The U.S. National Plant Germplasm System (1991)

T he size and organization of a program to manage genetic resources varies with the goals and policies of a nation and the resources it is willing to commit to that purpose. There are, however, basic elements essential to all national programs for managing plant genetic resources. This chapter describes the components of the National Plant Germplasm System (NPGS) in the context of the elements of the model outlined in the preceding chapter (Figure 1-1).

The basic mission of the national system is to make available plant germplasm to scientists in the United States and worldwide for plant improvement, research, teaching, or extension programs. Activities include exploration, exchange, collection, and introduction; increase or regeneration; evaluation; documentation; preservation or maintenance; and distribution.

Prior to reorganization of the Agricultural Research Service (ARS) in 1972, the major components of the germplasm system were administered through the ARS New Crops Research Branch within the U.S. Department of Agriculture (USDA), or, for some major collections, through the specific USDA-ARS branch dealing with that commodity. In 1974, following the reorganization, germplasm activities were grouped into the National Plant Germplasm System to provide an umbrella system for germplasm acquisition, preservation, preliminary evaluation, and distribution. With the ARS holding lead administrative responsibility,

the system was designed “to provide, on a continuing basis, the plant genetic diversity needed by farmers and public and private plant scientists to improve productivity of crops and minimize the vulnerability of those crops to biological and environmental stresses” (Jones and Gillette, 1982:1).

The NPGS is composed of several stations, repositories, or laboratories with varying responsibilities and locations throughout the United States (Table 2-1). The system holds more than 380,000 germplasm accessions representing more than 8,700 species.

While most of the activities of the U.S. germplasm program take place within the NPGS, no single site is solely responsible for all of them. The National Seed Storage Laboratory (NSSL), for example, is a specialized facility for long-term preservation that serves as a security backup to the active collections around the country. Seed viability testing and long-term storage are the major functions of the NSSL. Regeneration, characterization, evaluation, and distribution are performed elsewhere. While routine requests for seeds are often received at the NSSL, they are generally forwarded to and filled by a curator responsible for the particular active collection.

Cooperation and coordination among the components of the national system are essential. Individual sites may not be suitable for managing all of the accessions for which they have primary responsibility, and some accessions must be grown or regenerated at another site. Thus a station may maintain several species for which it has primary responsibility and may also grow plants for other collections. Materials that require short day-lengths to flower are problematic because the locations or other resources needed for regeneration are not always readily available.

Although spread over numerous locations, the NPGS is intended to function as an integrated national system for germplasm management. Plant materials entering or distributed from the NPGS follow predictable lines (Figure 2-1) of acquisition, conservation, management, and utilization. These activities are described below.

The Plant Introduction Office (PIO) and its predecessors have been responsible for the acquisition of germplasm since 1898. Closely allied to this office are plant exploration activities and the health and quarantine of imported plant materials. Accession documentation begins when the PIO records passport data that accompanies new accessions for entry into the Germplasm Resources Information Network (GRIN) database. All of these activities are conducted through the National Germplasm

Resources Laboratory (NGRL) at the Agricultural Research Service's Beltsville, Maryland, area.

Although plant exploration is an important part of the activities of the NPGS, the bulk of its collections have been acquired through exchange. Germplasm received from other collections worldwide accounts for an estimated 75 percent of acquisitions (S. Dietz, U.S. Department of Agriculture, personal communication, June 1989).

Not all of the germplasm that enters the United States comes in through the PIO. University researchers, botanical gardens and arboreta, companies, and private individuals all import plants and seed. Much of this material may not be duplicated or documented in NPGS collections.

Collections of unique germplasm, much of it privately held and not fully represented in the NPGS, form a considerable reservoir of diversity (Office of Technology Assessment, 1985). For example, groups such as the Seed Savers Exchange maintain heirloom or older varieties of vegetables, fruit, and flowers and distribute them primarily to individuals for personal use rather than for breeding new cultivars (Office of Technology Assessment, 1985; Shell, 1990). Their holdings are generally not part of NPGS collections and are not documented by it.

The PIO is responsible for cataloging incoming germplasm accessions, assigning plant introduction (PI) numbers, and distributing new acquisitions to appropriate curators. The PIO publishes an annual inventory, listing the materials that have been assigned PI numbers (e.g., U.S. Department of Agriculture, 1988a,b), and coordinates germplasm exchanges.

The PIO may receive germplasm entering the United States for transfer to the appropriate NPGS site or record that a particular site has received such materials. All documentation for accessions is verified at the PIO. A PI number is assigned by the NGRL once passport data are verified. The PIO then distributes the material to appropriate sites.

The office also monitors some of the germplasm that enters the United States through avenues outside the NPGS (e.g., industry, botanical gardens, researchers), especially those plants or seeds receiving the attention of plant quarantine. If any of these latter materials are of importance to the NPGS, the importer is contacted by the PIO and invited to provide samples to the appropriate NPGS site or collection.

With the aid of the National Program Staff of the ARS, PI numbers are also assigned to new crop varieties, parental and advanced breeding

lines, and even genetic stocks that are registered by public and private plant breeders. Registrations for agronomic crops that document origin and important traits of a particular material may be published in Crop Science (Burgess, 1971; White et al., 1988).

lines, and even genetic stocks that are registered by public and private plant breeders. Registrations for agronomic crops that document origin and important traits of a particular material may be published in Crop Science (Burgess, 1971; White et al., 1988).

Not all of the germplasm in the NPGS is assigned a PI number. Each collection site may maintain materials it received through channels other than that of the PIO. Currently, of the more than 372,000 accessions listed on the GRIN database about 70 percent are identified by PI numbers. Most of the balance carries identification numbers assigned by the NPGS sites.

