrass

Unfertilized switchgrass

Production Ethanol

Production

Gg y
1

Break-even e $25 53,755 532 Break-even 81,508 807 Break-even þ $25 1,33,479 1,321

Scenario

Ethanol

dam3 y
1 Gg y
1

dam3 y
1

169 257 420

62 94 154

195 296 485

a Price is relative to the average break-even price for switchgrass to produce the same profit as maize grain.

Fig. 1. Gross value of crop production and of nitrogen reduction for each scenario.

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

30

160

25

140

20

120

15

100

Cost ($ Mg-1)

Bay N loading reduction (Gg y-1)

8

10 5 0

B

B

A

A $111

A

80 60 40 20 0

Scenario Fig. 2. Potential annual reduction in nitrogen loading to the Chesapeake Bay: comparison of target to four scenarios. Grey bars show maximum potential nitrogen loading reduction on available area for each scenario. Note that the maize BMP benefit could be additional to the MaizeþCover Crop scenario. The black bar shows the nitrogen loading reduction if switchgrass is grown in the area deemed economically viable as predicted by the POLYSYS model with the ecosystem service payment to the grower of $25 Mg
1.

switchgrass should be analyzed in the context of competing alternatives for reducing N loading from maize fields. Three such options are (a) improving N fertilizer management, (b) using winter cover crops, and (c) changing annual cropland into set-aside nonharvested land, often as riparian buffers adjacent to streams and rivers. There have been many techno-economic analyses of the potential to produce ethanol from cellulosic biomass, especially switchgrass (e.g. [30e32]). However, it is not enough to demonstrate that such a value chain is feasible. It is also critically important to assure that bioenergy production is feasible in competition with other crops that could be produced on the same land, such as maize. For this reason, we determined the break-even cost for switchgrass production for the Northeast US region that would provide the same profit as maize over the last decade, adjusted for inflation to year 2015. We estimated this break-even cost to be 111 $ Mg
1 of switchgrass (Table 3). In Fig. 3, we show that this value is lower than previously published estimates for the North Central and Northeast region and is lower than eight of 11 published values for all regions of the USA (see Appendix Table A-1 for details and citations). However, this price is higher than estimates of what bio-refineries would be willing to pay for feedstock (e.g., 21e69 $ Mg
1 reported [33]). These estimates from the literature have not been updated for inflation (if so these figures would be slightly higher). Because this price is higher than bio-refineries may be willing to pay, another key question is: how large can an ecosystem service payment be for the water quality benefit of replacing maize with switchgrass? This payment could be paid to the grower, potentially reducing the gap between the biomass price required by the grower to equal the profit from maize and the price a bio-refinery is willing to pay for switchgrass. 4.1. Estimating the PES requires (1) estimating the N loading reduction from replacing maize with switchgrass, (2) determining whether current average practice or a best management practice is an appropriate baseline for such a calculation, and

Midwest

North Central & Northeast

Southeast

Fig. 3. Summary of comparative break-even farmgate cost for switchgrass compared to maize in regions of the USA. Bars with the same letter above are from the same study. The bar with the price above is the current study.

(3) estimating the price that States or other entities are willing to pay to reduce N loading to the Bay. For maize using the average grower N fertilizer rate, we estimated that N leaching below the rooting zone was 36 kg ha
1 y
1. This is somewhat lower than data from the literature for similar production systems, e.g. 40 kg ha
1 y
1 [3] and 49 kg ha
1 y
1 [17] but these studies had somewhat higher fertilizer rates of 179 and 180 kg ha
1 respectively. However, based on on-farm field trials, we estimated that reducing the N fertilizer rate also reduces N leaching to at least 24 kg ha
1 y
1 (Table 1). We suggest that this BMP rate is a more appropriate baseline than current average practice, because this rate is feasible and would increase grower profits, and thus it should be implemented before making a PES. Using this baseline in combination with the average N leaching from switchgrass fields from the literature and the price paid in Maryland for N reduction from a cover crop (13.29 $ kg
1 N, Table 4) results in a PES of 24 $ Mg
1 switchgrass for the lowest-cost switchgrass scenario (with adequate N fertilizer, Table 4). This price of 13.29 $ kg
1 is lower than the range for agricultural practices of 18e100 $ kg
1 N from previous publications for the CBW and is also lower than the range for point sources of 22e40 $ kg
1 N [34,35]. While these calculations provide a reference price required for growers to replace maize with switchgrass, the replacement of thousands of hectares requires determining the extent to which such replacement affects the market for other agricultural products (including maize) not only in the CBW but at a national level. If all of the available maize land (i.e., land that does not already have an N conservation practice) were converted to switchgrass, 31% of the Bay N loading reduction target of 26.9 Gg y
1 could be met (Table 4, Fig. 2). However, based on POLYSYS model results to account for biomass and crop market effects, only 11% of the N loading target could be met (Fig. 2). Furthermore, this area of switchgrass on former maize land does not represent the total switchgrass area of the CBW in POLYSYS, because switchgrass at such prices also displaces other crops for a total area of switchgrass of 216,000 ha, or 60% more than the area of switchgrass on only former maize land. Although not a

