5.1 Background on Energy Crops

Beginning in the late 1970s, numerous woody and perennial grass crops were evaluated in species trials on a wide range of soil types across the United States.50 One key outcome of this research was the development of crop management prescriptions for perennial grasses and woody crops. Some highlights of this research are presented below for representative energy crops deemed to have high potential. These crops include three perennial grasses, an annual energy crop (high-yield sorghum), and four woody crops, managed either as a single rotation (i.e., harvest before replanting) or managed as a multi-rotation (i.e., coppicing) crop.

5.1.1 Switchgrass and Other Perennial Grasses

Breeding and selection research on native perennial grasses such as switchgrass (Panicum virgatum), big bluestem (Andropogon gerardii), and indian grass (Sorghastrum nutans) started in 1936 when the USDA at Lincoln, Nebraska, began breeding native grasses to revegetate land damaged by the drought of the 1930s. In 1949, the first cultivar, 'Nebraska 28' switchgrass, was released jointly by the USDA and the University of Nebraska. Since that time, USDA and other scientists have evaluated native collections and selected and bred improved cultivars for most areas of the United States. These initial cultivars were developed for forage and conservation purposes. A full array of establishment and management practices has been steadily refined and improved. Millions of acres of these grasses have been planted. Past and continuing genetic research is leading to the development of bioenergy-specific cultivars with substantial genetic gains. Switchgrass is widely considered the model perennial grass for bioenergy production.

Several characteristics make switchgrass, big bluestem, and indian grass desirable biomass energy crops. They are broadly adapted and native to North America, which reduces the concerns for becoming invasive species. Each has consistently high yields with minimal inputs and is well-suited to marginal land. Additionally. they are relatively easy to establish from seed, and a seed industry already exists. Long-term plot trials and farm-scale studies indicate switchgrass is productive, enhances and protects environmental quality, and is potentially profitable given the establishment of a viable cellulosic biofuels market. Although stands can be maintained indefinitely, they are expected to last at least 10 years, after which time the stands could be renovated and replaced with new, higher-yielding cultivars (Figure 5.1). Currently there are additional public and private breeding programs throughout the United States.

Biology and adaptation. Switchgrass, big bluestem, and indian grass are perennial warm-season grasses that are native to most of North America, except for areas west of the Rocky Mountains and north of 55°N latitude. They grow 3 to 10 feet tall with most of the root mass located in the top 12 inches of the soil profile, according to the long-standing literature. There is variation by species, but the root depth can reach 10 feet with new varieties even deeper. More than 70 years of experience with these grasses used as hay and forage crops demonstrates that they are productive and sustainable on rain-fed marginal land east of the 100th Meridian (Mitchell et al, 2010). This meridian matches the western boundary of Oklahoma (excluding the panhandle), and bisects North Dakota, South Dakota, Nebraska, Kansas, and Texas. They are adapted to a wide range of habitats and climates and have few major insect or disease pests. In addition to potential bioenergy production, these grasses are used for pasture and hay production, soil and water conservation, and wildlife habitat.

Figure 5.1 Switchgrass research plot

(Courtesy of SD State University)

Switchgrass has distinct lowland and upland ecotypes and two primary ploidy levels (chromosome numbers). Tetraploid plants have 36 chromosomes, while octaploid plants have 72 chromosomes. All lowland ecotypes are tetraploids, whereas upland plants can be tetraploids or octaploids. Tetraploids and octaploids do not cross. Additionally, switchgrass ecotypes are differentiated by the latitude of their origin. Ecotypes from the southern United States are not well adapted to the northern United States because of winter kill, and northern ecotypes moved to the southern United States have low productivity. Upland ecotypes occur in upland areas that are not subject to flooding, whereas lowland ecotypes occur on flood plains and low-lying areas (Vogel, 2004). Generally, lowland plants have a later heading date and are taller with larger and thicker stems. Tetraploid lowland and upland ecotypes have been crossed to produce true F1 hybrids that have a 30% to 50% yield increase over the parental lines (Vogel and Mitchell, 2008). These hybrids are promising sources for high-yielding bioenergy cultivars. The lowland ecotypes and the lowland x upland hybrids have the most potential for bioenergy production because of their high yields.

Production and agronomics. In perennial grasses, successful stand establishment in the seeding year is mandatory for economically viable bioenergy production systems (Perrin et al., 2008). Weed competition during establishment is a major reason for stand failure. For example, acceptable switchgrass production can be delayed by at least 1 year due to weeds and poor stand establishment (Schmer et al., 2006). No-till planting has significant cost and environmental benefits. After the establishment year, well-established switchgrass stands require limited herbicides. Nitrogen fertilizer is not recommended during the planting year since nitrogen encourages weed growth, increases establishment cost, and increases economic risk associated with establishment if stands should fail (Mitchell et al., 2008; 2010). In most agricultural fields, adequate levels of phosphorus and potassium will be in the soil profile (Mitchell et al., 2010). Good weed management and favorable precipitation will produce a crop equal to about half of potential production, which can be harvested after frost at the end of the planting year with 75% to 100% of full production achieved the year after planting.

Although switchgrass can survive on low-fertility soils, nitrogen fertilizer is required to optimize yield. The optimum nitrogen rate for switchgrass managed for biomass varies (Mitchell et al, 2008; 2010), but biomass yield declines over the years if inadequate nitrogen is applied, and yield will be sustainable only with proper nitrogen application. Vogel and others (2002) found that for one variety, applying 100 pounds of nitrogen per acre per year optimized biomass, with about the same amount of nitrogen being applied as was being removed by the crop. A general nitrogen fertilizer recommendation for the Great Plains and Midwest region is to apply 20 pounds of nitrogen per acre per year for each ton of anticipated biomass if harvesting during the growing season, with nitrogen rate reduced to 12 to 14 pounds per acre per year for each ton of anticipated biomass if harvesting after a killing frost. The nitrogen rate can be reduced when the harvest is after a killing frost because less nitrogen is removed from the system and some nitrogen is recycled into the roots. Nitrogen is applied as switchgrass greens up in the spring to minimize cool-season weed competition. Spraying herbicides to control broadleaf weeds is typically only needed once or twice every 10 years in established, well-managed switchgrass stands (Mitchell et al., 2010).

Figure 5.2 Baling switchgrass

(Courtesy of ORNL)

Switchgrass can be harvested and baled with commercially available haying equipment (Figure 5.2). Self- propelled harvesters with rotary heads are preferred for harvesting high-yielding (greater than 6 tons per acre) switchgrass fields. Harvesting switchgrass within 6 weeks before killing frost or leaving a stubble height shorter than 4 inches can reduce stand productivity and persistence, whereas harvesting after a killing frost will not damage stands. A single harvest per growing year generally maximizes switchgrass yields, and harvesting after a killing frost ensures stand productivity and persistence. Proper management maintains productive stands for more than 10 years. Round bales tend to have less storage losses than large square bales when stored outside uncovered, but square bales tend to be easier to handle and load without road width restrictions. After harvest, poor switchgrass storage conditions can result in storage losses of 25% in a single year and can reduce biomass quality. Covered storage (e.g., net wrap, tarp, or structure) is necessary to protect the harvested biomass.

Potential yield and production costs. Switchgrass yield is strongly influenced by precipitation, soil fertility, location, and genetics. Most plot- and field-scale switchgrass research has been conducted on forage-type cultivars, selected for other characteristics in addition to yield. Consequently, the forage-type cultivars in the Great Plains and Midwest are entirely represented by upland ecotypes which are inherently lower yielding than lowland ecotypes. Yield data comparing forage-type upland cultivars like Cave-In-Rock, 'Shawnee,' 'Summer,' and 'Trailblazer' do not capture the full yield potential of switchgrass. For example, high-yielding F1 hybrids of 'Kanlow' and Summer produced 9.4 tons per acre annually, which was 68% greater than Summer and 50% greater than Shawnee (Vogel and Mitchell, 2008). New biomass-type switchgrass cultivars will be available soon for the Great Plains and Midwest. In a 5-year study in Nebraska, the potential ethanol yield of switchgrass averaged 372 gallons per acre and was equal to or greater than that for no-till corn (grain + stover) on a rain-fed site with marginal soils (Varvel et al., 2008). These results were based on switchgrass cultivars developed for grazing. Significantly greater yields are expected by the next generation of biomass-specific cultivars.

An economic study based on the 5-year average of 10 farms in Nebraska, South Dakota, and North Dakota indicated producers can grow switchgrass at a farmgate price of $60 per ton (Perrin et al., 2008). Producers with experience growing switchgrass had 5-year average costs of $43 per ton, with a low of $38 per ton. These costs include all expenses plus labor and land costs. This research from nearly 50 production environments indicates that growing switchgrass for cellulosic ethanol could be economically feasible in the central and northern Great Plains, with sufficiently cost-effective fuel conversion and distribution. Fuel and land prices have increased since this study, so the cost increases for those inputs need to be considered when determining switchgrass production costs.

Sustainability. Sustainability is crucial for biomass energy crops. Switchgrass protects soil, water, and air quality, sequesters atmospheric carbon, creates wildlife habitat, increases landscape diversity, returns marginal farmland to production, and could potentially increase farm revenue. In a 5-year study, Liebig et al. (2008) reported that switchgrass stored large quantities of carbon (C), with four farms in Nebraska storing an average of 2,590 pounds of soil organic carbon (SOC) acre/year when measured to a depth of 4 feet across sampled sites.

The energy-efficiency and sustainability of cellulosic ethanol from switchgrass has been modeled using net energy value (NEV), net energy yield (NEY), and the petroleum energy ratio (PER) (Schmer et al., 2008). Switchgrass fields in the Midwest produced 540% more renewable energy (NEV) than non-renewable energy consumed in production over a 5-year period (Schmer et al., 2008). The estimated on-farm NEY was 93% greater than human-made prairies and 652% greater than low-input switchgrass grown in small plots in Minnesota (Tilman et al., 2006). The on-farm study had an estimated PER of 13.1, equivalent to producing 93% more ethanol per acre than human-made prairies and 471% more ethanol per acre than low-input switchgrass in Minnesota (Schmer et al., 2008).

Implementing switchgrass-based bioenergy production systems will require converting marginal land from conservation plantings or annual row crops to switchgrass. Growing switchgrass on marginal sites likely will enhance ecosystem services more rapidly and significantly than on productive sites. There is concern of soil carbon loss associated with converting conservation grasslands such as those in the Conservation Reserve Program to bioenergy crops such as switchgrass. Recent research on converting grasslands to no-till corn demonstrates that using no-till revegetation practices results in no measurable soil carbon loss (Follett et al., 2009; Mitchell et al, 2005).

Switchgrass is the leading perennial grass biofuel feedstock option for the Great Plains and Midwest. Some have questioned if switchgrass is the best choice from an ecological perspective, and contend that diverse mixtures of native plants are ecologically more beneficial and should be considered for biomass production. However, feedstock selection will be determined by the amount of available land and the ability of producers to profit by its production. Managed switchgrass monocultures can produce 1.5 to 5 times more biomass than native tallgrass prairies and seeded polycultures (Table 5.1), which translates into less land being required to produce the necessary biomass and more profit potential for the producer.

An Oklahoma study compared monoculture and polyculture feedstock production managed in a low-input system (no nitrogen fertilizer) (Griffith et al., 2011). The monocultures were switchgrass, sand bluestem (Andropogon hallii Hack.), Old World bluestem (Bothriochloa ischaemum L. Keng), and big bluestem. The polycultures were four grasses, four grasses and four forbs, eight grasses and eight forbs, and Old World bluestem with alfalfa (Medicago sativa L.). Average yield was 2.8 tons per acre for the monocultures and 2.4 tons per acre for the polycultures. For each polyculture, a dominant species emerged by year three, indicating that over time polycultures may be similar to monocultures. These low-input systems produce about half the biomass of managed systems.

