Published March 6, 2014 Crop Economics, Production & Management Profitability of Cellulosic Biomass Production in the Northern Great Lakes Region Bradley J. Kells and Scott M. Swinton* ABSTRACT Producing bioenergy feedstocks on non-crop land can largely avoid the food price feedbacks of energy biomass production on cropland. The U.S. northern tier grassland-to-forest ecotone offers large areas of marginal land that is not currently cropped. In this ecological transition zone, the relative profitability of grassy vs. woody sources of energy biomass is little studied. This paper reports an exploratory investment analysis of cellulosic biomass production in the northern Great Lakes region. It compares two short-rotation tree crops, willow (Salix sachalinensis F. Schmidt)and hybrid poplar (Populus nigra L. X P. maximowiczii A. Henry), and switchgrass (Panicum virgatum L.) (a native prairie grass) to conventional mixed hay. Because biomass markets are not yet well developed, this study calculates threshold prices and yields at which biomass crops become at least as profitable as mixed grass hay. At 2010–2012 prices and available production technologies, none of the cellulosic crops is competitive with the hay baseline system. The breakeven price of energy biomass ranges from $90–100 per oven-dry Mg–1 for all three energy crops. Breakeven yields are much more variable, due to the high cost of harvesting woody biomass. At 2010–2012 prices, necessary biomass yield increases range from 3.5-fold for switchgrass and willow to over 25-fold for poplar. While the ratio of input costs to revenue remains relatively constant between the northern and southern Great Lakes regions, the opportunity cost of active cropland in the southern zone is much higher, implying an economic comparative advantage for marginal land of the northern tier of the Great Lakes region. Public concerns about U.S. energy independence and climate change risks associated with fossil fuel use have given rise to public laws mandating bioenergy use both for liquid transportation fuel and for electricity generation. The U.S. Energy Independence and Security Act of 2007 (EISA) (U.S. Congress, 2007) requires ambitious increases in the use of biofuels for transportation use, while over a score of state-level laws set targets for renewable portfolio standards (including biomass) to produce electricity (USDOE-EERE, 2012). Accompanying the scientific and engineering research into cost-effective innovation in bioenergy production has been a series of economic studies exploring the conditions for commercial viability of a bioenergy industry, starting with the production of energy biomass. As a corn grain ethanol industry already exists, the literature has focused on cellulosic biomass for conversion into ethanol or for direct combustion for electricity. Research has focused chiefly on how current cropland could shift to produce more energy biomass. Studies have ranged from breakeven analysis of threshold biomass prices and yields B.J. Kells, Dep. of Economics, George Mason Univ., Fairfax, VA 22030; S.M. Swinton, Dep. of Agricultural, Food and Resource Economics, Michigan State Univ., East Lansing, MI 48824. All authors are also affiliated with DOE Great Lakes Bioenergy Research Center, Michigan State University. Received 17 Aug 2013. *Corresponding author (swintons@msu.edu). Published in Agron. J. 106:397–406 (2014) doi:10.2134/agronj2013.0397 Available freely online through the author-supported open access option. Copyright © 2014 by the American Society of Agronomy, 5585 Guilford Road, Madison, WI 53711. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. required for profitable production (Mooney et al., 2009; James et al., 2010) to regional and national supply analyses that capture the relative opportunity costs of displacing alternative current crops either at current prices (Egbendewe-Mondzozo et al., 2011, 2013) or by simulating price feedbacks triggered by biomass expansion (Hertel et al., 2010; Khanna et al., 2011). In particular, the EISA law and similar legislation in other nations have increased demand for feedstocks to make liquid biofuels, triggering cropland shifts and grain price rises (Baier et al., 2009). The USDA estimates that more than 35% of the total corn supply grown in the United States will be used for biofuel during the next decade (USDA, ERS, 2012). This increased demand caused at least a 25% increase in corn prices from 2006 to 2008 (Baier et al., 2009; Tyner, 2008). And price increases in corn affect other goods: Crops that compete with corn for land (soybean, wheat), grain-intensive products (beef, poultry), and substitute goods all face increased prices (Baier et al., 2009; Hayes et al., 2009; Tyner, 2008). Higher grain prices have two pernicious effects. First, they reduce food consumption, especially among the world’s poor (Hertel et al., 2010). Second, higher food prices create a market incentive to bring new lands under production, both in the United States and the rest of the world (Chen and Khanna, 2012). More producing land helps mitigate food price rises, but expanding the area of land under cultivation increases emission of greenhouse gasses. This indirect land use change (ILUC) effect of biofuel mandates that are met from current cropland can lead to potentially large greenhouse gas releases along with Abbreviations: NPV, net present value; PLS pure live seed. A g ro n o my J o u r n a l • Vo l u m e 10 6 , I s s u e 2 • 2 014 397 global land degradation (Fargione et al., 2008; Hertel et al., 2010; Khanna and Crago, 2012; Searchinger et al., 2008). These findings have shifted the attention of scientists and policymakers to the potential for non-crop marginal lands to produce energy biomass (Hill et al., 2006; Keoleian and Volk, 2005; Swinton et al., 2011). Energy biomass production on marginal land minimizes competition with cropland and thus avoids putting pressure on crop and cropland prices (Campbell et al., 2008; Hill et al., 2006; Lemus and Lal, 2005). It may also decrease the environmental impact of expanded biofuel production, as many non-crop cellulosic biomass crops are perennial, and so have smaller environmental footprints than cropbased biofuels because of their greater efficiency at C sequestration (Fargione et al., 2008; Lemus and Lal, 2005; Adler et al., 2007), and greater N conversion efficiency (Crutzen et al., 2008). Non-crop marginal land in the United States is largely found where it gets dry enough to make crop production risky in the Great Plains, where it gets cold enough for grain yields to be risky, across the northern tier of the grain and dairy belt, and where soils are relatively unproductive, as in parts of the wooded Southeast (Swinton et al., 2011). In the wake of the declining U.S. pulpwood industry (Ince, 2009), energy biomass production from the wooded regions potentially offers the attractions of low opportunity cost and local economic development. The northern tier of the Great Lakes, notably northern Michigan, Minnesota, and Wisconsin, includes large areas of poor quality, glacial till-based soils covered with scrub brush, forest, and hay for dairy cows (Bos taurus). Land prices tend to be low compared to more productive cropland. It remains an empirical question whether dedicated energy crops are more economically feasible in this setting than in areas of food, fiber, and feed crops. Two types of energy crops are of particular interest: perennial grasses and short-rotation woody biomass crops like willow and poplar. Studies have shown perennial grasses like switchgrass and Giant Miscanthus (Miscanthus × giganteus J.M. Greef & Deuter ex Hodkinson & Renvoize)can give very high biomass yields on good quality cropland. On the other hand, willow