The North American grasslands contained hundreds of species of plants (Gould 1968; Weaver 1954). Sunflower (Helianthus annuus L.) is the only major cultivated crop that has been developed from a species native to the North American grasslands (Harlan 1975). Wild rice (Zizania aquatica L.) is from North America but it originates from marshy areas, not grasslands (Hitchcock 1951). The native grasslands and rangelands have and continue to provide food via meat production by native ruminants such as bison and introduced, domestic livestock for native North Americans and the subsequent European, African, and Asian immigrants who settled the continent. This report evaluates the potential of North American grasses to additionally benefit humans as new forage, grain, biomass, turf, and horticultural crops.
The initial evaluation and breeding work at the various experiment stations followed a similar pattern. A large array of accessions (ecotypes or strains) of various native grasses was collected from a general geographical area. These collections were then evaluated in uniform nurseries for various agronomic traits. The better accessions of the superior grasses were increased for testing in additional environments and based on these tests released directly to the public without any additional breeding work. This procedure was used by many of the state experiment stations and the U.S. Department of Agriculture in developing the initial grass cultivars for different geographical regions of the country (Barker et al. 1985; Jacobson et al. 1985; Hassell and Barker 1985; Vogel et al. 1985a; Voigt and Oaks 1985; Heizer and Hassell 1985). Examples of grass cultivars developed by this procedure include 'Blackwell' and 'Nebraska 28' switchgrass (Hanson 1972).
Although several hundred grasses are native to the United States, after several decades of testing, only a few have proven to have value as cultivated forage grasses. Cultivated forage grasses need to be easily established, persist under grazing, have high forage yields and good forage quality, possess good disease and insect resistance or tolerance, and produce adequate seed yields. Turf grasses need to have good turf quality instead of high forage yields and quality (Vogel et al. 1989). The native grasses that meet these criteria are primarily those which were subject to heavy grazing periodically by bison (Bison bison). Some of these grasses, which are primarily the principal species of the tall-, mid-, and shortgrass prairie, are currently being used as cultivated species and have potential to be used to an even greater extent.
Grasses can be classified as either cool-season (C3) or warm-season (C4) based on their photosynthetic mechanism (Waller and Lewis 1979). Cool season grasses produce most of their growth during the cool months of spring and fall and become essentially dormant during hot, summer months. Warm-season grasses grow most efficiently during the warm months of summer and usually are relatively unproductive under cool growing conditions.
Cool-season grasses that are widely used in the United States for forage and turf include such major species as Kentucky bluegrass (Poa pratensis L.), tall fescue (Festuca arundinaceae Schreb.), smooth bromegrass (Bromus inermis Leyss.), and wheatgrasses including Agropyron and Thinopyrum species (Barker and Kalton 1989; Meyer and Funk 1989). These are all introduced grasses from Europe, North Africa, and Asia where they evolved under centuries of intensive grazing by domestic livestock. The most widely utilized native cool-season grass is western wheatgrass [Pascopyrum smithii (Rydb.) Love]. It currently has minor use in revegetating agricultural lands in the Great Plains to which it is native. Other native cool-season grasses are utilized to a very minor extent and then usually because of legal or regulatory requirements that only native species can be used to revegetate specific sites. We believe that native cool-season grasses will have only limited use in the future as forage or turf grasses even if extensive breeding work on them is conducted. The introduced cool-season grasses can be significantly improved by breeding and the economic returns from the breeding effort would be greater than for native cool-season grasses.
The principal warm-season grasses that are being used in the southern regions of the United States are introduced grasses from Africa or Asia where they evolved under heavy grazing pressure from either wild or domestic animals. Important species are bermudagrass [Cynodon dactylon (L.) Pers], lovegrasses (Eragrostis spp.), old world bluestems (Bothriochloa spp.), and zoysiagrass (Zoysia spp.) (Burton 1989; Busey 1989). Most of these grasses cannot survive the winters in the central and northern latitudes of the United States and consequently, native warm-season grasses such as big bluestem (Andropogon gerardii), indiangrass (Sorghastrum nutans), and switchgrass (Panicum virgatum) have and are being increasingly utilized as cultivated grasses in these regions. These native warm-season grasses provide excellent pasture and forage during hot summer months when cool-season grasses are relatively unproductive and have poor forage quality.