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

complete solution to N loading, these results demonstrate that replacing some maize with switchgrass, including a feasible PES, could help meet CBW nutrient targets. In addition, this replacement can produce from 154 dam3 (S-NþH) to 420 (Scenario SþNþH) of ethanol per year (Table 5). The potential 24 $ Mg
1 PES could theoretically be applied in different ways to decrease feedstock costs, increase offered feedstock prices, or a combination of both. For example, it could offset 24 $ Mg
1 of the 111 $ Mg
1 profit-matching farm-gate price to landowners, in effect reducing feedstock cost to the biorefinery to 85 $ Mg
1. In this scenario, the CBW could produce approximately 807 Gg by 2025, which could potentially be converted to 257 dam3 of ethanol. In another example, bio-refineries could increase an offered price from 111 $ Mg
1 to 134 $ Mg
1. At this higher price, switchgrass production by 2025 could potentially increase to 1.3 Tg. However, just as there is competition between maize markets and biomass markets, there is also competition between different management practices that can reduce N loading from maize fields. We found that either planting a winter rye cover crop or implementing BMPs for N fertilizer would each provide a similar total N loading reduction benefit as would replacing maize with switchgrass under the constraints imposed by POLYSYS (Fig. 2). Even greater benefits would occur if both of these practices were implemented together, i.e. replacing a fraction of the maize area that does not have N conservation practices (about 0.2 million hectares for an 11% reduction in N loading) and adding cover crop and BMP N fertilizer management in the remaining maize area. Furthermore, the total gross value of both crop production and N loading reduction on a perhectare basis would be greater for these improved N management scenarios for maize as compared to any of the switchgrass scenarios (Fig. 1. However, they would not entail any fuel production, unless maize stover is harvested, which requires other environmental considerations. Conversely, another practice in the CBW is to remove cropland from production and replace it with set-aside non-harvested land (Scenario S-N-H). For example, a current goal for Pennsylvania is to achieve 38,400 ha additional riparian forest buffer by 2025 [36]. In comparison to this practice, both switchgrass harvest scenarios (with or without N fertilizer), would be more cost-effective, demonstrating the value of obtaining multiple ecosystem services (provision of improved water quality and feedstock for bioenergy) from the same land. The methodology used in our calculations requires several simplifying assumptions. We used maize area and N leaching from maize fields to benchmark the benefits obtained by improving N fertilizer management or substituting switchgrass for maize regardless of the crop rotation or the position in the landscape where this substitution occurs. Clearly, displacing maize implies displacing other components of the crop rotation such as soybean, but our analysis focuses only on the maize portion of the rotation. In addition, our POLYSYS estimate of the plausible switchgrass area coupled with our estimate of the average N leaching avoidance benefits can be considered an average baseline benefits for any location. Targeting switchgrass in sensitive areas of the landscape would be expected to result in larger than average reduction in N leaching to streams. There are of course other ecosystem services that can be provided by replacing maize with switchgrass fields that are more biologically diverse (e.g. Ref. [37]). More research is needed to assess the value and potential PES of other externalities, for example phosphorus and sediment reduction,

