Paper presented at BioEnergy '98: Expanding Bioenergy Partnerships, Madison, Wisconsin, October 4-8, 1998.

ABSTRACT

USDA and University of Nebraska research on switchgrass was expanded in 1990 to include development of switchgrass into a biomass fuel crop. Initial research evaluated all available cultivars and elite strains at three Midwestern locations and identified cultivars and strains that had the most potential for use as biofuel crops. The highest yielding strains produced over 14,000 kg/ha biomass per year and over 10,000 kg/ha of cellulose + hemicellulose which would yield over 5000 l/ha ethanol (500 gal/acre) with a conversion effeciency of 75%. Susequent research has focused on specific breeding, genetics, and production problems. Optimum stage of growth and time period for harvest of switchgrass biomass in the Midwest was a three week period after plants were fully headed (about July 20 to August 15) and the optimum fertilization rate was 120 kg N/ha. It was determined that switchgrass is a micorrhizae (VAM) dependent plant, but inoculation will not be necessary because of VAM levels in agricultural soils. Information was developed that will allow a new herbicide to be registered for weed control during the establishment year. Genetic information to develop switchgrass hybrids was developed as well as the first molecular genetic markers for switchgrass. The breeding goal is to develop cultivars that can produce annual yields of 22 Mg/ha in the Midwest.

Keywords: Switchgrass, Panicum virgatum, breeding, genetics, management

INTRODUCTION

(Conversions are as follows: 2.24 Mg/ha=1 ton/acre; 1.12 kg/ha=1 lb/acre)

Switchgrass research, both breeding and management, has been conducted in cooperative U.S. Department of Agriculture and University of Nebraska research since the mid-1930's. This research produced the first switchgrass cultivar released in the United States, 'Nebraska 28' as well as several other switchgrass cultivars (Alderson and Sharp, 1994; Vogel et al. 1996). It has developed much of the basic information that is available today on switchgrass establishment, management as a forage crop, and breeding and genetics information (Moser and Vogel, 1995). In 1990, this long term program was expanded to include research on switchgrass as a biomass fuel crop. A successful crop production system requires reliable establishment information and technology, cultural and production information including optimal time of harvest and fertilization rates, and high yielding, adapted cultivars. The purpose of this report is to summarize the research that has been completed since 1990 to provide the required information and technology to develop switchgrass into a biomass fuel crop for the Midwestern states.

CULTIVARS and GERMPLASM

The first phase of this research, 1990-1992, evaluated all available cultivars and elite strains at three Midwestern locations, Mead, NE, Ames, IA, and West Lafayette, IN (Hopkins et al., 1995a). The research plots were seeded in 1990 and harvested in 1991 and 1992. They were fertilized with 112 kg/ha N. The only other cultural practices were the harvesting operations. The highest yielding strains produced 11 to 14.9 Mg/ha dry matter (Table 1). Despite significant genotype x environment interactions or changes in relative rank of strains across locations, some strains such as Cave-in-Rock and Shawnee consistently ranked high in biomass yields at all locations. Shawnee was developed from Cave-in-Rock by breeding for high forage digestibility (Vogel et el., 1996). These results indicate that it is possible to develop switchgrass cultivars with high biomass yields that are stable over broad geographical regions. These results also indicate that existing switchgrass cultivars have the potential to produce significant amounts of liquid fuels per hectare if adequate conversion technology is developed. About 75% of the biomass produced in these trials was cellulose and hemicellulose (Hopkins et al., 1995b). Assuming 75% conversion of the constituent cellulose and hemicellulose to ethanol, these yields would result in ethanol production of over 5000 l/ha (500 gallons/acre).

Switchgrass seed was collected in 1989 from remnant prairie sites in the North Central USA and evaluated in replicated space-planted trials at Mead, NE, Ames, IA, and West Lafayette, IN during the period 1990 to 1992. The objective of the trials was to determine the extent of genetic variation among native germplasm sources and their potential for improving the biomass yield of switchgrass in breeding programs. Across locations and years, significant variation among accessions was observed for biomass yield at heading and heading date (Hopkins et al, 1995b). Some accessions, such as IA34 and IL62 were comparable in biomass yield to check strains and should be useful genetic sources of variation for this trait (Table 2). Genotype x environment interactions were significant but some strains and accessions again consistently produced high yields over environments. Selection among and within accessions from different maturity zones for biomass yield should be an effective approach to utilize genetic variation in switchgrass from remnant prairie sites in the USA and Canada. High yielding cultivars such as Cave-in- Rock that contain significant genetic variation also are valuable germplasm sources for breeding switchgrass for increased biomass yields.