Since 1898 more than 500,000 accessions have been received by the PIO (Table 2-2). Before the late 1940s, introductions went directly to interested researchers or breeders without any requirement that they be maintained beyond their usefulness to the individual. Duplicate or back-up samples were not held by the USDA since no facilities for that purpose existed until 1948. Consequently, most germplasm accessions obtained before 1948 are no longer available. However, many important breeder's lines and cultivars contain genes derived from them. Since 1948, an increasing proportion of germplasm introductions have become part of the NPGS collections and are listed on the GRIN (Table 2-2).

Government-sponsored exploration for the purpose of collecting new germplasm is coordinated through the NGRL. The primary mission is the planning and implementing of plant explorations, especially in foreign areas (Purdue and Christenson, 1989). Plant exploration is a deliberate effort by the NPGS to seek and acquire specific kinds of germplasm.

Exploration proposals may be developed by one or more individual researchers who submit a formal proposal through the appropriate crop advisory committee, which is an NPGS advisory group specific to a crop. Scientists need not be employees of USDA or the national system to make such requests. Alternatively, proposals may be developed by a crop committee, the NGRL, or ARS National Program Staff. Once proposals are approved, qualified scientists undertake the exploration.

Some U.S. scientists conduct explorations using funds from other government sources, such as the National Science Foundation (NSF), or in cooperation with botanical gardens or arboreta. These activities

The grasses native to the Altai region of the south central Soviet Union are surveyed as part of an effort to collect wild rye species for U.S. germplasm collections. Credit: U.S. Department of Agriculture, Agricultural Research Service.

support botanical research, plant conservation, or other goals that may be unrelated to the perceived needs of the NPGS.

Exploration can also be accomplished through cooperation with individuals or organizations outside the United States. The NPGS occasionally cooperates with the International Board for Plant Genetic Resources (IBPGR) to collect materials of mutual interest. Cooperative exploration almost always involves scientists within the country where exploration occurs. In accordance with IBPGR practice, duplicate samples of collected germplasm are provided to germplasm collections within the country of origin.

When germplasm is acquired from other areas or regions, pests or pathogens may be introduced that could endanger domestic agriculture. Quarantine regulations are intended to reduce this risk, but they are not intended to facilitate the movement and entry of germplasm into the country. The role of expediting the quarantine process falls to the National Plant Germplasm Quarantine Center in Beltsville, Maryland, which recently was subsumed within the NGRL. The center represents a cooperative effort of the Animal and Plant Health Inspection Service (APHIS) and the ARS. ARS scientists and technical personnel provide expertise to propagate materials under quarantine and to test them for the presence of pathogens and other pests of quarantine significance. APHIS scientists and personnel certify such materials for release once quarantine regulations have been satisfied.

Obtaining pathogen-free, healthy plants for a number of clonally propagated crops, particularly fruit trees, can require years of quarantine. For this reason selected materials are maintained as virus-free clones that can be distributed without the delay associated with lengthy quarantine.

The Interregional Research Project-2 (IR-2), located in Prosser, Washington, functions as a national center for virus-free cultivars of deciduous fruit trees and selected ornamentals. It maintains more than 1,000 virus-free cultivars and holds the only known virus-free clones of a few germplasm materials. The IR-2 collections consist primarily of commercially important cultivars for use by researchers and industry. The program also develops methods for detecting and eliminating viruses

from infected fruit tree cultivars. IR-2 maintains primarily the stocks of current importance for fruit production; it is not considered by many people to be a part of the NPGS.

After acquisition by the NPGS, accessions are sent to the appropriate curator or collection for conservation. Conservation activities include increasing the sample size through grow-out if there are few seeds or plants, maintaining the material to preserve its genetic integrity, and ensuring that there is sufficient material for use.

In general, two kinds of conservation collections exist. Active collections, such as those at the regional plant introduction stations, multiply the material and are the primary sites for its distribution, characterization, evaluation, and general management. Base collections are back-up reserves of the active collections held under conditions of long-term storage. For seeds in the NPGS, base collections are held at the NSSL. For perennial woody plants, and some selected herbaceous species, which are usually propagated asexually, there are no back-up or base collections, and the national clonal germplasm repositories serve as the sites for both active and back-up collections.

The central elements for managing and maintaining the germplasm of the NPGS are the many active collections throughout the United States. They are responsible for maintaining the germplasm, characterizing and evaluating it, and producing viable seed or planting materials, and they are the primary sources of material for distribution and exchange. The active collections of the NPGS include those of the regional stations and clonal germplasm repositories, and several commodity or special collections (Table 2-1). Other active collections, that are not part of the NPGS, exist in private or institutional collections at colleges, universities, and state agricultural experiment stations. Still other collections are held by industry, nonprofit organizations, botanical gardens, and arboreta. There is no precise information regarding the number, size, or condition of many of these mostly private collections, but it has been suggested that they probably represent a substantial germplasm pool (Office of Technology Assessment, 1985), some of which may be of considerable importance.

Four regional stations have overall responsibility for maintaining the major seed-reproducing species held by the national system. These are the North-Central Regional Plant Introduction Station, Ames, Iowa; the Northeast Regional Plant Introduction Station, Geneva, New York; the Western Regional Plant Introduction Station, Pullman, Washington; and the Southern Regional Plant Introduction Station, Griffin, Georgia (Table 2-3). They are operated jointly by the ARS and state agricultural experiment stations through the Cooperative State Research Service (CSRS). Collectively, they hold approximately 135,000 accessions of nearly 4,000 species.

