Food, Fuel, and Plant Nutrient Use in the Future

Number 51 March 2013

Predictions of a world population of nine billion by 2050 necessitate careful stewardship of current food, fuel, and plant assets; a major part of that challenge involves managing what is beneath the surface. (Photo by Colette Kessler, USDA Natural Resources Conservation Service.)

bstract world’s growing population and look Because of various circumstances, A into the best practices for how to grain production will need to in- Current conditions and future move forward. crease by approximately 50% during trends show that adequate food pro- To meet global food demand, the the next four decades. Current U.S. duction will require increases in the use of genetics to improve crop pro- growth rates in cereal yields should use of fertilizer nutrients. With a ductivity, promote soil conservation meet 2050 demands, but greater ce- growing population, dwindling ar- and management, and use nutrients real yields per unit land area require able land, and an increased demand efficiently is necessary. The key to increases in fertilizer nutrient use, ad- for biofuels,1 the world cannot count these endeavors lies in supporting vances in genetics, and improved soil on an expansion of harvested area research and development in these and crop management technologies. to fill the demands. Scientists and areas. Other topics in this paper include food producers need to look at the This paper looks at the back- issues dealing with cellulosic biofuel way land is currently used to feed the ground leading to the current situa- production. According to projections, tion and addresses the resulting re- land availability is not a constraint to 1 Italicized terms (except genus/species names quirements as world food production biofuel production, and the United and published material titles) are defined in the Glossary. develops during the next 40 years. States has the capabilities to decrease

This material is based upon work supported by the U.S. Department of Agriculture’s (USDA) National Institute for Food and Agriculture (NIFA) Grants No. 2010- 38902-20899, No. 2009-38902-20041, No. 2008-38902-19327, No. 2007-31100-06019/ISU Project No. 413-40-02, and No. 2006-38902-03539. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the views of USDA–NIFA or ISU. CAST Issue Paper 51 Task Force Members

Authors Paul E. Fixen, International Plant Peter Scharf, Division of Plant Nutrition Institute, Brookings, South Sciences, University of Missouri, Dakota Columbia David Zilberman (Chair), Depart- Harry Vroomen ment of Agricultural and Resource John L. Havlin, Department of Soil , The Fertilizer Economics, University of California, Science, North Carolina State Uni- Institute, Washington, D.C. versity, Raleigh Berkeley CAST Liaison Bruce E. Dale, Department of Reviewers Chemical Engineering and Materials Todd A. Peterson, WinField, John- Science, Michigan State University, Chris van Kessel, Department of Plant ston, Iowa East Lansing Sciences, University of California, Davis

dependency on imported oil. Efficient the 1990–2010 period to 0.8 and 0.4% analysis of this problem. Moreover, land use is a key, and cover crops will for the 2010–2030 and 2030–2050 the stress put on food production by play a significant part in this process. periods, respectively (Table 1). The increasing oil prices significantly In the United States, however, the growth rates are different for devel- challenges the agricultural system. removal of the three primary plant oped and developing countries, as Bioenergy has been identified as a nutrients—nitrogen, phosphorus, and shown in Table 1. major component in developing alter- potassium—has been increasing, and The question of how to feed man- native energy sources in the United there is a need for increased fertiliz- kind by 2050 is a very important one States to achieve some level of energy er use and more recovery and recy- and brings many concerns to policy- independence. But the use of cere- cling from farm and nonfarm systems. makers and researchers alike. Three als in the production of bioenergy has Continued advances in nutrient use major components deserve special created concerns of competition due efficiency will moderate increased nu- attention based on the world’s cur- to the increased demands of cereals to trient demand. rent food system: estimation of food feed the world. Future food, fiber, and fuel de- needs, availability of land to grow Increasing population and demands will not be met by expanding food, and the nutrients that are re- mand for cereals for food and feed, cropland area. The authors use data to quired to increase world food pro- increasing use of cereals and other analyze factors influencing crop production. One model on which to base agricultural products for bioenergy, duction now and indications of what projections for future food demands is and limited land resources require in- is to come. With a growing population cereal production, because cereals are creasing yields on current agriculture and increased demand for food and used not only as a main human source land or using land with limited and/ fuel, research regarding nutrient use, of energy but also in feeding animals or decreased nutrients. Historically it recovery, and recycling is crucial. that will be consumed as meat and has been documented that to increase dairy products. crop yields the use of fertilizers and Introduction The United States is and will con- other nutrient sources must also in- tinue to be a major producer of food crease, even though fertilizer use ef- The world’s population is expect- for the world in the form of cereals ficiency has improved in recent years. ed to increase to 9.2 billion people and animal production. Increasing Therefore, the objective of this paper by 2050. But it is estimated that the food demands and shrinking agricul- is to obtain a better understanding of annual population growth rates are de- tural land in the United States and the factors influencing future fertilizer creasing considerably from 1.2% for other parts of the world necessitate an nutrient requirements and availability.

Table 1. Differences in population growth and annual population growth rate between developed and developing countries (UN 2007) Population (billion) Average Annual Growth Rate (%) 1950– 1970– 1990– 2010– 2030– 1950 1970 1990 2010 2030 2050 1970 1990 2010 2030 2050 World 2.54 3.70 5.29 6.91 8.32 9.19 1.99 1.74 1.21 0.82 0.41

Developed countries 0.81 1.01 1.15 1.23 1.26 1.25 0.96 0.59 0.32 0.06 –0.09 Developing countries 1.72 2.69 4.14 5.68 7.06 7.94 2.41 2.08 1.41 0.96 0.49

2 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY This, in turn, would support recom- in developing nations will still be composition reflecting changes in mendations to identify future policy higher (0.5%) than those in industri- prices. In addition to large project- implications and research requisites. alized nations (-0.1%) (Table 1) (UN ed increases in population by 2050, The authors have addressed the issues 2007). In developing countries, most world food consumption per capita in the following sections using a target of the decline in population growth also will continue to increase dur- of year 2050: rates is related to improved education, ing the next 40 years; it is, however, economic development, and increased • Population Dynamics. Estimation projected to stabilize at approximately agricultural productivity primarily in of population in developed and 3,100 kilocalories (kcal)/capita/day South and East Asia, an area that rep- developing countries that needs to by 2050, an increase of 300 kcal from resents nearly 50% of the total world be fed by 2050. current consumption (FAO 2006). The population (UN 2007). Population largest increases in food consumption • Food Needs to Sustain the World growth rates are projected to remain will occur in developing countries Population. Increased food demand high in sub-Saharan Africa where where 30 years ago per capita con- associated with consumption pat- poverty, suppressed economic oppor- sumption was approximately 2,100 terns of cereals needed as food and tunity, and low agricultural productiv- kcal and is projected to increase to feed, and the impact on U.S. grain ity will continue to persist during the approximately 3,000 kcal/capita/day production. next 40 years (Bruinsma 2009; FAO by 2050. In contrast, food demand in • Impact of Energy and Biomass 2010). developed countries will increase only Production. Demand of cereals for Despite declining world popula- slightly from current levels of 3,400 biofuels as well as the land avail- tion growth rates, total world popula- kcal and remain constant at approxi- ability in the United States, and tion will still increase nearly 35% to mately 3,540 kcal/capita/day. These the potential that the production of more than 9 billion people by 2050 increases in caloric intake primarily second-generation biofuels pre- (UN 2007). Currently, approximately reflect increases in the consumption of sents, with two scenarios and the 80 million people worldwide are add- meats. impact on environmental issues. ed annually, which will slowly decline Whereas per capita consumption during the next four decades to an • Land Use and Productivity. The (in kcal) is projected to stabilize dur- estimated 40 million people per year. worldwide demands for land to ing the next 40 years, diet composi- Total population will remain relatively grow food considering space tion will change substantially. World constant at approximately 1.25 billion and soil quality and the need for consumption of cereals for food (ki- in the developed countries because increased yields based on historical lograms [kg]/capita/year [yr]) is not of their low and declining population yield increases since 1950. projected to increase over current lev- growth rate (Table 1). Therefore, most els; per capita consumption of animal • Applied Nutrients and Nutrient of the population growth will occur in products (meat and dairy), however, Availability. Historical nutrient the developing countries. In the least will expand by more than 30% from use of nitrogen (N), phosphorus developed countries (e.g., countries current consumption levels through (P), and potassium (K) and crop in sub-Saharan Africa), populations 2050 (Figure 1). Cereal consumption removal in the United States, cur- will more than double to 1.8 billion (all uses) will increase by approxi- rent soil fertility status and trends, in 2050. Although population growth mately 10% during the same time pe- future nutrient requirements, in the remaining developing coun- riod, which represents an increase in and availability of fertilizer raw tries will be less rapid, population will cereals used for feed grain. In devel- materials. increase from 4.5 billion to 6.1 bil- oped countries, per capita consump- • Conclusions and Recommendations. lion in 2050. It is interesting to note tion of food cereals is expected to that population in developed nations remain at current levels, but consump- will decrease from 32% of total world tion of animal products will exhibit a Population Dynamics population in 1950 to approximately modest 8% increase during the next From the early 1960s, the world 13% by 2050 (calculated from data in 40 years (Figure 1). Total cereal con- annual population growth rate of Table 1). sumption in developed countries will slightly more than 2% has decreased increase by 12.5%, most of which rep- by 50% to current levels of slightly Food Needs to Sustain resents an increase in cereals used for more than 1% (UN 2007). Projections animal feed. to 2050 show global population the World Population In contrast, substantial changes in growth rates will decrease again by The factors determining future diet will continue to occur in develop- 50% to approximately 0.5% per year. global food demand include changing countries (Figure 1). Although per Whereas annual growth rates in all es in population (food demand), per capita cereal consumption for food nations will decrease by 2050, those capita energy consumption, and diet will remain stable, consumption of

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 3 Figure 2. Actual and projected annual consumption of cereals and animal products (meat + dairy) from 1970 to 2050 (Mt = metric tons). Total consumption represents cereals used for food and animal feed (FAO 2006). Figure 1. Actual and projected changes in per capita annual cereal and animal (meat + dairy) product consumption from 1970 to 2050. Total cereal consumption countries during the next 40 years will represents cereals used for food and animal feed increase 59, 89, and 173%, respective- (FAO 2006). ly. Annual consumption worldwide shows that food and total cereals will animal products will increase by 70% food consumption measured across increase by 47% and 65%, respec- from current levels to those expected an increasing population better illus- tively, from current levels to 2050, in 2050. Per capita cereal consump- trates future food needs. Thus, annual whereas global animal product con- tion for all uses will increase approxi- consumption in million metric tons sumption (meat + dairy) will double mately 17%, which again reflects ad- (MMt) is calculated by multiplying (97% increase). ditional use of cereals for animal feed. annual per capita consumption (Figure A number of assessments of global Most of the increase in consumption 1) by the projected population (Table food demand and supply have docu- of animal products is related to in- 1) for a given year. In the developed mented that current food supply is creased meat consumption in China, countries, annual consumption of ce- sufficient to meet demand, despite the India, and several other countries in reals remains level during the next fact that approximately 900 million or South and East Asia (Evans 2009). four decades, whereas total cereal and 12% of the world population is under- Although increases in per capi- animal product consumption will in- nourished (FAO 2010; IFPRI 2002; ta consumption and changes in diet crease by 17% and 13%, respectively Rosen et al. 2008). A large proportion composition are important indicators (Figure 2). In contrast, annual con- of these people reside within a subset of future food demand (especially sumption of food cereals, total cereals, of 32 nations (primarily sub-Saharan in developing countries) (Figure 1), and animal products in developing Africa) wherein approximately 40%