Table 5.1 Reported Perennial Grass Yield and Acres Required for a 50-Million Gallon Cellulosic Ethanol Plant

Note: Feedstock requirements for a 50 million gallon biorefinery require 588,000 dry tons of feedstock at a conversion rate of 85 gallons of ethanol per dry ton. Ethanol conversion rate from Biomass Multi-Year Program Plan (U.S. Department of Energy 2011).

a. Low-input, high-diversity man-made prairies (Tilman et al., 2006).

b. Native tallgrass prairie burned in late spring (Mitchell, 1992).

c. Shawnee is an upland forage-type switchgrass cultivar released in 1995.

d. Lowland bioenergy-specific switchgrass in the cultivar release process.

e. F1 hybrid of 'Summer' and 'Kanlow' switchgrass (Vogel and Mitchell, 2008).

Adding perennial grasses into a landscape provides habitat improvements over corn and soybeans, even if the areas are mowed every year. If the grasses fill the landscape, special management practices can be used to optimize the habitat value. These include early summer harvest with regrowth prior to dormancy, or leaving some material standing during winter to provide winter cover and spring nesting habitat.

Conclusions. Characteristics that lead to potential adoption of new crops include profitability for the producer, ability to fit within existing farming operations, ease of storage and delivery to the end user, and availability of extension information on best management practices. Each of these exists for switchgrass. Switchgrass can be harvested after frost when many farmers have completed corn and soybean harvests. The operational aspects of perennial herbaceous cropping systems are fully developed and accepted by farmers, and the economic opportunities on small, difficult to farm, or marginally productive fields are attractive to many farmers (additional considerations are provided in Text Box 5.1).

Large-scale switchgrass monocultures evoke concerns of potential disease and insect pests and the escape of switchgrass as an invasive species, especially since little research has been conducted on these topics. However, the genetic diversity available to switchgrass breeders, the initial pathogen screening conducted during cultivar development, and the fact that switchgrass has been a native component of U.S. grasslands for centuries will likely limit negative pest issues.

Available cultivars and production practices reliably produce 5 tons per acre in the central Great Plains and Midwest, and 10 tons per acre in the Southeast. Improved cultivars and agronomics will increase yields similar to the yield increases achieved in corn in the last 30 years.51 Hybrid switchgrass makes producing 10 tons per acre a reality in the central United States. The availability of adequate land area and the profit potential in a region will determine the success of growing switchgrass for bioenergy. Production practices and plant materials are available to achieve sustainable and profitable biomass production.

TEXT BOX 5.1 | IRRIGATION OF ENERGY CROPS

Irrigation of energy crops can be a contentious issue. Water in the western United States has to meet a number of competing off-stream uses, such as municipal, agriculture, and industrial, as well as providing for hydropower generation and minimum in-stream flows for fisheries. In the West, the majority of water comes as winter precipitation, as rain or snow, and usually water for summer use comes from snow melt or storage.

In the western United States, most crops, including hay crops, are grown under irrigation. Irrigated energy crops will never compete economically with high-value irrigated crops, such as fruits and vegetables, but may be able to compete with lower valued crops such as hay and small grains. One potential energy crop species for irrigation in the western United States is switchgrass (Panicum virgatum) (Fransen, 2009). It is a C4 plant, and as such has higher water use efficiency than C3 plants such as wheat. It is native to many western states except for the Pacific Coast states.

An arena where energy crops may be able to utilize water in the West, without competing with food crops is to utilize water that cannot be used for crops for human consumption, such as from treated sewage waste, food processing, and mining and other industries. Significant quantities of produced water are extracted with the oil, gas, and coalbed methane. Produced water can range from being nearly fresh to being hypersaline brine. There are opportunities to improve the quality through treatment or use the better quality water for synergistic energy co-production (U.S. Department of Energy, 2006). Some of the produced water could be used in feedstock production, especially as new fossil-related extraction systems are developed that use less recycled water in enhanced recovery. In Wyoming, a coal bed methane well produces about 15,000 gallons (0.046 acre-feet) of wastewater per day. This may result in over a million acre-feet of wastewater produced per year in Wyoming. In California, 240,000 acre-feet of municipal waste water was used for agriculture in 2003. There is a goal of utilizing an additional 1 million acre-feet by 2020 and 2 million acre-feet over 2002 levels by 2030. Energy crops may be able to utilize marginal lands, including saline-affected land. In addition to the issue of water use, there is the issue of land competition. In California, 200,000 to 300,000 acres are classified as saline.

For high-valued crops, it may not be desirable to grow these crops 2 years in a row on the same land. While energy crops will not displace high valued crops, there may be opportunities to rotate some annual energy crops with some high-valued crops. Large irrigated acreages in the West are devoted to traditional agronomic crops (e.g. small grains, oilseeds, and forages) that often have low profit margins for the grower. For example, in California, low-value crops are grown on 5.5 out of 9 million acres. There may be opportunities to integrate energy crops into forage/grain/oilseed/sugar crop rotations. Some grasses may be able to produce biomass under limited irrigation, when other traditional crops might not produce a product (e.g. feed suitable for livestock feed). Grasses response to limited irrigation is species specific. Of course, the decision by producers as how to utilize their land and water will be market-and value-based.

Because irrigated lands can be highly productive, land rents are high (e.g. can be $200 per acre in the Columbia basin). This requires high yield from energy crops. For switchgrass, a yield of 11 dry tons per acre is achievable in the Columbia Basin. Water can cost $15 to $50 per acre plus costs for repairs, labor, and energy. Total irrigation costs can be in the range of $120 to 140 per acre. Presupposing the availability of water, profitable and competitive energy crop production requires high yields to offset irrigation costs.

5.1.2 Giant Miscanthus – Miscanthus x giganteus

High levels of biomass production shown in U.S. studies are a major reason that Miscanthus x giganteus (Greef & Deuter ex Hodkinson & Renvoize; hereafter referred to as Mxg or Giant Miscanthus) is an attractive feedstock (Figure 5.3). Mxg also exhibits many other characteristics that allow it to meet or exceed the criteria for desirable biomass crops. As a perennial, it typically requires fewer yearly agronomic inputs than annual row crops. After establishment, time spent in the field is usually limited to a single annual harvest. In some years, in some locations, additional field time may be spent applying fertilizer, but applications have neither been shown to be required every year, nor in every location. For example, Christian et al. (2008) reported no yield response to nitrogen applications to a 14-year-old stand in England, while Ercoli et al. (1999) did see a nitrogen response when nitrogen was applied to Mxg in Italy. Thus, Mxg crops have not needed annual planting, pest controls, or fertilization in ongoing studies. Its perennial growth also controls soil erosion. As it becomes established and grows, Mxg develops an extensive layer of rhizomes and mass of fibrous roots that can hold soil in place. Finally, the belowground growth can contribute soil organic carbon levels as shown in Germany (Schneckenberger and Kuzyakov, 2007) and Denmark (Foereid et al., 2004).

Biology and adaptation. Giant Miscanthus is a sterile triploid hybrid resulting from the cross of the diploid M. sinensis and tetraploid M. sacchariflorus (Scally et al., 2001). Originally discovered in Japan, Mxg was thereafter introduced into the United States as a landscape plant (Scally et al., 2001).

M. sinensis and M. sacchariflorus are native to regions in eastern Asia with overlapping ranges in the same areas of Japan (Stewart et al., 2009). There are several forms of M. sinensis and the species can be found in mountainous areas, mid-level grasslands, and in low-lying waste areas (Clifton-Brown et al., 2008). It is usually a rhizomatous clump-former of variable size that spreads by seed. M. sacchariflorus is a vigorously rhizomatous species that can spread both by seed and by rhizomes and is often found on the margins of rivers or marshes (Barkworth et al., 2007). It has escaped cultivation and individual clumps can cover more than 20-32 square feet in escaped roadside settings.

Figure 5.3 Miscanthus growth in August

(Courtesy of the University of Illinois)

The Mxg clone used in University of Illinois feedstock research originated from rhizomes obtained from the Chicago Botanic Gardens in 1988 and has been part of a landscape demonstration planting at the University since that time (Pyter et al., 2009). In addition to this common landscape clone, there are now other Giant Miscanthus types being developed and marketed specifically for biomass production. For example, 'Freedom' Giant Miscanthus was developed at Mississippi State University and is being produced for commercial planting by SunBelt Biofuels of Soperton, Georgia. Cantus Bio Power Ltd. of North Leamington, Ontario and Vancouver, British Columbia, Canada lists 'Amuri' Giant Miscanthus and 'Nagara' Giant Miscanthus as very cold-tolerant, high-yielding grasses. Both Mendel Biotechnology, Inc. of Hayward, California, and Ceres, Inc. of Thousand Oaks, California, have included feedstock types of Miscanthus spp. as part of their research activities.

Also of importance is the fact that Mxg grows efficiently in a variety of settings. Established Mxg stands have survived air temperatures of -20°F in Illinois (Pyter et al., 2009). The temperature for optimum photosynthesis is 86°F (Naidu et al., 2003; Naidu and Long, 2004), but it has the ability to photosynthesize at temperatures as low as 47°F (Naidu et al, 2003; Farage et al., 2006). While both are C4 grasses, Mxg produced 61% more biomass than maize in an Illinois study (Dohleman and Long, 2009) even though maize has a higher photosynthesis rate in midsummer. This was due to the ability of Mxg to begin growing earlier in the growing season and continue later in the season.

Established plants have exhibited tolerance to summer drought. While substantial water is necessary for high yields (Beale et al., 1999), soils—given adequate moisture—have not shown to effect Mxg biomass production (Pyter et al., 2009), but have affected establishment rates. In fertile soils, establishment is usually 2 to 3 years, while it may take 3 to 5 years in less fertile sites (Pyter et al., 2009).

Production and agronomics. In Illinois, new shoots emerge from scaly underground stems (rhizomes) in April, and the grass grows to approximately 6.6 feet by the end of May (Pyter et al., 2007). Growth continues through summer into autumn with sterile flowers emerging in late September, and it goes dormant with the onset of killing frosts, usually in October after it has reached approximately 13 feet (Pyter et al., 2007). With the onset of freezing temperatures, leaves drop and minerals are returned to the belowground portions of the plant. The senesced stems are harvested from mid-December through late March; however, the standing biomass can be harvested before a killing frost if necessary.

In established Mxg plantings, there are approximately 5 to 10 shoots per square foot (Pyter et al., 2009). Harvestable stems resemble bamboo and are usually 0.5 to 0.78 inches in diameter and approximately 9.5 feet long (Pyter et al., 2009). Given that the original University of Illinois demonstration plot was planted in the late 1980s and has continued to produce large amounts of biomass for the past 20+ growing seasons, it is anticipated that commercial plantings of Mxg will provide good yields for at least 10 to 15 years

A major drawback to Mxg is increasing the planting stock. Because it is sterile, seed propagation is not an option, and Mxg is typically propagated by tissue culture, plugs, or by rhizome division. In Europe, tissue culture-produced plants were more expensive and less winter hardy during the initial growing season than rhizome-produced plants (Lewandowski, 1998). Thus, tissue culture has not been widely used to propagate large numbers of Mxg in Europe, nor in the United States.

Rhizome propagation entails digging dormant root-rhizome (underground stem) clumps, separating the rhizomes into smaller pieces, and replanting the newly divided rhizomes. Healthy, 1- or 2-year plants work well for propagation. In central Illinois, a 1-year plant usually yields 7 to 10 rhizomes and a 2-year plant normally yields 25 or more usable rhizomes (Pyter et al., 2009). Thus, 2 years after planting, an acre of well-tended Mxg can produce enough rhizomes to plant 25 or more acres. An acceptable planting rate is 4,250 rhizomes per acre (Pyter et al., 2009). A planting depth of 4.0 inches is recommended (Pyter et al., 2010) for both propagation and final planting. There are ongoing efforts to develop seed sources because of the high cost of vegetative propagation. These would be crosses of various varieties. The costs are expected to be significantly lower, as much as half and even more over time.

Lastly, there appears to be little or no insect or disease pest problems associated with Mxg. There have been no reports of pests in commercial plantings in Europe. In the United States, however, several aphids (Bradshaw et al., 2010) have been reported recently and Mxg may be a site of oviposition and emergence for the western corn rootworm, a major pest of corn in the Midwest (Spencer and Raghu, 2009). Also, a leaf spot disease (Ahonsi et al., 2010) has been reported on Mxg plantings in Kentucky. It remains to be seen if these recently identified pests develop into commercial problems.

