Cotton Crop Germplasm Committee Meeting, May 1997

INTRODUCTION

Cotton, Gossypium spp., is grown as a source of fiber, food, and feed. Presently, 49 species constitute this genus, with a new species being discovered on an average of once every four years for the past 30 years. Of the 49 species now described, four are cultivated, G. hirsutum and G. barbadense which are tetraploid (2n=4x=52), and G. arboreum and G. herbaceum which are diploid (2n=2x=26). Gossypium hirsutum is by far the most widely grown species worldwide.

In the United States, cotton is grown in seventeen states and is a major crop in thirteen. Cotton production spans the southern half of the U.S., from the Carolinas to California. Cotton is produced on 13-15 million acres with a return of over 5 billion dollars annually for its fiber and seed by-products. Although cotton is grown mostly for fiber, its seeds are an important source of food oil, and the meal is a protein-rich by-product used as feed for ruminant livestock. There also is interest in feeding whole cottonseed and in eliminating seed gossypol to extend the usage of cotton's high protein meal to non-ruminant consumers. Cotton seed meal could become an important source of protein, especially in third world countries, where low protein diets are the rule.

Historically, market prices for cotton fiber have been quite volatile. Cotton, being a global commodity, can have its price affected by international swings between over and under supply. Production problems in China, Pakistan, and India in recent years have resulted in favorable prices for U.S. producers. International changes in clothing fashion also affect cotton prices. A recent trend in fashion toward natural fibers has impacted favorably prices received. Another factor which can have and has had an effect upon cotton prices are international crude oil prices. The price of synthetic fiber, cotton's direct competitor, is directly related to oil prices. In the recent past, short staple cotton prices have averaged between 60 and 80 cents per pound, with a high of one dollar per pound. Prices for extra-long staple (ELS) cotton, primarily Pima and Egyptian cultivars, have fluctuated during this same period from a low of 90 cents per pound to a high of $1.65 per pound for grade number one. Currently, about one half of the annual U.S. production of cotton is exported, bringing in revenue of better than three billion dollars.

PRESENT GERMPLASM ACTIVITIES

A. Collection

Since the last assessment of the status of cotton germplasm, ten acquisition trips have been made (Table 1). Seven of these were funded totally or in part by USDA, ARS. Eight of the trips were primarily in situ explorations, while two of the trips were made primarily in an effort to obtain germplasm held ex situ by other countries. Collecting trips included two trips to Australia, three trips to Mexico, and one trip each to the Galapagos Islands, the Caribbean Islands, and Northeast Brazil. The two germplasm exchange trips included visits to India, China, Russia, and Uzbekistan. During this time period twelve new species of Gossypium were described and an additional new species description is in manuscript. Seven of the new species were identified as a result of the collections listed in Table 1.

B. Evaluation

Evaluation efforts fall into two categories. The first is a systematic collection of descriptors, conducted principally but not exclusively by the curator of the working collection of the National Collection of Gossypium Germplasm. The second category of evaluation efforts are studies conducted as research projects or adjuncts to other research by public and private investigators. These latter evaluations usually involve subsets of varying sizes of the germplasm collection and often are not systematic nor exhaustive in their approach. The latter type of investigation, being highly goal oriented, often does not report 'negative' results from evaluations. Numerous evaluations are being conducted by university and federal investigators on problems ranging from plant biochemistry, to insect and nematode pests, to fungal and bacterial pathogens, to physiological stresses. Ongoing or recent evaluation activities are listed below.

1. Morphologic and agronomic trait evaluations -- USDA, College Station, TX, and Phoenix, AZ; and AAES, Fayetteville, AR.

2. Cytogenetic -- TAES, College Station, TX, and AAES, Fayetteville, AR.

3. Biochemical (gossypol) -- USDA, Phoenix, AZ and College Station, TX ; TAES, San Angelo, TX.

4. Insect resistance --

a. Boll weevil -- USDA and MAFES, Miss. State, MS.
b. Pink boll worm -- USDA and AAES, Phoenix, AZ.
c. Heliothis -- USDA, and TAES, College Station, TX; MAFES, Miss. State, MS; and AAES, AR.
d. Lygus -- USDA and MAFES, MS; CAES, CA; and AAES, AR.
e. Whitefly -- USDA and AAES, AZ; and USDA and TAES, TX.

