Raw Material Resource Base
The United States has abundant forests and croplands, favorable climates, accessible capital, and sophisticated technologies for a strong biobased industry. As agriculture productivity and silviculture productivity continue to increase, more biomass will be available to support a biobased industry. Advances in biotechnology will keep a continuous supply of new crops flowing into the marketplace. The United States has substantial resources to invest in a carbon economy based on renewable resources.
Conversion of industrial production to the use of renewable resources will require abundant and inexpensive raw materials. The three potential sources of such materials are agricultural and forest crops and biological wastes (e.g., wood residue or corn stover). The amount of each resource available for biobased production will depend on how much these crops are consumed by competing uses and how much land is dedicated to crops grown for industrial uses. The land and other agricultural resources of the United States are sufficient to satisfy current domestic and export demands for food, feed, and fiber and still produce ample raw materials for biobased industrial products except for massive fuel production.
Forests are a major source of raw materials for the production of wood products. The amount of land supporting the nation's forests has remained relatively constant since 1930 (USDA, 1995). Heightened public
interest for forest preservation has led to government policies that support conversion of federal forest lands to special uses such as parks and wildlife areas that prohibit timber production. As these competing uses for national federal forests intensify, increases in timber harvesting on private forestlands will have to offset timber production declines on U.S. public lands (NRC, 1997; USDA, 1995).
Productivity from silviculture and timber harvests has increased on forest lands. The average volume of standing timber per hectare (800 × 109 cubic feet in 1991) is now 30 percent greater than it was in 1952 (USDA, 1995). Forest growth nationally has exceeded harvest since the 1940sa trend that was accelerating until very recently. In 1991 forest growth exceeded harvest by 22 percent, even though the harvest was 68 percent greater than in 1952. More recently (1986 to 1991), the proportion of timber harvested from the total forested land has increased, primarily as a result of increased harvesting on industrial forestlands. The U.S. Department of Agriculture (USDA) Forest Service is forecasting further increases in the nationwide volume of harvested timber from slightly over 16 billion cubic feet in 1991 to nearly 22 billion cubic feet in 2040 (USDA, 1995).
Production capacity of timbered forestland may be underused. Softwood residues are generally in high demand as feedstocks, but hardwood timber residues have less demand and fewer competing uses. Under-utilized wood species include southern red oak, poplar, and various small-diameter hardwood species (USDA, 1995).
In the future forestlands may be planted to silviculture crops for use in bioenergy production. Bioenergy crops may confer a number of benefits such as low maintenance requirements, high yields, and environmental advantages. The USDA and the U.S. Department of Energy (DOE) have field tested several short-rotation woody crop species (harvested on a cycle of 3 to 10 years), including hybrid poplar, black locust, eucalyptus, silver maple, sweet gum, and sycamore. Certain woody feedstocks have yields averaging 4.5 to 7.5 dry tons per acre per yeartwo to three times the yields normally achieved by traditional forest management in the United States. Even higher yields occur under certain conditions. Recent results show potential yields that consistently reach 8.9 dry tons per acre per year in several locations (Bozell and Landucci, 1993).
Cropland acreage is the third major use of land in the United States. The most notable trend in cropland use is the movement of cropland from crop idling programs into crop use and out again (ERS, 1997a). Four principal cropscorn, wheat, soybean, and hayaccounted for nearly 80
percent of all crops harvested in 1996. Current use of commodity crops for industrial uses is low. Coproduction of grain crops such as corn for both food and ethanol fuels will help reduce any future conflicts inherent in allocating renewable resources to two important human needs: food and fuel.
The United States has long been the world's largest producer of coarse grains. Recent data indicate that domestic grain production makes up approximately 67 percent of the world's grain supply (USDA, 1997a). In 1998 the United States exported over 37 million metric tons of corn grain (ERS, 1999). According to the USDA, U.S. feed grain production is projected to increase steadily through 2005. Expected increases in production are due to increasing yields, except for corn, where more acreage also accounts for gains in some years. Corn yields are expected to increase 1.7 bushels per acre per year based on the long-term trend. Corn plantings are expected to remain at or above 80 million acres throughout the next decade (USDA, 1997a). Continuing gains in U.S. agricultural productivity will extend the resource base available for biobased crop production.
U.S. corn yield nearly tripled between 1950 and 1980 from an average of about 35 bushels per acre to over 100 bushels per acre (OTA, 1980). Based on these gains, analysts predicted that corn yields would exceed 120 bushels per acre by 1995. This has, in fact, occurred: the 1992 and 1994 corn yields exceeded 130 bushels per acre (NASS, 1994; USDA, 1997a). Of the 252 million metric tons of corn produced in the 1996 to 1997 marketing year, approximately 19 million metric tons (7 percent of the total corn grain production) were allocated to industrial uses (industrial starch, industrial alcohol, and fuel alcohol) (ERS, 1997b). To the extent that we understand the many factors contributing to crop yield, productivity increases in many cases will be enhanced by improvements in plant genetics, pest management, and soil quality. At the current rate of growth, another 19 million metric tons of corn could become available by the turn of the century.
