Douglas L. Karlen1 and David R. Huggins2

INational Laboratory for Agriculture and the Environment, USDA Agricultural
Research Service, U. S.A.

2Land Management and Water Conservation Research Unit, USDA Agricultural
Research Service, U. S.A.

7.3 Overview

Crop residues (e. g., corn stover and small grain straw) are sometimes excluded when discussing cellulosic energy crops per se, but because of the vast area upon which they are grown and their current role in the development of cellulosic energy systems, this chapter will review several important attributes of this “herbaceous” feedstock. Crop residues are potential feedstock sources for second-generation biofuel production. These materials, along with dedicated energy crops (e. g., switchgrass [Panicum virgatum L.], Miscanthus [Miscanthus x giganteus]), are considered to have greater potential for biofuel production than current first-generation feedstock (i. e., corn grain) [1-3]. Production of ethanol and other fuel sources from these lignocellulosic materials is receiving increased financial support for research and development [4-6]. Furthermore, biofuel production from crop residues provides a multipurpose land use opportunity where grain can be harvested to meet food and feed demands, while a sustainable portion of the residues provide a potentially available biofuel feedstock.

Corn stover, the aboveground plant material left in fields after grain harvest, was identified as an important biomass source in the Billion-Ton Study (2005 BTS) [7]. The vast area from which this feedstock could potentially be harvested was confirmed by USDA National Agricultural Statistics Service (NASS) data showing that between 2005 and 2011, corn was harvested in the U. S.A. from an average of 32 460 000 ha each year [8]. Wheat straw was the other dominant residue identified in the 2005 BTS, and from 2005 through 2011, wheat was harvested in the U. S.A. from an average of 20 037 000 ha each year. Based on these

Cellulosic Energy Cropping Systems, First Edition. Edited by Douglas L. Karlen. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd.

vast harvest areas, the 2005 BTS projected total annual corn and wheat residue production to be approximately 250 and 90 million Mg, respectively, with a sustainable removal of 82 and 12 million Mg after accounting for that needed to mitigate wind and water erosion.

The 2005 BTS projections of available crop residue immediately raised concern among many soil scientists because harvesting residues as a biofuel feedstock or for any other purpose (e. g. animal feed) will decrease annual carbon input and may gradually diminish soil organic carbon (SOC) to a level that threatens the soil’s production capacity [9]. Concerns within the U. S. Corn/Soybean Belt were accentuated knowing that for many soils artificial drainage, intensive annual tillage, and less diverse plant communities have already reduced SOC by 30-50% when compared to pre-cultivation levels [10]. Returning a portion of crop residues to replenish SOC was deemed essential for sustainability [11-16] because crop residues influence many vital soil, water, and air functions. Many scientists stated that caution must be used to ensure that harvesting residue for any use does not compromise ecosystem services or decrease overall soil productivity. Furthermore, others argued that for several current cropping systems, soil erosion and organic matter depletion indicate that crop residue returns to the soil are already insufficient [17,18].

As a result of soil resource sustainability concerns raised by the 2005 BTS, a follow-up report (2011 BT2) was developed by the U. S. Department of Energy (DOE) to include (1) a spatial, county-by-county inventory of potentially available primary feedstocks, (2) price and available quantities (i. e. supply curves) for individual feedstocks, and (3) a more rigorous treatment and modeling of resource sustainability [19]. The 2011 BT2 recognizes the importance of crop yield variation and the need to balance the economic drivers with ecologically limiting factors [20]. Table 8.1 presents some of the estimated feedstock supplies for various crop residues at selected price levels. These values are also consistent with several other estimates including those used for the U. S. National Academy of Science (NAS) study on Liquid Transportation Fuels from Coal and Biomass [21]. The 2011 BT2 also provides a more realistic overview of total crop residue availability and sets some achievable research and development goals for available feedstock supplies by creating various production scenarios that strive for higher crop yields and integrate multiple cellulosic energy crops into potential production systems.

Several assessments examining the multiple roles that crop residues have for maintaining multiple ecological functions have been published since the 2005 BTS [22-30]. Therefore, this chapter focuses on current corn stover and wheat straw research designed to address

concerns raised by those previous reviews and to help ensure that commercial bioenergy develops in an economically, environmentally, and socially acceptable manner.

Several key bottlenecks in the willow crop production system have been overcome during the past few years, making deployment on a large scale possible. One of those barriers has been availability of large quantities of shrub willow planting stock. Over the past few

years a commercial nursery in western New York (Double A Willow) planted over 60 ha of willow nursery beds to meet the projected annual demand for millions of planting stock cuttings and several other nurseries are being planned.

Another significant bottleneck has been how to efficiently and economically harvest the crop and produce a consistent quality product that is acceptable to end users. Since 2004, Case New Holland (CNH) has been working with SUNY-ESF and other partners to develop a harvesting system for willow biomass crops based on a New Holland self-propelled forage harvester and a header that is designed to cut short rotation woody crops. Trials with the latest version of this system, based on a New Holland FR series self-propelled harvester and a 130 FB coppice header, indicate that for three or four-year-old willow biomass crops with the majority of stems <75 mm in diameter, consistent high quality chips (>95% of the chips being smaller than 37.5 mm) can be produced at a harvest rate of about 0.8-1.8 ha h-1. Additional actions are being tested to further improve this machine [35]. As noted above, the improved production rates that are possible with this harvester will have a direct impact on the delivered cost of willow biomass.