9

greenhouse gas mitigation, or biodiversity benefits, that could also help realize combined environmental, social, and economic objectives in the CBW. Valuing these services may change the results obtained solely by focusing on a break-even price compared with maize and on N reduction. Also they may further highlight the benefits of switchgrass when replacing maize, or generate a larger portfolio of new beneficial scenarios that need to be optimized. 5. Conclusions Replacing row crops with perennial bioenergy feedstocks like switchgrass can reduce N loading to the CBW. Replacing maize area that is not currently under a specific N conservation practice with fertilized switchgrass could reduce N loading to the Bay by 18 kg ha
1 y
1 with 31% of the target being met (Scenario SþN). A similar reduction is obtained with unfertilized switchgrass, but with a lower biomass production and therefore a higher cost if the unrealized biomass payment is added to the N reduction benefit. The cost per unit of N is 13.29 $ kg
1 N for N fertilized switchgrass, but 39.50 $ kg
1 N for unfertilized and unharvested switchgrass. The break-even price of fertilized switchgrass at 9.9 Mg ha
1 yield to provide the same profit as maize in the region is 111 $ Mg
1. Growers replacing maize with switchgrass could receive an ecosystem service incentive payment of 148 $ ha
1 based on the price paid in Maryland for planting a rye cover crop. For our estimated average switchgrass yield and the greater N loading reduction of switchgrass compared to a cover crop this equates to 24 $ Mg
1 biomass. Production of 1.3 million Mg in 130,000 ha formerly in maize in the CBW under the (SþNþH) scenario could meet 11% of the N loading reduction target while also producing 420 dam3 y
1 of ethanol. This demonstrates that multiple ecosystem services could be achieved. In summary, while there is substantial potential for PES to both improve water quality and produce bioenergy in comparison to current practices, there is competition from other methods of reducing N loading from maize fields, specifically, improving N fertilizer management and/or planting a winter cover crop. However, in comparison to the practice of removing cropland from production as set-aside land, either of the switchgrass harvest scenarios would provide water quality benefits at a lower cost, because biomass could be harvested to produce valuable bioenergy or bio-products. Acknowledgements This work was supported by the USDA National Institute of Food and Agriculture Grant # 2012-68005-19703, The Northeast Woody/ Warm-Season Biomass Consortium: Building Sustainable Value Chains for Biomass Energy and by the EPA Grant# RD835568, Center for Integrated Multi-Scale Nutrient Pollution Solutions. This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-publicaccess-plan).

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

10

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11

Appendix

Table A-1 Comparative break-even cost for switchgrass compared to maize in the USA. Region

Break-even cost, switch-grass vs.maize

State

Notes

$Mg
1 MW MW MW

129 146 124

MW

115

MI

NC/NE NC/NE

148 111

PA

NC/NE SE SE

141 111 100

SE SE SE

87 112 56

Switch-grass yield

Maize yield

Mg ha
1

Mg ha
1

Year

Maize price

Soy price

$Mg
1

$Mg
1

Also includes risk averse. I1, MO

Cave-in-rock on non-eroded soil. Compared to maize-soy rotation. Unrealistically assume same switchgrass yield on all soils, but high yields on this good soil reasonable. Central Claypan Region. Continuous maize, includes sale of stover for feedstock. Does not include storage cost. Experimental study location.

8.2

9.8

2009

138

321

[51] [52] [53]

9.0

8.5

2009-2010

132

(n/a)

[54]

(This manuscript)

9.9

8.5

2009

169

11.7

7.5

2010

176

13.3

7.5

Also includes risk averse. MS Farmgate price for half of rowcrop land (not sure which crops). Also includes marginal and pasture land. TN SC

Citation

Also includes risk averse. Coastal Plain Pee Dee Region. 15-y switchgrass stand life.