and poplar may be better adapted to the poorer quality soils and short growing season of the northern tier. Tree crops have the added advantage of storing their own biomass until needed, as compared to grasses that must be harvested annually and stored until use. Emerging results from agronomic and forestry experimental sites in the northern tier of Michigan and Wisconsin paired with cost of production estimates create the opportunity to compute biomass yield and price thresholds for these dedicated biomass crops compared to current land-income opportunities in the area. In analyzing returns to investments in dedicated energy crops compared to alternative long-term investments in forage crops, this paper develops improved, dynamic analyses of breakeven comparative prices and yields. OBJECTIVES Assuming that land owners attempt to maximize the expected profitability of their land, this study aims to estimate the relative profitability of converting land from mixed hay (baseline system) to produce switchgrass, hybrid poplar, or willow. The specific research objectives of this study are as follows: 398 1. To estimate costs of production and potential returns of each crop compared to a current baseline system. 2. To develop an appropriate method for comparative breakeven analysis in a dynamic setting where yields evolve over time. 3. To compute the biomass price threshold necessary for each biomass system to break even with the profitability of the baseline system. 4. To compute the biomass yield threshold necessary for each biomass system to break even with the profitability of the baseline system. 5. To compare the profitability of energy biomass production in the northern Great Lakes to recent results from the southern part of this region. MATERIALS AND METHODS This study examines the potential profitability of expanded energy biomass production from switchgrass, willow, and hybrid poplar in the northern Great Lakes region. The general approach is to characterize likely commercial production systems and to evaluate their profitability using capital budgeting (investment analysis) over a 16-yr time horizon. None of these crops are currently produced commercially for energy use, though all are being researched for energy use and all have other current uses (e.g., paper pulp and wood products, wildlife habitat). Profitability is evaluated both individually (do likely revenues cover likely costs?) and comparatively (can net revenues compete with the best current alternative land use?). In Michigan’s Upper Peninsula in particular and on marginal land in the northern Great Lakes region in general, a common current land use is mixed hay production. As a means to represent existing revenue streams from marginal lands in the region, we chose a mixed grass hay production system. This rotation serves as the baseline land use against which energy biomass systems are compared. Production Practices and Yield Levels The basic production systems to be evaluated are synthetic systems built up from the best available research data and commercial production data to represent likely commercial production systems for energy biomass. They are described below, beginning with the mixed grass hay baseline system. A complete agronomic management protocol was created for each cropping system as it would be practiced under commercial conditions (Table 1). These protocols were a mix of current research findings and best practice guides based on commercial production information for each individual crop. The associated production costs are summarized in Table 2. Mixed hay: The mixed hay production guide was based on Michigan State University (MSU) crop production budgets (Stein, 2011a). Because these budgets were created for central and southern Michigan, the data was adjusted to fit northern Great Lakes conditions (D. Min, Extension Forage Specialist; C. Kapp, Research Technician, personal communication, Michigan State University (MSU), Upper Peninsula Research Station (UPRS), July 2012). In the northern Great Lakes region, many farmers who use marginal land currently grow a cool weather mixed grass hay (Corace et al., 2009) (D. Undersander, Professor, University of Wisconsin-Madison, personal communication, January 2013) (C. Shaeffer, Professor, personal communication, Univ. of Agronomy Journal • Volume 106, Issue 2 • 2014 Table 1. Agronomic protocol. Crop Mixed grass hay Willow Poplar Switchgrass Planting material per hectare per planting 15 15,650 2,170 9 Unit kg PLS† ha–1 cuttings ha–1 cuttings ha–1 N 0 0 0 P 0 0 0 K 0 0 0 kg PLS ha–1 120 0 0 Pest control None Oxyflourfen 2.5L ha–1, simazine 2.5L ha–1 Pendimethalin 5L ha–1 Glyphosate 5L ha–1, 2, 4-D 2.5L ha–1, imazetherapyr 283 g ha–1, atrazine 2.5L ha–1, dicamba 1.25L ha–1 † PLS, pure live seed. Minnesota, February 2013) (C. Kapp, personal communication, January 2013). Our budgets used a grass mix consisting of tall fescue (Festuca arundinacea Schreb.) and red clover (Trifolium pratense L.), planted with a mechanized seed drill. There is typically no replanting. Since most farmers in the northern Great Lakes region aim to keep input costs low, fertilizer is applied in small amounts, and thus we assume the application of 19–19–19 fertilizer four times over the 16-yr time horizon of these capital budgets. Harvest is yearly. Pesticides are rarely applied to mixed grasses, and therefore do not figure into our model. Willow: Willow is a short-rotation woody biomass crop under intensive research in the northern Great Lakes region. The willow production system described here is based on the State University of New York Environmental Science and Forestry Willow Biomass Producer’s Handbook (Abrahamson et al., 2010) for most basic data and for the management regime, adjusted to northern Great Lakes conditions by regional experts (R. Miller, Director; B. Bender, Operations Forester, MSU, Forestry Biomass Innovation Center (FBIC), personal communication, July 2012). Woody biomass in this region is rarely produced with the application of N, so our willow budget excludes fertilizer, but it includes weed control using glyphosate (N-(phosphonomethyl)glycine), 2,4-D (2,4-dichlorophenoxy) acetic acid), oxyflourfen (2-chloro-1-(3-ethoxy-4-nitrophenoxy)4-(trifluoromethyl)benzene), and simazine (6-chloro-N,N9diethyl-1,3,5-triazine-2,4-diamine) (Table 1). Current research suggests that coppiced willow production is the most efficient means of producing biomass from willow, with harvests occurring every 4 yr to keep biomass growth at its most efficient (R. Miller; Brad Bender, personal communication, August 2012). Variety trials are underway, and yields are consistently improving, but SX-61, the varietal selected for this study, is a current leader in biomass yield production at MSU’s Forest Biomass Innovation Center (FBIC) in Escanaba, MI (Wang and MacFarlane, 2012). As willow is harvested by coppicing, no replanting is necessary. Land preparation includes mowing the field, spraying with 2,4-D and glyphosate, and tilling, before planting with a mechanical Egedal planter. Hybrid poplar: Hybrid poplar, another short-rotation woody biomass crop, follows a longer rotation and different harvest regime than willow, because it is harvested by whole tree cutting as opposed to coppicing. The poplar budgets are based on the University of Minnesota’s hybrid poplar best management practices (Zamora et al., 2011), modified by regional experts (R. Miller; B. Bender, personal communication, July 2012). Poplar is harvested after 8 yr and then replanted for a second harvest 8 yr later, at the end of the 16-yr time horizon. The variety chosen for our budgets, NM-6, is a popular poplar clone with good yields (Wang and MacFarlane, 2012) and good disease resistance. Poplar yields