ESTABLISHMENT

Weed competition is a primary constraint to switchgrass establishment. Prior research has demonstrated that switchgrass establishment can be greatly enhanced and reduced seeding rates can be used if weeds can be controlled by herbicides during establishment (Martin et al., 1982; Vogel, 1987). Research is in progress to identify herbicides that can control annual grass and broadleaf weeds during establishment of switchgrass cultivars. Selected herbicides were applied to research plots near Clay Center and Lincoln, NE in 1996 within 5 days of planting Cave-In-Rock, Trailblazer, and Shawnee switchgrass. Switchgrass stand frequency (the number of sampling grids in which rooted switchgrass plants occurred and expressed as a percentage of the total number of grids evaluated) and herbage yield were measured the year after planting to determine establishment success (Tables 3 and 4).

Switchgrass establishment was usually improved by application of herbicides shortly after planting at both study sites. At Clay Center, stand frequencies were > 69% where herbicides were applied the year of planting compared to a stand frequency of 2% where no herbicide was applied (Table 3). There were some subtle differences in yield response of the three switchgrass cultivars to the herbicides at this site. Cultivar herbage yields ranged from 8.9 to 12.9 Mg/ha where the herbicides were applied and yields were <0.2 Mg/ha where no herbicide was applied (Table 3). At Lincoln, switchgrass cultivars responded in a similar manner to the herbicides. Switchgrass stand frequencies were > 55% where all herbicides, except isoxaflutole (BALANCE® ), were applied and were greater than the 34% switchgrass frequency that occurred where no herbicide was applied (Table 4). Switchgrass herbage yields were > 9.1 Mg/ha where most herbicides were applied and only 1.6 Mg/ha where no herbicide was applied (Table 4).

All the herbicides evaluated have the potential to be used during switchgrass establishment. Of the herbicides evaluated, only atrazine, imazapic (PLATEAU® ), and imazethapyr (PURSUIT® ) are registered for use during switchgrass establishment on roadsides and land enrolled in the Conservation Reserve Program. However, PLATEAU and PURSUIT may injure switchgrass seedlings if applied at > 0.07 kg active ingredient (ai)/ha. In this study, we found that PLATEAU applied at 0.035 kg ai/ha provided excellent weed control and negligible injury to switchgrass seedlings. Based on 1998 prices, PLATEAU at 0.035 kg ai/ha cost $10/ha, which is less than atrazine at 2.2 kg ai/ha ($16/ha) or PURSUIT at 0.07 kg ai/ha ($51/ha). Another advantage is that PLATEAU generally provides annual grass and broadleaf weed control that is superior to weed control with atrazine or PURSUIT.

HARVEST MANAGEMENT AND FERTILIZATION

Research to determine the optimum stage and time of harvest and N fertization rate was conducted using 'Cave-in-Rock' switchgrass plots planted in the spring of 1993 at Mead, NE and Ames, IA. No cultural practices or treatments were applied the establishment year other than the application of herbicides for weed control. In the fall of 1993, all forage was removed from the plots after a killing frost. In the spring of 1994, soil samples were taken at 0.3m increments to a depth of 1.5m at both locations. Six N treatments were applied (0, 60, 120, 180, 240, 300 kg/ha N) at both locations the last week of May, 1994. Harvests were made at seven weekly intervals beginning 28 June, 1994 at Ames and 15 July at Mead, NE. Differences in starting date was due to differences in spring growth. A final harvest was made after a killing frost at each location. Regrowth on the previously harvested plots was made at the same time as the final harvest. The same procedures were used in 1995. The first harvest in 1995 at Ames and Mead were June 29 and 30, respectively. The experiment design was a randomized complete block split-plot design with fertilizer treatments as main plots and cutting date as subplots. Each treatment was replicated four times at each location. Samples for biomass were collected from each plot at each harvest.

The results were similar for both locations so only the results for the Ames, IA location are presented (Figure 1). As expected, there were significant differences among both nitrogen fertilization rates and harvest dates at both locations. The highest yields were obtained once the plants had headed and harvest during a 4 week harvest period after heading produced similar yields. Delaying harvest until after a killing frost, significantly reduced biomass yields over all fertilizer levels at both locations. The results indicate that N fertilizer rates of 120 to 180 kg/ha are needed to obtain optimum yields depending upon year and site. At nitrogen rates of 120 kg/ha, about all the applied N is removed with the biomass. At higher fertilization rates, N began to accumulate in the rooting zone. In some years depending on the price of switchgrass biomass, adequate regrowth would be produced to warrant a second harvest.

