Increasing Crop Productivity to Meet Global Needs for Feed, Food, and Fuel

Global demand and consumption of agricultural crops for food, feed, and fuel is increasing at a rapid pace. This demand for plant materials has been expanding for many years. However, recent increases in meat consumption in emerging economies together with accelerating use of grain for biofuel production in developed countries have placed new pressures on global grain supplies. To satisfy the growing, worldwide demand for grain, two broad options are available: (1) The area under production can be increased or (2) productivity can be improved on existing farmland. These two options are not mutually exclusive and both will be employed to produce the additional 200 million tonnes/year of corn (Zea mays) and wheat (Triticum aestivum) estimated to be needed by 2017. Both options will alter the environmental footprint of farming. Of the two options, increasing productivity on existing agricultural land is preferable as it avoids greenhouse gas emissions and the large-scale disruption of existing ecosystems associated with bringing new land into production. In the United States, breeders, agronomists, and farmers have a documented history of increasing yield. U.S. average corn yields have increased from approximately 1.6 tonnes/ha in the first third of the 20th century to today’s approximately 9.5 tonnes/ha. This dramatic yield improvement is due to the development and widespread use of new farming technologies such as hybrid corn, synthetic fertilizers, and farm machinery. The introduction of biotechnology traits and development of new breeding methodology using DNA-based markers are further improving yields. Outside the United States, similar farming practices have been adopted in some agricultural nations, but in many major grain-producing countries, yields still lag well behind world averages. By continuing to develop new farming technologies and deploying of them on a global basis, demand for feed, fuel, and food can be met without the commitment of large land areas to new production. Estimates of global meat consumption and grain-based biofuel production. A, Global meat consumption estimates from OECD-FAO (2008). Meat consumption outside of the OECD is expected to increase by 48 million tonnes/year in the next decade. B, Global grain-based biofuel production estimates from FAPRI (2008). Grain-based biofuel production is expected to increase by 28 billion liters/year in the next decade. Global corn and wheat production and consumption estimates from FAPRI’s 2008 U.S. and World Agricultural Outlook All values are in million tonnes/year. Global corn and wheat production and consumption estimates from FAPRI’s 2008 U.S. and World Agricultural Outlook All values are in million tonnes/year. The Food and Policy Research Institute (FAPRI) estimates that an additional 6 million ha of corn and 4 million ha of wheat plus a roughly 12% increase in global corn and wheat yields will be used to produce this additional 200 million tonnes of grain. Both increases in planted area and increases in yield are likely to be needed to meet global demand for grain. However, improving yield on existing agricultural land will have a lower environmental impact than bringing new land into production. Cultivation of new acreage requires land clearing and subsequent tillage that results in significant greenhouse gas emissions (Fargione et al., 2008) also has negative impacts upon biodiversity and water quality (Foley et al., 2005). Increasing the productivity of existing agricultural land will also have environmental consequences (Tilman et al., 2002), but the negative consequences are generally less onerous and in some cases can be positive, depending upon how the land was previously used. Increased use of nitrogen fertilizers, a concern with both methods of increasing production, can increase nitrous oxide emissions, reduce water quality, and increase the size of hypoxic zones (Donner and Kucharik, 2008). However, incremental yield increases can be achieved on existing agricultural land through conservation tillage or transgenic insect control. Conservation tillage can decrease erosion, conserve soil moisture, and increase soil organic matter (Lal, 2004), and transgenic insect control can reduce broad spectrum insecticide use (Qaim and Zilberman, 2003; Cattaneo et al., 2006). Annual corn yield averages and area planted in the United States and the world. Yield rate of gain in the United States from 1961 to 2007 was 0.11 tonnes ha−1 year−1. Global yield rate of gain was about half of this at 0.06 tonnes ha−1 year−1. Global corn area harvested has been increasing at the rate of 0.93 million ha/year. In the United States, corn area harvested increased by approximately 5 million ha in 2007 and 2008, although the long-term trend is much lower at 0.15 million ha/year (FAO, 2008). Lines indicate yield trend line. Average corn yields from high- and low-yielding countries Values are 5-year averages from 2003 to 2007 (FAO, 2008). Yield, harvested area, and production were averaged independently and do not necessarily sum across this table. Average corn yields from high- and low-yielding countries Values are 5-year averages from 2003 to 2007 (FAO, 2008). Yield, harvested area, and production were averaged independently and do not necessarily sum across this table. Higher crop prices, prompted in part by rising demand, have increased costs for urban consumers, especially those in poorer countries. However, higher crop prices will also provide farmers with the economic incentive to invest in farming methods and technologies that improve crop yields (von Braun, 2007; Gallagher, 2008). Raising corn yields in the 10 largest, below average, corn-producing countries to just the world average will result in the production of an additional 100 million tonnes of corn or about 80% of the projected growth in demand by 2017. Implicit in this scenario is the idea that rising yields will markedly diminish the global need for new crop acreage. U.S. corn yield averages compared to yields obtained with widely used hybrids in test plots. A, National and Iowa corn yield averages (USDA ERS, 2008); Duvick’s era hybrid average yields are from Duvick (1997). Iowa state averages are included as the era hybrid experiments were conducted in Iowa. The yield differential between test plots grown in Iowa and the Iowa state averages can be seen to decrease over time from approximately 3 tonnes/ha in 1936 to 1942 to approximately 1.8 