Harvesting Mxg biomass usually begins after the grass is fully senesced and should be completed prior to the onset of spring growth (Pyter et al., 2009). There is not enough first-year growth to warrant harvesting, and second-year crops usually deliver yields of about half of fully established plantings. In quality soils, established Mxg in the third and subsequent years usually reaches plateau yields.Several commercial manufacturers are evaluating and developing equipment specifically designed to harvest and handle Mxg biomass, but at present, hay mowers, conditioners, and balers are used. In Illinois production, hay equipment from several different manufacturers has worked well, but slowly, due to stem toughness and density. There is a need for specialized harvesting equipment that can handle Mxg more efficiently than commercial hay equipment.

Potential yield and production costs. European study (Lewandowski et al., 2000) of 3-year-old and older stands from 19 variously distributed sites reported that dry-matter yields of spring-harvested Mxg ranged from 1.8 to 15.2 tons per acre. The lowest reported yield was obtained from Central Germany (50-52° N)—the site is similar latitude of Saskatoon, Saskatchewan. The highest yield was from northwest Spain (43° N) at 15.2 dry tons per acre, latitude similar to Saginaw, Michigan. This high-yielding site was fertilized with nitrogen, although there was no fertilizer effect. Another European site with a high yield was in Southern Italy (37° N – 15.2 tons per acre). This site's latitude is similar to that of Lexington, Kentucky.

Plot yields in 2004, 2005, and 2006 at three Illinois sites have varied depending on the latitude and weather during the growing season. In replicated studies of unfertilized Mxg planted in 2002 using small potted plants, the average hand-harvested yields over the 2004, 2005, and 2006 growing seasons were 9.8 tons per acre in northern Illinois (latitude 41.85N), 15.4 tons per acre in central Illinois (latitude 40.12N), and 15.5 tons per acre in southern Illinois (latitude 37.45) (Pyter et al., 2007). In the same 3-year period, yields for unfertilized upland switchgrass, 'Cave in Rock', seeded in 2002 were 2.2, 5.2, and 2.7 tons per acre at the same northern, central, and southern Illinois sites, respectively. A separate demonstration plot in Urbana, Illinois, yielded approximately 14.1 tons per acre of dry Mxg biomass in 2006 at the end of the third growing season (Pyter et al., 2007). Based on average yields of 13.2 tons per acre, it would require approximately 31.2 million acres of Mxg to produce 35 billion gallons of ethanol in the United States, compared to 83.5 million acres of switchgrass producing 4.6 tons per acre (Heaton et al., 2008).52

Figure 5.4 Harvesting miscanthus

(Courtesy of the discoversolarenergy.com)

Mxg yield data were collected after 2 years at the DOE/ Sun Grant Herbaceous Partnership sites in Kentucky. Nebraska, and New Jersey (Table 5.2). The plots at these sites receive 0, 54, and 107 pounds of nitrogen per acre. Second-year biomass yields increased at the New Jersey site with fertilization, but it did not increase yields in Kentucky or Nebraska. In fact, Nebraska yields of Mxg went down with increasing nitrogen levels. Further analysis will likely reveal that the differences were the result of native soil fertility or climate differences. These yields are impressive given that yields usually increase until the grass is fully established, which takes 3 to 5 years (Lewandowski et al, 2000).

Table 5.2 2009 (Second-Year) Miscanthus x giganteus Biomass Yields (Tons per Acre)

Developing a crop of Mxg will likely be expensive. Jain et al. (2010) estimates a cost of $1,197 to establish an acre of Mxg planting rhizomes at a rate of 4,000 rhizomes per acre in Illinois. Following establishment, Lewandowski et al. (2000) estimated the annual breakeven cost of producing Mxg in Denmark to be approximately $85 per ton (based on an exchange rate of $1.35 per Euro). In 2008, Khanna and others estimated the annual breakeven farm-gate price to produce Mxg to be between $37 and $52 per ton in the United States. Finally, Jain et al. (2010) most recently estimates the annual breakeven cost to produce Mxg to be between $46 per ton in Missouri and $139 per ton in Minnesota.

Future plantings may involve a seeded option closer to the establishment costs of grasses. Miscanthus from seed may reach maturity in the second year and have a higher yield because of higher plant density. Finally, because the variety is fertile (as opposed to the sterile rhizomes), risk of invasion would need to be managed if planted on any scale.

Sustainability. Most investigations of Mxg grown as a biomass feedstock have been positive because it is a long-lived perennial, produces high biomass yields, and has been shown to require minimal inputs in some studies. In fact, Heaton et al. (2004) summarized that Mxg stores carbon in the soil, has low fertilizer requirements, high water-use efficiency dries in the field, and has the ability to stand in the field during winter prior to harvest. While these positive attributes make Mxg an attractive feedstock, there are concerns about the invasiveness of Mxg and other Miscanthus species. A search of the literature revealed no substantiated settings where Mxg has been invasive. Other types of Miscanthus, including the parents of Mxg, have been reported to be invasive. Czarapata (2005) lists M. sinensis as a lesser invader of natural grasslands and prairies in the northeastern, southeastern, and eastern Midwest and Kaufman and Kaufman (2007) indicate that it will take over in roadsides and burned pastures in areas of the United States where soils are naturally moist. The Minnesota Department of Natural Resources (2009) has identified M. sacchariflorus to be invasive in moist areas of the upper Northeast and Midwest, where it can be found in roadsides and openings or edges of wooded areas.

Conclusions. Miscanthus x giganteus is receiving much attention as a potential biomass crop. In Europe, work using the grass as a feedstock began in the 1980s. While a great deal of research has been conducted recently, there are still barriers to commercial production. For example, developing low-cost, reliable, commercial-scale propagation methods are critical to developing the crop. Identifying the genotypes best suited to a given region is also critical. Gaining a better understanding of the relationships between the grass and mineral fertilization and the grass and pests and pest controls is also necessary. Finally, developing efficient harvesting methods and equipment will be necessary to remove the crop from the fields and into storage in a low-cost, timely fashion (Figure 5.4).

Mechanical equipment is becoming available that will plant more than 25 acres per day. Improvements in Mxg genetics, agronomy, and harvesting are also coming quickly. Long-term production from a single planting, modest-input requirements, and carbon capture, coupled with realistic commercial biomass yields of 9 to 16 tons per acre per year, makes Miscanthus x giganteus a candidate feedstock for addressing the U.S. renewable bioenergy demand.

5.1.3 Sugarcane

Sugarcane is a large-stature, jointed grass that is cultivated as a perennial row crop, primarily for its ability to store sucrose in the stem, in approximately 80 countries in tropical, semi-tropical, and sub- tropical regions of the world (Tew, 2003). It is one of the most efficient C4 grasses in the world, with an estimated energy in: energy out (I/O) ratio of 1:8 when grown for 12 months under tropical conditions and processed for ethanol instead of sugar (Bourne, 2007; Macedo et al., 2004; Muchow et al., 1996). Under more temperate environments, where temperature and sunlight are limited, I/O ratios of 1:3 are easily obtainable with current sugarcane cultivars if ethanol production from both sugar and cellulosic biomass is the goal (Tew and Cobill, 2008). In addition, sugarcane ethanol cuts GHGs at least 60% compared to gasoline—better than any other biofuel produced today. EPA has confirmed sugarcane ethanol's superior environmental performance by designating it as an advanced renewable fuel. Most of the sugarcane grown in the United States is dedicated to the production of sugar. As an energy cane industry has not developed to date, one can assume that the production practices of the mature sugarcane industry can be modified to ensure the sustainable production of energy cane as well.

Biology and adaptation. Sugarcane (Saccharum spp.) is a genetically complex crop with a genomic makeup that results from successful interspecific hybridization efforts, primarily involving S. officinarum and S. spontaneum (Tew and Cobill, 2008). Improvement of sugarcane for increased energy efficiency and adaptability to a wide range of environments is considered by many geneticists as synonymous with "genetic base broadening" (i.e., utilization of wild Saccharum germplasm), particularly S. spontaneum in sugarcane breeding programs (Ming et al., 2006). S. spontaneum, considered a noxious weed in the United States, can be found in the continents of Africa, Asia, and Australia in environments ranging from the equator to the foothills of the Himalayas. This makes it an excellent source of a number of valuable genes (Mukherjee, 1950; Panje and Babu, 1960; Panje, 1972; and Roach, 1978). The USDA-ARS's Sugarcane Research Unit (SRU) at Houma, Louisiana, in the 1960s took on the role of introgressing desirable genes from sugarcane's wild and near relatives (Miscanthus and Erianthus) to build new parents for utilization in its commercial sugarcane varietal development program in what is referred to as the SRU's basic component of its breeding program. Because these wild accessions contain only small amounts of sugar, three or four rounds of backcrossing to elite sugarcane varieties must be done to obtain a commercially acceptable sugarcane variety, which has high-sugar yields with minimum amounts of fiber.

Production and agronomics. Sugarcane is grown as a monoculture with fields being replanted every 4 or 5 years. This type of culture is conducive to the development of perennial weeds like johnsongrass and bermudagrass, since the row top (i.e., raised bed) remains relatively undisturbed for the 5-year crop cycle. The selective control of these weeds within the crop is difficult with currently registered herbicides once these weeds produce rhizomes. To minimize this risk, growers disk the old stubble fields in the winter or early spring and fallow the fields until they are planted to sugarcane again. Frequent disking and/or the use of multiple applications of glyphosate are used to deplete the soil of weed seed and rhizomes during the fallow period.

Sugarcane is vegetatively planted by laying 6- to 8-foot long stalks end-to-end in a planting furrow along rows spaced 5 to 6 feet apart and covering the stalks with 2 to 4 inches of soil. The wide row spacing is needed to accommodate mechanical harvesting. One acre of seedcane can plant 6 to 10 acres of sugarcane, depending on the length and number of stalks at the time of harvesting the "seedcane" for planting, and the number of stalks per foot of row being planted. When harvesting seedcane for planting, stalks are cut at the soil surface and at the last mature node at the top of the stem. An alternate method of planting is to plant 12- to 18-inch stalk pieces (billets) that can be harvested with the same chopper harvester used to harvest sugarcane for delivery to sugar mills. Once the stalks are planted, a broad spectrum preemergence herbicide is applied to control seedling weeds. New plants emerge from the axillary buds located in the nodal regions along the stalk. Growers produce most of their own seed cane for planting; hence, planting is generally done a few weeks prior to the beginning of the harvest season to ensure that stalks are plentiful and tall (Figure 5.5).

Figure 5.5 Energy cane research plots

(Courtesy of Ed Richard, USDA-ARS)

Vegetative planting of sugarcane is often considered a drawback by growers who are accustomed to planting large areas of seeded crops relatively quickly with one tractor and one planter. It is an expensive process as it requires considerable labor and equipment, is relatively slow, and requires that the grower plant sugarcane that would normally be sent to the raw sugar factory for processing. However, with energy cane, 20% to 30% higher planting ratios can be expected, because stalk numbers and heights are higher. In addition, at least two additional harvests per planting can be expected, significantly lowering the number of acres requiring planting each year. It is estimated that the cost to plant one acre of sugarcane is about $500 when the grower uses seedcane from the farm. Because planting costs are spread over four annual harvests, the annual cost would be approximately $125 per acre. With energy cane, one acre of seedcane would plant about 13 additional acres reducing per acre planting costs to $346. If spread over the anticipated six annual fall harvests, annual planting costs would be $58 per acre.

Vegetative planting also has advantages, especially when planting must be done under conditions of less than ideal seedbed preparation. The crop emerges 14 to 21 days later and continues to grow until the first heavy frost of the fall. The first production year (plant-cane crop) actually begins in the spring following planting with the emergence of the crop from winter dormancy. Herbicides are applied each spring to the subsequent ratoon crops (first- through third-ratoon crop) to minimize early weed competition. Nitrogen is applied at rates of 70 to 90 pounds per acre to the plant-cane crop (first growing season) and 90 to 120 pounds per acre to the subsequent ratoon crops. These applications are generally made in the spring about two months into the growing season.