References [1] MEA. Millennium Ecosystem Assessment, 2005, Ecosystems and Human Wellbeing: Synthesis, Island Press, Washington, DC., 2005, p. 155. [2] P. Kareiva, H. Tallis, T.H. Ricketts, G.C. Daily, S. Polasky, Natural Capital: Theory and Practice of Mapping Ecosystem Services, Oxford University Press, 2011. [3] G.F. McIsaac, M.B. David, C.A. Mitchell, Miscanthus and switchgrass production in central Illinois: impacts on hydrology and inorganic nitrogen leaching, J. Environ. Qual. 39 (5) (2010) 1790e1799. [4] W.K. Dodds, W.W. Bouska, J.L. Eitzmann, T.J. Pilger, K.L. Pitts, A.J. Riley, et al., Eutrophication of US freshwaters: analysis of potential economic damages, Environ. Sci. Technol. 43 (1) (2008) 12e19. [5] E.W. Boyer, C.L. Goodale, N.A. Jaworski, R.W. Howarth, Anthropogenic nitrogen sources and relationships to riverine nitrogen export in the northeastern USA, Biogeochemistry 57 (1) (2002) 137e169. [6] K. Blankenship, Region lags behind on nitrogen goals; phosphorus progress questioned, Bay J. (2015). http://www.bayjournal.com/article/region_lags_ behind_on_nitrogen_goals_phosphorus_progress_questioned. [7] USEPA, Interim evaluation of Pennsylvania's 2014-2015 milestones and WIP progress, U. S. Environ. Prot. Agency (2015) 6. [8] B. Moebius-Clune, H.M. van Es, J. Melkonian, Adapt-n Increased Grower Profits and Decreased Environmental N Losses in 2011 Strip Trials. What's Cropping up?, Department of Crop and Soil Sciences, Cornell University, 2012. [9] B. Moebius-Clune, M.C. Ball, H.M. van Es, J. Melkonian, Adapt-n Boosts Profits and Cuts N Losses in Three Years of On-farm Trials in New York and Iowa. What's Cropping up?, Department of Crop and Soil Sciences, Cornell University, 2014 p. Sept/Oct 57e60. [10] Melkonian JJ, van Es HM, DeGaetano AT, Joseph L. ADAPT-N: Adaptive nitrogen management for maize using high-resolution climate data and model simulations. In: Kosla R, editor. 9th International Conference on Precision Agriculture. Denver, CO 2008. [11] Penn State Extension, The Agronomy Guide 2015-2016, Pennsylvania State University, 2015. http://extension.psu.edu/agronomy-guide. [12] J.L. Wightman, Z.U. Ahmed, T.A. Volk, P.J. Castellano, C.J. Peters, S.D. DeGloria, et al., Assessing sustainable bioenergy feedstock production potential by integrated geospatial analysis of land use and land quality, BioEnergy Res. 8 (4) (2015) 1671e1680. [13] J.L. Wightman, J.M. Duxbury, P.B. Woodbury, Land quality and management

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25]

[52] This manuscript [51] [55] [56]

[57] [51] [58]

practices strongly affect greenhouse gas emissions of bioenergy feedstocks, BioEnergy Res. 8 (4) (2015) 1681e1690. P.B. Woodbury, R.H. Howarth, G. Steinhart, Understanding nutrient cycling and sediment sources in the upper Susquehanna River basin, J. Contemp. Water Res. Educ. 139 (2008) 7e14. S.W. Ator, J.W. Brakebill, J.D. Blomquist, Sources, fate, and transport of nitrogen and phosphorus in the Chesapeake Bay watershed: an empirical model, U.S. Geol. Surv. (2011) 27. €ckle, A.R. Kemanian, R.L. Nelson, J.C. Adam, R. Sommer, B. Carlson, C.O. Sto CropSyst model evolution: from field to regional to global scales and from research to decision support systems, Environ. Model. Softw. 62 (2014) 361e369. C.M. Smith, M.B. David, C.A. Mitchell, M.D. Masters, K.J. Anderson-Teixeira, C.J. Bernacchi, et al., Reduced nitrogen losses after conversion of row crop agriculture to perennial biofuel crops, J. Environ. Qual. 42 (1) (2013) 219e228. A.L.M. Daigh, X. Zhou, M.J. Helmers, C.H. Pederson, R. Horton, M. Jarchow, et al., Subsurface drainage nitrate and total reactive phosphorus losses in bioenergy-based prairies and corn systems, J. Environ. Qual. 44 (5) (2015) 1638e1646. G. Hernandez-Ramirez, S.M. Brouder, M.D. Ruark, R.F. Turco, Nitrate, Phosphate, and Ammonium Loads At subsurface drains: agroecosystems and nitrogen management, J. Environ. Qual. 40 (4) (2011) 1229e1240. USDA-ERS, Commodity Costs and Returns, United States Department of Agriculture, Economic Research Service, 2015. http://www.ers.usda.gov/ datafiles/Commodity_Costs_and_Returns/Data/Recent_Costs_and_Returns_ Maize/rncmaize.xls. C. Zulauf, N. Rettig, U.S. Have, Farm Input Prices Followed U.S. Crop Prices? Farmdoc Daily, Department of Agricultural and Consumer Economics, University of Illinois at Urbana-Champaign, 2015. C.R. Stoof, B.K. Richards, P.B. Woodbury, et al., Untapped potential: opportunities and challenges for sustainable bioenergy production from marginal lands in the Northeast USA, Bioenergy Res. 8 (2015) 482e501. M. Laser, H.M. Jin, K. Jayawardhana, L.R. Lynd, Coproduction of ethanol and power from switchgrass, Biofuels Bioprod. Biorefining-Biofpr 3 (2) (2009) 195e218. MACS. Maryland’s 2014-2015 Cover Crop Sign-up, Maryland Agricultural Water Quality Cost-share Program, 2014, p. 2. D.G. De La Torre Ugarte, D.E. Ray, Biomass and bioenergy applications of the POLYSYS modeling framework, Biomass & Bioenergy 18 (4) (2000) 291e308.