tonnes/ha in 1988 to 1991. B, Genetic gain study of DEKALB commercial hybrids released from 2001 through 2006 in the 110-day relative maturity group (RM110), a region of the corn belt stretching across central Iowa. New RM110 commercial hybrids introduced from 2001 through 2006 were tested at 20 locations/year from 2005 through 2007 to produce the reported yield averages. All seed was from the same nursery and none of the hybrids contained biotechnology traits (Trevor Hohls, personal communication). Annual yield improvement was estimated at 0.24 tonnes ha−1 year−1 for this group of hybrids and average yields were 3 tonnes/ha greater than Iowa and 4.2 tonnes/ha greater than the U.S. national average for this set of hybrids. Breeding rates of gain for a multitrait index for 248 corn populations initiated across 3 years. The multitrait index is weighted toward yield, but also incorporates other agronomic traits such as grain moisture and stalk strength (Eathington et al., 2007). Yield advantage of triple-stack corn. Corn hybrids expressing either three biotechnology traits (YieldGard Plus with Roundup Ready Corn2 or YieldGard VT Triple) or without any biotechnology traits were tested in yield trials at the indicated number of locations across the United States in 2005, 2006, and 2007. Average yield values are shown in the bars and the yield difference between triple-stack and nontransgenic corn is indicated in the text above the bars. These are average values from yield trials run across corn-growing regions in the United States. Values can be significantly higher in regions with more insect pressure. Nontransgenic corn was treated with insecticide to control corn rootworm. Yield increase in corn plants expressing cspB, a cold shock protein from Bacillus subtilis. Hybrids from a single transgenic event were tested in yield trials over 3 years at managed stress locations. Yield of the transgenic hybrid (green circles) and nontransgenic isogenic hybrids (white circles) at individual locations are plotted against the yield of all entries tested at that location (Castiglioni et al., 2008). While breeders, agronomists, and farmers are working to increase yields, a number of factors that may reduce yields must be considered. Over the next two decades, climate change effects in the central United States are predicted to increase night air temperatures, the number and severity of adverse weather events, and increase the incidence of insect pests and disease. The result could be a drag on crop yields (Hatfield et al., 2008). Rapid adaptation of crops to changing climatic conditions may help mitigate these effects. Such rapid adaptation may occur for crops supported by strong breeding programs that continuously develop and introduce superior, locally adapted hybrids and varieties. New, higher-yielding hybrids produced from Monsanto’s North American corn breeding program currently have a product half-life of approximately 4 years and are completely turned over about every 7 years, raising hopes that, for corn and possibly soybean, new varieties adapted to local conditions will be produced as a part of the ongoing breeding program. Unfortunately, this is not the case for crops such as wheat and rice (Oryza sativa) that lack the support of large, private breeding programs. Accordingly, increased public support for crop improvement efforts are sorely needed if new wheat and rice varieties are to be adapted to changing local climatic conditions. This is particularly true for regions of the world predicted to undergo more dramatic near term changes in climate than the central United States. Nitrogen is another factor that may limit crop yields. Nitrogen may become less available as the cost of fertilizer rises and the continued growth of eutrophic dead zones and nitrous oxide emissions leads to changes in the way fertilizer is used (Donner and Kucharik, 2008). Nitrogen use efficiency, defined as the amount of crop produced per unit of input, has steadily improved in the United States since the 1980s (Frink et al., 1999). More precise nitrogen applications and genetic improvements in crops are likely to sustain improvements in nitrogen use efficiency although there is a limit to how far nitrogen application can be reduced. A 10 tonnes/ha corn crop contains around 100 kg nitrogen/ha as protein and at least this amount of nitrogen must be added back to the field to maintain fertility. Lastly, a sharp downturn in the global economy could restrict demand for both meat and fuel in ways that reduce the economic incentive to increase crop yields (IMF, 2008). Anticipated impact of improvements in agronomics, breeding, and biotechnology on average corn yields in the United States. Rate of yield improvement due to breeding is extrapolated from observations such as those shown in Figure 3B, using data extending across maturity groups from Monsanto’s North American corn breeding program. Agronomic (planting density, fertilizer use efficiency, improvements in soil management) contributions to the rate of yield improvement are considered to proceed at current historical rates based on estimates in Duvick (2005). Rate of yield improvement for biotechnology traits is a combination of the effects of current yield-protecting biotechnology traits, the introduction of biotechnology traits for drought tolerance, and additional yield-enhancing biotechnology traits. Biotechnology contributions to yield from herbicide tolerance, corn borer, and corn rootworm protection are estimated from the data presented in Figure 5. Biotechnology contributions to yield from drought tolerance are estimated from data presented in Figure 6 and an assumption that drought conditions strong enough to reduce yield will be seen on approximately 10% of the planted acres. Biotechnology contributions from yield-enhancing transgenes assume the introduction of three new biotechnology traits with effects similar to those described by Padgette (2008) over the course of the next decade. In each case biotechnology trait adoption curves such as those observed for current commercially available biotechnology traits are assumed (Monsanto, 2008). Many colleagues at Monsanto have contributed to this review. John Anderson, Marsha Stanton, Beth Calabotta, Dusty Post, Trevor Hohls, and Paolo Castiglioni have all provided data, helpful suggestions on content, or editing. In addition, a much larger group of people have contributed to the research and development that has made the inventions described in this review possible.