Perennial grasses and legumes are being evaluated as potential energy crops (Hohenstein and Wright, 1994). These grasses include Bahia grass, Bermuda grass, eastern gama grass, reed canary grass, napiergrass, rye, Sudan grass, switchgrass, tall fescue, timothy, and weeping love grass. Legumes that have been tested include alfalfa, bird's-foot trefoil, crown vetch, flatpea, clover, and Sericea lespedeza. In 1992 about 150 million tons of hay (more than half of which was alfalfa and alfalfa mixtures) were harvested from 59 million acres of croplands in the United States (USDA, 1994). In 1994 hay was harvested from approximately 61 million acres in the United States (ERS, 1997a). Considerable preproduction research now focuses on the facile conversion of some of these materials into fermentable sugars. Thick-stemmed perennial grasses, such as en-
ergy cane and napiergrass, produce yields from 5.4 to 14.5 dry tons per acre per year. These are current yields and likely would increase following selection. They may one day be grown and used on a large scale.
Enhancing the Supply of Biomass
The amount of cropland that will actually be used to supply biobased processors depends on a demand for the final product, and the inputs used to make that product must be competitively priced. Industrial processors bid for corn and forages based on processing costs and product prices in the petrochemical and specialty chemical industries. Some industries that produce specialty starches and lactic acid plastics can bid grain and productive croplands away from food processors now. But some industrial products, such as grain-based ethanol, may not be able to compete with food producers even after considerable declines in grain or forage prices. Even with anticipated new technology, grain-based ethanol probably will not compete with petroleum fuels on a cost basis (Kane et al., 1989). Similarly, access to major commodity plastics markets, like ethylene, may require very low-cost feedstocks (Lipinsky, 1981). The amount of land devoted to crops for biobased industries will depend on economics, as tempered by agricultural policies.
Some resources that are not useful for food production may soon become more suitable for industrial products because processing technologies that use woody biomass are improving. Potential supplies from three sourcescrop residues, wood wastes, and Conservation Reserve Program landare discussed below. Other available biomass wastes (e.g., municipal solid waste) also may be potential sources of lignocellulosic materials. These reserves may provide the best odds for competitive production of biobased industrial products.
The United States produces abundant wastes that are potential raw materials for biobased products. It is estimated that 280 million metric tons per year of biological wastes are currently available (refer to Table 2-1). Much of this is crop residues, predominantly from cornabout 100 million metric tons of corn residues are produced annually (Gallagher and Johnson, 1995). To a lesser extent, paper mill, wood, and municipal solid waste also are important. Approximately 5.6 million metric tons of unused wood residue is generated in all U.S. sawmills (Smith et al., 1994). Crop residues represent a major untapped source of carbon-rich raw materials available onsite at a low to negligible cost. However, expenses for collection, storage, and transport must be considered in using these bulky,
low-valued residues. Sufficient biological wastes exist to supply the carbon for all 100 million metric tons of organic carbon-based chemicals consumed annually in the United States as well as to provide part of the nation's fuel requirements (Morris and Ahmed, 1992). Production of industrial products from agricultural wastes can reduce competition for agricultural resources.
Conservation Reserve Program
There is potential that land idled by the Conservation Reserve Program (CRP) could be used to grow biobased crops. This federal program was initiated in 1986 to help owners and operators of highly erodible crop-lands conserve and improve the soil and water resources on their farms and ranches through long-term land retirement. The CRP provides monetary incentives for farmers to retire environmentally sensitive lands from crop production for 10 to 15 years and to convert them to perennial vegetation. The 1996 Federal Agricultural Improvement and Reform Act limited enrollment to 36.4 million acres through the year 2002 (ERS, 1997a).
Some CRP lands may be suitable for harvest of perennial grasses and energy crop production while preserving soil and wildlife habitat. Judicious harvesting on a fraction of CRP lands might be consistent with wildlife and wetlands preservation. Field-scale studies are under way to quantify changes in soil and water quality and native biodiversity due to production of biomass energy crops on former agricultural lands (Tolbert et al., 1997; Tolbert and Schiller, 1996). Grass production on CRP lands could enhance biomass supply: at least 46 million tons of additional feed-stock would be available if one-half of CRP lands was available. This figure assumes low yields of biomass (approximately 2.5 tons per acre), and these yields could increase up to 10 tons per acre for some crops (e.g., switchgrass).
Land costs in the CRP are a barrier to the biobased industry. Land values are high because the federal government must at least match the opportunity of foregone profits from continued production of annual crops such as corn or wheat. Average rental costs under the 1997 CRP are about $40 per acre (Osborn, 1997). Without the CRP and with conservation requirements on these lands, biomass production might be competitive with lower land rental rates for pasture; comparable rental rates for midwestern pasture are about $20 per acre. If the rental rate for CRP lands fell to the pasture rate, the cost of producing switchgrass could decline; presently switchgrass production costs are about $40 per ton (Park, 1997). CRP revision for energy crop harvest is a contentious issue. Furthermore, some have argued that reduced land rental costs for energy crops are a de facto subsidy (Walsh et al., 1996). The potential of using CRP lands to grow biomass energy crops is a topic that merits further investigation.
The total biomass is sufficient to easily meet current demands for biobased organic chemicals and materials. High-value chemicals are not expected to require large acreages. Future demands for biobased commodity chemicals potentially can be met with biomass from waste resources and crops grown on some CRP lands. While biobased materials
such as lumber, cotton, and wool do have substantial markets, these products now compete successfully for land resources.
