Index | Search | Home | Table of Contents

Asay, K.H. 1993. Plant exploration for new forage grasses. p. 147-154. In: J. Janick and J.E. Simon (eds.), New crops. Wiley, New York.

Plant Exploration for New Forage Grasses*

K.H. Asay


The first and probably most important phase of any plant breeding program is to assemble the genetic resources from which breeding populations can be developed. The most elaborate facilities and selection procedures will not compensate for an inadequate germplasm base. Plant exploration has been particularly instrumental in providing the genetic diversity for breeding programs involving forage and turf grasses. Most important temperate grasses in North America were introduced from other continents. Plant breeders and other scientists working to improve these species must, therefore, rely heavily on germplasm collected from their native habitats or centers of diversity. Although the principal objective of most plant exploration expeditions is to provide germplasm resources for existing breeding programs, insight must also be used to identify new species or those that have not been evaluated on this continent.


Introduction of forage grasses in North America began in earnest with the arrival of the first colonists. Seed of many species were carried to the new world in the ballasts of ships or in livestock feed. Introductions also were made intentionally by settlers and by plant explorers such as N.E. Hansen and F.N. Meyer (Asay 1991). Germplasm from these early introductions remains an important component of the National Plant Germplasm System (NPGS) and is included in the parentage of several modern cultivars. The number of accessions included in the NPGS sheds some light on the relative impact of plant exploration on important grass genera. For example in 1989, 1,928 accessions of Panicum and 1,612 accessions of Festuca were included (GRIN 1989).

Several instances could be cited to illustrate the impact of plant exploration for forage grasses on American agriculture. Tall fescue (Festuca arundinacea Schreb.) in North America began with the release of two cultivars, 'Ky-31' and 'Alta'. Although the exact origin of 'Ky-31' is somewhat obscure, the parentage of both cultivars trace to European introductions. Parental germplasm for 'Ky-31' was obtained from a naturalized ecotype found growing on a hillside on the W.M. Suiter farm in Menifee County, Kentucky. This ecotype had apparently been established on this site since 1887. Under the direction of E.N. Fergus, seed was collected and after extensive evaluation, the cultivar was released by the Kentucky Agriculture Experiment Station in 1931 (Buckner et al. 1979; Hanson 1972). The tall fescue cultivar 'Alta', which was developed in cooperation with the USDA/ARS and the Oregon Agricultural Experiment Station, traces to three introductions from Germany (PI 19728, 24838, and 25206). Subsequent evaluation and selection led to the release of the cultivar in 1940 (Hanson 1972).

Since the release of these cultivars, the germplasm base of tall fescue has been expanded through plant exploration and hybridization (Asay et al. 1979) and it has become the predominant grass in the United States. It is estimated that the species occupies from 12 to 14 million ha in pure and mixed stands (Buckner 1985) and 44 cultivars have been registered by the Crop Science Society of America (CSSA) [Germplasm Resources Information Network (GRIN) 1991]. The turf potential of tall fescue is now being exploited by plant breeders. Selections have been made from products of plant exploration that are lower growing and have finer leaves and more dense tillers than typical tall fescue. Although the currently available tall fescue cultivars are somewhat coarser than Kentucky bluegrass, they have demonstrated more resistance to insects and diseases than typical Kentucky bluegrass cultivars.

The identification of an endophyte (Acremonium coenophialum Morgan-Jones & Gams) in tall fescue presents some interesting alternatives in plant exploration and breeding. From a negative perspective, the presence of the endophyte is associated with serious toxicity problems in livestock consuming the forage. On the other hand, alkaloids produced by the endophyte contribute to the vegetative vigor and resistance of the plant to pests and environmental stress (Bacon and Siegel 1988; Read and Camp 1986). Future plant exploration may contribute to research leading to the identification of an endophyte that will positively influence plant growth and persistence without the deleterious side effects.