The crop is susceptible to the rapid spread of a number of bacterial, fungal, and viral pathogens that can be spread easily by machinery and wind currents. These pathogens can affect the yield and ratooning ability (number of yearly harvests per planting) of the crop. Race changes of some of these pathogens are common and the industry is always susceptible to new diseases. Insects, primarily stalk borers, grubs, and aphids also plague the industry. The compactness of the industry and the fact that the crop is grown continuously as a monoculture makes sugarcane especially vulnerable to the rapid spread of diseases and insects. The planting of resistant varieties is the predominant means of managing diseases in sugarcane. For insects, mainly stalk borer, an effective integrated pest management program that involves field scouting, the use of tolerant varieties, and insecticides when established infestation thresholds are exceeded, is used. Additional research is exploring the use of multiple crop-production systems for year-round delivery of feedstocks and to respond to biotic stresses (McCutchen and Avant, 2008).

With each successive fall harvest during the crop cycle, yields tend to decline to the point that it is not practical to keep the old stubble for another year. The milling season in the more sub-tropical climates, like Louisiana, lasts about 100 days beginning in late October and ending in late December or early January. Given that the crop emergence in the spring depends on the date of the last killing frost, it is obvious that sugarcane harvested in December will produce higher yields than sugarcane harvested in September. For this reason growers try to have a balance in ratoon crop ages from 0 (plant cane that was never harvested) to 3 (third ratoon that was harvested three times previously) and begin the harvest with the third-ratoon fields that would lack the vigor of the plant-cane crop. In addition, early harvested crops tend not to yield as much the following year. A similar scenario is anticipated for energy cane with the exception that fourth- and fifth- ratoon crops would also be harvested.

In Louisiana, some energy cane varieties have produced average yields of 56 green tons per acre, with 8 to 14 tons per acre being fiber and 4 to 6 tons per acre being brix (soluble sugars) on a dry weight basis (see Figure 5.6).

Potential yields and production costs. The theoretical maximum for aboveground sugarcane biomass (total solids) yield is estimated to be 62 green tons per acre annually (Loomis and Williams, 1963). This is dependent on temperature and sunlight, and would probably occur under tropical conditions. Sugarcane breeding programs have reported sugar yield gains in the order of 1% to 2% per year (Edme et al., 2005). The economic sustainability of growing energy cane in non-traditional cane growing regions will require yearly biomass yield gains of this magnitude or greater, with a goal of ensuring that the I/O ratio of 1:8 projected for tropical countries can be met and ultimately exceeded under the sub-tropical cane growing conditions of the southeastern United States.

The "sun-dried crop" concept of allowing the crop to desiccate in the field and perhaps devoid itself of some of its leaves and moisture, as is proposed for many of the perennial grasses being considered for biofuels, is not an option for energy cane as the stalks are thick with a waxy coating on them and the new growing season should begin as soon after harvest as possible. Consequently, energy cane, like sugarcane, will have to be harvested green and dewatered if the fiber is to be stored and processed later in the year. The value of this liquid is in question because it will add to transportation costs. However, if water is needed for the digestion of the fiber or the maintenance of the bagasse under anaerobic conditions to minimize deterioration during outside storage, it would be present at no additional charge. What is also overlooked is the fact that the water contains sugar that is easily and much more cheaply converted to ethanol. Furthermore, in some conversion processes the yeast used in fermentation needs a substrate to grow and multiply on, and sucrose is an ideal substrate. Conceivably the biorefinery would have two processes for the production of biofuel, with one having sugar (brix) obtained from de-watering at the biorefinery as the feedstock and the other the fiber (bagasse). Economics would have to be considered with these options especially because of the very short cut-crush interval needed for sugar recovery and to prevent spoilage.

Figure 5.6 Average yields from four successive fall harvests of several candidate sugarcane varieties as compared to the standard variety, L 79-1002

(Courtesy of Ed Richard, USDA-ARS)

Figure 5.7 Harvesting sugarcane

(Courtesy of Ed Richard, USDA-ARS)

Sustainability. Green cane harvesting of sugarcane deposits 2.7 to 3.6 tons per acre of a mixture of brown and green leafy material and fragments of the stalks (Richard, 1999; Viator et al., 2006; 2009a;b). The fibrous extraneous matter generated during harvest has an energy value; however, the greatest value may be as mulch to: limit soil erosion, depress weed development, conserve moisture, and as a means to recycle nutrients. All are potential contributors to the sustainable production of energy cane (de Resende et al., 2006). It is estimated that the extraneous residue generated during the green cane harvesting of sugarcane contains 0.7% of N, 0.07% of phosporous, and 0.7% of potassium by weight. Using 2009 USDA economic data (USDA-ERS 2010b), 2.7 tons of residues would equate to a savings of approximately $50 per acre.

The critical amount of post-harvest residue that can be removed from the field has not been determined for energy cane. Sugarcane is harvested mechanically with a chopper harvester (Figure 5.7). These harvesters chop the sugarcane stalks into small 6- to 8-inch long pieces (billets) and use wind currents from an extractor fan to remove the leaves attached to the stalks. If used to harvest energy cane, the speed of the extractor fans could be adjusted to deposit a percentage of the extraneous matter on the soil surface while the remainder is collected with the stalks and used as an additional source of fiber.

Production cost estimates are complicated by the first-year production of seedcane crop and the number of ratoon crops before re-establishment. For example, different costs are associated with fallow field and seedbed preparation, seedcane planting and harvest, plant cane, and ratoon operations and harvest over multiple years (Salassi and Deliberto, 2011). Using the most common sugarcane assumptions from the production costs from Salassi and Deliberto, an extrapolated cost to produce and harvest the energy cane is estimated to be about $34 per dry ton. The assumed yield is about 14 dry tons per acre.

Energy cane will be grown as a perennial. Commonly discussed disadvantages of perennial feedstocks include the difficulty of establishing a perennial crop from seed and rhizomes, the control of weeds during the establishment period, and the fact that an economic return will not be realized the first year. Energy cane, like sugarcane, is vegetatively planted by placing the stalks in a planting furrow in mid- to late summer. Herbicides are labeled for at-planting preemergence applications in sugarcane; presumably, these herbicides can also be applied to energy cane with similar results. These herbicides are reapplied each spring during the course of a 3- to 4-year sugarcane production cycle. Because of the vigor (e.g., early spring emergence and high stalk population) of energy cane, use of these herbicides beyond the first spring after planting is not anticipated. The vigor of energy cane is also advantageous due to the fact that when the crop is planted in the summer, it emerges and produces a uniform stand quickly. The crop continues to grow until the aboveground portion is winter-killed in the fall. This aboveground material, which can equate to 2 to 4 dry tons per acre, could be harvested and converted to fuel in the first year of establishment.

The final consideration in the sustainability of energy cane production is the utilization of biorefinery byproducts to supply nutrients and reduce the impact of crop removal on soil health. These byproducts can include vinasse from the fermentation process, biochar from pyrolysis, and filter press mud and boiler fly ash from the squeezing of stalks. A positive synergistic response was observed when the application of fertilizer was combined with an application of filter press mud in Florida (Gilbert et al., 2008). In Brazil, vinasse from the ethanol distillery is typically returned to the recently harvested fields to supplement fertilizer requirements.

Conclusions. Looking to the future, the greatest needs to make energy cane a suitable feedstock for the cellulosic industry and extend its range of geographic distribution outside of the traditional sugarcane growing areas are cold tolerance for expansion outside of tropical areas; drought and flood (saturated soil) tolerance, as this crop will probably be grown on marginal soils that may be prone to flooding or where irrigation is difficult; insect and disease resistance; and a further exploitation of some varieties of sugarcane that encourages symbiotic relationships with nitrogen fixing bacteria.

The success of the sugarcane industry has been, and continues to be, dependent on the development of new hybrids with superior yields and increased resistance to many of the abiotic and biotic stresses previously mentioned. This formula will not change if the crop is grown as a dedicated feedstock for the production of liquid biofuels or electricity. Successful hybridization begins with the introgression of desirable traits from the wild relative of sugarcane, Saccharum spontaneum. Early generation progeny from these crosses with elite sugarcane clones exhibit high levels of hybrid vigor, which translates into increased cold tolerance, greater ratooning ability, enhanced levels of tolerance to moisture extremes, increased insect and disease tolerance, and more efficient nutrient utilization (Legendre and Burner, 1995). Much of the vigor of these early generation hybrids is lost in a conventional breeding program for sugar, as progeny from these crosses must be backcrossed with elite high-sugar-producing clones three to four times before a commercial sugarcane variety can be produced. These early generation hybrids would be considered ideal candidates as dedicated cellulosic biomass crops like energy canes. Most of these varieties can average over 15 dry tons per acre annually over at least four annual fall harvests.

By enhancing the level of stress tolerance through the conventional breeding techniques being employed by the basic breeding program of the USDA-ARS SRU, the geographic area of distribution could be expanded to more temperate regions of the United States to include those states in Hardiness Zone 8 (regions within Texas, Louisiana, Arkansas, Mississippi, Alabama, Georgia, South Carolina, North Carolina) and perhaps the extreme southern end of Hardiness Zone 7 (regions within Oklahoma, Tennessee, Virginia), where the annual low winter temperatures can approach 1° F. With this in mind, it is conceivable that the area devoted to this crop could be tripled, thus making it a more attractive feedstock for biotech companies with proprietary genes to further enhance the level of stress tolerance, or introducing genes for the production of saleable byproducts without the labeling restrictions encountered in food crops.

5.1.4 Sorghum

Of the crops recently identified for their potential bioenergy production, sorghum (Sorghum bicolor) has historically had the most direct influence on human development (Figure 5.8). Sorghum was domesticated in arid areas of northeastern Africa over 6,000 years ago (Kimber, 2000). Sorghum is traditionally known for grain production; it is the fifth most widely grown and produced cereal crop in the world (FAO, 2007). However, in many regions of the world, sorghum is just as (if not more) important as a forage crop. In addition to forage and grain, sorghum types high in stalk sugar content and extremely lignified types (for structural building) have been grown throughout the world.

Given the demands for renewable fuel feedstocks, sorghum is now being developed as a dedicated bioenergy crop. This designation is not new; sorghum was mentioned prominently as a bioenergy crop over 20 years ago (Burton, 1986). The interest in the crop is justifiable based on several independent factors that separately indicate good potential, but when combined, they clearly designate sorghum as a logical choice for bioenergy production. These factors include yield potential and composition, water-use efficiency and drought tolerance, established production systems, and the potential for genetic improvement by using both traditional and genomic approaches.

Biology and adaptation. Whether measured in grain yield or total biomass yield, sorghum is a highly productive C4 photosynthetic species that is well adapted to warm and dry growing regions. While sorghum is technically a perennial in tropical environments, it is planted from seed, then grown and managed as an annual crop. Sorghum has a long-established breeding history through which the productivity, adaptation, and utilization of the crop has continually been improved. These efforts have resulted in numerous cultivars and hybrids of sorghum that are used for various purposes.

The optimum type of sorghum to be grown for biofuels production is dependent on the type of conversion process that will be used. Sorghums are divided into distinct types based on the amount of different carbohydrates they produce.

Figure 5.8 Sorghum hybrid tests

(Courtesy of W. Rooney, Texas A&M University)

Grain sorghum hybrids produce large quantities of grain (approximately 50% of total biomass); the grain is composed primarily of starch (approximately 75%) and may be used as a food grain, feed grain, or ethanol substrate via starch hydrolysis and fermentation. If mechanically harvested, these hybrids are usually less than 2 meters tall. Residue is typically returned to the soil, but it is used as forage under drought conditions, and it could be used as a biomass feedstock as well.

Forage sorghums are usually one of two types. Sorghum-sudangrass hybrids are tall, leafy, thin-stalked hybrids used for grazing or hay production. Silage-sorghum hybrids are typically taller and thicker-stalked with high grain yield (25% of total biomass), and they are chopped and ensiled for animal feeding (Figure 5.8). Both of these types of sorghum have been highly selected for optimum forage production, palatability, and conversion in an animal system. There have been significant breeding efforts to enhance the forage quality of this material by incorporating the brown midrib trait into many forage hybrids. These brown-midrib hybrids have lower lignin and are more palatable, which increases conversion efficiency and consumption rate in ruminant feeding programs (Aydin et al., 1999; Oliver et al., 2004). These types may have application in certain energy sorghum applications where lower lignin content is desirable.