are lower than willow, but because they continue increasing over a longer time period poplar can be harvested less frequently. Fertilizer is not used in this region for wood production. Herbicides applied to poplar include glyphosate; 2,4-D; and pendimethalin (N-(1-ethylpropyl)-3,4-dimethyl-2,6dinitrobenzenamine). Land preparation is the same as for willow, and the poplar cuttings are planted with the same equipment. Switchgrass: Switchgrass is the only non-woody biomass crop being researched extensively for production in the northern Great Lakes region. A perennial grass crop, switchgrass is harvested annually, beginning the year after planting and continuing for 10 yr, with a replant in Year 11 taking production to the end of our 16-yr time horizon. The varietal Cave-In-Rock can withstand the cold winters of the region while maintaining high yields. Unlike the two woody crops, switchgrass requires N fertilizer. To aid with seedling establishment, it is treated after each planting with 2,4-D, imazetherapyr (2-[4,5-dihydro4-methyl-4-(1-methylethyl)-5-oxo-1H-imidazol-2-yl]-5-ethyl3-pyridinecarboxylic acid), atrazine (6-chloro-N-ethyl-N9(1-methylethyl)-1,3,5-triazine-2,4-diamine), and dicamba (3,6-dichloro-2-methoxybenzoic acid) (Min, 2011)). Switchgrass is assumed to be planted with a seed drill into a glyphosate-killed and plowed field. Yield data, summarized in Table 3, comes from the FBIC at Escanaba, MI, the Upper Peninsula Research and Extension Center at Chatham, MI, and the National Agricultural Statistics Service. Non-alfalfa hay is produced at a rate of about 3.5 Mg ha–1 in the northeastern counties of Wisconsin (USDA, NASS, 2010– 2012), a number that matches average hay yield in Michigan’s Upper Peninsula (K. Hillock-Vining, Chippewa County Farm Service Agency Executive Director, Sault Ste Marie, MI, personal communication, May 2012). The willow yields reported in Table 3 are based on reported yields at the Forest Biomass Innovation Center in Escanaba, MI, where they use similar harvest methods to the protocol here. Yields Table 2. Crop production input prices and data sources. Input Mixed grass seed (red clover/tall fescue) Switchgrass seed Willow Poplar Unit kg PLS† kg PLS Cutting Cutting Price $ 4.2/3.5 15.5 0.1725 0.16 Source Deer Creek Seed, Ashland, WI; De Bruyn Seed, Zeeland, MI Ernst Seed Co, Meadville, PN Double A Willows, Fredonia, NY Hramor Nursery, Manistee, MI † PLS, pure live seed. Agronomy Journal • Volume 106, Issue 2 • 2014 399 Table 3. Crop yields, prices, and data sources. Output Hay Yield Green Mg ha–1 yr–1 3.5 Price $ Mg–1 115 Source of yield value Source of price value Chippewa County Farm Service Agency, Phone Interview, May 2012. Wang and MacFarlane, 2012 Willow 20 23 Poplar 16 23 Wang and MacFarlane, 2012; Netzer et al., 2002 Switchgrass 10 38 Min, 2011; switchgrass for a bioenergy crop and livestock feed are reported in green Mg–1, equal to harvest weight. Because dry weight varies from one species to another, yields elsewhere are also reported in oven-dry Mg–1. Recent improvements in yield, documented in Wang and MacFarlane (2012), demonstrate willow’s potential to reach a mature yield of over 20 green (10 dry) Mg ha–1 per year. Though we calculate willow yields as only reaching maximum yield after the first coppice, this still leads to three harvests over 11 yr at full production capacity, with a total yield of more than 130 dry Mg ha–1 over 16 yr. Poplar yields of 16 green (8 dry) Mg ha–1 per year also come from Wang and MacFarlane (2012), adjusted to an 8 yr harvest cycle (R. Miller, B. Bender, personal communication, July 2012). The mature yield for poplar represents the total yield over 8 yr averaged out across the entire cycle. After this 8-yr period, yields begin declining at such a rate that failure to harvest will result in overall loss of profit (R. Miller, B. Bender, personal communication, August 2012). Switchgrass yields come from field trials conducted at the MSU experiment station in Chatham, MI. There is no current regional production of switchgrass by private individuals, but the yield data from Chatham is representative of the potential achievable in the northern Great Lakes region. Yields of 10 green Mg ha–1 have been achieved with the Cave-In-Rock varietal, the currently recommended varietal for the region (Min, 2011). Sensitivity analyses were conducted for each of the proposed biomass production methods to account for a range of outcomes in the northern Great Lakes region. For each biomass crop there is an “optimistic” and a “pessimistic” scenario in addition to the baseline. The optimistic scenarios assume a mix of yields, planting and harvest regimes, and chemical and fertilizer costs that favors profitability. The pessimistic scenarios entail higher costs and lower yields than the baseline. Details are included in Table 4. The optimistic willow scenario assumes that the trees require no fertilization, and they can grow for 4 yr at maximum productivity, requiring only four harvests during the budgeted period. The pessimistic willow scenario assumes that harvest is required every 3 yr instead of 4 to maintain its most efficient growth rate, which leads to five harvests over 16 yr. Application of 100 kg ha–1 of N is assumed to be required after each harvest in the pessimistic model (Abrahamson et USDA NASS survey data, 2010–2012. NRC 2011, modified to regional conditions. Price based on $45 dry Mg–1 converted to wet weight. NRC 2011, modified to regional conditions. Price based on $45 dry Mg–1 converted to wet weight. NRC 2011, modified to regional conditions. Price based on $45 dry Mg–1 converted to wet weight. al., 2010). The poplar optimistic and pessimistic scenarios also differ by whether fertilizer is required (Zamora et al., 2011). Poplar weed control costs also differ between scenarios, as the optimistic poplar model assumes a lighter chemical regime than the pessimistic poplar model (which, in addition to the chemicals previously discussed, includes additional applications of fluazifop {(±)-2-[4-[[5-(trifluoromethyl)-2-pyridinyl]oxy] phenoxy]propanoic acid}, imazaquin {5-dihydro-4-methyl-4-(1methylethyl)-5-[oxo-1H-imidazol-2-yl]-3-quinolinecarboxylic acid}, and oxyfluorfen). The optimistic model also assumes hand-planting instead of machine planting (R. Miller, personal communication, May 2013): due to the high fixed cost of currently available machinery, producers can hand plant low cutting-per-acre woody crops like poplar more cheaply than by mechanized methods. The optimistic switchgrass scenario assumes that switchgrass will survive the entire 16-yr time horizon of the study, whereas the pessimistic scenario assumes replanting every 6 yr to maintain high productivity, meaning two replantings over the 16 yr. The two scenarios also differ in weed control costs, as the pessimistic one requires herbicides and fertilizers to be applied at every planting, with a heavier rate of N used in the pessimistic scenario than in the optimistic one. Input Costs Whenever possible, the costs used in this study’s budgets are drawn from local primary sources. The herbicide prices are averages over the 2010–2012 period from Great Lakes Agri Service, a local chemical dealer (J. Sergant, Great Lakes Agri Service, Gladstone, MI, personal communication, July 2012). Fertilizer prices are also derived from a regional company (Ray’s Feed Mill, Bark River, MI, personal communication, August 2012) and adjusted to price per elemental pound. Seed and propagule prices are more complicated, but also based on primary data. Willow cutting and planting costs are derived from estimates given by a leading producer in New York (Sue Rak, Double-A Willow, Fredonia, NY, personal communication, July 2012). Costs for poplar cuttings come from a major Michigan Nursery (Hramor Nursery, Manistee, MI, www.hramornursery.com/ pricelist/hrab2012fall-2013spring.pdf, July 2013). Mixed grass hay seed is an average of regional prices (Debruyn Seed Co, Zeeland, Table 4. Sensitivity analysis differences by model. Willow optimistic Willow pessimistic Poplar optimistic Poplar pessimistic Switchgrass optimistic Switchgrass pessimistic 400 4 yr between harvests. No fertilizer applied. 