MICORRHIZAE

The objectives of this research were to

1. evaluate the level of dependence of switchgrass on vesicular-arbuscular mycorrhizae for nutrient uptake;
2. evaluate the effectiveness of different VAM populations to enhance switchgrass growth and nutrient uptake; and
3. determine whether switchgrass establishment and production can be enhanced by inoculating soils with more effective VAM isolates.

Switchgrass roots and rhizosphere soil were collected from native and seeded stands at 14 sites in six states. The effectiveness of the mycorrhizal populations in these soils to enhance switchgrass growth and nutrient uptake was evaluated in two greenhouse studies (Brejda et al, 1999). The cultivars Alamo, Cave-in-Rock, Kanlow, and Trailblazer were used in the greenhouse tests. Three-week old seedlings were transplanted into pots containing sand that had been innoculated with the inoculum sources or treated with an uninoculated control. All pots were fertilized weekly with a complete nutrient solution for 10 weeks and then harvested two weeks later. Averaged over 2 experiments, 14 inoculum sources and four cultivars, seedlings grown in inoculated pots had average shoot weights that ranged from 1.9 to 4.3 g/plot (3 seedlings per pot) while seedlings grown in uninoculated pots weighed only 0.4g/pot. Root weights and N and P uptake comparisons were similar clearly demonstrating the VAM dependence of switchgrass.

A field study was conducted to determine if VAM inoculation of seedlings with two different VAM sources had a significant effect on switchgrass growth in the establishment and post-establishment years. 'Shawnee' and 'Trailblazer' switchgrass seedlings were inoculated with highly effective, moderately effective, or indigenous populations of VAM and associated rhizosphere microflora, and grown in cone-tainers for 12 wk in a greenhouse prior to transplanting in a Sharpsburg silty clay loam (fine montmorillonitic mesic typic Argiudoll) and an Ortello loam (coarse loamy mixed mesic udic Haplustoll) soil near Mead, NE. Indigenous populations of AMF and associated rhizosphere microflora were obtained from the same fields collected prior to the start of the study. Each plot consisted of four rows of 12 plants each, with 30 cm between plants. Biomass was harvested after a killing frost the establishment year and after plants were headed the year after establishment (Brejda, 1996). There were no significant differences in switchgrass biomass yield or N or P uptake among the introduced and indigenenous VAM treatments at either field site. These results demonstrate that although switchgrass is micorrhizae dependent, indigenous micorrhizae in agricultural fields are effective with switchgrass and are apparently very competitive with applied micorrhizae. Micorrhizae inoculation of fields for switchgrass biomass production probably will not be needed.

BREEDING AND GENETICS

A series of genetic studies have been completed to develop basic and applied information that can be used to improve breeding effeciency, germplasm utilization, and potentially develop hybrid cultivars. Molecular genetics research with switchgrass chloroplast RFLP's determined that there are two distinct types of switchgrass based on chloroplast RFLP's and these two types are directly associated with the lowland and upland ecotypes (Hultquist et al., 1996). The cytotypes associated with the upland and lowland cytotypes have been designated the "U" and "L" cytotypes, respectively. Evaluation of germplasm collected from midwestern remnant prairies indicated that most of the Midwestern germplasm is the U cytotype but at one prairie site, both the U and L cytotypes were identified (Hultquist et al., 1997). Nuclear DNA content of most available switchgrass germplasms and cultivars has been determined using flow cytometry. Cytogenetic and flow cytometry research conclusively demonstrated that tetraploid strains have DNA contents of approximately 3pg/nuclei and octaploid strains have 6 pg/nuclei (Hopkins et al., 1996). The ploidy level of all available switchgrass cultivars and germplasms was subsequently classified using flow cytometry results (Hultquist et al, 1996: Hopkins et al., 1996). The flow cytometry research indicates that many remnant prairie sites in the Midwestern states contain both tetraploid and octaploid plants that may be separate breeding populations (Hultquist et al., 1997). An improved procedure for making controlled pollinations between switchgrass plants has been developed (Martinez-Reyna and Vogel, 1998). This procedure was used to make controlled crosses among lowland and upland switchgrasses and within ploidy level crosses. The research has demonstrated that tetraploid lowland and upland cytotypes are cross-fertile (Martinez-Reyna, 1998). These crosses and other crosses have provided clear evidence that a self-and cross-incompatibility system is present in switchgrass (Martinez-Reyna, 1998). Self-incompatibility will be needed to produce F1 switchgrass hybrid cultivars. In 1997, the progeny of reciprocal upland x lowland tetraploid were evaluated with RFLP probes to determine the inheritance of cytoplasm DNA in switchgrass. It was determined that chloroplast DNA in switchgrass upland and lowland crosses is maternally inherited (Martinez-Reyna, 1998). It was also determined that normally pairing occurs in meiosis of upland and lowland tetraploid hybrids indicating that they have the same genome (Martinez-Reyna, 1998).