As originally envisioned in the 1940s and 1950s, the regional stations were established to meet the germplasm needs of plant breeders and other scientists. They were to provide foreign and native plant germplasm to crop scientists, preserve and evaluate introduced materials, and serve as holding facilities for the nation's genetic resources. Their responsibilities were based mainly on the concerns of agriculture in

A field area at Central Ferry, Washington, on the Snake River is operated by the Western Regional Plant Introduction Station and used for germplasm regeneration, evaluation, testing, and maintenance. Credit: U.S. Department of Agriculture, Western Regional Plant Introduction Station, Pullman, Washington.

each of their respective geographic regions. As each station was established, it was incorporated into the ongoing federal plant introduction system, which in the late 1940s was headquartered at the Beltsville (Maryland) Agricultural Research Center. Site selection was based in part on accessibility, facilities, and by joint agreement of the directors of the respective state agricultural experiment stations and the ARS. It was agreed that the experiment stations would provide land, assist in establishing laboratories, greenhouses, and related facilities, and provide office space for staff. The ARS and CSRS supplied most of the funds for equipment, operating expenses, and staff.

In addition to the regional stations, four federal plant introduction stations were active during the 1940s and 1950s. Since then their status has changed. The station at Glenn Dale, Maryland, served as an introduction station and national quarantine center. It is being phased out during the 1990s. The station at Miami, Florida, was responsible for subtropical crops, particularly fruits, rubber, cacao, and coffee. One of the earliest introduction stations, it has now become one of the clonal germplasm repositories and continues to maintain its earlier collections. The facility at Savannah, Georgia, held accessions of bamboo, sweet potato, and several other crops, but it is no longer a federal plant introduction station. Its germplasm was transferred to other sites. The station at Chico, California, held germplasm of deciduous tree fruits and nuts adapted to semi-arid conditions and has since been closed. A portion of its accessions were transferred to clonal germplasm repositories.

The regional stations receive and distribute germplasm for most of the species that can be stored as dry seed. Thus, they maintain the active collections for much of the seed material in the national system. They are responsible for seed increase and for depositing back-up samples in the base collections of the NSSL. Curators at regional stations interact regularly with users concerning management and use of the species for which they have responsibility. Curators also characterize and evaluate germplasm, but such activities can be limited by insufficient funding or staff.

The clonal repositories (Table 2-4) contain active collections that hold and propagate agriculturally important germplasm, such as strawberries, raspberries, fruit trees, coffee, and nuts, that for a variety of reasons are not usually held in active collections as seed. Eight repositories are distributed over 10 sites, which together hold more than 27,000 acces-

The characteristics of fruit on an apple germplasm accession are examined at the National Clonal Germplasm Repository at Geneva, New York. Credit: U.S. Department of Agriculture, Northeast Regional Plant Introduction Station, Geneva, New York.

sions. Accession decisions are generally made following consultation with public and private sector representatives, crop advisory committees or technical advisory committees, and ARS officials.

The primary responsibilities of the repositories are to collect, identify, propagate, preserve, evaluate, document, and distribute clonal germplasm as part of the NPGS. This includes maintenance of an information file on each accession in the clonal collection (National Plant Germplasm Committee, 1986). For most material held in such collections, long-term storage is not feasible, so duplicate materials are generally maintained in field and greenhouse or screenhouse collections to provide some back-up protection against loss. The repositories are charged with developing active global collections of appropriate wild species and domestic cultivars, and to assemble a maximum level of genetic diversity possible for each genus and species for which they are responsible. They also conduct research to improve evaluation, propagation, characterization, and preservation of clonal germplasm (National Plant Germplasm Committee, 1986).

The national clonal germplasm repositories are intended to carry out the same function for vegetatively propagated crops as that carried out by the regional stations and the NSSL for seed crops. Unlike seeds held

at the laboratory, however, backups of clonal collections are limited to the same sites where the active collections are maintained. Many clonal crops can be conserved as seed, but they are impossible to maintain true to type by raising plants from seed. Many clonally propagated species take a long time to mature, and they are best preserved as mature live plants for plant breeding and research.

Clonal collections are expensive to establish, and they have many of the same problems that confront seed collections. Accessions must be maintained as plants in the field, which can require large tracts of land, or in screenhouses or greenhouses. Accessions may also be maintained as live sticks of budwood held under refrigeration or as tissue cultures. There may be losses during maintenance from insects and disease, freezing temperatures, electric power failures, or grazing animals. Clonal preservation is more expensive and labor intensive than seed storage. Clonal collections have been threatened as facility or land-use priorities have changed, as the principal scientists retired, died, or moved, or as funding declined. By establishing the national clonal germplasm repositories, a mechanism for stable, long-term maintenance for many important clonally propagated species has been provided.

The Interregional Research Project-1 (IR-1), located in Sturgeon Bay, Wisconsin, began in 1947 and is supported cooperatively by USDA through CSRS and ARS, and by the Wisconsin State Agriculture Experiment Station. It is the national repository for potato germplasm. The station has a collection of about 3,500 accessions of more than 100 wild and Cultivated potato species and is an important global resource.

IR-1 uses a variety of methods to maintain germplasm. True seed of potatoes, in vitro plantlets of selected clones, and tubers are maintained. The germplasm is propagated both for maintenance and distribution.

The National Small Grains Collection (NSGC) (Table 2-3) began in 1894 as a breeder's collection and is today the most widely used active collection in the NPGS (White et al., 1989). This collection, relocated in 1989 from Beltsville, Maryland, to Aberdeen, Idaho, holds more than 110,000 accessions of wheat, barley, oats, rice, rye, the wheat wild relative Aegilops, and the intergeneric wheat-rye hybrid, triticale. It is one of the world's largest collections with some of its accessions originating from nineteenth century plant explorations. About 100,000

samples are distributed yearly from the collection to breeders, researchers, and germplasm collections in the United States and abroad.