3 yr between harvests. 100 kg ha–1 N applied after each coppice. No fertilizer applied. Light chemical regime: Only 2,4-D, glyphosate, and pendimethalin applied. 50 kg ha–1 N applied every year. Assumes “light chemical regime” plus imazaquin, fluazifop, and oxyflourfen. No replantings required. 65kg ha–1 N per growing year Two replantings, every 6 yr. 150 kg ha–1 yr–1 N per growing year Agronomy Journal • Volume 106, Issue 2 • 2014 MI, www.debruynseed.com/new_page_5.htm, February 2013) (Deer Creek Seed, Ashland, WI, www.deercreekseed.com/forageseed/, February 2013). Cave-In-Rock switchgrass seed prices were acquired from a regional dealer (Ernst Seed Co, Meadville, PA, personal communication, July 2012). Machine costs are charged as custom hire operations, instead of separately accounting for all machinery ownership and operating costs. This assumption recognizes that expensive planting and harvest machinery for willow and poplar would typically not be efficient for ownership by any farming operation that does not function on a very large scale. For the more common pieces of equipment, this study based its custom costs on Michigan State University’s 2011 estimates for custom machine costs (Stein, 2011b). Certain pieces of equipment are too rare to have published custom hire costs, such as the Egedal planter, a machine that plants woody biomass cuttings and is used in our study for poplar and willow planting. In this case, we calculated custom hire equivalent cost from ownership costs, assuming a 5% rate of return to capital (Erickson et al., 2004), plus skilled equipment operator labor at $17/hour (Stein, 2011b). Poplar was assumed to be harvested on a whole tree system where trees are harvested in a manner similar to logging operations, so costs per acre were found for clean-cut-harvesting softwood plantations, divided into felling, skidding, and chipping costs. Felling, skidding, and chipping costs are based on average costs of forestry equipment use (A. Srivastava, D. Abbas, C. Saffron, and F. Pan. 2011. Economic analysis of woody biomass supply chain logistics for biofuel production in Michigan, Final report. Michigan State Univ., Dep. of Biosystems and Agric. Eng., East Lansing), modified to include the same custom harvesting profit rate assumed by Stein (2011b). Willow harvest costs are based on the State University of New York Environmental Science and Forestry Department’s (SUNY-ESF) EcoWillow model (SUNY-ESF, Syracuse, NY, www.esf.edu/willow/download.asp, June 2012), available online, and include cutting, chipping, and in-field transport. The EcoWillow model is built around specialized harvesting equipment developed by SUNY and Case New Holland, but due to the lack of willow harvesting data with conventional machinery, it provides the best available data on potential willow harvest costs. Transport costs, calculated at 20 miles of shipping, differed between the woody and grassy biomass crops, and separate costs for each, derived from Searcy et al. (2007), form the basis for this study’s assumed transport costs. Annualized costs for each production system are reported in Table 5. All cost calculations for switchgrass, poplar, and willow are calculated on a per green Mg–1 basis. Drying of the wood occurs at the very end of the process, so all the other steps (harvest, skidding, chipping) occur with raw green biomass. This means that the extra weight of green material adds to the costs of production. Storage costs are omitted, though Brechbill et al.’s (2011) literature review reported a median weight loss of 8.8% over 6 mo storage for plastic-wrapped round bales of switchgrass. While lower storage loss values are likely for woody species, no data were available. Output Prices The mixed hay baseline price of $115 Mg–1 is the average of prices paid in the 2010–2012 period for non-alfalfa hay across Michigan (USDA, NASS, 2010–2012). While in most of Michigan’s Upper Peninsula, hay is consumed on-farm and not sold, this price represents the price that could be received if farmers desired to sell their product. Since switchgrass, poplar, and willow are being analyzed as bioenergy sources, their value is based entirely on the amount of dry biomass they produce per hectare per year. For willow and poplar, the weight ratio of green to oven-dry biomass is 2:1 (50% dry matter). Switchgrass is 85% dry matter when harvested after senescence (K. Thelen, Professor, MSU, personal communication, September 2012). Woody biomass grown has alternative uses as fuelwood and pulpwood, worth $35 to $55 dry Mg–1 in the southern United States (National Research Council, 2011). As a reference price received at the biorefinery, we use the median of this range, $45 dry Mg–1 of biomass. As noted above, storage losses are omitted from the analysis. In sensitivity analysis, we also calculated the profitability of each biomass crop at $30 and $60 per oven-dry Mg–1. Capital Budgeting for Relative Profitability Analysis This study develops capital budgets for each system over a period of 16 yr, using them as the basis for calculating comparative breakeven yields and prices relative to the benchmark production method of mixed grass hay. The capital budgeting approach uses discounted cash flows, following Boehlje and Eidman (1984), based on revenues minus the costs that vary across production systems (CIMMYT, 1988). Costs that are the same across systems, such as land rental rates, are omitted because they do not affect the relative profitability of the systems included. While they would affect a simple breakeven price or yield calculation, they do not affect the comparative breakeven values calculated here. The capital budgets are based on 2010–2012 prices prevailing in the northern Great Lakes region. Comparative Breakeven Cellulosic Feedstock Price Analysis for Changing Crops The goal of this analysis is not simply to calculate the minimum returns from raising a biomass crop that covers its direct costs of production, but rather to calculate the level of returns that would cover both those direct costs plus the opportunity cost of giving up net income earned from the traditional dominant crop regime in the region. Comparative breakeven analysis calculates the biomass price or yield that would be necessary for the producer to earn as much from the energy biomass crop as from traditional crop production (mixed grass hay, in this case). This study also estimates breakeven costs Table 5. Present value of input costs by crop and by category, per hectare, over 16 yr. Crop Hay Willow Poplar Switchgrass Harvest $1519 $2280 $3160 $2980 Machine $80 $300 $510 $310 Pest control $0 $110 $280 $150 Agronomy Journal • Volume 106, Issue 2 • 2014 Fertilizer $680 $0 $0 $1860 Planting material $50 $2570 $560 $210 Total $2340 $5250 $4510 $5510 401 for both optimistic and pessimistic models, giving a range of likely prices necessary for biomass crops to equal the economic value of mixed hay production to a potential biomass farmer. The discount rate applied is 5%, based on Erickson et al. (2004), who found that to be a typical real rate of return to capital on U.S. farms. Assuming constant real prices over the 16-yr period, results are reported in 2010–2012 dollars (present value at time of study). To calculate annualized costs, revenue, and net profit from present values, a standard