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024

P.B. Woodbury et al. / Biomass and Bioenergy xxx (2017) 1e11 [26] U.S. Department of Energy, 2016 Billion-Ton Report: Advancing Domestic Resources for a Thriving Bioeconomy, M. H. Langholtz, B. J. Stokes, and L. M. Eaton (Leads), ORNL/TM-2016/160, in: Volume 1: Economic Availability of Feedstocks, Oak Ridge National Laboratory, Oak Ridge, TN, 2016, 448pp.. [27] M. Langholtz, L. Eaton, A. Turhollow, M. Hilliard, 2013 feedstock supply and price projections and sensitivity analysis, Biofuels Bioprod. Biorefining-Biofpr 8 (4) (2014) 594e607. [28] J.A. Larson, B.C. English, D.G.D. Ugarte, R.J. Menard, C.M. Hellwinckel, T.O. West, Economic and environmental impacts of the corn grain ethanol industry on the United States agricultural sector, J. Soil Water Conservation 65 (5) (2010) 267e279. [29] C. Hellwinckel, C. Clark, M. Langholtz, L. Eaton, Simulated impact of the renewable fuels standard on US Conservation Reserve Program enrollment and conversion, Glob. Change Biol. Bioenergy 8 (1) (2016) 245e256. [30] M. Haque, F.M. Epplin, J.T. Biermacher, R.B. Holcomb, P.L. Kenkel, Marginal cost of delivering switchgrass feedstock and producing cellulosic ethanol at multiple biorefineries, Biomass Bioenergy 66 (2014) 308e319. [31] M. Khanna, X. Chen, H. Huang, H. Oenal, Supply of cellulosic biofuel feedstocks and regional production pattern, Am. J. Agric. Econ. 93 (2) (2011) 473e480. [32] E. Gnansounou, A. Dauriat, Techno-economic analysis of lignocellulosic ethanol: a review, Bioresour. Technol. 101 (13) (2010) 4980e4991. [33] Y. Jiang, S.M. Swinton, Market interactions, farmers' choices, and the sustainability of growing advanced biofuels: a missing perspective? Int. J. Sustain. Dev. World Ecol. 16 (6) (2009) 438e450. [34] K. Stephenson, S. Aultman, T. Metcalfe, A. Miller, An evaluation of nutrient nonpoint offset trading in Virginia: a role for agricultural nonpoint sources? Water Resour. Res. (2010) 46. [35] J.E. Compton, J.A. Harrison, R.L. Dennis, T.L. Greaver, B.H. Hill, S.J. Jordan, et al., Ecosystem services altered by human changes in the nitrogen cycle: a new perspective for US decision making, Ecol. Lett. 14 (8) (2011) 804e815. [36] PADEP, The Pennsylvania Chesapeake Bay Strategy, 2015. Accessed March 2016. [37] B.P. Werling, T.L. Dickson, R. Isaacs, H. Gaines, C. Gratton, K.L. Gross, et al., Perennial grasslands enhance biodiversity and multiple ecosystem services in bioenergy landscapes, Proc. Natl. Acad. Sci. U. S. A. 111 (4) (2014) 1652e1657. [38] USDA-nass (U.S. Department of Agriculture, National Agriculture Statistics Service). Accessed October 8th, 2015. https://www.nass.usda.gov/. [39] F.M. Brandi-Dohrn, R.P. Dick, M. Hess, S.M. Kauffman, D.D. Hemphill, J.S. Selker, Nitrate leaching under a cereal rye cover crop, J. Environ. Qual. 26 (1) (1997) 181e188. [40] T.C. Kaspar, D.B. Jaynes, T.B. Parkin, T.B. Moorman, Rye cover crop and garnagrass strip effects on NO3 concentration and load in tile drainage, J. Environ. Qual. 36 (5) (2007) 1503e1511. [41] T.C. Kaspar, D.B. Jaynes, T.B. Parkin, T.B. Moorman, J.W. Singer, Effectiveness of oat and rye cover crops in reducing nitrate losses in drainage water, Agric. Water Manag. 110 (2012) 25e33.