Coproduction of human food and animal feed products such as protein with biobased products is expected to help prevent future conflicts between production of food and biobased fuels. Corn-based refineries, for example, yield protein for animal feed and oil, starch, fiber, and fuel alcohol products. In the case of pulp and paper mills, pulp, paper, lignin byproducts, and ethanol can be produced while recycling waste paper in a single system. Current demands for liquid fuel are being met with current production of corn grain. If policymakers chose to increase ethanol fuel production beyond the capacity for coproduction of food and liquid fuel, biobased crops grown for energy uses could compete for land with food production. Opportunities for coproduction of food, feed, liquid fuels, organic chemicals, and materials are described in more detail in Chapter 4.
Filling the Raw Material Needs of a Biobased Industry
The foundation of a biobased industry depends on an abundant supply of plant materials. Raw materials such as starches, cellulose, and oil can already be extracted from plants for the production of biomaterials, chemicals, and fuels. The committee envisions that many more plant substances (e.g., biopolymers or chiral chemicals) may serve as raw materials for industrial applications in the future. While conventional breeding methods continue to play an important role in developing new crops and cultivars, genetic engineering of existing crops will greatly enhance the number and precision of such modifications and the variety of plant products available for industrial use. Introduction of new crops for bio-based production will be limited without an adequate infrastructure for cultivar research, development, and commercialization.
Satisfying the raw material needs of expanding biobased industries will require crops with the following characteristics: contain biomolecules and biochemical systems with potential industrial applications; can be manipulated to produce desirable molecules; can sustain a high level of predictable raw material production; and are supported by an infrastructure for biomass harvesting, transfer, storage, and industrial processing.
Renewable resources have been used for a wide variety of industrial purposes. For example, siliviculture crops have been used for many years
as construction materials; wood can undergo additional processing to yield a variety of other products such as paper and textiles. Because agricultural and silvicultural crops are highly variable, plant parts become more valuable when they can be further separated into their biochemical components. In the long term, plant parts that can be converted to sugars for fermentation are likely to become a major feedstock for the production of biobased chemicals, fuels, and materials.
Woody Plant Parts (Lignocellulosics)
Woody plant parts are an abundant biological resource for the bio-based industry. Wood is a complex material composed of carbohydrate and lignin polymers that are chemically and physically intertwined. Considerable energy is required to separate the wood polymers from each other. Much of the harvested wood is used for lumber, and both wood and other woody materials are used for pulp production.
Improvements in wood processing are leading to new biomaterials that may replace plastic products currently produced from petrochemical sources. For instance, improved resistance to insects and decay fungi, dimensional stability, hardness, and other properties result when wood is esterified, cross-linked, and impregnated (Stamm, 1964). Wood that has been plasticized by decrystallization and esterification can be further shaped by injection- and extrusion-molding processes. Although these processes are currently economically impractical, they might provide many new products from wood-flour and wood-fiber-reinforced thermoplastic composites in the future. Scientists and engineers will continue to refine industrial processes that can be used to produce useful biomaterials from wood.
Separated Plant Components
Plant matter contains hundreds of components that are useful inputs for biobased production. Plants are primarily carbohydrates (cellulose, other polysaccharides such as starch, and sugars), lignins, proteins, and fats (oils). Starch and sugar polymers are end products of photosynthesis and dominate a plant's carbohydrate reserves. Accessing these vast carbohydrate reserves will be key to maintaining a renewable source of raw materials that can substitute for petrochemicals.
Cellulose is a carbohydrate polymer composed of glucose and constitutes about 45 percent of woody plant parts. Cellulose can be isolated by
pulping processes and then further processed to yield such chemicals as ethanol and cellulose ethers; cellulose acetate, rayon, and cellulose nitrate; cellophane; and other cellulosics. Many of these derivatives have only specialty applications because their cost is high relative to that of petrochemical-derived polymers.
Numerous sources of cellulose pulp can be used for chemical production. The primary source of wood cellulose pulp comes from conifer species (Smith et al., 1994), but hardwood uses have increased in the past two decades. Flax residue (flax tow) and kenaf are grown commercially for pulp production. In other countries, pulp is made from crop residues such as straw and sugarcane bagasse. Because of its dominating abundance in plants, cellulose will always be a primary feedstock of any biobased industry.
Hemicelluloses are composed of carbohydrates based on pentose sugars (mainly xylose) as well as hexoses (mainly glucose and mannose). Hemicelluoses make up 25 to 35 percent of the dry weight of wood and agricultural residues; they are second only to cellulose in abundance among carbohydrates. While use of hemicellulose is currently limited, quantities of hemicelluloses, pectins, and various other plant polymers are abundant in residues and have great potential in the production of chemicals and materials.
Lignin is a phenylpropane polymer that holds together cellulose and hemicellulose components of woody plant matter. Lignin constitutes about 15 to 25 percent of the weight of lignocellulose. Lignin has not yet been used as a raw material for industrial use in large quantities. Concerted attempts by pulp and paper research laboratories to develop new markets for byproduct lignins have had only limited success (Bozell and Landucci, 1993). Production of low-molecular-weight compounds from kraft lignin (phenols in particular) similarly has not yet proved commercially competitive. This reflects the chemical complexity of lignin and its resistance to depolymerization. Nevertheless, a recent DOE study concluded that pyrolysis of lignocellulosics (lignin, cellulose, and hemicellulose plant tissues) could make production of phenolics and anthraquinone from lignin competitive, and the potential also exists to produce benzene, toluene, and xylenes from lignin via pyrolysis (Bozell and Landucci, 1993). Lignocellulose pretreatment receives special attention in this report because it will be a key step for realizing the presently untapped potential of abundant lignocellulosic materials found in wood and other perennial crops.