financial annuity formula was used over the 16 yr sum of net returns. The formula used (Ross et al., 2008) is as follows: æ r(1 + r )t ÷ö ÷ A = NPV çç çè (1 + r )t -1÷÷ø [1] where A is annuity, NPV is net present value of production system, r is the discount rate (5%), and t is time horizon (16 yr). Due to the long time horizon used in this study, the year that harvest occurs has a significant impact on the profitability of any operation. While hay and switchgrass are harvested yearly, and thus generate yearly revenue, willow and especially poplar are harvested at much longer time intervals. Thus the present value of revenue from these tree crops is reduced significantly when calculated as a present value. To calculate comparative breakeven prices and yields, this study innovates on the comparative breakeven methodology for energy biomass crops used in James et al. (2010). The methodology used in that work calculated breakeven yield as the annual average yield that would be required, assuming a static biomass price and a 5% discount rate, to match the annualized NPV of the status quo crop. The method is sound if crop rotations are balanced over time, meaning that the same area is planted each year, so that average yields represent a weighted average over time. However, this approach tends to underestimate the comparative breakeven yields of perennial crops when plantings are not evenly spaced over time and harvests are infrequent. The underestimation is slight for crops that are harvested frequently, such as perennial grasses. The distortion is much larger for woody crops with intermittent harvests, when no biomass is harvested in many years but large yields occur in a few years. To correct for this time-dependent distortion when new plantings are not evenly spaced over the entire time horizon, this study develops a modified methodology to calculate comparative breakeven price relative to a time-discounted yield, and breakeven yield relative to a time-discounted biomass price. We calculate breakeven price (P BE) as a function of the NPV of a “defender” crop (NPV D) plus the cost (ct) of producing the new biomass crop, divided by an assumed biomass yield (y). Because P BE depends on the timing of when yields are harvested, the biomass yield of the “defender” crop (yDt) must discounted and subtracted from the discounted yield achieved by the “challenger” biomass crop (yCt). This ensures that the biomass yield numbers of both crops receive the same temporal weighting. The resulting equation becomes: 402 PBE = æ c ö NPVD + å t çç t t ÷÷÷ èç (1 + r ) ø÷ æy -y ö å t èççç (1Ct+ r )Dtt ø÷÷÷÷ [2] Note that unlike the James et al. (2010) study, in which the defender crop of corn for grain could also produce stover for energy biomass as a byproduct, in the current study, the defender rotation of mixed grass hay produces no additional energy biomass, so yDt = 0. The total breakeven yield (TY BE) is also calculated as a function of the NPV of the “defender” crop (NPV D) plus the cost of producing the new biomass crop (ct), divided by an assumed biomass price (Pt) that is discounted. Calculations are again made for both optimistic and pessimistic models. To simplify calculation, yield dependent costs (ydct) per Mg–1 are subtracted from the assumed biomass price, while acreage dependent costs (adct) are kept in the numerator. The resulting equation becomes: æ adct ÷ö NPVD + å t çç ÷ çè (1 + r )t ÷÷ø TYBE = æ P - ydc ö å t çççè (1t + r )t t ÷÷÷÷ø [3] This method gives a correctly discounted average total yield over the entire time horizon. For ease of interpretation, we convert the value to mature annual yield, by taking TY BE (total breakeven yield) and dividing it by the sum of the ratios of current year yield to mature yield for each year (Rt) (e.g., if yield in the first year is half of mature yield, then R1 = 0.5), the breakeven mature annual yield (YBE) can be calculated as follows. YBE = TYBE å t ( Rt ) [4] The breakeven mature annual yield is suitable target value for researchers seeking to develop new varieties that can break even with the “defender” cropping system at the assumed biomass price. Fig. 1. Annualized net present value for energy biomass crops and mixed grass hay ($ ha –1). Agronomy Journal • Volume 106, Issue 2 • 2014 RESULTS Investment Analysis: Relative Profitability Using a biomass price of $45 dry Mg–1, relative profitability was calculated for the mixed grass hay system and each of the three dedicated biomass crop systems. We calculate the average annual net revenue over selected costs of a mixed grass hay rotation at $200 ha–1. In the comparative breakeven budgets, this value is treated as the opportunity cost of giving up mixed grass hay. The three energy biomass systems not only failed to match the profitability of mixed grass hay, but they also failed to cover their own direct variable costs. The annualized net present values per hectare of each crop are illustrated in Fig. 1. Comparative Breakeven Prices for Replacing Mixed Grass Hay Using the average measured yields collected for this study, the comparative breakeven prices indicate that the farm-gate prices of $45 dry Mg–1 that prevail for pulpwood (~$22.50 green Mg–1 for wood) would be insufficient to equal the net return from the production of mixed grass hay. Willow has the highest breakeven price, at $100 Mg–1, while poplar and switchgrass both have breakeven prices around $92 Mg–1. Although switchgrass has lower yields and higher costs than poplar or willow, it manages to remain price competitive due to the frequency of its harvests, causing its revenues to face less discounting over time. Poplar has significantly lower costs than willow or poplar, as the higher costs of N fertilization of switchgrass and planting of willow more than offset the comparatively greater harvest costs of poplar. Comparative Breakeven Yield for Replacing the Mixed Grass Hay Baseline Assuming a price of $45 dry Mg–1 of biomass, the estimated mature yields necessary for biomass production to break even with net returns from hay production are shown in Fig. 2. Switchgrass and willow have significantly lower breakeven prices than poplar, due to the much higher yield dependent costs of poplar harvest. Since poplar has a longer growth period between harvests than willow, the wood to be harvested is significantly thicker than willow and thus requires more expensive equipment. Harvesting poplar, therefore, is significantly more expensive than harvesting willow or switchgrass, which results in significant differences in breakeven yields, even when breakeven prices are not extremely different. Sensitivity Analysis Breakeven prices differ by $30 to $40 dry Mg–1 per year between the optimistic and pessimistic scenarios. Even in the optimistic models, the breakeven prices are double the assumed current biomass price dry Mg–1, while in the pessimistic models, they are much higher. The problem is exacerbated in the breakeven yield calculations, where even the optimistic willow and switchgrass mature yields are about three times the current yield. The poplar budgets, due to their much higher yielddependent costs, would require a nearly 25-fold yield increase, even under the optimistic scenario (Tables 6 and 7). In Table 8, we compare annualized NPV results at optimistic and pessimistic biomass price forecasts. At a price of $30 dry Mg–1, all three biomass crops face annualized NPV losses of between $190 and $270 ha–1. At $60 dry Mg–1, poplar covers with its own variable costs, but does not make enough additional profit to compensate for the opportunity cost associated with switching away from a current land use such as mixed grass hay production. Willow and switchgrass both fall short of covering their own variable costs. Even at the optimistic biomass prices, all of the biomass crops fall more than $150 short of breaking even with the $200 ha–1 earning potential (opportunity cost) of a traditional mixed grass hay rotation. DISCUSSION Breakeven Analysis Harvest and planting costs play important roles in limiting the potential profitability of biomass production in the northern Great Lakes region. Harvest costs offset more than 65% of the expected revenue from the three biomass crops (Fig. 3). This stands in stark contrast to hay production, where <40% of revenue is needed to cover harvest costs. Hence, although breakeven prices only exceed the baseline biomass price by 100 to 150%, breakeven yields generally require an increase of more than 300% from baseline yields. This problem is most extreme Table 6. Sensitivity analysis of breakeven biomass prices under optimistic, baseline, and pessimistic scenarios for three energy crops ($ dry Mg –1) compared to mixed grass hay.