Important genetic resources are conserved and maintained at other federal, state, and university sites (Table 2-5). They include several crop

species collections, such as soybeans and other legumes, cotton, and grasses. Some crop species, such as the native grasses in Brookings, South Dakota, or clovers in Lexington, Kentucky, are maintained in collections by individuals with particular interests in them or by sites with histories of studying them. Maintenance support is generally provided by the NPGS, or by state or other federal sources. Many of these collections are not duplicated and are primary sources for the materials they hold.

Several colleges, universities, and state agricultural experiment stations (Table 2-6) hold collections of germplasm that support breeding programs or the research interests of individuals. Much of this germplasm was gathered independently and may not be part of NPGS collections. Some of these collections are not listed in any official inventory, because they are small or only used by a few individuals.

The committee could not assess the degree to which these collections, often widely disseminated, duplicate NPGS collections. They may hold proportionately smaller numbers of wild species, landraces, and prim

itive materials than NPGS collections, but, as with the Charles M. Rick Tomato Genetics Resource Center (formerly the Tomato Genetics Stock Center), they can be important sources for wild or primitive materials.

This report defines genetic stocks as accessions with unique genetic or cytological characteristics that frequently make them of particular value in basic research. They do not include elite or advanced breeders ' lines, in contrast with its definition by the Commission on Plant Genetic Resources of the Food and Agriculture Organization.

Genetic stocks differ from other germplasm stocks in several respects. Genetic stocks carry mutant genes or chromosomal rearrangements, deletions, or additions, and are often difficult and costly to maintain because they require specialized care and trained personnel. Stocks carrying recessive lethal traits must be maintained as hybrids. Cytological screening (chromosome examination) is needed to verify stocks with chromosomal abnormalities. Also, mutant stocks with lethal or deleterious genes often have short-lived seeds. These traits are often maintained in special genetic backgrounds or as groups of linked genetic markers.

Many genetic stock collections receive partial support from USDA through the ARS and from state sources. The NSF has been an important source of support, but it no longer funds maintenance of collections.

Genetic stock collections commonly held by the individuals or groups who developed them can be difficult to maintain because of their unique nature and frequently complex cytogenetics. Such collections are too easily lost when the person responsible for them moves on or retires. The NPGS has made efforts to determine where genetic stock collections are, to monitor them, and in some cases to provide funds through the ARS for maintenance.

The importance of genetic stock collections to crop development is well illustrated by the Charles M. Rick Tomato Genetics Resource Center at the University of California at Davis (Genetic Resources Conservation Program, 1988). The ARS provides partial support to the center, and an endowment fund has been established to support its work. It holds 2,750 accessions, about 1,000 of which are wild species related to the cultivated tomato. These wild species are important sources of resistance to 28 tomato diseases and pests ( Table 2-7), as well as tolerance to other environmental stresses, such as temperature extremes, salinity, drought, and waterlogging. The tomato has also become a popular research model system in molecular studies of higher plant genomes. The detailed

linkage map of the tomato genome was largely developed using the center's stocks.

Some genetic stock centers are expanding to include the maintenance of cloned DNA (deoxyribonucleic acid) sequences for use as molecular markers (restriction fragment length polymorphisms) and as specifically cloned genes. As these technologies develop and are increasingly used in research and breeding, the NPGS will have to consider how best to manage the data and DNA.

The National Arboretum, established in 1927 in Washington, D.C., by an act of Congress, is primarily concerned with research and public education on trees, shrubs, and related plants. It maintains its own herbarium (separate from the National Herbarium of the Smithsonian Institution) of more than 500,000 dried specimens with emphasis on economic and cultivated plants, including voucher specimens from USDA plant explorations. It cooperates closely with NPGS on an informal basis. An ARS plant germplasm staff person at the arboretum has the primary responsibility of developing a germplasm collection of woody landscape species.

The arboretum has an active worldwide germplasm collection program and distributes materials to users in both the public and private sectors.

It maintains about 60,000 accessions, primarily of ornamental trees and shrubs. Most are conserved as plants in outdoor plantings and in screenhouses or greenhouses. The arboretum has introduced many widely used ornamental plants. For example, the popular ornamental crapemyrtle hybrids, resistant to the powdery mildew fungus, were developed and introduced by the National Arboretum.

Some NPGS genetic resources are also held in base collections in long-term storage as a reserve to the active collections. If samples in active collections are no longer available or cannot be regenerated, the germplasm held in the base collection can be used to replace them.

Dried specimens of woody plants are among the materials held in the herbarium of the National Arboretum, Washington, D.C. Credit: U.S. Department of Agriculture, Agricultural Research Service.

Germplasm in a base collection must be maintained under conditions that promote long-term storage. For seeds this generally entails maintenance at low temperatures and low relative humidity. It may also include cryopreservation (storage in or suspended above liquid nitrogen between −150°C and −196°C) of seeds, pollen, in vitro cultures, or dormant buds, in the case of clonal germplasm.

The NSSL is the principal site of long-term seed storage of genetic resources in the United States (see Table 2-3). It provides base collection storage facilities for the national system and holds samples of other important seed accessions (Table 2-8). Consideration is being given to storing cell and tissue cultures, pollen, and DNA.