Eastern gamagrass [Tripsacum dactyloides (L.) L.] may also have considerable potential. Its use in the past has been restricted by limited seed production. The recent discovery of a mutant that has been reported to increase seed yields by up to 20-fold (Dewald and Dayton 1985) should alleviate this problem. Eastern gamagrass is noted for high forage yields but its persistence under grazing and intensive management needs further testing. Other native grasses that are and will be utilized as cultivated grasses on a smaller scale include grama grasses (Bouteloua spp.) and little bluestem [Schizachyrium scoparium (Michx.) Nash].
There are millions of hectares in the Central Great Plains and adjacent Midwest states that need to be in permanent vegetation for erosion control. In the current Conservation Reserve Program (CRP), millions of hectares of this land was seeded to permanent vegetation. It has been predicted that over 80% of this land may be plowed when the CRP program terminates (Heimlich and Kula 1989). The production systems of the participating farmers were not adequately taken into consideration in the crash CRP program and many of the grasses and other plants used for revegetation will not be productive enough to be profitable for farmers and ranchers.
Marginal land currently in crop production often produces grain for meat production. The same land can be used to produce meat via livestock grazing in sustainable and environmentally benign integrated crop-grassland-livestock production systems if adequate plant materials are available. Research to develop sustainable crop production systems for erodible lands in this region have been initiated. An alternative approach would be the development of sustainable agricultural systems in which livestock and grasslands are integral components of the system.
Class I and II land (Klingebiel 1958) in a region can be used for grain crop production without serious erosion problems if properly managed. Land classified as having higher erosion potential should be in permanent grasslands which is the best and most cost effective method of reducing erosion to acceptable levels (Vogel et al. 1985b). Livestock are needed to utilize the forage produced by these grasslands to make the systems economically viable. Permanent grasslands could maintain livestock herds during the spring, summer, and fall months. Crop residues on the cultivated land and harvested hay would be used to maintain the livestock during the winter months. The components of this system would change from region to region. Native warm-season grasses such as big bluestem and switchgrass would provide the summer grazing component for this system in the Central and Northern Great Plains and Midwest.
One-half to one-third of the potential grasslands in Midwest and central and northern Great Plains region could be seeded to grasses such as big bluestem and switchgrass which would be approximately 5 to 10 million ha. In addition, approximately one-third of the grasslands in the fescue belt or about 5 million hectares could be converted to warm-season pastures to provide high quality grazing during the summer months by reseeding them with switchgrass and big bluestem. These grasses may also be used in the Northeastern states of the United States on marginal cropland.
Although forage yields vary with locations due to precipitation, growing season, and soil fertility, forage yields of 10 to 20 Mg/ha can be expected from switchgrass and big bluestem in the Midwest and Great Plains states. Animal gains per hectare will depend upon pasture and livestock management. Results from grazing trials indicate that the economic return from properly managed native warm-season pastures can be similar to that of grain crops grown on the same land (Table 1). The numbers of hectares of land that will be converted into grasslands seeded to native prairie grasses will be determined by economic factors including farm programs of the United States.
The potential benefits of perennial grain crops would include reduced soil erosion due to reduced tillage, reduced production costs, and reduced energy use. Jackson (1990) has developed the concept of producing perennial grains in polycultures in grain producing "prairies" in which nitrogen fixing legumes would be part of the system.
Numerous grass species have been tested for their potential as perennial grain crops including grasses native to the North America. At the present time the seed yields of these grasses are only about a tenth to a fifth of wheat (Triticum aestivum L.) or maize (Zea mays L.) grown on similar land (Jackson 1990; Wagoner 1990). The proponents of perennial grain crops believe that the seed yields of these perennials can be improved by breeding to the level where they are economically competitive with grain crops. Seed yields can undoubtedly be improved but it will take a considerable sustained effort to double let alone quadruple seed yields. In the meantime, the seed yields of existing annual grain crops will not remain static. Annual genetic gains for maize from 1950 to 1980 averaged 92 kg/ha (Duvick 1984) while for wheat they increased 0.74% per year for the period 1958 to 1980 (Schmidt 1984). The increase in wheat and maize yields were due to the combined efforts of many geneticists, pathologists, entomologists, and other scientists working throughout the United States. Scientific resources of this magnitude will not be available to develop perennial grasses into grain crops, so it is doubtful if increases in seed yields of potential perennial grain crops would be equal to those that are being achieved in corn and wheat.