† Crop\Scenario: Willow Poplar Switchgrass Optimistic 95 84 79 Baseline 102 92 93 Pessimistic 116 123 108 † NB: For green weight values, divide by 2 for willow and poplar and by 1.18 for switchgrass. Table 7. Sensitivity analysis of breakeven biomass yields by optimistic, baseline, and pessimistic scenarios for three energy crops (dry Mg ha–1 yr–1).† Crop\Scenario: Willow Poplar Switchgrass Optimistic 48 212 30 Baseline 48 225 36 Pessimistic 67 327 41 † NB: For green weight values, multiply by 2 for willow and poplar and by 1.18 for switchgrass. Table 8. Sensitivity analysis of annualized net present value of three biomass crops at low, baseline, and high prices ($ ha –1 yr –1).† Fig. 2. Comparative breakeven mature yields of three biomass crops compared to mixed grass hay (dry Mg ha –1 yr –1). Crop Willow Poplar Switchgrass Agronomy Journal • Volume 106, Issue 2 • 2014 Low $30 dry Mg–1 (261) (193) (255) Baseline $45 dry Mg–1 (180) (113) (163) High $60 dry Mg–1 (37) 29 (1) 403 Fig. 3. Harvest cost as percent of total revenue. in poplar, where harvest costs account for more than 90% of revenue at baseline prices. The cost of planting willow is the other influential cost driving profitability findings. With a planting rate of over 15,000 cuttings per hectare at a cost of more than 17 cents per cutting, planting materials account for almost 50% of total willow costs (see Table 5). In comparison, planting costs make up only 2% of total costs for the hay baseline. Indeed, apart from the costs of planting materials, the willow scenarios have total costs only $400 greater than those of the hay rotation. But when planting material costs are included, the willow cost per hectare is almost $3,000 dollars greater than the total cost of a mixed grass hay rotation. Comparing Productivity in the Northern and Southern Great Lakes Offsetting the potential advantages of producing biomass on marginal lands—lower land cost and lack of food price feedbacks with their associated indirect land use effects on climate—the chief potential disadvantage is lower productivity. Mooney et al. (2009) found that switchgrass grown on marginal land in Tennessee had a higher simple breakeven price than switchgrass grown on cropland, implying that marginal land may not be a cost-effective means of increasing cellulosic biomass output. To analyze the claim that biomass production on marginal land in the northern tier of the Great Lakes region will have a comparative advantage to production on higher quality land in the southern tier, this study compared its northern tier budgets to those produced by James et al. (2010) studying biomass production on cropland in southern Michigan. To make a reasonable comparison, the James et al. (2010) budgets were adjusted to incorporate the price and cost parameters, the 16-yr time horizon, and the same breakeven calculation method (Eq. [2]–[4]) as in this paper. A number of details were changed in the James et al. (2010) numbers to fit the new situation. Poplar planting costs, instead of using hand planting costs, use the costs of the Egedal planter. Received biomass prices are standardized across the two studies, both using the $45 Mg–1 biomass price. Switchgrass costs are increased to account for the assumption that harvest will occur while the material is not yet fully dry. For costs that do not vary across regions (machine costs, imported seed, or cuttings), the prices used in this paper were used across the board. For costs that do differ (fertilizer, chemical application, transport) the numbers from the James et al. (2010) budgets are used. Since the procedure used for harvesting poplar and switchgrass does not 404 significantly differ from one area to another, most machine costs remain constant across the two budgets. In adjusting yields, one significant change that had to be made was to adjust the James et al. (2010) poplar yields up from the original values that came from the Forest Biomass Research Center in Escanaba, MI, to more recent yield data from southern Michigan. Yields currently being achieved at the Tree Research Center in East Lansing are currently in the range of double those at the FBIC (P. Bloese, Research Manager, Tree Research Center, MSU, personal communication, August 2012). This makes a significant difference in the relative profitability of poplar, and also sets poplar apart from switchgrass, which has roughly comparable yields in both the northern and southern Great Lakes region. To calculate the opportunity cost in the James et al. (2010) budgets, corn prices had to be increased to account for the recent increased value per hectare of growing corn. The price of corn has increased significantly, from $3.50 bu–1 in 2007–2009 (James et al., 2010) to $6.25 bu–1 in 2010–2012 (USDA, NASS, 2010–2012), resulting in a significant increase in land opportunity cost for producing biomass. So, while the ratio of variable cost: revenue is fairly constant across the northern and southern Great Lakes regions (that is, costs and revenue change proportionately with latitude), the opportunity cost of land decreases faster than predicted profitability as production moves farther north. Thus, while biomass production is not currently profitable in either region, production of biomass in the northern Great Lakes region has the advantage of being comparatively less unprofitable. To show the comparative advantage of the northern Great Lakes region for biomass production, a breakeven analysis was run for both of the two regions, using the adjusted southern Great Lakes budgets, for both poplar and switchgrass. At 2010–2012 prices the necessary breakeven price for the Lower Peninsula was higher than that of the Upper Peninsula (about $11 green Mg–1 ($22 dry Mg–1) higher for poplar, and about $30 green Mg–1($35 dry Mg–1) higher for switchgrass). The difference was even more stark in the breakeven yields: Compared to baseline yields, poplar in the southern Great Lakes region required a percentage increase that was 45% greater than in the northern Great Lakes region; for switchgrass, the increase relative to baseline was 66% greater. Results of the comparative analysis of the northern to the southern Great Lakes region is reported in Tables 9 and 10. A key factor driving the comparative breakeven analyses between the two regions is the land use assumed to be the baseline “defender” case, corn for grain in the southern Great Lakes region and mixed grass hay in the north. The reference price corn nearly doubled from the 2007–2009 period used in the James et al. (2010) to the 2010–2012 period used in the comparative study reported here, rising proportionately much more than the price of mixed hay. In fact, with corn at its $3.50 bu–1 mean price of 2007–2009 but all other prices at 2010–2012 levels, the southern Great Lakes