Starch is the principal carbohydrate reserve of plants. Corn starch currently is a primary feedstock for starch-based ethanol, plastics, loose-fill packing material, adhesives, and other industrial products. Approximately 600 million bushels of corn went into production of industrial products during the marketing year 1995 to 1996; of that total, 395 million bushels were used to produce fuel ethanol (ERS, 1996b). While the supply of corn starch has been sufficient to meet current demands, primarily anhydrous motor fuel grade and industrial ethanol, other supplies of sugar feedstocks are being evaluated to meet anticipated increases in demand for oxygenated fuels and chemicals.
Proteins are the primary means of expressing the genetic information coded in DNA. These polymers are based on building blocks of amino acid monomers whose sequence is predetermined by a genetic template. The sequence diversity of proteins is responsible for the wide array of functions performed by proteins in living organisms (OTA, 1993) (see Box 2-1). A variety of plant proteins might one day be commercially exploited as materials, but current understanding of the structural properties of most plant proteins is limited.
One of the few well-understood plant proteins is zein, an abundant protein in corn seeds. Zein makes up 39 percent of the kernel protein, or about 4 percent of the kernel weight. The protein has several properties of industrial interest, such as the ability to form fibers and films that are tough, glossy, and grease and scuff resistant. Zein resists microbial attack and cures with formaldehyde to become essentially inert. In addition, it is water insoluble and thermoplastic.
The USDA Northern Regional Research Laboratory in Peoria, Illinois, developed zein into a textile fiber in the late 1940s. Scientists generated the fiber by dissolving zein in alkali, extruding the solution through spinnerets into an acid coagulating bath, and then curing the product with formaldehyde. Zein fibers are strong, washable, and dyeable and possess other desirable properties. The Virginia-Carolina Corporation commercialized zein-based fiber as ''Vicara," producing about 5 million pounds in 1954. However, the company discontinued manufacture-shortly thereafter, perhaps because of the advent of comparable synthetic fibers. Zein's main use today is as a water-impermeable coating for pharmaceutical tablets, nuts, and candies. It also functions as a cork binder for gaskets and bottle-cap liners, a binder in ink, a varnish, and a shellac substitute. The advantageous properties of zein suggest that its industrial usefulness merits reexamination (Wall and Paulis, 1978). There is potential to alter
the structural properties of zein by genetic engineering to produce novel characteristics.
Many crops can serve as sources of plant oils; currently soybeans account for 75 percent of the vegetable oil produced in the United States.
Soybean crops are a major target of plant oil research. Approximately 305 million pounds of soybean oil were used in nonfood applications such as livestock feed and the manufacture of resins, plastics, paints, inks, and soaps in 1996 (ERS, 1997b). Fatty acids derived from soybean oil are being converted into surfactants, emulsifiers, and alkyd resins for paints. Soybean oil can be chemically transesterified to produce biodiesel (methyl esters). In the future biodegradable lubricants may be produced from soybean oil; genetically modified soybean varieties hold the promise of yielding lubricant products that outperform petroleum-based lubricants (ERS, 1997b). Additional oilseed crops, some yielding oils with unusual properties, could be grown in the United States (e.g., petroselenic acid from coriander oil) (see Box 2-2). Over the near term the volume of oil produced for such uses will remain small relative to petrochemical sources.
Fermentable sugars are by far the largest feedstock that might support a biobased chemicals industry in the United States. A wide range of fermentable sugars can be found in crops and wastes from agriculture and silviculture. Major feedstocks include corn, wheat, sorghum, potato, sugarbeet, and sugarcane; other sources include potato-processing residues, sugarbeet and cane molasses, and apple pomace (Polman, 1994). Sugars can be produced directly or derived from polysaccharides (such as cellulose and starch) and then, via microbial fermentation, used to produce a wide range of commodity and specialty chemicals. Existing commercial fermentations primarily utilize glucose (6-carbon sugar) to produce ethanol, acetic acid, amino acids, antibiotics, and other chemicals. Over the long term new sources of glucose will be required to meet the demands of a biobased industry. Growth of a biobased chemicals industry will depend on production of cellulose-rich crops, including those currently under production (e.g., corn and alfalfa) and others that presently are not grown commercially (e.g., switchgrass and hybrid poplar).
Significant increases in glucose reserves are available from lignocellulosic substances found in most plants, crop residues, and waste paper. Cellulose can be hydrolyzed by acid to glucose, although much of the glucose is destroyed during this process. The second most abundant sugar, found in hardwood and agricultural residues, is xylose derived from xylan hemicelluloses. Xylose is relatively easily recovered by acid or enzymatic hydrolysis but can be fermented to ethanol only by a few naturally occurring organisms or recombinant microbes. The practical sugar yield from lignocellulosics would increase significantly if commercial fermentations could utilize xylose (a 5-carbon sugar or pentose) as well as glucose (a 6-carbon sugar or hexose). Novel genetically engi-
neered microorganisms will eventually play a key role in the direct conversion of cellulose oligomers and 5- and 6-carbon sugars to ethanol.
To avoid destruction of sugars from lignocellulosic materials by acid treatment, enzymatic hydrolysis using mixtures of enzymes (cellulases and hemicellulases) is used. These enzymes, when combined with effec-
tive pretreatments of lignocellulosics, provide high yields of glucose, xylose, and other fermentable sugars with minimal sugar losses. However, these enzymes are currently too costly to use in large-scale conversion of lignocellulosic materials to fermentation substrates.