Breeders and production agronomists working with native North American grasses initially faced considerable seed production problems that needed to be resolved before these grasses could even be utilized as forage crops. Empirical (Schumacher 1962) and formal research (Smika and Newell 1966; Canode 1965; Kassel et al. 1985) has resulted in greatly increased seed production. By using improved seed production practices, experienced seed growers in Nebraska can produce from 250 to over 1,000 kg of seed/ha under dryland conditions and up to 2,000 kg/ha under irrigated conditions (Nebraska Crop Improvement Association pers. commun.). Since only 6 to 14 kg of pure live seed/ha are needed for grassland plantings, one seed production hectare will plant 25 to 90 hectares of grasslands which is similar to that of most cultivated crops (Vogel et al. 1989). This means that native grasses have adequate seed yields for use as forage crops. They do not have adequate seed yields for use as perennial grain crops.
In addition to the problem of increasing seed yields there are also problems dealing with insect and disease pests. The problems that can occur with seed pests in perennial grass production fields indicate that this problem alone may make perennial grain crops impractical. A bromegrass seed midge and a big bluestem seed midge can reduce yields of these grasses by over 50% (Neiman and Manglitz 1972; Carter et al. 1988). The biology of the bromegrass seed midge has been investigated but less is known about the biology of the bluestem seed midge (Vogel and Manglitz 1989). There are no known controls for either insect. It is very likely that similar insects exist on other prairie grasses. Based on existing knowledge of the insect, it is likely that insecticides could be used to control the insect but to date no adequate controls are available. Boe et al. (1989) recently reported evidence that there may be genetic variation for infestation by the big bluestem seed midge. The genetic differences were small and a tremendous amount of long-term breeding work would be required to produce strains that had economically improved tolerance or resistance. Because the life cycle of the seed midges matches that of the host grasses, it is apparent that they co-evolved. These insects would undoubtedly be a serious problem if a perennial grass was grown on extensive areas of land as a seed crop.
New, improved forage cultivars are already economically competitive with annual grain crops in the Great Plains (Table 1). Using these grasses in integrated crop-livestock-grassland production systems would achieve the same goals of conserving soils advocated by the proponents of perennial grain crops (Jackson 1990) and these goals can be achieved now rather than in the distant future. Additional breeding work on grasses as forage crops will only improve the economics of integrated production systems.
In addition to the production and breeding problems associated with perennial grain crops, there is also the marketing and utilization problem. This problem can be best understood by evaluating the new grain crop triticale (xTriticosecale rimpaui Wittm.). This new crop has not achieved its promise because no specific national markets or market channels for it have developed. It is likely that the same fate would await a perennial grain crop. If advocates of perennial grain crops want to pursue the development of a perennial grain crop in spite of these obstacles, we suggest that Canada wildrye (Elymus canadensis L.) and its relatives would have more potential as a perennial grain crop than other prairie grasses.
Ethanol fuel production from biomass will be dependent upon the development of economical processes for converting the cellulose and hemicellulose in plant cell walls to extractable ethanol. Conversion technology has not reached this stage of development. We are currently evaluating switchgrass germplasm for its potential as a biomass fuel crop. This evaluation includes herbage yield, other agronomic traits, and the stability of these traits over three midwestern environments. We are unable to evaluate switchgrasses for ethanol conversion traits because development of the conversion process technology has not reached the stage at which the herbage traits most important for ethanol production can be characterized (Janet Cushman and Anthony Turhollow pers. commun. Oakridge National Laboratory). Since the important steps in the conversion process will involve biological reactions we assume that traits which improve in vitro dry matter digestibility (IVDMD) will be similar to those needed for ethanol production from biomass.
Ethanol production from herbage of perennial grasses has considerable promise and if the conversion technology can be developed, American farmers will be growing millions of acres of switchgrasses and other grasses for biomass fuel production by the year 2020. It has been estimated that ethanol could be produced from herbage using existing technology for $0.36/liter ($1.35/gallon) (Bull 1989). The use of native prairie grasses as biomass fuel crops will probably depend upon United States government energy programs and policies.
Many of the native grasses could be used as ornamental plants because of their striking appearance. Some landscape architects are currently using them in their planting plans and these grasses are beginning to be available in nurseries. No ornamental cultivars of these grasses have been released to date. The use of prairie grasses as ornamentals will increase because of their natural beauty and their low maintenance requirements, but the extent of use as ornamentals cannot be predicted at this time.