Sweet sorghum is a unique type of sorghum that accumulates high concentrations of soluble sugars. Traditionally, these sorghums were grown for the stalk, which was milled to extract the juice. The juice was then cooked down, and the resulting syrup was used as sweetener. While these types of sorghum continue to be grown for syrup on an artisan level, there has been significant interest in the development of sweet sorghum as a dedicated bioenergy crop using a sugarcane system model. In the mid-1970s, significant research was conducted to explore the development of sweet sorghum as a bioenergy source for biofuels and energy production, and breeding programs were initiated to develop high-yielding sorghum specifically for ethanol production (McBee et al., 1987).

Dedicated biomass sorghums are the most recent class of sorghum that has been developed in response to the interest in bioenergy crops. These sorghums are highly photoperiod sensitive, meaning that they do not initiate reproductive growth until well into the fall season of the year. Consequently, in temperate environments like most of the United States, these sorghums will not mature. This absence of reproductive growth reduces sensitivity to periods of drought and allows the crop to effectively photosynthesize throughout the entire growing season. This results in higher yields of primarily lignocellulosic biomass that is completed in a single annual season. While phenotypically similar to forage sorghums, these biomass sorghums are distinctly different in that they are not selected for animal palatability, which results in plants with larger culms and flexible harvest schedules, which minimizes nitrogen extraction at the end of the season.

Production and agronomics. Biomass yield potential of sorghum is strongly influenced by both genetic and environmental factors. For example, grain sorghum is commonly grown in more arid regions of the country, and the plant itself is genetically designed to be shorter to facilitate mechanical harvesting. Alternatively, specific dedicated biomass sorghums are very efficient at producing large amounts of lignocellulosic biomass. Finally, both sweet sorghum and forage sorghum are prolific when the environmental conditions allow the plants to reach full genetic potential. Hallam et al. (2001) compared perennial grasses with annual row crops and found that sweet sorghum had the highest yield potential, averaging over 17 tons per acre (dry weight basis) and also performing well when intercropping with alfalfa. Rooney et al. (2007) reported biomass yield of energy sorghum in excess of 44.6 tons per acre (fresh weight) and 13.4 tons per acre (dry weight). They reported that potential improvements could extend the potential of these types of hybrids to a wide range of environments. Under irrigation in the Texas panhandle, McCollum et al. (2005) reported yield of commercial photoperiod sensitive sorghum hybrids as high as 36 tons per acre (65% moisture) from a single harvest. In subtropical and tropical conditions, single cut yields are generally lower, which is likely due to increased night temperatures, but cumulative yields are higher due to the ratoon potential of the crop. Total biomass yields as high as 13.4 tons per acre (dry weight basis) were reported near College Station, Texas (Blumenthal et al., 2007).

Composition of sorghum is highly dependent on the type that is produced, such as grain sorghum, sweet sorghum, forage, and cellulosic (high biomass) sorghum. Sorghum grain is high in starch, with lower levels of protein, fat, and ash (Rooney, 2004). Significant variation in the composition of grain is controlled by both genetic and environmental components, making consistency in composition a function of the environment at the time of production; consequently, these factors influence ethanol yield (Wu et al., 2007). Juice extracted from sweet sorghum is predominantly sucrose with variable levels of glucose and fructose, and in some genotypes, small amounts of starch are detectable (Clark, 1981; Billa et al., 1997). In forage and dedicated biomass sorghums, the predominant compounds that are produced are structural carbohydrates (lignin, cellulose, and hemi-cellulose) (McBee et al., 1987; Monk et al., 1984). Amaducci et al. (2004) reported that the environment influences sucrose, cellulose, and hemicellulose concentrations, while lignin content remains relatively constant.

Potential yield and production costs. Sorghum has a long history as a grain and forage crop, and production costs range from $200 to $320 per acre (USDA-ERS, 2010a). This history provides an excellent basis for estimating crop production costs for energy sorghums with a few modifications. Seed costs, planting costs, and production costs will be similar to grain and/or forage sorghum. Fertilizer rates will likely be less than forage sorghum on a production dry-ton basis (due to the reduced nitrogen content in the mature culm), but it is expected that yields will be higher, so total nitrogen requirements will be equalized. Production is expected under rain-fed conditions; therefore, no additional costs are added for irrigation. On a dry ton basis, given an average production of 10 dry tons per acre and assuming $400 per acre cost of production for dedicated biomass sorghums, biomass sorghum will cost $40 per dry ton at the farmgate.

Production practices for dedicated biomass and sweet sorghum are similar to traditional sorghum crops with some minor modifications. For both types of energy sorghums, it is expected that plant populations will be lowered relative to grain and certainly compared to forage sorghum. This drop will allow the plants to produce larger culms and reduce the potential for lodging and interplant competition. Pests and diseases of sorghum are well known and described, and there are some that will require management plans and effective deployment of host plant resistance for control. Of particular note is the disease anthracnose (caused by Colletotrichum graminicola), which is prevalent in the southeastern United States and is capable of killing susceptible sorghum genotypes. Fortunately, there are many sources of genetic resistance to the disease, and effective control relies on effective integration of these anthracnose resistance genes.

Harvesting and preprocessing of energy sorghums is an area of significant research and will likely require the greatest amount of modification compared to grain sorghum. Sweet sorghum will be harvested and moisture extracted for soluble sugars at a centralized location. For dedicated biomass sorghums, the forage harvest systems work very well, but there is a need to reduce moisture content to minimize transportation and storage costs.

The range of sorghum production varies with the type being produced. Both sweet sorghum and dedicated biomass sorghums grow well throughout the eastern and central United States as far north as 40° latitude, and the range of dedicated biomass sorghums is considered to be composed of most of the eastern and central United States. In the western United States, productivity will be directly related to available moisture from rainfall or irrigation. It is unlikely that the crop (or any crop) will be economically viable as a biomass crop in regions with less than 20 inches of available moisture annually. Dedicated biomass sorghums have shown yields of 7-13 dry tons per acre in the northern areas of the United States, with even higher yield potential possible in a southern environment due to the longer growing seasons. Therefore, dedicated biomass sorghum should find wide adaptation throughout most of the country that is suitable for herbaceous biomass production from an annual crop.

While sweet sorghum is productive at northern latitudes, the logistics of processing make the production of the crop unlikely in more temperate latitudes. Because soluble sugars are not stable for long periods, a processor requires a long harvest and processing window for effective use of capital equipment. The farther north the production, the shorter the growing season; hence, the harvest season is further reduced and the ability to consistently grow the high-yield potential sweet sorghum varieties is limited (Wortmann et al., 2010). Consequently, the areas of the United States that process sugarcane are also ideal locations for the production of sweet sorghum. Production in other regions will be dependent on detailed economic analysis of the cost of processing versus the length of the processing season.

Sustainability. Sorghum is unique among the dedicated bioenergy crops because it is an annual crop. The general opinion is that bioenergy crops should be perennial for sustainability purposes. While most of the bioenergy crops are perennial, there are several reasons why annual bioenergy crops are necessary. First, annual crops deliver large yields in the first year, as compared to most perennial crops, which typically increase annual yield in subsequent years following establishment. Given the challenges of propagating and establishing perennial crops, annual crops can provide insurance and production stability to industrial processors in the early phases of bringing a new processing facility online if perennial crop stand failures or establishment problems are encountered. Second, for the most part, the U.S. farming system has been based on annual crop production systems. Farmers, bankers, and processors are much more familiar and accepting of these systems, and while this will eventually be overcome, annual energy crops will be needed for that transition. Finally, annual crops are much more tractable to genetic improvements through breeding due to the simple fact that breeding is accelerated by multiple generations per year.

There are several traits of specific importance to sorghum improvement, as it relates to bioenergy production. These include, but are not limited to, maturity and height, drought tolerance, pest tolerance and/or resistance, and composition and/or quality. Improvements in these areas will increase yield potential, protect existing yield potential, and enhance conversion efficiency during processing.

While the reason for producing bioenergy feedstock is to produce renewable fuel, one of the critical components in their production will be water. Thus, both drought tolerance and water-use efficiency are critical, as many of these feedstocks will be produced in marginal environments where rainfall is limited and irrigation is either too expensive or would deplete water reserves. Sorghum is more drought tolerant than many other biomass crops. Depending on the type of biomass production in sorghum, both pre- and post-flowering drought tolerance mechanisms will be important. In sweet sorghum, both traits are important, but there has been little research regarding the impact of drought stress on sweet sorghum productivity.

For high-biomass, photoperiod-sensitive sorghums, preflowering drought tolerance is critical because, in most environments, this germplasm does not transition to the reproductive phase of growth. Each type of tolerance is associated with several phenotypic and physiological traits; these relationships have been used to fine map QTL (quantitative trait loci) associated with both pre- and post-flowering drought tolerance. Traits that have been associated with drought resistance include heat tolerance, osmotic adjustment (Basnayake et al., 1995), transpiration efficiency (Muchow et al., 1996), rooting depth and patterns (Jordan and Miller, 1980), epicuticular wax (Maiti et al., 1984), and stay green (Rosenow et al., 1983). Combining phenotypic and marker-assisted breeding approaches should enhance drought tolerance breeding in energy sorghums.

Unlike perennial bioenergy crops, sorghum will require crop rotation to maintain high yields and soil conditioning. Continuous cropping studies of sorghum have confirmed that yields will drop in subsequent years unless additional nitrogen is provided to maintain yields (Peterson and Varvel, 1989). Therefore, it is critical to consider rotations when accounting for potential land area needs in energy sorghum production. The exact rotation sequence and timeframe will vary with locale, but sorghum production once every 2 or 3 years will be acceptable in most regions. Failure to rotate may result in reduced yields and quality, as well as increased weed, insect, and disease problems. Research is needed to determine the appropriate rotation for energy sorghum in the target regions for production.

The molecular genetic resources available in the sorghum species are the most advanced among all of the potential energy crops. Combining these molecular genetic resources with traditional breeding approaches, it should be possible to rapidly develop and deploy improved, dedicated energy sorghum that meets the needs of both crop and biofuel producers.

Conclusions. Sorghum benefits from a long-established production history, existing research infrastructure, and a relatively simple genetic system. All of these factors allow for the rapid modification of the crop and delivery of specific sorghum types that are developed specifically for bioenergy and adapted to the target areas of production. The future of energy sorghum is based on development of energy-specific genotypes in which composition and productivity are optimized, while minimizing inputs like insecticide and fertilizer. In this scenario, resistance/tolerance to both biotic and abiotic stresses is critical. Fortunately, adequate genetic resources and technology are available to make these modifications in an efficient and timely manner. Composition and yield are obviously important, and continual enhancement of these factors will rely on the full use of genomic technology.

5.1.5 Poplar

The following section provides a brief description of the genus Populus, with attention to the biology, potential yield, production costs, and sustainability issues related to deploying an efficient, biomass-producing woody crop. Much more extensive information is available in other resources [e.g., Stettler et al. (1996) and Dickmann et al. (2001)].

Biology and adaptation. Today, the genus Populus includes almost 30 species and represents several taxonomic sections distributed throughout the northern hemisphere. Poplar species are an ancient and well-established component of the native North American landscape. Hybrids within and among species belonging to two sections, Aigeiros and Tacamahaca (cottonwoods), are commonly referred to as "hybrid poplars" (Figure 5.9).

Commercial deployment of hybrid cottonwood plantations for the production of fiber for paper and other products and biofuels and for the purpose of environmental remediation (e.g., phytoremediation) is a reality in many forested and agricultural landscapes, including those that lie within the temperate regions of the United States. Genotypes of eastern cottonwood (P. deltoides Bartr. ex Marsh) and hybrids between eastern cottonwood and Asian black poplar, European black poplar, and western black cottonwood (P. suaveolens Fish, subsp. maximowiczii A. Henry, P. nigra L., and P. trichocarpa Torr. & Gray, respectively) capable of producing in excess of 7 tons per acre per year by age 6-years have been identified by field tests, even in the harsh climate of the North Central region of the United States (Riemenschneider et al., 2001a; Zalesny et al, 2009).