4 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY (29 to 72% range) of the population tural research. Regulatory burden has area cultivated for cereal production is undernourished (<2,200 kcal/capi- been another cause for the decrease in (Figure 3). In 2007, cereal production ta/day). In these countries, the cur- productivity and, in particular, slow- was approximately 2,350 MMt pro- rent population of 580 million likely er development of new innovation. duced on approximately 700 million will increase to 1.4 billion by 2050. Graff, Zilberman, and Bennett (2009) ha, which represents 3,350 kg/ha of Inequitable food distribution perpetu- argue that the European practical ban total cereal grain yield. ated by ongoing issues (distressed of genetically modified (GM) crops Estimating future food consump- economies, troubled education sys- has slowed the development of this tion requires modeling of dynamic tems, poor agricultural productivity, technology. Sexton and Zilberman processes that affect the demand and and possible internal conflicts) will (2011) argue that adoption of GM supply of food. The data suggest that continue to require substantial food varieties in corn increased yields sig- there are significant differences in de- aid to meet future population growth nificantly, especially in developing mand and supply between developed and accompanying food demand. countries, and lessened the increase in and developing countries and that diet Since the 1800s, as scientific mea- food prices that was associated with is changing across regions. Although sures affected agricultural production, an increase in demand in developing these differences in consumption pat- cheap food combined with improved countries. They suggest that added terns exist, research on food demands health contributed to population adoption of GM varieties in Europe suggests that common forces affect growth. Overall, the rate of growth and Africa would further increase consumption patterns globally. Rising in food exceeded the rate of popula- food availability and significantly income levels within a range tend tion growth, and the relative price of lower some of the recent increases in to change diets by increasing reli- food has declined over time (Federico food prices. Thus, continued capacity ance on meats rather than grains. This 2005). Commodity programs in the to provide food to meet growing de- growing demand for meat products United States (including land diver- mands requires an increase in invest- strongly increases the demand for sion programs) were actually designed ment of research as well as regulatory cereal. The recent economic growth to decrease excess supply (Gardner policies that will enable implementa- in China was a driver behind the 1987). Currently, world food production of new technologies. increased demand for grain in that tion is meeting food demand (USDA– Worldwide cereal yields have country as well as the increase in the ERS 2008). Historically, per capita increased linearly by approximate- price of grain commodities globally cereal production has kept pace with ly 2,000 kg/hectare (ha) since 1960, (Hochman et al. 2011). Continued population growth, although since the which represents an average annual economic growth in the developing mid-1980s, population growth has ex- increase of 43.6 kg/ha (Figure 3). world, as well as population growth, ceeded cereal production (FAO 2008). The increase in world cereal produc- is likely to increase the demand for These data show annual per capita ce- tion since 1960 is primarily related food globally, but the exact estima- real production worldwide decreased to increased yield per unit of land tion of future demand is subject to from approximately 370 kg to 350 area compared with increased land much uncertainty. kg/person, a decline of 5.7%. Dyson (1999) concludes that even though per capita cereal production increased slightly in developing countries, it was offset by declines in developed nations. Alston, Beddow, and Pardey (2009) argue that the rate of growth in demand for agricultural commodities (stemming from population growth and economic growth in Asia as well as expansion of biofuel) has been faster than the rate of growth of agricultural supply, which has trig- gered an increase in agricultural commodity prices. The rate of growth of supply suffered from the decline in the rate of growth of agricultural productivity that was affected to a large extent by underinvestment in agricul- Figure 3. Historical trends in cereal production (MMt), cereal yield (kg/ha), and cereal production area (million ha) (FAO 2008).

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 5 Future supply is subject to uncer- relatively equal (USDA–ERS 2008), cereal demand times projected popu- tainties regarding future crop pro- which reflects low carryover or year- lation) provides an estimated cereal ductivity, land use patterns, weath- end cereal reserves. Second, the annu- consumption demand of approxi- er conditions, and policies. Lobell, al per capita cereal production remains mately 3,400 MMt by 2050 (Figure Schlenker, and Costa-Roberts (2011) constant at approximately 350 kg/per- 4). This represents a 45% increase in suggest that climate change may re- son (FAO 2008). Although increases cereal consumption from 2007 (2,350 sult in decreased yield because of in income may lead to increased con- MMt) to 2050, whereas cereal pro- susceptibility to heat. Agricultural sumption of meats, and thus of grain duction (kg/ha) will increase by 55% systems may adapt, however, mini- per capita, declining efforts associated over the same period if yield increases mizing crop yield losses (Nelson et with modern life may decrease the de- continue at historical rates (Figure 3). al. 2009). Intensive development and mand for calories per capita (Deaton Using an estimated 5,200 kg/ha cereal adoption of transgenic technologies and Drèze 2009). This is somewhat yield in 2050 (Figure 3) averaged over can provide an avenue to increase higher than in Dyson (1999), who sug- approximately 700 million ha results yield and adapt to climate change, gested that this amount may decrease in approximately 3,650 MMt of cereal but it is constrained by regulatory to 330 kg/capita. Third, world agricul- production by 2050. Using the same requirements (Potrykus 2010; Sexton tural land area used for crops remains procedure, cereal production in 2025 and Zilberman 2011) that may be relatively constant. And finally, cereal is estimated to be approximately 2,825 modified in the future. yields continue to increase at current MMt, which is similar to the estimate Thus rigorous prediction of food rates (~40 kg/ha/yr) (Figure 3). As provided by Dyson (1999). These fig- yields and consumption trends re- suggested earlier, there is much uncer- ures suggest that world grain produc- quires quantification of the processes tainty about the evolution of yield per tion will meet world grain demand if that affect it. This is consistent with a acre. Although climate change is likely the historical trend in yield increase recent analysis of Gitiaux, Reilly, and to decrease yield both directly (Lobell, can be maintained. Although there are Paltsev (2011), who examined alterna- Schlenker, and Costa-Roberts 2011) many uncertainties in any analysis of tive yield growth models using data and because of difficulties of adapta- future food demand, these data sug- from 1961 to 2009 and showed that tion (Zilberman, Zhao, and Heiman gest that grain production will need to linear growth is quite limited in its 2012), higher prices of food are likely increase by approximately 50% during ability to explain much of the varia- to increase investment in agricultural the next four decades, assuming cur- tion in productivity growth. There is research and intensify agricultural pro- rent projected population growth and significant evidence that structural duction. Furthermore, higher prices of relatively constant agricultural land breaks, due to both technological and food may increase pressure to stream- area. policy reasons, significantly affect line regulation of GMOs. Thus, exist- Although the estimated growth in the evolution of yields per acre. Yield ing trends are used as a starting point. cereal demand cited earlier has been estimation can improve by using A calculation (350 kg/capita annual verified by other authors (Bruinsma Bayesian procedures or by identifying causal models that explain structural change and relate it to yield growth. These efforts are beyond the scope of this analysis, which will use a simple, linear projection to calculate food consumption in the future, recogniz- ing its limitations and realizing that the contribution of this work is the qualitative implications rather than precise numerical predictions. Since grains represent the ma- jority of food consumed by the poor (directly) and the rich (indirectly through the consumption of livestock), future food use will be estimated by projecting cereal use (wheat, rice, corn, etc.). Several as- Figure 4. Estimated world cereal production demand. Open circles represent pro- sumptions are required. First, without jected population multiplied by an estimated average annual per capita consumption of 350 kg. The open squares represent FAO (2008)-reported using sophisticated forecasting mod- world cereal production (1961–2007). The linear estimate is taken from els, production and consumption are Figure 3 (y = 30.108x + 58079).

6 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 2009; FAO 2006; IWMI 2007; Ringler production should increase by at least meet 2050 cereal demand, assuming 2006), clearly wide variability in any 50% as well, which implies that the current distribution of cereal grain use estimate is likely due to uncertainties United States will require production in the United States. in estimating future population growth, of 600 MMt by 2050.2 Obviously, Unfortunately, agricultural land diet composition, potential income, many factors influence commodity area has been decreasing at an annu- and nonfood crop use. While doubling import-export projections; it is likely, al rate of 0.15 million ha since 1960 of food production by 2050 has been however, that the dominant role of the (Figure 5). This trend is somewhat widely communicated (Beachy 2010; United States in meeting world cereal misleading, as the annual decrease Diouf 2009), it is important to distin- grain demand will continue (USDA– from 1990 is approximately 0.54 mil- guish between future food and cereal ERS 2008).3 lion ha. This decrease is likely an demand. In this analysis the focus is Currently in the United States, ap- overestimate because cereal cropland on future cereal demand, which in- proximately 60 million ha are harvest- loss at this rate would result in ap- cludes cereals used for human con- ed for cereal production at an aver- proximately 35 million ha of cereal sumption and animal feed. In contrast, age yield of 6,500 kg/ha (Figure 5), cropland in 2050. Lubowski and col- future food demand includes projected which represents nearly 400 MMt of leagues (2006) estimated approxi- growth in all plant and animal food cereal production. Linear extrapola- mately 0.40 million ha/yr of total sources. In this regard, the doubling tion of current growth in U.S. cereal cropland loss to predominately rural of food production by 2050 is largely production shows that cereal yield residential uses. Similarly, the USDA based on estimates of increased meat will be nearly 9,700 kg/ha by 2050. (2009) estimated nearly 0.5 million consumption driven by potential in- Maintaining current land area (60 mil- ha/yr of total cropland loss between creases in personal income in China, lion ha) in cereal production results 1982 and 2007. With approximately India, and other East Asian countries. in approximately 580 MMt of cereal 70% of total U.S. cropland in cereals, While studies question the estimated production in 2050. Thus, the current annual cereal cropland loss would be doubling of meat demand by 2050 in growth rate in cereal yields should 0.25–0.30 million ha. If the conserva- East Asia (Ma, Huang, and Rozelle tive estimate of 0.25 million ha/yr is 2004), estimates of total cereal de- used, then 10 million fewer ha of cere- 2 Because income levels in Asia are lower, income mand in 2050 have been consistent elasticities of food demand in Asia are higher, and al cropland will be available in 2050. and agree with Food and Agriculture that will contribute to higher relative increases With approximately 50 million ha of Organization of the United Nations in their demand for food compared to the United cereal production, cereal yields need (FAO) data (FAO 2006, 2008). States (Rosengrant et al. 2001). to increase to more than 11,600 kg/ 3 The increase in grain prices in the new millennium (from approximately $100 in the 1990s to approxi- ha, compared with 9,700 kg/ha under Impact on U.S. Grain mately $150 between 2005 and 2012, with rapid current cereal yield increases on 60 increases) suggests increasing food scarcity, and thus million ha. To achieve an additional Production the prediction of future demands that undervalue 2,000 kg/ha cereal yield by 2050, an- Historically, North America and the the recent trend of price increases may be somewhat conservative. nual growth rate in cereal yield must European Union have been the major suppliers of grain to many food- poor nations. The United States is currently providing approximately 60% of world food aid (Shapouri and Rosen 2004) and meeting nearly 30% of cereal import needs (FAO 2008). Current U.S. exports are approximately 100 MMt, which represents 25% of total U.S. cereal production (FAO 2008; USDA–ERS 2008). Of the total world cereal production of 2,350 MMt in 2007 (Figure 3), 100 MMt of U.S. cereal exports represent approximately 4.3% of world cereal production. Assuming that global cereal grain production increases by 50% to approximately 3,500 MMt during the next four decades (Figure 4), and expect- ing the United States to maintain its Figure 5. Historical and projected cereal yield and production area in the United States (FAO 2008). share in global production, U.S. cereal

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 7 be increased from 81 kg/ha/yr to ap- increase the cost of all field operations Plant matter—particularly cellulosic proximately 123 kg/ha/yr. To achieve requiring diesel fuel, and rising coal materials such as grasses, straws, and this increase, substantial advances in and natural gas prices increase the woody substances—are now much genetics and soil/crop management cost of electricity. Therefore, the cost less expensive than oil on a dollars technologies will be required. of processing and refrigerating food per megajoule (MJ) basis. Figure 6 increases. It is clear that energy prices shows the cost of cellulosic biomass mpact of nergy and will not return to the low levels of the in dollars per dry Mt plotted against I E twentieth century. While new discov- the cost of oil in dollars per barrel. Biomass Production eries, and especially the drastic in- The dashed horizontal line represents A New World of Food and crease in known natural gas reserves, the approximate price of large-scale Fuel are likely to increase the supply of sources of cellulosic biomass, which energy, demand for oil will continue should be available at approximately This is a time of profound transi- to increase because of population and $65 per Mt (Sokhansanj et al. 2009). tion in how the world will be fed and income growth. Much of the new nat- (Some sources of cellulosic biomass fueled. At the root of the movement is ural gas resources can substitute for are available at much lower prices.) the fact that the age of stable, cheap coal to help address climate change The solid diagonal line shows where oil is over. The balance between oil concerns. Furthermore, the new eco- the price of the energy content of the supply and oil demand is tight enough nomics of oil and gas are subject to biomass (essentially, the heat released that economic growth will put steady high degrees of uncertainty (Paltsev when it is burned) is equal to the price pressure on oil prices, magnifying any et al. 2011). Scarcity and anxiety in of the energy content of the oil. effects of speculation and threatened the fuel markets will spill over to food It is not possible to create or de- or actual loss of supply (IEA 2007). markets, amplifying uncertainties in stroy energy, only to change its form. Unfortunately, it may be that low oil these markets, threatening both total The goal of biofuel production is to prices will remain low only during a food production and prices. convert the energy content of plant recession. Thus we arrive at a poten- material into a form that can substi- tially grim future described by the fol- tute for petroleum. Thus, when oil lowing painful cycle: recession leads The Promise of Biofuels was cheap ($20 per barrel or less), it to lower oil prices; lower oil prices Is there a way to resolve this di- was only affordable to purchase the promote economic recovery, thereby lemma in which food prices rise along energy content of biomass. There increasing oil demand; oil prices rise with energy prices? Are there ways was no economic margin left over to in turn, aborting the nascent recovery that energy and food production can meet the cost of processing biomass and returning us once again to reces- complement and not compete with to energy products. Now that oil is in sion. In short, not a pretty picture for each other? Cellulosic biofuels may a new, higher price range, it becomes an oil-dependent world. help resolve these crucial questions. Furthermore, because all energy carriers can be substituted to some de- gree, high and unpredictable oil prices will increase the price and volatility of all other energy carriers, including natural gas and coal. Increased energy costs are an important driving force behind recent worldwide increases in the cost of food and agricultural commodities (Anderson et al. 2008). The world has had cheap food in no small part because it has had cheap energy, led by cheap oil. The production, processing, and distribution of all agricultural and food commodities are intimately linked with the price of energy. For example, natural gas is the major feedstock for fertilizer-N production. Change in natural gas prices, therefore, can impact the cost of fer- tilizing many crops. Rising oil prices Figure 6. Cost of energy in plant biomass versus cost of energy in oil.