CRAPEMYRTLE PI 237884 Lagerstroemia fauriei Koehne GRIN Data Origin: Unknown Acquisition: Japan Common name: Crapemyrtle Year PI assigned: 1957 Specimens of Lagerstroemia fauriei can be found at the National Arboretum, Washington, D.C. Credit: U.S. National Arboretum. A wild species (Lagerstroemia fauriei), related to the popular ornamental tree, crapemyrtle (L. indica), illustrates the importance of wild plants as germplasm for improving and protecting horticultural plants. On November 26, 1956, when Dr. John Creech collected seeds of L. fauriei from its native habitat near a rocky streambed at an elevation of 1,000-feet (300 meters) on the small Japanese island of Yakushima, the common crapemyrtle was already

The NSSL currently holds in excess of 230,000 accessions of more than 1,800 species. It also holds reference specimens of germplasm registered through the Crop Science Society of America, and it is the sole site for some genetic stock collections such as the Jimson weed (Datura) collection of A. F. Blakeslee, important to early studies of chromosomes and plant development. Seed samples of plant varieties protected under the Plant Variety Protection Act are also held by the NSSL.

Selected collections of the national system have been designated by the IBPGR as part of its international network of base collections (Table 2-9). These may be exchanged as a part of the international collaboration between the NPGS and IBPGR. Many of these collections existed before the IBPGR designation.

The NSSL provides back-up storage for some major international

collections. For example, it holds accessions of rice from the International Rice Research Institute (National Research Council, 1988), but these are not inventoried as part of the NPGS collection. The seed accessions are maintained in bond, under quarantine in their original shipping cartons. Their disposition is at the discretion of the originating institution. The NPGS deposits duplicates of its barley and oat collections with the National Plant Gene Conservation Center in Ottawa, Canada, and is seeking arrangements with other nations to hold backups of other collections (H. L. Shands, U.S. Department of Agriculture, personal communication, September 1989).

As part of its comprehensive seed storage function, the NSSL periodically tests seeds to ensure continued viability. Regional stations, curators, and other cooperators are expected to regenerate supplies of seed as requested by the NSSL, although additional funds are not provided for this purpose. The NSSL does not regenerate, evaluate,

Back-up accessions of the collection of the International Rice Research Institute are kept in subfreezing storage at the National Seed Storage Laboratory.

enhance, or distribute germplasm as part of its mandate. The laboratory has begun efforts to rescue valuable seed samples of low viability. At present there are no effective plans for regenerating certain seed collections held at NSSL that are unadapted to conditions anywhere in the continental United States or Puerto Rico, even though the NPGS has accepted international base collection responsibility for such collections. This underscores the need for additional facilities and the importance of cooperating with other countries in the regeneration of such germplasm.

The present buildings of the NSSL were completed in the late 1950s and are still structurally sound. However, technological advances in seed storage and the need for more storage space have made the construction of a new, enlarged, and modernized laboratory imperative. Recommendations for the design of an expanded NSSL facility are given in a separate committee report (National Research Council, 1988). Until such facilities are completed, space will continue to be a major constraint on NSSL activities.

Germplasm held by the national clonal germplasm repositories and the IR-1 is not often stored in base collections as seed. In general, there

are no back-up or base collections of clonally propagated species, although clonal repositories typically maintain duplicates of field-grown accessions in screenhouses or greenhouses. For some species, seed storage is possible and some of these are held as part of the NSSL base collections. However, seeds from a heterozygous, clonally propagated plant, while preserving the genes of the accession, do not preserve the specific gene combinations that determine the plant's precise characteristics. It may be desirable to conserve selected gene combinations for some species, such as those that require several years to reach maturity.

For materials that must be maintained as clones, the active collections

usually keep back-up samples as field plantings or in screenhouses or greenhouses. Back-up accessions also may be held as in vitro tissues, plantlets, or other propagative material (e.g., budwood) held in a quiescent state, under special conditions. These back-up holdings are not regarded as base collections because their storage is neither long term nor secure from environmental loss (particularly for field-maintained duplicates). However, they provide some security against complete loss of the primary active collection. As tissue culture and cryopreservation become more reliable as preservation methods, it may be possible to provide back-up holdings of clonal materials at sites separate from the active collection. For example, the NSSL maintains a set of cryopreserved strawberry cultures as an experimental backup to the field collection at the repository in Corvallis, Oregon.

Germplasm management must ensure the continued high viability of all accessions and prompt, accurate distribution of samples in response to requests. Information about accessions, including passport, characterization, and evaluation data, must be in an accessible form to enable users to select from a large array of materials. This is accomplished in the national system through the cooperation of several sites, and it is facilitated by the GRIN database. The elements of this cooperation include the testing of seed viability, increase (regeneration) of seed and clonal germplasm, and distribution.

Germplasm maintained as seed must be tested regularly to monitor viability. This is carried out by the responsible NPGS site or in cooperation with the NSSL.

Seed germination tests under controlled conditions are performed at intervals that vary with the physiology of the species. Lettuce (Lactuca) and onion (Allium), for example, are more short-lived than many cereal grains or beans and need more frequent testing. Chemical tests of seed viability are less reliable and not generally practiced at NSSL; there is no reliable technique for the nondestructive assessment of viability.

Accessions received with low seed viability pose several difficulties, in part because seed with low viability may not survive well under storage (Priestley, 1986). If the sample size is small, prompt and careful regeneration must be performed to avoid further loss of genetic diversity.

The NSSL and IBPGR standards for viability at one time were different.

Seed is packaged and sealed in foil-laminated moisture-proof bags for storage at the National Seed Storage Laboratory. Credit: U.S. Department of Agriculture, Agricultural Research Service.