The susceptibility of the cottonwoods to vegetative propagation was, and continues to be, in large part, a factor to their commercial value and domestication. One of the most economical means of plantation establishment is to plant dormant hardwood cuttings capable of developing adventitious roots (Heilman et al., 1994; Zalesny et al., 2005). As needed, rooted cuttings can be used to enhance survival. Thus , the vegetative propagation of poplars can confer significant genetic advantage during all stages of the breeding and selection strategy to an aggregate phenotype with high commercial utility and stability (Eriksson, 1991; Orlovic et al., 1998; Zalesny et al., 2005).

There are many possible breeding strategies that can be applied to the development of a hybrid poplar woody biomass crop (Riemenschneider et al., 2001b). Yet, all breeding strategies derive from the need for a commercial variety to possess several attributes simultaneously such as an adventitious root system, rapid growth, and resistance to pests. Eastern cottonwood, when planted in the southern United States, is an example of such a species.

Figure 5.9 Hybrid poplar plantation in Pacific Northwest

(Courtesy of ORNL)

Elsewhere, interspecific hybridization may be necessary. For example, in the upper Midwest, eastern cottonwood cuttings root erratically in the field, and hybridization between that species and another more easily rooted species is necessary to achieve an economical silvicultural system (Zalesny and Zalesny, 2009). This need for an aggregate genotype possessing all required commercial attributes gives rise to the several breeding programs found throughout North America and elsewhere in the world. Commercial genotypes in use today have most, if not all, of the important traits affecting production. However, the number of commercial genotypes in use today is relatively low, and diversification, as well as yield improvement, is a goal of breeding programs.

Production and agronomics. Plant propagation for commercial plantation establishment is generally via cuttings, which are produced in densely planted "stool beds." In the South, eastern cottonwood roots readily under field conditions, which makes for economical commercial deployment. In the North, eastern cottonwood roots erratically under field conditions, and it is more common to utilize a hybrid between eastern cottonwood and European black cottonwood (Populus nigra) or one of the Tacamahaca poplars. Of these, Populus nigra is preferred as a hybrid parental species because of the reduced probability of stem canker disease. Cuttings are harvested from stool beds in the winter during the dormant period, stored under refrigeration, and then planted in the field when soil temperatures reach levels appropriate to specific regions and genotypes.

Poplar can be managed in a number of ways, depending on the desired end product and target rotation age. Plantations grown for the production of larger-diameter trees used in the manufacture of paper and lumber are typically planted at spacings ranging from 8 feet by 8 feet (680 trees per acre) to 12 feet by 12 feet (302 trees per acre). Plantations of this type are currently managed commercially in Minnesota for pulpwood production and Oregon and Washington for a mix of products, including sawtimber and pulpwood. Poplar has the ability to resprout from established stumps after harvest, and thus could be managed on repeated coppice rotations. In light of the development of new genotypes and increased interest in dedicated energy feedstock production systems, the repeated coppice management option is a subject of renewed interest, and field research is recommended to identify optimal plant spacing and biomass production of such systems.

After planting, it is necessary to eliminate weed competition. As poplar plantings are mostly established on marginal agricultural land that has been under prior cultivation, weeds are mostly herbaceous and can be managed by preemergence herbicides, contact herbicides, or by cultivation. Weed control is needed until tree canopy closure—usually by the end of the second or third year of tree growth.

It is important to protect poplar plantings from insects and diseases. Various chemicals are available to control common pests, such as the cottonwood leaf beetle (Mattson et al., 2001; Coyle et al., 2008 ) and other insects, and the possibility of genetic selection for resistance, landscape-level deployment strategies, and other integrated pest management strategies can be considered (Mattson et al., 2001). Disease incidence and severity often depend on region (Newcombe et al., 2001). For example, Septoria stem canker is a serious problem in the Midwest on some hybrids, while much less of a concern in the Northwest, which places serious constraints on parental poplar species selection in the Midwest. Genetic selection among parental poplar species, among specific parental genotypes, and within hybrid poplar breeding populations is practiced in nearly all breeding programs (Newcombe et al., 2001).

Figure 5.10 Harvesting poplar plantation

(Courtesy of B. McMahon, University of Minnesota)

Harvesting of poplar plantations can be accomplished by using the same timber harvesting equipment found in standard forest pulpwood systems or by using purpose-designed equipment that combines felling and chipping or bundling in a single machine (Figure 5.10). Selection of equipment and method of harvest depends on average tree size and age at harvest, which are, in turn, determined by plantation density. A wide array of possibilities can be envisioned.

Potential yield and production costs. Yields from commercial plantations are proprietary and not readily available; therefore, most yield data is from research plots (Figure 5.11). A series of plot (10 x 10 tree square plots) yield trials conducted in Wisconsin, Minnesota, North Dakota, and South Dakota from 1987 demonstrated yields as high as 5.0 tons per acre per year by age 7 years (Netzer et al., 2002). Yields of newly selected genotypes in smaller plot experiments have exceeded 7.0 dry tons per acre annually on good agricultural soil in southern Wisconsin and Iowa (Riemenschneider, 1996; Zalesny et al., 2009). In general, sustainable average yields of 4.5, 6, and 9 tons per acre annually (dry weight, stem, and branches) are expected in the midwestern, southern, and northwestern United States, respectively. With appropriate research and development investment, over time these yields could be significantly increased, even doubled (Volk et al. 2010) (Figure 5.11).

Using cash flow models of production costs and expected yields, costs of poplar biomass are comparable to other dedicated biomass production systems and range from $25 to $60 per dry ton depending on site quality and site-specific inputs. Using cash flow models developed by the University of Minnesota for the north-central United States, the total discounted cost of all inputs (assuming a 12-year rotation pulpwood-oriented system) is $450 per acre or roughly $36 per acre annually. Breakeven price of biomass in this system is approximately $16 per dry ton, including input costs only. The question of where woody energy crops will be deployed depends less on the breakeven price of the energy crop itself and more on the profitability of the crop being replaced. Based on data from a survey of production costs conducted by the University of Minnesota (2010), per-acre profits are estimated to range from $50 per acre in the case of wheat to $200 per acre in the case of corn production. Thus, energy crops will have to be priced at a level in which profits to growers are at least equal to competing crops.

Figure 5.11 Minnesota poplar plantation

(Courtesy of ORNL)

Sustainability. Perennial woody crops provide multiple benefits when managed sustainably, such as biological diversity, conservation of soil and water, maintenance of site productivity, carbon sequestration, and socioeconomic values (Ruark et al., 2006). In a summary paper on the subject published by Tolbert et al. (2000), several trends are identified. Soil structure, total organic content, and infiltration rate is shown to the agricultural system being replaced. Inputs of leaf litter and lack of annual site disturbance are thought to be contributing factors. Nutrient content and water yield of short rotation poplar plantations were found to be similar to older, natural aspen stands in Minnesota. Increased soil carbon has been documented under short rotation systems, particularly in those regions of the country where inherent soil organic content is low, like it is in the South. Over the long term, soil carbon is expected to increase under perennial woody crops due to inputs of leaf and root biomass and lack of disturbance of the soil surface. Oxidation of carbon from upper soil layers has been shown to be a major factor, accounting for differences between perennial energy crops and annually tilled agricultural crops. Studies of wildlife effects of hybrid poplar plantings in Minnesota have shown increased diversity in bird populations compared with row crops (Hanowski et al., 1997). Small mammal abundance was found to be a function of canopy closure, with younger plantations being more similar to grasslands. Research done to date indicates that perennial woody crops will not mimic natural forest stands, but will contribute to diversification of habitat in agriculturally dominated landscapes.

Conclusions. The widespread natural range of eastern cottonwood, plus the possibility of extending the adaptive range by interspecific hybridization, points to the fact that poplar is one of the most promising species groups for woody crops development nationally. High rates of biomass productivity, amenability to clonal propagation and agricultural management, as well as coppicing ability, are factors that make poplar a desirable crop to produce biomass for energy as well as other products. Past research has documented acceptable yields of these systems using genetic material that is essentially one generation away from native populations. Genetic improvement research underway in Iowa, Minnesota, and the Pacific Northwest has demonstrated significant gains in biomass yield and the benefits of a concerted breeding and field testing effort. Continued research in genetics and stand management is needed to improve yield and extend the range of high-yielding varieties to all regions where biomass crops may be planted.

Figure 5.12 Harvesting willow with a one-pass cut and chip forage harvester

(Courtesy of T.Volk, SUNY)

5.1.6 Willow

Interest in shrub willows (Salix spp.) as a perennial energy crop for the production of biomass has developed in Europe and North America over the past few decades because of the multiple environmental and rural development benefits associated with their production and use (Borjesson, 1999; Volk et al., 2004; Rowe et al., 2008). Initial trials with shrub willows as a biomass crop were conducted in the mid-1970s in Sweden with the first trials in the United States starting in 1986 (Volk et al., 2006). Since the initial trials in upstate New York in the mid-1980s, yield trials have been conducted, or are underway, in 14 states (Delaware, Indiana, Illinois, Maryland, Michigan, Minnesota, Missouri, New Jersey, New York, Pennsylvania, South Carolina, Virginia, Vermont, and Wisconsin) and six provinces in Canada.

Biology and adaptation. Willow shrubs have several characteristics that make them an ideal feedstock for biofuels, bioproducts, and bioenergy: high yields that can be sustained in 3- to 4-year rotations, ease of propagation from dormant hardwood cuttings, a broad underutilized genetic base, ease of breeding for several characteristics, ability to resprout after multiple harvests, and chemical composition and energy [3-year-old willow stems averaged 8,340 Btu per dry pound (Miles et al., 1996)], similar to other northern hardwood species.

Production and agronomics. The shrub willow cropping system consists of planting genetically improved varieties in fully prepared open land where weeds have been controlled. The varieties of shrub willow that have been bred and selected over the past two decades in New York can be grown successfully on marginal agricultural land across the Northeast. Midwest, and parts of the Southeast. This range could be expanded with the development of new varieties. Weed control usually involves a combination of chemical and mechanical techniques and should begin in the fall before planting if the field contains perennial weeds, which is often the case with marginal land. Willows are planted as unrooted, dormant hardwood cuttings in the spring as early as the site is accessible at about 6,070 plants per acre using mechanized planters that are attached to farm tractors and operate at about 2.0 acres per hour. To facilitate the management and harvesting of the crop with agricultural machinery, willows are planted in a double-row system with 5 feet between double rows, 2.5 feet between rows, and 2 feet between plants within the rows. Following the first year of growth, the willows are cut back close to the soil surface during the dormant season to force coppice regrowth, which increases the number of stems per stool from 2-4 to 8-13 depending on the variety (Tharakan et al., 2005). After an additional 3 to 4 years of growth, the stems are mechanically harvested during the dormant season after the willows have dropped their leaves (Figure 5.12). Forage harvesters with a specially designed cutting head cut the willow stems 2-4 inches above the ground, feed the stems into forage harvester, and produce uniform and consistent sized chips that can be collected and delivered directly to end users with no additional processing (Abrahamson et al., 2002; Volk et al., 2006). The chipped material is then delivered to end users for conversion to bioenergy, biofuels, and/or bioproducts.

Figure 5.13 Spring resprout after fall cutting of willow

(Courtesy of T. Volk, SUNY)

The plants will sprout again the following spring when they are typically fertilized with about 90 pounds per acre (Figure 5.13) (Abrahamson et al., 2002; Adegbidi et al., 2003) of commercial fertilizer or organic sources like manure or biosolids. The willows are allowed to grow for another 3- to 4-year rotation before they are harvested again (Figure 5.12). Projections indicate that the crop can be maintained for seven rotations before the rows of willow stools begin to expand to the point that they are no longer accessible with harvesting equipment. At this point the crop can be replanted by killing the existing stools with herbicides after harvesting and the killed stools are chopped up with a heavy disk and/or grinding machine followed by planting that year or the following year.