8 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY economically possible to convert bio- Energy Markets plants. Biofuels do use some water mass to petroleum substitutes (Dale (three to four gallons per gallon of Grain-based biofuels are competi- 2008). Economic forces alone will fuel) in the biorefinery (the process- tive with oil without subsidies when tend to drive the use of plant matter ing facility that converts crops to fu- the price of a bushel of corn is less to replace petroleum and other energy els), which is similar to the amount of than approximately 5% the price of carriers. Plant matter can be burned water consumed per gallon of gaso- a barrel of oil. For example, without to provide electricity or natural gas line produced in oil refineries (Aden subsidies, corn-based ethanol com- substitutes or processed to produce 2007). petes well with $100-per-barrel oil petroleum substitutes such as ethanol Water requirements should not when corn is $5 per bushel or less (a gasoline replacement) or biodiesel. present a real problem for the ex- (Tyner 2008). Corn-based ethanol, First-generation biofuels such as etha- pansion of cellulosic biofuels apart however, probably will never sup- nol from corn or sugarcane or biodie- from isolated local exceptions in ply more than approximately 10–15% sel from soy oil will, over time, be some watersheds because of exces- of the U.S. gasoline market. There decreased by second-generation biofu- sive draw-down of important aquifers, simply is not enough corn to produce els made at a much lower cost and in or perhaps decreased flows of sea- more than that amount. In compari- much larger volumes from cellulosic sonal streams resulting from high- son, cellulosic biofuels can be com- materials (Dale 2008). productivity energy crops. On a global petitive when oil is approximately $50 There is a large body of literature basis, expansion of the water recy- per barrel. When cellulosic ethanol on the impact of biofuels—particu- cling and water purification ecosys- technology matures, it will be possible larly first-generation biofuels such tem services provided by plants would to deliver ethanol to the pump at less as ethanol from corn or sugarcane— seem to be a strong point in favor of than $2 per gallon (Laser et al. 2009). on energy markets, food production, biofuels, as long as irrigation from Worldwide, cellulosic biofuels can be water use, land availability, and the endangered aquifers or limited surface produced in very large quantities. In environment (Khanna, Scheffran, water supplies is avoided. Put bluntly, the United States alone, there is suf- and Zilberman 2010; Naylor et al. plant biomass production, particularly ficient cellulosic biomass to produce 2007). Learning by doing (Hettinga through nonirrigated perennial grasses enough ethanol to replace all import- et al. 2007) has decreased the cost and trees, is generally good for the ed oil rather easily (USDOE/USDA of corn ethanol substantially in the environment and for human societies. 2005). The stranglehold that oil has United States, and with the rising Plants enhance both the quantity and over economic development and na- price of fuel, the economics of both quality of water available in the bio- tional security for many nations will corn and sugarcane ethanol have been sphere. The same thing cannot be said only be broken when oil has serious improved.4 for the effects of oil production on The capacity of first- competition as a provider of liquid fu- water quality and quantity. generation biofuel is limited, however, els, and cellulosic biofuels are an es- and the potential of second-generation sential part of that competition. biofuel must be unleashed to cap- Net Energy ture the much larger national security, Many people have heard that corn economic, and environmental benefits Water Use ethanol has a negative “net energy” (Greene 2004). Water is used for biofuels both to (Pimentel and Patzek 2005). Net en- This section will explain briefly grow crops and to process them into ergy is defined by these authors as how cellulosic biofuels are different fuels. Approximately 15% of corn is the lower heating value of the fuel from first-generation biofuels and how irrigated, perhaps drawing down wa- divided by the sum of all the fossil en- these differences are important rela- ter tables in some areas. Cellulosic ergy (coal, petroleum, and natural gas) tive to the previously stated issues. biofuel crops, however, probably will required to produce the fuel. By this The section concludes by explaining not be irrigated, and most of the water way of thinking, ethanol is a poor fuel how cellulosic biofuel systems likely used in crop production will be rain because “it takes more energy to pro- will be designed to coproduce animal and snow. Plants use a process called duce the ethanol than you get when feeds and also how they might be de- evapotranspiration to take water from you burn the ethanol in your car.” But signed to recover and recycle mineral the soil, pass it through the plant, and net energy, if misused, can become an nutrients, an important focus of this release water vapors into the atmo- irrelevant and misleading datum. Net report. sphere. These water vapors can then energy analysis is founded on the idea 4 Because of the instability of food prices, biofuel form clouds and fall again as rain or that 1 MJ of petroleum is equivalent producers may be vulnerable during periods of high snow somewhere else. Water recy- to 1 MJ of natural gas is equivalent to food prices and low fuel prices and may have to cling through evapotranspiration is a adjust their behavior accordingly (Hochman, Sexton, 1 MJ of coal. This is obviously untrue; and Zilberman 2008). key part of the ecological function of otherwise people would not be paying

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 9 so much more for 1 MJ of petroleum 11:1. It is customary, however, to ac- biofuel production (Christian, Riche, than for 1 MJ of coal (Dale 2008). count for electricity’s higher energy and Yates 2008). The energy value of coal is simply not quality by multiplying the electricity Finally, taking into account the as useful or as versatile as the energy output by a factor of three, yielding likely land efficiency savings of co- content of oil, so coal is less valuable. an EROI of more than 30:1 for this producing animal feeds and cellu- Second, the net energy argument system. If the willow had been con- losic biofuels (see sections to follow), can be misleading if the net energy of verted to ethanol and electricity in a second-generation biofuel production ethanol is not properly compared with biorefinery (Laser et al. 2009), the may actually free up current agricul- the net energy of gasoline. Calculated overall EROI would have been at least tural land for other uses such as con- on the same basis, gasoline has a low- 15:1. By way of comparison, bitumen servation and protection of biodiver- er net energy than ethanol from corn oil from the tar sands has an estimated sity. Contrary to popular belief, land (Farrell et al. 2006). Ethanol from cel- EROI of approximately 5:1 and deep- availability is not currently a con- lulosics will have an even better net water oil’s EROI is approximately straint to biofuel production. energy profile than ethanol from corn. 12:1 (Hall, Balog, and Murphy 2009). More importantly, both corn and cel- Thus, even in their current relatively Food vs. Fuel: Adapting lulosic ethanol produce more than 20 underdeveloped state, cellulosic bio- Agriculture for Large-scale, times more liquid fuel per unit of pe- fuels offer better EROIs than their Second-generation Biofuel troleum “invested” to make them than highly developed fossil counterparts. is produced in converting petroleum Production to gasoline. In essence, bioethanol Land Availability It is important to distinguish be- stretches domestic petroleum supplies tween the impact of biofuel on fi- Large amounts of crop and for- into the future by leveraging oil to nal food prices in the United States, est residues (straws, forest thinnings, make much more liquid fuel: ethanol. which is quite low (3%–4.5%) be- slash, etc.) are available for cellu- A related concept called energy cause commodities account for a losic biofuel production without any return on investment (EROI) is im- small share of the final food price new land required. At least 1.5 billion portant and useful and should not be (Hochman et al. 2011), and the impact tonnes of such residues are available confused with net energy analysis, as of biofuel on commodity food prices, worldwide annually (Kim and Dale net energy has commonly been un- which, however, is more substantial. 2004), enough to make approximately derstood and misused in the biofuels Zilberman and colleagues (2012) sug- 560 billion liters of ethanol, almost context. The EROI is a measure of the gest that biofuel contributed between the entire volume of gasoline con- total useful energy output of a sys- 25% and 40% of the increases in corn sumed each year in the United States, tem divided by the total energy input and soybean prices for the period or approximately 70% of the energy required to produce that output. The 2001–2008, which was less than the content of all U.S. gasoline. “excess” energy (output minus input) contribution of economic growth and Additionally, at least 400 mil- is what is available to run the rest of more than the contribution of rising lion ha of former agricultural land society—education, cultural events, energy prices. Furthermore, the food are abandoned or unused (Campbell health care, and so on. It is obvious price peak in 2008 was affected by et al. 2009). It should be possible to that all human activities require en- various governments’ inventory man- produce an average of 4 Mt/ha/yr ergy. The higher the EROI, the more agement policies. As a significant por- of cellulosic biomass on these lands activities not related to energy pro- tion of commodity food production is with minimal inputs, enough to make duction can be undertaken by society. allocated to biofuel, it contributes to approximately 750 billion liters of The EROI for corn ethanol is approxi- rising food prices, yet this is not a ma- ethanol per year, roughly the ener- mately 2:1, while sugarcane ethanol’s jor concern where supplies are ample gy equivalent of the annual amount EROI is approximately 10:1. Based and overall food prices are reasonable. of gasoline consumed in the United on near future technology, cellulosic Concerns about biofuel’s contribu- States. Furthermore, potential en- ethanol has an EROI of approximately tions to food price inflation, however, ergy crops such as switchgrass and 18:1 with the potential of 35:1 or so as are during periods of high food prices Miscanthus (crops grown specifically 5 technology improves (Hall, Dale, and as in 2008. to capture solar energy rather than for Pimentel 2011). Second-generation biofuels are their food or feed value) have received A pioneering study (Heller, much less likely to impact food prices. little agronomic attention to increase Keolian, and Volk 2003) showed that their yields. Significant yield gains electricity production from biomass should be expected, increasing the 5 Actually, one of the benefits of biofuel during (plantation willow) produced 11 units productivity of both abandoned and periods of ample supply is that it raises agricultural of electrical energy for every unit of commodity prices and rural incomes and decreases active agricultural lands for cellulosic fossil energy consumed, an EROI of the need for commodity support programs.