5. Seed Quality -- Contract, USDA or TAES, College Station, TX.

6. Disease resistance- USDA, TAES, College Station, TX.

7. Nematode resistance evaluation -- USDA, Miss. State, MS, and College Station, TX; AAES, AR.

8. Stress evaluation.

a. Water use efficiency -- USDA and TAES, Lubbock, TX.
b. Temperature stress -- USDA and AAES, Phoenix, AZ.
c. Soil salinity tolerance - USDA and CAES, Shafter, CA.

9. Fiber properties -- USDA, College Station,TX; Miss. State and Stoneville, MS.

10. Taxonomic evaluation.

a. Australian species - AAES, Fayetteville, AR.
b. Variation patterns in the various species collections analyzed (Isozyme, RFLP, RAPD) - Iowa State Univ., Ames, IA; USDA-ARS, Phoenix, AZ, and College Station, TX; AAES, AR.

C. Enhancement

Enhancement, or developmental breeding, in cotton is an on-going pursuit within private, state, and federal agencies. As specially funded public programs, enhancement projects generally have had goals with regional or beltwide impact and broad appeal. Since many enhancement projects out of necessity begin with identification of desirable or suitable genetic materials, the line between evaluation and enhancement becomes blurred. A partial listing of current or recent enhancement programs and projects include:

1. Metabolic Efficiency Studies

a. Water use efficiency - USDA, Lubbock, TX. Development of germplasm adapted to the semiarid regions where water shortage is becoming a limiting factor.
b. High temperature adaptability -USDA-ARS and AAES, AZ. Development of germplasm with ability to withstand stresses related to high temperature in both G. hirsutum and G. barbadense.

2. Photoperiodic Flowering Response Conversion

Much of the G. hirsutum and G. barbadense collections, being of tropical origin, are photoperiodic in flowering response and require conversion to a day-neutral status before the variability of these collections can be exploited. The USDA-ARS at State College, MS (funded) and Phoenix, AZ (unfunded) have ongoing conversion programs.

3. Host Plant Resistance -

a. Multi-adversity - TAES, College Station,TX; and AAES, Fayetteville, AR. Development of germplasm possessing bacterial blight resistance, seedling disease resistance, and resistance to various insect and nematode pests.
b. Diploid germplasm identified as possessing HPR is being hybridized for introgression of HPR into tetraploid cottons - AAES, AR.

4. Molecular Genetics, Mapping, Foreign Gene Identification and Cloning

a. Mapping - USDA-ARS and TAES, College Station, TX; AAES, AR
b. QTL and population characterization - USDA-ARS and TAES, College Station, TX; NMSU, Las Cruces, NM; USDA-ARS, Florence, SC.
c. Foreign gene identification - AAES, AR

5. Fiber Trait Quality Improvement -

a. QTL and molecular approaches - Texas A&M University and USDA-ARS, College Station, TX.
b. Traditional breeding methodologies (including species introgression) - USDA-ARS at Florence, SC, and Phoenix, AZ; TAES, Lubbock, TX; NMSU, Las Cruces, NM.

6. Diversification of cytoplasm - AAES, Fayetteville, AR. Introgression of wild species cytoplasms into cultivated cotton and evaluation of cytoplasmic/nuclear interactions.

D. Preservation

The National Collection of Gossypium Germplasm is housed at the Southern Crops Research Laboratory, Crop Germplasm Research Unit, College Station, TX. At present the collection contains 6052 accessions representing 49 species from 74 countries and/or political jurisdictions, assigned to three germplasm pools (Table 2). The various components of the National Collection are:

1. Varietal Collection - 1937 accessions, primarily of Gossypium hirsutum.

2. Primitive Race Collection - 2098 accessions, primarily of wild or primitive races of G. hirsutum.

3. G. barbadense Collection - 1062 accessions of primitive germplasm, obsolete cultivars and current cultivars.

4. Asiatic Collection - 430 accessions of the diploid species G. arboreum (285 accessions) and G. herbaceum (145 accessions).