Improving Plant Raw Materials
The discoveries occurring in plant and microbial genomics will advance the fundamental biological research needed to support a biobased industry. Scientific investigations are under way to decipher the genetic code of a flowering plant, Arabidopsis thaliana; the genetic map is complete for the microbial organisms Saccharomyces cerevisiae (common yeast), Bacillus subtilis (gram-positive bacteria), Escherichia coli (gram-negative bacteria), and Caenorhabditis elegans (nematode). In addition, it is expected that complete genomic sequences of Drosophila, humans, and several other eukaryotic species will become available in the foreseeable future. The genetic information collected on these organisms will provide researchers with insights on the genes that control plant traits and cellular processes (NRC, 1997). For example, understanding of the functions of the Arabidopsis genes will permit identification of a desirable gene, which on transfer to a different plant will have the gene's functions expressed there (e.g., the manufacture of a particular chemical).
Genetic engineering is perhaps the most significant development in plant biology in the past two decades. It has profound implications for understanding the fundamental processes of plant growth, development, and metabolism and for generating new agricultural and forest products. It is now possible to modify an organism genetically so that the modified plant or microbe produces greater quantities of a particular polymer. It is also possible to transfer an entire biological process into new organisms.
Biomass production can be improved by development of new cultivars and crops with enhanced agronomic traits. Researchers will need to use both traditional plant breeding and genetic engineering techniques to improve yield and pest resistance of traditional crops and new crops. Techniques like genetic fingerprinting and markers can be used to facilitate classical plant breeding. The identification of crop strains and the durability of genes that confer resistance to pests and environmental stress will contribute to enhanced productivity, as it already does with food and feed crops.
Significant screening efforts are needed to identify carbohydrates, lipids, and proteins with industrial potential. Development of commercial plant sources for these compounds through breeding and genetic engineering will require elucidation of the genes and enzymes responsible for the production of compounds in the source organism (plant,
(table continued on next page)
animal, or microbe). As our understanding of plant metabolism continues to improve, scientists will be able to manage more sophisticated manipulations of these systems to produce the desired biochemicals in the desired quantities. Separations of plant components for industrial uses can also be improved by genetic engineering.
Biochemical pathways and genes can be mobilized within plants to create new products based on molecules that originate from nonplant sources such as microorganisms. Further, biomolecules often can be modified to facilitate purification. Such capabilities have no parallel in
(table continued from previous page)
coal and petroleum-based feedstock systems and are a major potential advantage for biobased products.
Well-developed technologies do exist for transferring genes into plants using bacterial plasmids or particle acceleration (see Box 2-3). Many plant species have thus far been transformed. Nevertheless, more efficient methods for transformation and regeneration would improve development of commercial products. Most current transformation methods rely on the time- and resource-intensive process of producing large numbers of transgenics that are subsequently screened for gene activity and "true-to-type" plants. The pace of discovery and product introduction could be
accelerated if transgenes could be inserted at specific sites in the plant genome that consistently yield the desired gene activity level and that have little or no effect on the plant's agronomic fitness.
Transformation of "elite" germplasm is proving to be another methodological difficulty in genetic engineering. The present approach is to transform varieties most amenable to the process and then transfer the transgenes to elite germplasm by plant breeding. This is an inefficient process because evaluation and selection phases of research and development are prolonged, creating delays in the commercialization phase. Direct transformation and site-specific gene insertion might be especially useful for perennial plants because cycle times for evaluation and breeding of these plants are long, laborious, and impractical.
The current battery of promoters for gene expression (see Box 2-3) will be insufficient to meet the sophisticated expression profiles of the future. For example, redirection of carbon from starches and oils to other biopolymer (e.g., polyhydroxy alkanoate) production may be desired in specific cells of seed tissue during specific stages of seed formation. Further, subsequent degradation of the biopolymer during germination may be essential to permit carbon use for seedling growth. Such "fine tuning" of plant metabolism will require an extensive set of promoters from which appropriate selections can be made.
Biobased research should support improvement of plant productivity. First, any advances in research to improve crops will be relevant to biobased industrial crops. Second, the modern techniques used here for dealing with insect pests, pathogens, weeds, and stress are mostly transferable to plant design. High-level plant productivity on a consistent basis will be essential for supplying biomaterials for industrial production. Genetic engineering can enhance plant productivity by the introduction of traits that reduce farm inputs (e.g., pesticides, fertilizers, water), increase farm productivity, or modify biochemical content. Plant biotechnology products such as NewLeaf potato and Roundup Ready soybean are the precursors of a new generation of products to be commercialized over the next decade. Most of these products have been designed to improve agricultural crop productivity for food and feed uses. It is anticipated that the next generation of crops will be introduced to a broader range of markets, including biobased industries.
Environmental impacts from the release of genetically modified organisms continues to be an area of concern. An earlier report of the National Research Council concluded that crops modified by molecular and cellular methods should pose no risks different from those modified
by classical genetic methods for similar traits (NRC, 1989). A finding of relevance is that established confinement options are as applicable to field introductions of plants modified by molecular and cellular methods as to introductions of plants modified by classical genetic methods.