region would have lower biomass breakeven prices and yields than the north (sensitivity analysis available from authors). CONCLUSIONS Production of energy biomass on marginal land in the northern Great Lakes region is not currently competitive with conventional mixed hay at current prices and yields. Even the Agronomy Journal • Volume 106, Issue 2 • 2014 optimistic price of $60 dry Mg–1 is not enough for returns from production of energy biomass to break even with returns to alternative uses of land, as represented by mixed grass hay production. However, biomass production on marginal lands in the northern Great Lakes region is relatively more attractive than it is on cropland in the southern Great Lakes. Although biomass yields decline with the move to more marginal, northerly sites, at recent prices of alternative crops (mixed grass hay in the north; corn grain in the south), the decline in income flows from alternative uses of these lands is proportionately greater, giving the northerly marginal areas greater profit potential. There are two cost bottlenecks where targeted research could sharply trim production costs. The first involves automation of harvest. The equipment needed to harvest short rotation woody crops cost effectively is not yet in common usage. While companies like John Deere and Case New Holland are partnering with universities to work on producing effective biomass harvesting equipment (Abrahamson et al., 2008, Meadows et al., 2010), the resulting equipment is still experimental. Using the poplar budget as an example, if poplar harvest costs could be reduced from 95% of each dollar of revenue to 70% (the percentage of each dollar revenue from willow lost to harvest costs using specialized harvesting equipment), the breakeven yield could be reduced from almost 225 dry Mg–1 to just over 25 dry Mg–1 (55 green Mg–1, or less than a 250% increase over current yields), and breakeven price can be reduced from more than $90 dry Mg–1 to <$80 Mg–1. Such mechanical innovations have succeeded before. In 1960 professors at the University of California-Davis, working with a local equipment manufacturer, released a commercial tomato harvester that reduced tomato harvesting cost by nearly half. When this invention was combined with a new tomato variety that could withstand the force of mechanical processing, the labor requirements and overall cost of producing tomatoes decreased by 92%. This cost reduction was partially responsible for the quadrupling of tomato production in California from 1960 to 1990 (Thompson and Blank, 2000). Similar discoveries are possible in the woody and grassy biomass industries, and could have similar effects on production costs. The second bottleneck is the cost of willow cuttings as propagules, made especially costly in high density plantings. More efficient propagation methods and/or intensified price competition among firms could reduce costs. Technological innovation aside, another means to increase the potential profitability of energy biomass from short rotation Table 9. Northern vs. southern tier of Michigan: Comparative breakeven prices ($ dry Mg –1).† Crop Poplar Switchgrass North $92 $93 South $114 $128 † NB: For green weight values, divide by 2 for willow and poplar and by 1.18 for switchgrass. woody crops would be to develop production systems with multiple products, rather than to expect them to turn a profit on energy biomass alone. Drawing a lesson from the relative profitability of biomass byproducts from grain crops like corn stover and wheat straw (Egbendewe-Mondzozo et al., 2011), the challenge would be to separate out higher value timber products. Short rotation woody plantations might, over a growing cycle, produce biomass for bioenergy, pulp for area paper mills, and veneer bolts. This process has not yet been well studied, but a three step thinning cycle, with the first thinning being harvested for biomass, the second thinning harvested for pulpwood, and the third and final harvest being used for veneer, has the potential to generate more revenue than production exclusively of energy biomass (B. Bender, personal communication, August 2012). More research needs to be done into the economics of multi-product timber production. The findings reported here are based on capital budgets, which are point estimates of representative investment returns over a designated time period. These budgets are designed to be representative of potential costs and returns in the northern tier of the Great Lakes region, as represented by the eastern Upper Peninsula of Michigan during 2010–2012. But apart from the baseline hay system, they are based on production systems that do not currently exist on a commercial basis. Important changes would have to occur in the costs and returns of these systems for them to become commercialized. Our findings indicate that although the production of energy biomass on marginal lands in the northern Great Lakes region is not currently profitable, the profitability potential is greater in this zone than it is on cropland in the southern tier of the Great Lakes. For that profitability potential to be realized, one of four circumstances would have to obtain: The relative price of energy biomass would have to rise sharply, yields of dedicated biomass crops would have to rise dramatically, costs of production would have to drop (notably via mechanical planting and harvest), or an attractive multiple product production strategy would have to develop (e.g., wood product sales for biomass, paper pulp, and veneer). Should sufficient gains be accomplished in one or a combination of these areas, dedicated energy biomass production would become profitable in an agriculturally marginal area like the northern Great Lakes, before it would successfully compete for land that is already producing food and feed crops. In order for such a transition to occur, further research is needed into cost-reducing production methods and multi-product woody crop production and marketing. Acknowledgments This work was funded by the US Department of Energy Great Lakes Bioenergy Research Center (DOE Office of Science BER DE-FC0207ER64494) with additional support from MSU AgBioResearch and the USDA National Institute of Food and Agriculture. For data and information, the authors thank Brad Bender, Paul Bloese, Kaye Table 10. Northern vs. southern tier of Michigan: Current yields, comparative breakeven yields, and percent change needed to break even (Mg ha –1 yr –1). North Crop Poplar Switchgrass Current yield 8 8 Breakeven yield 225 36 South Percent change 2713% 350% Agronomy Journal • Volume 106, Issue 2 • 2014 Current yield 16 11 Breakeven yield 640 75 Percent change 3900% 582% 405 Hillock-Vining, Christian Kapp, Raymond Miller, Doo-Hong Min, Sue Rak, David Rothstein, Jay Sergeant, Craig Shaeffer, Kurt Thelen, and Dan Undersander. For comments (but no responsibility for the final product), we thank Christian Kapp, Raymond Miller, Daniel F. Mooney, and David Rothstein. References Abrahamson, L.P., T.A. Volk, E. Priepke, J. Posselius, D.J. Aneshansley, and L.B. Smart. 2008. Development of a willow biomass crop harvesting system in New York. Powerpoint. Univ. of New York, College of Environmental Sci. and Forestry. http://www.esf.edu/outreach/pd/2010/srwc/documents/ willowharvestingtalkSRWCOWG2010.pdf (accessed 17 Dec. 2013. Abrahamson, L.P., T.A. Volk, L.B. Smart, and K.D. Cameron. 2010. Shrub willow biomass producer’s handbook. State Univ. of New York, College of Environmental Sci. and Forestry. www.esf.edu/willow/documents/ProducersHandbook.pdf (accessed 15 Aug. 2013). Adler, P.R., S.J. Del Grosso, and W.J. Parton. 2007. Life-cycle assessment of net greenhouse-gas flux for bioenergy cropping systems. Ecol. Appl. 17:675– 691. doi:10.1890/05-2018 Baier, S., M. Clemens, C. Griffiths, and J. Ihrig. 