IBPGR set a standard of 85 percent germinability for seed, while the NSSL's minimum standard was 60 percent, but it was recently raised to match IBPGR. For a few accessions, such as some chromosomally aberrant genetic stocks, a lower standard may be appropriate because of their inherently low seed viability, even in freshly grown seed. Current viability data from the NSSL are based on a germination rate of 65 percent. Table 2-10 shows that 164,603 accessions (71 percent) in the collection have a germination rate greater than 65 percent. However, 45 percent of the accessions contain less than 550 seeds.

More important than viability, however, is whether the aged seed survives as a consequence of the genes it possesses. Over time selection may favor the survival of particular genotypes. Clearly, viability testing, sampling, and rejuvenation policies and procedures are critical to maintaining the integrity of accessions.

Viability tests, seed distribution, and the decline of viability in storage all reduce the numbers of viable seeds of an accession. New accessions with few or poorly viable seeds require immediate increase. Increase may be necessary when a large amount of seed is needed for special evaluations. Seed regeneration or increase is normally the responsibility of those holding active collections, such as the regional stations, but it may be requested either from the scientist who contributed the materials

Germplasm accessions of chickpea, lentils, and broad bean are grown to produce fresh seed at the Western Regional Plant Introduction Station. Credit: U.S. Department of Agriculture, Western Regional Plant Introduction Station, Pullman, Washington.

or from a third party, often on an unsupervised contract or cooperative basis.

Regeneration techniques should be designed to minimize changes in the relative abundance (frequencies) of genes within an accession. Under the extremes of drought, cold, or disease, for example, some individuals in an accession could survive or even thrive, while others would die, resulting in changes in the overall genetic composition of the sample (i.e., genetic shift). Because of the small population size of an accession, random genetic change (i.e., genetic drift) can also have a major effect. To prevent these genetic changes, parental populations of 100 to 200 plants are grown in an environment to which they are adapted to avoid unintended selection. Other protocols call for the use of multiple sites in varying environments or over succeeding years. Nonetheless, genetic changes may occur during the replenishment of aging seed through inadvertent selection during regeneration and seed aging (Roos, 1984a,b). Methods of storage and testing that reduce the frequency of regeneration are particularly important.

No procedure ensures that samples returned from regeneration have the same genetic composition as those that were sent out. New techniques, such as the comparison of gene frequencies before and after regeneration using restriction fragment length polymorphisms, or older technologies such as electrophoretic analysis of seed proteins, could provide some measure of assurance. The common practice is to assume that seed for regeneration has been grown correctly and protected from genetic change or contamination.

Clonal crops have specific requirements for germplasm maintenance and increase. Temperate tree fruits may require isolation to prevent loss due to pests or diseases. Screened area protection, quarantine isolation, and standard orchard management practices are among the methods employed. Specific environments may also be needed to ensure the normal development of plants or fruit. Herbaceous plants, such as potatoes, may require greenhouse culture, specific propagation methods, or unique cultural practices. As a consequence, clonal repositories can be very costly.

Distribution of germplasm to researchers and breeders in the United States and throughout the world is an important part of the activities of the national system (Table 2-11). Germplasm requests to the NPGS from outside the United States (except those from Canada) are processed through the Plant Introduction Office. Distribution for U.S. and Cana-

dian requests is done by the appropriate collection. A considerable amount of germplasm, particularly genetic stocks and breeding lines, is directly exchanged between scientists.

The PIO coordinates most of the international transfer of germplasm by the national system. In cooperation with the responsible active collection or curator, the PIO staff might work to determine which materials are appropriate to fill a request too large to be of practical use (e.g., a request for all of the accessions of a crop that may number in the thousands). Where quarantine regulations require certification or declaration of phytosanitary status of the material, the PIO, in cooperation with APHIS, completes the needed paperwork. This is frequently accomplished at the local level by qualified state or federal personnel.

Germplasm is not distributed from the NSSL unless it is unavailable from active collections or elsewhere in the NPGS. In this case, the NSSL generally sends the seed to an appropriate active collection site or the original supplier of the germplasm for regeneration. In cases of extreme need or emergency and if sufficient seed is available, the NSSL may distribute samples directly to users. If it is necessary to regenerate samples, distribution of requested germplasm can be delayed for months or, more rarely, years.

The number of seeds distributed depends on the species and the amount available in storage. In general enough seed is sent for the user to grow a row approximately 15 meters (50 feet) long. From this, requesters are expected to reproduce additional seed. Individuals requesting seed for personal use, as in gardens, are generally referred to other sources. Recipients of NPGS germplasm are asked to acknowledge its source in reports and publications, and to report performance or evaluation information to the NPGS curator for that crop.

To be useful and accessible to breeders and researchers, information about the germplasm in a collection should be well described for specific characters and should indicate origin. It will enable users to know which accessions are likely sources for particular genetic traits. The maintenance of this information in a retrievable form is referred to as documentation.

Passport data provide the basic documentation, including taxonomic designation and information about where an accession was collected. These data are provided by the collector or donor institution. They are reviewed upon arrival at the PIO and entered into the database record when a PI number is assigned by that office. Unfortunately, such minimal data are not readily available for many of the accessions of the

NPGS, particularly those received before 1978. While passport data do exist for many accessions, the NPGS has been slow in adding this information to its computerized database.

Characterization involves the assessment of a varying number of descriptors ranging from morphological to biochemical. These descriptors, intended to describe an accession in relation to others in the collection, are determined by crop advisory committees and curators. The information can be gathered by the curators of active collections as materials are regenerated or assessed by cooperating scientists during evaluation for resistance to disease, environmental responses, or other traits. The gathering of crop descriptor data is an important part of the work of regional stations, repositories, and commodity collections because such data better define germplasm holdings and aid re-identification during regeneration.