10 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY In fact, this issue of food vs. fuel re- Additional feed protein also may in land use efficiency might be to em- quires careful thinking because it of- be readily coproduced with biofuels ploy crop residues (e.g., corn stover fers potentially important synergies via the spent yeast cell mass follow- and wheat straw) to produce improved between food and biofuel producing the fermentation to produce fuels ruminant feeds. Since roughly half of tion. Throughout the United States (Matthews et al. 2011). Therefore, the corn plant is grain, this approach and much of the rest of the world, it is likely that increased cellulosic approximately doubles the per-acre people do not actually “grow food.” biofuel production will be accompa- productivity of land for animal feed. Instead, they grow animal feed and nied by large increases in quantities of Increased double cropping has then consume the meat, milk, eggs, ruminant animal feeds, leading to an already played a major role in meet- and so forth produced by the animals. increase in both protein and digestible ing the elevated demand for feed as- Approximately 85–90% of the best energy. The effect of these two chang- sociated with population growth. In U.S. agricultural land—more than 300 es will be to use land more efficiently particular, much of the increases of million ha of cropland and pasture—is to meet both food (actually animal soybean acreage, which has doubled used to feed animals, not directly to feed) and biofuel needs. Some pos- since 1990, is due to double cropping feed humans. sibilities for integrating animal feed of soybean and grains, primarily in Animals, like humans, require two and biofuel production and their land Argentina but also in Brazil. Much of primary food components: protein and use consequences, including effects this expansion was fueled by adop- digestible energy (calories). Also, ap- on greenhouse gas (GHG) production, tion of herbicide-tolerant varieties and proximately 70% of the protein and are outlined in the following pages. was associated with adoption of low- calories fed to livestock is fed to dairy Winter double crops are annual tillage practices (Trigo and Cap 2003). and beef cattle, and approximately 90 grasses or legumes (e.g., winter rye This increase in supply produced suf- million cattle are raised in the United or clover) planted in the fall after the ficient output to meet the growing de- States, a nation of 300 million human corn or soybean crop is harvested, mand for soybeans in China (Sexton beings (Dale et al. 2009). Cattle and typically to provide a green animal et al. 2009). Expanding double crop- other ruminant animals can use grass- feed in the spring and/or for soil im- ping in the United States may lead to es or straws as a source of calories. provement. Double crops increase higher productivity without increasing Most such cellulosic materials are not soil fertility (by sequestering carbon), the ecological footprint of agriculture. very digestible, however, so in the largely eliminate wind and water ero- Double crops, leaf protein, and United States the tendency has been sion, and significantly decrease leach- enhanced cellulosic animal feeds were to feed grains, primarily corn and corn ing of nitrates and other pollutants. In recently analyzed for their potential to silage, to cattle. Almost all soybean spite of these positive environmental provide current levels of food (mostly production, including the associated benefits, double crops are grown on animal feed) as well as significantly vegetable oil, goes to providing pro- much less than 10% of corn acres. larger amounts of feedstocks for bio- tein in animal diets. This is largely due to the fact that they fuel production (Dale et al. 2010). The The so-called pretreatment pro- represent a cost to the farmer without analysis was based on 114 million ha cesses that “unlock” sugars (cellulose much associated revenue. If double (approximately 70% of U.S. cropland), and hemicellulose) in cellulosic ma- crops became a source of farmer rev- assuming that one-third of U.S. corn terials for conversion to ethanol and enue—for example, for LPC and/or and soybean land was used to produce other biofuels may also unlock these cellulosic biofuel production—more a winter double crop. According to this sugars for digestion by ruminant ani- double crops would undoubtedly be analysis, total biomass (grain plus cel- mals (Dale 2008). The same facility planted. The LPC could replace some lulosic biomass) increased 2.5 times that produces pretreated feedstock for soy meal protein, thereby freeing up over current levels. This is enough biofuel production may be able to pro- land for additional biofuel production biomass to produce approximately 400 duce enhanced ruminant animal feeds. without impacting food supplies. gigaliters (GL) of ethanol per year, Furthermore, grasses and legumes As mentioned earlier, the cellulos- roughly the energy equivalent of all such as alfalfa can be harvested in the ic fiber remaining after LPC produc- imported petroleum used for gasoline early spring when their protein con- tion can replace some existing animal production, while still providing all tent is high (15–20%), and this pro- feeds or serve as a raw material for the food and feed currently produced tein, so-called leaf protein concentrate biofuel production. Furthermore, the on this acreage. This approach to in- or LPC, can be recovered as animal presence of the double crop protects tegrated food and biofuel production feed using well-known technology soil, thereby allowing increased re- also decreases total U.S. GHG emis- (Pirie 1978). The residual cellulosic moval of corn stover. These effects sions by approximately 700 teragrams fiber left behind after the protein is would increase the total biomass pro- (Tg) carbon dioxide (CO2) equivalents removed is suitable for animal feed or ductivity per acre without decreas- per year, roughly 10% of the total U.S. biofuel production. ing food supplies. Another increase GHG emissions.

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 11 Figure 7. Current allocation of 114 million ha of U.S. agricultural land to produce feed, food, and fuel (left-hand side) versus a land- efficient allocation of those same acres to produce much more biofuel (right-hand side). (Reprinted with permission from Dale et al. [2010]. Copyright 2010 American Chemical Society.)

Figure 7 is a “spaghetti” diagram corn grain yields increase to approxi- ing corn and soybean acres per year, showing current land use patterns on mately 13.4 Mt grain/ha/yr (250 bush- which is more than double the esti- the left and, on the right, how this els per acre per year), the amount of mate used earlier of approximately 73 same amount of land might be recon- cellulosic biomass that could be pro- million dry Mt per acre. Searchinger figured to provide all of the food/feed duced on existing acres would expand and colleagues (2008) suggest that it currently supplies while also provid- further. the expansion of agricultural land due ing enough additional biomass to pro- A very recent analysis of double to rising prices associated with bio- duce a very large amount of biofuel. crop coproduction in corn and soy- fuel and the indirect land use change Sensitivity analysis indicates that bean agriculture (Richard et al. 2010) (ILUC) affiliated with this extensi- the most important variables are the shows that the arbitrary estimate of fication is a major drawback of the fraction of land in cover crops, yields double cropping on one-third of corn technology. Recent studies have found of perennial grasses, and animal feed and soybean acres used earlier is prob- these estimates to be overstated, but requirements. The results outlined in ably too conservative. Richard and concerns still remain (Khanna and Figure 7, however, are robust across his colleagues at Pennsylvania State Crago 2012). Since the current prod- a wide range of assumptions, as University and the U.S. Department ucts of the land continue to be gener- shown in Figure 8. The bottom line of Agriculture (USDA) conclude that ated, this approach of producing addi- is that very large amounts of cellu- approximately 180 million dry Mt of tional biomass on existing land avoids losic biomass for second-generation double crops could be grown on exist- the so-called ILUC effect. biofuel production can be provided on existing land without compromising domestic food production or agricultural commodity exports. Production of more double crops and perennial grasses on a fraction of the existing land would decrease U.S. GHG production by approximately 10%, increase biodiversity, and improve soil fertility, and, if done correctly, could decrease nitrate emissions to groundwater and hence to the Gulf of Mexico, limiting the anoxic (“dead”) zone there. Looking to the future, if the yield of grasses increases to approximately Figure 8. Sensitivity of ethanol production and GHG reduction to various assumptions about animal feed replacements (EtOH = ethanol; AFEX = ammonia 27.5 tonnes/ha/year during the next fiber explosion). (Reprinted with permission from Dale et al. [2010]. Copy- decade (Sokhansanj et al. 2009) while right 2010 American Chemical Society.)

12 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY This analysis only addresses the As noted, the technologies that Kucharik 2008). technical potential of these changes. provide most of the benefit to food and Second-generation biofuels, par- Multiple drivers would be required biofuel production are extensive double ticularly those based on perennial to actually institute them. Changes in cropping and large-scale production of grasses, are likely to provide signifi- land use patterns are more likely to diverse cellulosic crops appropriate to cant environmental improvements, occur if they are economically attrac- different regions of the country. These especially if the overall system is tive to farmers, livestock producers, are not exotic, expensive, or high-risk designed to supply environmental, as and the biofuel industry. Chen, Huang, technologies. Considering their large well as economic, benefits. Perennial and Khanna (2012) assessed adop- benefits to energy and climate security, grasses tend to build soil organic mat- tion and impact of various second- extensive double cropping and produc- ter over time, thereby increasing soil generation biofuels in the United tion of diverse cellulosic crops deserve fertility; they essentially eliminate soil States using different assumptions and more study for widespread application erosion, provide better wildlife habi- showed that adoption patterns would in integrated biofuel and animal feeding tat, improve water quality, and effec- depend on policies as well as param- systems than they have received to date. tively capture N and other nutrients. eters of the technology. For example, The United States is the world’s Furthermore, ethanol from cellulosics the Renewable Fuels Standard 2 and largest petroleum user and also a decreases life cycle GHG emissions the Low Carbon Fuel Standard are significant exporter of agricultural by approximately 90% compared with more likely to enhance the adoption commodities. The analysis of this gasoline (Farrell et al. 2006). A ma- of Miscanthus, whereas a carbon tax paper shows that the United States jor advantage of the double-cropping may lead to decreased GHG emissions can produce very large amounts of system described earlier is that it pro- primarily by enhancing conservation, biofuels, maintain domestic food sup- vides some features of perenniality yet low adoption, of second-generation plies, continue the contribution to (e.g., year-round ground cover) while biofuel because of its cost. The high international food stock, increase soil preserving farmers’ flexibility regard- price of conversion of feedstock to fertility, and significantly decrease ing planting decisions. biofuel was a major barrier for the GHG emissions. If so, then integrat- expansion of second-generation bio- ing biofuel production with animal Plant Nutrient Issues in fuel in the Chen, Huang, and Khanna feed production may also be a path- Biofuel Production (2012) study, and policy instruments way available to many other coun- Given the need for fuels and the were required for the introduction of tries. Resolving the apparent “food availability of plant matter at energy these technologies to make economic versus fuel” conflict seems to be more equivalent prices well below those of sense. Thus, much further exploration a matter of making the right choices petroleum, it will almost certainly be of the technologies mentioned earlier, rather than hard resource and techni- possible to overcome the technical as well as the economics of combined cal constraints. With investment in obstacles that currently limit expan- food/feed/fuel systems, is required. technology and policy commitments, sion of biofuels. Efficient use of land Continued policy emphasis and adaptation of the agricultural system to grow plants for biofuel production, incentives tied to improving the envi- to produce food, animal feed, and sus- however, will require increased nutri- ronmental performance of biofuels as tainable biofuels is possible. ent inputs overall and possibly limit well as animal feed production also the potential of biofuel production would be needed to drive the desired Environmental Issues (Murrell et al. 2011). changes, at least during the period of First-generation biofuels do have Regarding the environmental is- adjustment. Combining double crops some environmental advantages and sues discussed previously, it is neces- with increased corn stover harvest drawbacks. For example, corn etha- sary to think of the whole system and is a key driver because of the large nol provides GHG reductions rela- work to improve its overall perfor- amounts of cellulosic biomass made tive to gasoline, and gasoline-ethanol mance rather than focusing exclu- available with concurrent improve- blends tend to burn more cleanly than sively on small pieces of the system. ments in several environmental pa- straight gasoline.6 Increased corn The fundamental fact regarding plant rameters. The use of double cropping production for ethanol (particularly nutrient issues is that the important in biofuel systems can be induced by on poorer soils), however, will tend fuel atoms are carbon and hydro- policies that support these practices as to increase soil erosion and could in- gen. Ideally, there should be no other well as policies that minimize N emis- crease N losses to streams, rivers, and chemical elements in our fuels. Thus, sions for the maximum environmental eventually the Gulf of Mexico, enlarg- there will be a strong incentive to benefits. Design of these policies will ing the anoxic zone there (Donner and recover and recycle plant nutrients require further analysis of their envi- (N, P, K, etc.) during the biorefining ronmental and economic impact, as 6 Recent studies (Khanna and Crago 2012) suggest process in which raw plant matter is well as their direct cost. that even when taking into account indirect land use, biofuel emits less GHG than gasoline in many cases. converted to biofuel. For example,