11

[42] D.V. McCracken, M.S. Smith, J.H. Grove, C.T. Mackown, R.L. Blevins, Nitrate leaching as influenced by cover cropping and nitrogen-source, Soil Sci. Soc. Am. J. 58 (5) (1994) 1476e1483. [43] W.F. Ritter, R.W. Scarborough, A.E.M. Chirnside, Winter cover crops as a best management practice for reducing nitrogen leaching, J. Contam. Hydrology 34 (1e2) (1998) 1e15. [44] K.W. Staver, R.B. Brinsfield, Patterns of soil nitrate availability in corn production systems - implications for reducing groundwater contamination, J. Soil Water Conservation 45 (2) (1990) 318e323. [45] K.W. Staver, R.B. Brinsfield, Using cereal grain winter cover crops to reduce groundwater nitrate contamination in the mid-Atlantic coastal plain, J. Soil Water Conservation 53 (3) (1998) 230e240. [46] J.S. Strock, P.M. Porter, M.P. Russelle, Cover cropping to reduce nitrate loss through subsurface drainage in the northern US Corn Belt, J. Environ. Qual. 33 (3) (2004) 1010e1016. [47] M. Jacobson, Z. Helsel, NEWBio Switchgrass Budget for Biomass Production, The Pennsylvania State University, State College, PA, USA, 2014. [48] Barnhart, S., M. Duffy, and R. Owen. Accessed 12 September, 2016. Estimated Costs of Pasture and Hay Production. File A1-15. Iowa State University. Available at: . [49] CBP, Reducing Pollution Indicators Analysis and Methods Documentation. Chesapeake Bay Program, 2015, p. 14 (Excel workbook). [50] USDA-NRCS, Impacts of Conservation Adoption on Cultivated Acres of Cropland in the Chesapeake Bay Region, 2003-06 to 2011, United States Department of Agriculture, Natural Resources Conservation Service, 2013, p. 113. [51] R.Q. Miao, M. Khanna, Are bioenergy crops riskier than corn? Implications for biomass price, Choices Mag. Food, Farm, Resour. Issues 29 (1) (2014). [52] A.K. Jain, M. Khanna, M. Erickson, H.X. Huang, An integrated biogeochemical and economic analysis of bioenergy crops in the Midwestern United States, Glob. Change Biol. Bioenergy 2 (5) (2010) 217e234. [53] G.W. Landers, A.L. Thompson, N.R. Kitchen, R.E. Massey, Comparative breakeven analysis of annual grain and perennial switchgrass cropping systems on claypan soil landscapes, Agron. J. 104 (3) (2012) 639e648. [54] L.K. James, S.M. Swinton, K.D. Thelen, Profitability analysis of cellulosic energy crops compared with corn, Agron. J. 102 (2) (2010) 675e687. [55] H. Kim, P.B. Parajuli, Economic analysis using SWAT-simulated potential switchgrass and miscanthus yields in the Yazoo river basin, Trans. ASABE 55 (6) (2012) 2123e2134. [56] B.E. Sharp, S.A. Miller, Estimating maximum land use change potential from a regional biofuel industry, Energy Policy 65 (2014) 261e269. [57] A. Okwo, V.M. Thomas, Biomass feedstock contracts: role of land quality and yield variability in near term feasibility, Energy Econ. 42 (2014) 67e80. [58] J.F. Chamberlain, S.A. Miller, Policy incentives for switchgrass production using valuation of non-market ecosystem services, Energy Policy 48 (2012) 526e536.

Please cite this article in press as: P.B. Woodbury, et al., Improving water quality in the Chesapeake Bay using payments for ecosystem services for perennial biomass for bioenergy and biofuel production, Biomass and Bioenergy (2017), http://dx.doi.org/10.1016/j.biombioe.2017.01.024