12.2 Conclusions

In order to meet the projected demand for biomass for the production of bioenergy, biofuels, and bioproducts in the United States, perennial energy crops will need to be developed and deployed across millions of hectares over the next 25-30 years. Over the past few decades, research in Europe and North America has resulted in the development of a shrub willow production system. Thousands of hectares of shrub willow crops have been deployed in Europe, and the system is beginning to be expanded in the United States, but the future of it as a sustainable system will depend on continued research on biological, ecological and socioeconomic factors, development of a feedstock production and supply infrastructure, and supportive renewable energy policies.

Many characteristics of shrub willows, and the production system that has been devel­oped, contribute to sustainability of the system. The perennial nature of willows, their extensive diffuse root system and the coppice management approach that has been devel­oped, result in a crop that can be maintained and productive for more than two decades after it is planted. These characteristics create tight nutrient cycles and a permanent crop on the landscape that will improve soil and water quality as well as biological and landscape diversity relative to traditional annual agricultural crops. The potential of the system to sequester greenhouse gases and its high-energy return on investment are other key features that contribute to its sustainability.

Under existing policy structures, the economics of willow biomass crops are marginal because of the relatively high cost of establishment, low prices for woody biomass, and limited experience with the crop. In addition to optimizing the production system and improving yields, changes in policies to support commercial deployment of willow and other perennial energy crops in the near term are necessary to transition these crops to commercially viable systems. Recent development of the USDA BCAP project in northern New York is a positive step forward. As this and subsequent expansions occur, potential socioeconomic benefits associated with producing a marketable product from marginal agricultural land should begin to accrue to rural areas. Since biomass from shrub willow crops will be integrated with woody biomass from other sources, such as low-grade material from forests and residues from forest harvesting operations, there is potential for benefits to accrue to local communities from revitalization of those sectors as well.

The future challenges are to simultaneously optimize willow biomass crop production, increase interest from potential producers, and develop long-term markets for willow and other sources of woody biomass. To accomplish this and develop a sustainable system, strong links between researchers, potential producers, and end users are required. With these links in place, development of a vibrant willow biomass enterprise will play an important role in bolstering the farm and forestry sectors, while increasing energy independence, providing environmental benefits, and mitigating pollution problems.

Rob Mitchell1, D. K. Lee2, and Michael Casler3

1 Grain, Forage, and Bioenergy Research Unit, USDA Agricultural Research Service, U. S.A.
2Department of Crop Sciences, University of Illinois, U. S.A.

3 U. S. Dairy Forage Research Center, USDA Agricultural Research Service, U. S.A.

5.1 Overview

Switchgrass (Panicum virgatum L.) is aperennial warm-season grass native to the grasslands of North America, is a model perennial grass for bioenergy, and is the most advanced herbaceous perennial bioenergy feedstock. Best management practices have been developed for switchgrass bioenergy production for the agroecoregions to which it is adapted. Field production of switchgrass likely will occur on cropland that is marginally productive for row crops, similar to land that was enrolled in the Conservation Reserve Program. Long-term, field-scale research demonstrates that switchgrass for bioenergy is productive, profitable for the farmer, and protective of the environment.

Switchgrass was selected by the Bioenergy Feedstock Development Program (BFDP) at the U. S. Department of Energy (DoE) as a model herbaceous species because of its potential to simultaneously meet energy demands and address global climate change [1]. It is a perennial, warm-season (C4) grass native to North America that is broadly adapted throughout the United States and is found in every state east of the Rocky Mountains [2]. Like many perennial C4 grasses, switchgrass is highly tolerant to abiotic stresses such as drought, temperature extremes, and salinity. For that reason, it is being recommended for biomass production on marginally productive cropland where it would have minimal land use competition with commercial food crops [3].

5.2 Phylogeny, Growth, Yield and Chemical Composition

Switchgrass is a highly polymorphic species with considerable morphological and physio-
logical variation. Much of this variation can be explained by ecotype, the main taxonomic

Cellulosic Energy Cropping Systems, First Edition. Edited by Douglas L. Karlen. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd.

subdivision named largely for phenotypic differentiation based on habitat [4]. The two ecotypes (upland and lowland) were initially distinguished by phenotypes but now can be separated by cytotypes and gene cluster, using numerous genetic markers [5]. Common upland cultivars are “Shawnee” and “Summer”, whereas common lowland cultivars are “Alamo” and “Kanlow”.