Potential yield and production costs. A rapid growth rate is one of the attributes that makes shrub willows an appealing biomass crop. Yields of fertilized and irrigated, unimproved varieties of willow grown for 3 years have exceeded 12 dry tons per acre per year (Adegbidi et al., 2001; Labrecque and Teodorescu, 2003). Due to the costs associated with irrigation and the relatively low value for biomass, irrigation will probably not be used for most large-scale production operations, with the exception of situations where willow crops could be irrigated with wastewater as part of a nutrient management plan. First-rotation, non-irrigated research-scale trials, with unimproved varieties in central New York, have produced yields of 3.8 to 5.2 dry tons per acre per year (Adegbidi et al, 2001; 2003; Volk et al., 2006). Second rotation yields of the five best-producing varieties in these trials increased by 18% to 62% compared to first-rotations (Volk et al., 2001), and in subsequent rotations, yields are maintained and largely dependent on weather conditions. The most recent yield trials using improved varieties of willow that have been bred and selected for biomass production in New York are showing yield increases of 20% to 40%.

The large genetic diversity across the genus Salix and the limited domestication efforts to date provide tremendous potential to improve yield and other characteristics, such as insect and disease resistance and growth form of willow biomass crops. The species used in woody crop systems are primarily from the subgenus Caprisalix (Vetrix), which has over 125 species worldwide (Kuzovkina et al., 2008). Breeding and selection of willow biomass crops in the United States began in the mid-1990s and has continued with various levels of effort since that time (Smart et al.. 2008). Selection trials of new varieties from the initial rounds of the breeding programs in the late 1990s have produced yields that are up to 40% greater in the first rotation than the standard varieties used in early yield trials. Second rotation results from these same trials indicate that the yield of some of the new willow varieties is more than 70% greater than the standard varieties. These results indicate that there is a large potential to make use of the wide genetic diversity of shrub willows to improve yields with traditional breeding and selection.

The economics of willow biomass crops have been analyzed using a cash flow model (EcoWillow v. 1.4 (Beta) that is publically available (Buchholz and Volk, 2011). The model incorporates all the stages of willow crop production from site preparation and planting through harvesting over multiple rotations to transportation of harvested chips to an end user. For the base case scenario in Eco Willow, the internal rate of return of willow biomass crops over seven 3-year harvest cycles (22 years) is 5.5%, and the payback is reached in the 13th year at assumed sale price of $60 per dry ton. Harvesting, establishment, and land rent are the main expenses associated with willow biomass crops over their entire lifespan, making up 32%, 23%, and 16% of the total undiscounted costs. The remaining costs, which include crop removal, administrative costs, and fertilizer applications, account for about 29% of the total costs.

The development of new harvesting technology is reducing costs by optimizing productivity. Another approach is to reduce the frequency of harvesting operations. Increasing the rotation length from 3 to 4 years reduces harvesting costs by 14% (from $14.79-$12.70 per dry ton) and increases the IRR by 11% (from 5.5%-6.2%).

Establishment costs are the second largest cost in the willow biomass crop production system and account for 23% of the total cost. Over 63% of these costs are for planting stock, so decreasing this input cost will affect the overall economics of the system. For instance, decreasing costs from a cutting from $0.12 to $0.10 reduces establishment costs by $106 per acre and increases the IRR of the system from 5.5% to 6.5%.

Several other components of the system need to be developed to improve the overall economics of willow biomass crop systems, and one of the main ones is yield. Increasing yields from the base case of 5.4 dry tons per acre annually by 50% to 8 dry tons per acre per year increases the IRR from 5.5% to 14.6%. With ongoing breeding and selection, as well as efforts to improve crop management, these levels of yield increases should be possible in the near future.

Sustainability. Willow biomass crops are being developed as sustainable systems that simultaneously produce a suite of ecological and environmental benefits in addition to a renewable feedstock for bioproducts and bioenergy (Volk et al., 2004; Rowe et al., 2009). The perennial nature and extensive fine-root system of willow crops reduce soil erosion and non-point source pollution relative to annual crops, promote stable nutrient cycling, and enhance soil carbon storage in roots and the soil (Ranney and Mann, 1994; Aronsson et al., 2000; Tolbert et al., 2000; Ulzen-Appiah, 2002). In addition, the crop is constantly in its rapid juvenile growth stage, so the demand for nutrients is high, which results in very low leaching rates of nitrogen, even when rates of applications exceed what is needed for plant growth (Adegbidi, 1999; Mortensen et al., 1998; Aronsson et al., 2000). The period with the greatest potential for soil erosion and nonpoint source pollution is during the first 1.5 years of establishment of the crop when cover is often limited because weeds need to be controlled and the willow canopy has not closed. The use of a winter rye cover crop has proven to be effective at providing cover for the soil without impeding the establishment of the willow crop (Volk, 2002). Since herbicides are only used to control weed competition during the establishment phase of willow biomass crops, the amount of herbicides applied per hectare is about 10% of that used in a typical corn-alfalfa rotation in upstate New York.

Nutrient removal from willow biomass crops is limited because only the aboveground woody portion of the crop is harvested during the dormant season after the leaves have dropped and most nutrients have been translocated to the root system. Nutrients not translocated from the foliage are returned to the system in litter. For most soils in the region where willow is being deployed, the only nutrient addition that is recommended is nitrogen, which is typically added at the rate of about 100 pounds of nitrogen per acre once every 3 to 4 years in the spring after the crop is harvested. However, research is ongoing to address concerns about nutrient management across a range of sites with new varieties of willow.

The recommended planting scheme for willow biomass crops is designed to maintain both genetic and structural diversity across a field and the landscape. Blocks of four or more willow varieties from different diversity groups should be planted in each field so that the structural and functional diversity of the system across the field is improved and any potential impact associated with pests and diseases in the future is reduced (Figure 5.13). At the landscape level, willow biomass crops will be in different stages of growth each year because they are managed on a three-year coppice cycle, which will further increase the structural diversity of the system.

Birds are one indicator of the biodiversity supported by willow biomass crops that have been studied in the United States. A study of bird diversity in willow biomass crops over several years found that these systems provide good foraging and nesting habitat for a diverse array of bird species (Dhondt et al., 2007). Thirty-nine different species made regular use of the willow crops and 21 of these species nested in them. The study found that diversity increased as the age of the willows and the size of the plantings increased.

It also found that birds have preferences for some varieties of willow over others (Dhondt et al., 2004). The number of bird species supported in willow biomass crops was similar to natural ecosystems, such as early succession habitats and intact eastern deciduous forest natural ecosystems. Willow biomass crops will increase diversity, especially in contrast to the open agricultural land that it will replace, rather than creating monocultures with a limited diversity across the landscape Lifecycle analysis of willow biomass crops has shown that they are low carbon fuels because the amount of CO2 taken up and fixed by the crop during photosynthesis is almost equal to the amount of CO2 that is released during the production, harvest, transportation, and conversion of the biomass crop to renewable energy (Heller et al., 2003). The cycle is balanced for all the CO2 inputs into the atmosphere from the system because only the aboveground portion of the willow biomass crop is harvested and used in the conversion process. When willow biomass is used to offset fossil fuels, it can help reduce the amount of CO2 emitted to the atmosphere. If the 99 million acres of available land in the United States were planted and harvested with short rotation woody crops to offset coal use for power production, up to 76% (11 quadrillion tons carbon per year) of the carbon offset targets for the United States under the Kyoto Protocol could be met (Tuskan and Walsh, 2001).

The low input intensity of willow biomass crops relative to agricultural crops and their perennial nature result in a large, positive net energy ratio for the biomass that is produced. Accounting for all the energy inputs into the production system, starting with the nursery where the planting stock is grown through to the harvesting of biomass, converting it to chips and delivering it to the side of the field, results in a net energy ratio of 1:55 (Heller et al., 2003). This means that for every unit of nonrenewable fossil fuel energy used to grow and harvest willow, 55 units of energy are produced and stored in biomass. Replacing commercial nitrogen fertilizers, which are produced with large inputs of fossil fuels, with organic amendments, such as biosolids, can increase the net energy ratio to 73-80 (Heller et al., 2003). Transporting the woody biomass 24 miles from the edge of the field to a coal plant where it is co-fired with coal to generate electricity results in a net energy ratio of 1:11. If a gasification conversion system is used, the net energy ratio is slightly higher (Keoleian and Volk, 2005).

Conclusions. Shrub willows have the potential to be grown on marginal agricultural land as a dedicated energy crop across a large range in the United States. The decades of research in Europe and North America provide a solid foundation for the large-scale deployment of the crop. This transition has begun with new varieties of shrub willow being scaled up in commercial nurseries in the United States and Canada, and the engagement of agricultural equipment manufacturers, like Case New Holland, in the development of harvesting systems. The continued optimization of the willow crop production system, a strong breeding and selection program, and quantification of the environmental and socioeconomic benefits associated with the crop are important for the effective and successful expansion of willow biomass crops. The proper deployment of willow biomass crops has the potential to put millions of acres of marginal agricultural land back into production, annually produce millions of tons of biomass, create thousands of rural jobs, and produce an array of environmental benefits.

5.1.7 Eucalyptus

Eucalyptus spp. is the world's most widely planted hardwood species. Its fast, uniform growth, self-pruning, and ability to coppice (regrow after harvest) make it a desirable species for timber, pulpwood, and bioenergy feedstocks (Figure 5.14). It has been domesticated for various products and has been widely commercialized in the tropics and subtropics.

In the United States, eucalyptus was introduced as early as the 1850s on the West Coast to produce dimension lumber and has been produced commercially in Florida since the 1960s. Though eucalyptus has naturalized in areas of the Southwest raising concerns of invasiveness, there is no evidence of spreading in the Gulf South. In anticipation of an increased role in biomass production, ongoing efforts aim to develop eucalyptus cultivars for improved yield and frost resistance in the southern United States.

Biology and adaptation. There are over 700 species of eucalyptus, adapted to various ecological conditions across its native range of Australia. Less than 15 species are commercially significant worldwide. In the South, genetic improvement programs are selected for fast growth, cold tolerance, desirable growth form, and reduced lignin. Genetic improvement programs aim to improve varieties for various growing conditions (Gonzalez et al., 2010; Rockwood and Carter, 2006a).

Production and silviculture. Eucalyptus production practices in different parts of the world vary with site conditions, desired products, and scale of commercialization. Genetic selection has led to commercialization of genotypes with unique advantages in different applications. They are commercially propagated by both seed and cloning of tissue culture. For conventional pulpwood production, stands are typically established at a planting density of 600-1,000 trees per acre, and harvested every 6-10 years. They may be replanted at harvest, which can benefit from improved genetic material, or regenerated from coppice growth, which eliminates the cost of replanting. Economically optimum time between harvests may be 3-4 years, with replanting after 2-5 harvests, on stands with initial densities of 3,400 trees per acre in Florida (Langholtz et al., 2007).

Figure 5.14 Eucalyptus plantation in Florida

(Courtesy of ArborGen and R. Gonzalez, NCSU)

Silvicultural strategies in the United States continue to evolve with changing markets, genotypes, and applications.

Because of high growth rates and tolerance to a range of growing conditions, eucalyptus can be produced in innovative ways, providing non-market benefits. For example, research trials demonstrate that E. grandis and E. amplifolia can be used for restoration of phosphate-mined lands (Rockwood and Carter, 2006b, Langholtz et al., 2007; 2009). Eucalyptus spp. has been shown to be effective at phytoremediation of reclaimed wastewater, municipal waste, storm water, and arsenic-and trichloroethylene-contaminated sites (Rockwood et al., 2004; Langholtz et al., 2005). Eucalyptus plantations that provide these types of environmental services may be viewed more favorably by the public, and compensation for non-market environmental services would improve the profitability of these systems.

Potential yield and production costs. Eucalyptus yields are influenced by precipitation, fertility, soil, location, and genetics. Eucalyptus spp. yielded 7.6-14.3 dry tons per acre annually after 3-5 years of growth on a clay settling area in central Florida, comparable to 8.9-13.8 dry tons per acre estimated for eucalyptus in Florida (Rahmani et al., 1997), but higher than the estimated 4-7.6 dry tons per acre estimated by Klass (1998), who observed that yields could be improved with SRWC development in the subtropical South. E. grandis is a high-yielding species in southern Florida, while E. amplifolia has the advantage of being more frost tolerant, with current trials as far north as South Carolina. The subsequent analysis in this report assumes a conservative annual yield average of 6.0 dry tons per acre.