Resistance to Insect Pests
Biotechnology companies are commercializing transgenic seed with resistance to insects in some major agricultural crops. Current approaches to insect resistance are based on expression of genes from the bacterium Bacillus thuringiensis (Bt) in plants. Bt genes produce toxins that can control caterpillars and certain families of beetles and mosquitoes. Bt genes are target specific and not efficacious against other major pests that cause significant yield losses. Additional genes will therefore be required to successfully manage the range of insects that are crop pests or that transmit plant viruses (e.g., whitefly, aphids). Moreover, pests continuously evolve, and the current generation of genes cannot provide the spectrum and durability of resistance required over the long term (NRC, 1996a).
Resistance to Plant Pathogens
Plant breeding will continue to be a predominant tool of defense against many plant diseases. While classical breeding techniques are fundamental to disease management, genetic engineering is becoming increasingly important. The first example of genetic engineering for disease resistance involved a gene encoding the coat protein of the tobacco mosaic virus (TMV) that was introduced into tobacco plants through an Agrobacterium vector (Abel et al., 1986). These plants were resistant to TMV as well as other closely related viruses. Expression of coat protein genes has become increasingly important in developing new varieties of resistant crops. Research is under way to transfer viral-coat proteins into other horticultural and field crops (Fitchen and Beachy, 1993). In 1995 a transgenic squash seed variety conferring this resistance characteristic was commercialized, and virus-resistant papaya, melon, tomato, potato, and other crops are in advanced stages of development. It is anticipated that transgenic techniques will enlarge the pool of disease-resistant genes that can be introduced into susceptible crop varieties (NRC, 1996a).
Managing plant fungal disease has attracted significant attention from biotechnologists, but progress has been slow. Bacterial chitinase reportedly confers resistance to Botrytis, a major fruit tree pathogen. Researchers claim that introduction of the resveratrol pathway into tobacco confers Rhizoctonia and Botrytis resistance (Hain et al., 1993). Several antifungal proteins have been identified in plant roots and seeds (AFP1
and AFP2 from radish and osmotin from tobacco), and their expression appears to delay infection by fungi (Terras et al., 1992). Whether the resistance demonstrated experimentally will be commercially useful remains to be seen. Based on the successes with insect and viral-vector control, the major missing link is identifying several efficacious antifungal genes (either proteinaceous or nonprotein) that can be used concurrently.
Researchers have identified some genes responsible for disease resistance in plants (Staskawicz et al., 1995). Natural disease-resistant genes generally produce proteins that recognize a ligand produced by an invading pathogen, causing cells exposed to the pathogen to undergo "hypersensitive response" (HR) and cell death. This process physically contains the pathogens and also produces one or more signals that activate the whole plant's defense system. This "systematic acquired resistance" (SAR) apparently confers broad protection against diverse pathogens. The mechanisms of HR and SAR are viable targets for engineering disease resistance into plants. The specificity of R genes to specific ligands, however, makes it difficult to confer multirace resistance to pathogens using these genes unless R genes can be identified that recognize certain basic ligands present in and essential to the pathogen. An alternative approach would be to gain an understanding of the HR and SAR mechanisms and introduce these ''activation" mechanisms by way of other systems. Sophisticated promoters that can turn the HR or SAR processes "on" might be essential in these efforts.
Weeds cause major losses in crop productivity and lower crop quality. Introduction of conservation-tillage, reduced-tillage, and no-tillage practices to lower soil erosion have been accompanied by use of broad-spectrum herbicides to control weeds previously managed by mechanical means. It is expected that in the near term weed management will be dominated by pesticides and by use of genetically engineered or classically bred crop varieties. Research leading to development of new herbicides and transgenic crops will involve efforts of scientists from industry and academe. Several herbicides with low risks to human health and the environment have been developed (e.g., glyphosate, sulfonylureas, imidazolinones, and glufosinate; Kishore and Shah, 1988).
Some low-toxicity herbicides can only be used safely on a narrow spectrum of crops, thus limiting their utility. Researchers have successfully introduced resistance genes into crop varieties to protect crops against herbicides. For example, sulfonylurea-tolerant soybean, imidazolinone-resistant corn, and glyphosate-resistant soybean varieties have been commercialized.
Resistance to Environmental Stress
In addition to pests, various abiotic agents lower plant productivity. Extremes in temperature, water, and salinity create plant stress and lead to declines in plant growth, productivity, and quality. In the 1980s, drought caused corn yield losses and led to contamination with aflatoxins in many parts of the United States. High soil moisture was responsible for major crop losses in the Midwest in 1993. Knowledge of the genes conferring resistance to abiotic stresses is improving, and further research can be expected to minimize environmentally induced yield losses. Recent research has identified certain genes that might mitigate these impacts, such as the biosynthetic genes for fatty acids, sugars, and amino acids (Yoshida et al., 1995; Tarczynski et al., 1993). Some genes may be introduced into plants via conventional or molecular techniques to increase a plant's tolerance to various stress conditions. Other genes might reduce costs associated with weather-related crop damage and result in yield gains. For example, fertilizer inputs may be reduced by growing plants that make more efficient use of nitrogen, phosphorus, and sulfur or that can fix atmospheric nitrogen. Development of such cultivars could potentially reduce nitrate contamination of groundwater caused by fertilizer seepage. A fundamental understanding of plant nitrogen metabolism will be essential to identify genes controlling plant utilization of nitrate and nitrogen. Researchers have recently cloned some important genes (e.g., glutamine synthetase, asparagine synthetase, glutamateoxoglutarate aminotransferase, nitrate reductase, nitrate and ammonia carriers) and are now beginning to elucidate their ectopic expression on plant growth and metabolism (Tsai and Corruzzi, 1993).