5. Wild Species Collection - 541 accessions representing over 40 species of wild diploid and tetraploid Gossypium spp.

6. Genetic Marker Collection - G. hirsutum and G. barbadense.

7. Base Collection (permanent storage of materials from the above) - maintained at the National Seed Storage Laboratory, Colorado State University, Fort Collins, Colorado.

In addition to the germplasm accessions listed above, 1180 Asiatic accessions have recently been acquired from India and are in the process of being increased. After seed increase these materials will be assigned to the appropriate subcollections.

Five locations are used for seed stock increase and renewal. The primary location is the Cotton Winter Nursery in Tecoman, Colima, Mexico. This location is used primarily for increasing long seasoned or photoperiodic accessions of G. hirsutum or G. barbadense. Varietal accessions can be increased at Monte Alto, Texas through an agreement with Rio Farms (a private research facility). Accessions of G. arboreum and G. herbaceum are grown at College Station, Texas when funds and field space allow. Several hundred lines or accessions are grown in the greenhouse at College Station, Texas each year. A trial seed increase was performed in Costa Rica to evaluate alternate sites for a winter nursery, but the cotton program at this site has since been terminated.

Seed renewal of accessions comprising the germplasm collection is currently being performed at the discretion of the curator. Previously, all accessions were renewed on a seven year schedule. Prior to that (1993), seed increases were performed on a five year schedule, or as needed for rescue. The increasing or indefinite time intervals between seed renewals results from static funding and increasing costs.

Sixty-five thousand dollars of the USDA, ARS, SCRL, CGRU recurring funds for cotton germplasm maintenance are dedicated to seed increase and renewal, and a further 100 thousand dollars of the cotton research budget has been directed recently to maintain the Cotton Winter Nursery operated by the Cotton Winter Nursery Steering Committee. This nursery is a vital component of the germplasm preservation program and its continued existence is crucial.

Table 2. List of currently recognized Gossypium species, organized by germplasm pools.
Species	Genome1	Notes
Primary (1o) Germplasm Pool
hirsutum	AD1	Current & obsolete cultivars, breeding stocks, primitive and wild accessions.
barbadense	AD2	Current & obsolete cultivars, breeding stocks, primitive and wild accessions.
tomentosum	AD3	Wild, Hawaiian Islands.
mustelinum	AD4	Wild, NE Brazil.
darwinii	AD5	Wild, Galapagos Islands.
Secondary (2o) Germplasm Pool
herbaceum	A1	Cultivars and land races of Africa and Asia Minor; one wild from Southern Africa.
arboreum	A2	Cultivars and land races from Asia Minor to SE Asia, & China; some African.
anomalum	B1	Wild, two subspecies from Sahel and SW Africa
triphyllum	B2	Wild, SW Africa
capitis-viridis	B3	Wild, Cape Verde Islands
trifurcatum	(B)	Wild, Somalia
longicalyx	F1	Wild, trailing shrub, East Central Africa
thurberi	D1	Wild, Sonora Desert
armourianum	D2-1	Wild, Baja California
harknessii	D2-2	Wild, Baja California
davidsonii	D3-d	Wild, Baja California
klotschianum	D3-k	Wild, Galapagos Islands
aridum	D4	Wild, arborescent, Pacific slopes of Mexico
raimondii	D5	Wild, Pacific slopes of Peru
gossypioides	D6	Wild, South central Mexico
lobatum	D7	Wild, arborescent, SW Mexico
trilobum	D8	Wild, West central Mexico
laxum	D9	Wild, arborescent, SW Mexico
turneri	D10	Wild, NW Mexico
schwendimanii	D11	Wild, arborescent, SW Mexico
Tertiary (3o) Germplasm Pool
sturtianum	C1	Wild, ornamental, Central Australia
robinsonii	C2	Wild, Western Australia
bickii	G1	Wild, Central Australia
australe	(G)	Wild, North Transaustralia
nelsonii	(G)	Wild, Central Australia
costulatum	(K)	Wild, decumbent, North Kimberleys of W Australia
cunninghamii	(K)	Wild, ascending, Northern tip of NT, Australia
enthyle	(K)	Wild, erect, N Kimberleys, Australia
exgiuum	(K)	Wild, prostrate, N Kimberleys, Australia
londonerriense	(K)	Wild, ascending, N Kimberleys, Australia
marchantii	(K)	Wild decumbent, Australia
nobile	(K)	Wild, erect, N Kimberleys, Australia
pilosum	(K)	Wild, ascending, N Kimberleys, Australia
populifolium	(K)	Wild, ascending, N Kimberleys, Australia
pulchellum	(K)	Wild, erect, N Kimberleys, Australia
rotundifolium	(K)	Wild, prostrate, N Kimberleys, Australia
anapoides	(K)	Wild, erect, N Kimberleys, Australia
stocksii	E1	Wild, Arabian Peninsula and Horn of Africa
somalense	E2	Horn of Africa and Sudan
areysianum	E3	Arabian Peninsula
incanum	E4	Arabian Peninsula
bricchettii	(E)	Somalia
benadirense	(E)	Somalia, Ethiopia, Kenya
vollensenii	(E)	Somalia
1The genomic grouping of the Australian species is under study. Where used, ( ) indicate provisional genomic placement for the species in question.