Orchardgrass (Dactylis glomerata L.) was introduced over 200 years ago from central and western Europe, but its value was not fully recognized in the United States until about 1940. This productive and nutritious grass was first cultivated in Virginia and is now one of our most widely accepted forage grasses. According to GRIN (1991), 12 cultivars have been registered by CSSA. In his review of the parentage of 29 orchardgrass cultivars, Hanson (1972) found that 12 were developed from introduced accessions or PIs and five were introduced as cultivars developed in foreign countries. The remaining cultivars were derived from selections from existing cultivars, old pastures, and naturalized strains (Hanson 1972).

Smooth bromegrass (Bromus inermis Leyss.) was apparently first introduced into North America from Hungary in the late 1800s and was known early as Hungarian bromegrass. In 1898, N.E. Hansen obtained seed from the Penza Region of the USSR, which led to the predominance of the northern strains in the United States and Canada. Following the devastating drought of the 1930s, smooth bromegrass became a popular grass in the Midwest. Since then, naturalized southern strains and cultivars developed in breeding programs have become an important component of American seed trade (Carlson and Newell 1985). The cultivar 'Lincoln', which was derived from selections made from a productive stand in Nebraska, was the early standard (Barker and Kalton 1989) and according to GRIN, 17 additional cultivars have since been registered with CSSA.

Because it is so well adapted and widely distributed in the temperate regions of North America, Kentucky bluegrass (Poa pratensis L.) is often considered to be a native species. It is now known that this multipurpose grass is indigenous to Europe where it is called smooth meadow grass. It has been suggested that Kentucky bluegrass was introduced into North America through Labrador or Alaska; however, this has not been documented. The species was probably first introduced along the Atlantic Coast by the early immigrants and transported inland during the westward migration or indirectly by grazing animals. Plant exploration coupled with public and private breeding programs have led to development of several cultivars beginning with the release of 'Merion' in the early 1930s. Forty-five cultivars have been registered by CSSA (GRIN 1991) and the species is now found as a component of pastures in over 16 million ha and in some 40 million lawns throughout the United States and Canada (Duell 1985).

Like Kentucky bluegrass, Timothy (Phleum pratense L.) probably crossed the Atlantic with the early settlers in hay litter or ballast from ships. The colonists were convinced that the grass was native to North America because it was so well adapted; however, it was later shown to be introduced from Europe. It was originally called Herd grass after John Herd who is thought to have found it along the Piscataqua River near Portsmouth, New Hampshire about 1911 (Childers and Hanson 1985). The ultimate name sake for Timothy was Timothy Hanson, who promoted its use in Maryland, North Carolina, and Virginia. The name was first recorded in a letter from Benjamin Franklin to Jared Eliot in 1747 stating that the Herd grass he had received was "mere Timothy." The grass was so well accepted that in the early 1800s, it was considered to be the most important hay grass in America (Hoover et al. 1948; Childers and Hanson 1985). The genetic base of Timothy has since increased through plant exploration and in 1989, 573 accessions were included in the NPGS (GRIN 1989).

Crested wheatgrass (Agropyron cristatum, L., Gaertner and Agropyron desertorum, Fisch. ex Link, Schultes) has had a major impact on American grasslands. Since its introduction from USSR, this versatile grass has become the most widely used grass for revegetating depleted rangelands in western North America. During the "dust bowl" era of the middle 1930s, it was particularly instrumental in stabilizing abandoned wheatlands in the Northern Great Plains (Lorenz 1983).

Crested wheatgrass was first introduced into North America in 1898 by N.E. Hansen of the South Dakota Agricultural Experiment Station following an exploration trip to Russia and Siberia (Dillman 1946). He observed the grass in evaluation trials at the Valuiki Experiment Station on the Volga River near what is now Volgograd. Seed of five accessions were obtained and assigned PI numbers 835, 837, 838, 1010, and 1012. Seed from these accessions was distributed to Agricultural Experiment Stations in Alabama, Indiana, Michigan, Colorado, and Washington, and apparently to an experiment station in Highmore, South Dakota. No record is available regarding further increase or distribution of these introductions (Dillman 1946; Lorenz 1983).