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 13 the fermentation residue from biofuel tively constant since 1990, whereas moisture and temperature stresses oc- production will likely be burned to re- cropland, arable and permanent crop- cur on more than 52% of the land area cover its energy value and increase the land, has increased slightly (~0.22%/ (Wiebe 2003). A number of soil physi- overall energy efficiency of biofuel yr), likely into these less productive cal and chemical properties, either production. Depending on the system areas (FAO 2008). natural or anthropogenic, also limit design, the plant mineral nutrients Per capita cropland use decreased cropland productivity. could be concentrated in the resulting nearly 50% from 0.44 ha/person in Expanding cropland into remain- ash stream and then might be recycled 1960 to 0.23 in 2007. By 2050, world ing agricultural lands that are sub- to the land. cropland use will further decrease stantially less productive than cur- Other contributions to this is- by approximately 30% to 0.16 ha/ rent croplands will limit global crop sue deal with increased production of person, assuming constant cropland production growth. For example, plant nutrients and greater efficiency area. If the annual increase in crop- Wiebe (2003) suggests that 80% of in plant nutrient uptake. Biofuel pro- land area from 1996 to 2007 continues future increases in crop production in duction, however, influences a rapidly (~3.38 million ha/yr or 0.22%), then developing countries will come from growing demand for agricultural and per capita cropland use will be nearly intensification instead of expansion forestry products and potentially a 0.18 ha/person. Per capita cropland of cropland. The poorest cropland, way to reverse the decades-long eco- assessments are misleading, how- unfortunately, occurs in regions with nomic decline in agricultural com- ever, because of the changing distri- the greatest need to expand produc- munities around the world. Therefore, bution of human populations in rural tion (Wiebe 2003). In developing both research and systems analysis are and urban areas. Similar to population countries (countries in sub-Saharan needed to design and implement ef- growth rate, the rate of urbanization Africa, Latin America, etc.), projected fective means by which plant nutrients has been decreasing, but the absolute croplands will increase only 0.3%/yr are recovered and recycled in biofuel urban population is increasing. For the or 120 million ha/yr during the next production processes. This is another first time in history, more than 50% of several decades, which is less than prime area for increased attention and the world population in 2008 lived in in previous decades, whereas little research funding. urban areas. By 2050, more than 60%, or no increases are expected in de- or nearly 6 billion people, will live veloped countries (Bruinsma 2009). Land Use and Productivity in urban areas. Therefore, the impact Accounting for the decrease in crop- of increasing population on conver- land in developed nations, the net gain Rapid human population growth sion of cropland to urban uses is less- in cropland is estimated to be only since 1950 (Table 1) has caused ened by the disproportionate expan- ~70 billion ha. rapidly growing demands for food, sion of urban areas. Urban population Because cropland expansion will water, timber, fiber, and fuel. These growth, however, commonly occurs have minimal impact on total global demands have impacted ecosystems on highly productive lands, and urban crop production, increased production more extensively than in any compa- expansion in developing countries de- on existing cropland areas must occur, rable time period in human history. creases cropland by 0.5 million ha/yr which is limited by continued land Approximately 12% (1.55 billion ha) (Rosengrant et al. 2001). degradation. Land degradation rep- of total world land area and 32% of Future food, fiber, and fuel de- resents deterioration of one or more agricultural land (4.93 billion ha) is mand obviously will not be met by ex- land properties decreasing land qual- current cropland (FAO 2008). The panding cropland area. Unfortunately, ity or the ability of the land to sustain remaining 3.38 billion ha of agricul- only an estimated 13% of the global a specific function such as crop pro- tural land, primarily (90%) in Latin land surface can be considered prime duction (Karlen et al. 1997; Lindert America and sub-Saharan Africa, are cropland (Class I) or lands with few 2000). Because soils are the funda- in forests, permanent pasture, and problems limiting sustainable grain mental component of land, soil degra- other noncrop uses. Wiebe (2003) production (Classes II and III) (Table dation leads to land degradation. Soil estimates these remaining agricul- 2). Approximately 76% of the global degradation, both natural and anthro- tural lands represent only 20% of the population resides on the least pro- pogenic, represents a change in chem- yield potential of the most productive ductive lands (Classes IV–IX). This ical, physical, or biological properties cropland, thus cropland expansion in may seem alarming; however, with that individually or collectively low- these areas occurs at a large economic equitable food input-export systems ers agricultural productivity or some cost (poor soil fertility, soil depth, low and policies, crop production from other ecosystem service. Soil erosion rainfall, etc.) and great risk to biodi- the most productive lands can meet by water or wind, nutrient depletion, versity, soil erosion, and other factors and has met global food and other re- compaction, desertification, saliniza- impacting ecosystem function. The source demands. Although many fac- tion, and acidification are examples of total agricultural land has been rela- tors limit grain crop productivity, soil processes that degrade soils.

14 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY Table 2. Distribution of global lands and populationa in land quality classesb (Beinroth, Eswaran, and Reich 2001)

Land Area Population Land Quality Classb Million ha % Millions % I 409 3.13 337 5.87 II 653 5.00 789 13.75 III 589 4.51 266 4.63 Decreasing IV 511 3.91 654 11.40 Land V 2,135 16.35 1,651 28.77 Productivity VI 1,722 13.19 675 11.76 VII 1,165 8.92 639 11.13 VIII 3,696 28.30 103 1.79 IX 2,178 16.68 625 10.89 Total 13,058 100.00 5,739 100.00 aPopulation only between 72°N and 57°S latitudes. bIncludes risk for sustainable grain crop production: Class I, <20%; Class II, 20–30%; Class III, 30–40%; Classes IV–VI, 40–60%; Class VII, 60–80%; and Classes VIII–IX, >80%. Class I: Few soil limitations restricting use for crop production. Class II: Moderate soil limitations restricting crop choice or requiring moderate conservation practices. Class III: Severe soil limitations restricting crop choice and/or requiring special conservation practices. Class IV: Very severe soil limitations restricting crop choice and/or requiring very careful management. Class V: Soils subject to little or no erosion but have other limitations, impractical to remove, that restrict use to pasture, rangeland, forestland, or wildlife habitat. Class VI: Severe soil limitations; generally unsuitable for cultivation; use restricted to pasture, rangeland, forestland, or wildlife habitat. Class VII: Very severe soil limitations; unsuitable for cultivation; use restricted to grazing, forestland, or wildlife habitat. Class VIII: Soils and miscellaneous areas have limitations precluding commercial plant production; use restricted to recreational, wildlife habitat, watershed, or esthetic purposes. Class IX: Mainly the deserts where biomass production is very low.

Continued degradation of world Water erosion represents the most Wiebe (2003) estimated a 0.1 to 0.3% soil productivity threatens the ability important mechanism for loss in crop- global average annual erosion-induced to meet future global food and fiber land productivity, whereas overgraz- decline in cropland productivity, demands. The Global Assessment of ing, deforestation, and agricultural depending on level of adoption of Soil Degradation (GLASOD) esti- activities are the greatest causes of conservation and other management mated that nearly 2 billion ha (15% soil degradation worldwide (Oldeman, practices. of total global land area; 23% of 8.7 Hakkeling, and Sombroek 1991). Although it is difficult to quantify billion ha used for crops, pasture, and Using GLASOD data, Crosson and few studies are available, one can forests) had been degraded by hu- (1997, 1998) concluded that from use the estimates of Wiebe (2003) to man activity (Oldeman, Hakkeling, 1945 to 1990 average annual produc- evaluate the potential impact of soil and Sombroek 1991). Although soil tivity declined 0.4%. Of the 8,700 degradation on cropland productivity. degradation varies widely between million ha used by humans, however, Assuming an average annual produc- regions (developed vs. developing), cumulative productivity decreased 5% tivity loss of 0.3%, combined with the approximately 38% of the world’s or 0.1%/yr during the 45-year period. 1961–2007 cereal yields (Figure 3), cropland has degraded (USDA–ERS Oldeman (1998) also reported cumu- the potential increase in cereal yield 2008). By region, 65% of cropland in lative cropland and pasture productiv- can be estimated assuming the 0.3%/ Africa, 51% in Latin America, 38% ity losses from 5 to 9% or 0.1 to 0.2%/ yr yield loss had not occurred. Using in Asia, and 25% in North America, yr, although in some regions crop- linear estimates of total cereal produc- Europe, and Oceania have been de- land losses were considerably higher tion (MMt) and yield (kg/ha), a 0.3%/ graded. Approximately 2 million ha (25% or 0.5%/yr in Africa and 37% yr adjustment was made for 1961– of rain-fed and irrigated agricultural or 0.7%/yr in Central America). In 2050 (Figure 9). lands are lost to production every North America cropland productivity These data demonstrate that in year because of severe land degrada- losses are much lower, with estimates 2007 an additional 400 kg/ha cereal tion, which increases the productivity ranging from 0.0 to 0.1%/yr (0.04%/ yield could have been produced. By demand on the remaining croplands yr average) depending on crop, slope, 2050, approximately 1,000 kg/ha or while increasing pressure on convert- and management (Alt, Osborn, and 19% more cereal yield was possible. ing less-productive land into cropland Colacicco 1989; Crosson 1986; Pierce Adjusting for world cropland area in (Oldeman 1994). et al. 1983). Averaging over regions, cereals, an additional 278 MMt could

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 15 significant proportion of field crop residues for fuel production also may lower soil organic matter and accelerate soil degradation, which lessens soil productivity.

Applied Nutrients and Nutrient Availability Plant nutrients are essential inputs for all forms of crop production, and meeting future societal needs will require careful attention to plant nutrition. A recent review of long-term studies showed that estimates of the contribution of commercial fertilizers to food production generally ranged Figure 9. Historical projected trends in cereal production (MMt) and cereal yield from 40 to 60% in the United States (kg/ha) under current levels of soil degradation and those corrected for a and England and tended to be much 0.3%/yr loss in cereal productivity (FAO 2008; Wiebe 2003). higher in the tropics (Stewart et al. 2005). Estimates made at the end of have been produced in 2007, whereas nearly 470 kg/person/yr was possible the millennium were that fertilizer-N more than 700 MMt of cereal grains without the deleterious effects of soil alone was responsible for supplying could have been available in 2050. degradation. the basic food needs of at least 40% As documented previously, the FAO Because future increases in food of the population and that popula- (2006) projects a global cereal de- production will come from increased tion growth and increasing prosperity mand of 350 kg/person/yr (Figure 1). yield per unit of land area instead of would eventually increase that esti- Under projected growth in population increased arable land area, it is im- mate to at least 60% (Smil 2001). Any and cereal production (assuming rela- perative that efforts to sustain and en- meaningful evaluation of the future tively constant cereal cropland area), hance soil productivity be increased, of food or biofuels must consider nearly 400 kg/person/yr will be need- especially in developing countries. the plant nutrients involved in their ed and likely can be produced (Figure Continued trends in soil degradation production. 10). Adjusting these projections for will jeopardize the capacity to meet the 0.3%/yr loss in productivity, future food demand. Removal of a Historical Nutrient Use and Crop Removal in the United States Realistic contemplation of future nutrient use is facilitated by an understanding of past and current nutrient use, especially in light of nutrients removed in crop harvest. Fertilizer consumption in the United States increased rapidly from the early 1960s to 1980, then experienced a few turbu- lent years resulting in a decline in use in the mid-1980s (Figure 11) (Slater and Kirby 2011). Since approximately 1986, N fertilizer consumption has been increasing linearly at a rate of 97,000 Mt per year, reaching nearly 12 MMt in 2007. Consumption of P and K during this same period has Figure 10. Estimated per capita world cereal consumption. Open squares represent historical and projected cereal production divided by population (UN been nearly constant except for a 2007) to obtain kg/person/yr. The open circles represent kg/person/yr sharp decline in 2009 resulting mostly adjusted for the 0.3%/yr loss in productivity. from a spike in fertilizer prices.

16 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY N consumption has exceeded crop- N removal (excluding alfalfa, soybeans, and peanuts) by approximately 3.5 MMt or 45% of current average removal. In contrast, P removal has exceeded P fertilizer use since approximately 1990, and K removal has always exceeded K fertilizer use due primarily to many soils in the western United States that are indigenously high in plant-available K and generally unresponsive to K fertilization. Livestock manure is another potentially significant nutrient source for crop production, although it does not represent new nutrients in agricultural systems but rather the recycling of nutrients within systems (Figure 14). Manure nutrients are difficult to account for in nutrient budgets on a national scale because of the par- Figure 11. Fertilizer consumption in the United States and trends from 1986 to 2010 tial decoupling of livestock and crop (P2O5 = phosphate; K2O = potash). production. The resulting geographic separation of feed production location From 1965 to 2010, crop produc- from the soil or other sources and, for and consumption location frequently tion increased markedly, driven by the soils that are at optimum nutrient causes accumulation of nutrients in re- increasing crop yields (USDA 2012). levels, replace the nutrients removed gions of high livestock density, result- Increased yields resulted in increased by crop harvest. It is instructive to ing in low nutrient use efficiency. nutrient uptake by crops and increased directly compare fertilizer consump- The USDA–Natural Resources removal of nutrients from farm fields. tion to crop nutrient removal (Figure Conservation Service (Kellogg et al. Nutrient removal can be estimated us- 13). Since the late 1970s, fertilizer- 2000) has estimated that 55 to 65% ing standard crop removal coefficients (amount of nutrient contained in each tonne of crop removed from the field) (IPNI 2012; Johnston and Usherwood 2002). Figure 12 shows the history of nutrient removal by crops in the United States.7 Nitrogen removal by alfalfa, soybeans, and peanuts is not included because they are legume crops that fix their own N from the atmosphere and typically receive little N fertilizer. Removal of all three nutrients across the United States has been increasing linearly during this entire period with annual rates of increase of 70, 51, and 67 thousand Mt for N, P2O5, and K2O, respectively. The primary roles of commercial fertilizer use are to supply plants with the nutrients they cannot obtain

7 Removal coefficients used for most major crops are from the IPNI reference. Crops not found in this Figure 12. Nutrient removal by crops in the United States (N removal by alfalfa, soy- source are from the older Johnston and Usherwood reference. beans, and peanuts excluded).

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 17 negative even with all recoverable manure-K included.