STATUS OF CROP VULNERABILITY

Genetic diversity is believed to provide a buffer against adverse effects of sudden increases in the virulence of pathogens or pests. An oft-cited example of the danger of genetic uniformity is the Southern corn leaf blight crisis of the 1970's. The majority of corn hybrids at that time shared a common cytoplasm that was used because it greatly facilitated hybrid seed production. This cytoplasm, and all hybrids using the cytoplasm, proved to be highly susceptible to a race of Southern corn leaf blight. Since that time, greater attention to the preservation and expansion of genetic diversity has been encouraged in all crops, although the encouragement is not always heeded.

May et al. (Crop Sci. 35:1570-1574) evaluated the genetic diversity among upland cotton (G. hirsutum) cultivars released in the U.S. between 1980 and 1990 by calculating coefficients of parentage (CP) among these cultivars. They found a relatively high degree of diversity among this group of cultivars. Mean CP among cotton cultivars for this period was 0.07, compared to 0.15 for soft red winter wheat and 0.22 for hard red winter wheat in 1984 (Cox et al., Proc. Natl. Acad. Sci. 83:5583-5586), and 0.32 for runner peanuts and 0.21 for Virginia peanuts (Knauft and Gorbet, Crop Sci. 29:1417-1422). Cluster analysis (May et al., Crop Sci. 35:1570-1574) identified groups of cotton cultivars that were more closely related. With-in cluster mean CP ranged from 0.17 to 0.34. Cultivars within each cluster generally were used in the same geographical area.

The genetic diversity among all available cultivars (i.e., available diversity) generally is higher than the diversity that exists among widely grown cultivars (i.e. diversity in current use) because producers tend to plant a few, preferred cultivars. Information in Table 3 reflects the relatively low degree of diversity in use during 1995. Acreage accounted for by the leading cultivars was taken directly from the USDA-AMS report, "Cotton Varieties Planted 1995 Crop." In the Southeast, Deltapine Acala 90 was the most popular cultivar with 18.5% of the acreage. Three other cultivars each occupied over 10% of the acreage. Two of the three were closely related to Deltapine Acala 90 with r > 0.6. In the South Central Region, Deltapine 50 was the most widely grown cultivar with 19.1% of the acreage. Two other cultivars, both closely related to Deltapine 50 (r = 0.52 to .075) each occupied over 10% of the acreage. In the Southwest, Paymaster HS26 and Paymaster HS200 occupied 33.2% and 16.9% of the acreage, respectively. These two cultivars are not closely related (r = 0.01). In the West, Acala Maxxa accounted for 52.7% of the acreage. The next most popular cultivar, Deltapine 5415 (not closely related to Acala Maxxa, r= 0.05), accounted for 17.1% of the acreage. Other cultivars occupied <5% of the acreage in the West. In all regions, one-half to one-third of the acreage was occupied by a single cultivar or closely related cultivars.