In 1906, N.E. Hansen made a second importation of crested wheatgrass. This seed, which was obtained from the same source as the original introductions, consisted of five lots labeled Agropyron desertorum (Fisch.) Schult. These lots were assigned PI numbers 19537-19541. An additional lot, labeled Agropyron cristatum (L.) Gaertn. was assigned PI number 19536. Seed from one or more of these introductions was distributed to 15 experiment stations from 1907 to 1913. The greatest enthusiasm for crested wheatgrass was most evident in the northern Great Plains, particularly South and North Dakota. Nurseries established at the Belle Fourche Experiment Station, Newell, South Dakota, the Northern Great Plains Field Station, Mandan, North Dakota, and the Dickinson Substation, Dickinson, North Dakota were particularly noteworthy in the early evaluation and distribution of crested wheatgrass in North America. The planting of PI 19538 at Mandan, North Dakota has been maintained and is still productive.

Other introductions have since contributed to the crested wheatgrass gene pool in North America. In 1910, N.E. Hansen first introduced Siberian wheatgrass (Agropyron fragile, Roth, Candargy), which is native to the dry steppes of western European Russian and western Siberia. Several other explorations have since been made to USSR, China, Turkey, Iran, and other Asian countries (Lorenz 1983). More recently, D.R. Dewey and co-workers at the USDA/ARS Forage and Range Research Laboratory, Logan, Utah, have been actively involved in exploration for crested wheatgrass germplasm. Since 1972, 12 expeditions have been conducted by members of this research group to Europe and Asia to collect germplasm of crested wheatgrass and other species of interest.

The first documented introduction of crested wheatgrass into Canada occurred in 1911. Seed of PI 19536 (A. cristatum) and 19540 (A. desertorum) was received by John Bracken of the University of Saskatchewan at Saskatoon (Lorenz 1983; Rogler 1960). Nurseries were established from this seed in 1916 by L.E. Kirk, then a graduate student at the University. The cultivar 'Fairway', which has since become an important component of the Canadian grass seed trade, was derived from selected plants of PI 19536 in these nurseries. The release of 'Fairway' was delayed until 1927 by a fire that destroyed the building where the original "breeder" seed was stored. Fortunately, the parental clones were still in the field and new seed was produced the following season (Lorenz 1983). 'Fairway' was the first cultivar of crested wheatgrass to be released in North America (Elliott and Bolton 1970).

Cytological studies have determined that crested wheatgrass is essentially an autoploid series of diploid (2n = 2x = 14), tetraploid (2n = 4x = 28), and hexaploid (2n = 6x = 42) forms (Dewey 1966). The diploids are represented by the cultivar 'Fairway' and cultivars subsequently released, 'Parkway' and 'Ruff' (Hanson 1972; Asay and Knowles 1985). Prominent tetraploid cultivars include 'Nordan', 'Hycrest', 'Ephraim', and 'P-27'. 'Nordan' was developed by the USDA/ARS Northern Great Plains Research Center at Mandan, North Dakota from plants in an old seeding at Dickinson, North Dakota. It has been the dominant cultivar in the United States. 'Hycrest' was released in 1984 by the USDA/ARS in cooperation with the Utah Agricultural Experiment Station and the USDA/SCS. It was derived through hybridization between induced tetraploid A. cristatum and natural tetraploid A. desertorum (Asay et al. 1985b). This cultivar has demonstrated superior establishment characteristics on harsh range sites and is rapidly becoming a major component in seeding mixtures, particularly in the Intermountain Region (Asay et al. 1986).

Plant materials recently obtained through plant exploration have contributed to breeding efforts in crested wheatgrass. Species of the complex are normally caespitose (bunch type); however, accessions recently received from Turkey, Iran, and China develop extensive rhizomes. These plants also are shorter in stature and have finer leaves and greater tiller density than typical crested wheatgrass. Progeny lines selected from these populations have been entered in a breeding program to intensify these characteristics. The research objective is to develop cultivars that are adapted for lawns, along roadsides, soil stabilization, and similar applications in water-limited environments and other areas where water conservation is a major concern (Asay 1991).