Lowland ecotypes are mostly tetraploids (2n = 4x = 36), whereas upland ecotypes are commonly tetraploid (2n = 4x =36) or octoploid (2n = 8x =72) with hexaploid (2n = 6x = 54) reported rarely [6,7]. Aneuploids appear to be common in switchgrass, particularly at higher ploidy levels [8]. Many molecular methods have been developed and used for studying the genetic relationship between upland and lowland cytotypes. The genetic relationship among 14 populations of upland and lowland switchgrass ecotypes has been characterized by using 92 polymorphic RAPD markers [9]. Hultquist et al. [10] used chloroplast DNA restriction fragment length polymorphisms to show that these upland and lowland ecotypes are genetically different in chloroplast DNA. A deletion of 49 nucleotides in trnL-UAA introns was identified in lowland cp genome [11]. Several recent studies have investigated nuclear polymorphisms using simple sequence repeats derived from expressed sequences tags (EST-SSRs) and identified several lowland and upland subpopulations [7,12-15].

Lowland ecotypes generally are taller, coarser, and more caespitose in growth form than upland ecotypes. Generally, they are better adapted to wetter and warmer environments, whereas upland ecotypes are best adapted to drier and colder environments [4,16]. In general, lowland ecotypes have greater biomass and better disease resistance than upland ecotypes [4,16]. Both ecotypes are largely self-incompatible and plants are cross-pollinated by wind [16].

Switchgrass has the typical anatomical and physiological characteristics of a C4 grass [16]. Seedling development has three phases: germination, emergence, and adventitious root development [17]. Optimum temperature for switchgrass seed germination and seedling growth is between 20 and 30°C, while germination and seedling growth are significantly reduced at soil temperature <20°C [18,19]. Seed germination is initiated with the radicle protrusion and the coleoptile emergence from the seed coat. Once the coleoptile emerges, it is pushed to the soil surface by elongation of the subcoleoptile internode, typical of the panicoid seedling development [17]. When the coleoptile reaches the soil surface, the subcoleoptile internode elongation stops, adventitious roots form, and water uptake and photosynthesis begin for plant growth. This is why proper seeding depth is critical for successful switchgrass establishment. Seeds planted deeper than 1 cm can result in poor establishment because seedling energy reserves are used for subcoleoptile elongation and adventitious root development is delayed [17]. Several tillers may be produced within six weeks of emergence.

Switchgrass growth during the establishment year varies depending on region, weather, soil fertility, and competition with weeds [20], but in general it is feasible to produce and harvest 50% of the cultivar’s yield potential after a killing frost. Furthermore, in the first full growing season after seeding, it is very feasible to produce and harvest 75-100% of the cultivar’s yield potential [20-22] with many fields in the central Great Plains approaching full production of 8-13 Mg ha-1 [23].

New growth in post-establishment years starts in early spring, with new tillers being ini­tiated from axillary buds on the crown and/or rhizomes [24-26]. Moore et al. [27] presented the phenologic development of switchgrass by maturity stages: emergence, vegetative/leaf development, stem elongation, reproductive/floral development, and seed development and ripening. Although the durations of each stage are dependent on genetics, both photoperiod and temperature play a critical role on vegetative growth and reproductive development [28-30]. Mitchell et al. [29] and Castro etal. [31] indicated that photoperiod is the primary determinant of switchgrass development, but temperature or heat units can significantly modify reproductive development.

Switchgrass biomass yield is influenced by agroecoregion and management practices, such as ecotype, cultivar, fertilization, and harvest timing. Maughan [32] reported a meta­analysis of 106 sites from 45 studies covering the eastern two thirds of the United States and southeastern Canada. Switchgrass biomass yield across all regions of the study, including both lowland and upland ecotypes, averaged 6.6 ± 3.0 Mg ha-1 during the establishment year, increased to 9.1 ± 5.5 Mg ha-1 in the second year, and reached a maximum of

10.9 ± 5.2 Mg ha-1 in the third year. During the post-establishment years, biomass yield for lowland and upland ecotypes was 11.1 ± 6.1 and 6.7 ± 3.2 Mg ha-1, respectively. Among regions, the lower central region, equivalent to U. S. Plant Hardness Zones 6 and 7, had the highest biomass of 6.7 ± 3.2 Mg ha-1 and the north region, equivalent to U. S. Plant Hardness Zones 3 and 4, had the lowest biomass yield of 7.3 ± 3.1 Mg ha-1. High-yielding cultivars developed for biomass yield in the Great Plains and Midwest are in the release process for commercial availability.

Lignocellulosic biomass is composed primarily of structural carbohydrates, cellulose and hemicellulose, and lignin, polyphenols, with a lower concentration of other proteins, nutrients, acids, salts, and minerals. Structural carbohydrates, which generally comprise two-thirds of the dry biomass, can be hydrolyzed to sugars and those sugars can be fermented to ethanol or other forms of liquid fuel. Even though lignin is not converted to fuel by the fermentation process, other conversion technologies, such as gasification and fast pyrolysis, could use lignin as an energy source. Biomass yield is the most important characteristic for sustainable bioenergy production. However, feedstock chemical composition and its consistency, which directly influence conversion process yield, are also very important.