Table 3. Genetic diversity among U.S. Upland cotton cultivars in 19951.

Coefficient of parentage among 7 leading cultivars

	Acreage of 7 leading	Acreage of 2 leading		Weighted by proportion of total
Region	cultivars	cultivars	Mean	acreage

-------------------%---------------

-----------------r-----------------

Southeast	72	31	0.31	0.25
South Central	75	35	0.35	0.25
Southwest	72	50	0.10	0.24
West	82	70	0.14	0.34

1 Acreage estimates are from the USDA-AMS report, "Cotton Varieties Planted 1995 Crop". Coefficients of parentage are from the pre print of USDA-ARS Technical Bulletin, "Coefficients of Parentage for 260 Cotton Cultivars Released Between 1970 and 1990".

Coefficients of parentage were drawn from the pre print of USDA-ARS Technical Bulletin, "Coefficients of Parentage for 260 Cotton Cultivars Released Between 1970 and 1990". Mean CP's were calculated as the arithmetic mean of pair-wise CP among the seven leading cultivars in each region. In the Southeastern and South Central Regions, the mean CP for the seven leading cultivars was >0.30. In the West and Southwest, greater diversity existed among the leading seven cultivars. However, when diversity was weighted by acreage of the cultivars, as described by Cox et al. (Proc. Natl. Acad. Sci. 83:5583-5586), acreage-weighted CP for the West was highest at r = 0.34 and other regions had r > 0.23.

While the diversity represented in commercial cotton production in any year and/or region may be somewhat limited, we do not appear to be in a situation similar to the corn industry of the 1970's when virtually all commercial production shared a common cytoplasm. If the leading cotton cultivar in a region was lost to disease or insect susceptibility, there is certainly the possibility that closely related cultivars (totaling approximately 50% of the acreage in most regions) also would be lost. "Lost", as used here, means that the cultivar(s) would no longer be commercially useful due to disease or pest susceptibility; it is unlikely that such susceptibility would result in the total loss of production of affected cultivars.

In the event that a portion of the currently popular cultivars were lost to disease or insect susceptibility, it appears at the present time that there are adequate genetic resources available to address the situation. Replacement cultivars could include cultivars that are grown currently on a minor acreage, cultivars from other regions, or older cultivars that have been recently replaced. Such loss of current popular cultivars would cause a temporary disruption in the industry as necessary adjustments were made.

The view that we presently have adequate genetic resources does not mean that breeders should relax their vigilance to maintain and expand genetic diversity. It is only through discovering or creating new variability that progress through breeding can be achieved. Judging by the diversity among clusters in the study by May et al. (Crop Sci. 35:1570-1574), it may be possible to both increase diversity within regions and achieve greater genetic gains in yield and fiber quality by tapping genetic resources outside one's geographic area. Incorporating variability from wild cottons and related species should remain a priority so that new variability is available for future generations.

The diversity of commercially produced Pima (G. barbadense L.) cotton is not known with the same precision as that of G. hirsutum. Until recently, it was the practice of the American Pima cotton industry to grow a single cultivar (sometimes two) at any given time. This circumstance left the commercial Pima industry vulnerable to any "new" disease, insect, or other adversity which might arise. The advent of private Pima breeding programs in the 1990's has increased both the number of available cultivars and the apparent diversity of commercial germplasm. Countering this apparent diversity is the fact that, at present, all commercial cultivars originate from a single germplasm pool maintained by the USDA-ARS. This germplasm pool has a complex developmental history, and is thought to be genetically diverse. The pool includes sources of Egyptian, Sea Island, Coastland, Pima, and upland germplasm.

Less is known about the two commercially produced diploid species, other than germplasm resources and diversity are thought to be large. Research to identify superior agronomic genotypes suitable for production in the U.S. has been initiated recently.