Crested wheatgrass has been criticized for the rapid decline in the quality of its forage as plants approach maturity during the summer. Soon after anthesis, leaves normally wilt and die back leaving a preponderance of stems until later in the season when new tillers are developed. Promising hexaploid accessions, recently obtained from Kazakhstan in the USSR, may provide the genetic resources to alleviate this concern. One particular accession, designated in the NPGS as PI 406442, has exceptionally broad leaves. In addition, leaves of this accession remain on the plant and retain their green color longer in the growing season than typical crested wheatgrass. This hexaploid accession also has larger seeds and seedling vigor advantages that are often associated with this trait. Hexaploid breeding populations have been established from selections within the Soviet accession and hybridization with other hexaploids obtained from Iran.

A breeding program also has been initiated to combine the leafiness and seedling vigor attributes of the Soviet hexaploid accession with the positive attributes of tetraploid 'Hycrest'. Hybrids between the two ploidy levels were readily obtained and the pentaploid progenies were relatively fertile. Moreover, the broadleaf character was easily detected in hybrid plants. Pentaploid hybrids, selected largely on the basis of leaf width and length were crossed among themselves and successively backcrossed to Hycrest clones. Relatively fertile and genetically stable tetraploid and hexaploid forms have been identified in these breeding populations, indicating that it is feasible to combine the genetic resources from the three ploidy levels in crested wheatgrass at either the tetraploid or hexaploid level. It is also evident that interploidy breeding schemes are valuable tools for more effectively utilizing genetic resources obtained through plant exploration to improve the forage quality, seedling vigor, and other characteristics of this valuable range grass.

Russian wildrye [Psathyrostachys juncea (Fisch.) Nevski] is a cool-season perennial grass that is native to the steppe and desert regions of USSR and China. The species was introduced into the United States in 1927, but its value in reseeding depleted rangelands was not fully recognized until about 25 years later (Hanson 1972). Although it is particularly noted for its productivity of palatable and nutritious forage during the spring and early summer, its nutritive value is retained better during the late summer than many other cool-season grasses. Russian wildrye has been widely used in rangeland seeding programs; however, its acceptance has been somewhat impeded by problems associated with seedling establishment, particularly on harsh range sites. Accordingly, improved seedling vigor is a major objective of breeding programs with this species (Asay and Johnson 1980; Berdahl and Barker 1984; Lawrence 1979).

Plant exploration and associated breeding programs have led to the release of several Russian wildrye cultivars. The cultivar 'Vinall' set the early standard in the United States. It was developed by the USDA/ARS at Mandan, North Dakota from collections made in the USSR (PIs 75737, 108496, and 111549) and released in 1960 (Hanson 1972). In 1978, 'Swift' was released by Agriculture Canada at Swift Current, Saskatchewan. Improved establishment vigor was emphasized during its development. More recent releases include 'Bozoisky-Select' by the USDA/ARS at Logan, Utah (Asay et al. 1985a) and 'Mankota' by the USDA/ARS at Mandan, North Dakota (Berdahl and Barker 1991b). 'Bozoisky-Select' was derived from PI 440627, an introduction from the USSR. It is significantly more robust and productive than 'Vinall' in the seedling as well as the more advanced growth stages. It is rapidly establishing itself as the dominant Russian wildrye cultivar in the Intermountain West. The parental germplasm for Mankota was obtained from PIs 314675 and 272136 and an experimental breeding population. Its performance has been superior to other Russian wildrye cultivars adapted to the northern Great Plains.

Russian wildrye is normally a diploid; however, promising tetraploid accessions were obtained from Soviet scientists during a recent plant exploration in the republic of Kazakhstan. Tetraploid forms are typically characterized by larger seeds, better seedling vigor, and a more robust growth habit than their diploid counterparts (Berdahl and Barker 1991a; Lawrence et al. 1990). Breeding populations have been developed from these accessions through selection and hybridization with induced tetraploids from promising diploid cultivars.