Switchgrass has a similar feedstock composition to other lignocellulosic feedstocks. Lee et al. [33] reported that switchgrass biomass has 37% cellulose, 29% hemicellulose, 19% lignin, 3% crude protein, and 6% ash when harvested in late autumn or after a killing frost. They also indicated that the chemical composition of switchgrass is relatively simi­lar to other crop residues, such as corn (Zea mays) stover and wheat (Triticum aestivum) straw. However, growth environment and genetics cause significant variation in feedstock composition [34]. Feedstock composition also has a significant impact on conversion effi­ciency, with one study demonstrating a range in potential ethanol production from 61 to 127 mg g-1 [34]. The range of composition data collected from multiples studies explained this variation, with cellulose, hemicellulose, and lignin varying from 31 to 45%, 22 to 25%, and 18 to 22%t, respectively.

Harvest timing is a major cultural practice affecting feedstock composition [33,35-39]. Delaying harvest to after a killing frost provided biomass with higher structural carbo­hydrates and lignin as well as lower protein and ash compared to biomass harvested at anthesis. Further delaying harvest to the following spring reduced ash and protein concen­trations even more [33,35,36]. Dien et al. [40] reported switchgrass mineral components were related to plant maturity (Table 5.1). Other studies indicate that either late season or

Table 5.1 Stage of maturity is the primary factor controlling switchgrass biomass composition within a cultivar.

post-frost harvest are likely to provide biomass with lower nitrogen, phosphorus, potas­sium, and chlorine [36,38,41]. Consequently, cultural practices can be used to provide a feedstock with the most desirable composition profile [21].

The recommended number of trees to plant per hectare varies depending upon the objectives of the landowner. In general, the number planted per hectare in Europe is higher than in New Zealand and the United States. Due to tradition, pines in Europe may be planted at 2500/ha and in the southern United States the number may be 1000-1400/ha; in New Zealand, the number may be 800-1000/ha [8]. When the delivered cost of pine logs is the same (per green tonne) for pulpwood and for fuelwood, then the economically optimum planting density (for a given species, location and discount rate) should be nearly the same.

When researchers establish a bioenergy pine plantation, they often plant more seedlings than would be recommended by an economic analysis. In one short-rotation experiment in Florida, Pinus clausa was planted at 26 900 seedlings per hectare [9]. Two factors that are often ignored are the cost of establishment and the cost of harvesting. In the southern United States, the cost of harvesting and transporting one tonne of pulpwood to a mill is about two-thirds the price paid at the mill. For example, a pulp mill might purchase green wood for $42 per Mg ($14 to the landowner, $10 for harvesting, and another $14 for transportation). Much of the time in harvesting is spent handling and processing each tree, so tree size is an important determinant in harvesting cost (Figure 10.1). For some planting densities, harvesting one green tonne might require less than 5.5 trees while planting 3000/ha may require cutting twice that many (Table 10.3). In many cases, the target planting density is selected without considering the economic impact of both establishment and harvesting costs.

Regardless of the type of sorghum being grown, the basic growth and development of the crop is similar. Differences between grain sorghum, forage sorghum and energy sorghum are due to specific changes in a few very important traits within the species. Those traits include tillering and regrowth, plant height, stem type, maturity and grain yield potential.

All commercial sorghums have a fibrous non-rhizomatous root system. For a mature sorghum plant, roots can extend at least 1.5 m in all directions from the base of the plant. In addition, tillering is especially important in sorghum, but tiller characteristics and their desirability depend on which type of sorghum is being produced. Sorghum has the potential for both basal and nodal tillering. In general, profuse tillering is desirable in forage sorghum while reduced tillering is more desirable for both grain and energy sorghum. If ratoon crops are desired, then basal tillering is essential to facilitate regrowth after cutting or grazing. Increased tillering is also associated with reduced stem size [17]. Temperature and plant density also influence tillering potential [18].

Stem thickness and composition are additional defining characteristics of different sorghum types. In general, thicker stems are desirable in energy and grain sorghum while thinner stems are appropriate for forage sorghum. Stem thickness in sweet sorghum is highly variable, depending on regional preferences and the primary use of the cultivar (i. e., syrup, feed or ethanol production). While there is significant genetic variation for this trait, stem diameter is strongly influenced by management and environmental conditions. For example, planting density is the largest factor affecting stem diameter. Numerous studies show that plant density and stem thickness are negatively correlated: the greater plant density, the finer the stem [19].

In addition to stem diameter, stem morphology and sugar concentration in the stems also define sorghum type. Sweet sorghum cultivars have a very juicy stem while biomass sorghum typically has a pithier stalk. Juicy and pithy stemmed characteristics are both found in grain and forage sorghum lines. A single gene locus D was reported to control juiciness of the stem [20] but more recent analysis indicates that, while this gene is important, it is not the sole genetic factor influencing juiciness of the stalk [21]. Sugar concentration is critical for sweet sorghum but less so for other sorghum types. Sugar concentrations are also strongly influenced by both genotype and environment. For example, sugar concentrations typically peak in sweet sorghum approximately 20-30 days post anthesis and different genotypes with unique profiles of fermentable sugars have been identified [22].

Plant height is a primary factor in overall biomass yield potential; higher yields are correlated with taller plants [23,24]. However, there are limits to height as taller plants are more prone to both stalk and root lodging, so the benefits of height must be effectively balanced with the risks [25]. Sorghum height is easily manipulated; four major height genes were identified [26] and many modifiers of these genes have been subsequently described [27,28].