2009. Biofuels impact on crop and food prices: Using an interactive spreadsheet. International Finance Discussion Paper no. 967. Board of Governors of the Federal Reserve System, Washington, DC. Boehlje, M.D., and V.R. Eidman. 1984. Farm management. Wiley, New York. Brechbill, S.C., W.E. Tyner, and K.E. Ileleji. 2011. The economics of biomass collection and transportation and its supply to Indiana cellulosic and electrical utility facilities. BioEnergy Research 4(2):141–152. doi:10.1007/ s12155-010-9108-0 Campbell, J.E., D.B. Lobell, R.C. Genova, and C.B. Field. 2008. The global potential of bioenergy on abandoned agriculture lands. Environ. Sci. Technol. 42:5791–5794. doi:10.1021/es800052w Chen, X., and M. Khanna. 2012. The market-mediated effects of low carbon fuel policies. AgBioForum 15(1):1–17. CIMMYT. 1988. From agronomic data to farmer recommendations: An economics training manual. Centro Internacional de Mejoramiento de Maíz y Trigo (CIMMYT), Mexico. www.fao.org/sd/erp/toolkit/books/fromagronomic_manual.pdf (accessed 5 Oct. 2012.) Corace, R.G., III, D.J. Flaspohler, and L.M. Shartell. 2009. Geographical patterns in openland cover and hayfield mowing in the Upper Great Lakes region: Implications for grassland bird conservation. Landscape Ecol. 24:309–323. doi:10.1007/s10980-008-9306-8 Crutzen, P.J., A.R. Mosier, K.A. Smith, and W. Winiwarter. 2008. N2O release from agro-biofuel production negates global warming reduction by replacing fossil fuels. Atmos. Chem. Phys. 8:389–395. doi:10.5194/ acp-8-389-2008 Egbendewe-Mondzozo, A., S.M. Swinton, B.D. Bals, and B.E. Dale. 2013. Can dispersed biomass processing protect the environment and cover the bottom line for biofuel? Environ. Sci. Technol. 47:1695–1703. Egbendewe-Mondzozo, A., S.M. Swinton, C. Izzauralde, D.M. Manowitz, and X. Zhang. 2011. Biomass supply from alternative cellulosic crops and crop residues: A spatially explicit bioeconomic modeling approach. Biomass Bioenergy 35:4636–4647. doi:10.1016/j.biombioe.2011.09.010 Erickson, K.W., C.B. Moss, and A.K. Mishra. 2004. Rates of return in the farm and nonfarm sectors: How do they compare? J. Agric. Appl. Econ. 32:789–795. Fargione, J., J. Hill, D. Tilman, S. Polasky, and P. Hawthorne. 2008. Land clearing and the biofuel carbon debt. Science (Washington, DC) 319:1235– 1238. doi:10.1126/science.1152747 Hayes, D.J., B.A. Babcock, J.F. Fabiosa, S. Tokgoz, A. Elobeid, T. Yu et al. 2009. Biofuels: Potential production capacity, effects on grain and livestock sectors, and implications for food prices and consumers. J. Agric. Appl. Econ. 41(2):465–491. Hertel, T.W., A.A. Golub, A.D. Jones, M. O’Hare, R.J. Plevin, and D.M. Kammen. 2010. Effects of US maize ethanol on global land use and greenhouse gas emissions: Estimating market-mediated responses. Bioscience 60:223– 231. doi:10.1525/bio.2010.60.3.8 Hill, J., E. Nelson, D. Tilman, S. Polasky, and D. Tiffany. 2006. Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels. Proc. Natl. Acad. Sci. USA 103:11206–11210. doi:10.1073/ pnas.0604600103 Ince, P.J. 2009. Status and trends of U.S. pulpwood market. Climate change and its causes, effects, and prediction series. Nova Science Publ., New York. p. 101–117. 406 James, L.K., S.M. Swinton, and K.D. Thelen. 2010. Profitability analysis of cellulosic energy crops compared with corn. Agron. J. 102:675–687. doi:10.2134/agronj2009.0289 Keoleian, G.A., and T.A. Volk. 2005. Renewable energy from willow biomass crops: Life cycle energy, environmental and economic performance. Crit. Rev. Plant Sci. 24:385–406. doi:10.1080/07352680500316334 Khanna, M., X. Chen, H. Huang, and H. Onal. 2011. Supply of cellulosic biofuel feedstocks and regional production patterns. Am. J. Agric. Econ. 93(2):473–480. Khanna, M., and C.L. Crago. 2012. Measuring indirect land use change with biofuels: Implications for policy. Annual Review of Resource Economics 4:161–184. doi:10.1146/annurev-resource-110811-114523 Lemus, R., and R. Lal. 2005. Bioenergy crops and carbon sequestration. Crit. Rev. Plant Sci. 24(1):1–21. doi:10.1080/07352680590910393 Meadows, S., T. Gallagher, and D. Mitchell. 2010. Project summary: Application of a trailer-mounted slash bundler for southern logging. www.srs.fs.usda. gov/pubs/37375 (accessed 19 June, 2012). Min, D.H. 2011. Switchgrass for a bioenergy crop and livestock feed. Michigan State Univ. Ext. http://msue.anr.msu.edu/news/switchgrass_for_a_bioenergy_crop_and_livestock_feed (accessed 13 Sept. 2013). Mooney, D.F., R.K. Roberts, B.C. English, D.D. Tyler, and J.A. Larson. 2009. Yield and breakeven price of ‘Alamo’ switchgrass for biofuels in Tennessee. Agron. J. 101:1234–1242. doi:10.2134/agronj2009.0090 National Research Council. 2011. Renewable fuel standard: Potential economic and environmental effects of U.S. biofuel policy. The National Academies Press, Washington, DC. Netzer, D.A., D.N. Tolsted, M.E. Ostry, J.G. Isebrands, D.E. Riemenschneider, K.T. Ward. 2002. Growth, yield, and disease resistance of 7- to 12-year-old poplar clones in the north central United States. Gen. Tech. Rep. NC-229. USDA, Forest Serv., North Central Res. Stn., St. Paul, MN. Ross, S.A., R.W. Westerfield, J. Jaffe, and B.D. Jordan. 2008. Modern financial management. 8th ed. McGraw-Hill, Boston, MA. Searchinger, T., R. Heimlich, R.A. Houghton, F.X. Dong, A. Elobeid, J. Fabiosa et al. 2008. Use of US croplands for biofuels increases greenhouse gases through emissions from land-use change. Science (Washington, DC) 319:1238–1240. doi:10.1126/science.1151861 Searcy, E., P. Flynn, E. Ghafoori, and A. Kumar. 2007. The relative cost of biomass energy transport. Appl. Biochem. Biotechnol. 137:639–652. doi:10.1007/ s12010-007-9085-8 Stein, D. 2011a. Crop and livestock enterprise budgets. Michigan State Univ. Ext. www.msu.edu/~steind/budgets_crop_production.html (accessed 16 Aug. 2013). Stein, D. 2011b. Custom machine and work rate estimates. Michigan State Univ. Ext. www.msu.edu/user/steind/1_2012%20Cust_MachineWrk%20 10_31_11.pdf (accessed 15 June 2012). Swinton, S.M., B.A. Babcock, L.K. James, and V. Bandaru. 2011. Higher US crop prices trigger little area expansion so marginal land for biofuel crops is limited. Energy Policy 39:5254–5258. doi:10.1016/j.enpol.2011.05.039 Thompson, J., and S. Blank. 2000. Harvest mechanization helps agriculture remain competitive. Calif. Agric. 54(3):51–56. doi:10.3733/ ca.v054n03p51 Tyner, W.E. 2008. The US ethanol and biofuels boom: Its origins, current status, and future prospects. Bioscience 58(7):646–653. doi:10.1641/B580718 U.S. Congress. 2007. Public Law 110-140: Energy Independence and Security Act of 2007. U.S. Gov. Print. Office, Washington, DC. www.gpo.gov/ fdsys/pkg/BILLS-110hr6enr/pdf/BILLS-110hr6enr.pdf (accessed 15 July 2012) USDA, ERS. 2012. USDA Long-term Projections, February 2012. USDA, Economic Res. Serv. www.ers.usda.gov/media/273331/oce121d_1_.pdf (accessed 29 Aug. 2012) USDA, NASS. 2010–2012. Survey of agriculture, 2010, 2011, 2012. USDA, Natl. Agric. Statistics Serv., Washington, DC. Accessed via Quick Stats 2.0, 2013. USDOE, EERE. 2012. State Laws and Incentives. U.S. Dep. of Energy, Office of Energy Efficiency and Renewable Energy. www.afdc.energy.gov/laws/state (accessed 21 Dec. 2012). Wang, Z., and D.W. MacFarlane. 2012. Evaluating the biomass production of coppiced willow and poplar clones in Michigan, USA, over multiple rotations and different growing conditions. Biomass Bioenergy 46:380–388. doi:10.1016/j.biombioe.2012.08.003 Zamora, D., G. Wyatt, and J. Isebrands. 2011. Hybrid poplar best management practices. Univ. of Minnesota Ext. www.extension.umn.edu/distribution/ naturalresources/M1313.html (accessed 15 July 2013). Agronomy Journal • Volume 106, Issue 2 • 2014