Exploration for grasses native to North America has provided valuable germplasm to the NPGS. There is justifiable concern that in some instances, we may be in danger of losing valuable native germplasm to activities associated with population growth as well as industrial and agricultural development. Grass breeders, primarily in the public sector, have exploited genetic diversity in native populations and several improved cultivars have been released. A crusade exists, particularly in the Intermountain West, to use only native species in rangeland seeding mixtures. Although the controversy centers on public lands, other areas are affected as well. Multiple demands including those imposed by livestock, wildlife, and recreation, have significantly altered the environmental forces in the plant community. These environmental changes have a profound influence on the optimum vegetative climax associated with a particular range site. It is therefore, not always in the interest of good land management to restore the vegetative ecosystem to its native state or to what we presume it to have been a few hundred years ago. New combinations of native and introduced plant species will be essential if we are to enjoy the maximum benefits of our natural resources. It is regrettable that decisions of this nature are often made in the political arena, with little consideration of the complex biological interactions involved.

Introduced germplasm has had and will continue to have a major impact on American agriculture. Grasses indigenous to Asia, Europe, and Africa have for the most part been subjected to more intense grazing pressure than our native species. It is not surprising that natural selection under such conditions would generate germplasm that is better adapted to this type of management even when transported to another continent. This is not to imply that native species do not have a role to play in rangeland improvement. They most assuredly do, but we must utilize other sources as well. Adaptability to the environment or intended use should be a more valid criterion for evaluating plant materials than their nationality.


Many of our native temperate grasses are closely related to European and Asian species. These introduced relatives can be valuable genetic donors to their native counterparts. Genetic introgression from introduced germplasm has been used to improve bluebunch wheatgrass [Pseudoroegneria spicata (Pursh) Löve]. Bluebunch wheatgrass is a cool-season native rangegrass. This caespitose species is drought resistant and produces nutritious forage; however, it is selectively grazed in mixtures with other species and stands are often depleted under heavy grazing (Hafenrichter et al. 1968; Mueggler 1975). Quackgrass [Elytrigia repens (L.) Nevski] has proven to be a valuable genetic donor in crosses with bluebunch wheatgrass in the USDA/ARS research program at Logan, Utah. The positive contribution of quackgrass in any breeding program may be surprising, as this aggressive species is often considered to be a noxious and troublesome weed. However, it has many desirable attributes and is considered to be a valuable forage in many temperate regions of the world. It is a productive, long-lived perennial grass, with moderate salinity tolerance, and because of its extensive rhizome development, it has excellent soil-binding characteristics.

The F1 hybrid between hexaploid (2n = 42) quackgrass and the tetraploid form (2n = 28) of bluebunch wheatgrass was disappointing (Dewey 1967). It was a pentaploid (2n = 35), meiotically irregular, and largely sterile. The hybrid also was plagued with deleterious traits and in general had poor vegetative vigor. Selected plants from the hybrid population were included in a breeding program in 1974 with the objective of combining the caespitose growth habit, drought resistance, and forage quality of bluebunch wheatgrass with the persistence, durability, productivity, and salinity tolerance of quackgrass. Eight generations after the initial cross, a population (designated as RS hybrid) with relatively good fertility and a stable chromosome number of 2n = 42 was obtained. Characteristics of both parental species were evident in the population, which was released as the cultivar 'NewHy' in 1989 (Asay et al. 1991). Other introduced relatives such as Pseudoroegneria stipifolia (Czern. ex Nevski) may be valuable sources of genetic diversity for improving bluebunch wheatgrass.

A promising accession of quackgrass (Elytrigia repens) was recently collected in Turkey. Most plants in this collection produced extensive rhizomes; however, genetic segregation for the caespitose growth was also evident. A recurrent selection program was initiated with this population and after four cycles, a caespitose form of quackgrass was obtained. This breeding population is productive and leafy, and has excellent salinity tolerance. The original parental germplasm may trace to a hybrid between quackgrass and an Asian form of bluebunch wheatgrass. As a result of plant exploration, a much better alternative is now available for revegetation of saline sites on semiarid rangelands.

Hybridization between introduced and native Leymus (wildrye) species is also a promising breeding approach. Leymus germplasm recently collected in the Soviet Union has been hybridized with a native relative, Great Basin wildrye [Leymus cinereus (Scrib. & Merr.) Löve]. The hybrid population is extremely robust and appears to be a promising source of germplasm for extending the grazing season during the late fall and winter on temperate rangeland.