Of all sorghum traits, maturity is likely the most important because it influences so many agronomically important characteristics. Sorghum maturity is controlled by genes that are influenced by both day length (photoperiod sensitivity) and temperature [16].

Photoperiod sensitive sorghum cultivars require a defined length of darkness to induce panicle differentiation and development [15]. As sorghum was moved to more temperate climates and away from lower latitudes, photoperiod insensitive sorghums were necessary for the crop to produce grain before the growing season ended. Much of the early work to develop photoperiod insensitive sorghum in the United States was completed by producers who would identify and save seed of individual, early maturing mutants or segregants [26]. In these photoperiod insensitive sorghums, maturity is primarily influenced by temperature.

Six maturity genes have been described in photoperiod sensitive and insensitive sorghum and there are likely many more [29,30]. These loci interact to produce an array of matu­rities in all types of sorghum cultivars. Maturity is important in forage sorghum because forage quality tends to decrease once the crop flowers and starts to produce grain. This is particularly critical in sorghum-Sudan grass hybrids that are grown for hay production. Because of the relationship between maturity and quality, sorghum breeders have devel­oped photoperiod sensitive sorghum-Sudan grass hybrids and forage sorghum hybrids that maintain forage quality with yields similar to photoperiod insensitive hybrids [30].

Andrzej Klasa1 and Doug Karlen2

department of Agricultural Chemistry and Environmental Protection, Warmia and Mazury
University in Olsztyn, Poland

2National Laboratory for Agriculture and the Environment, USDA Agricultural
Research Service, U. S.A.

11.1 Introduction

Populus consists of 25-35 species and among them hybridization is common. The genus itself has a large genetic diversity with some species growing 50 m tall with trunks of up to 2.5 m in diameter. All Populus species from family Saliceae are common in temperate climate zones but they are limited in tropical zones because the maximum temperature they can tolerate is approximately 30°C. Among the various species, Populus alba grows primarily in southern and central Europe, P. tremula in Europe and Asia (mainly in India), and P. tremuloides in North America with their northern border being Alaska. There are some reports concerning Populus deltoides growing in India [1], and although some Populus genotypes can be successfully grown on saline-sodic and alkaline soils, some tested clones could not survive those soil and climatic conditions in Uttar Pradesh province. In an age of globalization there is an increasing tendency for farmers, foresters, and owners of recreation areas to introduce different poplar species in non-native environments. Therefore, numerous Populus species are often found outside their natural borders. In Europe, for example, the most frequently grown poplar is hybrid Populus x euroamericana (Populus deltoides x P. nigra).

Berndes et al. [2] identified poplar short-rotation plantations (SRPs) as one of the most important sources of biomass for energy purposes, pointing out their energy and environ­mental soundness. In the United States, poplar species have been grown commercially for more than 100 years [3] and although, to date, business conditions have restricted their

Cellulosic Energy Cropping Systems, First Edition. Edited by Douglas L. Karlen. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd.

cultivation to specific geographic locations, those developments have provided proof of concept for growing short-rotation trees using intensive agricultural techniques. Therefore, looking to the future and to the great opportunity to develop a viable biomass energy feedstock supply, the poplar industry in many countries is well positioned to play a signif­icant role in meeting global energy needs. Poplar species and their numerous hybrids are very productive genotypes due to their rapid foliar production, very high leaf area index and high photosynthetic rates. It was stressed [4] that the Populus genus, which contains more than 30 species worldwide, has remarkable clonal variation for both biomass pro­duction and distribution. Therefore, some species and clones are much more suitable for lignocellulose biomass production than others because they can allocate more biomass to aboveground plant parts than to roots. By understanding genetic variability among poplar species, Wullschleger et al. [4] pointed out that there is a starting point for future breed­ing programs aimed at obtaining clones of high and reliable yield level with a desirable distribution of assimilates in plant biomass.

Poplar is also one of the so-called energy crops that can, theoretically, improve relations between agriculture and environment. For example, studies on 12 poplar genotypes grown on marginal and agricultural soil in Hungary [5] confirmed wide genetic variability with average biomass yields of 27 Mg ha-1 ODM (oven-dry matter) and average energy yields of 309 GJ ha-1. With regard to carbon, the authors reported that the balance was always positive (i. e. the amount of carbon emitted to the atmosphere as carbon dioxide during combustion was always less than the amount of carbon sequestrated in SRP poplar biomass) because of the high content of carbon in roots, litter and stools. Therefore, under Hungarian conditions, short-rotation poplar can be an effective carbon sink and source of renewable biofuel. Many other environmental benefits associated with growing poplar for energy have been presented from a United States perspective [6], with the prediction that in the United States growing poplar in short-rotation systems will link phytoremediation of contaminated soils with bioenergy production.

The fertilizer requirements, and in particular nitrogen requirements, of biomass crops have significant implications for the carbon footprint and environmental impact of biomass production systems.