The genetic variation that exists within both native and introduced cross-pollinated grasses consists of between ecotype (endemic strain) or synthesized strain variability and within strain variability. Ecotypes or endemic strains found in specific regions and sites have evolved by the genetic mechanisms of mutation, migration, selection, and random drift or chance resulting in between ecotype or endemic strain genetic variability (Falconer 1981). Eberhardt and Newell (1959) documented this between strain variability in switchgrass and we are currently conducting similar studies using germplasm of switchgrass, big bluestem, indiangrass, and Canada wildrye collected from remnant midwestern prairies in 1989. Based upon data from the first evaluation year, the between accession genetic variability is substantial.
Most of the initial breeding work with cross-pollinated grasses utilized this between strain genetic variability (Vogel et al. 1985a; Hanson 1972). Within strain genetic variability consists of the proportion of the plant-to-plant (phenotypic) variability that exists between plants of a strain that is due to genetic (genotypic) differences among plants (Falconer 1981; Hallauer and Miranda 1981). This variability is very difficult to observe or measure in a typical pasture or rangeland situation. However, if seed is harvested from the individual plants in a common native prairie site and planted in a space-planted nursery under uniform conditions, phenotypic variation among plants can be readily distinguished. By using quantitative genetics procedures, it is possible to determine the total genetic variation and the additive genetic variation in specific populations or accessions for specific traits and the heritability of those traits.
The genetics studies that have been conducted in switchgrass, big bluestem, eastern gamagrass, and indiangrass indicate that there is substantial genetic variation for important agronomic traits studied to date (Eberhardt and Newell 1959; Newell and Eberhardt 1961; Talbert et al. 1983; Vogel et al. 1981a,b; Godshalk and Timothy 1988; Godshalk et al. 1986, 1988; Gabrielsen et al. 1990; Ross et al. 1975; Glewen and Vogel 1984; Riley 1981; Boe and Ross 1988; Boe et al. 1989; Wright et al. 1983). The conclusions from these studies can be summarized as follows: (1) there is substantial genetic variability both between and within strains of these grasses for most agronomic traits including those that affect or determine forage yield and quality; (2) heritability values for most important traits range from 20 to 40% which should make it possible to improve these grasses by breeding; and (3) correlations among most desirable traits are usually positive but when negative they are usually not large indicating it should be possible to simultaneously improve several traits without adversely affecting other traits.
Almost all important native prairie grasses are cross-pollinated by wind (Hanson and Carnahan 1956). They have small florets that are difficult to emasculate, and effective mechanisms for producing hybrids such as cytoplasmic male sterility have not been developed. Thus breeders are largely limited to procedures that utilize additive genetic variability and that do not require emasculation. Fortunately, there is substantial additive genetic variability for most traits in these grasses, and breeding methods that do not require emasculation are some of the most efficient that are available. The expected gain from selection that can be made by using the breeding procedures or schemes that have been developed to date are described by Empig et al. (1972), Nguyen and Sleper (1983), and Hallauer and Miranda (1981). The two breeding methods that have the most potential of exploiting additive genetic variation are Restricted Recurrent Phenotypic Selection or RRPS (Burton 1974, 1982) and a modified form of between and within family selection (Vogel 1988; Asstveit and Asstveit 1990). Since biomass production requires traits similar to those needed for forage crops, the potential of improving these grasses as biomass fuel crops appears to be excellent.
Anderson et al.(1988) demonstrated that an increase in forage digestibility of less than two percentage units in 'Trailblazer' switchgrass resulted in about a 20% improvement in animal gains and beef production per hectare as compared to the cultivar 'Pathfinder' which had similar yields. The increase in beef production due to the small change in digestibility had a value of $89/ha ($35/acre) averaged over three years (Vogel et al. 1989). In a grazing trial completed in 1990 at Mead, Nebraska, by K.P. Vogel, K.J. Moore, B.A. Anderson, and T.J. Klopfenstein, an experimental intermediate wheatgrass produced significantly higher gains than the two leading cultivars. Averaged over a two year period, this increased gain had a value of over $50/ha. These gains were made with only initial breeding work to improve forage quality. The potential for significant economic gains by breeding for improved forage grasses appears to be the best method to improve the economic value of forage grasses. Theoretical studies evaluating the effect of increasing forage quality on animal gains indicates that exponential gains in animal performance can be achieved by breeding for improved forage quality (Fig. 1).
|Beef yearling gains||350||290|
|Gross return ha-1|
|Fig. 1. Animal production increases exponentially with improvements in forage digestibility. Digestibile energy (DE) intake values are for 300 kg beef steer. The darker shaded area represents maintenance requirements.|