Evaluation is a lengthy, often repetitive process of examining accessions for traits of significance to potential users. Screening for disease resistance is an example. It may take many years to test all of the accessions for a very large collection in a sufficiently wide range of environments. Although preliminary evaluation for genetically stable traits is generally considered an NPGS activity, more detailed evaluation for characters such as disease resistance or production qualities is generally part of the research that accompanies a breeding program. In a few cases, such as alfalfa at the western regional station, NPGS sites have funded researchers at other locations to conduct evaluations. Private companies and other users of NPGS germplasm also evaluate material, and the NPGS has occasionally provided funds for evaluating certain traits.

The Soil Conservation Service (SCS) also cooperates with the NPGS through a memorandum of understanding with the ARS to perform evaluations on certain materials. One purpose of the SCS is to foster acquisition, evaluation, and distribution of plants important to soil and water conservation.

The usefulness of new accessions depends on the user being able to retrieve information about them. In 1977 ARS recommended setting up a central repository for genetic resources information with standardized crop descriptors that would improve the management of NPGS collections (Mowder and Stoner, 1989). The Germplasm Resources Information Network manages all of the data associated with acquisition, evaluation, regeneration, inventory, and other records of the NPGS collections. It

is a centralized computer database for managing and operating the national system and for informing scientists and other users about the location and characteristics of NPGS germplasm (Database Management Unit, 1987).

The GRIN runs on a Prime 9955 Model II computer and uses a software package and database management system distributed by Prime Computer, Inc. Additional programs are written to meet specific needs. This system is transportable only to other, compatible Prime computers. Smaller portions of the data can be formatted for use on microcomputers using commonly available software packages. Searches of the database with printed results can be requested.

The GRIN holds three kinds of information. Passport data are recorded by the PIO when the PI number is assigned. They include the name of the collector, collection site data, taxonomy, and collection longitude, latitude, and elevation. Characterization and evaluation data are of value to potential users and include general plant descriptions, agronomic responses, disease and insect pathogen susceptibility or resistance, quality, and yield. Finally, the GRIN provides inventory and seed request processing data to NPGS sites as an aid to managing their collections.

While administering the network is the responsibility of the GRIN staff in Beltsville, Maryland, most sites also have personnel trained to enter their evaluation and other site data directly into the network.

The NPGS centers, for the most part, do not develop new cultivars or improve breeding materials. Nonetheless, a considerable amount of public funds are dedicated to germplasm enhancement, or the transfer of useful traits into an agronomically appropriate genetic background. For some commodity collections at experiment stations or universities, germplasm conservation may be an adjunct to enhancement.

Many public institutions are involved in germplasm enhancement. State agricultural experiment stations, public and private colleges and universities, and private industry use NPGS collections for breeding and enhancement to produce improved parental materials. Breeders do not use these collections frequently, but find them invaluable for particular genetic characters not readily available to them (Duvick, 1984; Peeters and Galwey, 1988). For example, the appearance of the Russian wheat aphid in the United States in 1986 caused more than $130 million in losses to grain crops in 1988 (Peairs et al., 1989). An immediate search for host plant resistance in the national wheat collection has already revealed some promising sources of resistance.

Russian wheat aphids feed on wheat leaves, which then roll and impede the effectiveness of conventional pesticide applications. Genetic resistance to the aphid is found in some barley, rye, wheat, and forage grass germplasm and in a few wild species related to wheat. Credit: U.S. Department of Agriculture, Agricultural Research Service.

Colleges and universities have made much use of germplasm for breeding and basic research with the aid of public sector funds. Although the private sector now plays a major role in producing modern varieties of such food crops as corn, sorghum, and soybeans, it is unlikely to assume a larger role in evaluating basic germplasm stocks.

Scientists in public and private colleges and universities have contributed to genetic resources activities, including plant exploration, evaluation, and regeneration. For some investigators, the USDA has provided funds for specific research tasks. Other federal agencies (e.g., NSF) have sponsored the work of university scientists on germplasm evolution, population genetics, taxonomy, and physiology. Examples include the collections of maize assembled in the 1940s by E. Anderson and H. C. Cutler of Washington University and P. C. Manglesdorf of Harvard University; A. F. Blakeslee's Jimson weed (Datura) collection; and R. E. Cleland's Evening Primrose (Oenothera) collection, which provided benchmark studies in cytology and cytogenetics.

The maize collections of Anderson, Cutler, and Manglesdorf were

built on the remnants of an earlier collection developed by G. N. Collins and J. H. Kempton of the USDA and then left to languish. Similar examples can be cited for tomatoes, barley, peas, and their wild relatives. The NPGS has recently begun to assist in developing and maintaining several of these old, long-used collections.

Recently the NPGS has been able to support non-USDA curators of selected germplasm collections. These arrangements should reduce the probability of losing such valuable resources. They allow collections to be managed by researchers with specific knowledge of their contents who can use specialized techniques that may be necessary for maintenance. However, it is important to survey and inventory many other collections, before they are lost. The NPGS is attempting to do this (H. L. Shands, U.S. Department of Agriculture, personal communication, September 1989).