Cadoux et al. [90] reviewed 27 studies dealing with the nitrogen, phosphorus and potas­sium (NPK) requirements of Miscanthus x giganteus. While significant amounts of nitro­gen, phosphorus and potassium are taken up by the crop, only a fraction of this peak nutrient content is removed during the winter harvest due to translocation of nutrients from shoots to rhizomes during winter (Table 4.3). In addition, leaf fall and nutrient leaching from the stems returns some of the absorbed nitrogen, phosphorus and potassium to the soil [37, 38], a proportion of which is used by the crop in subsequent years, along with remobilized nutrients from the rhizomes [28, 37, 38, 91].

The yield response of M. x giganteus to nitrogen fertilization is limited and varies between sites [90, 92, 93]. Just over half of eleven studies reviewed by Cadoux et al. [90] concluded a positive response of M. x giganteus to nitrogen, while five showed an absence of a response. In the studies showing a positive response to nitrogen, the response was generally moderate except under irrigation, where the response was higher. The absence of response in the remaining studies can be explained by the plants having sourced their nitrogen via rhizome remobilization and from available soil mineral nitrogen. Crop age also impacts on the nitrogen response with young plants requiring little to no nitrogen in the first few years. For example, using a meta-analysis, Miguez et al. [63] found Miscanthus did

not respond to nitrogen fertilizer during the first two growing seasons and then responded only slightly to a nitrogen rate of 100 kg ha-1.

The environmental impacts of over fertilization must be taken into account when consid­ering the nitrogen requirements of Miscanthus. Christian et al. [46] compared the nitrogen balance of Miscanthus crops over 10 years. For crops not fertilized, the nitrogen balance after a decade was negative (-254 kg ha-1), implying an overall decline in soil nitrogen reserves. In crops treated with nitrogen at a rate of 120 kg ha-1 yr-1, biomass yield was not significantly different from unfertilized crops while the nitrogen balance was positive (790 kg ha-1 after 10 years) with 280% more nitrogen having leached from the system than the unfertilized treatment.

As stated by Cadoux et al. [90], fertilizer recommendations for Miscanthus will be a compromise between the needs of the crop and the need to maintain soil nutrient reserves while limiting nutrient losses. As the exact nutrient needs of Miscanthus are not yet known, Cadoux et al. [90] proposed that nutrient recommendations should be based on the amount of nutrients removed or likely to be removed at harvest (using expected yield and median concentrations given in Table 4.3). For nitrogen, they recommended no fertilization during the first two years of cultivation because nitrogen requirements are low at this stage while the risk of nitrate leaching is high [94].

While M. x giganteus shows only a small yield response to nitrogen input, some vari­ability in nitrogen response is likely among Miscanthus species, especially in relation to rhizome development and internal nutrient cycling. It would be valuable for breeding pur­poses to determine which progenitor species has contributed M. x giganteus’s efficient yield response to nitrogen.

Following the release of the 2005 BTS, a collaborative research team1 (Table 8.2) with mem­bers from the USDA-Agricultural Research Service (ARS) Renewable Energy Assessment Project (REAP) and several universities was established as part of the Sun Grant Regional Partnership (RP) to determine the amount of corn stover that could be harvested in a sus­tainable manner [31]. The core treatments included no tillage or the least amount possible for economic crop production [e. g. Coastal Plain soils near Florence, SC, have a naturally occurring hardpan (E horizon)], so in-row subsoiling is needed each year prior to planting], three residue removal rates (none, approximately half, and the maximum mechanically collectable amount), and four replications. Leveraging the Sun Grant Partnership funds with long-term ARS research expanded both the number of treatments being evaluated as well as the number of years of study. For example, at Mead, NE, the rainfed and irrigated studies were initiated in 1999 and 2001, respectively. At Morris, MN, the study was initiated in 2005, taking advantage of a tillage experiment established in 1995. At Ames, IA, two studies were initiated in 2005 and one in 2008. Additional management practices being evaluated at one or more of the locations include alternate tillage practices (e. g. chisel plow or strip-tillage), use of cover crops, rotation with soybean, harvesting of cover crops as well as the corn stover, and application of biochar.

For each experimental site, soil samples were collected to a depth of 1.0—1.5 m, divided into increments of 0-5, 5-15, 15-30, 30-60, 60-90 and 90-150 cm, and analyzed for several soil quality indicators [e. g. total organic carbon (TOC), total nitrogen, pH, bulk density, and soil-test phosphorus (P) and potassium (K)]. The Soil Management Assessment Framework (SMAF) was used to evaluate and combine the different indicators, and thus establish a baseline soil quality index that could be used to determine long-term effects of the various stover harvest rates [15]. To date, TOC and soil-test potassium have had the lowest indicator scores at several RP and other REAP sites [16]. Longer-term data leveraged from the REAP plots at Brookings showed that through the first eight years TOC decreased as residue removal rates increased (Figure 8.1). A more detailed examination of samples collected in 2008 showed higher organic carbon content in all aggregate size classes from the low removal treatment than in the high removal treatment (Figure 8.2). Higher total protein was also measured in soil samples from the low removal treatment than from the high removal treatment.