Research to enhance productivity, improve processing characteristics, and reduce the time required for harvest could lead to crops designed specifically for industrial applications. For example, a large quantity of biomass is left on the field in the form of plant residues. In the future it may be desirable to redesign plants to maximize harvested plant biomass for industrial processing. If dwarf corn plants could produce two to three ears instead of one or two, harvested biomass and grain yields might double from 150 to 250 bushels per acre up to 300 to 500 bushels per acre. While this scenario is unlikely in the near term, such yield enhancement could lower the costs of biobased production and enhance the competitiveness of biobased industrial raw materials. Conversely, corn fiber might be improved for industrial uses, in which case larger plants, rather than dwarf plants, might be desired.
Altered Biochemical Content
Genetic engineering allows scientists to manipulate the biochemical content of plants in unprecedented ways. The ability to access and alter the expression of biochemicals in seeds and other harvested components will be of particular importance to biobased crop research. The majority of grain crops do produce diverse carbohydrates, oils, and proteins. To exploit this chemical diversity, scientists will need to gain a more detailed knowledge of the plant genes and enzymes regulating these biochemicals.
Significant progress has occurred in altering the carbohydrate chemistry of agricultural crops. For example, potatoes containing novel carbohydrates have been commercialized and similar technologies are being applied to corn and wheat (Kishore and Somerville, 1993). While potatoes produce less starch than grain crops on a per-acre basis, starch-enhanced potatoes are being commercialized in some value-added markets such as food processing (see Box 2-4). Potatoes producing amylose-free starch are providing a desirable starch source for some markets because these modified potatoes contain only amylopectin rather than a mixture of amylose and amylopectin. Potatoes capable of fructan production will yield a new polymer based on fructose instead of glucose. Microbes can metabolize fructose and industrial processes can convert fructose to a wide range of organic compounds. Fructose is an example of the many sugars that are becoming important feedstock to the chemical industry.
Other polysaccharides such as cellulose, pullulan, hyaluronic acid, guavan, and xylans have interesting material, polymer, and fiber properties. The genes and enzymes involved in the biosynthesis of specific molecules will need to be identified and expressed in their natural hosts or engineered into other organisms. In some cases, genes may be modified to alter enzyme substrate specificity and accumulation of new polymers with different ionic charge, chemical reactivity and stability, solubility, melting, and other thermoplastic properties. The availability of genetic mutants in Arabidopsis, corn, and other plants will accelerate this research. A complete understanding of plant carbohydrate metabolism in unmodified and modified plant tissues will promote application of these sophisticated engineering technologies to other applications.
Researchers have identified several plant genes and enzymes involved in lipid metabolism over the past five years, enabling development of
modified oilseed rape cultivars for the biotechnology industry. Because of the similarity in lipid metabolism across plant species, similar oil research is under way with soybean, sunflower, and corn crops. Genes have been identified that affect carbon chain length, degree of unsaturation, and substituents in the fatty acid hydrocarbon chain (Topfer et al., 1995). Numerous fatty acids have been identified by screening the composition of various plant species. These plant species will provide a gene source for creating cultivars of major agronomic crops that produce fats and oils of value to the chemical industry as lubricants, fuels, and detergents. Introduction of additional functionalities, such as hydroxy or epoxy groups or double- and triple-carbon bonds, into plant fatty acids will enable synthesis of new molecules in major oilseed crops.
Many benefits may be derived from research that improves productivity and biochemical characteristics of some oil-producing crops. For example, the ricinoleic acid present in castor bean oil can be used for the production of nylon 11; the erucic acid found in crambe and rapeseed oils can be used for nylon 13 production; and the petroselenic acid present in coriander can be used for the production of nylon 66. High-linolenic oils present in the seed can be used to produce various coatings, drying agents, and printing inks. Significant opportunities may exist to improve the agronomic productivity of some of these oil crops and develop applications for the fatty acids and byproducts.
The oil pathway of plants may serve as a platform for the production of novel biopolymers. Industrial scientists have recently succeeded in transferring genes from the bacterium Alcaligenes eutrophus into the plant Arabidopsis. This modification led to plant production of poly(hydroxybutyrate) (PHB); this is an example of biopolymer engineering that can be performed on plants (Poirier et al., 1995). PHB constitutes nearly 20 percent of leaf dry matter in the genetically engineered Arabidopsis. Researchers at Zeneca and Monsanto are now transferring PHB genes into oilseed rape and soybean seeds for production of the polymer. Scientists anticipate that cotton fiber quality could be improved if PHB genes could be introduced into cotton plants. Future work will expand beyond PHB and focus on the production of diverse polymers that vary in carbon chain length and substitution. This work will require a more detailed understanding of fatty acid metabolism and the microbial pathways involved in polymer formation. Related research leading to identification of inexpensive processes for the extraction and separation of these polymers also will be critical for developing industrial applications.
(table continued on next page)
Little genetic engineering research has focused on proteins other than enzymes, although there are several advantages to using protein polymers for biobased production:
• Plant proteins generally are more diverse than other plant polymers.
• Molecular weight and amino acid sequences of protein polymers can be precisely regulated.
• Proteins can catalyze reactions or be hydrophobic, hydrophilic, neutral, acidic, or basic reactants.