All switchgrass cultivars developed to date are synthetic varieties that have been developed by using population improvement or other breeding procedures that utilize primarily additive genetic variation (Moser and Vogel, 1995). These varieties have considerable within variety genetic variation which enables them to be stable or “buffered” across environments and years. Non-additive genetic variation or hybrid vigor is not utilized with synthetic cultivars. It may be feasible to develop hybrid switchgrass cultivars by using the self-incompatibility system in switchgrass. Research is currently in progress to determine if heterosis exists in switchgrass for agronomic traits including biomass yield and to evaluate the feasibility of producing hybrid seed of switchgrass using transplanted seed production fields. If the same levels of heterosis exist in switchgrass as in other cross-pollinated species, it may be possible to significantly increase biomass yields using hybrid cultivars.

REFERENCES

Alderson, J., and W.C. Sharp. 1994. Grass varieties in the United States. USDA-SCS Agric. Handb. 170. U.S. Gov. Print. Office, Washington, DC.

Brejda, John J. 1996. Evaluation of arbuscular mycorrhiza populations for enhancing switchgrass yield and nutrient uptake. Ph.D. Dissertation. University of Nebraska-Lincoln, Lincoln, NE. 160p.

Brejda, John J., L.E. Moser, and K.P. Vogel, 1999. Evaluation of switchgrass rhizosphere microflora for enhancing yield and nutrient uptake. Agron. J. (In Press).

Hopkins, A.A., K.P. Vogel, K.J. Moore, K.D. Johnson, and I.T. Carlson. 1995a. Genetic effects and genotype by environment interactions for traits of elite switchgrass populations. Crop Sci. 35:125-132.

Hopkins, A.A., K.P. Vogel, K.J. Moore, K.D. Johnson, and I.T. Carlson. 1995b. Genetic variability and genotype x environment interactions among switchgrass accessions from the Midwestern USA. Crop Sci. 35: 565-571.

Hopkins, Andrew A., Charles M. Taliaferro, Christopher D. Murphy, and D'Ann Christian. 1996. Chromosome number and nuclear DNA content of several switchgrass populations. Crop Sci. 36:1192-1195.

Hultquist, Sherry J., K.P. Vogel, D.J. Lee, K. Arumuganathan, and S. Kaeppler. 1996. Chloroplast DNA and nuclear DNA content variations among cultivars of switchgrass, Panicum virgatum L. Crop Sci. 36:1049-1052.

Hultquist, Sherry J., K.P. Vogel, D.E. Lee, K. Arumuganathan, and S. Kaeppler. 1997. Nuclear DNA content and chloroplast DNA polymorphisms among accessions of Panicum virgatum L. from remnant midwestern prairies. Crop Sci. 37:595-598. Martin, A. R., R. S. Moomaw, and K. P. Vogel. 1982. Warm- season grass establishment with atrazine. Agron. J. 74:916- 920. Martinez-Reyna, Juan M. 1998. Hybridization between upland and lowland and within upland cytotypes of switchgrass. Ph.D. Dissertation, University of Nebraska, Lincoln, NE.

Martinez-Reyna, Juan M. and K.P. Vogel. 1998 Controlled hybridization technique for switchgrass.. Crop Sci. 38:876:878. Moser, L.E. and K.P. Vogel. 1995. Switchgrass, big bluestem, and indiangrass. Chapter 32. p. 409-420.

In: Robert F. Barnes, Darrell A. Miller, and C. Jerry Nelson (eds.). Forages, 5th ed. Vol.I: An introduction to grassland agriculture. Iowa State Univ. Press., Ames, IA.

Vogel, K.P., A.A. Hopkins, K.J. Moore, K.D. Johnson, and I.T. Carlson. 1996. Registration of 'Shawnee' switchgrass. Crop Sci. 36:1713.

Vogel, K.P. 1987. Seeding rates for establishing big bluestem and switchgrass with pre-emergence atrazine applications. Agron. J. 79:509-512.