Whole plant samples were collected and fractionated into bottom, top, cob, and grain fractions. Plant parts lying on the ground within the sampling area (1.5 m[1] [2]) were also collected. Harvest index values and total nutrient uptake were collected using those samples. Stover was collected using a variety of mechanical harvesting techniques, all resulting in post-harvest soil surface cover differences, such as those shown for the Lamberton, MN,

site in the autumn (Figure 8.3) or the subsequent spring (Figure 8.4) following either conventional (chisel plow) or strip-tillage.

Additional data being collected at some but not all RP locations include greenhouse gas (GHG) emissions (CO2 and nitrous oxide, N2O), nitrate nitrogen (NO3-N) and phosphorus concentrations in water leaching through the soil profile, microbial biomass carbon, partic­ulate organic matter, glomalin-related soil proteins, the humic acid fraction of soil organic matter, aggregate stability, lignin, cellulose and other structural carbohydrates, and energy values for the various stover fractions. Collectively, these measurements are providing the data needed to develop the sustainable stover harvest strategies outlined through modeling in the 2011 BT2 report.

-Q

g 1.5

о

E?

Low cut — (> 4.5 t/ha)

High cut — (~3.4 t/ha)

Strip-Tillage Treatment

Conventional (Chisel Plow) Treatment

Strip-Tillage Treatment

Conventional (Chisel Plow) Treatment

One example (Figure 8.5) of the information being gathered shows the dependence of CO2 flux on soil temperature. The relatively strong logarithmic relationship suggests that a temperature-based interpolation method (Q10) will be most effective for estimating annual CO2 fluxes. These results also suggest that management practices which result in warmer soil temperatures, for example, through residue removal, may lead to higher CO2 fluxes. However, this effect will likely be offset by lower amount of available carbon substrate, that is, residue, so that the overall effect of stover harvest on annual CO2 flux will likely be a reduction in treatment differences.

With regard to N2O, Figure 8.6 shows that precipitation strongly influences the flux by reducing oxygen availability and stimulating denitrification. The lag between precipitation and maximum emission is evident, and is consistent with reports in the literature suggesting that the nitrous oxide flux is not maximized when the soil is saturated, but rather when water-filled pores space (WFPS) is about 65%. Annual sums of net N2O emission at this site were highest for the non-removal treatment and lowest for the maximum collectable treatment. They were also positively correlated with cumulative soil respiration, indicating that carbon availability was a controlling factor with respect to denitrification.

As recognized in the 2011 BT2 report, crop yield is a major driver associated with the availability of stover as a potential cellulosic bioenergy feedstock. Corn produces the highest volume of residue of all the major crops grown in the U. S.A. and because of the approximate 1:1 relationship between grain yield and aboveground biomass, the volume of

Figure 8.6 Nitrous oxide and rainfall relationships at the Rice County, MN, site in 2010. The "Ch. 1 to Ch. 6" designations simply refer to the six chambers used for the measurements. (Figure provided by John Baker, USDA-ARS).

available residue is directly proportional to grain yield (Table 8.3). To date, the RP studies have shown variable crop yield responses associated with stover harvest. This includes (1) no detectable short-term (3-year) effects at the Brookings, Florence, Morris, or University Park sites; (2) trends for increased yield when stover is harvested from no-till treatments at Ames and Mead; and (3) inconsistent site-differences at the Lamberton, Bauer Farm, and Rosemount sites in Minnesota. Another five-year assessment of stover removal effects near Ames, IA [16], showed that the most consistent grain yield response was a 21% lower average for continuous corn than for rotated corn. That study also showed that harvesting corn stover increased the average NPK removal by 29, 3 and 34 kg ha-1 for continuous corn and 42, 3, and 34 kg ha-1 for rotated corn, respectively, when compared to harvesting only the grain. Furthermore, it showed that the lower half of the corn plant contributed very little to the total available feedstock biomass because of its high water content and that it was not a desirable feedstock because of its high potassium, chloride, and ligin content, as well as an increased amount of soil contamination that interferes with both biochemical and thermochemical conversion processes.

So, what is the bottom line with regard to harvesting corn stover as a cellulosic feedstock? Firstly, producers must know their land. Prior to initiating any harvest strategy they should have good soil-test and nutrient management records for any areas from which crop residues may be harvested. Obviously, any land with erosion problems must be excluded and efforts should be made to use available stover in those areas to restore and rebuild the soil. Harvesting stubble will remove additional nutrients and could affect long-term soil organic matter levels, erosion rates, and water conservation. Producers should have and be using

long-term nutrient management and soil conservation plans. They should also be using the least amount of tillage possible. Again, avoid stover harvest from highly erosive areas and use routine soil-test and plant analyses to monitor the response on a routine basis. Finally, consider adopting other conservation practices, such as the inclusion of annual or perennial cover crops, buffer strips, and crop rotation, in order to enhance the sustainability of stover harvest.

John S. Cundiff

Department of Biological Systems Engineering, Virginia Tech, U. S.A.

12.3 Introduction

Society collects raw materials using the following extraction activities: mining (which includes drilling for oil and natural gas), logging, farming, fishing, and hunting. Mining is a “point source” extraction of raw materials. Some geological event has deposited a high concentration of the desired raw material at a given location, and a mine is established at this location. Typically, a rail line, pipe line, or marine port is built to cost-effectively transport the raw material to distant utilization points.