The effect that genetic engineering may have on genetic diversity in the future is unclear. On one hand, these new technologies make it possible to incorporate traits (variability) that are not otherwise available. However, the current trend in applying these technologies may serve to decrease background genetic diversity. The reliance of the present generation of transgenic lines on tissue culture regeneration from a very limited "regenerable" germplasm has restricted genetic variability. Further, most programs utilizing transgenes rely heavily or exclusively on backcrossing to incorporate foreign genes into a desired genotype--a strategy that does nothing to expand genetic diversity. Backcrossing programs act as a drag to the replacement of old cultivars with substantially different new cultivars; instead creating cultivars that differ by only a few alleles from their predecessors. By incorporating transgenes into traditional cross-and-select breeding programs, the industry can benefit from novel and valuable traits while maintaining background genetic diversity.

Our greatest opportunity to reduce or minimize genetic vulnerability in cotton lies in greater and more efficient use of our feral and/or exotic germplasm. Public sector programs, both federal and state, must take the lead in developing and enhancing this valuable wealth of germplasm. Cooperative research projects and information exchange groups, such as S-258 and SRIEG-61 , are instruments through which federal and state workers maintain and study the Gossypium germpool and coordinate research on the genetics, cytogenetics, pathology, taxonomy, and entomology of cotton. Increasing the scope of this work will assist in a better understanding of the diversity existing in Gossypium, of the heritable systems of the cotton plant, and of the systematics of the genus. More complete evaluations and enhancements of all the materials in the existing collection, and new materials added to the collection, would generally increase available germplasm for use by cotton plant breeders as needs arise in an ever-changing cotton industry.

GERMPLASM NEEDS

A. Collection

Collection efforts in the last ten years have improved the diversity available in some germplasm pools. In others, particularly in the dooryard or commensal accessions of the cultivated tetraploid species, modernization and the development of a global economy threaten their continued existence. Small personal plantings and dooryard plants have been abandoned in favor of commercial cultivars - setting up a now-or-never situation for collection. High priority should be placed on obtaining threatened commensal or dooryard accessions of the tetraploid species. In all germplasm pools, unplanned opportunities may arise to obtain germplasm filling previously identified deficiencies. Such 'targets of opportunity' as exchange projects with countries which maintain collections should receive high priority when they arise, and should be aggressively pursued. Below are listed some of the germplasm collections, categories, or geographical regions identified as germplasm acquisition priorities.

1. Germplasm Exchange -

a. Russian Republic, Uzbekistan, and other cotton growing republics of the former Soviet Union - Recent contacts indicate that it may be possible to develop an exchange agreement with the above countries. A problem will be funding. It is highly likely that we will have to supply all funding for such a project - which may entail such activities as seed renewal at the collections' present location prior to exchange.

b. Colombia, Greece, Turkey, etc. - These countries are known to maintain collections. However at present we are not aware of the numbers or type of materials in these collections or of the possibility of exchange agreements.

2. South America - The greatest deficiency in the primary germplasm pool of the germplasm collection is the under-representation of the diversity of G. barbadense and G. mustelinum from South America. A wild diploid species closely related to one of the subgenomes of the cultivated tetraploid cotton species, G. raimondii also is native to South America.

a. Gossypium barbadense - Feral and ruderal germplasm is rapidly being lost to development in an area which has been under-collected. This area roughly encompasses northern Argentina, eastern Bolivia, and parts of Paraguay and Brazi l. Collection in this area is a high priority and should be pursued as an independent objective.

b. Peru is the center of origin and diversity of G. barbadense and encompasses the total range of G. raimondii (the diploid species thought to be a close relative of one of the progenitors of tetraploid cotton). Obtaining germplasm from Peru is a priority and a contact has been established which may be highly productive.

c. G. mustelinum, a relic species resembling the original ancestor of tetraploid cottons, appears to be restricted to small and widely spaced locations in NE Brazil, where it may be threatened with extinction. U.S. scientists should cooperate with Brazilian counterparts to determine the true extent of its distribution and vulnerability as well as conserve its genetic diversity ex situ.