In general, the record of accomplishments of university scientists is excellent. Many of their collections were incorporated into USDA

SORGHUM PI 264453 Sorghum bicolor Moench GRIN Data Origin: Spain Acquisition: Spain PI assigned: 1960 Maintenance site: Southern Regional Plant Introduction Station Greenbug damage can be compared on sorghum hybrids (left) with and (right) without genetic resistance to the insect. Credit: A. Bruce Maunder. It may fairly be asked why large collections of germplasm should be maintained when an individual accession may be used only rarely, if ever. The value of a collection lies not in the frequency of its use, but in the resource it provides when, and if, it is needed. It is thus not surprising that some accessions may possess genetic traits that remain unknown or unrecognized for many years, but which later prove to be almost invaluable. In 1980, motivated by the threat of a new outbreak of greenbug (Schizaphis graminum) in the U.S. sorghum crop, scientists began to screen sorghum accessions in search of genes for resistance. An outbreak of a

collections, but some remain at risk. Increasing emphasis on molecular biology and genetic engineering at universities and colleges has raised concern about the declining interest in breeding, taxonomy, and evolutionary studies that could lead to abandonment of important collections. Recognition of the national importance of research on genetic resources has led to greater USDA-ARS support of at least some of these programs and scientists. There must be broader opportunities for the NPGS to harness the expertise of scientists in colleges and universities.

Chemical and pharmaceutical companies have supported plant collection activities to find new plant products. Pyrethrum from species of Chrysanthemum and anticancer compounds from Vinca species are examples. However, once the effective compound is isolated, identified,

and synthetically reproduced, there may be little economic incentive for the companies to maintain the germplasm.

The confectionery, perfume, beverage, and specialty food industries are interested in exotic plants and plant products. On occasion, these industries have funded collections of particular genera or species. Collections produced by these activities, however, generally do not become part of the NPGS.

A few, large U.S. seed companies, by contrast, have often contributed to the national system. These companies have regenerated collections in field plots for the NPGS. By growing materials obtained from many countries in a disease-free environment, the seed produced can be assured to be safe for importation. The multinational research programs of some seed companies have thus been used to meet the needs of U.S. genetic resources conservation and development.

Botanical gardens and arboreta are primarily intended to meet local needs. Their forms and functions may range from university gardens with academic and public service responsibilities, to privately sponsored gardens that are governed by a board of directors. They may or may not have research capabilities. Although many were centers of plant introduction (Plucknett et al., 1987), they are no longer primary repositories of crop genetic resources. They have become increasingly concerned with conserving wild plant species, particularly those that are rare or endangered. On a global basis this has been encouraged by the International Union for Conservation of Nature and Natural Resources (Bramwell et al., 1987) through its Botanic Gardens Secretariat.

The Center for Plant Conservation (CPC), an association of U.S. botanical gardens, maintains rare and endangered U.S. plant species. Cooperating gardens acquire and maintain plants or seeds of these species in their respective regions. The NPGS helps the CPC by providing back-up storage of seeds for its collections at the western regional station and at the NSSL. The CPC's collections are not, however, part of the NPGS.

Farmers, gardeners, and hobbyists have become important conservators of old, obsolete, or heirloom varieties of vegetables, fruits, and

The diversity of heirloom varieties of winter squash is shown in the collections of the Seed Savers Exchange. Credit: David Cavagnaro.

flowers (Office of Technology Assessment, 1985). Grass-roots groups share information, seeds, and plant materials. For example, the North American Fruit Explorers brings together people interested in conserving antique or heirloom fruit and nut varieties. A more organized effort, the Seed Savers Exchange (SSE), maintains a large collection of its own, forms networks of gardeners around the country, and publishes several books related to conserving heirloom genetic resources (Office of Technology Assessment, 1985). Native Seeds/SEARCH (Southwestern Endangered Aridland Resource Clearing House) in Tucson, Arizona, seeks to integrate cultural and biological resources in an effort to conserve locally adapted crops and wild species through in situ and ex situ efforts. Such groups are not formally part of the NPGS, but they may hold valuable germplasm not in its collections.

The SSE is one of the largest and most active nongovernmental groups that preserve plant genetic resources. It is based on a small farm in rural Iowa and depends on individuals who maintain seeds of numerous obsolete heirloom varieties of vegetables and crops. Like the NPGS, the SSE recognizes the problems of duplication, inconsistencies in nomenclature, contamination, inadvertent crossing between varieties, and record keeping. It has developed a database system, trained curators,

and produced a variety of written instructional materials. The Garden Seed Inventory (Seed Savers Exchange, 1989a) lists all nonhybrid seeds offered for sale by mail-order seed companies in the United States and Canada. It itemizes cultivars that are decreasing in availability and encourages readers to preserve them in their gardens. SSE also has produced a similar listing for fruit, berry, and nut crops available by mail order in the United States (Seed Savers Exchange, 1989b).

Nongovernmental collections generally differ from those of the national system. The SSE preserves old, obsolete, open-pollinated vegetable varieties not available commercially. Active members will often select materials to assure that the desired type is maintained, rather than to preserve genetic diversity. “Off” types are culled rather than preserved for the unusual or rare genes they may possess. Seed is produced every 1 to 3 years and is thus subject to regular environmental selection pressures each time a seed increase is made. In contrast, NPGS accessions are meant to be preserved without genetic change and are held as the basis for breeding and research activities. Thus NPGS collections generally contain a greater proportion of wild species, landraces, and breeding lines. Regeneration and storage methods are designed to reduce the number of times regeneration is needed. The SSE focuses on preserving the general, visually discernable characteristics of an accession, whereas the NPGS is more concerned with preserving all of its genes.

The Living Historical Farms also holds a variety of germplasm. These collections can be sources of old, obsolete varieties that are no longer commercially available. Recent interests in specialty vegetables has raised awareness of many older varieties and led to their commercial use.

The genetic resources maintained by many of these collections are at risk and are not always readily available for research purposes. They should be part of the national collections. Accuracy of nomenclature, authenticity, uniqueness, the extent of internal duplication, and overlap with existing national collections should be determined and assessed. Limited budgets, facilities, and personnel will constrain this work, but these collections should be encouraged and assisted. The NPGS agreement to provide back-up storage for CPC collections is an example of this cooperative effort.