Biomass is unique among the raw materials used by society in that it is a distributed resource. It must be collected from production fields/forests and accumulated for processing at a central location. This chapter primarily covers farming with some minor reference to logging. Fishing and hunting (relatively small in developed countries) are not covered.

This chapter is written to guide the reader through the thought process they will use if they are designing, or specifying, a logistics system for a bioenergy plant. Logistics systems have been designed for many agricultural and forest products industries. Thus, it is wise to use the lessons learned in these commercial examples. Each of these industries faces a given set of constraints (length of harvest season, density of feedstock production within a given radius, bulk density of raw material, various storage options, quality changes during storage, etc.) and the logistics system was designed accordingly. Typically, none of the existing systems can be adopted in its entirety for a specific bioenergy plant at a specific location, but the key principles in their design are directly applicable.

Our definition of biomass logistics is “a series of unit operations that begin with biomass standing in the field and ends with a stream of size-reduced material entering a bioen­ergy plant for 24/7 operation”. (One example is shown in Figure 13.1.) All the required

Cellulosic Energy Cropping Systems, First Edition. Edited by Douglas L. Karlen. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd.

unit operations are linked together in this chapter. In general, it is not wise to end the analysis of a logistics chain when a truckload of raw biomass enters the plant gate, as is often done.

The term “feedstock logistics” is used to indicate that the focus is on the movement of raw biomass from the field to a bioenergy plant. This plant may be a “Regional Biomass Processing Depot” (RBPD) as outlined by Eranki et al. [1]. The depots (size range 100­1000 ton/d) are envisioned as locations where raw biomass is received and some prepro — cessing/pretreatment is done to create an intermediate product, or products, that are shipped to a large-scale biorefinery where the final processing is done. The term “biomass logistics” is reserved for the entire logistics chain from field to biorefinery, thus our use of the term “feedstock logistics” for the logistics from field to plant. Throughout this chapter, keep in mind that the “plant” can be an end user or a RBPD.

Managing weeds is one of the most important factors for sustainable switchgrass biomass production. Since switchgrass seedlings develop more slowly than annual weeds, control­ling weeds immediately after planting is critical for successful establishment. Additionally, the economic feasibility of switchgrass for bioenergy is dependent on establishing stands with a harvestable yield in the planting year [42]. Poor establishment caused by weed pressure can delay full production of biomass for two or more years [43]. Well-established switchgrass is less likely to have weed issues.

Weed pressure can be minimized in the establishment year by no-till seeding into glyphosate-tolerant soybean stubble, which provides an excellent seedbed. Switchgrass seed germination is slow and seedling vigor is low compared to annual grassy weeds. Consequently, it is important to plant high quality seed in properly prepared seedbeds [22]. If heavy weed pressure is expected, delay seeding until the first flush of weeds, then apply a broad-spectrum herbicide like glyphosate [N-(phosphonomethyl) glycine) before planting.

Applying pre-emergent and post-emergent herbicides shows a significant effectiveness in controlling and reducing weed populations during the establishment year. Normally, switchgrass establishment is not interrupted by broadleaf weeds. Herbicides, such as 2,4-D (2,4-dichlorophenoxyacteic acid) can control the broadleaf weed effectively, but should not be applied until after the switchgrass seedlings have reached the four — or five-leaf stage [21]. Early grassy weed growth is reduced with herbicides. The herbicides utilized in the forage industry will control weeds but label directions are critical for safe and proper application. The forage industry has used atrazine [2chloro-N-ethy-N/-(1-methylethyl)-1,3,5-triazine — 2,4-diamine] as a pre-emergent to control cool-season annual grasses and broad leaf weeds [44]. Quinclorac (3,7-dichloro-8-quinlinecarboxylic acid) is another common herbicide that can be used as a pre-emergent or post-emergent herbicide to control most of annual warm — season grass weeds [23,45]. Mitchell et al. [23] reported that a combination of atrazine and quinclorac applied immediately after planting provided the best weed control and most rapid establishment for upland and lowland switchgrass ecotypes in Nebraska, South Dakota, and North Dakota, U. S.A. Although some herbicides work well for establishing switchgrass, not all herbicides are labeled for application on the crop in all states. Always read and follow label directions.

Weed competition in post-establishment years is not a major issue for well-established stands, but if stands are poor during the establishment year, they typically have increased weed pressure in subsequent years. Switchgrass stands with seedling densities below 10 plants per square meter are considered to be poor and should be over-seeded or reseeded [21]. As long as adequate switchgrass frequency of occurrence (i. e., >40%) has been achieved in the seeding year, weed control is relatively easy during the post-establishment years. One of the most effective methods to control cool-season annual or perennial weeds is with the application of a broad-spectrum herbicide, such as glyphosate. Switchgrass must be dormant when glyphosate is applied, either prior to spring green-up or after senescence in late summer or early autumn. Once switchgrass starts to grow, canopy development is much faster than annual warm-season weeds and summer annual weeds are usually not an issue.