3. Secondary germplasm pool -

a. A-genome species (G. arboreum and G. herbaceum). Countering the recent acquisition of 1180 accessions of G. arboreum from an Indian collection is an apparent high level of redundancy in this material. Regional diversity known to exist among the A-genome diploid species, but poorly represented in the collection are:

1) South Africa - The only known wild representatives of the G. herbaceum species, assumed to be the progenitors of the cultigen, occur in South Africa. These populations are very poorly represented in the collection and should be extensively collected.

2) Southeast Asia (other than India). G. arboreum is known to exist throughout SE Asia (Nepal, Bangladesh, Burma, Thailand, Loas, Vietnam, Cambodia, Malaysia, Indonesia) as dooryard plants or in small family plots. The representation of these genotypes in the U.S. collection is very poor, and the amount of diversity is unknown. Exchanges, regional surveys, and collections of this germplasm is a high priority.

3) Other areas - A-genome cottons have spread to other areas of the world by human activities and have been locally adapted in small plot cultivation. Primary areas for additional acquisitions are the African Sahel, and the countries of Asia Minor eastward (Turkey, Iraq, Iran, Pakistan). These should be obtained as opportunity arises.

b. D-genome wild diploid species - Collecting expeditions in the last 10 years have provided good representation of G. davidsonii, G. armourianum, G. harknessii, G. thurberi, and G. turneri. The most widely distributed and most taxonomically diverse species, G. aridum, is surprisingly one of the least represented. Special effort should be made obtain representative seed samples from throughout the range of this species. Concurrently, additional populations of G. lobatum, G. laxum, G. schwendimanii, and G. gossypioides should be collected in this effort. Special attention should be given to obtaining samples from the range of G. trilobum, which has an extensive range but is poorly represented in the U.S. collection.

c. B- and F- genome species. These African species (5 total) are represented by four or less accessions in the Wild Species Collection, all of unknown provenance. One species (G. trifurcatum) recently described from herbarium specimens is not represented in the collection. By virtue of their status in the secondary germplasm pool, their proven record of possessing a range of host plant resistances, and sparse representation in the collection, acquisition of documented accessions of these species is important.

4. Tertiary germplasm pool (species of Australia, and Afro-Arabia) Collections in Australia, including the addition of 7 new Gossypium species and extensive representation of the geographic range covered by the genus, have made significant contributions to the available diversity. However, Afro-Arabian species are very poorly represented. No accessions are available for three Afro-Arabian species, and the remaining four are represented by 3 or fewer accessions, of unknown provenance in most cases. The Horn of Africa, where many of these species occur, appears to be an area of diversity for Gossypium and should be explored extensively when the political environment allows this to be done safely. Species on the Arabian peninsula should be collected as soon as opportunity and inclination of potential collectors coincide.

B. Evaluation

Collection efforts in the past ten years have improved the diversity available in the tetraploid species and in several diploid species. However, germplasm evaluation has lagged significantly. The report of the long-range plan for the National Plant Germplasm System, 1980, recommended funding for cotton evaluations - to be split about 50% in-house and 50% extramural. To date, cotton has no funds provided for planned evaluation. Areas in need of greater emphasis and/or support include:

1. None of the collections are completely cataloged for all the descriptors of interest. Successful germplasm acquisition has added to the backlog. With no funding available for evaluation, a system of prioritizing the collection of descriptor information should be considered, and funding for these activities should be sought.

2. Molecular genome evaluation and mapping- A low density molecular marker map has been produced using an interspecific population of the two tetraploid species. To be useful in germplasm evaluation and enhancement efforts, a map of greater density is needed. A collection of molecular markers and probes should be established and a system of maintenance and distribution should be created.

3. Recent advances in fiber measurement instrumentation promises greater precision in measuring fiber properties contributing to superior end-products. However, fiber properties measured by the new instrumentation do not always correspond to traits previously measured. There is a need for inheritance studies of the fiber properties being measured by new instrumention and an evaluation of the germplasm for these traits.