Although this review is concerned primarily with temperate grasses, it is evident that plant exploration has contributed to the improvement of warm-season grasses as well. Warm-season grasses can be divided into two groups, western and southern, based on their adaptation to soil and climatic factors. Pasture improvement in the southern states has relied heavily on introduced species, whereas in the West, several native warm-season species have merited substantial breeding effort (Burton 1989).

Domestic exploration and subsequent breeding has led to significant genetic advance in switchgrass (Panicum virgatum L.). The cultivar 'Pathfinder' was developed from collections made in the Midwest and released in 1967 by the USDA/ARS and the Nebraska Agricultural Experiment Station. Subsequent research led to the development of an improved cultivar 'Trailblazer'. This cultivar represents a significant advance in terms of animal performance (Vogel et al. 1991). The genetic base of switchgrass and other native western warm-season grasses continues to expand as a result of plant exploration. These include big bluestem (Andropogon gerardii Vitman var gerardii), sand bluestem (A. gerardii var paucipilus (Nash) Fern), and indiangrass [Sorghastrum nutans (L.) Nash] (K.P. Vogel pers. commun.).

Germplasm introduced from foreign countries has contributed in a significant way to the improvement of grasses adapted to the southern states. The most prevalent warm-season grass in the South is common bermudagrass [Cynodon dactylon (L.) Pers.]. Bermudagrass is native to Africa and was probably introduced to America in livestock feed by the Spaniards. Similar to quackgrass, this aggressive grass has been a troublesome weed, but it has provided germplasm for several improved forage and turf cultivars. The most notable of these is the cultivar 'Coastal' (Burton 1989). This cultivar was developed by the USDA/ARS at Tifton, Georgia from a hybrid between 'Tift' common bermudagrass and two introductions from South Africa (Burton 1947). Coastal has been vegetatively propagated on 5 million ha in the South and has served as a germplasm resource in the development of other cultivars including 'Midland', 'Tifton 44', 'Coastcross-1', and 'Tifton 78'. More recent introductions are being used in crosses to improve the winter hardiness and extend the range of bermudagrass further north (Burton 1989).

Other warm-season grasses that have reached American shores through plant introduction include bahiagrass (Paspalum notatum Flugge), which was first introduced from South America in 1913 by the Florida Agricultural Experiment Station (Watson and Burson 1985). Germplasm obtained from Africa has contributed to improved forage quality and disease resistance in pearl millet [Pennisetum glaucum (L.)] (Burton 1989). Buffelgrass (Cenchrus ciliaris L.) has significantly improved the productivity of 1 million ha of American grasslands. This apomictic species was introduced to south Texas from southern Africa in 1946 (Voigt and MacLauchlan 1985). Common dallisgrass (Paspalum dilatatum Poir), a native of South America, has become an important grass in the United States, although its popularity has declined somewhat in recent years (Watson and Burson 1985). Weeping lovegrass [Eragrostis curvula (Schrad.) Nees] was introduced from Africa in 1927 and introductions have contributed to its continued improvement since then (Voigt and MacLauchlan 1985; Voigt 1971; Hanson 1972). Tragic circumstances are associated with the introduction of centipedegrass [Eromochloa ophiuroides (Munro) Hack.]. Original seed of this species was found in the baggage of the prominent plant explorer, F.N. Meyer, who had been collecting in the Hunan Province in China. He apparently fell overboard from a steamer in the Yangtze River.


It is evident that plant exploration has and will continue to have a significant influence on the quality and productivity of American grasslands. Early introductions were often made by accident such as inclusion in the ballasts of ships or imported livestock feed. Limited exploration provided some genetic diversity, but most pasture and rangeland seedings were made with unimproved accessions that were often poorly suited for the particular environment or intended use. In some instances, adapted strains were generated through natural selection; however, the most meaningful genetic improvement was achieved through purposeful plant exploration and subsequent breeding. With more areas in the world opened to exploration and the scientific community more amenable to germplasm exchange, the outlook for continued improvement has never been better.


*Cooperative investigations of the USDA/ARS and the Utah Agricultural Experiment Station, Logan. Paper no. 4260.
Last update April 8, 1997 aw