• Proteins can form higher-order structures such as multimers of polymers.
• Plant proteins are generally inexpensive.
• Certain proteins, such as silk and wool, have long histories in the textiles industry.
(table continued from previous page)
The biochemicals that may be manufactured by plants are not limited to carbohydrates, oils, and proteins. Rubber is an important hydrocarbon produced by certain plants, and genetic engineering could enhance the amount and quality of rubber from these sources. Worldwide demand for rubber is growing at a dramatic rate as automobiles are becoming more common in the emerging economies of Asia. The opportunity to meet this need from renewable resources is real and deserves attention. Thus, genetic engineering could significantly enhance the biochemical diversity of the plant world and address some major issues in plant-based industries.
Although this chapter focuses primarily on plant biotechnology, developments in microbial biotechnology will also be key to the expansion of biobased production. Research on microbial systems is addressing the processing of plant products as well as the handling of society's wastes. Increased understanding of metabolic control in microorganisms is point-
(table continued on next page)
(table continued from previous page)
ing to new ways to genetically modify organisms for industrial conversion of plant-derived feedstocks.
Introduction of New Crops
Most of the agricultural research in the United States focuses on major agronomic crops and developing applications for existing biochemicals extracted from these crops. Although this work is worthwhile and should continue, it will be equally important to develop new crops that have the potential to produce desired biochemicals. In addition to selecting and designing crops to meet certain industrial chemical needs, enhanced productivity should be a major goal of crop development efforts. This section briefly explores issues and opportunities associated with the introduction of new crops. Other analyses have described potential markets and botanical details of alternative crops (e.g., Harsch, 1992). Table 2-2 summarizes data for some new crops that have received initial scientific and commercial investments.
Basic technical factors often create difficulties in the commercialization of new crops. Many alternative crops lack characteristics that would result in high yields because they have been neither intensively cultivated nor subject to research and development for improved agronomic traits. Well-tended experimental plots may demonstrate useful genetic potential. However, the absence of directed plant breeding and underlying scientific knowledge generally makes for a long development period prior to major production of a new crop. In contrast the value of traditional crops is continuously being enhanced by technological advances resulting from major investments in research and development.
Biobased crops should be selected based on their productivity of the desired product as well as specific biochemical characteristics. Plant breeding may be necessary to introduce new biomolecules or enhance total biomass production. Sugar cane provides a useful example. Cultivars of sugar cane, called "energy cane," have been developed for ultimate conversion of the sucrose, cellulose, and hemicellulose contents to ethanol. Plant breeders developed cultivars having lower sucrose contents because the higher biomass yields of the cultivars more than compensated for the lower sucrose levels, thereby reducing the raw material's final cost.
Plants of natural origin and genetically engineered crops should be considered during the crop selection process. Castor provides an instructive example. The large-scale reintroduction of this crop is largely driven by a desire to replace the $30 million annual importation of castor oil with a reliable, cost-effective, domestic supply of ricinoleic acid. However, crops that produce high levels of oleic acid, such as sunflower or rape-
seed, are being engineered to contain the gene required to produce hydroxyleic acid, thereby yielding the desired ricinoleic acid in an established agronomic crop.
Over the near term the acreage of traditional crops will continue to dwarf that of new crops. In the long-term, alternative crops can make important contributions in the industrial and agricultural sectorsif they can compete in the marketplace with traditional crops. Industrial crops that will be successful will be those with sufficient registered crop protection chemicals, appropriate infrastructure, optimized manufacturing processes and equipment, and byproduct utilization systems.
If appropriate and sufficiently low-cost processing technologies were developed, there is enough unused biomass to satisfy all domestic demand for organic chemicals that can be made from biological resources (approximately 100 million tons per year) and all of the nation's oxygenated fuel requirements (use of oxygenated gasoline) in areas that did not meet the federal ambient air standard for carbon monoxide as mandated by the Clean Air Act Amendments of 1990. Production of biobased crops on land presently idled could, given low-cost technologies for converting these crops, provide an additional source of U.S. liquid fuels. A few new crops have received initial scientific and commercial investments, but various factors impede their commercial adoption. Nevertheless, certain nontraditional crops, such as switchgrass and hybrid poplar, are valuable because of their high yields.
Classical plant breeding and genetic engineering techniques will continue to be used by scientists for the development of new crops and improvement of well-established crops. Genetic engineering offers unprecedented opportunities to manipulate the biochemical content of specific plant tissues and design a raw material for easier processingan advantage not enjoyed by fossil feedstocks. However, much more remains to be done to provide the raw materials for expanding biobased industries.
Over the long term, a major research priority is to maintain a commitment to fundamental and applied research in the biology, biochemistry, and genetics of plants and microorganisms. It is necessary to gain an understanding of underlying processes associated with gene expression, growth and development, and chemical metabolism. Improved methods of plant transformation and new promoters that further refine gene expression are needed to hasten the development of crops suitable for biobased industries. A sound scientific base in these fundamental areas will be critical to formulating strategies to supply future raw materials for biobased industries.
Future development of agricultural and forest crops for a biobased industry will strengthen the ties between agriculture and industrial production. The change will depend not only on continued improvement of traditional crops but also on the development of alternative crops, genetically engineered cultivars, and separation and fermentation processes that can make use of biomass. Making the transition to a competitive biobased industry will require close coordination between plant scientists and process engineers to develop cost-effective biological and industrial processes for the conversion of raw materials into value-added products.