4. Taxonomic evaluation - Countering the considerable taxonomic activity utilizing biochemical and molecular approaches that has occurred in the past 10 years is the loss of a permanent herbarium collection housed with the ARS in College Station, TX and the loss of a taxonomic specialist dedicated to the genus. Recent discoveries of new species, the likelihood that additional taxa will be discovered, and the continued redefinition of species' distribution; all indicate the need of continued support of taxonomic activities focussed on the Gossypium genus.

C. Enhancement

Some of the more immediate needs encompassing enhancement and requiring specific funding are as follows:

1. Metabolic Efficiency Studies Recommendations

a. Water use efficiency - Develop germplasm adapted to semiarid regions where water shortage is becoming the limiting factor.

b. High temperature adaptability - Develop germplasm with ability to withstand stresses related to high temperature in both G. hirsutum and G. barbadense.

2. Gene Transfer

a. Introgression of genes from wild races and through interspecific hybridization - Introgress desirable heritable traits (insect, pathogen, and environmental stress resistance; unique fiber properties; biochemical properties; etc) from primitive races into acceptable agronomic genotypes. Transfer desirable characters between species through interspecific hybridization work. With the latter, it may be necessary to develop the breeding methodology to accomplish this.

b. Patterns of genetic and geographical variation in the collections need to be analyzed in order to better screen for desirable genes. Genes that have similar functions seem to be grouped in material from specific geographic areas.

c. A major stumbling block to the efficient utilization of genetic engineering methodologies in foreign gene transfer to cotton has been the reliance on techniques which require tissue culture regeneration from a very limited "regenerable" germplasm. Genetic engineering methods which do not require plant regeneration or regeneration methodologies which are genotype independent are required.

d. Identification and transfer of useful genetic variation is impeded in a large portion of the germplasm collection by the photoperiodic nature of tropical accessions. Conversion of photoperiodic G. hirsutum race stocks to a day-neutral flowering status in is being conducted by a funded USDA, ARS program at State College, Mississippi. Conversion of G. barbadense germplasm is being performed by a USDA, ARS project located in Phoenix, Arizona. The scope and pace of G. barbadense conversion has been impeded by the lack of any allocated funding for this purpose.

As previously stated, many enhancement projects out of necessity begin with identification of desirable or suitable genetic materials, blurring the line between evaluation and enhancement efforts. Meaningful enhancement efforts for insect, pathogen, or environmental stress resistance, fiber property improvement, performance characteristics, etc., are dependent upon adequate characterization of the germplasm.

D. Preservation

1. With the accession number in the collection approaching 7000 entries, there is a need to increase storage space. This would entail expanding vault carrying capacity, increasing drying and refrigeration, and increasing labor to handle the additional volume. Storage space limitations will become acute by the summer of 1998 with the increase of the additional of the 1183 accessions acquired from India,

2. Due to static budgets and increasing costs, seed renewal and increase is now performed on an 'as needed basis' at 'the discretion of the curator'. With this new, indefinite criteria for seed renewal, it would be prudent to set up facilities to do periodic germination testing and establish guidelines for the periodicity of testing. Such testing would insure that accessions are increased on a timely basis and might save money by eliminating some unnecessary seed renewals.

3. One of the primary functions of descriptor data is to insure that the integrity of accessions in the collection are being maintained. The use of molecular and biochemical analysis could be a very useful addition - especially in materials that must be grown at remote sites and require trips to collect descriptor data. Molecular and biochemical analysis also might save money by allowing the identification and elimination of duplicate accessions in the collection.

RECOMMENDATIONS

Priority of action categories are outlined as follows:

A. Evaluation

B. Preservation

C. Collecting of Germplasm

D. Enhancement

Specific priority activities within each category have been outlined in the previous section, Germplasm Needs.

SOURCE OF FUNDING

It is unlikely that any substantial funding for NPGS, outside of ARS, will be forthcoming in the foreseeable future, from either state or private organizations that have an interest in promoting cotton. All of the cotton producing states are reported operating under austere budgets. However, it would be prudent to have a long-range plan for germplasm needs in place, anticipating that the economic woes will change for the better.