Current Soil Fertility Status and Trends The implications of the partial nutrient budgets discussed previously regarding fertilizer needs in the near future depend in part on current soil P and K fertility. If more nutrients are removed by crops than are applied, declining soil fertility occurs and at some point in the future when a critical level is reached, nutrient application would need to be increased for yield levels to continue to improve and for the system to remain sustainable. The higher the current level of fertility, the further into the future that point will be encountered. Figure 13. Fertilizer nutrient consumption in excess of crop nutrient removal in the Soil testing is the primary means United States (N removal by alfalfa, soybeans, and peanuts excluded). by which farmers evaluate soil fertility levels in the United States. Fixen and colleagues (2010) and the International Plant Nutrition Institute (2010) sum- marized soil test P and K levels in North America from an extensive study of 4.4 million soil samples. These studies revealed that approximately 45% of the soil samples tested below P or K critical levels, meaning that they required annual fertilization to avoid profit loss in most major crops. Similarly, P or K application could be less than crop removal for one or more years for 55% of the samples. Similar summaries of soil fertility in North America were completed in 2001 and 2005. Comparing soil test levels across the period from 2001 to 2005 at a state or province scale dem- onstrated that most of North America showed no change in soil-P levels, Figure 14. Comparison of nutrient removal by crops in the United States to nutrient suggesting that the current practices applied as fertilizer, recoverable manure, or fixed by legumes (average of were roughly maintaining existing 2006–2008). soil-P fertility (Fixen et al. 2006). From 2005 to 2010, however, soil-P of recoverable manure is farm-level of recoverable manure-N in the na- fertility declined significantly in the excess and may not be usable as a nu- tional nutrient budget increases the N Corn Belt where P budgets were nega- trient source because of transportation in excess of nonlegume removal by tive. Potassium changes across the de- costs. Recent increases in fertilizer 1.3 MMt, for a total excess of approx- cade were more complex. Nearly the prices have raised the affordability of imately 4.8 MMt. Recoverable ma- entire Great Plains showed decreases manure transport, but higher fuel costs nure-P more than balances P removal, in soil-K levels, whereas the eastern have had the opposite effect. Inclusion but the K budget remains substantially states or provinces showed no change,

18 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY a slight increase, or a slight decrease. o Advances in biotechnology of However: The highly negative K budgets in the corn and other crops combined o This negative P balance would Great Plains will eventually lower with advances in crop manage- be greater than all the recover- soil-K to critical levels, at which point ment may accelerate the rate of able manure-P in the United K fertilization will need to increase to yield increase over that of the States. Because it is highly sustain improvement in productivity. past several decades and increase doubtful that all manure-P could crop-N demand/removal, result- be transported the distances A Baseline for Future Nutrient ing in upward pressure on fertil- needed to enable it to be used izer use compared with the past. Needs only where crop needs exist, and This could be offset if the seed because most soil-P levels have These historical trends in nutrient industry succeeds in developing either been constant or declin- use and crop removal combined with crop varieties exhibiting increased ing under the current P budget, current status can be used to estab- N use efficiency or if manage- this additional deficit likely will lish a baseline for evaluating nutrient ment practices are adopted that cause loss of soil-P fertility, re- use into the future (Table 3). All six lead to increased efficiency. sulting in higher fertilizer-P rec- parameters included show a decidedly o Expansion of biofuel production, ommendations and higher P use linear trend line from 1986 through stimulating additional increases than the baseline projections. 2010, simplifying projections of sta- in corn acreage, increased crop tus quo relationships into the future. o Higher P prices should encour- biomass removal, or shifting of These are not projections of expected age greater use of soil test- Conservation Reserve Program nutrient needs or nutrient balance, but ing, better soil sampling, more land to crops managed for bio- rather a projected baseline from which sophisticated P placement, and mass could all increase N use and such estimates of needs and balance more judicious general P man- removal, resulting in additional can be made. agement, which should increase upward pressure on fertilizer use. P application accuracy and ef- • Baseline fertilizer-N consumption o The net effect would be that ficiency. This should lessen the is increasing at a slightly higher these changes could offset mak- magnitude of P use increases. rate than N removal and should ing the original baseline projec- increase by 44% in 2050, which o As with N, advances in crop tion of a 44% increase in fertil- would increase N in excess of crop genetics and technology could izer-N application reasonable, or removal from 3.6 MMt currently to increase crop removal and efficiency gains could be large 4.8 MMt. However: fertilizer-P use. enough to lessen this increase. o Higher N prices and environ- o As with N, biofuel developments • Baseline fertilizer-P consump- mental concerns are likely to could increase fertilizer-P use. tion is increasing at a much slower speed both development and o rate than crop removal and should The net effect indicates that adoption of efficiency-enhancing increase only 12% by 2050, which it seems highly probable that N sources and practices (IPNI would further increase the an- fertilizer-P use will need to 2012) and will likely slow this nual fertilizer-P O deficit from increase more than the baseline rate of fertilizer growth. 2 5 -0.8 MMt currently to -2.5 MMt. would predict.

Table 3. Baseline projections to 2050 of the balance between U.S. fertilizer consumption and crop nutrient removal using rates of change from 1986 through 2010

2050 Balance 2007 Trend Line 1986–2010 Rate of Change* 2020 Projection 2050 Projection (applied-removed) Million metric tons (MMt) Fertilizer-N applied 11.3 0.0974/yr 12.6 15.5 4.8 Crop-N removal 7.7 0.0702/yr 8.6 10.7

Fertilizer-P2O5 applied 4.1 0.0113/yr 4.2 4.6 –2.5 Crop-P2O5 removal 4.9 0.0511/yr 5.5 7.1

– Fertilizer-K2O applied 4.6 0.0057/yr 4.5 4.3 –6.5 Crop-K2O removal 8.0 0.0665/yr 8.8 10.8

*2009 year excluded for P and K.

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 19 • Baseline fertilizer-K consump- for food resulting from population lasting trend (The Economist 2012). tion is essentially constant, which growth and increases in income in Commercial P2O5 fertilizers are by 2050 would increase the annual developing countries as well as the processed from mined P2O5 rock fertilizer-K2O deficit to -6.5 MMt emergence of the biofuel industry concentrated in geologic deposits in compared with the current level of and the growing role of agriculture various parts of the world. At ap- -3.4 MMt. However: in providing both fuel and food. The proximately 26 MMt/yr (7.5 MMt o This substantial negative K bal- growing demand for food and fuel, P2O5 equivalent), the United States ance with accumulative effects however, is likely to raise the income currently is second only to China in should continue to mine indige- generation capacity of U.S. farmers, P2O5 rock production and contributes nous K from soils of the west- who will become a major contribu- more than 15% of the world’s P2O5 ern Corn Belt and Great Plains. tor in improving the U.S. balance of rock. Based on rock value, cost of The resulting soil-K depletion trade. Continued investment in im- extraction, and 2009/2010 mine pro- will expand the area where K proved genetic materials and other duction, the U.S. Geological Survey fertilization is common practice agricultural technologies is likely to has estimated U.S. P2O5 reserve life and increase fertilizer-K con- enable U.S. agriculture to keep pace at 53 years (USGS 2011a). Life es- sumption above the baseline with growing demand for agricultural timates increase markedly if higher projection. commodities and continue its leading value is attributed to the P2O5 rock role in the global food and fuel sup- as more of the U.S. P resources be- o As with P, higher K prices ply system. The expected growth in come economically minable. The should encourage improved K agriculture, however, will increase the U.S. Geological Survey has estimated management and should have a amount of nutrients consumed in pro- world rock reserve life at 380 years at positive impact on fertilizer-K duction. The extent to which potential 2009/2010 production levels. A study effectiveness. growth of agricultural production will of world P2O5 rock reserves and re- o Genetics, biotechnology, and occur depends on the ability to obtain sources by the International Fertilizer biofuel developments—as the nutrients required through differ- Development Center provided similar with N and P—could increase ent means. estimates for U.S. and world reserves fertilizer-K use. of 69 years and 351 years, respective- o The net effect, as with P, indi- Availability of Fertilizer Raw ly, based on 2009/2010 production levels (Van Kauwenbergh 2010). cates that it seems highly prob- Materials able that fertilizer-K use will Potash is also mined from ex- need to increase more than what The raw materials needed for fer- tensive geological deposits, with the the baseline would predict. tilizers are natural resources. Nitrogen largest known economically min- fertilizer production uses N from the able deposit located in Canada. More This baseline evaluation indi- air and, in most instances, natural gas than one-fourth of the world’s K2O cates that P and K fertilizer use in to create ammonia. The ammonia is production comes from Canada, and the United States will likely need to then either directly applied or used to nearly half the world’s known K2O increase in the future even without manufacture other N fertilizer prod- reserves are located within its bor- adjustments for increases in biofu- ucts. Coal, fuel oil, and naphtha can ders (USGS 2011b). United States el feedstock production or exports. substitute for natural gas. The cost K2O production is currently 0.8 MMt/ Adjustments from the baseline for N of natural gas accounts for 70–90% yr, equivalent to about one-fifth of have greater uncertainty. The follow- of the production cost of ammonia, domestic consumption. World K2O ing section will quantify the impacts so the competitiveness of an ammo- reserve life is estimated at 353 years of biofuels and growth in global food nia plant in global markets is largely at 2009/2010 production levels. As demand on plant nutrient needs in the determined by the local cost of natu- 8 for P2O5, however, the reserve life es- United States. ral gas (TFI 2008). The high cost of timates increase significantly if higher natural gas in the United States rela- value is attributed to the K2O ore. Food and Fuel Impacts on tive to other regions prior to 2010 has The world supply of raw materi- Future Plant Nutrient Needs caused numerous ammonia plant clo- als needed for fertilizers should be United States agriculture is chal- sures. New discoveries of natural gas sufficient to meet anticipated growth lenged by global increases in demand reserves, mostly through the develop- in demand. Future P and K needs in ment of “fracking,” and the prospect the United States will be met to an of further discoveries, however, have 8 Predictions are consistent with the Association of increasing extent by imported raw American Plant Food Control Officials and The Fer- decreased the price of natural gas materials or final products, whereas tilizer Institute concerning N and somewhat higher and are likely to decrease the price of future N needs could be met primar- concerning P and K fertilizers. N-based fertilizers, reversing a long- ily by North American production