Sustainability. Intensive management of eucalyptus, characterized by short rotations of genetically uniform monocultures, has dramatically increased yields over recent decades. These tree plantations maintain some sustainability attributes associated with forested landscapes, while at the same time facing sustainability challenges common in agriculture (Binkley and Stape, 2004). Infrequent tilling in tree plantations reduces risk of soil erosion associated with annual crops, and carbon sequestered in eucalyptus stand biomass exceeds the amount of carbon sequestered in herbaceous crops.

Conclusions. Eucalyptus has proven to be one of the most productive and economically viable biomass crops in the world, with expansive commercialization on all populated continents. As with other biomass crops, high yields require fertilization and water. Intensively managed plantations offer both environmental benefits over conventional agricultural systems and potential environmental downsides if native ecosystems are displaced. It is expected that eucalyptus will continue to be produced commercially in the United States and will play an increasing role as a feedstock for bioenergy systems.

5.1.8 Southern Pines

Pines comprised 32% of the tree species planted for production purposes around the world in 2005 (FAO, 2007) and 83% of tree species planted in the southern United States (USDA Forest Service, 2007b). Softwoods in the southern United States already contribute 40% of the total annual industrial wood supply of roundwood (USDA Forest Service, 2007c) and 40% of southern softwoods are used for pulpwood and composites. Because the fiber industry has long used both bark and black liquor to produce energy for running the pulp mills, southern pines are already a significant contributor to U.S. biomass energy.

Biology and adaptation. Loblolly pine (Pinus taeda L.) is the most important and widely cultivated timber species in the southern United States. Because it grows rapidly on a wide range of sites, it is extensively planted for lumber and pulpwood (Figure 5.15). This tree is dominant on 30 million acres and comprises over half of the standing pine volume in the South (USDA Forest Service, 2007c). A medium lived loblolly matures in about 150 years, with select trees reaching 300 years in age. Other pine species are found in the South, including slash pine (Pinus elliottii Englem), longleaf pine (Pinus palustris Mill), and shortleaf pine (Pinus echinata Mill); hybrids of loblolly and the three other species are also found (Peter, 2008). Of these, loblolly and slash pine are most frequently planted, and loblolly is the most important southern pine for bioenergy feedstock production. Loblolly shows a strong growth response to management inputs and is the best choice on good sites with better-drained soils where hardwood competition is a problem.

Figure 5.15 Pine plantation

(Courtesy of William M. Ciesla, Forest Health Management International)

Production and silviculture. Improvements in pine silviculture have resulted in improving southern U.S. pine productivity by a factor of about 6 since the 1940s and increasing the number of planted acres of all pines from zero in 1940 to 37.66 million acres by year 2006 (USDA Forest Service, 2007a). The change from relying on natural pine stands to establishing and intensively managing pine plantations for fiber production is one of the major success stories in plantation forestry (Fox et al., 2007b). Loblolly pines are now deemed to be one of the most productive species that could be used in the southern United States for supplying bioenergy resources (Gonzalez et al., 2009).

Loblolly pines are normally planted as 1-year-old bare-root seedlings, though the more expensive containerized seedlings offer several advantages, including better survival (Taylor, 2006). Production of bare-root seedlings involves planting seed in specialized beds with controlled conditions for 8-12 months, top pruning, lifting, and grading. Currently 0.8 to 1.0 billion loblolly and slash pine bare-root seedlings are sold annually for forest planting. Essentially all of the seed is genetically improved for growth and disease resistance, with 70% of the seedlings being loblolly pine and 30% slash pine (Peter, 2008).

Many steps have contributed to improving the productivity of loblolly pine in the South (Stanturf et al., 2003a; Fox et al., 2007a). Naturally regenerated forests were the common practice from the 1920s through the 1950s, with very low annual productivity. Improved nursery and field planting practices began in the 1950s with continued improvement through the 1970s, and as a result, whole tree aboveground yields tripled. Seed orchards dedicated to seed improvement were first established in the late 1950s. The first generation improved seeds increased value of plantation wood by 20%, and second generation improved seeds being used now are adding another 14%—23%. The importance of hardwood competition control was recognized by the early 1970s. First methods of control were entirely mechanical, but by the late 1970s herbicides were added, and by 1990, chemical site preparation was predominate with limited mechanical site preparation involved. Fertilization of pine plantations was initiated in the late 1960s, but was implemented slowly during the 1970s and 1980s (Albaugh et al., 2007). Average productivity increased rapidly from the 1970s to 1990s primarily as a result of implementing use of improved site preparation, hardwood competition control, and genetically improved seeds.

Implementation of silviculture and genetic improvements very much accelerated in the 1990s as a result of the non-proprietary research conducted by university-industry cooperatives. In 1999, there were 23 research cooperatives at nine southern universities (Stanturf et al, 2003b). During the 1980s and 1990s, cooperative research clearly confirmed the benefits to pine productivity of fertilizing with both nitrogen and phosphorus, especially in mid-rotation. Further research published since 2000 has shown the need for micronutrients on certain soil types (Fox et al., 2007a;b; Kyle et al., 2005). Other recent studies have compared the effects of management intensity levels (Borders et al., 2004; Cobb et al., 2008; Martin and Jokela, 2004; Roth et al., 2007; Samuelson et al.. 2008; Will et al., 2006), clearly showing the potential for much higher yields. Since third generation seeds from selected parents were beginning to be deployed in the early 2000s (McKeand et al., 2003), several of the recent research trials have included a higher performing genotype that resulted in enhanced yields.

At present, most loblolly pines stands in the South are managed for a combination of pulp and timber so that thinning is incorporated into the management. The stands are planted on average at about 600 seedlings per acre (~1480 seedlings per hectare), planning for a 25-year rotation with a thinning at age 15 (Gonzalez et al., 2009). With many studies showing the benefits of weed control and fertilization, mid-rotation fertilization has become considerably more common (Albaugh et al., 2007). Average operational yields in the southeastern United States were reported in 2003 to be about 4 dry tons per acre annually total aboveground oven-dry weights (Stanturf et al., 2003b). Current yield potential is assumed to be higher with the recent deployment of third generation loblolly pine seedlings on sites with site preparation treatments that ensure adequate survival and rapid early growth. Future management techniques are predicted to include "clonal plantations, whole rotation resource management regimes, use of spatially explicit spectral reflectance data as a major information source for management decisions, active management to minimize insect and disease losses, and more attention to growing wood for specific products" (Allen et al., 2005).

Potential yield and production costs. Loblolly pine research plots managed with site preparation and weed control but no fertilizers have produced total aboveground biomass yields (stem, branches, and foliage) of 3.3 to 3.8 dry tons per acre per year. Research plots with site preparation, weed control, and fertilization only at planting have produced total yields in the 3.6 to 5.2 dry tons per acre per year range. Addition of higher levels of fertilizers plus irrigation in some cases has bumped yields to 5.1 to 7.3 dry tons per acre per year of biomass. Very intensive management with selected loblolly pine genotypes, annual fertilization, irrigation (in some cases), excellent site preparation, and weed control has increased biomass yields to 5.4 to 8.5 dry tons per acre per year. Based on recently reported research results, companies are predicting future operational yields of 6 to 8 dry tons per acre per year when greater management intensity is used. However, it is unlikely that yearly fertilization will be economically viable or indeed it may not be necessary for high-yield achievement.

Various ideas have been proposed on how to manage southern pines for bioenergy production. Both Gonzalez et al. (2009) and Scott and Tiarks (2008) have recently described management plans for producing both timber and bioenergy products. Both involve a combination of rows of widely spaced trees and tightly spaced rows for bioenergy. The bioenergy rows would be harvested in 5 to 8 years and a widely spaced row for lumber production to be harvested at 18 to 22 years. While this might be a reasonable transition strategy, an efficient harvesting strategy for removing the bioenergy trees has not been discussed. Planting and harvesting can be much more efficient when pine plantations are dedicated entirely to supplying bioenergy feedstocks. Such plantations are likely to be planted at higher densities and managed on shorter rotations similar to poplars and eucalyptus.

The age of optimal stand harvest has not yet been determined for higher density loblolly pine plantings. Recent intensive management studies planted at stand densities of 454 to 670 trees per acre show total aboveground biomass continuing to increase between 10 and 15 years of age (Samuelson et al., 2008; Borders et al., 2004). However, those same studies also show density-dependent mortality beginning at basal areas of about 153 square feet per acre on fertilized wet sites, which correlates to an age range of about 9 to 10 years. The highest density study with 1,210 trees per acre showed a slowing of the current annual increment by age 5, but the mean annual increment was still increasing (Roth et al., 2007). The cost of planting will depend on initial planting density and the amount of replanting needed (Taylor et al., 2006). Advanced generation, bare-root seedlings were reported to cost $47.50 per thousand seedlings in 2006. Over the planting ranges mentioned above, and including culls and extra seedlings needed for replanting, seedling costs could be expected to range from about $40 to $60 per acre. Planting with current planting equipment is expected to cost about $65 to $100 per acre.

Harvesting of small-diameter trees has been a significant cost barrier to using southern pines for energy (Peter, 2008) but the results of intensive management studies are showing that excellent growth can be achieved at densities low enough to allow individual trees to achieve an economically harvestable size. Consequently, harvest and handling costs (to roadside) using currently available equipment should be similar to current pulp harvesting costs or about $20 per dry ton.

Economically optimal fertilization strategies will vary for each planting site. Intensive culture studies produce higher yields with high annual fertilization fairly consistently, while financial returns depend on the magnitude of the growth response obtained, the product mix, stumpage prices, cost of fertilization, and the length of time before harvest (Fox, 2007b). As with hardwoods, first fertilization with nitrogen and phosphorus should be delayed a year or two to avoid stimulating weed competition, but no later than stand closure. Mid-rotation fertilization applications of both nitrogen and phosphorus (at 200-pound nitrogen per acre and 25-pound phosphorus per acre applied at time of stand closure) have shown very positive stand responses lasting for several years in lower density stands, but more frequent fertilization at lower levels may be needed in higher density loblolly stands (Fox et al, 2007a).

Sustainability. Use of intensive management to produce wood specifically for bioenergy is generally only economically viable when the total aboveground portions of the trees are removed. This has raised concern about long-term site productivity impacts. Research and analysis of intensive pine production has shown that good site preparation, chemical control of non-crop vegetation, and fertilizer application at levels and times that optimize utilization by the trees, increases biomass yields in an energy-efficient manner, while maintaining or improving long-term site productivity (Scott and Dean, 2006). Allen et al. (2005) argue for use of a fully integrated management approach starting with good site selection followed by excellent early competition control and additional inputs, as needed. Such management practices will not only create economically sustainable woody production systems, but will also minimize the potential for adverse environmental effects.

Conclusions. In the near term, pine bioenergy feedstocks are most likely to be obtained by thinning existing loblolly pine stands that are planted for multiple uses (fiber and energy). If loblolly pines are planted specifically for energy, then they will be grown at relatively dense spacings and short (8-10 year) rotations. Research studies suggest that the lowest planting density under intensive management that might be expected to achieve an economically harvestable size within that time period is about 726 trees per acre. Average yields of about 5.5 dry tons per acre annually in the Southeast, Atlantic Coast, and Delta regions are obtainable with appropriate management. This includes plowing, disking, and application of a total kill herbicide once or twice before planting. Non-crop vegetation is controlled during the first 2 years, primarily with herbicide applications. In the southern United States, phosphorus and potassium are usually added to high-yield stands in the planting year, and nitrogen additions of about 89 pounds per acre are added in years 2 through 6, based on foliar analysis studies showing nitrogen demand levels (Will et al., 2006). Economically viable harvest is expected to occur as early as the eighth year. Both traditional and molecular genetics need to continue to be aggressively pursued to improve the productivity potential of loblolly and other pines, and substantial yield improvements are expected between now and 2030.

50 Much of the research was conducted under the Department of Energy's Biomass Feedstock Development Program. More than 150 woody and 35 herbaceous crops including nearly 20 perennial grasses were screened and evaluated as potential energy crops. A historical perspective on herbaceous and woody crops can be found in Wright (2007) and Wright (2012).

51 Corn grain yields have risen at an average annual increase of 1.7 bushels per acre even while fertilizer inputs have declined (Dobermann et al., 2002).

52 Assuming an ethanol yield of about 85 gallons per dry ton (U.S. Department of Energy, 2011).