20 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY if current natural gas prices remain K fertilizer use will need to increase glucose units that constitutes the globally competitive. or more nutrients will need to be remain part of the cell walls of plants. covered and recycled from farm and Double crops. Planting a second crop Conclusions and nonfarm waste streams. after the first has been harvested. If fertilizer-N use follows the trend Evapotranspiration. The process that ecommendations R of the last 20 years, it will increase takes water from the soil, passes it Escalating population is the prima- 44% by 2050. Improved N use effi- through the plant, and releases wa- ry driver in increasing crop production ciency could lessen this increase but ter vapors into the atmosphere. and nutrient use. In addition to popu- likely will not eliminate it. Therefore, Fracking. High-volume hydraulic lation gains, per capita caloric intake U.S. agriculture will conceivably be fracturing or horizontal industrial- in developing regions will increase more dependent on commercial fertil- scale gas drilling to access deep much more than in developed regions, izer N in 2050 than it is today. Higher shale formations. further enlarging food and nutrient prices of natural gas in the near past demand. decreased the domestic capacity of N, Societal food and fuel needs are yet it has to be expanded to meet the Literature Cited growing, and in order to increase cur- challenge of the future. Application of Aden, A. 2007. Water usage for current and rent scale of production under de- future N may need to be more precise future ethanol production. Southwest creasing amounts of land, a reliable to address environmental concerns. Hydrol 6 (5): 22–23. supply of plant nutrients to replace Phosphate reserves in the United Alston, J. M., J. M. Beddow, and P. G. Pardey. 2009. Agricultural research, productivity, those removed by cropping is es- States are estimated to be equivalent to and food prices in the long run. Science sential. Commercial fertilizers are a 50- to 70-year supply at current pro- 325 (5945): 1209–1210. responsible for 40 to 60% of current duction rates but increase substantially Alt, K., C. T. Osborn, and D. Colacicco. 1989. Soil Erosion: What Effect on Agricultural U.S. food production. For more than at higher rock prices. Domestic K2O 40 years, removal of the three primary production is equivalent to only ap- Productivity? Agriculture Information Bulletin No. 556. Economic Research plant nutrients by crops in the United proximately 20% of consumption with Service, U.S. Department of Agriculture, States has been increasing linearly at world reserves equivalent to 350 years Washington, D.C. rates of 70,000, 51,000, and 67,000 or more at current production rates. Anderson, D. P., J. L. Outlaw, H. L. Bryant, Mt of N, P2O5, and K2O per year, re- The increased demand for biofu- J. W. Richardson, D. P. Ernstes, J. M. spectively. If crop yields increase at a els has a direct impact on the use of Raulston, J. M. Welch, G. M. Knapek, B. K. Herbst, and M. S. Allison. 2008. The faster rate in the future, these rates of nutrients. There is an urgent need to Effects of Ethanol on Texas Food and removal will likely increase further. support research and development to Feed. Research Report 08-1. The Agri- For more than 20 years, P and K decrease total land requirements for cultural and Food Policy Center, Texas fertilizer use in the United States has biofuel production by integrating ani- A&M University, College Station. been increasing at considerably lower mal feed production with cellulosic Beachy, R. 2010. American Association for rates than crop removal, with increas- the Advancement of Science, http://www. biofuel production and to recover and aaas.org/news/releases/2010/0625beachy. es of 11,000 Mt of P2O5 per year and recycle key plant nutrients during bio- shtml (12 September 2012) essentially no change in K2O use. The fuel production. Beinroth, F. H., H. Eswaran, and P. F. Reich. combination of soil nutrient reserves 2001. Global assessment of land quality. Pp. 569–574. In D. E. Stott, R. H. Mohtar, indigenously present or developed lossary from past fertilization and the use G and G. C. Steinhardt (eds.). Sustaining the Global Farm—10th International of recoverable manure nutrients has Anoxic zone. An area deficient in oxygen. Soil Conservation Organization Meeting. supplemented the negative nutrient West Lafayette, Indiana. 1,169 pp. budgets of the past 20 years to sustain Anthropogenic. Relating to the influ- Bruinsma, J. 2009. The resource outlook to the productivity increases during this ence of human beings on nature. 2050: By how much do land, water and period. Evaluation of existing soil fer- Biofuel. A fuel composed of or crop yields need to increase by 2050. FAO Expert Meeting on How to Feed the tility levels and the supply of recov- produced from biological raw World in 2050, Rome, Italy, June 24–26. erable manure nutrients suggest that materials. Campbell, J. E., D. B. Lobell, R. C. Genova, the nutrient budget deficits must be Biomass. Plant materials and animal and C. B. Field. 2009. The global poten- lessened in the future to sustain cur- waste used especially as a source tial of bioenergy on abandoned agricul- rent productivity increases. Any addi- of fuel. ture lands. Environ Sci Technol 42 (15): 5791–5794. tional nutrients removed from crop- Biorefinery. The processing facility Chen, X., H. Huang, and M. Khanna. 2012. ping systems as a result of bioenergy that converts crops to fuels. Land use and greenhouse gas implica- production will intensify the need for Cellulosic. Of, relating to, or made tions of biofuels: Role of technology and more balanced P and K budgets. The policy. In Agricultural and Applied Eco- from cellulose, a polysaccharide of net effect would be that future P and nomics Association 2011 Annual Meeting,

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 21 Pittsburgh, Pennsylvania, July 24–26. & Phosphate Institute, http://www.ipni. Hicks. 2007. Technological learning in Christian, D. G., A. B. Riche, and N. E. Yates. net/ipniweb/portal.nsf/0/20dee100daa8 U.S. ethanol production. Pp. 2232–2237. 2008. Growth, yield and mineral content 6f668525727300746df5/$FILE/Fertil- In 15th European Biomass Conference of Miscanthus x giganteus grown as a ity%20of%20North%20American%20 and Exhibition, Berlin, Germany, May biofuel for 14 successive harvests. Ind Soils.pdf (12 September 2012) 7–11. Crop Prod 28:320–327. Fixen, P. E., T. W. Bruulsema, T. L. Jensen, R. Hochman, G., S. E. Sexton, and D. Zilberman. Crosson, P. 1986. Soil erosion and policy Mikkelsen, T. S. Murrell, S. P. Phillips, 2008. The economics of biofuel policy issues. Pp. 35–73. In T. T. Phipps, P. R. Q. Rund, and W. M. Stewart. 2010. The and biotechnology. J Agr Food Ind Org Crosson, and K. A. Price (eds.). Agricul- fertility of North American soils, 2010. 6 (2). ture and the Environment. Washington, Better Crops 94 (4): 6–8, http://www.ipni. Hochman, G., D. Rajagopal, G. Timilsina, and D.C.: Resources for the Future. 295 pp. net/publication/bettercrops.nsf/0/2D03B D. Zilberman. 2011. The role of inventory Crosson, P. 1997. Will erosion threaten agri- 3D343C281738525797D0061C037/$FI adjustments in quantifying factors caus- cultural productivity? Environment 39 LE/Better%20Crops%202010-4%20p6. ing food price inflation.Policy Research (8): 4–31. pdf (12 September 2012) Working Paper Series 5744, The World Crosson, P. 1998. The on-farm economic costs Food and Agriculture Organization of the Bank. of soil erosion. Pp. 495–511. In R. Lal, United Nations (FAO). 2006. World International Energy Agency (IEA). 2007. W. H. Blum, C. Valentine, and B. A. Agriculture: Towards 2030/2050. Global Medium-Term Oil Market Report. July, Stewart (eds.). Methods for Assessment Perspectives Studies Unit, Interim Report. http://www.oilmarketreport.org (12 Sep- of Soil Degradation. CRC Press, Boca UN–FAO, Rome. tember 2012) Raton, Florida. 558 pp. Food and Agriculture Organization of the International Food Policy Research Institute Dale, B. E. 2008. Biofuels: Thinking clearly United Nations (FAO). 2008. FAOSTAT, (IFPRI). 2002. Sustainable Food Security about the issues. J Agr Food Chem http://faostat3.fao.org/home/index.html for All by 2020—Proceedings of an Inter- 56:3885–3891. (12 September 2012) national Conference, Bonn, Germany, 4–6 Dale, B. E., M. S. Allen, M. Laser, and L. R. Food and Agriculture Organization of the September 2001. A 2020 Vision for Food, Lynd. 2009. Protein feeds coproduction in United Nations (FAO). 2010. The State Agriculture, and the Environment. 296 pp. biomass conversion to fuels and chemi- of Food Insecurity in the World 2010. International Plant Nutrition Institute (IPNI). cals. Biofuel Bioprod Bior 3:219–230. Rome, Italy. 2010. Soil Test Levels in North America: Dale, B. E., B. D. Bals, S. Kim, and P. Eranki. Gardner, B. 1987. Causes of U.S. farm com- Summary Update. Publication No. 30- 2010. Biofuels done right: Land efficient modity programs. J Polit Econ 95 (2): 3110. IPNI, Norcross, Georgia. 42 pp. animal feeds enable large environmental 290–310. International Plant Nutrition Institute (IPNI). and energy benefits.Environ Sci Technol Gitiaux, X., J. Reilly, and S. Paltsev. 2011. 2012. 4R Plant Nutrition Manual: A 44 (22): 8385–8389. Future Yield Growth: What Evidence Manual for Improving the Management Deaton, A. and J. Drèze. 2009. Food and nutri- from Historical Data. Joint Program of Plant Nutrition, North American Ver- tion in India: Facts and interpretations. Report Series 199. Massachusetts Insti- sion. T. W. Bruulsema, P. E. Fixen, and Econ Polit Weekly 44 (7): 42–65. tute of Technology Joint Program on the G. D. Sulewski (eds.). IPNI, Norcross, Diouf, J. 2009. World must double food pro- Science and Policy of Global Change, Georgia. duction by 2050: FAO chief. World Food Cambridge, Massachusetts, http://global- International Water Management Institute Programme, http://www.wfp.org/content/ change.mit.edu/pubs/all-reports.php (12 (IWMI). 2007. Water for Food, Water world-must-double-food-production- September 2012) for Life: A Comprehensive Assessment 2050-fao-chief (12 September 2012) Graff, G. D., D. Zilberman, and A. B. Bennett. of Water Management in Agriculture. In- Donner, S. D. and C. J. Kucharik. 2008. 2009. Correspondence: The contraction ternational Water Management Institute, Corn-based ethanol production compro- of agbiotech product quality innovation. Earthscan, London, U.K. mises goal of reducing nitrogen export Nat Biotechnol 27 (8): 702–704. (Supple- Johnston, A. M. and N. R. Usherwood. 2002. by the Mississippi River. P Nat Acad Sci mentary information available on Nature Crop nutritional needs. Pp. 13–21. Plant 105:4513–4518. Biotechnology website, http://www.nature. Nutrient Use in North American Agricul- Dyson, T. 1999. World food trends and com/nbt/index.html [12 September 2012]). ture. PPI/PPIC/FAR Technical Bul. 2002- prospects to 2025. P Natl Acad Sci Greene, N. 2004. Growing Energy: How 1. International Plant Nutrition Institute 96:5929–5936. Biofuels Can Help End America’s Oil (IPNI), Norcross, Georgia. Evans, A. 2009. The Feeding of the Nine Bil- Dependence. Natural Resources Defense Karlen, D. L., M. J. Mausbach, J. W. Doran, lion: Global Food Security for the 21st Council, New York, New York. 78 pp. R. G. Cline, R. F. Harris, and G. E. Century. Royal Institute of International Hall, C. A. S., S. Balog, and D. J. R. Murphy. Schuman. 1997. Soil quality: A concept, Affairs, Chatham House, London. 2009. What is the minimum EROI that a definition, and framework for evaluation. Farrell, A. E., R. J. Plevin, B. T. Turner, A. sustainable society must have? Energies Soil Sci Soc Am J 61:4–10. D. Jones, M. O’Hare, and D. M. Kam- 2:25–47. Kellogg, R. L., C. H. Lander, D. C. Moffitt, men. 2006. Ethanol can contribute to Hall, C. A. S., B. E. Dale, and D. Pimentel. and N. Gollehon. 2000. Manure Nutrients energy and environmental goals. Science 2011. Seeking to understand the reasons Relative to the Capacity of Cropland and 311:506–508. for different energy return on investment Pastureland to Assimilate Nutrients: Spa- Federico, G. 2005. Feeding the World; an Eco- (EROI) estimates for biofuels. Sustain- tial and Temporal Trends for the United nomic History of Agriculture 1800–2000. ability 3 (12): 2413–2432. States. USDA/NRCS/ERS Publication Princeton University Press, Princeton and Heller, M. C., G. A. Keolian, and T. A. Volk. No. nps00-0579. Oxford. 2003. Life cycle assessment of a willow Khanna, M. and C. Crago. 2012. Measuring Fixen, P. E., T. W. Bruulsema, A. M. Johnston, bioenergy cropping system. Biomass indirect land use change with biofuels: R. L. Mikkelsen, T. S. Murrell, C. S. Bioenerg 25:147–165. Implications for policy. Ann Rev Res Snyder, and W. M. Stewart. 2006. The Hettinga, W. G., H. M. Junginger, M. M. Econ 4 (September). Also Agr Policy Fertility of North American Soils. Potash Hoogwijk, A. J. McAloon, and K. B. Briefs, Department of Agricultural and

22 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY Consumer Economics, University of national Food Policy Research Institute, Policy Research Institute, Washington, Illinois at Urbana-Champaign, http:// Washington, D.C. 19 pp. D.C. 224 pp. www.farmdocdaily.illinois.edu/2011/03/ Oldeman, L. R. 1994. The global extent of Searchinger, T., R. Heimlich, R. Houghton, F. measuring_indirect_land_use_ch.html land degradation. Pp. 99–118. In D. J. Dong, A. Elobeid, J. Fabiosa, S. Tokgoz, (12 September 2012). Greenland and I. Szabolcs (eds.). Land S. Hayes, and T-H. Yu. 2008. Use of U.S. Khanna, M., J. Scheffran, and D. Zilberman. Resilience and Sustainable Land Use. croplands for biofuels increases green- 2010. The Handbook of Bioenergy Eco- CABI, Wallingford, U.K. 576 pp. house gases through emissions from land nomics and Policy. Springer, New York. Oldeman, L. R. 1998. Soil Degradation: use change. Science 319:1238–1240. Kim, S. and B. Dale. 2004. Global potential A Threat to Food Security? Report Sexton, S. and D. Zilberman. 2011. Land for bioethanol production from wasted crops 98/01. International Soil Reference and food and fuel production: The role of and crop residues. Biomass Bioenerg Information Centre, Wageningen, the agricultural biotechnology. Pp. 269–288. 26:361–375. Netherlands. In J. S. G. Zivin and J. M. Perloff (eds.). Laser, M., H. Jin, K. Jayawardhana, B. Dale, Oldeman, L. R., R. T. A. Hakkeling, and W. G. The Intended and Unintended Effects and L. Lynd. 2009. Projected mature Sombroek. 1991. World Map of the Status of U.S. Agricultural and Biotechnology technology conversion scenarios for con- of Human-Induced Soil Degradation: Policies. University of Chicago Press, version of cellulosic biomass to ethanol An Explanatory Note. International Soil London, U.K. with coproduction of thermochemical Reference and Information Centre and Sexton, S., D. Zilberman, D. Rajagopal, and fuels, power and/or animal feed protein. United Nations Environment Programme. G. Hochman. 2009. The role of biotech- Biofuel Bioprod Bior 3:231–246. 41 pp. nology in a sustainable biofuel future. Lindert, P. H. 2000. Shifting Ground: The Paltsev, S., H. D. Jacoby, J. M. Reilly, Q. J. AgBioForum 12 (1): 130–140. Changing Agricultural Soils of China and Ejaz, F. O’Sullivan, J. Morris, S. Rausch, Shapouri, S. and S. Rosen. 2004. Fifty years Indonesia. MIT Press, Cambridge, Mas- N. Winchester, and O. Kragha. 2011. of U.S. food aid and its role in reducing sachusetts. 365 pp. The Future of U.S. Natural Gas Produc- world hunger. Amber Waves 2 (4). Taken Lobell, D. B., W. Schlenker, and J. Costa- tion, Use, and Trade. Report 186. MIT from Food Security Assessment, GFA-15, Roberts. 2011. Climate trends and global Joint Program on the Science and Policy of USDA–ERS, May 2004. crop production since 1980. Science 333 Global Change, Cambridge, Massachusetts. Slater, J. V. and B. J. Kirby. 2011. Commercial (6042): 616–620. Pierce, F. J., W. E. Larson, R. H. Dowdy, and Fertilizers 2010. Association of Ameri- Lubowski, R. N., M. Vesterby, S. Bucholtz, W. A. P. Graham. 1983. Productivity of can Plant Food Control Officials, Inc., A. Baez, and M. J. Roberts. 2006. Major soils: Assessing long-term changes due Columbia, Missouri. Uses of Land in the United States, 2002. to erosion. J Soil Water Conserv 38 (1): Smil, V. 2001. Enriching the Earth. MIT Press, USDA–ERS. Economic Information Bul- 39–44. Cambridge, Massachusetts. 640 pp. letin No. (EIB-14), May. 54 pp. Pimentel, D. and T. W. Patzek. 2005. Etha- Sokhansanj, S., S. Mani, A. Turhollow, A. Ku- Ma, H., J. Huang, and S. Rozelle. 2004. Reas- nol production using corn, switchgrass, mar, D. Bransby, L. Lynd, and M. Laser. sessing China’s livestock statistics: An and wood: Biodiesel production using 2009. Large-scale production, harvest analysis of discrepancies and the creation soybean and sunflower.Nat Resour Res and logistics of switchgrass (Panicum of new data series. Econ Dev Cult 14 (1): 65–72. virgatum)—Current technology and en- Change 52 (2): 445–473. Pirie, N. W. 1978. Leaf Protein and Its By- visioning and mature technology. Biofuel Matthews, J. A., H. Tan, M. J. B. Moore, and Products in Human and Animal Nutrition. Bioprod Bior 3:124–141. G. Bell. 2011. A conceptual lignocel- Cambridge University Press, Cambridge, Stewart, W. M., D. W. Dibb, A. E. Johnston, lulosic “feed + fuel” biorefinery and its U.K. 224 pp. and T. J. Smyth. 2005. The contribution application to the linked biofuel and cattle Potrykus, I. 2010. Regulation must be revolu- of commercial fertilizer nutrients to food raising industries in Brazil. Energ Policy tionized. Nature 466 (7306): 561. production. Agron J 97:1–6. 39 (9): 4932–4938. Richard, T. R., R. Baxter, G. Camargo, G. The Economist. 2012. Fracking: Landscape Murrell, T., P. E. Fixen, W. Huang, J. L. Kovar, Feyereisen, and J. Baker. 2010. Cover with well, http://www.economist.com/ and P. M. White Jr. 2011. Nitrogen, crops as energy crops: A hidden treasure. node/21558462 (6 December 2012) phosphorus, and potassium requirements Paper presented at the Global Conference The Fertilizer Institute (TFI). 2008. Sup- to support a multi-billion gallon biofuel on Sustainable Biofuels, Minneapolis, ply & Demand, Energy Drive Global industry. Pp. 160–176. In R. Braun, D. Minnesota. Fertilizer Prices, https://www.chsinc. Karlen, and D. Johnson (eds.). Sustain- Ringler, C. 2006. Global Food, Feed, Fiber com/portal/chs/server.pt/gateway/ able Alternative Fuel Feedstock Opportu- and Bioenergy Demand Prospects: Impli- PTARGS_0_9261_32114_0_0_18/ nities, Challenges and Roadmaps for Six cations for Natural Resources—Drawing CHS%20Supply%20and%20De- U.S. Regions. Soil and Water Conserva- on the Millenium Ecosystem Assessment. mand%20Brochure%20-%20march%20 tion Society, Ankeny, Iowa. International Food Policy Research Insti- 16%20-%20sm.pdf (26 April 2010) Naylor, R., A. Liska, M. Burke, W. Falcon, tute, Washington, D.C. Trigo, E. J. and E. J. Cap. 2003. The impact J. Gaskell, S. Rozelle, and K. Cassman. Rosen, S., S. Shapouri, K. Quanbeck, and of the introduction of transgenic crops in 2007. The ripple effect: Biofuels, food B. Meade. 2008. Food Security Assess- Argentinean agriculture. AgBioForum 6 security and the environment. Environ- ment, 2007. Outlook Report No. GFA-19. (3): 87–94. ment 49 (9): 30–43. USDA–Economic Research Service, Tyner, W. E. 2008. The U.S. ethanol and Nelson, G. C., M. W. Rosegrant, J. Koo, R. Washington, D.C. 55 pp., http://www. biofuels boom: Its origins, current status, Robertson, T. Sulser, T. Zhu, C. Ring- ers.usda.gov/Publications/GFA19/ (12 and future prospects. BioScience 58 (7): ler, S. Msangi, A. Palazzo, M. Batka, September 2012) 646–653. M. Magalhaes, R. Valmonte-Santos, Rosengrant, M. W., M. S. Paisner, S. Meijer, United Nations–Department of Economic and M. Ewing, and D. Lee. 2009. Climate and J. Witcover. 2001. Global Food Pro- Social Affairs, Population Division (UN). Change: Impact on Agriculture and Costs jections to 2020: Emerging Trends and 2007. World Population Prospects: The of Adaptation. Food Policy Report. Inter- Alternative Futures. International Food 2006 Revision. New York.

COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY 23 U.S. Department of Agriculture (USDA). U. S. Department of Energy/U.S. Department Resources. Technical Bulletin IFDC-T75. 2009. Summary Report: 2007 National of Agriculture (USDOE/USDA). 2005. International Center for Soil Fertility Resources Inventory. Natural Resources Biomass as Feedstock for a Bioenergy and Agriculture Development (IFDC), Conservation Service, Washington, D.C., and Bioproducts Industry: The Techni- Muscle Shoals, Alabama. and Center for Survey Statistics and cal Feasibility of a Billion-Ton An- Wiebe, K. 2003. Linking Land Quality, Agri- Methodology, Iowa State University, nual Supply. 78 pp., http://www.ornl. cultural Productivity, and Food Security. Ames, Iowa. 123 pp., http://www.nrcs. gov/~webworks/cppr/y2001/rpt/123021. Agricultural Economic Report No. 823. usda.gov/Internet/FSE_DOCUMENTS/ pdf (13 September 2012) Resource Economics Division, Economic stelprdb1041379.pdf (12 September U.S. Geological Survey (USGS). 2011a. Research Service, U.S. Department of 2012) Phosphate rock. Pp. 118–119. In Mineral Agriculture. U.S. Department of Agriculture (USDA). Commodities Summaries, http://minerals. Zilberman, D., J. Zhao, and A. Heiman. 2012. 2012. Crop Production Historical Track er.usgs.gov/minerals/pubs/commodity/ Adoption versus adaptation, with empha- Records: April 2012. National Agricul- phosphate_rock/mcs-2011-phosp.pdf (13 sis on climate change. Annu Rev Resour tural Statistics Service (NASS). 239 pp., September 2012) Econ 4 (1): 27–53. http://usda01.library.cornell.edu/usda/ U.S. Geological Survey (USGS). 2011b. Pot- Zilberman, D., G. Hochman, D. Rajagopal, nass/htrcp//2010s/2012/htrcp-04-25-2012. ash. Pp. 122–123. In Mineral Commodi- S. Sexton, and G. Timilsina. 2012. The pdf (13 September 2012) ties Summaries, http://minerals.er.usgs. impact of biofuels on commodity food U.S. Department of Agriculture– gov/minerals/pubs/commodity/potash/ prices: Assessment of findings.Am J Agr Economic Research Service (USDA– mcs-2011-potas.pdf (13 September 2012) Econ, http://ajae.oxfordjournals.org/con- ERS). 2008. http://ers.usda.gov (1 Van Kauwenbergh, S. J. 2010. World tent/early/2012/06/07/ajae.aas037.extract November 2012) Phosphate Rock Reserves and (1 November 2012)

CAST Member Societies, Companies, and Nonprofit Organizations AMERICAN ASSOCIATION OF AVIAN PATHOLOGISTS n AMERICAN ASSOCIATION OF BOVINE PRACTITIONERS n AMERICAN BAR ASSOCIATION, SECTION OF ENVIRONMENT, ENERGY, & RESOURCES–AGRICULTURAL MANAGEMENT n AMERICAN DAIRY SCIENCE ASSOCIATION n AMERICAN FARM BUREAU FEDERATION n AMERICAN MEAT SCIENCE ASSOCIATION n AMERICAN METEOROLOGICAL SOCIETY, COMMITTEE ON AGRICULTURAL AND FOREST METEOROLOGY n AMERICAN SOCIETY FOR NUTRITION n AMERICAN SOCIETY OF AGRICULTURAL AND BIOLOGICAL ENGINEERS n AMERICAN SOCIETY OF ANIMAL SCIENCE n AMERICAN SOCIETY OF PLANT BIOLOGISTS n AMERICAN VETERINARY MEDICAL ASSOCIATION n AQUATIC PLANT MANAGEMENT SOCIETY n COUNCIL OF ENTOMOLOGY DEPARTMENT ADMINISTRATORS n CROPLIFE AMERICA n DUPONT PIONEER n ELANCO ANIMAL HEALTH n IOWA SOYBEAN ASSOCIATION n MONSANTO n NATIONAL PORK BOARD n NORTH CENTRAL WEED SCIENCE SOCIETY n NORTHEASTERN WEED SCIENCE SOCIETY n NOVUS INTERNATIONAL n POULTRY SCIENCE ASSOCIATION n SOCIETY FOR IN VITRO BIOLOGY n SOCIETY OF NEMATOLOGISTS n SYNGENTA CROP PROTECTION n THE FERTILIZER INSTITUTE n UNITED SOYBEAN BOARD n WEED SCIENCE SOCIETY OF AMERICA n WESTERN SOCIETY OF WEED SCIENCE n WINFIELD SOLUTIONS, A LAND O’LAKES COMPANY

The mission of the Council for Agricultural Science and Technology (CAST) is to assemble, interpret, and communicate credible science-based information regionally, nationally, and internationally to legislators, regulators, policymakers, the media, the private sector, and the public. CAST is a nonprofit organization composed of scientific societies and many individual, student, company, nonprofit, and associate society members. CAST’s Board is composed of representatives of the scientific societies, commercial companies, nonprofit or trade organizations, and a Board of Directors. CAST was established in 1972 asaresultof meeting sponsored in 1970 by the National Academy of Sciences, National Research Council. ISSN 1070-0021

Additional copies of this Issue Paper are available from CAST. Carol Gostele, Managing Scientific Editor, http://www.cast-science.org.

Citation: Council for Agricultural Science and Technology (CAST). 2013. Food, Fuel, and Plant Nutrient Use in the Future. Issue Paper 51. CAST, Ames, Iowa.

24 COUNCIL FOR AGRICULTURAL SCIENCE AND TECHNOLOGY